5 15 672 https://www.nal.usda.gov/exhibits/speccoll/files/original/bb1834344c7745c2c615041f9240d6ac.pdf d469d5bfab7762ff9f1098f3a59b65f4 PDF Text Text Item D Number 05384 Author Erickson, Mitchell D. CorDOratB Author D MotSBannad Midwest Research Institute RflpOrt/ArtlClB TltlO Methods of Analysis for By-Product PCBs - Literature Review and Preliminary Recommendations Journal/Book Title Year 1982 Month/Day October^ D Color Number of Imagos 137 Doscrlnton NotBS Task 51 M • Interim Report No. 1, EPA Contract No. 68-01 -5915, MRI Project No. 4901 -A(51) Friday, March 08, 2002 Page 5384 of 5427 �United States Environmental Protection Agency Office of Toxic Substances Washington DC 20460 EPA-560/5-82-005 October, 1982 Toxic Substances oEPA Methods of Analysis for By-Product RGBs Literature Review and Preliminary Recommendations 100.0 i C12H6Cl4 11 14 100.0 C12D6Cl4 1 1 iL. 100.0 13c12H6Cl4 1 280 285 290 295 300 I 305 Gallons 310 315 320 325 �DISCLAIMER This document has been reviewed and approved for publication by the Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Approval does not signify that the contents necessarily reflect the views and policies of the Environmental Protection Agency, nor does the mention of trade names or commercial products constitute endorsement or recommendation for use. �METHODS OF ANALYSIS FOR BY-PRODUCT PCBs--LITERATURE REVIEW AND PRELIMINARY RECOMMENDATIONS By Mitchell D. Erickson and John S, Stanley Midwest Research Institute 425 Volker Boulevard Kansas City, MO 64110 TASK 51 INTERIM REPORT NO. 1 EPA Contract No. 68-01-5915 MRI Project No, 4901-A(51) October 12, 1982 For U.S. Environmental Protection Agency Office of Toxic Substances Field Studies Branch TS-798 Washington, DC 20460 Attn: Dr. Frederick W. Kutz, Project Officer Mr. Dayid P. Redford, Task Manager �PREFACE This report presents the results of a literature review and preliminary methods recommendations accomplished on MRI Project No. 4901-A, Task 51, "PCB Analytical Methodology Task," for the Environmental Protection Agency (EPA Prime Contract No. 68-01-5915). The review was performed and the document prepared by Drs. Mitchell D. Erickson and John S. Stanley, with assistance from Elliot Hirsch, Scott Meeks, Betty Jones, Kay Turman, Kathy Funk, Lanora Moore, Cindy Melenson, Carol Shaw, and Gloria Sultanik. MRI would like to thank the people listed in Appendix A for their cooperation, We are especially indebted to Phillip W. Albro, Thomas A- Bellar, Michael D. Crouch and associates, Ralph C. Dougherty, Larry F. Hanneman, Edo D. Pellizzari, James D. Petty, and David L. Stalling, J. Lawrence Robinson, and the CMA PCB Analytical Task Group through Robert J. Fensterheim who provided valuable written peer review comments on the draft document. In addition, the helpful comments of John Smith, DDE, OTS, EPA; Dave Redford and Ann Carey, FSB, OTS, EPA; and the members of the PCB Analytical Task Force of CMA are especially appreciated. MIDWEST RESEARCH INSTITUTE John Going, Head Environmental Analysis Section Approved: James L. Spigarelli, Director Analytical Chemistry Department iii �CONTENTS Preface Figures Tables . , List of Terms, Abbreviations, and Symbols 1. 2. 3. 4. 5. Summary . . , Introduction , Literature Review Sources of Information. , Review Procedure Review Articles on PCBs Standard Methods Sampling Extraction Cleanup Determination Data Reduction • Confirmation. Screening Techniques Quality Assurance By-Product Analyses Applicable Techniques Extraction. . . < , Cleanup . . . . . . . Determination Data Reduction Limit of Quantitation Data Reporting Confirmation Techniques Screening/Equivalent Methods Possible Analytical Schemes Issues to be Addressed Quality Assurance , iii vi . vii viii , . . . . , • • • • . .... 1 3 5 5 6 6 7 10 11 15 20 45 55 56 57 59 62 62 64 64 66 70 70 71 71 73 73 74 Appendices A. Personal Contacts B. Bibliography 76 81 v �FIGURES Number 1 Page Comparison of packed column gas-liquid chromatography and and capillary column gas-liquid chromatography with Aroclor standards 24 Comparison of packed column gas-liquid chromatography and capillary column gas-liquid chromatography with a milk extract 25 3 Comparison of PCB resolution on different columns 27 4 Electron capture detection of Aroclors 1242, 1260, and 5460 (400 pg each) chromatographed on an Apiezon L (WCOT) silanized Pyrex glass capillary, 0.20 mm i.d. x 50 m in length 29 Electron capture detection of PCB in an extract of yellow perch 29 Scanning capillary column gas-liquid chromatography/mass spectrometry analysis of a mixed Aroclor standard used to establish retention windows for the CGC/MS-SIM analysis of PCBs 30 Detection limits/dynamic range for several instrumental methods 32 Packed column gas-liquid chromatography/electron capture detector chromatograms illustrating potential interferences between pesticides and PCBs 36 Total ion current profiles for negative chemical ionization data (upper) and electron impact data (lower) obtained on a Finnigan 4000 glass capillary GC/MS system for Ashtabula River fish sample 43 Partial high resolution mass spectrum obtained from 2.5 x 10 10 g TCDD plus matrix from 10 g human milk, illustrating potential interferences in low resolution MS 44 2 5 6 7 8 9 10 vi �TABLES |if umber Page 1 Standard Methods of Analysis for PCBs 8 2 Reported Limits of Detection for PCBs 33 3 Relative Molar Responses of Electron Capture and Flame lonization Detectors to Some Chlorobiphenyls 37 Comparison of Relative Response Factors Between (GC)2-ECD, GC-EIMS (Molecular Ion) and (GC)2-NICIMS (m/z 35) for Homologous Series of PCBs ~ 38 Summary of Laboratory Techniques Used for the CMA Round-Robin Study 46 Summary of Laboratory Techniques Used for the CMA Round-Robin Study , 47 4 5 6 vii �LIST OF TERMS, ABBREVIATIONS, AND SYMBOLS Accuracy Closeness of analytical result to "true" value, Aroclor Trade name (Monsanto) for a series of commercial PCB mixtures marketed in the United States, Askarel Nonflammable synthetic chlorinated hydrocarbon insulating liquids used in capacitors, transformers, etc,; often containing PCBs. Byproduct PCBs PCBs generated as by-products or impurities in synthesis of other products (as opposed to commercial PCBs). CGC Capillary column gas-liquid chromatography (includes WCOT, SCOT, fused silica, glass, and metal). CI Chemical ionization (mass spectrometry), CIMS Positive chemical ionization mass spectrometry, Congener One of 209 PCBs, not necessarily the same homolog. Cutoff Lowest PCB concentration of regulatory concernf DDE l,l-Dichloro-2,2-bis(|>-chlorophenyl)ethylene. DDT 1,1,l-Trichloro-2,2-bis(g-chlorophenyl)ethane. BCD Electron capture detector. El Electron impact ionization (mass spectrometry). EIMS Electron impact mass spectrometry. Equivalent method Any method, certified against the primary method, which can be used for routine analysis of samples. Also, termed screening method. External standard Standards for calculation not added to the sample extract. FFAP Free fatty acid phase. viii �FID Flame ionization detector. FTIR Fourier transform infrared spectrometry. GC Gas-liquid chromatography (column type unspecified). GC/MS Gas-liquid chromatography/mass spectrometry (ionization mode unspecified) . GPC Gel permeation chromatography. HECD Hall electrolytic conductivity detector (other similar detectors such as the Coulson are included) . HEETP Height equivalent to an effective theoretical plate. Homolog One of the 10 degrees of chlorination of PCBs through C12Clio). HPLC High performance liquid chromatography. High resolution electron impact mass spectrometry. Internal standard Standards used expressly for quantitation added to sample extract immediately prior to the analytical determination. IR Infrared spectrometry. Isoraer One of up to 46 PCBs possessing the same degree of chlorination (3,4- and 4,4' -dichlorobiphenyl are different- isomers). KOH Potassium hydroxide. LMS Limited mass scanning (mass spectrometry). LOD Lower limit of detection (also MDL) . Lowest concentration which an analyte can be identified as present in the sample at a stated statistical confidence level. LOQ Lower limit of quantitation. Lowest concentration to which a value can be assigned at a stated statistical confidence level . MDL Method detection limit. Mean Arithmetic mean. MS/MS Mass spectrometry /mass spectrometry. m/z Mass-to-charge ratio. ix �NAA Neutron activation analysis. NCI Negative chemical ionization (mass spectrometry), NCIMS Negative chemical ionization mass spectrometry, NMR Nuclear magentic resonance spectrometry. PBB Polybrominated biphenyl. PCB Polychlorinated biphenyl (including monochlorobiphenyl, but excluding biphenyl). PCN Polychlorinated napthalene. PCT Polychlorinated terphenyl. PGC Packed column gas-liquid chromatography. Parts per billion (109). ppm Parts per million (10~6). Precision Reproducibility of an analysis, measured by SD of replicates. QA Quality assurance. An organization's program for assuring the integrity of data it produces or uses, QC Quality control. The specific activities and procedures designed and implemented to measure and control the quality of data being produced. Remand Rule Rulemaking for new PCB regulations for inclusion in 40 CFR, Part 761 in response to orders from the U.S. Court of Appeals from the District of Columbia on, October 30, 1980 and April 13, 1981. RI Retention index. RIA Radioimmunoassay or radio isotope dilution assay. RJC Reconstructed ion chromatogram. RMR Relative molar response. RSD Percent relative standard deviation (SD/mean x 100), SCOT Support coated open tubular (CGC column). x �SD Standard deviation. Sensitivity The slope of instrument response with respect to the amount of analyte. Also used colloquially in reference to lowest detectable amount of analyte. SIM Selected ion monitoring (also mid or mass fragmentography). Surrpgate Standard compounds added to the sample prior to any analytical manipulations for the express purpose of measuring recovery through extraction, cleanup, etc. TCP Thermal conductivity detector. TCDD Tetrachlorodibenzo-jj-dioxin. TIC Total ion current chromatogram. TLC Thin-layer chromatography. WCQT Wall coated open tubular (CGC column). xi �SECTION 1 SUMMARY The published literature on PCB analysis is critically reviewed. Several hundred references are cited in a bibliography. The review is subdivided into extraction, cleanup, determination, data reduction, confirmation, screening, quality assurance, and by-product analysis sections. The determination section includes TLC, HPLC, GC (PGC and CGC), GC detectors (BCD, FID, HECD, EIMS, and other MS) and nonchromatographic analytical methods (NMR, IR, electrochemistry, NAA, and RIA). Based on the review of the literature, personal communications with researchers in the field, and the authors' judgment, techniques applicable to analysis of commercial products, air, and water for by-product PCBs under the Remand Rule are discussed. Each individual analytical component (extraction, cleanup, determination, etc.) is separately discussed. The final section of this report presents the recommended overall primary analytical scheme: 1. Homogenize sample and subsample if necessary. 2. Incorporate surrogate compounds (e.g., four 13 C PCB congeners). 3. Dilute, extract, or clean up as required. 4. Concentrate or dilute to a known volume. 5. Analyze a known aliquot by GC/EIMS. 6. Identify PCBs by relative retention time and mass spectral characteristics. 7. Integrate the PCBs by homolog and calculate amounts of each homolog by normalizing the responses to responses for the surrogate compounds, using one or more homolog response factors. 8. Sum all 10 homolog concentrations to obtain a total PCB value. 9. Report on standard reporting form. 10. Follow specified routine QC (blanks, controls, duplicates, standard condition, instrument performance criteria, etc.). 11. Maintain appropriate records. �Several details in this scheme are subject to revision, as discussed in the report. Several unresolved issues are discussed, including the permissi^ ble flexibility in the method, the use of equivalent methods, and quality assurance. �SECTION 2 INTRODUCTION PCBs have been manufactured as commercial products since 1929 and marketed under trade names such as Aroclor (United States), Chlophen (Germany), and Kanechlor (Japan). These were all complex mixtures of many congeners and several homologs. They were used as dielectric fluids in capacitors and transformers, hydraulic fluids, fire retardants, plasticizers, and many other applications. Their manufacture and use have been frequently reviewed (Hutzinger et al., 1974; and many other reviews listed in "Review Articles on PCBs" section below). Beginning in 1966 (Jensen, 1972; anonymous, 1966), PCBs were found in environmental samples as interferents with chlorinated pesticides (e.g., DDT) analysis with increasing frequency. As the magnitude of the problem grew, the emphasis on PCBs gradually shifted from interferents to analytes. Concomitantly, their toxicology was being studied. While their toxicity varies among isomers and species, PCBs have been sufficiently implicated in animal and human toxicity to warrant their ban in the United States (Toxic Substances Control Act (TSCA), Public Law 94-469, October 11, 1976). The public outcry prior to TSCA and enforcement of the law thereafter have prompted increased scientific interest in all aspects of PCBs. Central to the environmental studies of PCBs is their analysis. Most of the early methods were direct adaptations of chlorinated pesticide procedures. As interest grew, analytical techniques improved and methods of analysis became increasingly sophisticated. However, PCB analysis is plagued by the fact that PCBs are not one specific compound like most pesticides, but in fact are a class of 209 congeners. Until very recently, all of the PCB analyses were directed toward commercial (e.g., Aroclor) products or their derivatives (metabolites, weathered samples, etc.). The complexity of the raw data (chromatograms) and lack of other standards have led scientists to report PCB findings in terms of Aroclor (or other products) calibration standards. This procedure is at best approximate when the sample resembles the standard and becomes hopeless if the sample does not "match" standards. The problem of PCB analysis has recently become more complex due to concern over incineration products and by-product PCBs, where the PCB mixture in no way resembles an Aroclor pattern. Incinerator products are of concern since U.S enforcement efforts have recently concentrated on destruction of existing PCB products rather than on landfill disposal. The concern over by-product PCBs may be traced to the opinion of U.S. courts that PCBs generated as impurities in other products are subject to the TSCA ban on PCB manufacture. �This document presents a review of analytical techniques used for the analysis of PCB and discusses which of these may be applicable to determination of by-product PCBs. A general analytical scheme is proposed with several options in areas where there is no clear best available technology. �SECTION 3 LITERATURE REVIEW This section is a critical review of the published literature on analytical techniques for PCBs. Where possible, comments have been made regarding the quality of the work and the utility of the technique. However, discussion of the relevance of techniques toward the PCB Remand Rule has been left to Sections 4 and 5. SOURCES OF INFORMATION Computerized and manual searches and relevant references in recent articles were used. Many documents were obtained from the personal files of MRI employees. Recent issues of several key journals (Analytical Chemistry, Journal of Chromatography. Journal of the Association of Analytical Chemists, Environmental Science and Technology, etc.) were searched manually to pick up any recent references not yet in the computer data bases. In addition, several known PCB researchers (Appendix A) were called to discuss approaches to the Remand Rule. In these discussions, they were asked to send copies or give references to any recent publications or preprints. The computer searches were done using DIALOG. Chemical Abstracts (CA) files were searched back to 1972, printing all references containing "PCB" and synonyms and keywords beginning with the following letters: "anal," "chromatogr," "mass spectr" and synonyms. A similar search was performed on the National Technical Information Service data base (now including Smithsonian Science Information Exchange). These searches printed out 349 citations, of which 188 were judged sufficiently relevant to obtain at least the CA abstract. In addition to nonanalytical articles, articles where PCBs were mentioned but were clearly not the major focus of the article, and articles which clearly contained only analytical results and not methods, the majority of citations not obtained were in obscure foreign language journals and apparently were similar to other available references. Once the primary search data had been digested, it became apparent that several authors were of primary interest and all of their recent (1980 and 1981) publications were retrieved by a CA search on their name. These authors were P. W. Albro, T. F. Bidleman, U. A. Th. Brinkman, 0. Hutzinger, R. G. Kaley, S. Safe, D. L. Stalling. Most of the new citations retrieved by this search were irrelevant (metabolism studies, synthesis, etc.) to this report. An additional manual cross-check was made with the document "Polychlorinated Biphenyls, Polybrominated Biphenyls, and Their Contaminants: A Literature Compilation 1965-1977" (Winslow and Gerstner, 1978). This document contains 1,880 PCB citations, although not all pertain to analytical methods. �References contained in review articles and primary literature were also checked to assure that no important articles were missed by the computer search. Several articles were added to the files by these searches. REVIEW PROCEDURE All articles cited in the bibliography (Appendix B) were surveyed for relevant analytical details. The salient features of each article were noted and any key subject areas were listed. Each citation was cross filed in any of 33 applicable key subject areas (PGC, CGC, EIMS, Review, etc.). REVIEW ARTICLES ON PCBs Any class of chemical compounds as often studied and as subject to regulatory pressure as PCBs has been the subject of review articles. A total of 27 books and articles were characterized as PCB reviews. The PCB review most often cited is probably the book The Chemistry of PCBs by Hutzinger et al., (1974). This book characterizes commercial PCBs; synthesis routes; chemical and photochemical reactions; metabolism; mass spectrometry; NMR, UV, and IR spectral properties; determination methods, and recent developments. Although dated, this review covers the general subject and the analysis of PCBs (in 1974) comprehensively. Reynolds (1969) reviewed the problems of PCB interference with the analysis of pesticide residues and later expanded on this in Residue Reviews (Reynolds, 1971). Risebrough (1971) reviewed analytical techniques for PCBs in environmental samples. Fishbein's (1972) review of the chromatographic and biological aspects of polychlorinated biphenyls discussed column chromatography, TLC, GC, and GC/EIMS in some detail, DeVos (1972) reviewed the methods for pesticides and PCBs in. wildlife samples used by various laboratories in the Netherlands. Lincer (1973) reviewed PCBs, again as interferents with pesticide residue analysis. - Safe (1976) reviewed the analytical problems and methods for PCBs as part of a national conference on PCBs (Ayer, 1976) which covered all aspects of the PCB issue (biology, metabolism, destruction, regulatory, etc.). Sherma (1975) reviewed GC of PCBs, including extraction, cleanup, GC systems, identification, confirmation, and quantitation. Chau and Sampson (1975), in an attempt to standardize quantitation, surveyed and reviewed the various methods for PGC/ECD quantitation. Lao et al. (1976) reviewed the application of GC/MS to PCB analysis. Extraction, chromatographic separation (cleanup), GC/ECD, and computer-aided data interpretation were discussed by Krull (1977). Brinkman et al. (1978) discussed various literature procedures for discrimination between PCBs and polychlorinated naphthalenes (PCN). Lawrence and Turton (1978) included PCBs in a tabulation of HPLC data for 166 pesticides. Environmental Health Perspectives devoted an entire issue (Rail, 1978) to PCBs. Pomerantz et al. (1978) reviewed the chemistry of PCBs; Matthews et al. (1979) reviewed metabolism and toxicity; Cordle et al. (1978) reviewed human exposure; Kimbrough reviewed animal toxicology; and the DHEW Subcommittee �of Health Effects of PCBs and PBBs (1978a, 1978b) provided general recommendations and general summary and conclusions which included discussions of the analytical aspects of PCBs. Stalling et al. (1979) reviewed PCS analysis with particular emphasis on their own work of automating the extraction and cleanup processes. The sampling and analysis of PCBs in air have been reviewed by Lao et al. (1976), Margeson (1977), and Fuller et al. (1976). The analytical aspects of PCBs have also been reviewed by Sherma (1981), Nose (1976), and Tanabe (1973). Finlay et al. (1976) reviewed PCB levels in the environment but did not discuss analysis. Other reviews not discussing analysis include Resource Planning Commission (1982), Fishbein (1979), and Kimbrough (1980). None of the review articles discussed above is current (the most recent was 1979), nor do they discuss the problems of determination of incidentally generated PCBs. In response to the PCB Remand Rule, the Chemical Manufacturers Association (1981) critically reviewed the PCB analytical techniques potentially applicable to the rule and recommended that GC/EIMS be the method of choice for analysis under this rule. STANDARD METHODS Table 1 lists the standard analytical methods available for PCBs. It should be noted that not all of the methods listed are sanctioned at this point. Many have interim status and some have been proposed but never endorsed by an organization. The standard methods provide analytical approaches for the measurement of PCBs in a variety of materials including water, wastewaters, soils, sediments, sludges, air, combustion and incinerator emissions, capacitor askarels, transformer'fluids, waste oils, mixtures of chlorinated benzenes, pigments, food, milk, and adipose tissues. Table 1 summarizes each of the standard methods. Extraction and cleanup procedures are presented in terms of the materials and reagents required for analysis. Less than half of the methods comment on the criteria required to make qualitative determinations for the presence of PCBs in sample extracts. The method of quantitation for each of the analysis schemes is presented along with the limit of detection (LOD), if specified. Additional analytical procedures to confirm the levels or presence of PCBs are also indicated. More than half of the standard methods mention quality assurance in some respect. The quality assurance steps include analysis of blanks, replicates, control samples, spiked additions, and accuracy, precision, and instrumental performance criteria. All of the are directed to DCMA (1981) and product PCBs in methods, except those provided by DCMA (1981) and Dow (1981) the analysis of PCBs as Aroclors or similar mixtures. The Dow (1981) methods were developed for the analysis of bycommercial products. �TABLE 1. Method Matrix ANSI Air (toluene impinger) ANSI Extraction Cleanup STANDARD METHODS OF ANALYSIS FOR PCBs Determination method Quant . method Qual. LOD Confirmation QA Reference - (H2S04) PGC/ECD (Saponif ication) (Alumina) No Single peak 2 ppb None Yes ANSI, 1974 Water Hexane (H2S04) PGC/ECD (Saponif ication) Alumina No Single peak or summed peaks 2 ppm None Yes ANSI, 1974 ANSI Sediment soil CH3CN/ hexane H2S04 Saponify/ alumina PGC/ECD No Single peak or summed peaks 2 ppm None Yes ANSI, 1974 AOAC ( 9 2) Food CH3CN/Pet. ether Florisil MgO/ PGC/ECD Celite saponif ication . So Total area or Ind. peaks NSa TLC paper chrom. No AOAC, 1980a AOAC, 1980b Paper and paperboard saponification Capacitor Askarels DIb None SCOT CGC/FID No Total area 2.8 x 10~8 mol/1 None No ASTM, 1980a PGC/ECD No Total area NS None Yes ASTM, 1980b PGC/ECD No 10 isomers ~ 1 ppm/homolog PGC/MS Yes DCMA, 1981 D3303-74 D3304-74 DCMA Air DI Water Soil, sediment Hexane H20/CH3CN 3 pigment types A. Hexane/ H2S04 B. CH2C12 Florisil 00 ( 2 0) HS 4 (Saponif ication) (alumina) None Devenish Water Hexane Alumina PGC/ECD No NS 106 ng/£ None No Devenish and Harling-Bowen 1980 DOW Chlorinated benzenes DI None PGC/EIMS Yes Total peak height/ homolog NS None Yes Dow, 1981 EPA Sludge (Halocarbon) Hexane/ CH2C12/ acetone (83/15/2) GPC S removed PGC/ECD Yes Peak area or peak height NS GC/MS Yes Rodriguez et al. , I960 EPA ( P P) Sludge CH2C12 (base/ neutral and acid fractions) GPC PGC/EIMS Yes NS NS None Yes EPA, 1979e EPA ( 0 h 34) Water Hexane/ CH2C12 (85/15) Florisil/ silica gel (CH3CN) (S removal) PGC/ECD or HECD Yes Summed areas or or WebbMcCall NS None Yes EPA, 1978 EPA (B100) Sludge CH2C12 ( fractions) 3 GPC Silica gel CGC/EIMS or PGC/EIMS Yes NS NS None Yes Ballinger, 19 (continued) �TABUS 1 (continued) Qnant. method Ttetenniflation method Qua I . CH3CH Florisil Silica acid PGC/ECD Yes Ind. peaks 50 ppb Perchlorination Yes Watts, 1980 Sherroa, 1981 Pet. ether/ CH3CN Saponif ication Florisil TLC No Semiquant. 10 ppm None No Watts, 1980 Dl ( 2 0) HS 4 (Florisil) (Alumina) (Silica gel) (GPC), (CH3CH) PGC/HECD or /ECD or /E1MS (CGC) No Total area or WebbMcCall 1 mg/kg None Yes EPA, 1981 Bellar and Ltchtenberg, 1981 Natural gas Hexane sampled with Florisil H2S04 PGC/ECD No Harris and Mitchell, 1981 608 Water CH2C12 (Florisil) (S removal) PGC/ECD No Area 0.04-0.15 JJg/£ None Yes EPA, 1979a 625 Water CH2C12 None PGC/EIMS (CGC) • Yes Area NS None Yes EPA, 1979b 10 ng GC/MS Method Matrix Extraction EPA (HERL) Milk Acetone/ hexane EPA (HERL) Adipose EPA ( i ) ol Transformer fluids or waste oils EPA (gas) Hexane Cleanup Total area 0.1-2 (Jg/m3 peak height or Webb-McCall (Perchlorination) 2S04 emissions and ambient air collected on Florisil LOD Confirmation N °ne QA Reference '-lHa i ic -_,! ana Baladi, 1977 nation PGC/ECD EPA Combustion sources collected on Florisil Pentane or CH2C12 (Florisil/ silica gel) PGC/HS Yes Area/ homolog 0.1 ng/inj None No Levins et a 1 . , 1979 NIOSH (air) (P&CAM 2 4 4) Air collected on Florisil Hexane None PGC/ECD No Peak height 0.01 mg/m3 or area from standard curve or Webb-McCall None No NIOSH, 1977a HIOSH (air) (P&CAM 253) Air collected on Florisil Hexane Hone PGC/ECD Perchlorination No Peak height 0.01 mg/m3 or area from standard curve Perchlorination Japan Food Several Silica gel Saponif ication (Florisil) PGC/ECD Yes Summed NS areas perchlorination None No Tanabe, 1976 PAM Food CH3CN/Pet. ether Silicic acid (Saponif ication) (Oxidation) (Florisil) PGC/ECD (PGC/HECD) (HP-TLC) (RP TLC) Ho Area TLC No FDA, 1977 a No specific details. b Direct injection. c Techniques in parentheses are described as optional in the protocol. NS Ho NIOSH, 1977b �The ANSI methods are based on techniques that were used by the Monsanto Industrial Chemical Company for the isolation and determination of PCps in water, soil, sediment, and biological materials. Packed column gas chromatog17 raphy with electron capture defection (POC/JSCD) is the designated method for quantitation pf PCBs as Aroclops in, the ANSI methods, Mass spectrometry, however, is recommended for eaph of the designated analysis schemes if confirma-» tion is required. The cleanup techniques are required only if interferences are noted for the PGC/ECD determination. The quality assurance procedures in the ANSI methods emphasize t;he number of theoretical plates and tailing factor for the packed gas chromatography column, The ASTM, AOAC, and EPA methods are generally designed for a particular matrix. The level of quality assurance procedures varies from method to method. The recent methods provide quality assurance programs of greater de"tail and require reasonable effort to maintain the accuracy and precision of the overall determination. The EPA method for analysis of PC$s in transformer °ils, and crude oils provide^ the most generalized approach, with respect to sample preparation, cleanup procedures, and^ instrumental analysis. Several cleanup procedures are provided as optional approaches in this protocol, and instrumental analy* sis by halogen specific, electron capture, pr mass spectrometry detectors are allowed, provided appropriate limits of detection can be achieved. A strong quality assurance program including control samples, daily quality control check samples, blanks, standard additions, accuracy and precision records, and instrumental and chromatographic performance criteria is required to support all data generated by the method. The Dow (1981) and DCMA (1981) procedures also require strong quality assurance programs for analysis of by-product PCBs. first $tep in any successful analysis is the collection of ta,tive samples, The selection of sampling sites, frequency of sampling, number of samples, measurement of physical a,n.d chemical parameters of the sample, and the ovejraU statistical design of sampling methods have been provided in extensive detail (Moser and Huibregtse, 1976; EPA, J976). The sampling design in roost cases is directly related to the objectives pf a specific research program as a regulatory action. The methods that are of interest in the context of this report are the procedures practiced for obtaining representative specimen of air, water, and solids f Aqueous and solid media are generally collected as grab samples, although aqueous samples have been preconqentrated on solid adsorbents prior to extraction. Water The solid adsorbents used for aqueous preconcentration methods include macroreticular resins (Cobum et al,, 1977; Seiber, 1974; Musty and Nickless, 1974; Webb, 1975; Picer and Picer, 1980), polyurethane foam and Chromosorb W 10 �coated with a mixture of undecane and Carbowax 4000 monostearate (Gesser et al,, 1971; Lawrence and Tosine, 1976; Bellar and Lichtenberg, 1975; Ahling and Jensen, 1970; Osterrpht, 1974; Musty and Nickless, 1976), as well as Tenax (Leoni et al,, 1976). Preconcentrating organic analytes from aqueous media allows the analyst to work with large volumes of water and thus lowers the method detection limit for the compounds of interest. The sorbent preconcentration method also can be used as a continuous on-site sampling procedure. Air Sampling The PCBs in ambient air and flue gas emissions have been collected on sqlid adsorbents including polyurethane foam plugs, Tenax, XAD-resins, Florisil, Chromosorb, and Poropak and combinations of these materials (Bidleman and Olney, J974; Bidleman et al., 1980; Bidleman, 1981; Bidleman et al., 1981; Billings and Bidleman, 1980, 1982; Burdick and Bidleman, 1981; Simon and Bidleman, 1978, 1979; Lewis et al., 1977a, 1977b; Lewis and MacLeod, 1982; Lewis and Jackson, 1982; Williams et al., 1980; Haile and Baladi, 1977; Giam et al., 1975; Haile et al., 1982; Stanley et al., }982; Harris and Mitchell, 1981). The solid adsorbents are effective for sampling PCBs in air although some problems have been encountered with breakthrough of the lower molecular weight PCBs at high flow rates for extended sampling periods. Doskey and Andren (1979) evaluated polyurethane foam coated with DC 200, Florisil, and Amberlite XAD-2 resin for their ability to sample airborne PCBs. The collection efficiency of the adsorbents was studied using carbon-14 labeled 2,5,2',5'-tetrachlorobiphenyl. The XAD-2 resin was found to have an excellent collection efficiency for the tetrachlorobiphenyl at a flow rate of 1 liter/min. Their sampling system yielded 96.5% collection efficiencies for the tetrachlorobiphenyl and 83.0% for Aroclor 1221. Further investigations demonstrated low retention efficiencies for monochlorobiphenyl (72%) and dichlorobiphenyl (86%), thus demonstrating that the sampling system was not equally effective for all PCB congeners. The analytical recoveries for tetrachlorobiphenyl, Aroclor 1242, and Aroclor 1221 were 85.5%, 80.1%, and 64.9%, respectively. Hanneman (personal communication, 1982) reported that PCBs were not retained at acceptable levels on common solid adsorbents when the flue gas temperature was greater than 150°C or in cases where the air contained an aerosol of a nonpolar material in which PCBs are very soluble, Hanneman reported successful collection of PCBs in these instances using polyurethane foam plugs coated with liquid polydimethylsilocane. Several plugs of the coated polyurethane were placed in a water-cooled jacket to sample the air at elevated temperatures, A PCB isomer was added to the surface of the foam plugs as a surrogate prior to sampling. A second PCB isomer was added to the foam plugs after sampling to monitor surrogate recovery and collection efficiency. EXTRACTION Reliable PCB analysis begins with the quantitative extraction of the analytes from the sample matrix. The extraction method is dependent on the sample type and the complexity of the matrix encountered. In general, the extraction methods require the use of solvents such as petroleum ether, hexane, methylene chloride, acetone, and acetonitrile. Digestion of the sample matrix with sulfuric acid or saponification with alcoholic potassium hydroxide is necessary 11 �in some instances to effectively extract incorporated PCBs. Dilution with suitable organic solvents prior to gas chromatography and even direct injection of some PCB-contaminated samples have provided suitable quantitative analysis. PCBs in air and flue gas emissions are typically collected on solid adsorbents and removed by extracting with a suitable solvent. Standard Methods A number of standard extraction methods for specific sample types are listed in Table 1. The standard methods include extraction techniques for transformer and capacitor oils, food, soils, sediments, dyes, milk, adipose tissue, sludge, wastewater, natural waters, emissions from combustion sources, and ambient air. Review Articles Extraction methods have been reviewed previously (Hutzinger et al., 1974; Sherma, 1975; Krull, 1977). Factors and problems that should be considered for a given extraction procedure (Albro, 1979) include the following: (a) each extraction method must be validated for each different matrix; (b) the nature of the sample matrix influences the effectiveness of a given extraction procedure through various matrix properties. These properties include the solubility of the matrix in solvent, the ease of homogenizing the sample for subsampling, the water content of the sample which greatly affects the extraction efficiency of the solvent, and the lipid content of tissue samples that governs the volume of solvent required for quantitative extraction; and (c) the incorporation of the analyte in a sample matrix and the most effective means of adding spikes to the sample for method evaluation and quality assurance measurements. Primary Literature A large number of extraction techniques provide quantitative recovery of PCBs from widely different matrices. The application of the various extraction procedures to specific sample types is discussed below. Air-- Simple and straightforward extraction procedures are used for extraction of adsorbents from air sampling. Ambient air and flue gas emissions have been collected on adsorbents including polyurethane filter plugs (Bidleman and Olney, 1974; Bidleman et al., 1980; Bidleman, 1981; Bidleman et al., 1981; Billings and Bidleman, 1980, 1982; Simon and Bidleman, 1977a, 1977b, 1977c; Lewis and MacLeod, 1982; Lewis and Jackson, 1982), Florisil (Harris and Mitchell, 1981; Williams et al., 1980; Haile and Baladi, 1977; Giam et al,, 1975), and Amberlite XAD-2 resin (Haile and Lopez-Avila, 1981; Stanley et al., 1982). PCBs were quantitatively recovered from these adsorbent materials via extraction with hexane, petroleum ether or benzene in a Soxhlet apparatus or as small chromatographic columns. Water and Wastewater-PCBs in aqueous samples, including natural waters, potable supplies, sewage effluents and industrial wastewaters, have been extracted by a number of 12 �procedures. Liquid-liquid extraction with hexane, cyclohexane or methylene chloride provide quantitative isolation of PCBs from aqueous samples (Brownrigg et al., 1974; Devinish and Harling-Bowen, 1980; EPA, 1979a, 1979b; Haque et al., 1974; Adams et al., 1979; Bellar and Lichtenberg, 1975; Caragay and Levins, 1979). In principle it is possible to extract all PCBs present in any water sample without large scale solvent extraction (Krull, 1977). Lower levels of detection of PCBs in aqueous samples have been achieved through the application of solid adsorbents for large volumes that cannot be effectively handled by classical liquid-liquid extraction procedures. The adsorbents are extracted with hexane, petroleum ether, and diethylether to recover the PCBs. Macroreticular resins (XAD-2, -4, -7, and -8) have been evaluated in a number of Studies (Coburn et al., 1977; Seiber, 1974; Musty and Nickless, 1974; and Webb, 1975; Picer and Picer, 1980). Polyurethane foams and Chromosorb W coated with a mixture of undecane and Carbowax 4000 monosterate as a reversed liquid-liquid partition method have also demonstrated successful isolation of PCBs from aqueous matrices (Gesser et al., 1971; Lawrence and Tosine, 1976; Bellar and Lichtenberg, 1975; Ahling and Jensen, 1970; Osterroht, 1974; Musty and Nickless, 1976). Two studies (Bellar and Lichtenberg, 1975; Webb, 1975) compared various extraction methods including batch liquid-liquid extraction, vortex stirring with an organic solvent, and adsorbent concentration on polyurethane and amberlite macroreticular resins. Extraction efficiencies of PCBs were greatest with liquid-liquid extraction. Continuous liquid-liquid extractors (Leoni, 1971; Ahnoff and Josefsson, 1973, 1974; Ahnoff et al., 1979) and steam distillation of aqueous samples with simultaneous liquid-liquid extractions of the distillate into pentane (Godefroot, 1982) have been studied as alternate means to lower detection limits for specific organic compounds including PCBs present in water. Wastewaters from some industrial processes have posed some interesting problems in measurement of total PCBs released. In particular, the cellulose fibers collected from paper mill effluents required hydrolysis with alcoholic potassium hydroxide and extraction with hexane or methylene chloride as well as extraction of the aqueous phase to effectively quantitate total PCBs (Delfino and Easty, 1979; Easty and Wabers, 1978). Dissolution of the residual fibers in paper mill effluents with 72% H2S04 followed by dilution with water and extraction with hexane was demonstrated to promote quantitative isolation of PCBs. Colenutt and Thorburn (1980) have discussed the application of gas stripping or purge and trap techniques for the analysis of PCBs in water. Spiked PCBs were removed from water samples by purging with nitrogen at ambient temperature. The PCBs were concentrated on activated charcoal and desorbed with a minimum volume (50 (Jl-1.0 ml) of organic solvent. Recoveries of greater than 90% were reported for Aroclors 1221, 1248, and 1254. The efficiency of this extraction technique is dependent on the gas flow rate, the time of stripping, adsorbent particle size, and the desorbing solvent. 13 �Soils, Sediments, and Sludges-Soxhlet extraction of soils and sediments with hexane-acetone, petroleum ether-acetone, or hexane has been a common method of isolating the PCBs (Bellar and Lichtenberg, 1975; Eder, 1976a; Goerlitz and Law, 1974; Jensen et al., 1977; Macleod, 1979; Seidl and Ballschsmiter, 1976; Adams et al., 1979; Hutzinger et al., 1974). Other methods that have been used for soils, sediments, and sewage sludges have included silica sonication (Chau and Babjak, 1979), blending with suitable solvent such as methylene chloride with centrifugation for separation of the phases (Rodriguez et al., 1980), and a column technique that required mixing sediment with Florisil and eluting with 10% water in acetonitrile. Ethanolic potassium hydroxide reflux prior to extraction of the sediment has been reported in at least one instance (Wakimoto et al., 1971). Soxhlet extraction, solvent shakeout, solvent blending, two column elution methods, and high frequency dispersion (Tissumizer) extraction techniques were compared for the same bottom sediment (Bellar and Lichtenberg, 1975). The results indicated that the highest recovery of PCBs was achieved by Soxhlet extraction of dried samples. The authors concluded that this should be the technique of choice for bottom sediments and sludges. Bellar et al. (1980) extended this study using Soxhlet extraction, Bonification, and steam distillation for the recovery of PCBs from environmentally contaminated lake and river bottom materials. The high frequency dispersion extraction technique was not used in this study because of excessive and rapid wearing of parts of the device and excessive breakage of glassware. Bottom sediments were spiked with Aroclor 1254 and extracted by the three techniques. The mean recovery of PCBs for spiked samples was 81-109% for the different methods, indicating that any of the three might be used for quantitative extraction. However, the Soxhlet extraction method yielded higher levels of PCBs from environmentally contaminated sediments than either steam distillation or Bonification. The results of this study are not conclusive since only one solid sample matrix was considered. Seidl and Ballschmiter (1976a) studied the recovery rates of PCBs from soil using Soxhlet extraction with hexane, acetone/acetonitrile, or ultrasonic extraction with acetone. Carbon-14 labeled Clophen A-30 was added to the soil to simplify the extraction studies. The authors concluded that Soxhlet extraction with acetone/acetonitrile yielded the best recoveries (greater than 95%) and Soxhlet extraction with hexane or ultrasonic extraction with acetone was not suitable for good recoveries of PCBs from soil. Biological Matrices-Considerable emphasis has been placed on the analysis of biological materials for the presence of PCBs. Tissues are generally homogenized and ground with sodium sulfate, sand, or Florisil and are either Soxhlet extracted (Bagley et al., 1970; Curley et al., 1971; deVos, 1972; Hattula, 1974; Holden, 1971; Kuehl et al., 1980; Stalling et al., 1972) or packed into a chromatographic column and the PCBs are eluted with an appropriate solvent (Bowes and Lewis, 1974; Call et al., 1974; Donkin et al., 1977; Erney, 1974b; Ernst, 1974; Hattula, 1974; Stalling, 1971; Wardall, 1977; Hutzinger et al., 1974; Stalling et al., 1972; Sawyer, 1973; Armour and Burke, 1970). Recently, microcontinuous 14 �liquid-liquid extraction combined with steam distillation has been shown to be an effective extraction procedure for tissues (Kuehl et al., 1980). Other tissue extraction techniques have included blending with chloroform and methanol mixtures followed by centrifugation (Sherma, 1975) and saponification of fat and animal tissue with methanolic potassium hydroxide followed by liquid-liquid extraction with hexane (Adams et al., 1979). PCBs in serum and blood samples have been extracted by dilution with methanol followed by liquidliquid extraction with either hexane or diethylether. Miscellaneous-The other materials that have been analyzed for PCB content include transformer oil, silicone fluids, chlorinated benzenes, paper, packaging materials, and pigments. Generally, transformer oils, silicone fluids, and chlorinated benzenes are simply diluted with solvent prior to cleanup or direct analysis (Adams et al., 1979; ASTM, 1980a; EPA, 1981; Bellar and Lichtenberg, 1981; Klimisch and Ingebrigtson, 1980; Dow, 1980). The paper and packaging materials have required homogenization by grinding before Soxhlet extraction with hexane or acetone (Kurastune and Masuda, 1972; Giacin and Gilbert, 1973; Serum et al., 1973) or hydrolysis with refluxing alcoholic potassium hydroxide and final extraction with petroleum ether (Burke et al., 1976; Easty, 1973; Shahied et al., 1973). In a method published by the DCMA (1980), extraction of phthalocyanine blue pigment required dissolution with sulfuric acid prior to hexane extraction to isolate incorporated PCBs. The PCBs in other pigments, such as phthalocyanine green and diarylide yellow are quantitatively extracted by high speed homogenization with methylene chloride (DCMA, 1980). Only the extraction methods described for the colored pigments (DCMA, 1980) were developed for PCBs as nonAroclor chlorinated compounds. All other methods were used in studies that considered the environmental impacts of the commercially distributed PCBs as Aroclors. CLEANUP In addition to PCBs, a large number of chlorinated compounds, lipid materials, and sulfur are extracted from aqueous, oil, tissue, sludge, and sediment samples by the methods described above. Hence, it is necessary in many cases to provide an additional sample preparation step or cleanup to remove the coextractants that may act as interferences. Cleanup techniques vary considerably according to the particular sample matrix and needs for final instrumenta1 analys i s. Review Articles Sample extract cleanup procedures for PCB analyses have been partially reviewed previously (Holden, 1971; Fishbein, 1972; deVos, 1972; Hutzinger et al., 1974; Krull, 1977; Albro, 1979). 15 �Standard Methods Table 1 listed the cleanup procedures that are typically used with a number of standard methods for these materials. These cleanup procedures include liquid-liquid partition of the sample extract, saponification with alcoholic potassium hydroxide, addition of concentrated sulfuric acid, gel permeation chromatography, and oxidation of interferences in the sample matrix. Primary Literature Adsorption Chromatography Cleanup-Perusal of the literature indicates that adsorption column chromatography is the most often practiced method of sample extract cleanup prior to instrumental analyses. The extent of sample cleanup required is dependent, in many cases, on the specificity of the detector used for identification and quantitation. Electron capture and halogen specific detectors (i.e., Hall electrolytic conductivity detector) require clean extracts since so many other compounds can interfere. Chromatographic column cleanup procedures have been used extensively with adsorbents such as Florisil, silica gel, and alumina. Cleanup procedures may require the use of only one column, combinations of adsorbent materials, use of an adsorbent following liquid-liquid partition, or cleanup after matrix destruction by sulfuric acid or saponification. Proper activation of adsorbed materials and the characterization of the degree of activation is essential for effective and reproducible cleanup of sample extracts by a particular chromatography procedure (Edwards, 1974; Zitko, 1972). Reproducibility of a column method requires avoidance of overloading the column, accidental deactivation of the adsorbent during cleanup, and use of pure solvents (Edwards, 1974). Optimum cleanup and separation of PCBs from interferences require fully activated adsorbents, and large eluent volumes (Edwards, 1974; Albro and Parker, 1980). The alternative to large, eluent volumes is to use only a fraction of the extract with microcolumn adsorbent procedures. A notable example is the Sep Pak marketed by Waters Associates. Several investigators have reported the application of Sep Pak for PCB cleanup as discussed below. The most commonly used adsorbents are Florisil and silica gel at various levels of activation. Florisil has been used to remove gross interferences from sample extracts from air, water, wastewater, tissues, and dairy products as well as paper, paperboard, and paper mill effluent (AOAC Methods, 1980; Adams et al., 1979; Delfino and Easty, 1979; Easty, 1973; EPA, 1979a; EPA, 1979b; EPA, 1978; EPA, 1980; Kamops et al., 1979; Kuehl et al., 1980; Modi et al., 1976; Price and Welch, 1972; Reynolds, 1971; Reynolds, 1969; Robbins and Willhite, 1979; Rodriguez et al., 1980; Stijve et al., 1974; Swift and Settle, 1976; Tessari and Savage, 1980; Yakushiji et al., 1978; Bagley et al., 1970; Bagley and Cromartie, 1973; Bellar and Lichtenberg, 1976; Chau and Babjak, 1979). Florisil has also been used to provide additional separation of sample extracts following initial cleanup of matrices by low temperature precipitation, acetonitrile partitioning, oxidation, sulfuric acid digestion, alumina chromatography, or gel permeation chromatography (Eden, 1976; Ernst et al., 1974; Kohli et al., 1979; Mes et al., 1977a, Mes et al., 1977b; Mulhern et al., 1972; Stanovick et al., 1973; Swift and Settle, 1976; Tessari and Savage, 16 �1980; Trotter, 1974; Uk et al., 1972; Bagley et al., 1970; Bagley and Cromartie, 1973; Copeland and Gohmann, 1982), Seidl and Ballschmiter (1976b) investigated the recovery and efficiency of cleanup methods for the isolation of PCBs from vegetable oils. Column chromatography on Florisil, matrix destruction via saponification and sulfuric acid treatment, and liquid-liquid partition with hexane/acetonitrile or hexane/ dimethylformamide were compared as cleanup techniques. The Florisil chromatographic column with hexane/methylene chloride (80:20) as the eluent and liquidliquid partition with hexane/dimethylformamide were shown to be the methods of choice with recoveries of greater than 90%. Silica gel or silicic acid has also been used for a large number of sample matrices including water, sediments, sludges, foodstuffs, tissues, and transformer fluids (Armour and Burke, 1970; Devinish, Harling-Bowen, 1980; Erickson and Pellizzari, 1977, 1979; Ernst, 1974; Giacin and Gilbert, 1973; MacLeod, 1979; Masumoto, 1972; Mes et al., 1976; Mes and Campbell, 1977; Nose, 1973; Ogata et al., 1980; Picer and Abel, 1978; Price and Welch, 1972; Sawyer, 1973; Stalling, 1971; Steichen et al., 1981, 1982; Balya and Farrah, 1980; Beezhold, 1973; Bellar and Lichtenberg, 1975; Coburn et al., 1977). Many applications have utilized the excellent separation properties of silica gel to remove other halogenated interferences from PCB fractions. The separation of a number of halogenated pesticides has been accomplished by using silica gel after preliminary cleanup of gross interferences and controlling the degree of activation and size of the column or by slightly modifying the adsorbent with silver nitrate or an oxidizing agent (Erney, 1974a; Erney, 1974b; EPA, 1978; Herzel, 1971; Huckins et al,, 1976; Kreiss et al., 1981; Kveseth and Brevick, 1979; Leoni, 1971; Mitzutani and Matsumoto, 1973; Musial et al., 1974; Needham et al., 1980; Public Health Services, CDC, Atlanta, 1980; Serum, 1973; Snyder and Reinert, 1971; Stratton, 1977, Swift and Settle, 1976; Trevesani, 1980; Underwood, 1979; United Kingdom, Department of Environment, 1979; Wakimoto et al., 1975; Bidleman et al., 1978). Cleanup and separation of interferences from PCBs has been accomplished with Sep Pak miniature silica gel cartridges for transformer (Gordon et al., 1982; Steichen et al., 1981), mineral, phosphate ester, glycol base, and sulfonated mineral oils (Balya and Farrah, 1980). Alumina has been used as an adsorbent in more recent applications for cleanup of matrices from oils, fats, and tissues (Donkin et al., 1977; Hattula, 1974; Kohli et al., 1979; Kveseth and Brevick, 1979; Ofsta'd et al., 1978; Teichman et al., 1978; Wardall, 1977; Zitko, 1976) and other biological materials such as blood and human milk (Musial et al., 1972; Siyali, 1973; Tuinstra and Traag, 1979a, 1979b; Welborn et al., 1974) and from oil, water, sediments, and vegetable materials (Goerlitz and Law, 1974; Lewis et al., 1977; Tuinstra et al., 1981; United Kingdom, Department of Environment, 1979). Dispersion of activated carbon on polyurethane foam, termed carbon foam chrpmatography, and use of activated charcoal have also proven to be effective means of isolating PCBs from complex matrices such as tissues and sediments (Chau and Babjak, 1979; Jensen and Sundstrom, 1974; Stalling et al., 1979a; Stalling et al., 1979b; Stalling et al., 1978; Stalling et al., 1975; Tiechman et al., 1978). 17 �Gel Permeation Chromatography-Gel permeation chromatography (GPC) has arisen as a popular cleanup procedure for complex matrices, especially samples containing macromolecular interferents such as biological materials containing high levels of lipid materials. The GPC method has been fully automated for large numbers of sample extracts. However, it is necessary to fully validate the GPC method for each different sample matrix as with other cleanup procedures. Gel permeation chromatography has been successfully used as a cleanup for matrices of high molecular weight content and has provided a cost effective approach towards automation of large numbers of samples (Albro, 1979; Caragay and Levins, 1979; Griffitt and Craun, 1974; Haile and Lopez-Avila, 1981; Hopper and Hughes, 1976; Kohli et al., 1979; Kuehl et al., 1980a, 1980b; Rodriguez et al., 1980; Stalling, 1971, 1976; Stalling et al., 1972, 1979; Tessari, 1980). Gel permeation chromatography of tissue and vegetable material extracts followed by carbon-foam chromatography provides a cleanup that is exceptionally selective for planar aromatic hydrocarbons and the chlorinated analogs such as PCBs (Dougherty et al., 1980). Lipidex was shown to separate PCBs and other semivolatile halocarbons from water, fat, butter, and milk (Egestad et al., 1982). Elution conditions could be adjusted to separate the halocarbons from steroids and fatty acids. The authors noted that a combination of partition, molecule sieving, and aromatic adsorption was involved in the separation. High Performance Liquid Chromatography and Thin Layer Chromatography— High performance liquid chromatography (Aitzetmiiller, 1975; Dolphin and Willmott, 1978; Rohleder, 1976) and thin-layer chromatography (Fishbein, 1971; Hattula, 1974; Koeniger, 1975) have also received limited attention as chromatographic cleanup methods. Recently, an HPLC cleanup method for determination of PCBs in oils and waste oils has been devised (Chesler et al., 1982). Acid Cleanup— Sulfuric acid is added as the first step of many cleanup procedures to remove gross interferences, although a number of studies found sulfuric acid cleanup alone sufficient for PCB analysis (Ahling and Jensen, 1970; Becker and Schulte, 1976; Haile and Baladi, 1977; Hattula, 1974; Mattson and Nygren, 1976; Murphy, 1972; Veierov and Aharonson, 1980; Harris and Mitchell, 1981). Losses of the mono- to trichlorobiphenyls were reported in two studies (DCMA, 1981; Lincer, 1973) which used a heated sulfuric acid cleanup. No losses were observed at room temperature (Haile and Baladi, 1977). The chemical destruction cleanup methods are extremely useful but care must be taken to ensure valid recoveries of the particular PCB isomers of interest. For example, mono-, di-, and trichlorobiphenyl isomers were not quantitatively removed from pigment matrices that were treated with concentrated sulfuric acid (DCMA, 1981; Lincer, 1973). Low recoveries were presumed to be due to sulfonation of the biphenyl ring (Lincer, 1973). Liquid-liquid Partitioning— Liquid-liquid partition is also used to remove large amounts of interferences from water, wastewater, milk, food products, packaging materials, silicone fluids, oil, and tissue extracts before final cleanup by a chromatography technique (EPA, 1978; Gordon et al., 1982; Leoni et al., 1973; Mulhern 18 �et al., 1972; Siyali, 1973; Swift and Settle, 1976; Tessari, 1977; Tessari and Savage, 1980; Klimisch and Ingebrigtson, 1980; Welborn et al., 1974). Large amounts of an interference in a sample extract, such as the lipid content of a tissue extract, has a pronounced effect on the efficiency of liquidliquid partition cleanups (Albro, 1979). Validation of a cleanup procedure using spiked blanks provides much better recovery values than can be achieved for an actual sample. Therefore, it is necessary to spike actual sample extracts to fully characterize the limitations of this type of cleanup step. Seidl and Ballschmiter (1976b) have demonstrated PCB recoveries of greater than 90% for cleanup of vegetable oil extracts by liquid-liquid partition with hexane and dimethylformamide. Saponification— Saponification of the sample matrix has been discussed as a method for extracting PCBs from certain materials. Saponification may also be considered a cleanup procedure (Lincer, 1973; Tatsukawa and Wakimoto, 1972; Trotter, 1974). Saponification of sample matrices with ethanolic potassium hydroxide has been accomplished without chemical change to the PCBs present (Young and Burke, 1972). Miscellaneous— Other cleanup procedures that have been shown to be effective but have limited use are low temperature precipitation of lipids from tissues and human milk samples prior to solvent extraction or liquid-liquid partitioning (Mes et al., 1977a, 1977b; Mes and Campbell, 1976). Oxidation of interfering chloronaphthalenes and chlorinated pesticides such as DDT and DDE with chromium trioxide or chromic acid has been shown to be effective when used in conjunction with a final chromatographic cleanup (Holmes and Waller, 1972; Mulhern et al., 1972; Underwood et al., 1979). However, Szelewski et al. (1979) have questioned the reliability of the chromium trioxide oxidation of interferences in sample extracts. Fish extracts, spiked with Aroclor 1016, 1221, and 1254, were treated with the chromium trioxide oxidation technique. Recoveries of Aroclor 1016 from this oxidation step ranged from 30 to 90% for eight replicates, while Aroclor 1254 recoveries ranged from 40 to 80% for eight replicates. Aroclor 1221 was reported to have 0% recovery from this oxidation step for six replicate samples. Szelewski et al. (1979) theorized that PCBs in the extracts were lost by oxidation, by volatilization due to the highly exothermic nature of the oxidative process, or a combination of the two. Steam distillation of water, sediments, and tissues provides relatively clean extracts that require little or no additional cleanup (Veith and Kiwus, 1977; Dougherty et al., 1980). Interference from elemental sulfur is a serious problem, especially for electron capture detector methods of analysis. The sulfur interference in water, wastewater, sewage sludges and sediments can be effectively removed by precipitation of sulfur with mercury (Bellar and Lichtenberg, 1976; Goerlitz and Law, 1974; Rodriguez et al., 1980) or by converting sulfur to thiosulfate by addition of tetrabutylammonium sulfate (Jensen et al., 1977). Recovery Measurement Regardless of the cleanup procedure required for PCB analysis from a particular sample, it is of utmost importance to document recovery of PCBs from the method. Routine quality assurance governs that the recovery be determined 19 �for each different sample matrix encountered. Also, recovery of the PCBs should be determined each time a given parameter of an established cleanup procedure is changed. For example, different batches of an adsorbent may differ greatly with respect to activation. Cleanup steps should be monitored with spiked samples to support overall quality of the data. In some instances, a visible marker such as azulene can be used as a real time monitor to determine the performance of column chromatography methods (Nowicki, 1981). DETERMINATION Thin-Layer Chromatography In addition to its use as a cleanup technique, thin-layer chromatography (TLC) has been used extensively as a determination technique. TLC was used early (latter 1960s, early 1970s) because HPLC was not readily available and the GC techniques were not well-developed. Most of the early TLC reports were normal phase (silica gel) and included elaborate cleanup steps to remove interferents (e.g., oxidation of DDE to a benzophenone derivative). In the mid-1970s when packed column gas-liquid chromatography/electron capture detection (PGC/ECD) became the method of choice, emphasis on TLC methods dwindled. Several articles have been published which take advantage of modern TLC techniques: high performance TLC, two-dimensional TLC, reverse phase TLC, and new detection methods. TLC has been shown to be an effective technique for determination of (Aroclor) PCBs in a wide variety of matrices. The advantages included its ease of use and the simplicity of the apparatus. The disadvantages include lack of resolution, moderate sensitivity, and specificity. Review Articles-TLC analysis of PCBs was reviewed by Fishbein (1972). Standard Methods— TLC is included as an alternate technique for "semiquantitation" analysis of PCBs in human adipose tissue in EPA manuals (Watts, 1980; Sherma, 1981). It is also included in the Association of Official Analytical Chemists methods for confirmation of identity (AOAC, 1980). TLC is mentioned by the Food and Drug Administration (1977) as a technique which they feel may also be useful in dealing with particular resin combinations . Primary Literature— Since the publication of a TLC method for PCBs by Mulhern (1968) and Mulhern et al. (1971), several researchers have used a similar method for analysis of PCBs in food (Stijve and Cardinale, 1974), animal feeds (Westoo and Noren, 1970), food packaging (Zimmerli et al.), bald eagles (Bagley et al., 1973), Aroclor mixtures (Willis and Addison, 1972), animal tissue (Collins et al., 1972; Koeniger et al., 1975; Bush and Lo, 1973; Hattula, 1974; Mes et al., 1977), human adipose tissue (Price and Welch, 1972; Lucas et al., 1980), and human milk (Savage et al., 1973a, 1973b; Mes and Davies, 1979). 20 �Many of these researchers employed TLC in conjunction with other techniques such as GC/ECD. Often, TLC was used as a confirmation technique. Several publications have reported developments claimed to improve the technique. Circular TLC reportedly improves sensitivity by an order of magnitude with a PCB limit of detection of about 0.05 M8 (Koch, 1979). Fused glass TLC has been reported as yielding longer plate life (Okamura et al., 1973). Reverse phase TLC has been reported to yield better separation of PCBs from interferences (deVos and Peet, 1971; deVos, 1972; Stalling and Huckins, 1973; Brinkman et al., 1976a). An impregnated silica gel plate has been reported (Bergman et al., 1976) which improves selectivity apparently on the basis of ion-pairing. The use of surfactant micellar solutions as the mobile phase is certainly novel and reportedly has potential for separation of chlorinated aromatics, including decachlorobiphenyl (Armstrong and Terrill, 1979). Improvements in detection have included an AgNOs spray followed by UV irradiation (deVos and Peet, 1971; deVos, 1972; Kawabata, 1974) and fluorescence (Kan et al., 1973; Ueta et al., 1979). A two-dimensional TLC method was developed which barely separated the DDT analogs from PCBs (Fehringer and Westfall, 1971). This last reference points to one of the major problems with TLC determination of PCBs. Many common interferences (e.g., DDE in biological tissues) have similar elution characteristics and are not easily resolved. One common technique for removal of DDE prior to TLC is oxidation of the DDE to dichlorobenzophenone with chromium trioxide or other oxidant (Biros et al., 1972; Sherma, 1981; Watts, 1980; Collins et al., 1972). Two studies (Bush et al., 1971; Collins et al., 1972) compared TLC and GC/ECD. In both studies the PCB values obtained were generally comparable, although in the study by Bush et al., the TLC results were generally lower than GC/ECD. Lucas et al. (1980) reported a statistical analysis of semiquantitative determinations of PCBs in human adipose tissue generated by the EPA's National Human Monitoring Program during FY 1972 to 1976. Results were reported only as ranges (not determined, < 1, 1 to 3, and > 3 ppm) for 5,259 samples. The EPA TLC technique (Watts, 1980; Sherma, 1981) was used in this study through November 1974 and a GC/ECD technique involving a single PCB peak quantitation was used thereafter. A total of 3,802 TLC results and 1,457 PGC/ECD results were compared and not found significantly different, High Performance Liquid Chromatography High performance liquid chromatography (HPLC), with ultraviolet and other detectors, has been reported in the characterization of commercial PCBs as a cleanup technique and as an analytical technique. Despite its general applicability in analytical chemistry, HPLC has not been as popular as gas chromatography (GC) for PCB analysis. The major reason is that GC detectors, especially those selective toward halogens, exhibit much lower limits of detection. Since HPLC is basically an instrumental version of the column chromatographic cleanup techniques, described above, it is applicable both as a cleanup 21 �and a determination technique. Some researchers have exploited this and combined cleanup and determination into one step with HPLC (Hanai and Walton, 1977; Van Vliet et al., 1979). Review Articles— Krull (1977) mentioned HPLC in a review of PCB analysis. Lawrence and Tuiton (1978) reviewed the HPLC data on pesticides, including PCBs in their tabulation. Standard Methods-No standard methods utilize HPLC. Primary Literature— Several authors (Brinkman et al., 1976a, 1976b; Veith and Austin, 1976; Albro and Parker, 1979; Brinkman and deVries, 1979) have used HPLC in characterization of commercial PCB products or establishing the chemical behavior of PCBs. HPLC has been used as a cleanup technique prior to gas chromatographic determination (Aitzenmiiller, 1975; Dark and Grossman, 1973; Rohleder et al., 1976; Krupcik et al., 1977; Dolphin and Willmott, 1978). More recently it has been used on a preparative scale to clean up waste and transformer oils prior to CGC/ECD determination (Anonymous, 1982; Chesler et al., 1981). In the course of these investigations, the researchers noted that the CGC/ECD limit of detection was about 100 times lower than the HPLC/UV limit of detection. As HPLC became increasingly popular in the early 1970s, Eisenbeiss and Sieper (1973) performed preliminary investigations of the use of HPLC for pesticide (and PCB) analysis. They concluded that HPLC can be regarded as an alternative or supplementary method to conventional methods such as gas chromatography. Hanai and Walton (1977) developed an HPLC/UV method for determining PCBs in water. No LOD was determined, but good recoveries were obtained for 250-|Jg/ liter Aroclor 1232 spiked into distilled water. The water was pumped directly through the HPLC system and the PCBs subsequently eluted by gradient elution. A similar application (Van Vliet et al., 1979) used an HPLC precolumn to concentrate PCBs from water and then elute them onto the analytical column for separation and determination. Belliardo et al. (1979) developed an HPLC/UV procedure for PCBs in oil and compared it with a PGC/ECD method. The HPLC method was judged suitable to approximate, but not quantitate, the PCB content. Seidl and Ballschmiter (1979) used HPLC/UV to detect biphenyl after dehydrochlorination of PCBs. Electron capture detection of HPLC effluents has been described (Willmott and Dolphin, 1974) for the analysis of PCBs. The LOD of HPLC/ECD is reported to be about 10 times higher than for GC/ECD (Brinkman et al., 1978). Stalling et al. (1980) gave a preliminary description of an HPLC/MS (presumably the chemical ionization MS mode) system for rapid screening for PCBs. 22 �Gas Chromatography Gas chromatography (GC), in combination with various detectors, has been by far the most popular and useful analytical procedure for PCBs. In recent years, capillary column GC (CGC) has been used increasingly, although most investigators still use packed column GC (PGC). The popularity of GC lies in its resolution and speed (most PGC analyses take less than 30 min) and the sensitivity (ECD), selectivity (BCD, HECD), and specificity (electron impact mass spectrometry) pf the available detectors. Separation-Packed column GC (PGC)—Over 215 references were abstracted which used PGC as the analytical separation technique. The vast majority of these used PGC in a routine manner with a common liquid phase. The quality of the chromatography (resolution and tailing) was generally adequate for low resolution separation of Aroclor-derived samples into a "fingerprint" for identification or quantitation. Since the Aroclor mixtures are too complex for resolution, little or no emphasis was placed on improving resolution by PGC. The most common PGC detector has been ECD. ECD has historically required isothermal GC operation (not so with modern instruments). Figures 1 and 2 present some PGC/ECD chromatograms (Mullin and Filkins, 1981). Review articles—By 1971 sufficient work in PCB analysis by PGC had been completed to merit a review (Reynolds, 1971). This was followed by several other reviews (Fishbein, 1972; Hutzinger et al., 1974; Fuller et al., 1976; Krull, 1977; Margeson, 1977; and CMA, 1981). One review by Sherma (1975) was devoted to PGC analysis of PCBs and related chlorinated aromatic pollutants. Standard methods—PGC is the recommended analytical separation method in all but one of the 11 standard methods listed in Table 1. Primary literature--Since over 215 citations on the use of PGC in the analysis for PCBs have been abstracted, a complete discussion of the pri-1 mary literature would be a formidable task. Most of these citations used PGC in a routine manner and included little or no discussion as to why a liquid phase or GC condition was chosen (if they were even mentioned). Several articles are worth noting. Albro et al. (1977) evaluated 13 packed columns ranging in polarity from Apiezon L to OV-225. The number of observed theoretical plates ranged from 491 to 3,833. None of the columns could successfully resolve all PCBs, In the best case, it was calculated that of the 21,945 theoretically possible pairs of PCB congeners, 465 would be indistinguishable using the best column tested. The researchers discussed the use of multiple columns for resolving indistinguishable pairs and concluded that five columns were necessary to resolve all isomers. Thus, using this scheme, each sample would have to be analyzed once on each of five PGC columns to resolve all congeners. 23 �UJ CO O a. CO 01 O O LU K J 34 LU O /A/.; TIME, min UJ CO z O H (Hi) - 4 5 rrn/sor Q. CO UJ QC o UJ H LU a 120 60 TIME, min INJ Figure 1. Comparison of packed column gas-liquid chromatography (top) and capillary column gas-liquid chromatography (bottom) with Aroclor 1242 and 1260 standards (Mullin and Filkins, 1981). 24 �LLJ CO 1 CO LU tr tr g ui UJ Q I 34 TIME, min INJ i1 I 1,(H,, - 45 , 1/Sl'l1 T -- 5 LU •-T z o a. CO LU cc cc . 1 .. ,,h , yj JJ o a • LU o !, 1 i i 100 70 30 "17 INJ TIME, min Figure 2. Comparison of packed column gas-liquid chromatography (top) and capillary column gas-liquid chromatography (bottom) with a milk extract (Mullin and Filkins, 1981). 25 �Albro and Parker (1979) applied this technique to the identification of the components in Aroclor 1016 and 1242. The identity of 44 congeners was reported. A report by Jensen and Sundstrom (1974) presents what may be the highest resolution PGC chromatogram of PCBs. Even though it was operated isothermally, this 5.2-m Apiezon L column resolved 59 peaks in a Chlophen mixture. Capillary column gas chromatography--CGC has not been nearly so popular as PGC, although 42 citations have been abstracted. In recent years the quality of the CGC separations reported have been truly impressive. Despite these advances, no CGC method reported or predicted will separate all 209 PCB congeners. As an example of the overlap problem, Pellizzari (1982a) reported CGC/ECD and CGC/NCIMS identification of PCB congeners in an Aroclor 1016/1254/ 1260 mixture. Of the 103 peaks listed, 73 were identified, but only 43 corresponded to a single congener. The others were possibly two (19 peaks), three (9 peaks), four (1 peak), or even six (1 peak) co-eluting congeners. In addition to the problem of congener separation, interferences may coelute. EIMS can readily discriminate against non-PCBs; however, co-eluting major components may affect the mass spectral response of PCBs. Review articles—CGC was briefly cited (five references) in one review (Sherma, 1975). An Aroclor CGC/ECD chromatogram was included in a CGC monograph (Jennings, 1978) as an application of CGC. Standard methods-1-One of the 11 standard methods in Table 1 recommends CGC, specifically a support coated open tubular (SCOT) column coated with FFAP (free fatty acid phase) for analysis of PCBs in capacitor Askarels. EPA Method 625 (EPA, 1979b) recommends PGC or if desired, capillary or SCOT columns may be used. Capillary gas chromatography is also allowed, if desired, for the analysis of PCBs in transformer fluids or waste oils (EPA, 1981). Primary literature—CGC was utilized in 43 articles abstracted for this review. The level of detail and column specifications span a wide range. Sissons, and Welti (1971) published an early article which characterized many of the PCB isomers in Aroclor 1254. Using an Apiezon L packed column, 23 peaks were resolved, while the same phase on a SCOT column (24,000 to 27,000 plates) separated 65 peaks. The next year, Webb and McCall (1972) performed similar experiments using an SE-30 SCOT column. Although the resolution was poor by today's standards, Biros et al. (1970) used CGC/EIMS to determine PCBs in human adipose even earlier. Krupcik et al. (1971) evaluated metal WCOT columns coated with Apiezon L or OV-101 and found them unsuitable. However, OV-101 on a glass WCOT column gave good results. An example of the separations obtained by CGC using liquid phases of different polarity is shown in Figure 3 (Krupcik et al., 1977). The quality of the chromatography is less than optimal because the GC was operated isothermally. Krupcik et al. (1982) have also reported on the optimization of experimental conditions for the analysis of complex mixtures by capillary gas chromatography. The optimization procedure for complex materials was demonstrated with Aroclor 1242. Forty PCBs were separated 26 �Separation of a PCS mixture by OLC on a glas$ capillary column coated with OV-101 at 200 °C (column E)f ULJ Separation of PCB mixture by GLC on a glass capillary column coated with Carbowax 20M at 200 °C (column FJ. Figure 3. Comparison pf PCB resolution on different columns (Krupcik et al., 1977). �at 170°C using a 40.0 m Carbowax 20M glass capillary column connected to a 75.6 m Apiezon L glass capillary column. Using a 50-m Dexsil 410 glass capillary Albro et al. (1981) have achieved 175,000 effective theoretical plates for 2,3,5,2',3',5'-hexachlorobiphenyl. Resolution of Aroclor 1260, which required an isothermal chromatogram of 5 h, generated 110 peaks, of which only 4 were unidentified. Even at this resolution, the Dexsil 410 did not resolve all congener pairs. Less efficient columns coated with Silar 5 C, Apiezon L, and OV-25 were used to provide different separations which resolved the congener pairs not previously resolved. Although no column performance parameters were given, Mullin and co-workers have achieved impressive resolution by temperature-programmed CGC/ BCD on C-87 columns (Mullin and Filkins, 1981; Mullin et al., 1981) and SE-54 columns (Safe et al., 1982). Pellizzari et al. (1981) have compared a number of capillaries (capillary material, pretreatment, and liquid phase). Apiezon L was judged to be the best of the liquid phases tested (SE-54, C-87, SP-2100, and Apiezon L), for PCB analysis, based on resolution, separation number, and HEETP. Two examples of this column's performance are shown in Figures 4 and 5. This conclusion supports that of several other investigators who have used and recommend Apiezon L (Sissons and Welti, 1971; Albro et al., 1977; Stalling et al., 1978; Albro et al., 1981; Jensen and Sundstrom, 1974; Nakamuna and Kashimato, 1977; United Kingdom Department of the Environment, 1979) or similar hydrocarbon phases for PCB analysis (Mullin and Filkins, 1981). Tuinstra and coworkers (Tuinstra and Traag, 1979a, 1979b; Tuinstra et al., 1980; Tuinstra et al., 1981) have explored the automation of CGC/ECD analysis with autoinjection onto a splitless injector. This approach, although not thoroughly presented in Tuinstra (no mention of sample throughput or automation of data recording and reduction is mentioned), should be pursued by laboratories facing large sample loads. Recently, bonded liquid phases have been made available on capillary GC columns. These exhibit low bleed and background and have long lifetimes. Figure 6 presents a CGC/EIMS chromatogram of a PCB standard on a DB-5 column. It is interesting to note that J&W Scientific, Analabs, and Supelco present CGC/ECD chromatograms of Aroclor mixture in their catalogs. This indicates that CGC is commercially available and that the capillary manufacturers consider their PCB separations good enough to advertise. Comparison of PGC and CGC—The relative merits of PGC and CGC are well-known and apply to the separation of PCBs. CGC provides better resolution, retention time precision, and higher qualitative reliability. PGC yields a simple chromatogram (less data reduction), permits higher sample loading (and therefore possibly lower LOQs), and is generally considered easier to use. Historically, PGC quantitation has been more precise, although it has not been established how much of the imprecision attributed to CGC has been due to poor technique on the part of the analyst. 28 �Figure 4. Electron capture detection of Aroclors 1242, 1260, and 5460 (400 pg each) chromatographed on an Aoiezon L (WCOT) silanized Pyrex glass capillary, 0.20 mm i.d. x 50 m in length. The carrier was 53 cm/s; capillary was temperature-programmed from 150 to 390°C at l°C/min (Pellizzari et al., 1981). Figure 5.. Electron capture detection of PCB in an extract of yellow perchSee Figure 4 for chromatographic conditions (Pellizzari et al., 1981). 29 �Sample: Combine^ Aroclor 1248,1254,1260 250ng/^r&DCB l^bngX/il, 1/il Injection 10 800 20:00 1000 25:00 1200 30:00 1600 40:00 1400 35:00 1800 45:00 2000 50:00 2200 SC;AN 55;QQ TIME Figure 6. Scanning capillary column gas-liquid chromatography/mflsis spectrometry analysis of a mixed Aroclor standard used to establish retention windows for the CGC/MS-selected ion monitoring analysis of PCBs. Instrumental parameters: column, 15-m, fused silica, DB-5; column temperature, 80°C for 2 min, 8°C/min to 300°C; helium carrier at 2,5 psl; J&W on column injector. (J. S. Stanley and C. L. Haile, Midwest Research Institute, personal communication, 1982). 30 �Figures 1 and 2 (Mullin and Filkins, 1981) present graphic comparisons of PGC and CGC results for PCBs. Similar results have been presented by Onsuka and Comba (1978). Detectors-GC detectors are classified as either universal or selective. The BCD and HECD are highly selective toward halogenated compounds. This selectivity, coupled with its extreme sensitivity, has made BCD very popular for analysis of trace levels (residues) of pesticides and PCBs and has, in fact, had a significant role in regulatory actions on these classes of compounds. FID is the most common GC detector and is a universal detector, giving similar responses for most organic compounds. Thus, FID would be unsuitable for detection of PCBs in a complex matrix. Mass spectrometry and Fourier transform infrared spectrometry (FTIR) are in essence both universal and selective GC detectors. By focusing on a spectral property characteristic of a compound or class of compounds, these detectors can be quite specific. However, by using the full spectral range, nearly any compound eluting from the GC will be detected. Due to the much higher information content of mass and infrared spectra, identifications by GC/MS or GC/FTIR are generally made with greater certainty than by other detectors . The analysis of PCBs generally requires selectivity and sensitivity. Usually, even after cleanup, PCBs are a minor component of the sample, mixed in with other halocarbons (e.g., DDE), hydrocarbons, lipids, etc. Thus, the detector must selectively detect PCBs in the presence of other compounds present at orders of magnitude higher concentration. Furthermore, the levels typically observed in food, biota, tissue, soil, and other matrices of interest are in the parts per billion range. These levels strain the capabilities of even the most sensitive detection device such as BCD, resulting in a large number of "not detected" values in many reports. The choice of detector often depends upon the level of analytes. LOW concentrations demand a detector capable of detecting low amounts (high sensitivity) . Figure 7 presents the typical range and detection limits for most of the GC detectors used in analysis for PCBs. The detection limit of the HECD is 10 lx g with a linear range up to about 10 2 g, as measured for lindane (Anderson and Hall, 1980). As can be seen, BCD exhibits the lowest limit of detection (LOD). The reported LOD for PCBs in a variety of matrices are listed in Table 2. Comparison of the reported LODs is difficult because no standard definition of LOD was used. Glaser et al. (1981) followed a rigorous definition and experimentally determined the LOD with a fair degree of confidence, while other investigators clearly guessed at the LOD based on this work. The issue is further clouded by inconsistency in discussing the LOD with respect to the instrumental determination versus the entire procedure. Some LODs are reported for standard solutions, while others take into account the interferences in the matrix which often raise the LOD considerably. 31 �Detection Mode Selectivity Grans NT13 10 "12 ID'11 10"10 HT9 KT* Ifl"7 10"* 18"s 10"* NT3 10"2 1 Thermal Energy Analyzer Photoionization EleclionXiptuie Mass Spectrometry Election Impact Multiple Ion Del. Neg. Chemical Ion. Fluorescence Rame lonization Thermal Conductivity FT/IR t)V Figure 7. Detection limits/dynamic range for several instrumental methods (Pellizzari, 1981) �TABLE 2. REPORTED LIMITS OF DETECTION FOR PCBs Converted LODa (Mg/g or ppm) Instrument Reported LOD GC/ECD 0.065 yg/£ 0.5 ppb 6.5 ppb 50 ppb 1-0.1 ppb/isomer 0.5 ppm 1 ppm 0.5 ppm 0.6 Mg/£ 1 ng/m3 0.1 ng/m3 0.000065 0.0005 0.0065 0.05 0.001-0.0001 3 MgM 1 ppm 0.003 1 ppm 0.5 1.0 0.5 O.g006 NA, NA Substance Matrix Aroclor 1242 Aroclors Aroclors Aroclors Isomers Aroclors Total PCB Total PCB Perchlorinated Perchlorinated Theoretical per isomer Aroclor 1260 10 homologs Dist. water River water Pure solution Milk Vegetable Transformer fluid Reference Air Air Glaser et al., 1981 Kuehl et al., 1980 Teichman et al., 1978 Tessari and Savage, 1980 Tuinstra et al., 1981 Kirshen, 1981 Chesler et al., 1981 Balya and Farrah, 1980 Stratton et al., 1979 Stratton et al., 1978 Lewis et al., 1977 Blood serum Pigments Kreiss et al., 1981 DCMA, 1980 Oil Oils Ground water GC/HECD 1 mg/kg 1.0 Aroclors Oil EPA, 1981 GC/EIMS 30 | g £ j/ 36 |ag/£ 0.01-0.2 pg/£ 0.030 0.036 0.01-0.2 Aroclor 1221 Aroclor 1254 Single isomer 5 ppm 5 Single isomer Dist. water Dist. water Industiral sample extract Chlorinated hydro carbons Glaser et al., 1981 Glaser et al., 1981 Tindall and Winninger, 1980 Collard and Irwin, 1982 HREIMS 10 ppb 0.01 Aroclors Biological extracts Safe et al. , 1975 GC/NCIMS None (continued) �2 ^continued) Converted Instrument Reported IOD Dir. Probe NCIMS •*• 1 ppb tLC 0.5 ppm < 0 0 ppm .4 0.1 pg 0.05 jig 0.2 jjg 1 }ig a -o 0.031 Substance NSC Reference Matrix 0.5 < 0.04 Aroclor Aroclor Aroclor Aroclor Aroclor Aroclor Biological Dougherty et al . , extracts 1980 Adipose Milk Animal tissue NS Animal tissue NS Bush and Lo, 1973 Savage, 1973 deVos and Peet, 1971 Koch, 1979 Mulhern et al. , 1971 Ismail and Bonnerj 1974 Converted to common units of micrograms per gram (parts per million) assuming 1 ml = 1 g density. b NA = not applicable. c (pg/g or ppffl) NS = not specified. �Reviews Articles-Every review article abstracted covered the subject of BCD detection of GC effluents (Riseborough, 1971; Reynolds, 1971; Fishbein, 1972; Linear, 1973; Hutzinger et al, 1974; Sherma, 1975; Fuller et al., 1976; Margeson, 1977; Krull, 1977; Safe, 1976). Fishbein (1972), Sherma (1975), and Hutzinger et al. (1974) all reviewed the use of electrolytic conductivity detectors for PCB determina*tion. Safe (1975) and Hutzinger et al. (1974) discussed the use of flame ionization detection (FID), mostly with respect to calibration of ECD or establishing BCD response factors. Hutzinger et al. (1974) did mention that for the mono- and dichlorobiphenyls FID and ECD sensitivities are comparable. Standard Methods— As noted in Table 1, most of the standard methods specify ECD as either the detector or one of the options. FID is the detector prescribed in the American Society for Testing and Materials (1980a) procedure for determining PCBs in capacitor Askarels. In this case, the matrix is well-characterized and generally contains no other compounds in the PCB retention window. HECD is permitted as an alternate detector in three procedures (EPA, 1978; EPA, 1981; FDA, 1977). Electron capture detection—Based on literature citations and number of samples processed, the electron capture detector (ECD) has been the most common detector for GC analysis of PCBs. ECD is extremely sensitive for PCBs. It does, however, detect many nonPCB compounds (halogenated pesticides, PCNs, chloroaromatics, phthalate and adipate esters, and other compounds) which may be differentiated from PCBs only on the basis of retention time. Figure 8 illustrates the potential interferences from chlorinated pesticides. A major disadvantage of ECD is the range of response factors (Tables 3 and 4) which different PCB congeners exhibit. This seriously inhibits reliable quantitation. The opposite trends in the two tables presumably result from differences in the equations used (i.e., whether the PCB response is in the denominator or numerator). The earlier PGC/ECD work (Table 3) has a range of about 1,400, while the later CGC/ECD work (Table 4) has a range of only about 120. This may be a function of the differences in detector design and GC column throughput. In addition, the CGC is temperature-programmed, while PGC data were presumably obtained isothermally. Despite these differences, both tables clearly illustrate that, even within a homolog, the % RSD is very large and would result in poor accuracy if quantitation involves extrapolation from one isomer to another. A recent report of the ECD relative response factors for all 12 octachlorobiphenyls showed a range from 0.007 to 2.644 with a RSD of 35% (Mullin et al., 1981). This also illustrates the problem of PCB quantitation by ECD. This subject is treated in more detail in the Quantitation section. Over 175 references were abstracted in which ECD was used as a GC detector. Any novel aspects of the articles dealt with qualitative or quantitative aspects of ECD and will be treated in the appropriate subsections below. 35 �1248 4%SE-30/6%OV-210 Chromatograms of three AROCLORS on column of \& SE-30 / 6$ OV-53.0. Column temp. 200°C., carrier flow 60 ml/min., % detector, electron, attenuation on an E-2 10 x 16; dotted line a rdxture of chlorinated pesticides, identity and injection concentration given below: 1. 2. 3. U, 5. 6. 1.5 ng Diazinon Heptechlor -- 0.03 Aldrin .Ol}5 Kept.Epox. .0? .09 p,p»-CDE Dieldrin — .12 ?. 8. 9. 10. 11. o,p'-DDT ~ O.A ng p,p«-DDD — .& p,pf-EBT — .30 Mian — .75 Hethoxychlor .60 u AROCIOR 1260 Figure 8, Packed column gas-liquid chromatography/electron capture detector chromatograms ill-Tistra-ting potential Interferences between, pesifcides and PCBs <Uatts, 1980). �TABLE 3. RELATIVE MOLAR RESPONSES OF ELECTRON CAPTURE AND FLAME IONIZATION DETECTORS TO SOME CHLOROBIPHENYLS3 Chlorobiphenyl 2342,2'~di 2,4-di 2,6-di 3,3'-di 3,4-di 4,4'-di 2,4,4'-tri 2,2' ,4,4'-tetra 2,2* ,6,6'-tetra 3,3' ,4,4'-tetra 3,3' ,5,5'-tetra 2,3,4,5-tetra 2,3,5,6-tetra 2,2' ,4,4',6,6'-hexa 3,3',4,4',5,5'-hexa 2,2',3,3',4,4',6,6-octa 2,2',3,31,5,5',6,6'-octa deca N Mean SD RSD ( ) % Relative molar response Electron capture Flame ionizatipn 1.00 0.20 1.10 5.16 17.7 32.0 6.10 15.2 5.97 135 106 20.6 396 320 367 259 347 726 1,180 1,150 1,410 21 310 438 140 a Taken from Hutzinger et al., 1974, and Safe, 1975. 37 1.00 0.92 0.87 0.99 0.86 0.91 0.94 0.86 0.81 0.78 0.87 0.90 0.87 0.85 0.87 0.71 16 0.88 0.07 8,3 �TAB1E 4. COMPARISOH OF RELATIVE RESPONSE FACTORS BETWEEB (GC)2-ECD, GC-EIMS (MOLECULAR ION) AJJD (GC)2-NICIMS { / 35) FOR HOMOLOGOUS SERIES OF PCBs* az Homolog series Range (GC)2-ECD3 Mean ± S.D. ( RSD) Hc % 1C1{3) 15.089-39.342 29.589 ± 12.78 (43) 2C1(12) 0.425-10.641 4.271 ± 3.83 (90) 3C1(24) 0.328-2.136 1.193 ± 0.68 (58) AC1(42) 0.385-2.229 1.074 ± 0.41 ( 8 3) 5C1(46) 0.462-8.481 1.266 ± 1.29 (102) 6C1(42) 0.391-1.912 0.973 ± 0.335 (35) 7C1(24) 0.402-2.432 1.220 ± 0.419 (34) 8C1{12) 0.925-2.602 1.514 ± 0.679 (45) 9C1(3) 1.005-1.816 1.291 ± 0.45 (35) 10C1(1) 1.168 U) 00 3 9 9 31 35 37 21 10 3 1 Range 0.456-1.787 2.881-21.199 0.721-10.901 0.102-4.267 0.465-1.216 0.369-1.440 0.236-1.192 0.241-1.116 0.066-0.565 - Overall : 0.328-39.342 ( 120:1) ~ 0.924 ± 8.343 ± 2.921 ± 2.058 ± 0.805 ± 0.817 ± 0.703 ± 0.573 ± 0.354 ± 0.418 0.75 6.17 3.64 1.02 0.27 0.29 0.30 0.26 0.26 Overall : 0.066-21.199 (y 3 0 1 2:) * From Pellizzari et al. (1982). a STI data. b Martelli et al., 1981. c N = number of PCS isomers included in measurement. A All values are relative to ottachloronaphthalene. e Responses were relative to lowest response for each group. f ( ) = number of theoretical isomers possible. (GC)2-HICIMS3 Mean ± S.D. ( RSD) N % (81) (74) (125) (50) (33) (36) (43) (46) (73) Range GC-EIMSb Mean ± S . . « RSD) N 0 3 1.000-1.090 8 1 . 000-2 . 062 7 1.000-1.627 16 1.000-2.146 12 1.000-1.013 16 1.000-1.321 13 8 1.000-1.359 3 1 1.050 1.736 1.400 1.549 1.004 1.153 ± ± ± ± ± ± 0.04 0.30 0.24 0.33 0.01 0.11 1.179 ± 0.25 - (3.8) 3 (17) 10 9 (17) (21) 11 (0.7) 3 (9.6) 7 0 2 (22) 0 0 �Flame ionization detection--Flame ionization detection (FID) is the most commonly used GC detector because of its sensitivity and universality. Although some investigators have used FID for PCB determination in samples, it has generally been used only for calibration of response factors or other method development work. FID has been used for determination of PCBs in environmental samples (Mizutani and Masayoshi, 1972; Modi et al., 1976; Lao et al., 1976; Onsuka and Comba, 1978). Biros (1971) split the GC effluent to FID for quantitation and EIMS for identification. Cook et al. (1978) and Zimmerli (1974) used FID to detect biphenyl following dehydrochlorination of PCBs; a technique termed carbon skeleton chromatography. Most of the FID applications have been in establishing response factors, characterizing Aroclors, or other method development areas (Webb and McCall, 1972; Ugawa et al., 1973; Dexter and Pavlou, 1976; Boe and Egaas, 1979; Albro and Parker, 1980; Albro et al., 1981; Stalling et al., 1982). An example of the use of FID is presented in Table 3, where the molar responses of FID and ECD were compared (Hutzinger et al., 1974; Safe, 1975). The rtna 1 conductivity (TCP)--Hirwe et al. (1974) used TCD to characterize Aroclor mixtures. This application is similar to many of the FID applications. Electrolytic conductivity--The Hall (and its predecessor, the Coulson) electrolytic conductivity detector (HECD) has been used often in PCB analysis. It is much less subject to interference from nonhalogenated compounds than ECD and the response is proportional to the number of chlorines. The high limit of quantitation and difficulty of operation are the disadvantages of this detector. Webb and McCall (1973) and Sawyer (1978) used Hall detection in the characterization of Aroclor standards. Serum et al. (1973) used it, ECD, and electron impact mass spectrometry as PGC detectors in analysis of paper products for PCBs and other compounds. Hofstaedter et al. (1974) determined that sulfur compounds in certain petroleum oils gave positive interferences in PGC/ECD determinations of PCBs. Flame photometric, microcoulo>metric, and Hall detectors were used to characterize the PCBs and interferences. Chesler et al. (1981) characterized oil products in the preparations of National Bureau of Standards standard reference materials for PCBs in oil. They used both ECD and HECD as CGC detectors. The ECD was found to be more sensitive than the HECD by two orders of magnitude and easier to maintain in a noncontaminated state. However, ECD response factors varied for different PCB isomers, whereas the molar response to chlorine which is obtained from the HECD appeared to be constant. The HECD exhibited a wider linearity range and is more selective as it responded only to halogenated compounds. An interesting, though tangential, use of HECD was presented by Dolan et al. (1972), Dolan and Hall (1973), and Su and Price (1973). By adjusting the HECD operating parameters they selectively detected organochlorine pesticides in the presence of PCB interferences. Electron impact mass spectrometry—Electron impact mass spectrometry (EIMS) ranks second only to ECD in popularity as a GC detector for PCBs. 39 �Electron impact has been and continues to be the most widely used MS ionization technique. While the chemical ionization (CI) and negative chemical ionization (NCI) techniques are often more sensitive, their operation is more complicated and variation in the spectra and response are higher. EIMS has been applied to PCB determination using both direct probe; and gas chromatography for sample introduction. Early work generally employed the then-exotic technique as a confirmation technique. In recent years, as GC/ EIMS has become more routine, more and more analysts have chosen GC/EIMS as the primary technique. Review articles—The application of EIMS to analysis for PCBs has been reviewed by Fishbein (1972), Oswald et al. (197A), Hutzinger et al. (1974), and Safe (1975). Standard methods—As listed in Table 1, several of the standard methods use GC/EIMS, either as the primary analytical technique or as the con™ firmatory technique. Primary literature—In 69 articles abstracted, EIMS was used. The applications ranged from confirmation to routine use, from direct probe to CGC, and from Aroclor characterization to analysis of dirty samples. Among the pioneers, Biros et al. (1970) used CGC/EIMS to determine PCBs in human adipose tissue; Sissons and Welti (1971) used CGC/EIMS in the characterization of Aroclor 1254; and Bonelli (1972) presented PGC/EIMS data for an Aroclor 1254/chlorinated pesticide mixture. In addition to Sissons and Welti (1971), Webb and McCall (1972, 1973), Ugawa et al. (1973), and Oswald (1974) employed GC/EIMS in characterization of commercial PCB products. Using both electron impact and chemical ionization mass spectrometry, Oswald et al. (1974) were able to differentiate some isomers in complex mixtures from their spectra. While full spectra provide the most qualitative information, the use of selected ion monitoring enhances the instrument sensitivity and seleC" tivity and simplifies data interpretation. Examples of this technique have been presented by Beggs and Banks (1976), Eichelberger et al. (1974), Martelli et al. (1981), Collard and Irwin (1982), Erickson and Pellizzari (1977, 1979) and Tressl and Wessely (1976). Especially with the more highly chlorinated homologs, several m/z values are available for monitoring. Eichelberger et al. (1974) addressed the criteria for selection—intensity and probability of interference from higher homologs or other compounds. A compromise between full scan and SIM techniques is mass chromatpgraphy. Full spectra are collected and then ion intensity versus file position plots are extracted from the data by the computer. Thus, mass chromatography has the ease of interpretation of SIM but higher LOQs since full spectra are collected. These full spectra are available for qualitative use if needed. Canada and Regnier (1976) presented a technique which used mass chromatography to monitor the ion ratios in the PCB isotopic clusters. 40 �Another compromise technique, limited njass scanning (WMS), involves (as the name implies) scanning the spectrometer only over the mass range of interest (e.g., molecular ion cluster). This permits the spectrometer to spend! more time on the ions of interest and thus achieve better sensitivity than the full scan mode. Tindall and Wininger (1980) utilized IMS in theip PGC/CIMS analysis of commercial products for incidental PCBs, Albro and Parker (1980) utilized PGC/EIMS as part of a general an^ alytical scheme for chlorinated aromatic pollutants. Positiye chemical ionization mass spectrometry—Positive chemical ioni» zation (Cl) mass spectrbmery (CIMS) is oneof the "soft" ionization techniques spectromery one of techniques which tend to produce fewer fragments. Thus, the spectra are simple and the molecular ion is generally one of the most intense peaks, However, with PCBs, the electron impact spectra generally exhibit good molecular ions, reducing the advantages of Cl. Another problem with Cl is that the ionization, process depends on a reagent gas introduced with the sample into the source. Slight changes in gas pressure, source temperature, and electronic conditions can affect the reaction conditions and thus the spectrum (both fragmentation pat* terns and overall intensity). Thus, Cl is not as reproducible as electron impact, either qualitatively or quantitatively. Several researchers have utilized GC/CIMS for determination of FCBs,* Oswald et al, (1974a), Sawyer (1978), and Cairns and Siegmund (1981) characterized standard samples. Oswald et al. (1974b), lida and Kashiwagi Stalling (1976), and Cairns and Jacobsen (1977) applied GC/CIMS to PCB tabolites , environmental samples , and food samples . Dougherty et al. (1973) reported the use of direct probe positive and negative CIMS for the analysis of human adipose tissues for PCBs. Stalling et al. (1980) reported an HPIC/MS technique for PCBs which is presumed to use the Cl mode. This preliminary report speculated that HPLC/MS could be useful as a screening technique for environmental samples. A related but often defined as separate technique, atmospheric pressure chemical ionization has been reported for PCB determination. Dzidic et al. (1975) reported subpicogram detection of 2,3,4,5,6-pentachlorobiphenyl, and Thomson and Roberts (1980, 1982) used the technique for in situ detection, of PCBfi in clay and soil. Negative chemical ionization mass spectrometry — Negative chemical tion (NCI) mass spectrometry (NCIMS) is similar to both CIMS and ECJD- The basic difference between negative and positive Cl is the polarity of the vari« ous voltage potentials in the spectrometer and the detector. Many of the chemical reactions in the NCI source and the BCD are the same. NCI and BCD exhibit similar detection limits and selectivities toward chlorinated compounds, thus the interest in NCI. The reproducibility problems of Cl are also present in NCI. The range of response factors found with ECD are also found with NCI. NCIMS, a relatively recent technique, isi still considered to be a research technique. 41 �The group led by Dougherty has published extensively on the methods iand application of (direct probe) NCIMS (Dougherty et al., 1973; Kuehl et al., 1980; Dougherty et al., 1980; and Dougherty, 1981a, 1981b), The technique is described as rapid and highly selective toward halogenated compounds. The latter advantage reduces the need for cleanup and, according to Dougherty (1981a), permits the analysis without the customary GC separation. Kuehl et al. (1980) used both CGC/EIMS and CGC/NCIMS to analyze fish samples for a variety of chloroorganics, including PCBs. The electron impact spectra were used for primary identification, although the NCI spectra were also of great value. Figure 9 presents the NCI and electron impact TICs for comparison. The NCI is much more selective toward the halogenated compounds, eliminating the broad hump which is presumably a complex mixture of lipids and oils from the fish matrix. Pellizzari et al. (1981) have used CGC/NCIMS for analysis of PCBs and have characterized the instrument operation parameters. The choice of reagent gas and its pressure markedly affect the relative intensities of the major peaks (m/z 35 and 37, molecular ion, etc.). The response factors for several PCB congeners are presented in Table 4 (Pellizzari et al., 1982). As with ECD, the range of the response factors is broad. This would probably make quantitation by extrapolation from a single calibration isomer inaccurate. While not directly used for PCB determination, CGC/atmospheric pressure negative chemical ionization mass spectrometry was shown to be both sensitive and selective for PCDDs in the presence of PCBs (Mitchum et al., 1982). Vith proper selection of masses and ionization conditions, this technique may be highly selective for PCBs. High resolution electron impact mass spectrometry--'HREIMS is capable of obtaining precise and accurate mass measurements of a peak. This known mass can correlate with only a few possible molecular formulas, As reviewed by Safe (1975), HREIMS is particularly useful for chlorinated cqmpounds because the chlorine mass defect clearly distinguishes a halocarbon from a molecule containing only carbon, hydrpgen, nitrogen, and oxygen. Safe (1976) and Safe et al. (1975) have reported the application of HREIMS to the analysis of crude goat urine extracts and other biological samples for PCB and PCT metabolites. The reported 10-ppb detection limit and the rapid analysis time (no GC separar tion is used) would appear to make this technique a suitable technique for rapid screening of samples for the presence of PCBs. The lack of work in this area, however, suggests that other considerations must reduce the applicability of HREIMS. Hass and Friesen (1979) illustrated the need for HREIMS to separate interferences (if not chromatographically separated). The spectrum in Figure 10 shows that DDE, TCDD, and PCB would have given one peak under low resolution conditions. 42 �NEGATIVE CHEMICAL IONI2ATION I 3 U c O 75 .o JL 1000 2000 Spectrum Number 3000 ELECTRON IMPACT 1000 2000 3000 Spectrum Number Figure 9. Total ion current profiles for negative chemical ionizatjion data (upper) and electron impact data (lower) obtained on a Finnigan 4000 glass capillary GC/MS system for Ashtabula River fish sample (Kuehl et al., 1980). 43 �1 322.000 321.900 T I 321.800 Figure 10. Partial high resolution mass spectrum obtained from 2.5 x 10"10 g TCDD plus matrix from 10 g human milk, illustrating potential interferences in low resolution mass spectrometry (Hass and Friesen, 1979). Nonchromatographic Methods— This section presents a variety of miscellaneous methods reported for the determination of PCBs, Nuclear magnetic resonance (NMR) spectrometry—Wilson and Anderson (1973) used both 13C and *13. nuclear magentic resonance (NMR) to characterize the chemistry of selected PCBs. No attempt at analysis of real samples was made. Levy and Hewitt (1977) reported the analysis of PCB mixtures by 13C NMR, but noted that the technique was not as useful for higher homologs. Hutzinger et al. (1974) included a discussion of the NMR characteristics of PCBs in their review book. Synthetic congeners have been characterized by proton NMR (Mullin et al., 1981). Infrared (IR) spectrometry—Hutzinger et al. (1974) discussed the infrared (IR) spectral properties of PCBs. Webb and McCall (1972) used IR and other techniques to identify 24 PCB congeners in Aroclor 1221. Radioimmunoa ssay--Albro and coworkers have reported preliminary results in the development of a radioimmunoassay (RIA) method for PCBs (Albro et al., 1979; Kohli et al., 1979; Luster et al., 1979, 1981). Suggestive evidence is presented indicating the feasibility of employing radioimmunoa$says for determining the Aroclor product number and concentration in environmental samples (Luster et al., 1979). The assay requires an antiserum for each isomer but is termed fairly specific. Other techniques—Interrupted-sweep voltametry has been applied to the identification of PCBs, yielding positive identifications (Farwell et al., 1975). Plasma chromatography has been reported to give characteristic qualitative data for PCBs (Karasek, 1971). One report utilized neutron activation analysis for the determination of PCBs in dosed rats (Mamri et al., 1971), 44 �identification and quantitation of PCBs (EPA, 1972; Brownrigg et al., 1974; Brownrigg and Hornig, 1974). The limit of detection was reported to be as low as 0.01 ppm. DATA REDUCTION Depending on the detection and output system, data may be presented to the analyst as strip chart recorder chromatograms, digitized chromatograras, numerical peak integrations, mass spectra, MS selected ion monitoring plots, MS total ion current plots, etc. Computers can easily reduce the analyst's work in data reduction and should be used for MS data acquisition and reduction. However, excessive reliance on data system output without interaction by a qualified analyst can yield spurious results. The first task in data reduction is to qualitatively identify the analyte. Quantitation can be attempted only after a positive qualitative identification. Qualitative The qualitative aspects of the analysis are all too often overlooked. Especially with PCBs, differences in the qualitative assessment of a sample can dramatically affect the quantitative results. In standard methods or other work where two or more analysts are expected to produce comparable results, the qualitative assessment of the data must be carefully specified. As an example of how qualitative interpretation of results can affect quantitation, a round robin study was recently conducted to assess the interlaboratory variability of PCB analysis in commercial products (Pittaway and Homer, 1982). Eleven data sets, generated by PGC/ECD, PGC/ EIMS, PGC/EIMS SIM, CGC/EIMS, and PGC/FID, were acquired (see Tables 5 and 6). While most of the techniques were sufficiently specific to differentiate PCBs from interferences, the PGC/FID was not. Since no attempt was made by the analyst to differentiate PCBs from interferences in this case, the PGC/FID quantitation was 28 times higher than the mean of the other analyses. Qualitative assessment of results depends to a large extent on sample type and its pretreatment. In samples where the presence of PCBs has been well-established (e.g., adipose tissue) the qualitative burden is not nearly so great as for samples in which PCBs are not expected. In many PCB procedures, the cleanup involves a rather specific liquid chromatographic separation which separates PCBs from most organochlorine pesticides. In addition, the use of specific detectors (BCD, HECD) reduces the probability of interferences and increases the confidence in identification. Better still, MS provides spectra of the eluent which may be compared with those of authentic compounds to give high confidence identifications. Finally, the retention time of the eluent should match that of a standard. CGC, and better yet high precision CGC, gives much more precise retention times than PGC and increases confidence. In the case of Aroclor (or similar mixtures) derived PCBs, a pattern of isomers usually resembles the pattern of a standard. This has been a common qualitative technique in residue analyses, especially when PGC/ECD is the analytical procedure, 45 �TABLE 5. SUMMARY OF LABORATORY TECHNIQUES USED FOR THE CMA ROUND-ROBIN STUDY (Erickson, 1982) MS calibration method GC calibration method PGC/ECb PGC/EIMS Confirmation only ? B & J Dil/Inje PGC/EIMS SIMf Autotune C Sonication/ Inj (some diluted) PGC/MSJ PFTBAJ D Sonication/ Inj (some diluted) PGC/FID E DI and Dil/Inj PGC/MS-* SIM PFTBA F Heat /Inj ; Dil/Inj; DI PGC/EIMS SIM G DI H Work- up technique Lab A DIa Analysis technique Integration technique Calculation equations LOD presented PHC No Yes EG8 (3 pts) Areas Yes Yes1 ES Ion Intensities No Yes No Yes ? ES ES (3 pts) Areas No Yes ? ES, RFk Area No Yes PGC/EIMS SIM ? ES (1 pt) Area Yes Yes Dil/Inj PGC/EIMS SIM ? RF Area Yes Yes I Dil/Inj CGCn/EIMS SIM ? RF Area Yes Yes K Acid Digest/ Extract/Inj ; Some heated CGC/MSJ IS, RF° Summed Areap No No See footnotes for Table 6. DFTPP �TABLE 6. SUMMARY OF LABORATORY TECHNIQUES USED FOR THE CMA ROUND-ROBIN STUDY (Erickson, 1982) Lab No. ions/homolog Isotope ratio mentioned Reporting units QA discussed? No. standard isomers Comments A 7 7 ppm No 18 Too many significant figures. little discussion. B &J 1 No mg/kg Yes 24 Some full scan confirmation; good discussion, good work overall. C 1 No mg/£; ppm No 10 Sample preparation is good. D " " mg/*; ppm No 5 E 3 Yes HS/8 (?) Yes 25 F 21 No H8/8 Yes iom G 4-5 Yes ppm Yes 20 Good ion/ retention time table; % recovery calculated from standard addition (were results corrected?) H I 1 1 No No ppm ppm No No 10 10 H and I are same organzation; no discuscussion of differences observed in two methods. No ppm Yes 12 Limited recovery study (•*• 100% recovery); injection precision measured for one sample. K Very Major interferences suspected — no caveats or discussion of divergent data with "Lab C." No individual isomers reported. Standard addition attempted but failed; use of Aroclor mixtures as standards is of dubious value. �FOOTNOTES FOR TABLES 5 AND 6 a Direct injection. b Packed column GC/electron capture detector, c Peak height. d Packed column GC/electron impact mass spectrometry (full scan). e Dilute and inject, f Selected ion monitoring, ^g External standard. h Details presented—two integration methods used, one designated "B"; one designted "J.1 i Estimated, no details, -P- j Unspecified operating mode, 00 k Response factor. 1 One ion from parent cluster and one ion from fragment cluster, m Aroclor mixtures. n Capillary column GC. o Internal standard (details unclear). p Areas of three (two for Cjo^Cl) most intense ions in parent cluster summed, q Limited scan technique used; all ions in parent cluster were observed. �Review Articles— Only five reviews mentioned the qualitative aspects of PCS analysis (Reynolds, 1971; Sherma, 1975; Safe, 1976; Fuller et al., 1976; and Margeson, 1977). Generally, the qualitative discussion was cursory. Reynolds (1971) advised use of pattern recognition (against the Aroclor standards). Sherma (1975) recommended a slightly more formal approach—comparison of retention times in samples and standards. Standard Methods— As noted in Table 1, less than half of the standard methods discuss qualitative analysis. In the PGC/ECD methods pattern recognition is the qualitative procedure. Often chromatography on two GC colums of different polarity is used to enhance the confidence of the verification. An example of good qualitative guidance for the analyst is found in the EPA protocol for analysis of PCBs in transformer fluids and waste oils (EPA, 1981). Locate each PCB in the sample chromatogram by comparing the retention time of the suspect peak to the retention data gathered from analyzing standards and interference free Quality Control Samples. The width of the retention time window used to make identifications should be based upon measurements of actual retention time variations of standards over the course of a day. Three times the standard deviation of a retention time for each PCB can be used to calculate a suggested window size; however, the experience of the analyst should weigh heavily in the interpretation of chromatograms. In methods where GC/EIMS is specified, both mass and retention time may be used for identification. None of the standard methods adequately address the selection of ions and the permissible abundance ratios. Primary Literature— Less than half of the 38 articles abstracted contained mention of the qualitative criteria used in identification of PCBs. A representative qualitative criterion was that the chromatogram exhibits a "typical Aroclor pattern" (Gordon et al., 1982; Giam et al., 1972; Ofstad et al., 1978; Kirshen, 1981). Several publications dating back to the landmark work of Sissons and Welti (1971) concentrated on the identification of PCB isomers in commercial (Aroclor, Chlophen, etc.) mixtures (Tas and deVos, 1971; Tas and Kleipool, 1972; Armour, 1972; Willis and Addison, 1972; Paasivirta and Pitkanen, 1975; Jensen and Sundstrom, 1974; Stalling et al., 1978; Neu et al., 1978; Zell et al., 1978; Ballschmiter and Zell, 1980; Pellizzari et al., 1981; Tuinstra et al., 1981). The objective of most of these papers was to characterize the commercial mixture as an aid to quantitation in environmental samples or for toxicological information. Several researchers mentioned the comparison of retention times or relative retention time in the sample and an Aroclor standard for PCB identification (Webb and McCall, 1972; Onsuka and Comba, 1978; Pellizzari, 1982; 49 �Tuinstra et al,, 1980). A much more specific identification scheme involves use of one of the retention index (RI) (e.g., Kovats) schemes. Several publications contain tabulations of RIs (Sissons and Welti, 1971; Zell et al., 1978; Neu et al., 1978; Ballschmiter and Zell, 1980; Albro et al., 1981; Albro and Fishbein, 1972a, 1972b). Since all 209 PCB congeners are not available, a scheme of predicting RIs has been developed based on the half-RI values for the various chlorination positons on one of the benzene rings. Sissons and Welti (1971) first proposed this system, which was expanded upon and further validated by Albro and Fishbein (1972) and Albro et al. (1977), The use of half RIs permits the analyst to qualitatively identify all 209 PCBs on the basis of their retention time, although the level of confidence in this identification is low. Using state-of-the-art chromatography, RI measurement precision of ± 0.05% has been reported for PCBs (Neu et al., 1978), and with full optimization, precision of ± 0.01% has been predicted (Neu and Zinburg, 1979). Recently Dunn et al. (1982) used a computerized pattern recognition technique to evaluate CGC/ECD data on PCBs in sediments, water, benthos, and fish. The computer program can detect incorrect assignments (i.e., Aroclor 1242 instead of 1260) and abnormal samples. This appears to be a most promising technique for interpretation of large numbers of PCB determinations. When a mass spectrometer is used as the GC detector, an additional qualitative dimension is available. The mass spectra of a PCB are distinctive due to the cluster of masses generated by the presence of two chlorine isotopes in nature. If sufficient material is present to obtain full mass spectra, the unknown can be reliably identified by comparison with spectra of standards or spectral compilations (Stenhagen et al., 1974; Mass Spectrometry Data Centre, 1970; Heller and Milne, 1978). The quality of the spectrum required for identification needs to be defined (Christman, 1982). As mentioned above, the natural isotopic abundance ratios yield a characteristic pattern. The use of these ratios in selected ion monitoring can provide qualitative information when full mass spectra are not obtained. The actual ratios have been tabulated for PCBs (Rote and Morris, 1973) or may be readily calculated. This approach has been utilized (Canada and Regnier, 1976) with complex samples (Erickson and Pellizzari, 1977, 1979). Even though the natural isotopic abundance ratios are constant, instrumental variances and interferences can affect the observed ratio. Thus, tolerance criteria need to be utilized by the analyst. Tindall and Wininger (1980), in a method designed to determine PCBs even if they are not "Aroclor-derived," established qualitative criteria: Each peak in the chromatogram is evaluated to determine if it is a PCB peak. Peaks must meet these criteria to be labeled PCB peaks for quantitation: (1) the peaks of the characteristic ions must maximize at the same retention time; (2) the peak must be in the proper retention time window; and (3) the relative peak intensities of the molecular ions must be within ± 15% of the theoretical ratio. This tolerance is arbitrary and can be made larger for very low concentrations of PCBs where statistical variations in peak intensity become large. 50 �Work by Collard and Irwin (1982) and Dow (1981) established similar qualitative criteria: Identify the chlorinated biphenyl homologs by their mass ion response, relative retention time, and ion intensity ratio (± 20% relative). Secondary confirmation of trichloro-through decachlorobiphenyl may rely upon the M-70+ ion response. These two works suggest a new awareness that qualitative criteria must be stipulated in any method if the results are to have any significance. In addition to the various gas chromatographic identification methods, thin-layer chromatography and high performance liquid chromatography yield qualitative information, as discussed above. More exotic techniques such as other mass spectrometric techniques, Fourier transform infrared spectrometry, and nuclear magnetic resonance spectroscopy are not used routinely in most laboratories, so have been included as confirmatory techniques which will be discussed below. Quantitative With most organic compounds, the quantitation is relatively straightforward. The instrumental response is calibrated using standards. The amount of unknown is measured by comparison of the signal it generates with the calibration factor or curve. Quantitation of PCBs is not nearly so simple because the analyte is not a single compound but rather a complex mixture of 209 possible congeners and standards of all 209 congeners are not available for calibra^ tion. Given these problems, analysts have devised alternate quantitation methods generally based on the similarity of the sample PCB mixture to a commercial product (e.g., Aroclor). Review Articles— Most review articles mention quantitation of PCBs briefly. Safe (1976) discussed the problems of BCD response variability discussed in detail above and suggested that perchlorination would be more consistently accurate. Hutzinger et al. (1974) reviewed the quantitation methods to that date, most of which involved relating the unknown to Aroclor standards. Sherma (1975) provided a similar, but more detailed, review which included a positive assessment of the Webb and McCall (1973) procedure. The other reviews (Riseborough, 1971; Fuller et al., 1976; Reynolds, 1971; Margeson, 1977) discussed quantitation in similar but less detailed fashion. Standard Methods— As shown in Table 1, all standard methods give at least cursory instructions on quantitation. At one extreme, the general purpose protocols (EPA, 1979a, 1979b; Ballinger, 1978; EPA, 1978) contain vague direction to "integrate the area under the peak." Much more complete quantitation guidelines are given in EPA's (Bellar and Lichtenberg, 1981; EPA, 1981) protocol for analysis of PCBs in transformer fluids and waste oils. Primary Literature-It is obvious from the data presented that PCBs were quantitated in most of the references abstracted. However, many articles neglect to mention how 51 �the PGC/ECD signal was converted into a concentration value. Some 80 articles abstracted mentioned the quantitation technique, although many only made a brief mention of "integration" or "comparison with Aroclor 1260 standard." Packed column gas liquid chromatography/electron capture detector--As noted in Table 3, the response factors for the individual PCB congeners vary widely, even within a homolog. This fact has typically been overlooked in quantitation procedures based on the use of Aroclor (or other commercial mixtures) standards. The most prominent GC/ECD quantitation method was originated by Webb and McCall (1973). The weight percent and homolog identification (relative proportions where more than one homolog was present) were determined for several Aroclors, and retention times relative to £,pJ-DDE were specified. The general procedure is as follows: Chromatograph known amounts of the standards and measure the area for each peak. Using the tables of data determine the response factor (ng PCB/cm2) for each peak. Chromatograph the sample and measure the area of each peak. Multiply the area of each peak by the response factor for that peak. Add the nanograms of PCB found in each peak to obtain the total nanograms of PCB present. Samples containing one Aroclor or more than one Aroclor can be quantitated by comparison with appropriate standards. Following an interlaboratory survey, Chau and Sampson (1975) recommended that the Webb and McCall method be adopted as the uniform quantitation method. They cited the general applicability, elimination of mixed standards, the more realistic results, and simplicity of the method as reasons for their recommendations. Exact replication of the method requires reproducing the chromatography and using the same lot of Aroclor standards as Webb and McCall used. These stringent requirements have led most researchers to characterize their quantitation practice as a modified Webb and McCall (Erickson et al., 1981; Kreiss et al., 1981; Harris and Mitchell, 1981; Steichen et al., 1980; Sawyer, 1978a). The calculations required can be easily automated using common GC integrators or data systems (Erickson et al., 1981; Kirshen, 1981). Ugawa et al. (1973) devised a quantitation method similar to a Webb-McCall except it was based on the Japanese commercial PCB, Kanechlor. Unfortunately, most samples are exposed to weathering, metabolism, differential adsorption, etc., and the PCB pattern, though originally one or more Aroclors, does not closely resemble that of the standard. This has been noted repeatedly and certain correction procedures have been proposed (Beizhold and Strout, 1973; Webb-McCall, 1973). Several researchers (Collins et al., 1972; Rote and Murphy, 1971; Bellar and Lichtenberg, 1975) and standard methods (AOAC, 1980; ASTM, 1980a, 1980b) advocate comparison of the total areas under the "Aroclor region" in the sample and standard chromatograms. This is a simple approach and has been recommended 52 �by Sawyer (1973) as the most reliable method for obtaining interlaboratory precision. In a later collaborative study (Sawyer, 1978b), the individual peak height (Webb and McCall, 1973; Sawyer, 1978a), total peak height and total peak area methods were all compared and gave similar results, although the individual peak method was judged slightly better. On this basis, AOAC (1980) permits either individual peak (Webb-McCall) or total area quantitation of PCBs. Bellar and Lichtenberg (1975) used either the total peak height for samples closely resembling Aroclors or Webb-McCall for patterns "not representing a single Aroclor." Wolff et al. (1982) used 2,4,4'-trichlorobiphenyl, 2,4,5,2',5'-pentachlorobiphenyl; and 2,4,5,2',4',5'-hexachlorobiphenyl as standards for CGC/ BCD quantitation of PCBs in plasma and adipose samples of occupationally exposed people. Quantitation by this sum of individual peaks method gave comparable results to the Webb-McCall method using P6C/ECD. Both methods gave lower values than the sum of peak areas method using PGC/ECD data. Zobel (1974) devised a computer fit routine which matched the sample chrpmatogram with various "co-added" Aroclor chromatograms to obtain a best fit. "Spuriously large or small peak heights, caused by interfering compounds or metabolism, are automatically sorted and rejected." The method reports results in terms of the different Aroclors and can be modified to generate an estimate of "premetabolism" PCB content. While providing no details on how the PCBs were quantitated, Giam et al. (1973) required at least 50% of the peaks in the sample chromatogram to match those in an Aroclor standard when analyzing marine biota for PCBs. Capillary column gas liquid chromatography/electron capture detection— The application of CGC to PCB determination complicates the already difficult quantitation problems: more peaks are present. Since the peaks are ostensibly single congeners instead of the mixture obtained by PGC, much emphasis has been placed on quantitation of single congeners. Boe and Egaas (1979) devised a calibration factor system that permits the analyst to calculate the ECD response factor for a given congener, once its structure is known. More recently, the response factors for 159 congeners were measured (Table 4). As discussed above, the ranges over homologs and within a homolog indicate that calibration with each congener would be necessary for accurate results. Despite the higher resolution with CGC and therefore more available information, simplistic quantitation routines are still used, Gordon et al. (1982) used three peaks from each Aroclor standard as their method for quantitating PCBs in transformer oil by CGC/ECD. While PCBs do not weather extensively in transformer oil and are thus more likely to resemble the parent Aroclor, this method utilizes only a small portion of the available information. Schulte et al. (1976) recommend quantitation of CGC/ECD chromatograms based on two selected charactistic peaks in food extracts. Albro et al. (1981) determined the relative molar percentages of the individual components of about 100 different PCB congeners in Aroclors 1248, 1254, and 1260. They recommend using the Aroclor mixtures as secondary standards to calibrate CGC/ECD responses. 53 �Gas liquid chromatography/electron impact ionization mass spectrom.etry"The ability of EIMS to easily sort PCBs by homolog has led to a tendency among GC/EIMS users to quantitate by homolog (e.g., summing all homolog peaks). Another major difference from analog (BCD) detector quantitation is the ability to quantitate using either a single PCB-specific m/z peak or the total ion current which corresponds to the analog detector output. The first reported GC/EIMS quantitation was a simple translation of classical PCB GC/ECD quantitation: comparison of "the area under one or more of the eight peaks to the area of a known amount of a standard" (Bonelli, 1972a,b). Eichelberger et al. (1974) chose what they termed the conventional approach and assumed that the PCB mixture was identified as one of the commercial mixtures. The total peak area for each SIM mass in the sample and standard were compared using an internal standard to normalize the peak areas. Erickson and Pellizzari (1977, 1979) quantitated PCBs in sludge based on a relative molar response (RMR) of each homolog. Using SIM techniques, the RMR for one isomer of each homolog was determined. Standards of hepta- through monochlorobiphenyl were too impure to use and RMRs for these homologs were interpolated. The RMR decreased dramatically (logarithmically) with increasing degree of chlorination, presumably due to the decreasing ionization crosssection. Williams and Benoit (1979) compared the summed total integrated area for six to eight selected peaks in samples and standard for quantitation of PCBs in several household products. Tindall and Wininger (1980), in one of the few papers addressing analysis of non-Aroclor PCBs, established homolog response ratios using an unspecified number of isomers per homolog. The highest and lowest response factors for a homolog were averaged to give the average response factor used in the calculation of PCB concentration in the unknown. An internal standard (tribromobiphenyl) was used to get relative responses. Martelli et al (1981) reported the EIMS relative response factors for 45 PCB congeners (Table 4). The relative standard deviation per homolog ranged from 0.7% (3 of 46 isomers) to 21% (11 of 42 isomers). They propose using these average response factors on GC/EIMS quantitation of PCBs by homolog. Collard and Irwin (1982) used an unspecified number of isomers per homolog to establish response factors for each homolog. A daily calibration plot at three concentrations was used for comparison of the summed peak heights for one homolog. Dow (1981), in a related protocol, specified 22 congeners to be used in a similar calculation. For homologs with more than one isomer in the standard a summed intensity was used. Miscellaneous--Cairns and Jacobsen (1977) advocated the quantitation of PGC/EIMS because of the reduction in interferences by other halogenated compounds such as DDE. Eichelberger et al. (1974) also mentioned PGC/EIMS as an alternate technique for dirty samples but did not discuss quantitation. 54 �As discussed in a separate section, perchlorination and dehydrochlorination, followed by PGC/ECD and PGC/FID determinations, respectively, have been proposed as PCB quantitation methods which eliminate the congener variability problems. TLC and HPLC have also been employed as quantitative techniques and are discussed in separate sections above. CONFIRMATION Confirmatory techniques have been frequently used in PCB analysis. The term confirmation may be loosely defined as any operation performed to increase the confidence of the results beyond the primary analysis. Qualitative confirmation is much more often reported than quantitative confirmation. Confirmatory techniques involve variation of the same technique (PGC/ECD on two dissimilar columns), confirmation by a lesser technique (PGC/ECD with TLC confirmation), or confirmation by a more advanced technique (PGC/ECD with PCG/ EIMS confirmation). Review Articles Hutzinger et al. (1974) devoted about two pages to the subject of confirmation. While mass spectrometry was briefly mentioned, most of the discussion centered on perchlorination. Standard Methods Table 1 listed all of the standard methods and notes the type of confirmation suggested. All of these confirmations are optional and qualitative. Primary Literature As early as 1969, the need for confirmation of findings was discussed (Reynolds, 1969). An exchange of comments following a presentation by Riseborough (1971) led to a proposal for confirmation by P. L. Diosady, covering, mass spectrometry, dechlorination, and perchlorination. Price and Welch (1972) are typical of many early investigators who backed up their PGC/ BCD analysis with a TLC confirmation (see also the standard methods: AOAC, 1980; FDA, 1977). Hannan et al. (1973) utilized a cumbersome ultraviolet irradiation method to confirm PGC/ECD results. Mes and coworkers have utilized a variety of confirmatory techniques, generally in concentration, in the analysis of adipose and milk samples (Mes et al., 1977; Mes and Davies, 1979; Mes et al., 1980). The methods include two dissimilar GC columns, perchlorination, and GC/EIMS. GC/EIMS was used by Biros et al. (1972) to confirm TLC results. GC/EIMS confirmation has also been reported (Musial et al., 1979; Teichman et al., 1978; Lucas et al., 1979; Rodriguez et al., 1980; Haile and Baladi, 1977; Erickson et al., 1981). HREIMS has been reported as a confirmatory technique (Safe et al., 1975; Safe, 1976; Musial et al., 1974). Kuehl et al. (1980) used CGC/NCIMS to qualitatively confirm their CGC/EIMS PCB identifications in 55 �fish. Mass and Friesen (1979), placing particular emphasis on polychlorinated dibenzodioxins, reviewed the advanced mass spectrometric techniques for both high sensitivity and high reliability analysis: HREIMS and NCIMS. SCREENING TECHNIQUES Screening techniques in this text are defined as methods used to identify the presence of PCBs qualitatively, semiquantitatively, or quantitatively without specification of the homologs in a sample extract. Screening techniques under this definition could include thin-layer chromatography, high performance liquid chromatography, and gas-liquid chromatography. These methods have been discussed in detail earlier in this review. Perchlorination and carbon-skeleton chromatography, however, are screening methods that have not been mentioned, although Table 1 indicates that perchlorination has been used as a quantitative method or PCB confirmation technique in several of the standard procedures. Perchlorination Perchlorination methods are based on the exhaustive chlorination of the biphenyl ring of the PCB congeners. The major disadvantage of the perchlorin1ation reactions is that biphenyl can also be perchlorinated. Thus, the presence of biphenyl can lead to erroneously high levels of quantitation. Quantitative analysis is typically accomplished by GC/ECD systems although GC/MS identification has been used in some instances. Perchlorination reactions are reportedly troublesome because of contamination of reagents with decachlorobiphenyl or brominated compounds (Trotter and Young, 1975). Perchlorination reaction methods were first studied using antimony pentachloride (Berg et al., 1972; Matsumoto, 1972; Armour, 1973) and thionyl chloride in the presence of aluminum chloride (Nose, 1972). Armour (1973) reported greater than 90% recovery of PCBs by perchlorination and found the technique comparable to PGC/ECD comparison with Aroclor standards. Nose (1972) reported approximately 100% conversion of tri-, tetra-, and hexachlorobiphenyls to decachlorobiphenyl with the thionyl chloride system. Antimony perchloride is apparently the most frequently used reagent as indicated by a review of the literature. Hutzinger et al. (1973) studied trichlorosulfur-tetrachloroaluminate to quantitatively convert Aroclor 1254 to decachlorobiphenyl. One of the major disadvantages of perchlorination arises from blank problems (Trotter and Young, 1975). This has resulted in the need to carefully characterize perchlorination reagents prior to reaction (Huckins et al., 1974). The other major disadvantage of perchlorination is the conversion of biphenyl to decachlorobiphenyl. Chlorine-37 labeled perchlorination reagents have been studied as a means to clarify this problem and at the same time distinguish the contribution of various PCB homologs to the final decachlorobiphenyl by computer assisted isotope dilution interpretation (Burkhard and Armstrong, 1982). This technique, although unique in approach, requires optimum reaction and MS conditions for successful analysis. Perchlorination has been used successfully for numerous studies in recent years (Kohli et al., 1979; Sherma, 1981; Stratton et al., 1978; Fulton et al., 1979; Crist and Moseman, 1977; Robbins and Willhite, 1979; Mes et al., 1977; Mes and Davies, 1979; Haile and Baladi, 1977; Vannuchi et al., 1976; Margeson, 1977; Brinkman et al., 1978; Kohli et al., 1979; Albro et al., 1980; Leoni et al., 1976; Trevisani, 1980), 56 �Carbon Skeleton Chromatography Carbon skeleton chromatography is based on the dechlorination of PCBs to biphenyl. Catalysts for the dechlorination are typically platinum or palladium. The disadvantage of carbon skeleton chromatography is that background levels of biphenyl in the sample extract will yield erroneously high concentrations of total PCBs as noted for perchlorination. Also, since the product of dechlorination is biphenyl, mass spectrometry must be used to reliably identify the compound, especially in extracts from complex matrices. Quantitative carbon skeleton chromatography by catalytic decomposition of the PCBs over platinum or palladium to biphenyl has been discussed in three articles (Berg et al., 1972; Zimmerli, 1974; Cooke et al., 1978). Zimmerli (1974) and Cooke et al. (1978) studied conversion of PCBs as well as halogenated terphenyls, napthalenes, dioxins, furans, and DDT. Effective catalysts were found to be effective as 3% palladium at 305°C and 5% platinum at 180°C. Reaction products for the various compounds were identified by GC/MS. On the other hand, false negative results have been observed using this technique for analysis of chlorinated bottoms (personal communication, M. D. Crouch, Toxicon Laboratories, Baton Rouge, Louisiana, 1982). QUALITY ASSURANCE A strong quality assurance (QA) program for PCB analysis should include use of pure standards, solvents, and glassware; an evaluation of method blanks for background PCB and interference levels; calibration of instrumental equipment; validation of the individual method steps as well as the overall method; and an evaluation of the overall performance of a method through replicates, interlaboratory comparisons, and/or standard reference materials. The data that should be provided by a strong QA program should include at a minimum, precision and accuracy measurements for each sample matrix. These parameters have been previously outlined by MacDougall et al. (1980) in an attempt to clarify needs for general data quality evaluation for comparison of trace organic results among numerous laboratories. The guidelines for data acquisition and data quality evaluation presented by MacDougall et al. (1980) were provided under the direction of the Americal Chemical Society Committee on Environmental Improvement and the Subcommittee on Environmental Analytical Chemistry. The guidelines were based on good analytical practices to assist analysts in obtaining data of requisite quality and to aid in the evaluation of the quality of the reported data. In addition to the QA parameters previously listed, MacDougall et al. (1980) have presented requirements for sampling to adequately characterize a sample and enhance reliability in the final result. These guidelines also discuss the necessity of detailed documentation of sample preparation and actual analysis such that other qualified analysts may duplicate the work. The demonstration of precision and accuracy of measurements through good laboratory practices, proven methodologies, low noise instrumentation, the use of standard reference materials, and participation in collaborative studies were also discussed as essential to strong QA programs. The guidelines also presented definitions of and criteria to establish limits of detection (LOD) and limits of quantitation (LOQ). Method validation, qualitative confirmation of validated measurements, risks in data interpretation from low recovery methods, reporting of interferences, and the 57 �appropriate presentation of the analytical results were discussed with respect to evaluation of data quality. Quality assurance in some form has been practiced in many of the studies abstracted for the PCB analysis literature review. However, few of the studies have implemented enough QA to allow comparison of the data from one matrix to the next. The EPA has taken steps to implement strong QA programs in various standard methods of analysis and as part of long-term project goals (EPA, 1979a, 1980a, 1980b, 1981; Bellar and Lichtenberg, 1981). Table 1 lists the standard methods that acknowledge the need to follow some type of QA program. Other than the standard methods and EPA guidelines, QA programs have been practiced for collaborative method studies for PCBs in different matrices (DCMA, 1981; Sawyer, 1973, 1978; Delfino and Easty, 1979; Devenish and Harling-Bowen, 1980). The QA program for the Dry Color Manufacturers Association (DCMA, 1981) round robin study included instrument calibration specifications, performance evaluation of the gas chromatography column with a standard mixture of PCBs, and measurement of sensitivity for PCBs by serial dilution of the standard, methods blanks, specification of quantitation procedures and validation of sample preparation procedure. The validation of sample extraction, cleanup and analysis included workup of blind and known spike samples. The results of the DCMA study indicate that variance in reported PCB levels between laboratories was signficantly reduced when a commercially prepared quantitation standard was used by all participating laboratories. Data from the DCMA report indicated relative standard deviations of 3.1 to 9.1% within a laboratory, 2.4 to 40% between laboratories, and values ranging from 7.3 to 41% representing the total reproducibility for analysis of three different pigments. The Chemical Manufacturers Association sponsored a round-robin study of PCB concentrations in five different samples that are indicative of matrices that will be regulated by the PCB Remand Rule (Pittaway and Horner, 1982). Eight different laboratories participated in the study. In contrast to the DCMA study, no defined protocol or QA programs were specified for analysis of the matrices. Each laboratory was allowed to choose the method of extraction, cleanup, instrumental determination and quantitation, and QA program if desired. This study indicated that there are many sources of potential error in the quantitative analysis of PCBs. A comparison of the reported levels of PCBs and precision of measurements between laboratories indicated a true need for a strong QA program that might allow some normalization of the data, The collaborative study reported by Delfino and Easty (1979) focused on the analysis of PCBs in paper mill effluents. The study consisted of two phases. The first phase was used to determine the comparability of PCB methodologies between six different laboratories and the abilities of the participating analysts to perform the basic operations required for PCB analysis. These factors were determined by direct injection and quantitation of a performance standard and the simple extraction and analysis of a spiked aqueous solution. The second phase required an actual validation of a sample method using known and blind samples. A modified EPA wastewater analysis protocol 58 �was followed by all participating laboratories. Some flexibility to the method protocol was allowed for column materials and exact quantitation procedures. The results of the first phase extraction from distilled water yielded an average recovery of 95.6% with a relative standard deviation of 14,7%. The relative standard deviation for direct injection of a standard solution was 15.6%, thus indicating that gas chromatographic analysis was the principle source of variance. The results for paper mill effluent yielded similar results with 93.6% average recovery with a 16.0% relative standard deviation, and indicated that the method was satisfactory for paper mill effluents . Sawyer (1978) conducted a collaborative study of PCB quantitation with BCD as the detector. Ten independent laboratories took part in the study and used existing AOAC methodology to study three ECD quantitation procedures. The average combined recovery in this study was approximatley 85% with a coefficient of variation of 15%. No significant difference was noted for the three different quantitation operations. A large void in most QA programs has been filled by the provision of standard reference materials of known PCB concentration (Chesler et al., 1982). Although the standard reference material will only be available as an oil, it leads the way in establishing further QA criteria for the analysis of PCBs in other media. The preparation of additional PCB standard reference materials as wet and dry reference materials has been discussed by Chau and Lee (1980), Chau et al. (1979), and Addison and Nearing (1982), although these materials are not currently available. Other policies that are of considerable concern and reflect the current attitudes toward QA have been presented by Glaser et al. (1981) concerning method detection limits based on confidence levels and the guidelines presented by Environmental Science and Technology (Christman, 1982) outlining information required to label compounds identified by mass spectrometry as "tentative" or "confirmed." BY-PRODUCT PCB ANALYSES Historically, analysis of PCBs has been concerned with commercial mixtures , such as Aroclors, and their dispersal in the environment and certain commercial products (packaging materials, paper products, transformer oils, etc.). The analytical approaches to identification and quantitation of these commercial PCB mixtures has already been discussed in this literature review. Health and environmental concerns have resulted in an increasing number of federal regulations (EPA, 1979d) controlling the manufacture, use, and disposal of PCBs. A recent report prepared by the Chemical Manufacturers Association for EPA (Pittaway et al., 1981) documents numerous commercial processes that will be affected by proposed federal regulation of by-product PCBs, These by-product PCBs are produced by diverse processes, few of which resemble the commercial PCB synthesis routes. Thus, the resultant PCB mixtures do not resemble the familiar Aroclors. The analysis of by-product PCBs was reviewed by Hodges et al. (1982). 59 �An extremely limited number of articles are available for review of byproduct analysis (Tindall and Wininger, 1980; Collard and Irwin, 1982; Pittaway and Horner, 1982; DCMA, 1981; Dow, 1981). These few references, however, describe some of the problems encountered in by-product analysis of commercial products as discussed below. Dry Color Manufacturers Association Pigment Analysis A major study was conducted by the Dry Color Manufacturers Association (1981) for the analysis of by-product PCBs in three different pigments. This study concluded that a universal cleanup procedure was not possible for accurate PCB analysis from all of the pigments. The use of GC/MS was recommended for establishing positive identification of the PCBs. Two of the pigments studied contained only one PCB isomer, while the third pigment was contaminated with several different isomers of the penta- and hexachlorobiphenyl homologs. A thorough quality assurance program was developed for the purpose of reducing interlaboratory variability. The quality assurance program included validation of each step of sample preparation, gas chromatographic performance standards, and mass spectrometer calibration and performance appraisal, as well as requirements for analysis of spiked blanks, replicates, and standard additions. Round-robin experiments were conducted under this study. The DCMA found that a large portion of the interlaboratory variance was due to the differences in preparation of the standard mixtures of PCB isomers used for establishing response factors. A calibration mixture obtained from a single source was found to greatly reduce interlaboratory variance. In addition, specification of PGC criteria with respect to retention times and resolution of specific isomers was required to promote comparability of results between laboratories. Chemical Manufacturers Association Round-Robin The Chemical Manufacturers Association has conducted a round-robin experiment for analysis of by-product PCBs (Pittaway and Horner, 1982) in chlorinated benzene waste streams, mixtures of chlorinated benzenes, blind spikes in the chlorinated benzenes, composite waste streams from a chlorinated aliphatic process, and a benzene column bottom sample. The round-robin studies defined some of the problems of by-product PCB analysis in commercial products and process waste streams. Many sources of potential error in the quantitative analysis of these samples were identified and include unknown interferences, inappropriate use of reference standards, inappropriate protocols, day-to-day variations in instrumental responses, calibration, and execution of analytical procedures, inappropriate collection of samples, contamination of samples, and limitations in instrumental methods. The round-robin study measured differences in analytical results between laboratories, differences due to variation in analytical methods, limitations of instrumental methods, and impact of analysis by random laboratories 60 �(Pittaway and Homer, 1982; Hodges et al., 1982). A total of eight laboratories (six industrial and two EPA) participated in the study, using a variety of techniques as shown in Tables 5 and 6. Guidelines were not given for methods of sample preparation, instrumental analysis, quantitation measurements or quality assurance practices. The results from the round-robin study showed a significant variance in data among laboratories, which one might expect from the lack of written protocol and quality assurance. This study demonstrated that there is a need for a common denominator in analytical protocol for analysis of by-product PCBs from a wide variety of simple to complex matrices. Other Studies Tindall and Wininger (1980), Collard and Irwin (1982), and Dow (1981) studied PGC/MS methods for analysis of by-product PCBs in commercial and environmental samples. The MS analysis method was based on limited mass scan ranges to qualitatively identify and quantitate any of the possible 209 PCB congeners by homologs. Tindall and Wininger (1980) reported that the criteria for PCB quantitation must include matching of characteristic ions at proper retention time windows. In addition, characteristic ions must maximize at the same retention time and the relative peak intensities must be within ± 15% of the theoretical ratio. The accuracy limiting step of the PGC/MS (limited mass scan range) methods (Tindall and Wininger, 1980; Dow, 1981; Collard and Irwin, 1982) is the selection of standards. In each case response factors were determined for a limited number of isomers for each PCB homolog with the underlying assumption that all PCBs of the same homolog have nearly the same response factor, Quantitation procedures varied between the studies. Tindall and Wininger (1980) used an internal standard, tribromobiphenyl, which responded to MS source changes much like a PCB. In addition, its molecular weight was great enough that interferences were rarely encountered. Collard and Irwin (1982) and the Dow method (1981) quantitated versus a calibration curve established at various concentrations using 10 congeners to represent each PCB homolog. No internal standards were used. The accuracy and precision of these PGC/MS methods are dependent on frequency of instrumental calibration and the extent that other compounds in the ion source of the MS affect the sensitivity during the course of an analysis. 61 �SECTION A APPLICABLE TECHNIQUES The objective of this section is to outline the best possible approaches for by-product PCB analysis in commercial products that will be regulated under the PCB Remand Rule. The proposed procedures are a result of the review of the available literature presented in the previous section. The analytical approaches provide versatility in terms of the wide spectrum of matrices represented by the proposed regulated products. The success of the proposed rule will rely heavily on a strong quality assurance (QA) program to monitor sample preparations and instrumental analysis. The proposed QA program will provide sufficient data to determine the quality of the quantitation data for each specific sample matrix encountered. Sample extraction, cleanup, instrumental determination, quantitation and data reduction, confirmation, screening, and the overall quality assurance program are discussed. EXTRACTION The literature review of extraction techniques describes several approaches to isolation of PCBs from various matrices. The extraction may be as simple as dilution of an organic liquid, batch extraction of aqueous solutions , or Soxhlet extraction of solids; or as complex as matrix destruction via saponification or with concentrated acid before extraction of the PCBs with an appropriate organic solvent. However, the analyst must keep in mind that totally unexpected reactions produced the trace levels of the by-product PCBs being determined. Hence, the use of vigorous or harsh chemical reactions may generate or destroy PCBs (L. F. Hanneman, Dow Corning Corporation, personal communication, 1982). Suitable organic solvents will include petroleum ether, hexane, and methylene chloride. The exact extraction procedure, however, is dependent on the specific matrix. The alternative to designating a specific extraction procedure for all solid and liquid samples that are of highly dissimilar matrices both chemically and physically is to formulate a rigid QA protocol before the extraction step and to continue it through all aspects of analysis. The QA protocol will require extensive homogeniaation of samples (solids, suspensions, liquids) by grinding and mixing. An aliquot of each homogenized sample will be spiked with a series of surrogate compounds. Final analysis of the sample extract for the surrogate compound recoveries will provide sufficient quantitative information to evaluate the effectiveness of the extraction procedure and or cleanup technique. The choice of surrogate compounds is critical for exact performance measures of any method. The surrogate compounds must maintain the exact chemical characteristics of the PCBs for extraction, cleanup, and quantitation purposes. The surrogates may be either a series of PCB congeners representing 62 �each of the chlorinated homologs or a selected number of stable massrlabeled (carbon-13 or chlorine-37) PCB isomers. The series of unlabeled PCB congeners surrogates would necessitate independent measures of spike recovery and analyte concentration, whereas the mass-labeled PCB surrogates would allow simultaneous determination and quantitation of the analyte PCBs and the surrogate compounds if EIMS is used as the GC detection. The implementation of the use of the mass-labeled surrogates would provide a strong QA program since recovery could be monitored for each sample analyzed and consistency between analytical laboratories and different sample matrices could be compared more readily by a regulatory agency. Mass-labeled PCB isomers as surrogates could be provided at a reasonable cost per sample analyzed (W. Duncan, Midwest Research Institute, personal communication, 1982). A series consisting of mono-, tetra-, octa-, and decachlorobiphenyl mass-labeled isomers would provide a sufficient set of surrogates. Carbon-13 labeled isomers of 99% purity for these homologs would provide sufficient differences in mass spectra patterns for differentiation from isomers of natural abundance. The major problem to consider in adding surrogate compounds is whether the incorporation of these compounds in a matrix will mimic the true analytes. Incorporation cannot always be achieved, especially with matrices that require exhaustive extraction methods. However, the measured recovery of surrogates from an extraction and cleanup procedure will provide information on degradation of PCBs by the analytical procedure. It may be desirable to design small scale experiments to incorporate the surrogate PCBs during a product process (L. F. Hanneman, Dow Corning, personal communication, 1982). Solid matrices, expecially those that may be intractable, should be of prime interest for this approach. The surrogate compounds could be incorporated before polymerization, vulcanization, curing, precipitation, or other processes. The recoveries of the surrogates could be used to determine if the proposed extraction technique is applicable for routine analysis of particular solid matrices. Exact extraction protocols could be designated for air and simple aqueous samples as shown in Table 1. Exact extraction protocols for commercial products could also be designated, but optimum performance for all matrices is highly unlikely. A specified extraction protocol would require rigorous methods for all samples and must consider possible adverse reactions of certain products to sulfuric acid digestions and alkaline saponification. Independent extraction procedures for different matrices combined with the use of surrogate compounds and thus validation of the method would be an effective alternative. Each independent laboratory, however^, must certify that an effective extraction procedure is practiced. The level of recovery considered sufficient and method of addition of internal standards for final quantitation are yet to be determined. 63 �CLEANUP Many cleanup techniques are applicable to commercial products. The nature of the sample, complexity of the matrix, and the chemical characteristics of other components dictate the requirements for any sample preparation. Cleanup for air and aqueous samples in effluents from commercial production facilities is achievable by applying standard methods (Table 1). However, cleanup of a wide range of product matrices will require application of many different techniques. A generic cleanup procedure may not suffice or be necessary in the majority of analyses because of the different chemical characteristics in the sample matrix. For example, sulfuric acid may provide sufficient cleanup and quantitative recovery of one product that contains only decachlorobiphenyl. However, this procedure will not be sufficient for a matrix that contains mono- through trichlorobiphenyl isomers, which may not be recovered quantitatively, likewise, designated adsorbent columns may not provide the separation of interferences necessary for good quantitative analysis for a large majority of matrices. As with the extraction step, one alternative is to allow the individual laboratory to develop the necessary cleanup procedure. Each laboratory, however, must follow the stringent QA program using spiked samples or surrogate compounds to validate the method at a determined level of proficiency. This will meet the special analytical needs of the individual analyst and at the same time provide the data necessary to determine analytical proficiency of the method and consistency between laboratories and matrices. DETERMINATION Gas-liquid chromatography is judged to be the only acceptable primary separation method. Capillary GC is preferred over packed GC. The injection system, type of liquid phase, column dimensions and operating conditions should not be specified, but performance should be maintained within established criteria. Electron impact mass spectrometry is the primary Operating conditions (SIM, full-scan, or limited mass teria, and other variables are still to be specified. BCD and HECD are considered too nonspecific for these for general application (NCIMS, MS/MS, FTIR). detection candidate. scan), performance criOther detection options, matrices or too uncommon Separation As clearly evidenced in the review of the literature, GC is by far the most popular technique for PCB determination. The relative merits of PGC and CGC are well-known and apply to the separation of PCBs. CGC provides better resolution, retention time precision, and higher qualitative reliability. PGC yields a simple chromatogram (less data reduction), permits higher sample loading (and therefore possibly lower LOQs), and is generally considered easier to use. Historically, PGC quantitation has been more precise, although it has not been established how much of the imprecision attributed to CGC was due to poor technique on the part of the analyst. �The high resolution of CGC is not required in this application for separation and identification of the individual PCB congeners since the final result needed is total PCB. However, the high resolution of CGC should aid the analyst in separating PCBs from interferences. In this respect, CGC is preferable. In many cases when the sample is amenable to "dilute and shoot" techniques, very high levels of matrix materials may be present, which will overload the column. Although CGC is more sensitive to column overloading, both CGC and PGC will overload with percent levels of matrix materials. The advantages of CGC, therefore, make it the technique of choice. However, PGC should also be allowed. Nearly every GC phase has been reported in PCB analysis. The most satisfactory separations have been achieved on nonpolar and semipolar phases (Apiezon L, methyl silicone, Dexsil, etc.). Since enforcement of very specific column parameters is difficult and since rigorous stipulations do not appear warranted, it is recommended that any nonpolar and semipolar capillary column be permitted. The choice of CGC injector (split, splitless or "Grob," and on-column) can substantially affect the amount of material transmitted through the system. Since enforcement of use of a particular injection would be difficult and since no clear choice is presented, any injector which meets performance criteria should be permitted. As part of the quality assurance program, a set of GC performance criteria should be established. The criteria should include number of effective plates, separation number (Tz), resolution, and peak asymmetry. Since PCBs are neutral, the acid/base characteristics of a column are of little interest; however, some measure of compound transmission through the system must be used. This may be achieved by monitoring the overall system response. Other options would entail additional work and are thus less favored. Detection The electron capture detector is the most sensitive candidate detector. However, it suffers from large differences in response factors for PCB congeners, which would result in very poor precision (Tables 3 and 4). Even more serious is its lack of specificity. Since many of the sample matrices will contain large amounts of halogenated nonPCB compounds, BCD would be overloaded throughout much of the chromatogram, making PCB identification and quantitation impossible. In certain cases, ECD may be a satisfactory detector, especially as a semiquantitative technique for screening samples prior to CGC/EIMS analysis. Electron impact mass spectrometry (EIMS) appears to be the method of choice. Mass spectrometry has sufficient selectivity that a chlorinated organic matrix will not generally interfere with PCB determination. While the El mode is not the most sensitive (NCI is much more sensitive), it is the most common ionization technique for GC/MS and, most importantly, is the most reproducible quantitatively. The precision of the EIMS depends upon the precision of the response factors. The most thorough evaluation of PCB response factors 65 �(Martelli et al., 1981) found up to + 20% RSD in selected isomers of one homolog. Since only 45 of the 209 congeners were characterized, the magnitude of the variation in the remaining response factors is not known. DATA REDUCTION Qualitative Since most matrices subject to regulatory analysis contain substantial amounts of halogenated compounds in addition to PCBs and since an "Aroclor pattern" will not usually be present, identification of PCBs is very important. Any misidentification of a nonPCB as a PCB will yield an erroneously high value. From the EPA's viewpoint, this poses no regulatory problem. However, it may result in needless effort and cost to the regulated manufacturers. In most cases, the EIMS data should provide sufficient confidence in the identification of PCBs for action. For those laboratories which choose to employ an equivalent technique for routine analysis, any samples with PCB values near the regulatory cutoff ("near" has not been defined) would have to be reanalyzed by the primary technique before a regulatory decision could be made. In cases where there are some doubts as to the identity of a peak as PCB by CGC/EIMS, any available confirmatory technique should be allowed, provided that the LOQ is equivalent to or lower than the CGC/EIMS LOQ. Positive confirmations present no regulatory problems to EPA. Any confirmations which show that a peak is not PCB must be well-documented with appropriate QA (for example, a spectrum of a PCB standard spiked into the matrix). In many cases, instrument responses are highly dependent on the matrix, so the response to standards in clean solvent cannot be equated with the response in the sample. In cases where a peak is shown to contain both a PCB and an interference, regulations should state that the entire peak must be quantitated as PCB unless the level of the interference can be precisely shown. The qualitative criteria for the EIMS data should address the following points: 1. Full spectra a. Number of ions which must be present. b. Background subtraction techniques permitted. c. Ion intensity tolerances. d. Relative retention time windows. 2. SIM a. Number and mass of ions to be monitored per homolog. b. Tolerance of ion intensity ratios. 66 �c. Relative retention time windows. d. Signal-to-noise ratio. 3. LMS a. Mass range to be scanned. b. Tolerance of ion intensity ratios. c. Relative retention time windows. d. Signal-to-noise ratios. Quantitative Assuming that GC/EIMS is to be used, digitized data will be obtained for use in quantitation. The areas or intensities of individual mass ions must be measured, compared with those for a standard, and converted to a concentration value. This process is complicated in PCB analysis by several factors: 1. Previous schemes based on Aroclor standards are not applicable to incidentally generated PCBs. 2. PCBs are a complex mixture, so the problem really involves up to 209 separate quantitations. 3. All 209 congeners are not available as standards. 4. Two or more congeners may co-elute. The ions for quantitation, selection of standards, and calculation procedures are discussed below. Quantitation Ions— The highest signal-to-noise ratio and therefore precision is a compromise between absolute ion intensity and background signal. Most analysts have chosen an ion in the molecular cluster for quantitation. Even though other ions may be more intense, the molecular cluster is at the highest mass and the background is generally lower. The use of the most intense ion in the molecular cluster is recommended for quantitation, with options of less intense ions from that cluster if interferences are encountered for the primary ion. Calibration— Four calibration methods are available: external standard, internal standard response factors, internal standard multi-point calibration, and direct use of the surrogates. External standard calibration—Calibration of the analysis system versus an external standard and then quantitation using the absolute intensities or areas of the peaks lacks precision (Haefelfinger, 1981) and is not recommended for gas chromatographic analysis. Often, the greatest source of imprecision is the reproducibility of injection volume. 67 �Internal standard response factors — In this method, internal standard(s) are added to the sample extract immediately prior to the instrumental deter" mination, and the analytes are quantitated using the ratio of the peak height or area of the analyte and internal standard. A previously determined response factor (essentially a two-point calibration curve, with an assumed intercept at the origin) is used in converting the response ratio to mass. For PCBs, which can span a large chromatographic range, three or four internal standards are recommended since the precision of the response factors, and therefore the final quantitation, is related to how close the analyte peak and internal standard elute (Haefelfinger, 1981; Bickford et al., 1980). Other factors to be considered in the selection of internal standards are chromatographic resolution from analytes and interferences, different mass spectral properties to assure identification in GC/MS analysis, chemical similarity to analytes (to minimize effects of changes in system selectivity), very low probability of occurring in samples, and chemical stability. Candidates for internal standards in PCB analysis include other halobiphenyls (e.g., fluorononachlorobiphenyl or dibromobiphenyl), related haloaromatics (e.g., chloronaphthalenes), and isotopically labeled PCBs (e.g., d6-3,4,3',4'-tetrachlorobiphenyl). Given the complexity of the matrices in which by-product PCBs may need to be determined, related chloroaromatics should not be used and halobi-^ phenyls must be selected judiciously. Since this technique is in essence a one-point calibration, the response factors must be determined at a concentration close to that of the analyte. Differences of more than one order of magnitude may induce significant errorRecovery surrogates added prior to any sample treatment may also be quantitated against the internal standard and, since the amount added is known, their recovery can be calculated. Knowledge of the percent recovery is useful in monitoring extraction/cleanup performance. The final number reported may be corrected for recovery if desired (or required), or the value found and percent recovery reported separately. It should be noted that if the recovery surrogates and analytes are not, in fact, equally recovered, then the reported recovery is meaningless. This will happen if the surrogates are not fully incorporated into the matrix. Internal standard multi-point calibration—This technique is essentially the same as the response factor technique, above, except multiple calibration points (typically three, spanning up to two orders of magnitude) are used to establish a calibration curve. This has the potential for greater precision, but requires much more time and several solutions. In cases where detector sensitivity (signal response versus amount or concentration) is either nonlinear or the curve does not intercept close to the origin, a multi-point curve is advisable. Surrogate calibration—Surrogates may be used as internal standards for quantitation. Using previously determined response factors (or calibration curves), the response ratio of the analyte and surrogate, and known masses and volumes, the mass of the PCBs may be calculated. When the surrogate recovery is less than 100%, this method automatically corrects for this loss and provides a built-in recovery correction. This method has the advantage of simpler 68 �calculation and requires fewer solutions. This technique is often referred to as isotope dilution. As with the internal standard technique, the surrogates must be incorporated into the matrix to assure equivalent recovery between the surrogate and analyte. Selection of Compounds for Calibration— Clearly, most of the reported quantitation methods which rely on relation of the sample peaks to those in an Aroclor standard are not applicable to by-product PCB determination. The options for compounds to be used in calibration are: 1. Establish and use relative responses for all 209 congeners. 2. Establish and use relative responses for all available (about 80) congeners and extrapolate the responses for the other isomers. 3. Establish and use relative responses for several congeners and extrapolate the responses for the other congeners. 4. Characterize a secondary standard using all available congeners. The secondary standard would be prepared from commercial mixtures to span the range of congeners. Option 1 is the ultimate technique. However, all 209 congeners are not available and synthesis/acquisition would be extremely expensive and require well more than a year for completion. Thus, even if this approach is to be pursued, an interim approach must be specified. Options 2 and 3 are compromises. Option 2 could be termed the best available method. Option 3 is easier to implement and utilizes a reasonable number of quantitation congeners. Option 4 would generate a well-characterized mixed Aroclor or similar mixture. This has the advantages of low cost (oace the characterization has been completed) and uniformity. The disadvantages are that (a) the concentrations of the congeners range over two or more orders of magnitude, so calibration of the instrument would be difficult and (b) many users will have only a limited range of PCB homologs (e.g., only dichlorobiphenyls) and would not want to use a standard requiring a full GC temperature program. Options 2 and 3 appear to be the most applicable. Whether Option 2 or 3 is more appropriate depends on the variability of the response factors and the precision desired by EPA. This area needs to be further investigated. Assuming that all 209 congeners are not characterized and used as calibration standards, analysts will have to extrapolate response factors from the standards used to other isomers. Guidelines for these extrapolations must be specified. A preliminary investigation by at least one laboratory to define the response factor variability among all available congeners is recommended. An estimate of the error associated with the extrapolation would be available from statistical evaluation of the resulting data. 69 �LIMIT OF QUANTITATION Extrapolation of the data in Table 2 to a proposed LOQ for the PCB Remand Rule is difficult because of several uacharacterized variables. The levels of interferences will vary widely. In addition, recoveries will vary, also affecting the LOQ. A major variable is the concentration factor in the workup (number of grams of sample concentrated or diluted to a given sample volume for injection on the GC). If the method works with a 1-g sample, an order of magnitude decrease in LOQ can be effected by using a 10-g sample. This requires additional effort and therefore cost. Taken to the extreme, it is possible to use very large samples (many kilograms) to lower the LOQ. On the other hand, the customary volume to concentrate a sample extract is 1 ml. One can, with some difficulty and loss of volume precision, further concentrate the extract to 100 pi, effecting a 10-fold decrease in LOQ. However, if the sample is still dirty, this further concentration can lead to solidification, which prevents injection onto a GC column. DATA REPORTING The final data report should list total PCB as the bottom line number, since this is the value of regulatory interest. The analytical reporting format should be specified to eliminate any ambiguity and should address the following issues: 1. Tabulation of individual congener quantitation. This may be difficult to achieve since congener identifications may not be available for more than a few congeners. 2. Tabulation of individual homolog quantitation. quired so that data reviewers can assess the data. This should be re- 3. Reporting units. These must be specified. Units such as micrograms per gram (solids), micrograms per liter (water), and micrograms per cubic meter (air) are recommended. 4. Recovery correction. If the final protocol specifies use of surrogates to monitor recoveries and if the method is validated, then recovery correction would be appropriate. While this conflicts with the customary practice in pesticide residue, priority pollutant, and other analyses, it is con~ sistent with the practice in many other fields (e.g., clinical analyses). Since near quantitative (> 70%) recovery cannot be assumed, recovery correction is strongly recommended. Although high recoveries may not be guaranteed, very low recoveries cannot be tolerated. The lower the recovery, the higher the overall method LOQ. In addition, very low recoveries are generally accompanied by poor precision. If, for instance, 10 ± 5% recovery is achieved, the uncertainty in the nominal 10X correction factor is huge (6.7-20X). Thus, a lower acceptable limit of recovery (e.g., 30%) must be stipulated for acceptable data. 5. Standard reporting format. A standard reporting form, including all equations to be used, intermediate calculations, etc., should be required. This improves quality assurance since the chance for error (e.g., using the 70 �wrong conversion factor) is reduced, and any errors would be detected more easily. In addition, data presented to regulatory personnel on standard forms are much easier to review. 6. Internal standards. If no recovery surrogates are available, internal standards would most certainly be advocated. Their use generally improves precision, and thus data quality, over the use of external standards. If surrogates are used, the need for internal standards is not so clear. One can use the surrogate response in the calculation of unknown concentrations. In this case, instrument response variability, losses in the instrument, and workup recovery are taken into account, all in one calculation. Thus, recoveries are corrected without any real knowledge of the percent recovery. On the other hand, the use of both internal standards and recovery surrogates would yield a fairly accurate assessment of the recovery and permit the analyst (or regulator) to decide if the recovery is unacceptably low. While knowledge of recoveries is not of central regulatory concern, it does provide the analyst with an estimate of the method performance. Therefore, if no internal standards are used, a seraiquantitative estimate of recovery must be made using the CGC/EIMS response (area counts or similar measure) to the surrogates. If this is below a certain threshold (say 30% of expected), then the sample preparation must be repeated or changed. The use of the surrogate compounds as standards for both workup and instrumental analysis is simple-, and since it is one step instead of two, it should be more precise. The price paid for this simplicity is that recoveries are not well characterized. In the interest of simplicity and better precision, the use of internal standards in addition to recovery surrogates is not recommended. CONFIRMATION TECHNIQUES Qualitative Alternate columns, detectors (HREIMS, FTIR, NCIMS, HECD, etc.), and techniques such as MS/MS, direct probe HREIMS, NMR, FTIR, HPLC, etc., will be permitted for confirmation of PCBs. Proper validation and demonstration of comparable or lower detection limit must be provided with any confirmation which overrules the GC/EIMS identification and eliminates the compound from quantitation as a PCB. Quantitative As part of the overall QA, quantitative confirmation is required. The options proposed include duplicate analyses and standard addition. Acceptance criteria for these confirmatory techniques are discussed in more detail in the Quality Assurance subsection. SCREENING/EQUIVALENT METHODS Alternate procedures to the designated protocol may be necessary to obtain rapid estimates of PCB concentration for commercial facilities operating on a continuous process basis or for small businesses relying on contract laboratories for analyses. The alternate procedures could possibly include perchlorination, dechlorination, TLC, PGC/ECD, PGC/HECD, CGC/ECD, CGC/HECD, etc. 71 �The data generated by these methods would be for the individual industry's use to determine if changes in process design or initial reactants are necessary to lower the levels of PCBs in the final product. However, compliance with the regulations must still be determined with the designated protocol unless EPA accepts the screening technique as equivalent to the protocol. Equivalency must be demonstrated in terms of sensitivity and selectivity for PCBs, limits of detection and quantitation, and interferences. A strong QA program must be implemented to establish and monitor the equivalency of an alternate method. The quality control program should include measurements of blanks, spiked blanks, and spiked samples (blind and known) to establish limits for precision, accuracy, and recovery of analyses from the sample matrix. Equivalent methods would be most applicable to continuous process operations with little system variance. Gross changes in any parameter of a continuous operation process should require further verification of equivalency of the alternate method. The levels of PCBs in a fraction of all samples should still be analyzed according to the proposed primary protocol for quality assurance. 72 �SECTION 5 POSSIBLE ANALYTICAL SCHEMES The purpose of this section is to discuss how all the analytical com" ponents presented in Section 3 can be integrated to produce an effective overall protocol. For discussion purposes, it is presumed here that a primary protocol will be established and that it will contain the following steps: 1. Homogenize sample and subsample if necessary. 2. Incorporate surrogate compounds (e.g., four 13C PCB congeners). 3. Dilute, extract, or clean up as required. 4. Concentrate or dilute to a known volume. 5. Analyze a known aliquot by CGC/EIMS. 6. Identify PCBs by relative retention time and mass spectral characteristics. 7. Integrate the PCBs by homolog and calculate amounts of each homolog by normalizing the responses to responses for the surrogate compounds, using one or more homolog response factors. 8. Sum all 10 homolog concentrations to obtain a total PCB value. 9. Report on standard reporting form. 10. Follow specified routine quality assurance (blanks, controls, duplicates, standard addition, instrument performance criteria, etc.). 11. Maintain appropriate records. ISSUES TO BE ADDRESSED If this or a similar protocol is specified, several issues must be addressed. Method Flexibility Some flexibility must be permitted in the method details (GC columns, solvent evaporation techniques, etc.) to accommodate different apparatus and 73 �laboratory practices. However, excessive flexibility will adversely affect the data quality since many operations are uncontrolled. With proper QA practices, the method can be flexible while still generating acceptable results. Thus, it appears that the best approach is to provide options and suggest rather than require for most method details. As long as the laboratory demonstrates it is within the performance boundaries specified by the QA guidelines, the optional approaches should be allowed. Substitute Methods Except when validated and routinely confirmed by CGC/EIMS, no substitute methods should be permitted. Equivalent Methods Equivalent methods should be permitted. Equivalent methods (TLC, GC/ECD, etc.) are defined as methods which have been validated against the primary method and yield comparable quantitative results and have LOQs comparable to the primary method or lower LOQ than the regulatory cutoff. Results obtained by an equivalent method must be confirmed by the primary method if significant interferences are suspected or the levels found are near the regulatory cutoff ("near" must be defined). Any use of an equivalent method would be subject to the additional QA provision that a specified number (e.g., every tenth) of samples be routinely run by the primary method and that the two results agree within specified tolerances (the agreement should be specified by homolog, not simply by total PCB, to avoid any method bias toward one end of the homolog lines). Since an equivalent method would be subject to validation and additional QA, it would be applicable only to routine monitoring of a process. Obviously any single batches or one-shot analyses would have to be done by the primary method. The major application of equivalent methods is projected for the company with a process at several plants which must be monitored periodically. Since most plants do not have GC/MS instrumentation and a slow turnaround from central research would either delay product shipping or permit untested product to be released to customers, use of an equivalent method appears to be the only acceptable alternative. QUALITY ASSURANCE The QA options to be addressed include: 1. Round robins: One or more round robins appear to be a good mechanism for improving the methodology and predicting the data quality. The objectives and execution of the round robin need to be addressed. Whether periodic round robins should be required is also of interest. 2. QA organization: Some organization must be designated to administer QA. The responsibilities and authority of the organization need to be specified. At one extreme, the QA organization periodically reviews the data submitted. On the other extreme, the QA organization would have laboratory 74 �facilities and confirm results on selected samples, prepare and send out performance audit samples, organize and execute round robins, conduct systems audits, and conduct method development efforts when necessary. Clearly the data quality and cost are roughly proportional to the amount of QA. A sensible compromise must be reached. 3. Systems audits: One standard QA practice is the systems audit. This is especially valuable in that the QA officer observes the personnel and facilities in operation and assesses their competence and performance. This is the only way the QA office can monitor the laboratory practices and review the raw data (chromatograms, mass spectra, magnetic tapes, etc.). The use of systems audits is desirable, but it requires personnel and travel fund commitments . 4. Performance audits: A performance audit is a quantitative analysis with a material of known PCB content. Performance audits consist of blanks and samples, blind or known, submitted by the QA lab and are generally analyzed along with routine samples. A performance audit system is mandatory as part of the overall laboratory certification program. 5. Laboratory certification; A laboratory certification program is recommended. The quality of the data and therefore the laboratory capabilities and performance must be assured. There are several methods for laboratory certification: round robin participation, performance audit participation, or submission to a systems audit. Most likely a combination of the three would be the most reasonable certification route. Following initial certification, all participating laboratories must be periodically recertified. Performance audits and systems audits are the most appropriate recertification methods. 75 �APPENDIX A PERSONAL CONTACTS Note; In preparation of this document, telephone, written, or personal discussions were held with the individuals listed. 76 �Ahmed, Karim. Natural Resources Defense Council, 122 E. 42nd Street, New York, New York 10169. Alford-Stevens, Ann. Environmental Monitoring and Support laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268 Bell, Robert A. Corporate Research and Development, General Electric Company, 1 River Road, K-l, 3B15, Schenectady, New York 12301, (518) 385-8505. Albro, Phillip. NIEHS, Research Triangle Park, North Carolina 27711, FTS 629-3264. Bellar, Thomas A. Environmental Monitoring and Support Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268, FTS 684-7311. Bidleman, Terry F. Department of Chemistry, University of South Carolina, Columbia, South Carolina 29208, (803) 777-4239. Breen, Joseph J. Office of Toxic Substances, U.S. Environmental Protection Agency, 401 M St., Washington, D.C. 20640, FTS 382-3569. Budde, William. Environmental Monitoring Support Laboratory, U.S, Environ" mental Protection Agency, Cincinnati, Ohio 45268, FTS 684-7309. Burgess, Ken. Dow Chemical Company, 1803 Building, Midland, Michigan, 48640, (517) 636-3177. Bursey, Joan T. Analytical Sciences Division, Research Triangle Institute, Research Triangle Park, North Carolina 27709, (919) 541-5928. Bumgarner, Joseph. Environmental Monitoring and Systems Laboratory, U(S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711, FTS 629-2434. Carey, Ann E. Office of Toxic Substances, U.S. Environmental Protection Agency, 401 M St., Washington, D.C. 20460, FTS 382-3569. Carra, Joseph. Office of Toxic Substances, U.S. Environmental Protection Agency, 401 M St., Washington, D.C. 20460, FTS 382-3900. Caspers, Horst. Stauffer Chemical Company, Dobbs Ferry, New York 10522, (914) 673-1200. Chesler, Stephen N. Organic Analytical Research Division, National Bureau of Standards, Washington, D.C. 20234, (301) 921-2153. Christman, Mark H. E. I. DuPont de Nemours, Wilmington, Delaware 19898, (202) 774-6443. Crouch, Michael D. Toxicon Laboratories, 3213 Monterrey Boulevard, Baton Rouge, Louisana 70814, (504) 925-5012. DaRoche, Maria. Sun Chemical Corporation, 441 Tompkins Avenue, Staten Island, New York 10305, (212) 981-1600 ext. 215. 77 �Dougherty, Ralph C. Department of Chemistry, Florida State University, Tallahassee, Florida 32306, (904) 644-5725. Ewald, Fred. PPG Industries, Inc., P.O. Box 31, Barberton, Ohio 44203, (216) 848-4600. Fensterheim, Robert J. (202) 887-1189. Gebhart, Judy. Ohio 43201. CMA, 2501 M Street, NW, Washington, B.C., Battelle Columbus Laboratories, 505 King Avenue, Columbus, Gunter, Bill. CCD, Office of Toxic Substances, U.S. Environmental Protection Agency, 401 M Street, SW, Washington, D.C. 20460, (202) 382-3933. Haile, Clarence L. Midwest Research Institute, 425 Volker Boulevard, Kansas City, Missouri 64110, (816) 753-7600. Hanneman, Larry F. Dow Corning, P.O. Box 1592, Midland, Michigan 48640, (517) 496-5003. Hass, J. Ronald. FTS 629-3463. NIEHS, Research Triangle Park, North Carolina 27711, Heggem, Daniel T. Field Studies Branch, Office of Toxic Substances, U.S. Environmental Protection Agency, TS-798, Washington, D.C. 20460, (202) 382-3584. Hensler, Charles. DuPont, Jackson Lab, Deepwater, New Jersey 08023, (609) 299-5000, ext. 3611. Hodges, Kent L. Dow Chemical Company, 574 Building, Midland, Michigan 48640, (517) 636-6544. Johnson, Larry. Industrial Environmental Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711, FTS 629-7943. Kaley, Robert. Monsanto Company, 800 North Lindbergh Boulevard, St. Louis, Missouri 63166, (314) 694-4964. Kingsley, Barbara. SRI International, 333 Ravenswood Avenue, Menlo Park, California 94025, (415) 326-6200. Kleopfer, Robert D. Region VII, U.S. Environmental Protection Agency, 1735 Baltimore Avenue, Kansas City, Missouri 64108, (816) 374-4285. Kutz, F. W. Field Studies Branch, Office of Toxic Substances, U.S. Environmental Protection Agency, TS-798, Washington, D.C. 20460, (202) 382-3569. Lewis, Robert G. Environmental Monitoring and Systems Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711, FTS 629-3065. 78 �Lopez-Avila, Viorica. Acurex, 485 Clyde Avenue, Mountain View, California 94042, (415) 964-3200. Moll, Amy. Regulatory Impacts Branch, Office of Toxic Substances, U.S. Environmental Protection Agency, TS-779, Washington, D.C. 20460 (202) 382-3715. Mullin, Michael D. U.S. Environmental Protection Agency, Large Lakes Research Station, 9311 Groh, Grosse lie, Michigan 48138, FTS 226-7011. Parris, Reenie. Organic Analytical Research Division, National Bureau of Standards, Washington, D.C. 20234, (301) 921-2153. Pellizzari, Edo D. Analytical Sciences Division, Research Triangle Institute, Research Triangle Park, North Carolina 27709. Petty, James D. Fish-Pesticide Research Laboratory, Fish and Wildlife Service, U.S. Department of Interior, Columbia, Missouri 65201, FTS 276-5399. Pfaffenberger, Carl D. Division of Chemical Epidemiology, School of Medicine, University of Miami, 15655 S.W. 127th Avenue, Miami, Florida 33177, (305) 255-3300. Redford, David. Office of Toxic Substances (TS-798), U.S. Environmental Protection Agency, Washington, D.C. 20460, FTS 382-3583. Robinson, J. Lawrence. Dry Color Manufacturers' Association, Suite 100, 1117 North 19th Street, Arlington (Rosslyn), Virginia 22209, (703) 525-9483Robinson, Thomas. Vulcan Materials Company, P.O. Box 7689, Birmingham, Alabama 35253, (205) 877-3556. Ronan, Richard. Versar, Inc., 6621 Electronic Drive, Springfield, Virginia 22151, (703) 750-3000. Safe, Stephen, Department of Physiology and Pharmacology, College of Veterinary Medicine, Texas A & M University, College Station, Texas 77843. Sawyer, Leon D. Food and Drug Administration, 240 Hennepin Avenue, Minneapolis, Minnesota 55401, FTS 725-2121. Slevon, Larry. Battelle Columbus Laboratories, 505 King Avenue, Columbus, Ohio 43201 Smith, John. DDB, Office of Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C. 20460, FTS 382-3900. Sonchik, Susan. Versar, Inc., 6621 Electronic Drive, Springfield, Virginia 22151, (703) 750-3000. Stalling, David L. Fish-Pesticide Research Laboratory, Fish and Wildlife Service, U.S. Department of Interior, Columbia, Missouri 65201, FTS 276-5399. 79 �Underwood, Joseph. Food and Drug Administration, 1009 Cherry Street, Kansas City, Missouri 64106, (816) 374-5524. Warren, Jackie. Natural Resources Defense Council, 122 East 42nd Street, New York, New York 10168. 80 �APPENDIX B BIBLIOGRAPHY 81 �BIBLIOGRAPHY Aaronson, M. J., J. D. Tessari, and S. M. Bergin, "PCBs in Transformer Oils," Bull. Environ. Contain. Toxicol., 28, 584-591 (1982). Acker, L., and F. Schulte, "Occurrence of Hexachlorobenzene and Polychlorinated Biphenylenes Besides Chlorinated Insecticides in Human Fat Tissues and Milk," Ernaehrungsforschung, 16(4), 559-67 (1972), Chem. Abst., 78, 80552c (1973). Adams, T. T., N. J. Williams, R. J. McElhaney, and R. A. Viator, "Analysis for Polychlorinated Biphenyls in Environmental Samples," Union Carbide report prepared for the U.S. Department of Energy, Contract No. W 7405 eng 26 (June 1979). Addison, J. B., and M. E. Nearing, "An Approach to the Preparation of Standard Reference Material for the Analysis of Polychlorinated Biphenyls in Marine Sediments," Intern. J. Environ. Anal. Chem., 11., 9-16 (1982). Afghan, B. K., and D. Mackay (Eds.), Hydrocarbons and Halogenated Hydrocarbons in the Aquatic Environment, Plenum Publishing Corporation, New York. Proceedings of International Symposium on the Analysis of Hydrocarbons and Halogenated Hydrocarbons in the Aquatic Environment, Ontario, Canada, 1978 (1980). Ahling, B., and S. Jensen, "Reversed Liquid-Liquid Partition in Determination of Polychlorinated Biphenyl (PCB) and Chlorinated Pesticides in Water," Anal. Chem., 42(13), 1483-1486 (1970). Ahnoff, M., and B. Josefsson, "Confirmation Studies on Polychlorinated Biphenyls (PCB) from River Waters Using Mass Fragmentography," Anal.Lett., 6(12), 1083-1093 (1973). Ahnoff, M. , and B. Josefsson, "Simple Apparatus for On-Site Continuous Liquid^ Liquid Extraction of Organic Compounds from Natural Waters," Anal. Chem., 46(6), 658-663 (1974). Ahnoff, M., G. Eklund, and B. Josefsson, "Analysis of PCB in Water from the Swedish West Coast," Acta hydrochim. Hydrobiol., 7 ( ) 171-178 (1979). .2, Ahnoff, M., and B. Josefsson, "Clean Up Procedures for PCB Analysis on River Water Extracts, Bull. Environ. Contain. Toxicol., 13(2), 159-166 (1975). Aitzetmuller, K., "Adsorption Liquid Chromatography of DDT and Polychlorinated Biphenyls,11 J. Chromatogr. , 107, 411-415 (1975). 82 �Albro, P. W., "Problems in Analytic Methodology: Sample Handling, Extraction, and Cleanup," Ann. N.Y. Acad. Sci., 320, 19-27 (1979). Albro, P. W., "Validation of Extraction and Cleanup Procedures for Environmental Analysis," Chapter 8 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, J. D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981). pp. 163-175. Albro, P. W., J. R. Hass, and W. B. Crummett, "Summary of the Workshop on Recent Advances in Analytical Techniques for Halogenated Aromatic Compounds," Ann. N.Y. Acad. Sci.t 320, 125-130 (1979). Albro, P. W., J. T. Corbett, and J. L. Schroeder, "Quantitative Characterization of Polychlorinated Biphenyl Mixtures (Aroclors® 1248, 1254 and 1260) by Gas Chromatography Using Capillary Columns," J. Chromatogr., 205, 103-111 (1981) Albro, P. W., and C. E. Parker, "Comparison of the Compositions of Aroclor 1242 and Aroclor 1016," J. Chromatogr., 169, 161-166 (1979). Albro, P. W., and C. E. Parker, "General Approach to the Fractionation and Class Determination of Complex Mixtures of Chlorinated Aromatic Compounds," J. Chromatogr., 197, 155-169 (1980). Albro, P. W., and L. Fishbein, "Quantitative and Qualitative Analysis of Polychlorinated Biphenyls by Gas-Liquid Chromatography and Flame lonization Detection. I. One to Three Chlorine Atoms," J. Chromatogr., 69, 273-283 (1972). Albro, P. W., and L. Fishbein, "Intestinal Absorption of Polychlorinated Biphenyls in Rats," Bull. Environ. Contain. Toxicol., 8(1), 26-31 (1972). Albro, P. W., J. K. Haseman, T. A. Clemmer, and B, J. Corbett, "Identification of the Individual Polychlorinated Biphenyls in a Mixture by Gas-Liquid Chromatography," J. Chromatogr.. 136, 147-153 (1977). American National Standards Institute, Inc., "American National Standard Guidelines for Handling and Disposal of Capicator- and Transformer-Grade Askarels Containing Polychlorinated Biphenyls," ANSI C107.1-1974. American Society for Testing and Materials, "Standard Method for Rapid Gas Chromatographic Estimation of High Boiling Homologues of Chlorinated Biphenyls for Capacitor Askarels," in Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania (1980). pp. 830-876, ANSI/ASTM D 3303 - 74 (Reapproved 1979). American Society for Testing and Materials, "Standard Method for Analysis of Environmental Materials for Polychlorinated Biphenyls," in Annua1 Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania (1980). pp. 877-885, ANSI/ASTM D 3304 - 77. Analabs, "Chromatography Chemicals and Accessories" (1981). 83 �Anderson, R. J., and R. C. Hall, "Hall Bipolar Pulse, Differential Electrolytic Conductivity Detector for GC," Amer. Lab., 108-124 (February 1980). Anonymous, "Method Improves PCB Measurements in Oil," Chem. Eng. News, ^9, March 1, 1982. Anonymous, New Scientist, 612 (1966), Araki, Y., and K. Tanaka, "Simple Screening Method for [Hepatic] Enzymeinducing Agents of Polychlorinated Biphenyl (PCB) Type," Fukuoka-Igaku-Zasshi, 65(1), 47-52 (1974); Cfaem. Abst.. 81, 146433n (1974). Armour, J. A., "Gas Chromatographic Data for Polychlorinated Biphenyl Components in Six Aroclors®," J. Cfaromato.gr.. , 72, 275-282 (1972). Armour, J. A., and J. A. Burke, "Method for Separating Polychlorinated Biphenyls from DDT and Its Analogs," J. Assoc. Offic. Anal. Chem., 53(4), 761-768 (1970). Armour, J. A., and J. A. Burke, "Behavior of Chlorinated Naphthalenes in Analytical Methods for Organochlorine Pesticides and Polychlorinated Biphenyls," J. Assoc. Offic. Anal. Chem., 54, 175-177 (1971). Armour, J. A., "Quantitative Perchlorination of Polychlorinated Biphenyls as a Method for Confirmatory Residue Measurement and Identification," J. Assoc. Offic. Anal. Chem.. 56(4), 988-993 (1973). Armstrong, D. W., and R. Q. Terrill, "Thin Layer Chromatographic Separation of Pesticides, Decachlorobiphenyl, and Nucleosides with Micellar Solutions," Anal. Chem., 51(13), 2160-2163 (1979). Arpino, P. J., B. G. Dawkins, and F, W. McLafferty, "A Liquid Chromatography/ Mass Spectrometry System Providing Continuous Monitoring with Nanogram Sensitivity," J. Chromatogr. Sci., 12, 574-578 (1974). Association of Official Analytical Chemists, "General Method for Organochlorine and Organophosphorus Pesticides," in Official Methods of Analysisofthe Assor ciation of Official Analytical Chemists, W. Horwitz (Ed.), 13th ed., Washington, D.C. (1980a). pp. 466-474. Association of Official Analytical Chemists, Official Methods of Analysis of the Association of Official Analytical Chemists, W. Horwitz (Ed.), 13th ed., Washington, D.C. (1980b). p. 475. Aue, W. A., and S. Kapila, "The Electron Capture Detector—Controversies, Comments, and Chromatograms," J. Chromatogr. Sci., 11, 255-263 (1973). Ayer, F. A. (Ed.), "National Conference on Polychlorinated Biphenyls (November 19-21, 1975, Chicago, Illinois)," Report of proceedings prepared for the U.S. Environmental Protection Agency, Office of Toxic Substances, Contract No. 68-01-2928, Report No. EPA-560/6-75-004 (March 1976). 84 �Bagley, G. E., W. L. Reichel, and E. Cromartie, "Identification of Polychlo-* rinated Biphenyls in Two Bald Eagles by Combined Gas-Liquid ChromatographyMass Spectrometry," J. Assoc. Offic. Anal. Chem., 53(2), 251-261 (1970). Bagley, G. E., and E. Cromartie, "Elimination Pattern of Aroclor 1254 Components in the Bobwhite," J. Chromatogr., 75, 219-226 (1973). Ballinger, D. G., "Test Procedures for Priority Organics in Municipal Wastewater and Sludges," U.S. Environmental Protection Agency, Cincinnati, Ohio, December 11, 1978. Ballschmiter, K., and M. Zell, "Analysis of Polychlorinated Biphenyls (PCB) by Glass Capillary Gas Chromatography. Composition of Technical Aroclorand Clophen-PCB Mixtures," Fresenius 2. Anal. Chem., 302, 20-31 (1980). Balya, B. R., and G. H. Farrah, "Column Chromatographic Procedure for the Determination of Polychlorinated Biphenyls in Certain Industrial Oils," Anal. Lett., 13(B8), 655-671 (1980). Bartle, K. D., "Identification of Polychlorobiphenyls by High Resolution Proton Magnetic Resonance," J. Assoc. Qffic. Anal. Chem.. 55(5), 1101-1103 (1972). Bauer, U., "Combination of Thin-layer and Gas Chroraatography for Identifying Halogenated Hydrocarbons," Vom Wasser, 38, 49-62 (1971); Chem. Abstr., 76, 136672x (1972). Becker, G., and E. Schulte, "Presence of PCB in Packaging Materials," Mitteilungsbl. GDCh-Fachgruppe Lebensmittelchem. Gerichtl. Chem., 30(6), 101-108 (1976); Chem. Abstr., 85, 107564u (1976). Beezhold, F. L., and V. F. Stout, "The Use and Effect of Mixed Standards on the Quantitation of Polychlorinated Biphenyls," Bull. Environ. Contain. Toxicol. 10(1), 10-15 (1973). Beggs, D. P., and D. B. Banks, "Qualitative Analysis of PCB by GC/MS," Application Note AN 176-20, Hewlett-Packard, Avondale, Pennsylvania (1976). Bellar, T. A., and J. J. Lichtenberg, "Some Factors Affecting the Recovery of Polychlorinated Biphenyls (PCB's) from Water and Bottom Samples," in Special Technical Publication 573, American Society for Testing and Materials, Philadelphia, Pennsylvania (1975). pp. 206-219. Bellar, T. A., and J. J. lichtenberg, "The Determination of Polychlorinated Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environmental Protection Agency, EPA-600/4-81-045 (1981). Bellar, T. A., J. J. Lichtenberg, and S. C. Lonneman, "Recovery of Organic Compounds from Environmentally Contaminated Bottom Materials," Chapter 4 in Contaminants and Sediments, Vol. 2, R. A. Baker (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1980). 85 �Belliardo, F., E. Gionchiglia, and G. M. Nano, "Analysis of Polychlorinated Biphenyl Residues in Waste Oils by High-Performance Liquid Chromatography," J. Liq. Chromatogr., 2(1), 77-83 (1979). Berg, 0. W., P. L. Diosady, and G. A. V. Rees, "Column Chromatographic Separation of Polychlorinated Biphenyls from Chlorinated Hydrocarbon Pesticides, and their Subsequent Gas Chromatographic Quantitation in Terms of Derivatives," Bull. Environ. Contain. Toxicol.. 7(6), 338-347 (1972). o Bergman, A., R. Gothe, and C. A. Wachtmeister, "Impregnation of Silica Gel with Tetraalkylammonium Salts in Adsorption Chromatography of Neutral Aromatic Compounds," J. Chromatogr., 123, 231-236 (1976). Bickford, B., J. Bursey, L. Michael, E. Pellizzari, R. Porch, D. Rosenthal, L. Sheldon, C. Sparacino, K. Tomer, R. Wiseman, S. Yung, J. Gebhart, L. Rando, D. Perry, and J. Ryan, "Preliminary Draft Report, Master Scheme for the Analysis of Organic Compounds in Water, Part III: Experimental Development and Results," EPA Contract No. 68-03-2704, March 1980. Bidleman, T. F., W. N. Billings, and C. G. Simon, "Evaluation of Solid Adsorbents for Collecting Atmospheric Chlorinated Hydrocarbons," EPA 600/2-80-167 (1980). Bidleman, T. F., "Interlaboratory Analysis of High Molecular Weight Organochlorines in Ambient Air," Atmos. Environ., 15, 619-624 (1981). Bidleman, T. F., E. J. Christensen, W. N. Billings, and R. Leonard, "Atmospheric Transport of Organochlorines in the North Atlantic Gyre," J. Marine Res., 39(3), 443-464 (1981). Bidleman, T. F., J. R. Matthews, C. E. Olney, and C. P. Rice, "Separation of Polychlorinated Biphenyls, Chlordane, and £,p_'-DDT from Toxaphene by Silicic Acid Column Chromatography," J. Assoc. Of fie. Anal. Chem., 6_1(4), 820-828 (1978). Bidleman, T. F., and Ross Leonard, "Aerial Transport of Pesticides over the Northern Indian Ocean and Adjacent Seas," Atmos. Environ., 16(1), in press (1982). Bidleman, T. F., E. J. Christensen, W. N. Billings, and R. Leonard, "Atmospheric Transport of Organochlorines in the North Atlantic Gyre," Reprint from Sears Foundation: Journal Marine Research, 39(3), 443-464, August 1981. Bidleman, T. F., and C. E. Olney, "High-volume Collection of Atmospheric Polychlorinated Biphenyls," Bull. Environ. Contam. Toxicol., 1^(5), 442-450 (1974), Billings, W. N., and T. F. Bidleman, "High Volume Collection of Chlorinated Hydrocarbons in Urban Air Using Three Solid Adsorbents," submitted to Atmos. Environ., 1982. Billings, W. N., and T. F. Bidleman, "Field Comparison of Polyurethane Foam and Tenax-GC Resin for High-Volume Air Sampling of Chlorinated Hydrocarbons," J. Amer. Chem. Soc., 14(6), 679-683 (1980). 86 �Biros, F. J., E. F. Enos, A. C. Walker, and J. F. Thompson, "Analytical Aspects of Polychlorinated Biphenyl Determinations in Human and Animal Adipose Tissue," Prepr. Pap. Natl. Meet., Div. Water, Air Waste Chem., Amer. Chem. Soc., 12, 61-66 (1972). Biros, F. J., A. C. Walker, and A. Medbery, "Polychlorinated Biphenyls in Human Adipose Tissue," Bull. Environ. Contain. Toxicol.. 5(4), 317-323 (1970). Biros, F. J., "An Integrated Gas Chromatograph-Flame lonization Detector-Mass Spectrometer System for Pesticide Residue Analysis," Pestic. Chem. Proc. Int. IUPAC Congr. Pestic. Chem. 2nd, 4, 149-169 (1971). Boe, B., and E. Egaas, "Qualitative and Quantitative Analyses of Polychlorinated Biphenyls by Gas-Liquid Chromatography," J. Chromatogr., 180, 127-132 (1979). Bonelli, E. J., "GC/MS Techniques for the Determination of Interferences in Pesticide Analysis," Ins t. Environ. Sci. Proc., 18, 251-254 (1972a). Bonelli, E. J., "Gas Chromatograph/Mass Spectrometer Techniques for Determination of Interferences in Pesticide Analysis," Anal. Chem., 4 ( ) 603-606 43, (1972b). Boniforti, L., G. Citti, and G. Laguzzi, "Identification and Quantitative Determination of Polychlorinated Biphenyls and Polynuclear Aromatic Hydrocarbons in Fly Ash from Municipal Incinerators by Gas Chromatography with ElectronCapture Detection and Gas Chromatography-Mass Spectrometry," Anal. Chem. Symp. Ser.. 7, 219-225 (1981); Chem. Abstr., 96, 90803y (1982). Bourke, J. B., vPCB's and Pesticides," N.Y. State Assoc. Milk and Food Sanitarians, Ann. Rept., 46, 41-43 (1972). Bowes, G. W., and J. A. Lewis, "Extraction of Polychlorinated Biphenyls: Evaluation of A Column Technique Applied to Polar Bear and Seal Tissue," J. Assoc. Offic. Anal. Chem., 57(1), 138-144 (1974). Bowes, G. W., M. J. Mulvihill, M. R. DeCamp, and A. S. Kende, "Gas Chromatographic Characteristics of Authentic Chlorinated Dibenzofurans; Identification of Two Isomers in American and Japanese Polychlorinated Biphenyls," J. Agric. Food Chem., 23(6), 1222-1223 (1975). Brinkman, U. A. Th., and G. De Vries, "Relationship Between Structure and Retention of Polyhalogenated Aromatics in Two Adsorption Chromatographic Systems," J. Chromatogr., 169. 167-182 (1979). Brinkman, U. A. T., and A. De Kok, "Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products. Production, Properties and Usage," Top. Environ. Health, 4, 1-40 (1980). Brinkman, U. A. Th., G. De Vries, A. De Kok, and A. L. De Jonge, "Discrimination Between Polychlorinated Naphthalenes and Polychlorinated Biphenyls," J. Chromatogr., 152, 97-104 (1978). 87 �Brinkman, U. A. Th., A. De Kok, G. De Vries, and H. G. M. Reymer, "Highspeed Liquid and Thin-layer Chroraatography of Polychlorinated Biphenyls," J. Chromatoer., 128, 101-110 (1976a), Brinkman, U. A. Th., J. W. F. L. Seetz, and H. G. M. Reymer, "High-speed Liquid Chromatography of Polychlorinated Biphenyls and Related Compounds," J. Chromatogr.. 116, 353-363 (1976b). Brown, D, P., and M. Jones, "Mortality and Industrial Hygiene Study of Workers Exposed to Polychlorinated Biphenyls," Archives of Environmental Health, 36(3), 120-129 (1981). Brownrigg, J. T., and A. W. Hornig, "Identification of PCB's in the Presence of DDT-Type Compounds Using Low Temperature Luminescence," Arch. Environ. Contam. Toxicol., 4, 175-182 (1976). Brownrigg, J. T., et al., "Estimation of Polychlorinated Biphenyls in the Presence of DDT-Type Compounds," Prepared for the U.S. Environmental Protection Agency, NTIS Publication No. PB-233 599 (June 1974). Burdick, N. F., and T. F. Bidleman, "Frontal Movement of Hexachlorobenzene and Polychlorinated Biphenyl Vapors through Polyurethane Foam," Anal. Chem., 53(12), 1926-1929 (1981). Bruggeman, W. A., J. van der Steen, and 0. Hutzinger, "Reversed-Phase ThinLayer Chromatography of Polynuclear Aromatic Hydrocarbons and Chlorinated Biphenyls. Relationship with Hydrophobicity as Measured by Aqueous Solubility and Octanol Water Partition Coefficient," J. Chromatogr., 238(2), 335-346 (1982). Burke, J. A., "Report on Chlorinated Pesticides," J. Assoc. Offie. Anal. Chem., 59(2), 338-340 (1976). Burke, J. A., "Report on Chlorinated Insecticides," J. Assoc. Offie. Anal. Chem.. 58(2), 233-235 (1975). Burkhard, L. P., and D. E. Armstrong, "Labeled Perchlorination Reagent for De* termination of Polychlorinated Biphenyls," Anal. Chem., 53(3), 523-528 (1981). Burse, V. W., R. F. Moseman, G. W. Sovocool, and E. C. Villanueva, "PCB Metabolism in Rats Following Prolonged Exposure to Aroclor 1242 and Aroclor 1016," Bull. Environ. Contam. Toxicol., 15(1), 122-128 (1976). Bush, B., and F.-C. Lo, "Thin-layer Chromatography for Quantitative Polychlorinated Biphenyl Analysis," J. Chroma togr., 77_, 377-388 (1973). Bush, B., F. Baker, R. Dell'Acqua, C. L. Houck, and F.-C. Lo, "Analytical Response of Polychlorinated Biphenyl Homologues and Isomers in Thin-layer and Gas Chromatography," J. Chromatogr., 109, 287-295 (1975). 88 �Cairns, T., and E. G. Siegmund, "Determination of Polychlorinated Biphenyls by Chemical lonization Mass Spectrometry," Anal. Chetn., 53(11), 1599-1603 (1981). Cairns, T., and E. G. Siegmund, "PCBs: Regulatory History and Analytical Problems," Anal. Cfaem., 53(11), 1183A-1193A (1981). Cairns, T, and R. A. Jacobson, "New Approaches to FDA Analytical Problems," J. Chem. Infor. Comp. Sci.. _T7(2), 105-109 (1977). Call, D. J., Y. A. Greichus, and J. J. Worman, "A Gas Chromatographic-mass Spectrometric Comparison of Polychlorinated Biphenyl Residues in the Japanese Quail Brain to an Aroclor Standard," Bull. Environ. Contain. Toxicol., 11(4), 333-338 (1974). Canada, D. C., and F. E. Regnier, "Isotope Ratios as a Characteristic Selection Technique for Mass Chromatography," J. Chromatogr. Sci., 14, 149-154 (1976). Caragay, A. B., and P. L. Levins, "Evaluation of Protocols for Pesticides and PCB's in Raw Wastewater," Final Report, Arthur D. Little, Inc., Contract No. 68-01-3857, Office of Research and Development, Environmental, Protection Agency, Report No. EPA-600/2-79-166, November 1979. Carey, A. E., and J. A. Gowen, "PCB's in Agricultural and Urban Soil," In; Proceedings of the National Conference on PCBs, pp. 195-198, EPA 560/6-75-004 (1976). Chau, A. S. Y., and L. J. Babjak, "Column Chromatographic Determination of Mirex, Photomirex, and Polychlorinated Biphenyls in Lake Sediments," J. Assoc. Off. Anal. Chem., 62(1), 107-113 (1979). Chau, A. S. Y., J. Carron, and H.-B. Lee, "Analytical Reference Materials. II. Preparation and Sample Integrity of Homogeneous Fortified Wet Sediment for Polychlorinated Biphenyl Quality Control Studies," J. Assoc. Off. Anal. Chem., 62(6), 1312-1314 (1979). Chau, A. S. Y., and H.-B. Lee, "Analytical Reference Materials. III. Preparation and Homogeneity Test of Large Quantities of Wet and Dry Sediment Reference Materials for Long Term Polychlorinated Biphenyl Quality Control Studies," J. Assoc. Off. Anal. Chem.. 63, 947-951 (1980). Chau, A. S. Y., and C. J. Sampson, "Electron Capture Gas Chromatographic Methodology for the Quantitation of Polychlorinated Biphenyls: Survey and Compromise," Environ. Lett., 8(2), 89-101 (1975). Chemical Manufacturers Association, "The Analysis of Chlorinated Biphenyls," submitted to the Environmental Protection Agency, August 21, 1981, in reference to OPTS-62017. Chen, P. H., K. T. Chang, and Y. D. Lu, "Polychlorinated Biphenyls and Polychlorinated Dibenzofurans in the Toxic Rice-Bran Oil that Caused PCS Poisoning in Taichung," Bull. Environ. Contain. Toxicol., 26, 489-495 (1981). 89 �Chesler, S. N., F. R. Guenther, W. E. May, and R. M. Parris, "SRM's for Accurate Analyses of PCBs in Oil," Abstract, Center for Analytical Chemistry, National Bureau of Standards, Washington, DC, 20234. Chesler, S. N., W. E. May, P. A. White, R. M. Parris, and F. R. Guenther, "Determination of Polychlorinated Biphenyls in Waste and Lubricating Oils," National Bureau of Standards Special Publication 584, Proceedings of Joint Conference on Measurements and Standards for Recycled Oil/Systems Performance and Durability, Gaithersburg, MD, October 23-26, 1979, pp. 295-297, NBS, (Issued November 1980). Christman, R. P., "Guidelines for GC/MS Identification," ES&T Editorial, Environ. Sci. Tecfanol., 16(3), 143A (1982). Claeys, R. R., "Analysis of PCBs in Pulp Mill Effluent," Special Report-NCASI (National Council of the Paper Industry for Air and Stream Improvement), No. 77-04, 1977. pp. 46-49. Coburn, J. A., I. A. Valdmanis, and A. S. Y. Chau, "Evaluation of XAD-2 for Multiresidue Extraction of Organochlorine Pesticides and Polychlorinated Biphenyls from Natural Waters," J. Assoc. Off. Anal. Chem., 60(1), 224-228 (1977). ' ~ ~ — Coleman, W. E., and R. G. Tardiff, "Contaminant Levels in Animal Feeds Used for Toxicity Studies," Arch. Environ. Contam. Toxicol., 8, 693-702 (1979), Colenutt, B. A,, and S. Thorburn, "Optimisation of a Gas Stripping Concentration Technique for Trace Organic Water Pollutants," Intern. J. Environ. Anal. Chem., 7, 231-244 (1980). Collard, R. S., and M. M. Irwin, Jr., "GC/MS Determination of Chlorinated Biphenyls in Complex Chlorinated Hydrocarbon Ptocess and Waste Streams," The Dow Chemical Company (1982). Collard, R. S., and M. M. Irwin, Jr., "Determination of Chlorinated Biphenyls (Incidental PCBs) in Chlorinated Hydrocarbon Process and Waste Streams," Abstract, 184th Annual American Chemicaly Society National Meeting, September 12-17, 1982. Collins, G. B., D. C. Holmes, and F. J. Jackson, "The Estimation of Polychlorobiphenyls," J. Chromatogr., 71, 443-449 (1972). Cooke, M., G. Nickless, A. M. Prescott, and D. J. Roberts, "Analysis of Polychlorinated Naphthalenes, Polychlorinated Biphenyls and Polychlorinated Terphenyls via Carbon Skeleton Gas-Liquid Chromatography," J. Chromatqgr.? jt5J>> 293-299 (1978). Cordle, F., P. Corneliussen, C. Jelinek, B. Hackley, R. Lehman, J. McLaughlin, R. Rhoden, and R. Shapiro, "Human Exposure to Polychlorinated Biphenyls and Polybrominated Biphenyls," Environ. Health Perspec., 24, 157-172 (1978). 90 �Corneliussen, P. E., "Report on Multiresidue Methods (Interlaboratory Studies)," J. Assoc. Off. Anal. Chem., 5jK2), 238-239 (1975). Copland, G. B., and C. S. Gohmann, "Improved Method for Polychlorinated Biphenyl Determination in Complex Matrices," Environ. Sci. Techno1., 16(2), 121-124 (1982). Crist, H. L., and R. F. Moseman, "Simplified Micro Perchlorination Method for Polychlorinated Biphenyls in Biological Samples," J. Assoc. Off. Anal. Chem., 60(6), 1277-1281 (1977). Curley, A., V. W. Burse, M. E. Grim, R. W. Jennings, and R. E. Linder, "Polychlorinated Biphenyls: Distribution and Storage in Body Fluids and Tissues of Sherman Rats," Environ. Res.. 4, 481-495 (1971). Dark, W. A., and L. W. Crossman, Jr., "Applications Highlights - Polychlorinated Biphenyls," Waters Associates, Milford, Massachusetts, AH-324, 1973. Das, K. G., and P. S. Kulkarni, "Gas-Liquid Chromatograpy," Chapter 2 in Pesticide Analysis, K. G. Das (Ed.), Marcell Dekker, Inc., New York (1981). De Kok, A., R. B. Geerdink, R. W. Frei, and U. A. Th. Brinkman, "The Use of Dechlorination in the Analysis of Polychlorinated Biphenyls and Related Classes of Compounds," Intern. J. Environ. Anal. Chem., 9, 301-318 (1981). De Kok, A., R. B. Geerdink, R. W. Frei, and U. A. Th. Brinkman, "Limitations in the Use of Perchlorination as a Technique for the Quantitative Analysis of Polychlorinated Biphenyls," Intern. J. Environ. Anal. Chem., 11, 17-41 (1982). Delfino, J. J., "Toxic Substances in the Great Lakes," Enviroa.Sci. Techno1., 13(12), 1462-1468 (1979). Delfino, J. J. (Ed.), Methods for Analysis of Organic Compounds in the Great Lakes, Proceedings of an Invitational Workshop, April 1-2, 1980, University of Wisconsin Great Lakes Research Facility, Milwaukee, Wisconsin, WIS-SG-80-236 (1980). Delfino, J. J., and D. B. Easty, "Interlaboratory Study of the Determination of Polychlorinated Biphenyls in a Paper Mill Effluent," Anal. Chem., 51(13), 2235-2239 (1979). Department of Commerce, "Standard Reference Material 1581, Polychlorinated Biphenyls in Oil," Summer 1982. Department of Health, Education and Welfare, "Final Report of the Subcommittee on the Health Effects of Polychlorinated Biphenyls and Polybrominated Biphenyls of the DHEW Committee to Coordinate Toxicology and Related Programs," Env. Health Persp.. 24 (1978). 91 �Department of Health, Education and Welfare, "DHEW Subcommittee on Health Effects of PCBs and PBBs, General Recommendations," Env. Health Persp., 24, 187-189 (1978a). Department of Health, Education and Welfare, "DHEW Subcommittee on Health Effects of PCBs, General Summary and Conclusions," Env. Health Persp., 24, 191-198 (1978b). Dequidt, J., F. Erb, C. Van Aerde, and P. Colein, "Determination of Polychlorinated Biphenyls in Biological Media," Bull. Soc. Pharm. Lille, 31(2-3), 149-155 (1975); Chem. Abst., 84, 39391n (1976T Devenish, I. W., and L. Harling-Bowen, "The Examination and Estimation of the Performance Characteristics of a Standard Method for Organo-Chlorine Insecticides and PCB," Environ. Sci. Res.. 16, 231-253 (1980). De Vos, R. H., "Analytical Techniques in Relation to the Contamination of the Fauna," TNO-nieuws, 27, 615-622 (1972). De Vos, R. H., and E. W. Peet, "Thin-Layer Chromatography of Polychlorinated Biphenyls," Bull. Environ. Contain. Toxicol., 6(2), 164-170 (1971). Dexter, R. N., and S. P. Pavlou, "Characterization of Polychlorinated Biphenyl Distribution in the Marine Environment," Bull. Environ. Contam. Toxicol., 1£>(4), 477-482 (1976). Dolan, J. W,, and R. C. Hall, "Enhancement of the Sensitivity and Selectivity of the Coulson Electrolytic Conductivity Detector to Chlorinated Hydrocarbon Pesticides," Anal. Chem., 45(13) 2198-2204 (1973). Dolan, J. W., R. C. Hall, and T. M. Todd, "Selective Detection of Chlorinated Insecticides in the Presence of Polychlorinated Biphenyls," J. Assoc. Off. Anal. Chem., 55(3), 537-538 (1972). Dolphin, R. J., and F. W. Willmott, "Separation of Chlorinated Dibenzo-£» dioxins from Chlorinated Congeners," J. Chromatogr., 149, 161-168 (1978). Donkin, P., S. V. Mann, and E. I. Hamilton, "Microcoulometric Determination of Total Organochlorine Pesticide and Polychlorinated Biphenyl Residues in Grey Seal (Halichoerus grypus) Blubber," Anal. Chim. Acta, 88, 289-301 (1977). Doskey, P. V., and A. W. Andren, "High-Volume Sampling of Airborne Polychlorobiphenyls with Amberlite XAD-2 Resin," Anal. Chim. Acta, 110, 129-137 (1979). Dougherty, R. C., "Chemical Geometrodynamics: Physical Fields can Cause Asymmetric Synthesis," Origins of Life, 11., 71-84 (1981). Dougherty, R. C., "Negative Chemical lonization Mass Spectrometry: Applications in Environmental Analytical Chemistry," Biomed. Mass Spectrom., 8(7), 283-291 (1981a). 92 �Dougherty, R. C., M. J, Whitaker, S.-Y. Tang, R. Bottcber, M. Keller, and D. W. Kuehl, "Sperm Density and Toxic Substances: A Potential Key to Environmental Health Hazards," Chapter 13 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, J, D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981b). pp. 263-278. Dougherty, R. C., J. D. Roberts, H. P. Tannenbaum, and F. J. Biros, "Positive and Negative Chemical lonization Mass Spectra of Polychlorinated Pesticides," *-n Mass Spectrometry and NMR Spectroscopy in Pesticide Chemistry, R. Haque and F. J. Biros (Eds.), New York-London (1973). pp. 33-47. Dougherty, R. C., M. J. Whitaker, L. M. Smith, D. L. Stalling, and D. W. Kuehl, "Negative Chemical lonization Studies of Human and Food Chain Contamination with Xenobiotic Chemicals," Environ. Health Persp.. 36, 103-118 (1980). Dow Chemical Company, "Determination of Chlorinated Biphenyls in the Presence of Chlorinated Benzenes," July 1, 1981. Dry Color Manufacturers Association, "An Analytical Procedure for the Determination of Polychlorinated Biphenyls in Dry Phthalocyanine Blue, Phthalocyanine Green, and Diarylide Yellow Pigments." Dunn, W. D., E. Johannson, D. L. Stalling, T. R. Schwartz, B. K. Marlow, J. D. Petty, and J. D. Hogan, "Application of SIMCA to the Characterization of PCB Residue Profiles in Sediment, Water, and Aquatic Biota," Abstract, 184th Annual Chemical Society National Meeting, September 12-17, 1982. Dzidic, I., D. I. Carroll, R. N. Stillwell, and E. C. Horning, "Atmospheric Pressure lonization (API) Mass Spectrometry: Formation of Phenoxide Ions from Chlorinated Aromatic Compounds," Anal. Chem., 4 ( ) 1308-1312 (1975). 78, Easty, D. B., "Quantitation of Mixtures of Polychlorinated Biphenyls Isolated from Paperboard Containing Recycled Fiber," Tech. Assoc. Pulp and Paper Ind., 56(5), 131-134 (1973). Easty, D. B., and B. A. Wabers, "Determination of Polychlorinated Biphenyls in Paper Mill Effluents," Techy Assoc. Pulp and Paper Ind., 61(10), 71-74 (1978). Eder, G., "Polychlorinated Biphenyls and Compounds of the DDT Group in Sediments of the Central North Sea and the Norwegian Depression," Chemosphere, 2, 101-106 (1976a). Eder, G., "Evaluation of Environmental Polychlorobiphenyls and DDE in Terms of Mixtures of Commercial Preparations from Peak Heights of Packed-Column Gas Chromatograms Using a Programmable Calculator," J. Chromatogr., 121, 269-277 (1976b). 93 �Edison Electric Institute, and The National Rural Electric Cooperative Association, "Comments and Studies on the Use of Polychlorinated Biphenyls in Response to an Order of the United States Court of Appeals for the District of Columbia Circuit," Vol. Ill, Reportof the Study of PCBs in Equipment Owned by the Electric Utility Industry, submitted to the U.S. Environmental Protection Agency, Resource Planning Corporation, Washington, B.C. (1932). Edwards, R., "Some Factors in the Separation of Polychlorobiphenyls (PCBs) from Organochlorine Pesticides by Column Chromatography Combined with GasLiquid Chromatography," Pestic. Sci., 5, 293-304 (1974). Egestad, B., T. Curstedt, and J. Sjovall, "Simple Procedures for Enrichment of Chlorinated Aromatic Pollutants from Fat, Water and Milk for Subsequent Analysis by High-Resolution Methods," Anal. Lett., 15(A3), 293-307 (1982). Eichelberger, J. W., L. E. Harris, and W. L. Budde, "Application of Gas Chroraatography-Mass Spectrometry with Computer Controlled Repetitive Data Acquisition from Selected Specific Ions," Anal. Chem., 46(2), 227-232 (1974). Eisenbeiss, F., and H. Sieper, "The Potential Use of High-Performance Liquid Chromatography on Residue Analysis," J. Chromatogr., 83, 439-446 (1973). Environmental Protection Agency, Office of Pesticides Programs, Benefits and Field Studies Division, "Sample Collection Manual, Ecological Monitoring Branch," Washington, D.C. (1976). Environmental Protection Agency, Office of Research and Development, Environmental Monitoring and Support Laboratory, "The Analysis of Polychlorinated Biphenyls in Transformer Fluid and Waste Oils," (February 1981). Environmental Protection Agency, "Organochlorine Pesticides and PCBs-- Method 608 »" Fed. Reg., 44, 69501-69509 (December 3, 1979a). Environmental Protection Agency, "Base/Neutrals, Acids, and Pesticides—Method 625 >" Fed. Reg., 44, 69540-69552 (December 3, 1979b), and subsequent revisions. Environmental Protection Agency, "Identification of Polychlorinated Biphenyls in the Presence of DDT-Type Compounds" EPA-R2-72-064 (NTIS PB-213 900) (October 1972). Environmental Protection Agency, "Appendix III - Example Quality Assurance and Quality Control Procedures for Organic Priority Pollutants," Fed. Reg., 44, 69553-69559 (December 3, 1979c). Environmental Protection Agency, Fed. Reg., 44, 31514-31568 (1979d) Environmental Protection Agency, Environmental Monitoring and Support Laboratory, "Analytical Protocol for Screening Publicly Owned Treatment Works (POTW) Sludges for Organic Priority Pollutants," (September 1979e). 94 �Environmental Protection Agency, "Methods for Benzidine, Chlorinated Organic Compounds, Pentachlorophenol and Pesticides in Water and Wastewater," Interim Report, September 1978. Environmental Protection Agency, "Guidelines and Specifications for Preparing Quality Assurance Project Plans," Draft, May 5, 1980a. Environmental Protection Agency, "Quality Assurance Program Plan" for the Office of Toxic Substances, Office of Pesticides and Toxic Substances, October 1980b. Erickson, M. D., "Review of the Report 'Statistical Analysis of Data from a Round Robin Experiment on PCB Samples,'" Letter Report, U.S. EPA Contract No. 68-01-5915, Task 06, February 1982. Erickson, M. D., and E. D. Estes, "Evaluation of Chlorinated Hydrocarbon Catalytic Reduction Technology" EPA-600/2-78-059, March 1978. Erickson, M. D., J. B. Goforth, M. A. Jones, M. C. Saylor, K. B. Tomer, and E. D. Pellizzari, "Analysis of Blood and Adipose Tissue Samples from GSA Workers for Polychlorinated Biphenyls (PCBs) Phase I: Packed Column GC/ECD and GC/MS Analyses," Interim Report, Contract No. 68-01-3849, Task 4, U.S. Environmental Protection Agency (May 1981). Erickson, M. D., and E. D. Pellizzari, "Identification and Analysis of Polychlorinated Biphenyls and Other Related Chemicals in Municipal Sewage Sludge Samples," Final Report, EPA-68-01-1978, Office of Toxic Substances, (August 1977). Erickson, M. D., and E. D. Pellizzari, "Analysis of Municipal Sewage Sludge Samples by GC/MS/Computer for Polychlorinated Biphenyls and Other Chlorinated Organics," Bull. Environ. Contain. Toxicol., 22, 688-694 (1979). Erney, D. R., "A Feasibility Study of Miniature Silica Gel Columns for the Separation of Some Polychlorinated Biphenyls, DDT, and Analogs," Bull. Environ. Contam. Toxicol.. 12(6), 710-716 (1974a). Erney, D. R,, "Rapid Screening Method for Analysis of Chlorinated Pesticide and Polychlorinated Biphenyl Residues in Fish," J. Assoc. Offie. Anal. Chem., 57(3), 576-579 (1974b). Ernst, W., R. G. Schaefer, H. Goerke, and G. Eder, "Aufarbeitung von Meerestieren fiir die Bestimmung von PCB, DDT, DDE, ODD, y-HCH und HCB," Z. Anal. Chem.. 272, 358-363 (1974). Farwell, S. 0., F. A. Beland, and R. D. Geer, "Interrupted-Sweep Voltammetry for the Identification of Polychlorinated Biphenyls and Naphthalenes, Anal. Chem.. 47(6), 895-902 (May 1975). Fehringer, N. V., and J. E. Westfall, "Separation and Identification of DDT Analogs in the Presence of Polychlorinated Biphenyl Compounds by Two-Dimensional Thin-Layer Chromatography," J. Chromatogr., 57, 397-405 (1971). 95 �Finlay, D. J., F. H. Siff, and V, J. DeCarlo, "Review of PCB Levels in the Environment," EPA-560/7-76-001 (1976). Fishbein, L., "Potential Industrial Carcinogens and Mutagens," Amsterdam; Elsevier, 286-294 (1979). Fishbein, L., "Chromatographic and Biological Aspects of Polychlorinated Biphenyls," J. Chromatogr.. 68, 345-426 (1972). Frank, R., M. Holdrinet, H. E. Braun, D. P. Dodge, and G. E. Sprangler, "Residues of Organochlorine Insecticides and Polychlorinated Biphenyls in Fish from Lakes Huron and Superior, Canada, 1968-76," Pesticides Moult, J., 12(2), 69-80 (1978). Frank, R., R. L. Thomas, M. Holdrinet, A, L. W. Kemp, H. E. Braun, and R. Dawson, "Organochlorine Insecticides and PCB in the Sediments of Lake Huron (1969) and Georgian Bay and North Channel (1973)," Sci. Total Environ,. 13, 101-117 (1979). Food and Drug Administration, "Pesticide Analytical Manual," Vol. I, Sec. 251, August 1, 1977. Fries, G. F., "Polychlorinated Biphenyl Residues in Milk of Environmentally and Experimentally Contaminated Cows," Environmental Health Perspectives (April 1972). Frimmel, F., "Differentiation of Polychlorinated Biphenyls from Pesticides with Combined Gas Chromatography-Mass Spectrometry," 2. Wasser Abwasser Forscht 6(4), 113-116 (1973). ' """ ~~~ Fuller, B., J. Gordon, and M. Kornreich, "Environmental Assessment of PCBs in the Atmosphere," U.S. Environmental Protection Agency (April 1976). Gast, R., "Studies into Quantitative PCB Determination," Muench. Beitr. Abwasser, Fisch.-Flussbiol.. 30, 215-225 (1978); Chem. Abst., 90, U5924e (1979). " " General Accounting Office, "EPA Slow in Controlling PCBs," GAO/CED-82-21, 39 pages, December 30, 1981. Gesser, H. D., A. Chow, and F. C. Davis, "The Extraction and Recovery of Polychlorinated Biphenyls (PCB) Using Porous Polyurethane Foam," Anal. Letters, 4(12), 883-886 (1971). ' Giacin, J. R., and S. G. Gilbert, "Analysis of Polychlorinated Biphenyls in Packaging Materials," Technical Advances in Packaging with Flexible Barrier Materials, ASTM STP 548, 10-19 (1973). Giam, C. S., H. S, Chan, and G. S. Neff, "Rapid and Inexpensive Method for Detection of Polychlorinated Biphenyls and Phthalates in Air," Anal. Chem., 47(13), 2319-2320 (November 1975). 96 �Giara, C. S., A. R. Hanks, R. L. Richardson, W. M. Sackett, and M. K. Wong, "DDT, DDE, and Polychlorinated Biphenyls in Biota from the Gulf of Mexico and Caribbean Sea - 1971," Pestic. Monit. J., 6(3), 139-143 (1972). Giger, W., M. Reinhard, C. Schaffner, and F. Ziircher, "Analyses of Organic Constituents in Water by High-Resolution Gas Chromatography in Combination With Specific Detection and Computer-Assisted Mass Spectrometry," in Identification and Analysis of Organic Pollutants in Water, L. H. Keith (Ed.)> Chapter 26, Ann Arbor Sci., Ann Arbor, Michigan (1976). pp. 433-452. Glaser, J. A., D. L. Foerst, G. D. McKee, S. A. Quave, and W. L. Budde, "Trace Analyses for Wastewaters," Environ. Sci. Technol., 15, 1426-1435 (1981). Godefroot, M., M. Stechele, P. Sandra, and M. Verzele, "A New Method for the Quantitative Analysis of Organochlorine Pesticides and Polychlorinated Biphenyls," J. High Resol. Chrom. and Chrom. Comm., 5, 75-79 (1982). Goerlitz, D. F., and L. M. Law, "Determination of Chlorinated Insecticides in Suspended Sediment and Bottom Material," J. Assoc. Off. Anal. Chem., 57(1), 176-181 (1974). Gordon, R. J., J. Szita, and E. J. Faeder, "Determination of Polychlorinated Biphenyls in Transformer Oils by Capillary Gas Chromatography," Anal. Chem., 54, 478-481 (1982). Gorski, T., and T. Syrowatka, "Simplified Method for the Separation and Determination of Residues of Organochlorine Pesticides and Polychlorinated Biphenyls in Human Fatty Tissue," Chem. Abst.. 91., 885W (1979). Gothe, R., "Oxidation with Tetrabutylammonium Permanganate for Quantitation of DDT Residues in GLC Determination of Chlorinated Hydrocarbons," Bull, of Environ. Contain. Toxicol., U ( ) 451-455 (1974). .5, Greichus, Y. A., J. J. Worman, M. A. Pearson, and D. J. Call, "Analyses of Polychlorinated Biphenyls in Bird Tissues and Aroclor Standards with Gas Chromatography and Mass Spectrometry," Bull, of Environ. Contain. Toxicol., H 2 , 113^-120 (1974). () Griffitt, K. R., and J. C. Craun, "Gel Permeation Chromatographic System: Evaluation," J. Assoc. Off. Anal. Chem., 57(1), 168-172 (1974). An Hadorn, H., and K. Zuercher, "Contamination of Food With Polychlorinated Biphenyls (PCB) by Packaging Material and Environmental Pollution," Gordian, 73(12), 458-462 (1973). Haefelfinger, P., "Limits of the Internal Standard Technique in Chromatography," J. Chromatogr.. 218, 73-81 (1981). 97 �Haile, C. L., and E. Baladi, "Methods for Determining the Total Polychlorinated Biphenyl Emissions From Incineration and Capacitor and Transformer Filling Plants," Midwest Research Institute, Contract No. 68-02-1780, Task 2, Report No. EPA-600/4-77-048 (November 1977). Haile, C. L. , and V. Lopez-Avila, "Development of Analytical Test Procedures for the Measurement of Organic Priority Pollutants in Sludge," Revised Draft Final Report, Contract No. 68-03-2695, for Environmental Protection Agency, Environmental Monitoring and Support Laboratory, Cincinnati, Ohio (1981). Haile, C. L., J. S. Stanley, R. M. Lucas, D. K. Melroy, C. P. Nulton, and W. L. Yauger, Jr., "Pilot Study of Information of Specific Compounds from Combustion Sources," Final Report, Contract No. 68-01-5915, for Environmental Protection Agency, Office of Pesticides and Toxic Substances, Field Studies Branch, Washington, D.C. (1982). Hanai, T., and H. F. Walton, "Chromatography of Chlorinated Biphenyls on an Ion-Exchange Resin," Anal. Chem.. 4_9(6), 764-766 (May 1977). Hannan, E. J., and D. D. Bills, "Separation of Aldrin from Aroclor 1254," Bull. Environ. Contam. Toxicol., 8(6), 327-328 (1972). Hannan, E. J., D. D. Bills, and J. L. Herring, "Analysis of Polychlorinated Biphenyls by Gas Chromatography and Ultraviolet Irradiation," J. Agr. Food Chem., 21(1), 87-90 (1973). Haque, R., and D. Schmedding, "A Method of Measuring the Water Solubility of Hydrophobic Chemicals: Solubility of Five Polychlorinated Biphenyls," Bull. Environ. Contam. Toxicol., U(l), 13-18 (1975). Haque, R., D. W. Schmedding, and V. H. Freed, "Aqueous Solubility, Adsorption, and Vapor Behavior of Polychlorinated Biphenyl Aroclor 1254," Environ. Sci. Tech., 8(2), 139-142 (February 1974). Haque, R., and D. Schmedding, "Studies on the Adsorption of Selected Polychlorinated Biphenyl Isomers on Several Surfaces," J. Environ. Sci. Health, BU(2), 129-137 (1976). Harris, R. W., C. G. Grainger, and W. J. Mitchell, "Validation of a Method for Measuring Polychlorinated Biphenyls in Natural Gas Pipelines," EPA 600/4-81-048; NTIS PB82-207556 (1981). Harvey, J., Jr., G. Zweig, R. Cannizzaro, H. Dishburger, and J. Sherma (Ed.), "Pesticide Analytical Methodology," ACS Symposium Series 136, Amer. Chem. Soc., (1980). Hass, J. R., and M. D. Friesen, "Qualitative and Quantitative Methods for Dioxin Analysis," Ann. N.Y. Acad. Sci., 320, 28-42 (1979). 98 �Mass, J. R., M. D. Friesen, and M. K. Hoffman, "Recent Mass Spectrometric Techniques for the Analysis of Environmental Contaminants," Chapter 11 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, J. D. McKinney (Ed.)> Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981). pp. 219-243. Hattori, Y., Y. Kuge, and M. Nakamoto, "The Correlation Between the ElectronCapture Detector Response and the Chemical Structure for Polychlorinated Biphenyls," Bull. Chem. Soc. Jpn., 54(9), 2807-2810 (1981). Hattula, M. L., and 0. Karlog, "Adsorption of Polychlorinated Biphenyls (PCB) on Glass Surfaces," Dan. Tidsskr. Farm., 45(7), 259-261 (1971). Hattula, M. L., "Simultaneous Clean Up of Fish Fat Containing Low Levels of Residues and Separation of PCB From Chlorinated Pesticides by Thin-Layer Chromatography," Bull. Environ. Contam. Toxicol., 12(3), 331-337 (1974). Hattula, M. L., "Some Aspects of the Recovery of Chlorinated Residues (DDTType Compounds and PCB) From Fish Tissue by Using Different Extraction Methods," Bull. Environ. Contam. Toxicol., 12(3), 301-307 (1974). Heeg, F. J., R. Zinburg, H. J. Neu, and K. Ballschmiter, "Retention Index Calculation Based on a Non-Linear Relationship Between Net Retention Time and Carbon Number of n-Alkanes," Chromatographia, 12(7), 451-458 (1979). Heller, S. R., J. M. McGuire, and W. L. Budde, "Trace Organics by GC/MS," Environ. Sci. Technol., 9(3), 210-213 (March 1975). Heller, S. R., and G. W. A. Milne, "1978 Indices to EPA/NIH Mass Spectral Data Base," U.S. G.P.O., Washington, D.C. (1978). p. 25409. Henderson, J. E., and W. H. Glaze, "GC/MS Analysis of XAD-2 Extracts of Superchlorinated Septage," Water Res., 1.6, 211-218 (1982). Herzel, F., "Polychlorinated Biphenyls and Their Discrimination from Organochlorine Insecticides," Vom Wasser, 38, 71-80 (1971). Hesselberg, R. J., and J. L. Johnson, "Column Extraction of Pesticides from Fish, Fish Food and Mud," Bull. Environ. Contam. Toxicol., 7, 115-120 (1972). Higuchi, K. (Ed.), PCB Poisoning and Pollution, Academic Press, New York (1976). Hirota, K., Y. Takata, Y. Arikawa, T. Yoshihara, H. Akimori, and S, Ganno, "Preparation and Application of Glass Capillary Columns," Bunseki Kagaku, 23(10), 1194-1201 (1974). Hirwe, S. N., R. E. Borchard, L. G. Hansen, and R. L. Metcalf, "Gas-Liquid Chromatography - Mass Spectrometric Characterization of Aroclor 1242 and 1254 Components," Bull. Environ. Contam. Toxicol., 12(2), 138-144 (1974). 99 �Hodges, K. L,, T. A. Robinson, C. J. Hensler, R. S. Kaley, H. Gaspers, F. Ewald, and A. R, Pittaway, "Exposure Studies Relating to Industrial Processes Containing Incidental PCB," in Recent Advances in Exposure, Health and Environmental Effects Studies of PCBs, Proceedings of Symposium, Bethesda, Maryland, May 12-13, 1982 (in press). Hofstader, R. A., C. A. Bache, and D. J. Lisk, "Interference in the ElectronCapture Technique for Determination of Polychlorinated Biphenyls by SulfurContaining Compounds in Petroleum Products," Bull. Environ. Contam. Toxicol., 11(2), 136-138 (1974). Holden, A. V., "The Analysis of Fish for Organochlorine Residues," International Symposium on Identification and Measurement of Environmental Pollutants, Ottawa, Ontario, Canada (June 1971). Holdrinet, M. V. H., "Preliminary Results of an Interlaboratory PCB Check Sample Program," Environ. Qual. Safety Suppl., 3 (Pesticides), 51-56 (1975). Holdrinet, M. V. H., "Confirmation of Mirex and cis- and trans-Chlordane in the Presence of Other Organochlorine Insecticides and Polychlorinated Biphenyls," Bull. Environ. Contam. Toxicol., 21, 46-52 (1979). Holmes, D. C., and M. Wallen, "A Simple Differentiation of Polychlorobiphenyl,s from Chlorinated Naphthalenes," J. Chroma tog r., 71_, 562-563 (1972). Hopper, M. L., and D. D. Hughes, "An Improved GPC System for Pesticides in Fats," Laboratory Information Bulletin, FSB/EDRO, p. 50-55 (July 1976). Huckins, J. N. , D. L. Stalling, and J. D. Petty, "Carbon-Foam Chromatographic Separation of Non-o,o'-Chlorine Substituted PCBs from Aroclor Mixtures," J. Assoc. Off. Anal. Chem., 63(4), 750-755 (1980). Huckins, J. N., J. E. Swanson, and D. L. Stalling, "Perchlorination of Polychlorinated Biphenyls," J. Assoc. Off. Anal. Chem., 57(2), 416-417 (1974). Huckins, J. N., D. L. Stalling, and J. L. Johnson, "Silicic Acid Chromatographic Separation of Polychlorinated Biphenyls and Pesticides: Some Contaminants and Limitations," J. Assoc. Off. Anal. Chem., 59_(5) , 975-981 (1976), Hutzinger, 0., S. Safe, and V. Zitko, "Preparation, Gas Chromatographic Behavior, and Spectroscopic Properties of Hydroxylated Chlorobiphenyls," J. Assoc. Off. Anal. Chem., 5_7(5), 1061-1067 (1974). Hutzinger, 0., W. D. J. Jamieson, S. S. Safe, and V. Z. Zitko, "Exhaustive Chlorination as a Technique in the Analysis of Aromatic Hydrocarbons," J. Assoc. Off. Anal. Chem., 56(4), 982-986 (1973). Hutzinger, 0., R. A. Heacock, and S. Safe, "Thin-Layer Chromatography and Colour Reactions of Some Hydroxylated Chlorobiphenyls," J. Chromatogr., 97, 233-247 (1974). 100 �Hutzinger, 0., S. Safe, and V. Zitko, "Analysis of Chlorinated Aromatic Hydrocarbons by Exhaustive Chlorination: Qualitative and Structural Aspects of the Perchloro-Derivatives of Biphenyl, Naphthalene, Terphenyl, Dibenzofuran, Dibenzodioxin and DDE," Intern. J. Environ. Anal. Chem., 2, 95-106 (1972). Hutzinger, 0., S. Safe, and V. Zitko, The Chemistry of PCB's, CRC Press, Cleveland, Ohio, 1974, 269 pp. lida, Y., and T. Kashiwagi, "Analysis of Polychlorinated Biphenyls by Gas Chromatography-Chemical lonization Mass Spectrometry," Seikei Daigaku Kogakubu Kogaku Hokoku. J.9, 1461-1462 (1975). Ismail, R. J., and F. L. Bonner, "New, Improved Thin Layer Chromatography for Polychlorinated Biphenyls, Toxaphene, and Chlordane Components," J. Assoc. Off. Anal. Chem., 57(5), 1026-1032 (1974). Jennings, W., "Gas Chromatography with Glass Capillary Columns," Academic Press, New York, San Francisco, London, p. 139-140 (1978). Jensen, S., Ambio, 1, 123 (1972). Jensen, S., L. Renberg, and L. Reutergardh, "Residue Analysis of Sediment and Sewage Sludge for Organochlorines in the Presence of Elemental Sulfur," Anal. Chem.. 49(2), 316-318 (1977). Jensen, S., L. Renberg, and R. Vaz, "Problems in the Quantification of PCB in Biological Material," Committee Proceeding PCB 2nd Conference, Wenner-Gren Center, 1972, 7-14 (1973). Jensen, S., L. Renberg, and R. Vaz, "Methods for Analysis of DDT and PCB in Environmental Samples Using Chromatographic Methods," AD Fisheries Technical Paper, Rome, FIRI/T137 (1975). pp. 229-236. Jensen, S., and G. Sundstrom, "Structures and Levels of Most Chlorobiphenyls in Two Technical PCB Products and in Human Adipose Tissue," Ambio, 3(2), 7076 (1974). Jung, J. D:, "A Review of Methods for PCB Analysis of Transformer Oil," Proc. Electr./Electron. Insul. Conf., 15th, 231-236 (1981); Chem. Abst., 97, I6335b (1981). Kaminsky, L. S., and M. J. Fasco, "High-Performance Liquid Chromatography of Polychlorinated Biphenyls," J. Chromatogr., 155, 363-370 (1978). Kamops, L. R., W. J. Trotter, S. J. Young, A. C. Smith, J. A. G. Roach, and S. W. Page, "Separation and Quantitation of 3,3',4,4'-Tetrachlorobiphenyl and 3,3',4,4',5,5'-Hexachlorobiphenyl in Aroclors Using Florisil Column Chromatography and Gas-Liquid Chromatography," Bull. Environ. Contam. Toxicol., 23, 51-56 (1979). 101 �Kan, T., K. Kamata, T. Ueta, R. Yamazoe, and T. Totani, "Fluorescence Reactions of Organohalogen Compounds. I, Fluorometry of Polychlorinated Biphenyls (PCB) with Diphenylamine on Thin-Layer Chroraatograms," Tokyo Toritsu Eisei Kenkyusho Kenky Nempo, 24, 137-145 (1973); Chem. Abst., 80, 115771w (1974). Kapila, S. and W. A. Aue, "Chemical Degradation of Single, Gas Chromatog^aphic Peaks for Confirmation of Pesticide Residue Identity," J. Chromatogr. Sci., 15, 569-572 (1977). Karasek, F. W., "Plasma Chromatography of the Polychlorinated Biphenyls," Anal. Chem., 43(14), 1982-1986 (1971). Kashimoto, T., H. Miyata, S. Kunita, T.-C. Tung, S.-T. Hsu, K.-J. Chang, S.-Y Tang, G. Ohi, J. Nakagawa, and S.-I. Yamamoto, "Role of Polychlorinated Dibenzofuran in Yusho (PCB Poisoning)," Archives of Environmental Health, 36(6), 321-326 (1981). Kawabata, J., "Simple Method for the Determination of PCBs [Polychlorinated Biphenyls] by a Combination of Thin-Layer Chromatography and UV Absorption," Kogai To Taisaku. 10(10), 1112-1116 (1974); Chem. Abst., 83, 20l652b (1975). Kawaguchi, T., S. Hiratsuka, T. Yoneshima, and Y. Nagao, "A Method of Analysis for Polychlorobiphenyl Gas Chromatograms," Tokyo-Toritsu Kogyo Gijutsu Senta Kenkyu Hokoku, 5, 57-60 (1975); Chem. Abst., 86, 83287y (1977). Karppanen, E., and L. Kolho, "The Concentration of PCB in Human Blood and Adipose Tissue in Three Different Re$earch Groups," PCB Conference II, 1972, Stockholm, Natl. Swedish Environ. Protection Board, Publication No. 1973:4E, 124-8 (1973). Kerkhoff, M. A. T, A. De Vries, R. C. C. Wegman, and A. W. M. Hofstee, "Analysis of PCBs in Sediments by Glass Capillary Gas Chromatography," Chemosphere, n ( ) 165-174 (1982). .2, Khan, M. A., A. F. Novak, and R. M. Rao, "Comparative Evaluation of Polychlorinated Biphenyl Solutions," Bull. Environ. Contain. Toxicol., 16(3), 360' 367 (1976). Kilikidis, S. D., J. E. Psomas, A. P. Kamarianos, and A. G. Panetsos, "Monitoring of DDT, PCBs, and Other Organochlorine Compounds in Marine Organisms from the North Aegean Sea," Bull. Environ. Contam. Toxicol., 26, 496-501 (1981) Kimbrough, R. D., (Ed.) Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products, Elsevier/North-Holland Biomedical Press, New York (1980). Kimbrough, R., J. Buckley, L. Fishbein, G. Flamm, L. Kasza, W. Marcus, S. Shibko, and R. Teske, "Animal Toxicology," Env. Health Persp., 24, 173-184 (1978). Kirshen, N. A., "PCBs in Transformer Fluids," Varian Instrum, Appl., 15(2), 1-10 (1981); Chem. Abst., 94(26), 219095J (1981). 102 �Kirshen, N. A., "PCBs in Transformer Fluids," Amer. Lab.. 65-69 (1981). Klimisch, H. M., and D. N. Ingebrigtson, "Determination of Polychlorinated Biphenyls in Silicone Fluids," Anal. Chem., 52, 1675-1678 (1980). Koch, R., "Circular Thin-Layer Chromatography as a Rapid Method for a Qual* itative Detection of Organochlorine Compounds," Acta Hydrochim. Hydrobiol., 7(3), 355-356 (1979); Chem. Abst.. 91, 101574z (1979). Kodama, H., N. Kawamura, and H. Ota, "A Simple Method for the Determination of PCB in Human Milk and Blood Using an Improved Essential Oil Distillator," Nippon Eiseigaku Zasshi, 31(6), 644-651 (1977); Chem. Abst., 87, 96853a (1977). Koeniger, M., P. Wallnoeer, and G. Engelhardt, "Residue Determination of Polychlorinated Biphenysl (PCBs) in the Presence of Chlorinated Hydrocarbon Pesticides in Eggs of Wild Birds," Nachrichtenbl. Dtsch. Pflanzenschutzdienstes, 27(9), 131-135 (1975); Chem. Abst., 84, 85033v (1976). Kohli, K. K., P. W. Albro, and J. D. McKinney, "Radioisotope Dilution Assay (RIDA) for the Estimation of Polychlorinated Biphenyls (PCBs)," J. Anal. Toxicol., 3, 125-128 (1979). Kovats, von E., "Gas-Chromatographische Charakterisierung Organischer Verbindungen. Teil 1: Retentionsindices Aliphatischer Halogenide, Alkohole, Aldehyde und Ketone," Helv. Chim. Acta, 41, 1915-1932 (1958). Kraul, I., and 0. Karlog, "Persistent Organochlorinated Compounds in Human Organs Collected in Denmark 1972-73," Acta Pharmacol. et Toxicol., 38, 38-48 (1976). ~ Kreiss, K., M. M. Zack, R. D. Kimbrough, L. L. Needham, A. L. Smrek, and B. T. Jones, "Association of Blood Pressure and Polychlorinated Biphenyl Levels," J. Amer. Med. Assoc.. 245(24), 2505-2509 (1981). Krull, I. S., "Recent Advances in PCB Analysis," in Residue Reviews, F. A. Gunther and J. D. Gunther (Eds.), 66, New York (1977). Krupcik, J., J. Kriz, D. Prusova, P. Suchanek, and Z. Cervenka, "Analysis of Polychlorinated Biphenyls by High-Performance Liquid Chromatography and Capillary Gas-Liquid Chromatography," J. Chromatogr., 142, 797-807 (1977). Krupcik, J., P. A. Leclercq, A. Simova, P. Suchanek, M. Collak, and J. Hrivnak, "Possibilities and Limitations of Capillary Gas Chromatography and Mass Spectrometry in the Analysis of Polychlorinated Biphenyls," J. Chromatogr., 119, 217-283 (1976). Krupcik, J., P. A. Leclercq, J. Garaj, and A. Simova, "Analysis of Alkylated Mixtures of Polychlorinated Biphenyls by Capillary Gas Chromatography-Mass Spectrometry," J. Chromatogr., 191, 207-220 (1980). Krupcik, J., J. Mocak, A. Simova, J. Garaj, and G. Guiochon, "Optimization of Experimental Conditions for the Analysis of Complex Mixtures by Gas Chromatography," J. Chromatogr.. 238. 1-12 (1982). 103 �Kuehl, D. W. , and E. N. Leonard, "Isolation of Xenobiotic Chemicals from Tissue Samples by Gel Permeation Chromatography," Anal. Chem., 50(1)> 182-185 (1978). Kuehl, D. W., E. N. Leonard, K. J. Welch, and G. D. Veith, "Identification of Hazardous Organic Chemicals in Fish from the Ashtabula River, Ohio, and Wabash River, Indiana," J. Assoc. Off. Anal. Chem., 63(6), 1238-1244 (1980). Kuehl, D. W., M. J. Whitaker, and R. C. Dougherty, "Micromethods for Toxic Residue Screening by Negative Chemical lonization Mass Spectrometry," Anal. Chem., 52, 935-940 (1980). Kurastune, M., and Y. Masuda, Env. Health Persp., 1, 61 (1972). Kuroki, H. , and Y. Masuda, "Determination of Polychlorinated Dibenzofuran Isomers Retained in Patients with Yusho," Chemosphere, 10, Ill-Ill (1978) Kutz, F. W., S. C. Strassman, and J. F. Sperling, "Survey of Selected Organochlorine Pesticides in the General Population of the United States: Fiscal Years 1970-1975," Annals New York Academy of Sciences, 320, 60-68 (1979). Kveseth, N. J., and E. M. Brevik, "Column Chromatographic Method for Cleaning Up Extracts from Biological Material and Simultaneous Separation of PCBs and DDE," Bull. Environ. Contain. Toxicol. , 2J., 213-218 (1979). Laitem, L., and P. Caspar, "A T.L.C. Procedure for Identification of DDT and Its Metabolites in Presence of PCB," Bull. Environ. Contain. Toxicol., 264-265 (1978). Lamparski, L. L., and T. J. Nestrick, "Micro Chlorination Procedure for Synthesis of Higher Chlorinated Dibenzo-p_-dioxins from [13C]-2,3,7,8-Tetrachlorodibenzo-£-dioxin," Anal. Chem., 54, 402-406 (1982). Lao, R. C., R. S. Thomas, and J. L. Monkman, "Application of Computerized Gas Chromatography-Mass Spectrometry to the Analysis of Polychlorinated Biphenyls," Dynamic Mass Spectrom., 4, 107-131 (1976). Lawrence, J., and H. M. Tosine^ "Adsorption of Polychlorinated Biphenyls from Aqueous Solutions and Sewage," Environ. Sci. Techno1., 1.0(4), 381-383 (1976). Lawrence, J. F., and D. Turton, "High-Performance Liquid Chromatographic Data for 166 Pesticides," J. Chromatogr., 159, 207-266 (1978). Leib, A. J., and D. D. Bills, "Influence of Storage Temperature of Florisil on Analysis of Polychlorinated Biphenyls," Bull. Environ. Contam. Toxicol., 12(3), 328-330 (1974). Leoni, V., "The Separation of Fifty Pesticides and Related Compounds and Polychlorobiphenyls into Four Groups by Silica Gel Microcolumn Chromatography," J. Chromatogr., 62, 63-71 (1971). 104 �Liddle, J. A., "Method Standardization and Interlaboratory Compatibility of Serum Polychlorinated Biphenyl Analysis," Abstract, 18Ath Annual Chemical Society National Meeting, September 12-17, 1982. Lincer, J. L., "Polychlorinated Biphenyls: Their Potential Interference with Pesticide Residue Analysis and Present Analytical Status," Progr. Anal. Chem., 5, 109-131 (1973). Luke, M. A., J. E. Froberg, and H. P. Masumoto, J. Assoc. Offjc. Anal. Chem., 58, 1020 (1975). Lucas, R. M., M. D. Erickson, P. V. Piserchia, and S. R. Williams, "PCS Residue Levels in Human Adipose Tissue, A Statistical Evaluation by Racial Grouping," Environmental Protection Agency, Contract No. 68-01-5848, Report No. EPA-650/13-79-015 (November 1980). Luckas, B., H. Pscheidl, and D. Haberland, "Determination of Chlorinated Pesticides and Polychlorinated Biphenyls by Derivatization Gas Chromatography," J. Chromatogr., 147, 41-46 (1978). Lunde, G., and E. B. Ofstad, "Determination of Fat-Soluble Chlorinated Compounds in Fish," 2. Anal. Chem., 282, 395-399 (1976). Luster, M. I., P. W. Albro, G, Clark, K. Chae, S. K. Chaudhary, L. D. Lawson, J. T. Corbett, and J. D. McKinney, "Production and Characterization of Antisera Specific for Chlorinated Biphenyl Species: Initiation of a Radioimmunoassay for Aroclors," Toxicol. Appl. Pharmacol., 50, 147-155 (1979). Luster, M. I., P. W. Albro, K. Chae, S. K. Chaudhary, and J. D. McKinney, "Development of Radioimmunoassays for Chlorinated Aromatic Hydrocarbons," Chapter 14 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, J. D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981). pp. 279-297. MacDougall, et al., "Guidelines for Data Acquisition and Data Quality Evaluation in Environmental Chemistry," Anal. Chem., 52, 2242-2249 (1980). MacLeod, K. E., "Sources of Emissions of Polychlorinated Biphenyls into the Ambient Atmosphere and Indoor Air," U.S. Environmental Protection Agency, Report No. EPA-600/4-79-022 (March 1979). MacNeil, J. D., S. Safe, and 0. Hutzinger, "The Ultraviolet Absorption Spectra of Some Chorinated Biphenyls," Bull. Environ. Contam. Toxicol., 15(1), 66-77 (1976). Manri, T., Y. Shinbori, S. Murakawa, and Y. Mochizuki, "Determination of Chlorobiphenyl in Rat Adipose Tissues by Neutron Activation Analysis," Radioisotopes. 20(8), 371-375 (1971); Chem. Abst., 76, 11738s (1972). Margeson, J. H., "Methodology for Measurement of Polychlorinated Biphenyls in Ambient Air and Stationary Sources - A Review," U.S. Environmental Protection Agency, Report No. EPA-600/4-77-021 (April 1977). 106 �Martelli, G. P., M. G. Castelli, and R. Fanelli, "Analytical Response of Single Polychlorinated Biphenyls Analysed by Selected Ion Monitoring," Bioroed. Mass Spectrom.. 8(8), 347-350 (1981). Mass Spectrometry Data Centre, in collaboration with ICI Ltd., Eight Peak Index of Mass Spectra, 2nd ed., Vol. 1-3, AWRE, Aldermaston, Reading RG 7/4PR, United Kingdom (1970). Masuda, Y., and H. Shiokawa, "Experimental File. 42. Chromatography. Three Experiments," Kagaku No Jikken. 23(7), 669-671 (1972); Chem. Abst.. 77, 87143k (1972). Masui, T., M. Yamamoto, and S. Kimura, "Problems in Polychlorinated Biphenyl Analysis," Shizuoka-ken Eisei Kenkyusho Hokoku, 19, 59-64 (1976); Chem. Abst., 90, 109442e (1979). Masumoto, H. T., "Total Polychlorinated Biphenyl Quantitation by Reduction," EDRO SARAP Res. Tech. Rep.. I, Paper 20-73, 30 pp. (1976); Chem. Abst., 88, 73094k (1978). Masumoto, H. T., "Study of the Silicic Acid Procedure of Armour and Burke for the Separation of Polychlorinated Biphenyls from DDT and Its Analogs," J. Assoc. Offie. Anal. Chem., 55(5), 1092-1100 (1972). Matthews, H., G. Fries, A. Gardner, L. Garthoff, J. Goldstein, Y. Ku, and J. Moore, "Metabolism and Biochemical Toxicity of PCBs and PBBs," Env. Health Persp.. 24, 147-155 (1978). Mattsson, P. E., and S. Nygren, "Gas Chromatographic Determination of Polychlorinated Biphenyls and Some Chlorinated Pesticides in Sewage Sludge Using a Glass Capillary Column," J. Chromatogr., 124, 265-275 (1976). McKinney, J. D., "Environmental Health Chemistry--Definitions and Interrelationships: A Case in Point," Chapter 1 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Health Hazards, J. D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981). pp. 3-15. McLafferty, F. W., R. Knutti, R. Venkataraghavan, P. J. Arpino, and B. G. Dawkins, "Continuous Mass Spectrometric Monitoring of a Liquid Chromatograph with Subnanogram Sensitivity Using an On-Line Computer," Anal. Chem., 47(9), 1503-1505 (1975). McLeod, H. A., and P. J. Wales, "A Low Temperature Cleanup Procedure for Pesticides and Their Metabolites in Biological Samples," J. Agr. Food Chem., 20(3), 624-627 (1972). Mes, J., D. S. Campbell, R. N. Robinson, and D. J. A. Davies, "Polychlorinated Biphenyl and Organochlorine Pesticide Residues in Adipose Tissue of Canadians," Bull. Environ. Contain. Toxicol., V7(2), 196-203 (1977). 107 �Mes, J., and D. J. Davies, "Variation in the Polychlorinated Biphenyl and Organochlorine Pesticide Residues During Human Breastfeeding and Its Diurnal Pattern," Chemosphere. 9, 699-706 (1978). Mes, J., and D, J. Davies, "Presence of Polychlorinated Biphenyl and Organochlorine Pesticide Residues and the Absence of Polychlorinated Terphenyls in Canadian Human Milk Samples," Bull. Environ. Contain. Toxicol., 2_1, 381-387 (1979). Mes, J., and D. S. Campbell, "Extraction Efficiency of Polychlorinated Biphenyl, Organochlorine Pesticides, and Phthalate Esters from Human Adipose Tissue," Bull. Environ. Contain. Toxicol., 16(1), 53-60 (1976). Mes, J., D. Davies, and J. Truelove, "The Determination of Polychlorinated Biphenyl in Small Samples of Monkey Milk and Tissues. I.," Intern. J. Environ. Anal. Chem., 8, 89-98 (1980). Mestres, R., C. Chevallier, C. Espinoza, and R. Cornet, "Application of Coupled Gas Chromatography-Mass Spectrometry in Methods for the Study and Determination of Pesticide Residues and Organic Micropollutants in Environmental and Food Materials," Ann. Falsif. Expert. Chim., 70(751), 177-188 (1977); Chem. Abst., 8_7, 63535v (1977). Mestres, R., C. Chevallier, and M. Pagnon, "Mass Spectrometric Confirmation of the Identity and Content of Pesticide and Chlorinated Biphenyl Residues in the Sediments of the Mediterranean Sea," Trav. Soc. Pharm. Montpellier, 36(2), 151-160 (1976); Chem. Abst., 85, 73422d (1976). Mestres, R., and C. Chevallier, "XXXII. Method of Research and Determination of PCB and Phthalate Pesticide Residues in Water," Ann, Falsif.Expert. Chipi., 70(750), 101-111 (1977); Chem. Abst., 87, 79446c (1977). Metcalf, R. L., "Screening Compounds for Early Warnings About Environmental Pollution," in Trace Substances in Environmental Health - VIII, D. D. Hemphill (Ed.), Proceedings of University of Missouri's 8th Annual Conference on Trac$ Substances in Environmental Health," Columbia, Missouri, June 11-13, 1974. pp. 213-217. Middleditch, B. S., S. R. Missler, and H. B. Hines, "Mass Spectrometry of Priority Pollutants, Plenum Press, New York (1981). Mieure, J. P., 0. Hicks, R. G. Kaley, and V. W. Saeger, "Characterization of Polychlorinated Biphenyls," in National Conference on Polychlorinated Biphenyls (November 19-21, 1975, Chicago, Illinois), F. A. Ayer (Ed.), prepared for U.S. Environmental Protection Agency, Contract No. 68-01-2928, Report No. EPA-560/6-75-004 (March 1976). Miki, T., "PCBs [Polychlorinated Biphenyls] and Organochlorine Pesticide Residues in Sugars," Seito Gijutsu Kenkyukaishi, 25, 32-38 (1975); Chem. Abst., 83, 149456z (1975). 108 �Millar, J. D., R. E. Thomas, and D. E. Johnson, "Determination of Pesticides and PCBs in Industrial and Municipal Wastewaters," U.S. Environmental Protection Agency, Contract No. 68-03-2606, Report No. EPA-600/4-82-023 (March 1982). Mitchum, R. K., W. A. Korfmacher, and G. F. Moler, "Capillary Gas Chromatography/ Atmospheric Pressure Negative Chemical lonization Mass Spectrometry of the 22 Isomeric Tetrachlorodibenzo-pj-dioxins," Anal. Chem., 54, 719-722 (1982). Mizutani, T., and M. Matsumoto, "Separation and Determination of Polychlorinated Biphenyls and DDE by Oxidation in a Solid Support," Eisei Kagaku, 19(5), 287-292 (1973); Chem. Abst., 80, Il6981b (1974). Mizutani, T., and M. Matsumoto, "Determination of Polychlorinated Biphenyls by an Exhaustive Chlorination Method," Shokuhin Eiseigaku Zasshi, 13_(5), 398404 (1972); Chem. Abst.. 78, 66686v (1973). Modi, G., M. Piccinini, and S. Bonciani, "Study of Pesticide Contamination in Some Tuscany Wines," Boll. Chim. Unione Ital. Lab. Prov., 2(3), 133-147 (1976); Chem. Abst.. 85, 141275k (1976). Moseley, M. A., and E. D. Pellizzari, "Development and Evaluation of Wall Coated Open-Tubular Columns for GC Analysis of Individual Polychlorinated Biphenyl Isomers," J. of HRC & CC, 5, 404-412 (1982). Moser, J. H., and K. R. Huibregtse, "Handbook for Sampling and Sample Preservation of Water and Wastewater," EPA-600/4-76-049 (1976). Moye, H. A., "High Speed Liquid Chromatography of Pesticides," J. Chromatogr. Sci., 13, 268-278 (1975). Mulhern, B. M., "An Improved Method for the Separation and Removal of Organochlorine Insecticides from Thin-layer Plates," J. Chromatogr., 34, 556-558 (1968). Mulhern, B. M., E. Cromartie, W. L. Reichel, and . A. Belisle, "Semiquantitative Determination of Polychlorinated Biphenyls in Tissue Samples by Thin Layer Chromatography," J. Assoc. Offic. Anal. Chem., 54(3), 548-550 (1971). Mullin, M. D., and J. C. Filkins, "Analysis of Polychlorinated Biphenyls by Glass Capillary and Packed-Column Chromatography," Chapter 11 in Advances in the Identification and Analysis of Organic Pollutants in Water, L. H. Keith (Ed.), Vol. 1, Ann Arbor Science Publishers, Inc., Ann Arbor, MI (1981), pp. 187-196. Mullin, M., G. Sawka, L. Safe, S. McCrindle, and S. Safe, "Synthesis of the Octa- and Nonachlorobiphenyl Isomers and Congeners and Their Quantitation in Commercial Polychlorinated Biphenyls and Identification in Human Breast Milk," J. Anal. Toxicol.. 5, 138-142 (1981). Murphy, P. G., "Sulfuric Acid for the Cleanup of Animal Tissues for Analysis of Acid-Stable Chlorinated Hydrocarbon Residues," J. Assoc. Offic. Anal. Chem., 55(6), 1360-1362 (1972); Chem. Abst., 78, 25095h (1973). 109 �Musial, C. J., and 0. Hutzinger, "Presence of PCB, DDE and DDT in Human Milk in the Provinces of New Brunswick and Nova Scotia, Canada," Bull. Environ. Contam. Toxicol.. 12(3), 258-267 (1974). Musty, P. R., and G. Nickless, "The Extraction and Recovery of Chlorinated Insecticides and Polychlorinated Biphenyls from Water Using Porous Polyurethane Foams," J. Chromatogr.. 100(1), 83-94 (1974). Musty, P. R., and G. Nickless, "Extractants for Organochlorine Insecticides and Polychlorinated Biphenyls from Water," J. Chromatogr., 120, 369-378 (1976). Musty, P. R., and G. Nickless, "Use of Amberlite XAD-4 for Extraction and Recovery of Chlorinated Insecticides and Polychlorinated Biphenyls from Water," J. Chromatogr., 8£, 185-190 (1974). Nakagawa, H., M. Kayama, and E. Ariyoshi, "Measurements of Polychlorinated Biphenyl (PCB) in Water, Sediments, and Organisms of Two Rivers in Hiroshima Prefecture," J. Fac. Fish Anim. Husb. Hiroshima Univ., 14(2), 253-259 (1975); Public Health Abst., 62(6), 34268 (1976). Nakagawa, J., M. Morita, and Y. Higuchi, "Analysis of Polychlorinated Dibenzofurans by Gas Chromatography Using Glass Capillary Columns," Tokyo-toritsu Eisei Kenkyusho Kenkyu Nempo, (30-1), 260-264 (1979); Chem. Abst., 93, 160694c (1980). Nakamura, A., and T. Kashimoto, "Studies on a Calculation Method for Polychlorinated Biphenyl (PCB) Isomers. Analysis Using an Apiezon L. Column," Shokuhin Eiseigaku Zasshi, 1J3(1), 1-12 (1977); Chem. Abst.. 87, 63524r (1977). Nakatsu, K., J. F. Brien, H. Taub, W. J. Raqz, and G. S. Marks, "Gram Quantity Synthesis and Chromatographic Assessment of 3,3',4,4'-Tetrachlorobiphenyl," J. Chromatogr., 239, 97-106 (1982). National Institute for Occupational Safety and Health, NIOSH Manual of Analytical Methods, Second Edition, Part I, NIOSH Monitoring Methods, Vol. 1, "Polychlorinated Biphenyls (PCB) in Air, Analytical Method P&CAM 244," U.S. Department of Health, Education, and Welfare, Cincinnati, Ohio, April 1977a. National Institute for Occupational Safety and Health, NIOSH Manual of Analytical Methods, Second Edition, Part I, NIOSH Monitoring Methods, Vol. 1, "Polychlorinated Biphenyls (PCB) in Air, Analytical Method P&CAM 253," U.S, Department of Health, Education, and Welfare, Cincinnati, Ohio, April 1977b. National Research Council, Polychlorinated Biphenyls, National Academy of Sciences, Washington, D.C. (1979). National Technical Information Service, "Polychlorinated Biphenyls in the Environment, 1980-September, 1981," PB82-803867, 70 pp. (1982). Needham, L. L., A. L. Smrek, S. L. Head, V. W. Burse, and J. A. kiddle, "Column Chromatography Separation of Polychlorinated Biphenyls from Dichlorodiphenyltrichloroethane and Metabolites," Anal. Chem., 52, 2227-2229 (1980). 110 �Neu, H. J., M. Zell, and K. Ballschmiter, "Identifizierung von Einzelkomponenten in komplexen Gemischen durch Retentions-Index-Vergleich nach Capillar-GasChromatographie mit Elektroneneinfangdetektor (ECD)," Fresenius Z. Anal. Chem., 293, 193-200 (1978). Neu, H. J., and R. Zinburg, "Are We Using the Full Resolving Power of Capillary GC?" J. High Resol. Chromatogr. & Chromatogr. Commun., 2, 395-399 (1979). Norstrom, R. J., R. W. Risebrough, and D. J. Cartwright, "Elimination of Chlorinated Dibenzofurans Associated with Polychlorinated Biphenyls Fed to Mallards (Anas platyrhynchos)," Toxicol. Appl. Pharmacol., 37, 217-228 (1976). Nose, K., "Decachlorination of Polychlorinated Biphenyls and Its Application to Soil and Rice Analyses," Nippon Nogei Kagaku Kaishi, 46(12), 679-681 (1972); Chem. Abst., 78, I46633r (1973). Nose, K., "Assay of Residual PCB [Polychlorinated Biphenyl] in Water, Soil, and Agricultural Products," Noyaku Seisan Gijutsu, 30, 29-36 (1973); Chem. Abst., 7i, 101513f (1973). Nose, K., "Determination of PCB (Polychlorinated Biphenyls)," Yukagaku, 25(10), 638-644 (1976); Chem. Abst.. 86, 26565p (1977). Nowicki, H. G., "Application of Azulene as Visual Aid to Monitor Column Chromatographic Fractionation of Samples for Pesticide and Polychlorinated Biphenyl Determination," J. Assoc. Offic. Anal. Chem., 64(1), 16-18 (1981). Nygren, S., and P. E. Mattsson, "Flow Programming in Glass Capillary ColumnElectron Capture Gas Chromatograpy by Using the Valve in the Splitter Line," J. Chromatogr., 123, 101-108 (1976). Oehme, M., and H. Stray, "Quantitative Determination of Ultra-Traces of Chlorinated Compounds in High-Volume Air Samples from the Arctic Using Polyurethane Foam as Collection Medium," Fresenius' Z. Anal. Chem., 311, 665-673 (1982); Chem. Abst.. 17, 60081q (1982). Ofstad, E. B., G. Lunde, and K. Martinsen, "Chlorinated Aromatic Hydrocarbons in Fish from an Area Polluted by Industrial Effluents," Sci. Total Environ., 1£, 219-230 (1978). Ogata, J. N., J. D. Okun, J. W. Hylin, and A. Bevenue, "Gas Chromatographic Method for the Analysis of Polychlorinated Biphenyls in Transformer Oil," J. Chromatogr.. 189, 425-427 (1980). Okumura, T., T. Kadono, and M. Nakatani, "Thin-layer Chromatography on Precoated Adsorbents Fixed with Fused Glass. V. Thin-layer Chromatography on Kieselguhr Sintered Plate," Yakugaku Zasshi, 93(1), 79-86 (1973); Chem. Abst., 78, 105744J (1973). Onuska, F. I., and M. Comba, "Identification and Quantitative Analysis of Polychlorinated Biphenyls on WCOT Glass Capillary Columns," Environ. Sci. Res., 16, 285-302 (1980). Ill �Osterroht, C., "Development of a Method for the Extraction and Determination of Non-polar, Dissolved Organic Substances in Sea Water," J. Chromatogr., 101, 289-298 (1974). Osterroht, C., "Dissolved PCB's and Chlorinated Hydrocarbon Pesticides in the Baltic, Determined by Two Different Sampling Procedures," Marine Chem., 5, 113-121 (1977). Oswald, E. 0., L. Levy, B. J. Corbett, and M. P. Walker, "Differentiation and Characterization of Isomeric Polychlorinated Biphenyls by Gas-Liquid Chromatography Coupled with Electron Impact and Chemical lonization Mass Spectrometry," J. Chromatogr., 9j3, 63-90 (1974). Oswald, E. 0., P. W. Albro, and J. D. McKinney, "Utilization of Gas-Liquid Chromatography Coupled with Chemical lonization and Electron Impact Mass Spectrometry for the Investigation of Potentially Hazardous Environmental Agents and Their Metabolites," J. Chromatogr., 98, 363-448 (1974). Ouw, H. K., G. R. Simpson, and D. S. Siyali, "Use and Health Effects of Aroclor 1242, a Polychlorinated Biphenyl, in an Electrical Industry," Arch. Environ. Health, 31, 189-194 (1976). Paasivirta, J., and M. Pitkanen, "Structure Analysis of the Main Components in PCB Mixture Clophen A 60," Finn. Chem. Lett., 89-93 (1975). Pacco, J. M., and A, K. Mukherji, "Determination of Polychlorinated Biphenyls in a Polymer Matrix by Gel Permeation Chromatography Using Micro-Styragel® Columns," J. Chromatogr., 144, 113-117 (1977). Pellizzari, E. D., "Analysis of Adipose and Blood Sera Samples for Individual PCB Isomers," MRI Subcontract No. 82-4900-21, EPA Contract No. 68-01-5915, Monthly Tech. Prog. Report No. 2, March 1, 1982. Pellizzari, E. D., "State-of-the-Art Instrumental Organic Analysis in Environmental Chemistry," Chapter 10 in Environmental Health Chemistry, James D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, MI (1981), pp. 195-218. Pellizzari, E. D., K. B. Tomer, and M, A. Moseley, "High-Resolution Gas Chromatography and Negative Ion Chemical lonization Mass Spectrometry of Polychlorinated Biphenyls," Chapter 12 in Advances in the Identification and Analysis of Organic Pollutants in Water, L. H. Keith (Ed.), Vol. 1, Ann Arbor Science Publishers, Inc., Ann Arbof, MI (1981). pp. 197-218. Pellizzari, E. D., M. A. Moseley, S. D. Cooper, J. V. Harry, and B. A. Demian, "Recent Advances in the Analysis of Polychlorinated Biphenyls in Environmental and Biological Media," in Recent Advances in Exposure, Health and Environmental Effects Studies of PCBs, Proceedings of Symposium, Bethesda, Maryland, May 12-13, 1982. Penville, P. E., "Dry Ice Homogenization Procedure for Fish Samples in Pesticide Residue Analysis," J. Agr. Food Chem., 18(5), 948-949 (1970). 112 �Pepe, M. G., and J. J. Byrne, "Adhesion-Binding of 2,2',4,4',5,5'-Hexachlorobiphenyl to Glass and Plastic: A Posible Source of Error for PCB Analysis," Bull. Environ. Contam. Toxicol., 25, 936-940 (1980). Peterman, P. H. , J. J. Delfino, D. J. Dube, T. A. Gibson, and F. J. Priznar, "Chloro-organic Compounds in the Lower Fox River, Wisconsin," in Hydrocarbons and Halogenated Hydrocarbons in the Aquatic Environment, B. K. Afghan and D. Mackay (Eds.), Plenum Publishing Corporation, New York (1980). pp. 145-160. Picer, M., and M. Ahel, "Separation of Polychlorinated Biphenyls from DDT and Its Analogues on a Miniature Silica Gel Column," J. Chromatogr., 150, 119-127 (1978). Picer, N., and M. Picer, "Evaluation of Macroreticular Resins for the Determination of Low Concentratons of Chlorinated Hydrocarbons in Sea Water and Tap Water," J. Chromatogr., 193, 357-369 (1980). Pittaway, A. R., R. S. O'Connor, and E. J. Heiden, "A Report of a Survey on the Incidental Manufacture, Processing, Distribution, and Use of Polychlorinated Biphenyl at Concentrations Below 50 ppm," Report prepared for Chemical Manufacturers Association by Regulatory Research Service (November 1981). Pittaway, A. R., and T. W. Horner, "Statistical Analysis of Data from a Round Robin Experiment on PCB Samples," Report prepared for Chemical Manufacturers Association, Washington, D.C., March 18, 1982. Polishuk, Z. W., M. Ron, M. Wassermann, S. Cucos, D. Wassermann, and C. Lemesch, "Pesticides in People. Organochlorine Compounds in Human Blood Plasma and Milk," Pesticide Monit. J., 1.0(4), 121-129 (1977). Pressley, T. A. and J. E. Longbottom, "The Determination of Organohalide Pesticides and PCBs in Industrial and Municipal Wastewater: Method 617," U.S. Environmental Protection Agency, Report No. EPA-600/4-82-006, 35 pp. (1982); NTIS PB82-156001 (1982). Risebrough, R. W., "Determination of Polychlorinated Biphenyls in Environmental Samples," Presented at the International Symposium on Identification and Measurement of Environmental Pollutants, Ottawa, Ontario, Canada, June 14-17, 1971. Riva, M., A. Carisano, and A. Gaghetta, "Rapid GLC-ECD [Gas-Liquid Chromatography-Electron Capture Detector] Determination of Chloroorganic Pesticide and Polychlorobiphenyl (PCB's) Residues in Animal Foods. General Considerations about Ecological Contamination," Riv. Ital. Sostanze Grasse, 50(12), 434-442 (1973); Chem. Abst., 81, 11924c (1974). Robbins, A. L., and C. R. Willhite, "A Rapid Method for the Analysis of Polychlorinated Biphenyls in Milk," Bull. Environ. Contam. Toxicol., 21, 428-431 (1979). Rodriguez, C. F., W. A. McMahon, and R. E. Thomas, "Method Development for Determination of Polychlorinated Hydrocarbons in Municipal Sludge," Final Report, Contract No. 68-03-2606, Environmental Protection Agency, EPA-600/280-029, March 1980. 113 �Rohleder, H., H. Staudacher, and W. Suemmermann, "High-pressure Liquid Chromatography for the Separation of Lipophilic Organochlorine Xenobiotics from Triglycerides in Trace Analysis," Fresenius Z. Anal. Chem., 279(2), 152-153 (1976); Chem. Abst.. 84, I60197g (1976). Ross, P. F., D. L. Osheim, and H. A. Nelson, "PCB Residues in Feedlot Steers," Bull. Environ. Contam. Toxicol., 26, 485-488 (1981). Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in Identification of Polychlorinated Biphenyls by Mass Spectrometry," J. Assoc. Offie. Anal. Chem.. 56(1), 188-199 (1973). Ruzo, L. 0., M. J. Zabik, and R. D. Schuetz, "Polychlorinated Biphenyls: Photolysis of 3,4,3',4'-Tetrachlorobiphenyl and 4,4'-Dichlorobiphenyl in Solution," Bull. Environ. Contam. Toxicol., 8 4 , 217-218 (1972). () Saeki, S., A. Tsutsui, K. Oguri, H. Yoshimura, and M. Hamana, "Isolation and Structure Elucidation of the Amino Component of KC-400 (Chlorobiphenyls)," Fukuoka-Igaku-Zasshi, 62(1), 21-24 (1971); Chem. Abst., 74, I46294y (1971). Safe, S., "Overview of Analytical Identification and Spectroscopic Properties," in National Conference on Polychlorinated Biphenyls (November 19-21, 1975, Chicago, Illinois), F. A. Ayer (Ed.), EPA-560/6-75-004, March 1976. Safe, S., and R. K. Boyd, "Pesticide Analysis by Mass Spectrometry," in Pesticide Analysis Conf. Proc., K. G, Das (Ed.), Dekker, New York, (1981) pp. 329-368; Chem. Abst., 95, 110007b (1981). Safe, S., M. Mullin, L. Safe, C. Pochini, S. McCrindle, and M. Romkes, "High Resolution PCB Analysis," in Proceedings: PCB Workshop, D. McKay (Ed.), Toronto, Ontario, December 1981, Ann Arbor Science Publications, Ann Arbor, Michigan, in press (1982). Safe, S., A. Parkinson, L. Robertson, R. Cockerline, L. Safe, S. Bandiera, and A. Okey, "PCBs as AHH Inducers," in Chlorinated Dioxins and Related Compounds , 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds.), Pergamon Press, New York, 383-392 (1981). Safe, S., N. Platonow, 0. Hutzinger, and W. D. Jamieson, "Analysis of Organochlorine Metabolites in Crude Extracts by High Resolution Photoplate Mass Spectrometry," Biomed. Mass Spectrom., 2, 201-203 (1975). Sans, W. W., "Multiple Insecticide Residue Determination Using Column Chromatography, Chemical Conversion, and Gas-Liquid Chromatography," J.Agr. Food Chem., 15, 192-198 (1967). Saperstein, M. D., R. J. Gordon, and E. J. Faeder, "PCB Contamination in Distribution Transformers," J. Environ. Sci. Health, A17(2), 241-251 (1982). Sauter, A. D., P. E. Mills, W. L. Fitch, and R. Dyer, "Interlaboratory GC/MS Response Factor Precision," J. High Resol. Chromatogr. & Chromatogr. Commun., 5, 27-30 (1982). 114 �Savage, E. P., J. D. Tessari, J. W. Malberg, H. W. Wheeler, and J. R. Bagby, "A Search for Polychlorinated Biphenyls in Human Milk in Rural Colorado," Bull. Environ. Contam. Toxicol., 9(4), 222-226 (1973a). Savage, E. P., J. D. Tessari, J. W. Malberg, H. W. Wheeler, and J. R. Bagby, "Organochlorine Pesticide Residues and Polychlorinated Biphenyls in Human Milk, Colorado—1971-72," Pestic. Monit. J., 7(1), 1-3 (1973b). Sawyer, L. D., "Collaborative Study of the Recovery and Gas Chromatographic Quantitation of Polychlorinated Biphenyls in Chicken Fat and Polychlorinated Biphenyl-DDT Combinations in Fish," J. Assoc. Offic. Anal. Chem., 56(4), 1015-1023 (1973). Sawyer, L. D., "Quantitation of Polychlorinated Biphenyl Residues by Electron Capture Gas-Liquid Chromatography: Collaborative Study," J. Assoc. Offic. Anal. Chem.. 61(2), 282-291 (1978). Sawyer, L. D., "Quantitation of Polychlorinated Biphenyl Residues by Electron Capture Gas-Liquid Chromatography: Reference Material Characterization and Preliminary Study," J. Assoc. Offic. Anal. Chem.. 61(2), 272-281 (1978). Schulte, E., H. P. Thier, and L. Acker, "Residual Analysis of Polychlorinated Biphenyls in Foods of Animal Origin: Experiences and Proposals for Standardization," Dtsch. Lebensm.-Rundsch.. 72(7), 229-232 (1976); Chem. Abst.. 85^, 121894h (1976). Schulte, E., and L. Acker, "Identifizierung und Metabolisierbarkeit von polychlorierten Biphenylen," Naturwissenschaften, 61, 79-80 (1974). Schulte, E., and L. Acker, "Gas-Chromatographie mit Glascapillaren bei Temperaturen bis zu 320°C und ihre Anwendung zur Trennung von Polychlorbiphenylen," Fresenius Z. Anal. Chem., 26_8(4), 260-267 (1974); Chem. Abst. , 81, 115377t (1974). Schwartz, T. R., J. D. Petty, D. L. Stalling, J. W. Hogan, M. K. Marlow, R. D. Campbell, and R. L. Little, "A MUMPs (DSM-11) Derived Data Base for Congener Specific PCB Analysis Leading to Pattern Recognition Studies," Abstract, 184th Annual Chemical Society National Meeting, September 12-17, 1982. Seiber, J. N., "Reversed-Phase Liquid Chromatography of Some Pesticides and Related Compounds. Solubility-Retention Relationships," J. Chromatogr., 94, 151-157 (1974). Seidl, G., and K. Ballschmiter, "Isolation of PCB's from Soil. Recovery Rates Using Different Solvent Systems," Chemosphere, 5, 373-376 (1976a). Seidl, G., and K. Ballschmiter, "Isolation of PCB's from Vegetable Oils: Recovery and Efficiency of 'Clean Up' Methods," Chemosphere, 5, 363-366 (1976b). 115 �Seidl, G., and K. Ballschmiter, "Quantitation of Polychlorinated Biphenyl (PCB)-Residues after Hydrodechlorination to Biphenyl Using Liquid Chromatography with UV-Detection," Fresenius Z. Anal. Chem., 296, 281-284 (1979). Serum, J. W., S. C. long, L. E. St. John Jr., C. A. Bache, D. R. Mertens, and D. J. Lisk, "Electron-Capturing Compounds and Selected Elements in Paper," Bull. Environ. Contain. Toxicol., 10_(2), 88-96 (1973). Shahied, S. I., R. P. Stanovick, D. E. Mclnturff, and E. Missaghi, "Determination of Polychlorinated Biphenyl (PCB's) Residues in Grades of Pulp, Paper and Paperboard," Bull. Environ. Contain. Toxicol., 10(2), 80-87 (1973). Shaw, G. R., H. C. Chapman, and G. W. Roache, "Determination of Organochlorine Pesticides and Polychlorinated Biphenyls in Sediment Samples Containing Oil Residues," Residue (Melbourne), 3(2), 2-8 (1976); Chem. Abst.. 88, 1362y (1978). Shaw, D. G., "Discrimination Between PCB and DDT Residues by a Gas Chromatographic-Mass Spectrometric Technique," Bull. Environ. Contam. Toxicol., 8(4), 208-211 (1972). Sherma, J., "Gas-Chromatography Analysis of Polychlorinated Biphenyls and Other Nonpesticide Organic Pollutants," Chapter 5 in Advances in Chromatography, J. C. Giddings, E. Gruehka, R. A. Keller, and J. Cages (Eds.), Marcel Dekker, Inc., New York (1975). pp. 141-176. Sherma, J., Manual of Analytical Quality Control for Pesticides and Related Compounds in Human and Environmental Samples, EPA-600/2-81-059, PB81-222721 (April 1981). Shirai, F., "Detection of Polychlorinated Terphenyls on Food Wrappers by Reversed-Phase Partition Thin-Layer Chromatography," Eisei Kagaku, 20(5), 282-286 (1974); Chem. Abst., 82, 84559t (1975). Simon, C. G., and T. F. Bidleman, "Sampling Airborne PCB with Polyurethane Foam--A Chromatographic Approach to the Determination of Collection Efficiencies," Presented at the 176th ACS National Meeting, Miami Beach, Florida, September 11-15, 1978. Simon, C. G., and T. F. Bidleman, "Sampling Airborne Polychlorinated Biphenyls with Polyurethane Foam--Chromatographic Approach to Determining Retention Efficiencies," Anal. Chem., 51_(8), 1110-1113 (1979). Sissons, D., and D. Welti, "Structural Identification of Polychlorinated Biphenyls in Commercial Mixtures by Gas-Liquid Chromatography, Nuclear Magnetic Resonance and Mass Spectrometry," J. Chromatogr., 60, 15-32 (1971). Siyali, D- S., "Polychlorinated Biphenyls, Hexachlorobenzene and Other Organochlorine Pesticides in Human Milk," Med. J. Aust., 2, 815-818 (1973). Snyder, D., and R. Reinert, "Rapid Separation of Polychlorinated Biphenyls from DDT and Its Analogues on Silica Gel," Bull. Environ. Contam. Toxicol., 6(5), 385-390 (1971). 116 �Spies, D. N., "Rapid Field Extraction Technique for Determining Extent of PCS Contaminated Sediments," Hazardous Materials Spills, Conference Proceedings , J. Ludwigson (Ed.), 435-437 (1982); Chem. Abst., 97, 60688t (1982). Stalling, D. L., J. N. Huckins, J. D. Petty, J. L. Johnson, and H. 0. Sanders, "An Expanded Approach to the Study and Measurement of PCBs and Selected Planar Halogenated Aromatic Environmental Pollutants," Ann. N.Y. Acad. Sci., 320, 48-59 (1979). Stalling, D. L., "Application of Analytical Methods Research to Monitoring Organic Residues in Fish," The Institute of Electrical and Electronics Engineers, Inc., Annals No. 75CH1004-I 7-5 (1976). Stalling, D. L., J. Johnson, and J. N. Huckins, "Automated Gel Permeation— Carbon Chromatographic Cleanup of Dioxins, PCBs, Pesticides, and Industrial Chemicals," in Environmental Quality and Safety, Supplement Vol. Ill, Pesticides lectures of the IUPAC, Third International Congress of Pesticide Chemistry, Helsinki, July 1974, F. Coulson and F. Korte (Eds.), G. Thieme Publ., Stuttgart, Germany (1975). pp. 12-18. Stalling, D. L., R. C. Tindle, and J. L. Johnson, "Cleanup of Pesticide and Polychlorinated Biphenyl Residues in Fish Extracts by Gel Permeation Chromatography," J. Assoc. Offic. Anal. Chem., 55(1), 32-38 (1972). Stalling, D. L., "Analysis of Organochlorine Residues in Fish: Current Research at the Fish-Pesticide Research Laboratory," Pestic. Chem.? Proc. Intl. IUPAC Congr. Pest. Chem., 2nd, 4, 413-438 (197lji Stalling, D. L., and J. N. Huckins, "Reverse Phase Thin Layer Chromatography of Some Aroclors, Halowaxes, and Pesticides," J. As s o c. Offi c. Ana1, Chem., 56(2), 367-372 (1973). Stalling, D. L., L. M. Smith, and J. D. Petty, "Approaches to Comprehensive Analyses of Persistent Halogenated Environmental Contaminants," Measurement of Organic Pollutants in Water and Wastewater. ASTM STP 686, C. E. Van Hall (Ed.), American Society for Testing and Materials (1979). pp. 302-323. Stalling, D. L., J. N. Huckins, and J. D. Petty, "Presence and Potential Significance of o-o'-Unsubstituted PCB Isomers and Trace Aroclor® 1248 and 1254 Impurities," in Hydrocarbons and Halogenated Hydrocarbons in the Aquatic Environment, B. K. Afghan and D. Mackay (Eds.), Plenum Press, New York (1978). pp. 131-139. Stalling, D. L., and J. N. Huckins, "Gas-Liquid Chromatography—Mass Spectrometry Characterization of Polychlorinated Biphenyls (Aroclors) and 3fe Cl-Labeling of Aroclors 1248 and 1254," J. Assoc. Offic. Anal. Chem., 54(4), 801-807 (1971). Stalling, D. L., J. D. Petty, G. R. Dubay, and R. A. Smith, "Application of LC/MS to Chromatographic Separation of Aromatics Using Carbon as the Stationary Phase," J. Chromatogr. Sci., 18 107-108 (1980). 117 �Stalling, D. L., J. D. Petty, L. M. Smith, and G. R. Dubay, "Contaminant Enrichment Modules and Approaches to Automation of Sample Extract Cleanup," Chapter 9 in Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, J. D. McKinney (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1981). pp. 177-193. Stan, H.-J., "Combined Gas Chromatography-Mass Spectrometry," Chapter 9 in Pesticide Analysis, K. G. Das (Ed.), Marcel Dekker, Inc., New York (1981). Stanley, J. S., C. L. Haile, G. R. Cobb, and T. Walker, "Comprehensive Assessment of the Specific Compounds Present in Combustion Processes," Draft Final Report, Contract No. 68-01-5915, prepared for U.S. Environmental Protection Agency, Office of Pesticides and Toxic Substances, Field Studies Branch (1981). Stanovick, R. P., S. . Shahied, and E. Missaghi, "Determination of Polychlorinated Biphenyl (Aroclor 1242) Migration Into Food Types," Bull. Environ. Contain. Toxicol., 10(2), 101-107 (1973). Steichen, R. J., R. G. Tucker, and E. Mechon, "Standardization of Aroclor Lots for Individual-Peak Gas Chromatographic Calibration," J. Chromatogr., 236, 113-126 (1982). Stenhagen, E., S. Abrahamsson, and F. W. McLafferty, Registry of Mass Spectral Data, Vol. 1-4, Wiley, New York (1974). Stijve, T., and E. Cardinale, "Rapid Determination of Chlorinated Pesticides, Polychlorinated Biphenyls, and a Number of Phosphated Insecticides in Fatty Foods," Mitt. Geb. Lebensmittelunters. Hyg., 65(1), 131-150 (1974); Chem. Abst.. 82, 137862V (1975). Stratton, C. L., and P. C, Geiszler, "Analysis of Polychlorinated Biphenyl (PCB) in Human Blood Serum Samples," Final Report, prepared for U.S. Environmental Protection Agency, Contract No. 68-01-3248, NTIS PB-291 358/8ST (October 1977). Stratton, C. L., J. M. Allan, and S. A, Whitlock, "Advances in the Application of the Perchlorination Technique for the Quantitation and Confirmation of Polychlorinated Biphenyls (PCBs)," Bull. Environ. Contain. Toxicol., 21 230-237 (1979). Stratton, C. L., K. L. Tuttle, and J. M. Allan, "Environmental Assessment of Polychlorinated Biphenyls (PCBs) near New Bedford, MA Municipal Landfill," Report prepared for U.S. Environmental Protection Agency, Contract No. 68-013248, Report No. 560/6-78-006 (May 1978). Stratton, C. L., S. A. Whitlock, and J. M. Allan, "A Method for the Analysis of Polychlorinated Biphenyl (PCB) in Air," Report prepared for U.S. Environ*mental Protection Agency, Contract No. 68-01-2978, Report No. EPA-600/4-78-048 (August 1978). 118 �Su, G. C. C., and H. A. Price, "Element Specific Gas Chromatographic Analyses of Organochlorine Pesticides in the Presence of PCB's by Selective Cancellation of Interfering Peaks," J. Agr. Food Chem.. 21 1099-1102 (1973). Sundaram, K. M. S, and P. G. Davis, "Analysis of DDT Residues in Animal Tissues and Soils Collected from Different Regions of Canada," File Rep. - Chem. Control Res. Inst., Ottawa. 3, 63 pp. (1974); Chem. Abst., 82, 93952t (1975). Sundstrom, G., and C. A. Wachtmeister, "Synthesis of 14C-Labelled and Unlabelled PCS Compounds," in PCS Conference II, National Swedish Environment Protection Board, Stockholm, Publications 1973:4E (1972). pp. 73-86. Sundstrom, G., 0. Hutzinger, F. W. Karasek, and J. Michnowicz, "Environmental Chemistry of Substitutes for Polychlorinated Biphenyls. I. Composition and Properties of an Alkylchlorobiphenyl Product," J. Assoc. Offie. Anal. Chem., 59(5), 982-988 (1976). Supelco, Chromatography Supplies Catalog 20, p. 26. Swift, R., and H. Settle, "A Procedure for the Determination of Polychlorinated Biphenyls in Fat Samples," Residue (Melbourne), 3(2), 9-12 (1976); Chem. Abst., 87, 199289a (1977). Sworzyn, E. M., and D. G. Ackerman, "Interim Guidelines for the Disposal/ Destruction of PCBs and PCB Items by Non-thermal Methods," Environmental Protection Agency, Contract No. 68-02-3174, Task 41, Report No. EPA-600/ 2-82-069 (1981). Szelewski, M. J., D. R. Hill, S. J. Spiegel, and E. C. Tifft, Jr., "Loss of Polychlorinated Biphenyl Homologues during Chromium Trioxide Extraction of Fish Tissue," Anal. Chem., 5J.(l4), 2405-2407 (1979). Takeshita, R., and T. Yamashita, "Reversed-Phase Partition Thin-Layer Chromatography of Chlorinated Pesticides and Polychlorinated Biphenyls," Eisei Kagaku, 18(6), 388-392 (1972); Chem. Abst., 78, I44l91w (1973). Tanabe, H., "Methods of Analysis of PCB [Polychlorinated Biphenyls] in Edible Oil and Fat," Yukagaku, 22(1), 35-40 (1973); Chem. Abst.. 78, 119783h (1973). Tanabe, H., "PCB Microanalysis," in PCB Poisoning and Pollution, K. Higuchi, (Ed.), Kodansha, Ltd., Tokyo, Academic Press, London, New York, San Francisco, (1976). p. 127-145. Tas, A. C., and R. H. de Vos, "Characterization of Four Major Components in a Technical Polychlorinated Biphenyl Mixture," Environ. Sci. Technol., 5_(12), 1216-1218 (1971). Tas, A. C., an R. J. C. Kleipool, "Characterization of the Components of Technically Polychlorinated Biphenyl Mixtures—II," Bull. Environ. Contain. Toxicol., 8(1), 32-36 (1972). Tatsukawa, R., and T. Wakimoto, "Pollution Analysis. Determination of Polychlorinated Biphenyl in Soil," Kogai Bunseki Shishin, 6, 45-56 (1972), Chem. Abst., 82, 11793e (1975). 119 �Teichman, J., A. Bevenue, and J. W. Hylin, "Separation of Polychlorobiphenyls from Chlorinated Pesticides in Sediment and Oyster Samples for Analysis by Gas Chromatography," J. Chromatogr., 151, 155-161 (1978). Tessari, J. D., "Analytical Methods for PCBs in Human Milk," Eny.Che.in: & Animal Health, 5, 227-229 (1977). Human Tessari, J. D., and E. P. Savage, "Gas-Liquid Chromatographic Determination of Organochlorine Pesticides and Polychlorinated Biphenyls in Human Milk," J. Assoc. Offic. Anal. Chem.. 63(4), 736-741 ( 9 0 . 18) Tessari, J. D., L. Griffin, and M. J. Aaronson, "Comparison of Two Cleanup Procedures (Mills, Onley, Gaither v£ Automated Gel Permeation) for Residues of Organochlorine Pesticides and Polychlorinated Biphenyls in Human Adipose Tissue," Bull. Environ. Contain. Toxicol., 25_, 59-64 (1980). Thielemann, H., "Diinnschichtchromatographische Nachweisgrenzen (semiquantitative Bestimmung) von polychloriertem Biphenyl (Chlophen A5o®) sowie Identifizierung des Insektizids Thiodan (Endosulfan)," Z. Chem., 1^, 147-148 (1978). Thielemann, H., "Separation and Identification of Polychlorinated Biphenyls and p,p'-DDE in Paraffin-Impregnated Commercial Sheets by Repeated Thin-Layer Chromatography," Z. Gesamte Hyg. Ihre Grenzgeb., 26(6), 463-464 (1980); Chem. Abst., 94, 202208a (1981). Thielemann, H., "Diinnschicht-chromatographische Auftrennung und Identifizierung eines Gemisches von polychlorierten Biphenylen (PCB) und p,p'-DDE an parraffinimpragnierten Kieselgur G-Schichten," Fresenius Z. Anal. Chem., 282, 144 (1976). Thomson, B. A., and J. R, Roberts, "The Development of an Analytical Technique for the Rapid in situ Detection of PCBs and Trichlorobenzene in Clay," in A Case Study of A Spill of Industrial Chemicals—Polychlorinated Biphenyls and Chlorinated Benzenes, National Research Council of Canada, Associate Committee on Scientific Criteria for Environmental Quality (NRCC No. 17586), Section VII (Appendix 3) (1980). 22 pp. Thomson, B. A., and J. R. Roberts, "A New Technique for the Rapid Analysis of Soil for the Presence of Polychlorinated Biphenyls," Intern. J. Environ. Anal. Chem., n, 139-151 (1982). Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass Spectrometry Method for Identifying and Determining Polychlorinated Biphenyls," J. Chromatogr., 196, 109-119 (1980). Tindle, R. C., and D. L. Stalling, "Apparatus for Automated Gel Permeation Cleanup for Pesticide Residue Analysis. Applications to Fish Lipids," Anal. Chero., 44(11), 1768-1773 (1972). Tressl, G., and K. Wessely, "Analysis of Chloropesticides and PCB's in Cheese by Glass Capillary GC/MS Using the Selected Ion Monitoring Technique," Application Note AN 176-22, Hewlett-Packard Company, Avondale, Pennsylvania (1976). 120 �Trevisani, G. R., "Simultaneous Studies on Chlorinated Pesticides, Polychlorinated Biphenyls, Polychlorinated Triphenyls, and Polychlorinated Naphthalenes in Fatty Foods of Various Types," Boll. Chim. Unione Ital. Labv Prov., Parte Sci., 6(2), 117-142 (1980); Chem. Abst.. 93, 202765J (1980). Trotter, W. J., and S. J. V. Young, "Limitation on the Use of Antimony Pentachloride for Perchlorination of Polychlorinated Biphenyls," J. Assoc. Offic. Anal. Chem., 58(3), 466-468 (1975). Trotter, W. J., "Removing the Interference of DDT and Its Analogs in the Analysis for Residues of Polychlorinated Biphenyls," J. Assoc. Offic. Anal. Chem., 58(3), 461-465 (1975). Tuinstra, L. G. M. Th., W. A. Traag, and H. J. Keukens, "Quantitative Determination of Individual Chlorinated Biphenyls in Milkfat by Splitless Glass Capillary Gas Chromatography," J. Assoc. Offic. Anal. Chem., 63(5), 952-958 (1980). Tuinstra, L. G. M. Th., and W. A. Traag, "Automated Glass Capillary Gas Chromatograpic Analysis of PCB and Organochlorine Pesticide Residues in Agricultural Products," J. High Resol. Chromatogr. & Chromatogr. Commun., 2, 723728 (1979). Tuinstra, L. G. M. Th., and W. A. Traag, "Automated (GC)2 Analysis of PCB and Organochlorine Pesticide Residues in Agricultural Products," Med. Fac. Landbouww. Rijksuniv. Gent, 44(2), 885-893 (1979). Tuinstra, L. G. M. Th., W. A. Traag, and A. J. Van Munsteren, "Determination of Individual Chlorinated Biphenyls in Agricultural Products by Automated Capillary Gas Chromatography. Determination in Cattle Feed and Its Relation to Milk Residues," J. Chromatogr., 204, 413-419 (1981). Ueta, T., K. Kamata, T. Kan, M. Kazama, and T. Totani, "Fluorescence Reactions for Organic Halogen Compounds. II. In Situ Fluorometry of Polychlorinated Biphenyls and Their Isomers on Thin-Layer Chromatograms Using Diphenylamine," Tokyo Toritsu Eisei Kenkyusho Kenkyu Nempo, 25, 111-118 (1974); Chem. Abst., 83, 21991c (1975). Ugawa, M., A. Nakamura, and T. Kashimoto, "Calculation Method for Polychlorinated Biphenyl (PCB) Isomers in Gas Chromatographic Determination," Shokuhin Eiseigaku Zasshi, 14(5), 415-424 (1973); Chem. Abst., 80, 90955u (1974). Ugawa, M., A. Nakamura, and T. Kashimoto, "Studies on a Calculation Method for Polychlorinated Biphenyl Isomers (PCBs)," in New Methods in Environmental Chemistry and Toxicology, Collection of Papers Presented at the Research Conference on New Methodology in Ecological Chemistry, Susono, Japan, November 23-25, 1973, F. Coulston, F. Korte, and M. Goto (Eds.), International Academic Printing Co., Ltd., Tokyo (1973). pp. 253-267. 121 �Uk, S., C. M. Himel, and T. Dirks, "Mass Spectrometric Identification of Mirex Residues in Crude Extracts and in the Presence of Polychlorinated Biphenyls," Bull. Environ. Contam. Toxicol., 8(2), 97-104 (1972). United Kingdom Department Polychlorinated Biphenyls Examination of Waters and Polychlorinated Biphenyls of the Environment, "Organochlorine Insecticides and in Waters 1978; Tentative Method. Methods for the Associated Materials," Organochlorine Insectic. Waters, 28 pp. (1979), Underwood, J. C., "Separation of Polychlorinated Biphenyls from DDT and Its Analogs Using Chromic Acid and Silica Gel," Bull. Environ. Contam. Toxicol., 21, 787-790 (1979). Uthe, J. F., J. Reinke, and H. O'Brodovich, "Field Studies on the Use of Coated Porous Polyurethane Plugs as Indwelling Monitors of Organochlorine Pesticides and Polychlorinated Biphenyl Contents of Streams," Environ. Lett., 6(2), 103-115 (1974). Vannucchi, C., R. Salvadori, R. Nottoli, G. Ghimenti, and G. Taponeco, "Identification and Quantification of Aromatic Polychlorinated Hydrocarbon Residues in the Environment," Riv. Sci. Tecnol. Alimenti Nutr. Urn., 6(2), 101-114 (1976); Chem. Abst., 85, 175663a (1976). Vannucchi, C., and M. Berlincioni, "Sampling and Analytical Method to Determine PCB Concentrations in the Air of the Working Areas of an Electrical Industry," Am. Ind. Hyg. Assoc. J., 41, 352-360 (1980). Van Vliet, H. P. M., Th. C. Bootsman, R. W. Frei, and U. A. Th. Brinkman, "On-Line Trace Enrichment in High-Performance Liquid Chromatography Using a Pre-Column," J. Chromatogr., 185, 483-495 (1979). Veierov, D., and N. Aharonson, "Economic Method for Analysis of Fluid Milk for Organochlorine Residues at the 10 ppb Level," J. Assoc. Offic. Anal. Chem., 63(3), 532-535 (1980). Veierov, D., and N. Aharonson, "Improved Cleanup of Large Lipid Samples for Electron Capture Gas Chromatographic Quantitation and Gas ChromatographicMass Spectrometric Confirmation of Organochlorine Residues," J. Assoc. Offic. Anal. Chem.. 63(2), 202-207 (1980). Veith, G. D., "Baseline Concentrations of Polychlorinated Biphenyls and DDT in Lake Michigan Fish, 1971," Pestic. Monit. J., 9(1), 21-29 (1975). Veith, G. D., and N. M. Austin, "Detection and Isolation of Bioaccumuable Chemicals in Complex Effluents," Chapter 20 in Identification and Analysis of Organic Pollutants in Water, L. H. Keith (Ed.), Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1976). pp. 297-302. Veith, G. D., and L. M. Kiwus, "An Exhaustive Steam-Distillation and SolventExtraction Unit for Pesticides and Industrial Chemicals," Bull. Environ. Contam. Toxicol., jj, 631-636 (1977). 122 �Veith, G. D., D. W. Kuehl, E. N. Leonard, K. Welch, and G. Pratt, "Polychlorinated Biphenyls and Other Organic Chemical Residues in Fish from Major United States Watersheds Near the Great Lakes, 1978," Pestic. Mon. J., 15, 1-8 (1981). Wakimoto, T., M. Fukushima, R. Tatsukawa, and T. Ogawa, "Separation of PCBs [Polychlorinated Biphenyls] from p,p'-DDE and Other Organochlorine Pesticides by a Newly Developed Silica Gel," Nippon Nogei Kagaku Kaishi, 49(10), 499-503 (1975); Chem. Abst.. 84, Il6448r (1976). Wakimoto, T., R. Tatsukawa, and T. Ogawa, Kogai to Taisaku, 7, 517 (1971); Fish Res. Bd. Can., Translation No. 2185. Ward, R. S., and A. Pelter, "The Analysis of Mixtures of Closely Related Naturally-occurring Organic Compounds Using High Performance Liquid Chromatography," J. Chromatogr. Sci.. 12, 570-574 (1974). Wardall, G. L., "Method for the Extraction of Organochlorine Compounds from Quail Eggs," Analyst. 102, 54-55 (1977). Wassermann, M., D. Wassermann, S. Cucos, and H. J. Miller, "World PCBs Map: Storage and Effects in Man and His Biologic Environment in the 1970s," Ann. N.Y. Acad. Sci., 320, 69-124 (1979). Watanabe, I., T. Yakushiji, K. Kuwabara, S. Yoshida, K. Koyama, I. Hara, and N. Kunita, "Studies on Polychlorinated Biphenyl (PCB) Isomers in Blood of Ordinary Persons, Oil Sickness Patients, and Occupationally Exposed Workers. 2. Apiezon L Column Analysis," Osaka-furitsu Koshu Eisei Kenkyusho Kenkyu Hokoku, Shokuhin Eisei Hen. 8, 51-58 (1977); Chem. Abst., 88, 183989? (1978). Watts, R. R. (Ed.), Analysis of Pesticide Residues in Human and Environmental Samples, A Compilation of Methods Selected for Use in Pesticide Monitoring Programs, U.S. Environmental Protection Agency, Contract No. 68-02-2474, Publication No. EPA-600/8-80-038, Research Triangle Park, North Carolina (June 1980). Webb, R. G., "Isolating Organic Water Pollutants: XAD Resins, Urethane Foams, Solvent Extraction," prepared for U.S. Environmental Protection Agency, NTIS Publication No. PB-245 647 (June 1975). Webb, R. G., and A. C. McCall, "Identities of Polychlorinated Biphenyl Isomers in Aroclors," J. Assoc. Offic. Anal. Chem.. 55(4), 746-752 (1972). Webb, R. G., and A. C. McCall, "Quantitative PCB Standards for Electron Capture Gas Chromatography," J. Chromatogr. Sci., 11, 366-373 (1973). Welborn, M. E., R. E. Borchard, L. G. Hansen, and R. L. Metcalf, "Extraction and Analysis of Aroclor 1254 Components in Blood," J. Assoc. Offic. Anal. Chem., 57(6), 1248-1252 (1974). 123 �Westoo, G., and K. Noren, "Determination of Organochlorine Pesticides and Polychlorinated Biphenyls in Animal Foods," Acta Chem. Scand., 24(5), 1639" 1644 (1970). Williams, D. T., and F. M. Benoit, "The Determination of Polychlorinated Biphenyls in Selected Household Products," Bull. Environ. Contain. Toxieol., 21, 179-184 (1979). "~T Williams, D. T., and B. J. Blanchfield, "Screening Method for the Detection of Chlorodibenzo-j>-dioxins in the Presence of Chlorobiphenyls, Chloronaphthalenes, and Chlorodibenzofurans," J. Assoc. Qffic. Anal. Chem.. 55(1), 93-95 (1972). Williams, D. T., and B. J. Blanchfield, "Thin Layer Chromatographic Separation of Two Chlorodibenzo-£-dioxins from Some Polychlorinated Biphenyl and Organochlorine Pesticides," J. Assoc. Offic. Anal. Chem., 54(6), 1429-1431 (1971). Williams, D. T., G. L. LeBel, and T. Furmanczyk, "Polychlorinated Biphenyl Contamination of Laboratory Air," Chromosphere, 9_, 45-50 (1980). Willis, D. E., and R. F. Addison, "Identification and Estimation of the Major Components of a Commercial Polychlorinated Biphenyl Mixture, Aroclor 1221," J. Fish. Res. Board Can., 29(5), 592-595 (1972). Willmott, F. W., and R. J. Dolphin, "A Novel Combination of Liquid Chromatography and Electron Capture Detecton in the Analysis of Pesticides," J. Chromatogr. Sci., 12, 695-700 (1974). Wilson, N. K., and M. Anderson, "Carbon-13 and Proton Magnetic Resonance Studies of Chlorinated Biphenyls," in Mass Spectrometry and NMR Spectroscopy in Pesticide Chemistry, R. Haque and F. J. Biros (Eds.), Plenum Press, New York (1973). pp. 197-218. Winslow, S. G., and H. B. Gerstner (Eds.), Polychlorinated Biphenyls, Poly brominated Biphenyls,, and Their Contaminants: A Literature Compilation 1965-1977, Oak Ridge National Laboratory, Oak Ridge, Tennessee (1978). Wolbach, C. D., W. F. Fitch, N. Flynn, and B. Markoja, "Analytical Results of a PCS Test Incineration," in Second Symposium on Process Measurements for Environmental Assessment (February1980, Atlanta), J. C. Harris (Ed,), Environmental Protection Agency, Contract No. 68-02-3111, Task 108, Report No. EPA-600/9-81-018 (1981). Wolff, M. S., J. Thornton, A. Fischbein, R. Lilis, and I. J. Selikoff, "Disposition of Polychlorinated Biphenyl Congeners in Occupationally Exposed Persons," Toxieol. Appl. Pharmacol., 62, 294-306 (1982). Yakushiji, T., I. Watanabe, K. Kuwabara, and S. Yoshida, "Identification of Low Chlorinated Biphenyls in Human Milk by Gas Chromatography-Mass Spectrometry," J. Chromatogr., 154, 203-210 (1978). 124 �Yarbrough, J. D., and M. R. Wells, "Vertebrate Insecticide Resistance: The in vitro Endrin Effect on Succinic Dehydrogenase Activity on Endrin-Resistant and Susceptible Mosquitofish," Bull. Environ. Contain. Toxicol., 6(2), 171-176 (1971). Young, S. J. V., and J, A. Burke, "Micro Scale Alkali Treatment for Use in Pesticide Residue Confirmation and Sample Cleanup," Bull. Environ. Contain. Toxicol., 7(2/3), 160-167 (1972). Zelenski, S. G., J. Tashiro, and L. R. Worthen, "A Gas Chromatographic Method of Quantitating DDT in the Presence of Interfering Polychlorinated Biphenyl," J. Chromatogr.. 84, 67-73 (1973). Zell, M., H. J. Neu, and K. Ballschmiter, "Identifizierung der PCB-Komponenten durch Retentionsindexvergleich nach Kapillar-Gaschromatographie," Chemosphere, 2/3, 69-76 (1977). Zell, M., H. J. Neu, and K. Ballschmiter, "Single Component Analysis of Polychlorinated Biphenyl (PCB)- and Chlorinated Pesticide Residues in Marine Fish Samples," Fresenius Z. Anal. Chem., 292, 97-107 (1978). Zimmerli, B., B. Marek, and H. Sulser, "Determination of Chlorinated Biphenyls and Chloride Pesticides," Mitt. Geb. Lebensmittclunters. Hyg., 64(1), 70-79 (1973); Chem. Abst.. 79, 124762a (1973). Zimmerli, B., "Beitrag zur Bestimmung von Umweltkontaminantien Mittels der Hydrierenden Reaktionsgaschromatographie," J. Chromatogr., 88, 65-75 (1974). Zitko, V., and P. M. K. Choi, "PCB and £,£'~DDE in Eggs of Cormorants, Gulls, and Ducks from the Bay of Fundy, Canada," Bull. Environ. Contam. Toxicol., 7(1), 63-64 (1972). Zitko, V., "Problems in the Determination of Polychlorinated Biphenyls," Intern. J. Environ. Anal. Chem.. 1, 221-231 (1972). Zitko, V., "Chromatography of Chlorinated Paraffins on Alumina and Silica Columns," J. Chromatogr., 81, 152-155 (1973). Zitko, V., "The Detection of Aromatic and Chlorinated Hydrocarbons in Marine Lipids," J. Am. Oil Chemists' Soc., 52 131A (1975). Zitko, V., "The Interference of Aromatic Hydrocarbons in the Determination of PCB's," in Proceedings of the Joint Conference on Sensing of Environmental Pollutants, 4th, New Orleans (1977). pp. 757-760. Zitko, V., 0. Hutzinger, and P. M. K. Choi, "Determination of Pentachlorophenol and Chlorobiphenylols in Biological Samples," Bull. Environ. Contam. Toxicol., 12(6), 649-653 (1974). Zitko, V., "Levels of Chlorinated Hydrocarbons in Eggs of Double-Crested Cormorants from 1971 to 1975," Bull. Environ. Contam. Toxicol., 16(4), 399405 (1976). Zobel, M. G. R., "Quantitative Determination of Polychlorinated Biphenyls— A Computer Approach," J. Assoc. Offic. Anal. Chem., 57(4), 791-795 (1974). 125 �TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 1. REPORT NO. 3. RECIPIENT'S ACCESSION NO, EPA-560/5-82-005 4. TITLE AND SUBTITLE 5. REPORT DATE Methods of Analysis for By-Product PCBs—Literature Review and Preliminary Recommendations 7. AUTMQR(S) October 1982 6. PERFORMING ORGANIZATION CODE 4901-A(51) 8. PERFORMING ORGANIZATION REPORT NO. Mitchell D. Erickson John S. Stanley Interim Report No. 1 9. PERFORMING ORGANIZATION NAME AND ADDRESS Midwest Research Institute 425 Volker Boulevard Kansas City, Missouri 64110 10. P R O G R A M E L E M E N T NO. 11. CONTRACT/GRANT NO, 68-01-5915 Task 51 12. SPONSORING AGENCY NAME AND ADDRESS U.St Environmental'Protection Agency Office of Toxic Substances Field Studies Branch, TS-798 Washington. D.C. 20460 13. TYPE OF REPORT AND PERIOD COVERED Interim (March-April 1982) 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES Frederick W. Kutz, Project Officer David P. Redford, Task Manager 16. ABSTRACT A review of the literature on polychlorinated biphenyl (PCB) analysis and recommendations for methods to determine by-product PCBs in commercial products and other matrices is presented. This report was prepared to assist EPA in formulating a rule regulating by-product PCBs. The published literature on PCB analysis is critically reviewed. Several hundred references are cited in a bibliography. The review if subdivided into extraction, cleanup, determination, data reduction, confirmation, screening, quality assurance, and by-product analysis sections. The determination section includes TLC, HPLC, GC (POC and COC) , GC detectors (ECD, FID, HECD, EIMS, and other MS) and nonchromatographlc analytical methods (NMR, IR, electrochemistry, NAA, and RIA). Techniques applicable to analysis of commercial products, air, and water for by-product PCBs are discussed. The final section of this report presents a recommended overall primary analytical scheme. KEY WORDS AND DOCUMENT ANALYSIS IZ: DESCRIPTORS 3. b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group 19. SECURITY CLASS (This Report) 21. NO, OF PAGES PCBs Polychlorinated Biphenyls Analysis By-product PCBs Review 18. DISTRIBUTION STATEMENT Unclassified 20. SECURITY CLASS (This page) Unlimited Unclassified EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE 135 22. PRICE � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 186 Folder The folder containing the original item. 5384 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Erickson, Mitchell D. John S. Stanley Description An account of the resource <strong>Corporate Author: </strong>Midwest Research Institute Date A point or period of time associated with an event in the lifecycle of the resource October 12 1982 Title A name given to the resource Methods of Analysis for By-Product PCBs - Literature Review and Preliminary Recommendations ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/62f7399c340eba5e215718304bfbac19.pdf e3f2b4b55454ed73c4ece6477b153742 PDF Text Text Ram D Number °5385 Author Erickson, Mitchell D, D ^ot Scanned Midwest Research Institute Report/Article TitlB Analytical Methods for By-Product RGBs - Preliminary Validation and Interim Methods Journal/Book TitlB Year 1982 Month/Day October 11 Color D Number of Images 246 UBSCriDtOn NOtBS Task 51, Interim Report No. 4, EPA Contract No. 68-01-5915, MRI Project No. 4901 -A(51) Friday, March 08, 2002 Page 5385 of 5427 �Office of Toxic Substances Washington DC 20460 United States Environmental Protection Agency EPA-560/5-82-006 October, 1982 Toxic Substances v>EPA Analytical Methods for By-Produet PCBsPreliminary Validation and Interim Methods 100.0 T —T"—T 100.0 100.0 l3c 12 H 6 Ci 4 T 280 285 290 T-—r 295 300 305 Da I tons 3tO 315 320 325 �DISCLAIMER This document has been reviewed and approved for publication by the Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Approval does not signify that the contents necessarily reflect the views and policies of the Environmental Protection Agency, nor does the mention of trade names or commercial products constitute endorsement or recommendation for use. �ANALYTICAL METHODS FOR BY-PRODUCTS PCBs—PRELIMINARY VALIDATION AND INTERIM METHODS By Mitchell D. Erickson, John S. Stanley, Kay Turman, Gil Radolovich, Karin Bauer, Jon Onstot, Donna Rose, and Margaret Wickham Midwest Research Institute 425 Volker Boulevard Kansas City, MO 64110 TASK 51 INTERIM REPORT NO. 4 EPA Contract No. 68-01-5915 MRI Project No. 4901-A(51) October 11, 1982 For U.S. Environmental Protection Agency Office of Toxic Substances Field Studies Branch TS-798 Washington, D.C. 20460 Attn: Dr. Frederick W. Kutz, Project Officer Mr. David P. Redford, Task Manager �PREFACE This report presents the results of a preliminary method validation accomplished on MRI Project No. 4901-A, Task 51, "PCB Analytical Methodology Task," for the Environmental Protection Agency (EPA Prime Contract No. 6801-5915) during the period April 24 to August 31, 1982. The document was prepared by Drs. Mitchell D. Erickson (Task Leader) and John S. Stanley and Ms. Kay Turman, with assistance from Kathy Funk, Cindy Melenson, and Gloria Sultanik. The laboratory work was conducted by Kay Turman, and Donna Rose, with assistance from Steven Turner. The gas chromatography/mass spectrometry analysis was performed by Gil Radolovich, Margaret Wickham, Jon Onstot, and Arbor Drinkwine. Statistical analysis of the data was provided by Karin Bauer. Editorial comments were provided by Rudena Mallory and Jeanne Robson. The EPA Task Manager, David Redford, has been especially helpful and encouraging. The helpful comments of Ann Carey, Frederick W. Kutz, and John Smith, all of EPA, are also appreciated. SEARCH INSTITUTE C J/mn E. Going, Head Environmental Analysis Section Approved: James L. Spigarelli, Director Analytical Chemistry Department ill �CONTENTS Preface Figures Tables 1. 2. 3. 4. 5. iii vii ix Introduction Summary Experimental Preparation of PCB stock solutions and working standards. Gas chromatography/electron impact mass spectrometry. . . Determination of PCB response factors (GC/EIMS) Validation of method steps Validation with product and product waste samples . . . . Method Validation Preparation of analytical methods Gas chromatography/mass spectrometry of PCBs Validation of selected method steps Validation of product and product waste method with samples Discussion References ' i 1 2 3 3 8 8 19 20 24 24 26 38 45 56 70 Appendix A - Supplementary GC/EIMS Data on PCB Congeners A-l Appendix B - Analytical Method: The Analysis of Incidentally Generated Chlorinated Biphenyls in Commercial Products and Product Wastes B-l Appendix C - Analytical Method: The Analysis of Incidentally Generated Chlorinated Biphenyls in Air C-l Appendix D - Analytical Method: The Analysis of Incidentally Generated Chlorinated Biphenyls in Industrial Wastewater D-l �FIGURES Number 1 2 3 4 5 6 1 8 9 Plot of average response factor versus homolog for 77 PCB congeners 27 Plot of response factor per isomer versus homolog for 77 PCB congeners 28 Plot of response factor per isomer versus homolog for 77 PCB congeners, determined on a single day 32 Retention times of 77 PCB congeners relative to 3,3*4,4'tetrachlorobiphenyl-de (RRT of 1.00) 36 Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS) chromatogram or the calibration standard solution required for quantitation of PCBs by homolog 39 Reconstructed ion chromatogram for SIM analysis of the CMA-A sample no. 2110 59 SIM ion plots for monochlorobiphenyls (188 and 190 Daltons) and the 13C6-monochlorobiphenyl surrogate (194 Daltons) in CMA-A sample no. 2110 60 SIM ion plots for dichlorobiphenyls (222 and 224 Daltons) in CMA-A sample no. 2110 61 SIM ion plots for trichlorobiphenyls (256 and 258 Daltons) in CMA-A sample no. 2110 62 10 SIM ion plots for tetrachlorobiphenyls (290 and 292 Daltons), 3,3',4,4'-tetrachlorobiphenyl-d6 (298 Daltons), and the 13C12-tetrachlorobiphenyl surrogate (304 Daltons) in CMA-A sample no. 2110 63 11 SIM ion plots for pentachlorobiphenyls (326 and 328 Daltons) in CMA-A sample no. 2110 64 SIM ion plots of hexachlorobiphenyls (360 and 362 Daltons) in CMA-A sample no. 2110 65 12 13 SIM ion plots of heptachlorobiphenyls (394 and 396 Daltons) in CMA-A sample no. 2110 66 vii �FIGURES (continued) 14 15 16 SIM ion plots of octachlorobiphenyls (428 and 430 Daltons) and the ^Cj^-octachlorobiphenyl surrogate (442 Daltons) in CMA-A sample no. 2110 67 SIM ion plots of nonachlorobiphenyl (464 and 466 Daltons) in CMA-A sample no. 2110 68 SIM ion plots of decachlorobiphenyl (498 and 500 Daltons) and the 13C12-decachlorbiphenyl (510 Daltons) in CMA-A sample no. 2110 69 Vlll �TABLES (continued) Number 34 35 Pagj PCB Concentration (pg/g) of CMA-A Samples Treated With Various Cleanup Procedures (Surrogate Compound Correceted) 55 Recovery ( ) of Carbon-13 Labeled Surrogate Compounds % From Diarylide Yellow and Phthalocyanine Blue and Green Pigments 57 XI �TABLES Number Page 1 Numbering of PCB Congeners 5 2 Working Solutions for PCB Response Factors 6 3 Approximate Concentration of Individual PCB Congeners in Dilute Working Standards 7 Concentrations of Congeners in PCB Calibration Standards (ng/ml) 9 4 5 6 7 8 9 Composition of Surrogate Spiking Solution (SS100) Containing 13C-Labeled PCBs 10 Operating Parameters for Capillary Column Gas Chromatographic System 11 DFTPP Key Ions and Ion Abundance Criteria for Quadrupole Calibration 12 Operating Parameters for Quadrupole Mass Spectrometer System 13 Operating Parameters for Magnetic Sector Mass Spectrometer System 14 10 Characteristic Single lion Monitoring (SIM) Ions for PCBs . 15 11 Limited Mass Scanning (LMS) Ranges for PCBs 16 12 Characteristic Ions for 13C-Labeled PCB Surrogates 17 13 Pairings of Analyte, Calibration, and Surrogate Compounds . 18 14 Commercial Product and Product Waste Stream Samples Received for Preliminary Method Validation Studies. . . . 21 15 Preliminary Method Validation Samples 22 16 Comparison of Average Relative Response Factors (RRF) for 77 Commercially Available PCB Congeners Measured Over Several Days as Four Replicates Each Versus Single Measurements of All Congeners in a Single Day 30 IX �TABLES (continued) Number 17 18 Page Average Relative Response Factors (RRF) for PCB Congeners in Solution 1 Measured as Replicates on a Single Day and as Single Measurements for Day-to-Day Basis Measured Average Response Factor (RRF) and Corresponding Upper and Lower 95% Confidence Limits 19 20 31 Relative Response Factors Measured Versus 3,3',4,4'-Tetrachlorobiphenyl-de by Electron Impact Mass Spectrometry Quadrupole (Finnigan 4023) and Magnetic Sector (Varian (MAT 311A) Instruments 34 35 Relative Retention Time (RRT) Ranges of PCB Homologs Versus d6-3,3' ,4,4'-Tetrachlorobiphenyl 37 21 Recovery Data for Acid Cleanup 40 22 Recovery Data for Florisil Column Protocol Cleanup 41 23 Recovery Data for Florisil Slurry Protocol Cleanup 42 24 Recovery Data for KOH Protocol Cleanup 43 25 Recovery Data for Alumina Protocol Cleanup 44 26 Uncorrected PCB Concentrations (pg/g) in CMA-A Samples. . . 46 27 Corrected PCB Concentrations (pg/g) in CMA-A Samples. . . . 47 28 Uncorrected and Corrected PCB Concentrations (|Jg/g) in CMA-E Sample (Dilution Preparation) 49 Uncorrected PCB Concentration (|Jg/g) in the CMA-A Sample Matrix (Internal Standard Calculation) 50 Corrected PCB Concentration ((Jg/g) in the CMA-A Sample Matrix 51 Uncorrected PCB Concentration (|Jg/g) of Spiked CMA-A Samples Determined by the Internal Standard Quantitation Method 52 Corrected PCB Concentration (|Jg/g) of Spiked CMA-A Samples Determined by Surrogate Recovery Correction 53 29 30 31 32 33 PCB Concentration (|Jg/g) of CMA-A Samples Heated With Different Cleanup Procedures (Internal Standard Quantitation) 54 �SECTION 1 INTRODUCTION The Environmental Protection Agency (EPA) is in the process of preparing rules for regulation of certain polychlorinated biphenyls (PCB) which are generated as by-products in the manufacture of commercial products (U.S. EPA, 1982). This regulation is under the Toxic Substances Control Act (PL 94-469), and EPA's Office of Toxic Substances has been assigned the task of preparing the rule. As part of the rule, EPA is suggesting analytical methods for PCBs in air (stack gas and fugitive emissions), wastewater, product waste streams, and final products to assist organizations seeking an exclusion under this rule. To assist EPA in this mission, Midwest Research Institute (MRI) was asked to prepare appropriate analytical methodologies. A literature review and recommendation of general analytical approaches (Erickson and Stanley, 1982; Stanley and Erickson, 1982) constituted the first phase. The second phase, reported here, covers initial method validation and preparation of interim methods. As part of the method validation, four 13C-PCB surrogates were synthesized and are reported separately (Roth et al., 1982). The third phase will involve interlaboratory validation and method refinement. This report presents the initial results of method validation for analysis of by-product PCBs in product and product waste samples. Specifically, gas chromatography/electron impact mass spectrometry retention time and response factor data for 77 PCB congeners for two different gas chromatography/ mass spectrometry systems, recoveries from several proposed cleanup steps, and recoveries from industrial samples using a variety of the method options are presented. �SECTION 2 SUMMARY The objective of this study was to present EPA with appropriate methodologies for the analysis of by-product PCBs in commercial products, product waste streams, wastewaters, and air. In addition, EPA requested preliminary analytical studies to provide data in support of the proposed methods. This document presents proposed analytical methods for the analysis of by-product polychlorinated biphenyls in commercial products and product waste streams (Appendix B), wastewater (Appendix C), and air (Appendix D). The proposed methods are based on determination of PCBs using gas chromatography/ electron impact mass spectrometry (GC/EIMS). Capillary column gas chromatography (CGC) and packed column gas chromatography (PGC) are presented as alternate approaches. The 13C-labeled PCB surrogates are added to samples prior to any sample preparation to allow method flexibility for a wide spectrum of matrices. Recovery of the surrogates will allow determination of the quality of analytical data. This method is valid only if the surrogates are thoroughly incorporated into the matrix. The analytical method for commercial products and product waste streams relies heavily on a strong quality assurance program consisting of use of four 13C-labeled surrogate PCBs, blanks, duplicates, spiked samples, and quality control samples. The analytical methods for water and wastewater are based on EPA Methods 608 and 625, revised to include the use of the 13Clabeled surrogates. Likewise, the air method is a revision of a proposed method for PCBs in air and flue gas emissions. This document presents relative response factors (RRF) of 77 PCB congeners which were used to determine the average RRF for PCBs by homolog. Statistical analysis of the data was performed to check the validity of the response factor data and to extrapolate RRFs for the unavailable congeners. Relative retention time (RRT) data for the 77 PCB congeners are also presented. The RRF and RRT data were determined on both magnetic sector and quadrupole mass spectrometer systems. Preliminary studies were undertaken to check the validity of the proposed methods for the analysis of PCBs in commercial products and product waste streams. Data are presented for analysis of individual cleanup procedures as well as for analysis of product and product waste samples. The data indicate that the proposed method is applicable and useful for analysis of the matrices studied. However, these studies are preliminary and additional validation is necessary and ongoing. �SECTION 3 EXPERIMENTAL The method validation was conducted in three stages: (a) determination of GC/EIMS parameters for 77 PCB congeners; (b) validation of individual method steps with clean matrices; and (c) validation of selected method options with real samples. PREPARATION OF PCB STOCK SOLUTIONS AND WORKING STANDARDS Source of Standards Seventy-seven PCB congeners were acquired from Ultra Scientific, Inc., Hope, Rhode Island, and Analabs, North Haven, Connecticut. Quality control gas chromatography/flame ionization detection (GC/FID) data for the specific isomers were requested to verify the 99% purity assigned to these compounds. The GC/FID data supported the reported purity. In addition, all available nuclear magnetic resonance spectra used for specific isomer identification were requested but not supplied. Weighing Procedures Accurate mass measurement required calibration of a Cahn microbalance with National Bureau of Standards (NBS) certified masses of 5 and 10 mg. The balance was calibrated with the NBS standards followed by calibration of an in-house working standard mass. The calibration of the microbalance with the NBS certified masses was witnessed by a representative of the MRI quality assurance office. The mass of the working standard was measured between all measurements of individual PCB isomers to ensure that the balance was operating accurately. A record of the measured working standard mass was kept in a laboratory notebook. The mean value for the working standard was 10.037 ± 0.002 mg (0.02% relative standard deviation). When all measurements were completed, the mass of the NBS certified standards was determined as a final measure of the accuracy of the Cahn microbalance. Preparation of Solutions Preparation of PCB standard stocks began after accurate performance of the Cahn balance was demonstrated with the certified NBS and daily working standard. An aluminum weighing pan was preshaped such that complete transfer of the weighing pan plus sample could be made directly into the appropriate dilution vessel. The Cahn balance was tared to compensate for the weight of the aluminum boat, and the PCB standards were added via a micro spatula. The mass of the particular PCB was determined with the Cahn balance. �The aluminum pan containing the PCB standard was transferred to the dilution vessel using clean forceps, taking care not to spill any of the sample. The dilution vessel was capped tightly until solvent was added. All PCB congeners were dissolved in toluene (Burdick and Jackson, distilled in glass). Masses of 0.1 to 5 mg were dissolved in a total of 1.0 ml toluene while masses of approximately 10 mg and greater were dissolved in 5.0 ml toluene. The solvent was delivered volumetrically by pipette. Room temperature and solvent temperature were recorded at the time of standard dissolution. Volumetric pipettes used for solvent delivery were calibrated so that the most accurate determination of analyte concentration could be calculated. Toluene was pipetted into a tared vessel, and the total mass was measured. Density of the solvent at the specific room temperature was used to calculate the actual volume dispensed. This calibration was performed for all pipettes used for volumetric delivery of solvent. The stock solutions were sonicated in an ultrasonic bath for at least 15 sec after the volumetric addition of toluene to ensure complete dissolution of the PCBs. The solution level was etched on the side of the dilution vessel as a means of detecting losses by evaporation. The individual PCB congeners were referred to by the congener number indicated in Table 1. The stable labeled PCBs, 3,3',4,4'-tetrachlorobiphenyl-d6, 4-chlorobiphenyl-13Cg, 3,3',4,4'-tetrachlorobiphenyl-13C12, 2,2',3,3',5,5',6,6,'• octachlorobiphenyl-13Ci2> and decachlorobiphenyl-13Ci2 were assigned congener numbers of 210 to 214, respectively, for the purpose of this work. Sample labels were generated in duplicate to identify the specific PCB isomer stock solution and to document entries in the laboratory notebook. Table 2 presents the dilute working solutions that were prepared for determination of the response factors for the PCB congeners. The working solutions were prepared as 10 ml total volume. Table 3 presents the approximate concentration of each congener that was in the dilute working standard used for response factor determination. Tetrachlorobiphenyl-dg was added to 1.0 ml of each solution as the internal standard. All stocks were added to the working solutions in volumes of 20, 200, 250, 400, 500, or 1,000 | l The syringes were calibrated at j. these volumes. Calibration of the 10-ml volumetric flasks used for working standards was accomplished by measuring the difference between the mass of the empty flask and the mass of the flask plus toluene added to the appropriate dilution mark. The density of toluene at the correct solvent temperature was used to calculate the final volume of each solution. The dilute working solutions were divided into multiple aliquots. One hundred micrograms of tetrachlorobiphenyl-de was added to each of the 1.0-ml aliquots of the solutions that were used to establish CGC/EIMS response factors. The remaining dilute working solutions were stored in at least four crimp seal vials and refrigerated. The solvent meniscus was marked in permanent form to note losses of solvents from evaporation or spills. All solutions, stock standards and working solutions, were stored in a refrigerator. All vials removed from storage were first brought to room temperature and then sonicated for at least 15 to 30 sec before removing any of the solution. �Mb. Structure No. Honoetilorobiohenyli 1 2 3 2 3 4 D<eh1arob1ph«ny!s 4 5 6 7 8 9 10 11 12 13 1415 2.2' 2.3 2,3' 2,4 2,4' 2,5 2,6 3,3' 3,4 3,4' 3,5 4,4' TrlehlaroMehtnyls 16 17 13 19 20 21 22 23 24 25 26 27 23 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 15 46 47 48 49 50 51 2,2', 3 2,2', 4 2,2', S 2,2', 6 2,3,3' 2,3,4 2,3,4' 2,3,5 2,3,6 2,3', 4 2, 3', 5 2, 3', 6 2,4,4' 2,4,5 2,4,6 2, 4 ' , 5 2, 4'. 5 2 ' , 3, 4 2', 3,5 3,3', 4 3,3',S 3,4,4' 3,4,5 3.4', S NTJMBERING OF PCB CONGENERS a Structure Ha, Tttnehl orofal phtnyl s 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 2,2'.5.5' 2,2'.5,6' 2,2', 5,6' 2.3.3',4 2,3,3'. 4' 2,3,3', S 2,3,3'. 5' 2.3.3'. 6 2.3,4,4' 2,3,4,5 2,3,4.6 2, 3, 4', 5 2, 3,4', 6 2,3,5,6 2, 3', 4, 4' 2. 3'. 4,5 2,3'. 4,5' 2,3'. 4, 6 2,3', 4 ' , 5 2,3' ,4'. 6 2,3' ,5, 5' 2,3',5'.6 2,4,4', 5 2,4,4' .6 2'. 3 4,5 3,3' 4,4' 3,3' 4,5 3,3' 4,5' 3,3' 5,5' 3. 4, 4 ' , 5 Pentachl orobl cheny 1 s 82 83 84 85 86 37 88 39 90 91 92 93 94 TttracMoromehwyli 95 96 97 2, 2', 3,3' 98 2 2' 3 4 2;2','3,4' 99 100 2. 2'. 3.5 101 2,2', 3, 5' 2,2',3,6 102 2.2' .3.6' 103 2,2',4,4' 104 2,2' ,4, 5 2,2', 4,5' 2,2', 4, 6 2,2', 4, 6' TABLE 1. Structure 2,2',3.3',4 2.2'. 3. 3', S 2,2', 3, 3' .6 2,2', 3, 4, 4' 2,2'.3.4.5 2, 2'. 3, 4, 5' 2,2', 3, 4, 6 2,2', 3, 4, 5' 2,2', 3, 4 ' , 5 2, 2', 3, 4'. 5 2.2* .3,5,5' 2,2', 3,5, 6 2,2'.3,5,5' 2, 2', 3.5', 5 2,2',3,6,6' 2,2'. 3'. 4.5 2,2',3'.4,5 2,2' ,4. 4'. 5 2 2' .4,4'. 6 2,2'. 4, 5, 5' 2,2' ,4, 5, 6' 2,2', 4,5'. 5 2,2' .4. 6,5' NO. Structure 161 162 163 164 165 166 167 168 169 2,3.3 I ,4,5',6 2, 3, 3'. 4 ' , 5, 5' 2, 3,3'. 4 ' , 5, 6 2, 3, 3', 4 ' , S ' , 6 2, 3,3', 5. 5 ' , 5 2. 3. 4, 4 ' , 5, 6 2,3', 4, 4 ' . 5. 5' 2,3',4,4',5'.S 3,3',4,4',5.S' Pentaehl orobi oheny 1 s 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 2,3, 3', 4,4' 2,3,3',4,5 2, 3, 3'. 4'. 5 2,3, 3'. 4, 5' 2.3, 3', 4, 6 2.3, 3'. 4' ,6 2.3.3' ,5,5' 2,3,3',5,6 2,3.3', 5', 6 2,3,4. 4'.5 2,3,4,4'. 6 2,3,4,5,6 2,3, 4 ' , 5, 6 2,3',4,4',5 2, 3', 4, 4 ' , 6 2,3' ,4, 5,5' 2,3', 4, 5 ' , 5 2' ,3, 3', 4,5 2' .3. 4, 4 ' . 5 2' .3. 4, 5, 5' 2'. 3, 4.5, 6' 3, 3' ,4, 4 ' , 5 3,3',4,5,5' Hexaehlorobiofienyls 128 129 130 131 132 133 134 135 '136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 2,2', 3,3', 4, 4 ' 2,2', 3, 3', 4, 5 2,2', 3,3', 4, 5' 2,2'. 3, 3'. 4,6 2,2', 3,3', 4, 6' 2,2'.3,3',5,5' 2,2'. 3, 3', 5, 6 2,2', 3, 3' ,5, 5' 2, 2'. 3,3'. 6, 6' 2,2* ,3,4,4', 5 2,2'.3,4,4 I .S 1 2, 2 ' , 3, 4, 4', 6 2, 2 ' , 3, 4, 4 ' . 6' 2,2', 3, 4, 5, 5' 2,2', 3, 4, 5, 6 2,2' ,3, 4. 5,6' 2,2', 3, 4, 5 ' , 6 2,2', 3. 4, 5, 5' 2,2'. 3. 4 ' , 5, 5' 2,2', 3. 4 ' , 5,5 2,2', 3, 4 ' . 5, 5' 2, 2' ,3, 4', 5 ' , 5 2,2',3,4',6,6' 2,2' .3,5, 5'. 5 2,2', 3, 5,6, 6' 2, 2 ' , 4, 4 ' , 5, 5' 2.2', 4, 4 ' , 5, 5' 2,2', 4 , 4 ' , 6, 5' 2,3,3', 4, 4 ' , 5 2,3, 3', 4, 4 ' , 5' 2,3, 3', 4, 4 ' . 5 2,3, 3'. 4, 5, 5' 2,3,3'. 4, 5.5 HexachlorobiBnefiyls Heotactilorobiohenyl s 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 2, 2', 3, 3' , 4 , 4 ' ,5 2, 2 ' , 3, 3', 4 , 1 ' , 5 2, 2', 3, 3', 1,5,5' 2, 2'. 3, 3 ' , 1,5, 5 2, 2 ' , 3, 3 ' , 1,5, 5' 2. 2'. 3, 3', 4 , 5 ' , 5 2, 2', 3, 3 ' , 4, 5, 5' 2.2',3,3',4',5,a 2, 2 ' , 3, 3', 5, 5 ' , 5 2, 2 ' , 3, 3 ' , 5, 5, 5' 2. 2'. 3, 4. 4 ' , 5, 5' 2, 2', 3, 4, 4 ' , 5,6 2, 2 ' , 3. 4, 4 ' , 5, 5' 2, 2'. 3, 4, 4 ' , = ' , 6 2, 2', 3, 4, 4 ' , i, 5' 2, 2 ' , 3, 4, 5, 5' ,6 2, 2 ' , 3, 4, 5. 6, 6' 2, 2 ' , 3, 4 ' , 5,5' ,5 2. 2 ' , 3, 4 ' , 5, 5, 6' 2, 3, 3', 4, 4 ' . 5,5' 2, 3, 3', 4, 4 ' , 5, 5 2, 3, 3' ,4, 4 ' , 5 ' , 5 2,3, 3', 4, S, 5 ' , a 2,3,3'.4',S,5',5 Octacnl orjb • cne«y 1 s 194 195 196 197 198 199 200 201 202 203 204 205 2 , 2 ' , 3, 3', 4 , 4' , 5 , 5 ' 2, 2', 3, 3' , 4 , 4 ' ,5,5 2 , 2 ' , 3 , 3 ' ,4,1', = , 5 ' 2, 2', 3, 3 ' , 4,1', 5, 5' 2, 2 ' , 3 , 3 ' , 1,5,5' ,5 2, 2 ' , 3, 3 ' , 4, 5, 5, 5' 2, 2 ' , 3,3', 4, 5' , 5 , 5 ' 2, 2 ' , 3, 3'. 4, 5. 5 ' . 5' 2, 2'. 3, 3', 5, 5' ,5,5' 2, 2 ' , 3, 4, 4 ' , 5, 5 ' . 6 2, 2 ' , 3, 4, 4' , 5 , 6 , 5 ' 2,3. 3', 4, 4 ' . 5, 5 ' , 5 Monsehtcrobi;ns''yl s 206 207 208 2. 2 ' , 3, 3 ' , 1,1', 5, s ' , 5 2. 2', 3,3', 4 , 1 ' , 5, 5, 5' 2, 2 ' , 3, 3'. 4, 5, 5 ' , 5,5' DecachloHjOionenyi 209 2,2',3,3'4,4',5,5'.5,5' Adapted from Ballschmiter K, Zell M. 1980. Analysis of polychlorinated biphenyls (PCB) by glass capillary gas chromatography. Composition of technical Aroclorand Clophen-PCB mixtures, Fresenius Z. Anal Chera 302:20-31. �TABLE 2. PCB Soln. homolog no. 1 Soln. no. 2 Soln. no. 3 Soln. no. 4 WORKING SOLUTIONS FOR PCB RESPONSE FACTORS Soln. no. 5 PCB congener no. Soln. Soln. Soln. Soln. no. 6 no. 7 no. 8 no. 9 Soln. Soln. Soln. Soln. no. 10 no. 11 no. 12 no. 13 1 2 3 Dichloro- 11 5 7 8 9 10 4 12 14 Trichloro- 29 21 31 26 24 28 18 33 30 Tetrachloro- 47 44 40 49 50 52 53 54 66 61 65 69 72 Pentachloro- 121 97 88 93 101 103 100 104 a 115 87 116 119 Hexachloro- 136 129 128 137 138 141 143 151 139 153 154 155 156 Heptachloro- 181 171 183 185 Octachloro- 195 194 198 200 202 204 Nonachloro- 207 208 206 Decachloro- 209 9 9 Soln. no. 14 Monochloro- 15 70,75,77 Total congeners a 10 7 6 6 5 5 4 4 3 3 3 3 Congener no. 112 was added to this solution but, on analysis, was determined to have a mass of 286 and appeared to be a diaminotrichlorobiphenyl. This congener was omitted from any further consideration. �TABLE 3. APPROXIMATE CONCENTRATION OF INDIVIDUAL PCB CONGENERS IN DILUTE WORKING STANDARDS5 PCB horaolog Concentration (pg/ml) Monochlorobiphenyl 50 Dichlorobiphenyl 50 Trichlorobiphenyl 50 Tetrachlorobiphenyl Pentachlorobiphenyl 100 Hexachlorobiphenyl 100 Heptachlorobiphenyl 100 Octachlorobiphenyl 200 Nonachlorobiphenyl 200 Decachlorobiphenyl a 100 200 Tetrachlorobiphenyl-de was added to all solutions as an internal standard at *> 100 |Jg/ml. �Preparation of Calibration Standard and Spiking Mixtures A mixture of 11 congeners was used for calibration. This solution was spiked into solvent for protocol step validation experiments and into product and product waste samples for standard addition experiments. These congeners were determined to be the best standards for quantitation calibration based on the average relative response factor for each PCB homolog, as will be discussed in Section 5. Table 4 presents the composition of the 11-component solutions that are specified as the calibration standards, CSxxx, where the xxx is used to encode the nominal concentration in nanograms per milliliter. A more concentrated solution was diluted as necessary to prepare spiked samples and the appropriate standards for GC/EIMS analysis. The internal standard, tetrachlorobiphenyl-dg, was added to all standards and final extracts before GC/ EIMS analysis. The standards contained the four 13C-labeled PCBs that were added from the spiking solution shown in Table 5. GAS CHROMATOGRAPHY/ELECTRON IMPACT MASS SPECTROMETRY The capillary gas chromatography parameters used are shown in Table 6. The quadrupole and magnetic sector mass spectrometer parameters used are shown in Tables 7 through 9. The characteristic ions for single ion monitoring and limited mass scanning are presented in Tables 10 through 12. All data generated for relative response factors and concentration levels of PCBs in sample extracts were calculated based on the area of the primary quantitation ion specified in Table 10. The quantitation ions for the 13Clabeled monochloro-, tetrachloro-, octachloro-, and decachlorobiphenyl were 194, 304, 442, and 510 Daltons, respectively. The pairings of analyte, calibration, and surrogate compounds are presented in Table 13. DETERMINATION OF PCB RESPONSE FACTORS (GC/EIMS) The response factors for 77 PCB isomers were determined by GC/EIMS using the working standards prepared as described in Tables 2 and 3. A high resolution capillary column (J&W Scientific Durabond DB-5, 15 m, 0.25 |Jm film thickness) was used for the separation of the PCB mixtures. Scanning mass spectrometry was used to calculate response factors for the PCB isomers present in each solution versus a known quantity of tetrachlorobiphenyl-dgThe quadrupole GC/EIMS system was tuned daily prior to any acquisition of data for PCB response factor calculations. The system was brought to operating temperature for at least 15 min. The fluorocarbon FC-43 was introduced to the ion source, and 176 and 502 Daltons were manually adjusted 'to a two-to-one ratio. This was accomplished by adjusting the multiplier voltage to 300 mV while monitoring 176 Daltons. A selected ion monitor acquisition was set up for 176 and 502 Daltons with a variance of 1 Dalton. The ratio of the two values was tuned to the two-to-one ratio as described above. The mass spectrometer was operated in the normal full scan acquisition mode after tuning with the FC-43. Approximately 100 ng of decafluorotriphenylphosphine was injected and the ratio of the values of 198/442 was monitored. �TABLE 4. CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)a Homolog Congener no. CS1000 CS100 CS050 CS010 1 1 1,040 104 52 10 1 3 1,000 100 50 10 2 7 1,040 104 52 10 3 30 1,040 104 52 10 4 50 1,520 152 76 15 5 97 1,740 174 87 17 6 143 1,920 192 96 19 7 183 2,600 260 130 26 8 202 4,640 464 232 46 9 207 5,060 506 253 51 10 209 4,240 424 212 42 4 255 255 255 255 1 211 (RS) 104 104 104 104 4 212 (RS) 257 257 257 257 8 213 (RS) 407 407 407 407 10 a 210 (IS) 214 (RS) 502 502 502 502 Concentrations given as examples only. �TABLE 5. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100) CONTAINING 13C-LABELED PCBs3 Congener no. Compound Concentration (|jg/ml) 211 104 212 ,3' (13C12)3 ,4,4'-tetrachlorobiphenyl 257 213 (13C12)2 ,2', 3, 3', 5, 5' ,6,6'-octachlorobiphenyl 214 a 1 13 (I1, 2', 3 ,4',5',6'- C6)4-chlorobiphenyl (13C12)decachlorobiphenyl Concentrations given as examples only. 10 395 502 �TABLE 6. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM Parameter Value Gas chromatograph Finnigan 9610 Column 15 m x 0.255 mm ID Fused silica Liquid phase DB-5 Liquid phase thickness 0.25 urn Carrier gas Helium Carrier gas velocity 45 cm/sec Injector On-column (J&W) Injector temperature Optimum performance Injection volume 1.0 Mlb Initial column temperature 110°C (2 min)c Column temperature program . 110° to 325°C at 10°C/min d Separator None Transfer line temperature 280°C (J&W) a Measured by injection of air or methane at 270°C oven temperature. b For on-column injection, follow J&W instructions regarding injection technique. c With on-column injection, the initial temperature equals the boiling point of the solvent; in this instance toluene. d C12Clio elutes at 270°C. Programming above this temperature ensures a clean column and lower background on subsequent runs. e Fused silica columns may be routed directly into the ion source to prevent separator discrimination and losses. 11 �TABLE 7. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA FOR QUADRUPOLE CALIBRATION Mass Ion abundance criteria 197 198 199 Less than 1% of mass 198 100% relative abundance 5-9% of mass 198 275 10-30% of mass 198 365 Greater than 1% of mass 198 441 442 443 Present but less than mass 443 Greater than 40% of mass 198 17-23% of mass 442 12 �TABLE 8. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM Parameter Value Mass spectrometer Finnigan 4023 Data system Incos 2400 Scan range 95-550 Scan time 1 sec Resolution Unit Ion source temperature 280°C Electron energy3 70 eV Trap current 0.2 mA Multiplier voltage -1,600 V Preamplier sensitivity 106 A/V a Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. 13 �TABLE 9. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER Parameter SYSTEM Value Mass spectrometer Finnigan MAT 311A Data system Incos 2400 Scan range 98-550 Scan mode Exponential Cycle time 1.2 sec Resolution 1,000 Ion source temperature 280°C pt Electron energy 70 eV Emission current 1-2 mA Filament current Optimum Multiplier -1,600 V a Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. 14 �TABLE 10. CHARACTERISTIC SINGLE ION MONITORING (SIM) IONS FOR PCBs Ion (relative intensity) Homo log Primary Secondary Ca2H9Cl 188 (100) 190 (33) CigHgCla 222 (100) 224 (66) 226 (11) C12H7C13 256 (100) 258 ( 9 9) 260 (33) Ci2H6Cl4 292 (100) 290 (76) 294 (49) Ci2HsCl5 326 (100) 328 ( 6 6) 324 (61) C12H4C16 360 (100) 362 (82) 364 (36) Ci2H3Cl7 394 (100) 396 ( 8 9) 398 ( 4 5) £12^2^-8 430 (100) 432 ( 6 6) 428 (87) Ci2HClg 464 (100) 466 (76) 462 (76) C 498 (100) 500 ( 7 8) 496 ( 8 6) 12CllO Source: a Tertiary _a Rote JW, Morris WJ. 1973. Use of isotopic abundance ratios in identification of polychlorinated biphenyls by mass spectrometry. J Assoc Offie Anal Chem 56(1):188-199. / None available. 15 �TABLE 11. LIMITED MASS SCANNING (LMS) RANGES FOR PCBs ft Mass range (Daltons) Compound C^Cl, 186-190 C12H8C12 220-226 C12H7C13 254-260 C12H6C14 288-294 C12H5C15 322-328 C12H4C16 356-364 Lj *i oXloL^J-V 386-400 \_, -i OJlOvjJ^Q 426-434 \j i O-H-W-L o 460-468 1210 494-504 C12D6C14 294-300 13 192-196 13 300-306 13 438-446 C612C6H9C1 C12H6C14 C12H2C18 13 506-516 C12C110 a Adapted from Tindall GW, Wininger PE. 1980. Gas chromatography-mass spectrometry method for identifying and determining polychlorinated biphenyls. J Chromatogr 196:109-119. 16 �TABLE 12. CHARACTERISTIC IONS FOR 13C-LABELED PCB SURROGATES Primary Ion (relative intensity) Secondary Tertiary 13 194 (100) 196 (33) a 13 304 (100) 306 (49) 302 (78) 13 442 (100) 444 (65) 440 (89) 13 510 (100) 512 (87) 514 (50) Compound C612C6H9C1 C12H6C14 C12H2C18 Ci2Cl10 a None available. 17 �TABLE 13. PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE Analyte Congener no. I 2,3 4-15 16-39 40-81 82-127 128-169 170-193 194-205 206-208 209 Calibration standard Compound 2-C12H9Cl 3- and 4-C12H9Cl C12HgCl2 C12H7C13 C^HsCls Cl^EUClg C12H3C17 C12H2Clg Ci2HCla C12C110 Congener no. 1 3 7 30 50 97 143 183 202 207 209 Compound 2 4 2,4 2,4, 6 2,2' ,4,6 2,2' ,3', 4, 5 2,2' ,3,4,5 ,6' 2,2' ,3', 4, 4',5',6 2,2' Q Q t 5,5' ,6,6' 2,2' 3 3' 4, 4', 5, 6, 6' r L Lior COMPOUNDS Surrogate Congener no. 211 211 211 212 212 212 212 213 213 213 214 Compound 13 C6-4 C6-4 13 C6-4 13 C12-3 ,3' ,4,4' 13 C12-3 ,3' ,4,4' 13 C12-3 ,3' ,4,4' 13 C12-3 ,3' ,4,4' 13 Ci2-2 ,2' ,3,3' ,5,5' ,6,6' 13 3 3* Ci2-2 ,2' ,0,0 ,5,5' ,6,6' 13 C12-2 ?' ,3,3' 5 5 r ,6,6' 1ft 13 �The response of 198 Daltons was 100% full scale and 442 Daltons was adjusted from 40 to 45% of the base peak. These criteria were met daily before data acquisition for response factor calculations was initiated. All working standards were brought to room temperature and sonicated before injection into the GC/MS system. Solution No. 1 was analyzed daily as a means of normalizing response factors calculated from day to day. This allowed some compensation for differences in sensitivity due to subtle changes in the mass spectrometer operation from day to day. Also, a solution of tetrachlorobiphenyl-de (internal standard) was analyzed separately. Four replicates of each working standard were analyzed to calculate variances of the response factors. The solutions were sonicated at least 15 sec prior to removal of sample for injection. The syringe and needle were rinsed with 200- to 300-|Jl of toluene between injections. The gas chromatograph was operated at 110°C for 2 min, and programmed at 10°C/min to 325°C. One microliter injections were made with a J&W on-column injection system. Helium carrier flow was adjusted to 45 cm/sec. The peak shape of the eluting PCBs was monitored. If excessive tailing was noted, the injection end of the fused silica capillary column was removed and shortened by at least 10 cm. Tables 6, 7, and 8 present the instrument and operating parameters that were used to measure the response factors for the individual PCB isomers in the working solutions. Response factors (RF) were calculated using the area of the peaks for these ions according to the equation: A M _ ~r PCB rj IS OT — Kr A IS WPCB A where MPCB = .IS = ..IS = PCB = Area of the quantitation peak of the specific PCB, Mass (in nanograms) of the internal standard injected, Area of the quantitation peak of the internal standard, and Mass (nanograms) of the specific PCB injected. All relative response factor data were subjected to Student's t-test at the 95% confidence level to test for significant differences for day-to-day and solution-to-solution variances. VALIDATION OF METHOD STEPS A limited number of experiments were completed as preliminary validation steps for the proposed method presented in Appendices B through D. The experiment included evaluation of several of the cleanup procedures using solvent spiked with the 13C-labeled surrogates and a mixture of PCB congeners representing each of the possible homologs. The laboratory cleanup procedures followed the protocol steps except where noted. One hexane solvent blank was analyzed by each procedure with the samples to monitor interferences and contamination. 19 �All samples were analyzed by CGC/EIMS in the full scan mode using the Finnigan 4023 system. Tables 6, 7, and 8 present the instrumental parameters. VALIDATION WITH PRODUCT AND PRODUCT WASTE SAMPLES Sources of Samples Product waste samples were received from Dow Chemical Company (Kent Hodges) and Vulcan Materials Company (Thomas Robinson) through the cooperation of the Chemical Manufacturers Association (Robert Fensterheim). These samples are aliquots of the materials used for the Chemical Manufacturers Association (CMA) round robin study (CMA, 1982). The CMA and associates supplied samples of chlorinated benzene waste streams, mixtures of chlorinated benzenes, composite waste streams from a chlorinated aliphatic process and a benzene column bottom sample. Table 14 presents an inventory of all the samples received. Product samples were received from the Dry Color Manufacturers Association (J. Lawrence Robinson and Maria DaRoche). These samples included diarylide yellow, phthalocyanine green, and phthalocyanine blue pigments that were used in the Dry Color Manufacturers Association (DCMA) round robin study of an analytical method, reported by the DCMA (1981). These samples are also included in the inventory in Table 14. The samples supplied by industry are examples of the samples which will be analyzed using the method in Appendix B. However, since no attempt was made to span the range of products and product wastes, the samples analyzed do not include all matrices which an analyst could encounter. Experimental Design Table 15 presents an overview of the preliminary method validation samples. The samples from Table 14 that were used for these studies included the chlorinated benzene waste stream, CMA-A; the benzene column bottom sample, CMA-E; and the yellow, blue, and green pigment samples, DCMA-1, DCMA-4, and DCMA-8, respectively. Blind quantitation standards and quality control samples were prepared by the MRI quality control staff either through spiked addition or by dilution of particular sample matrices. Other quality control procedures included the analysis of duplicate samples and blanks and the validation of cleanup steps. Two sets of samples were prepared and run at separate times. This first sample set us designated by numbers 10 through 110 and the second sample set is designated by numbers 2001 through 2210Q in Table 15. The sample preparations ranged from addition of the 13C-labeled surrogates followed by dilution and injection, to preparation of pigment samples via sulfuric acid dissolution and hexane extraction or methylene chloride extraction with Florisil cleanup. 20 �TABLE 14. COMMERCIAL PRODUCT AND PRODUCT WASTE STREAM SAMPLES RECEIVED FOR PRELIMINARY METHOD VALIDATION STUDIES3 Sample no. Quantity Sample description Sample source CMA-A 100 ml Chlorinated benzene waste stream CMA-B 100 ml Mixture of chlorinated benzenes Dow Chemical Co. with Aroclor 1254 spike CMA-C 100 ml Blind spike of CMA-B with the addition of 64 ppm of PCB isomers Dow Chemical Co. CMA-A 5 ml Vulcan Materials Co. CMA-B 5 ml CMA-C 5 ml CMA-D 5 ml CMA-E 5 ml Chlorinated benzene waste stream Mixture of chlorinated benzenes with Aroclor 1254 spike Blind spike of CMA-B with the addition of 64 ppm of PCB isomers Composite waste stream sample from a chlorinated aliphatic process Benzene column bottoms sample Diarylide yellow pigment Phthalocyanine green pigment Phthalocyanine blue pigment Phthalocyanine blue pigment Phthalocyanine green pigment DCMA DCMA DCMA DCMA DCMA DCMA-1 DCMA-4 DCMA-6 DCMA-8 DCMA-9 100 100 100 100 100 g g g g g Dow Chemical Co. Vulcan Materials Co. Vulcan Materials Co. Vulcan Materials Co. Vulcan Materials Co. a Aliquots of CMA-A, CMA-B, and CMA-C were received from two sources, who indicated that they were identical. MRI has assumed that both aliquots are the same. 21 �TABLE 15. PRELIMINARY METHOD VALIDATION SAMPLES Sample no. Description Preparation 10 20A 20B 60 110 2001 2005 2010 CMA-A CMA-A CMA-A Hexane blank CMA-E Hexane blank CMA-A3 CMA-A 0.1 g/10 ml hexane 0.1 g/10 ml hexane 0.1 g/10 ml hexane None None None • 0.1 g/1 ml hexane 0.1 g/1 ml hexane 0.1 g/1 ml hexane 0.5-0.2 g/1 ml hexane 0.1 g/1 ml hexane 0.1 g/1 ml hexane 0.1 g/1 ml hexane 0.1 g/1 ml hexane None DCMA-A DCMA-A (0.1 g) DCMA-B DCMA-B (0.1 g) Base Base (0.1 g) DCMA-B (1.0 g) DCMA-B (1.0 g) DCMA-B (1.0 g) DCMA-B (1.0 g) DCMA-B (1.0 g) DCMA-B DCMA-B DCMA-B DCMA-A DCMA-A DCMA-A None 2020 CMA-A 2025Q 2030 2040 2050 2060Q 2070Q 2080 2090 2100 2110 2120 2130 2135 2140 2150 2160 2170Q 2175 2180 2185 2190 2195 2200Q 2210Q CMA-A CMA-A + CS002 CMA-A + CS005 CMA-A + CS010, CMA-A + CSXXX CSxxx Blank, CMA-AD Blank, CMA-AD Blank CMA-A DCMA-13 DCMA-1 DCMA-1 DCMA-1 + no. 11 (50 ppm) DCMA-1 + no. 11 (20-80 ppm) DCMA-4 DCMA-4 DCMA-4' DCMA-8 DCMA-8* DCMA-8 CSxxx Dilution factor 1/100 1/100 1/100 None None None 1/10 1/10 1/10 1/10 1/10 1/10 1/10 1/10 None 1/10 1/10 1/10 1/10 1/10 1/10 1/100 1/100 1/100 1/200 1/200 1/100 1/100 1/100 1/50 1/50 1/50 None a No surrogates added to assess any background interferences for these compounds. b Prepared from aliquot received from Dow Chemical Company; all other CMA-A samples prepared from aliquot received from Vulcan Materials Company. 22 �The CMA-A and CMA-E samples were each analyzed after 1/10 or 1/100 dilution, depending on the operating sensitivity of the mass spectrometer. The CMA-A chlorinated benzene waste was the most extensively studied matrix of the available samples. Sample preparation included the simple dilution described above with and without the addition of the four surrogates. The samples prepared without surrogates allowed measurement of the background that might interfere with the four surrogate compounds. Duplicate samples of the CMA-A were analyzed at the same dilution in two separate experiments. The CMA-A matrix was also analyzed by standard addition methods with total spiked PCS levels of the 11-compound spiking solution (CS050) at approximately 70, 140, and 270 ng/sample. The CMA-A matrix was also prepared using the sulfuric acid and ethanolic KOH procedures discussed in Section 9.3.2 of Appendix D, Cleanup of the Analytical Method: The Analysis of By-Product Chlorinated Biphenyls in Commercial Product and Product Wastes (Appendix B). Variations of the analytical procedures used by the Dry Color Manufacturers Association (1981) for the analysis of PCBs in various pigments were also applied to the CMA-A matrix. The DCMA procedures included acid dissolution followed by hexane extraction from the acid (DCMA Preparation A) and Florisil treatment of the concentrated sample matrix (DCMA Preparation B). The homogenization and centrifugation steps required by the DCMA-B procedure were not included for the CMA-A matrix. All samples except those representing blanks were spiked with the surrogates at levels of 100 to 500 ng and were mixed thoroughly before beginning the sample preparation. The typical CMA-A sample size was 0.1 g. The diarylide yellow (DCMA-1), phthalocyanine green (DCMA-4), and pthalocyanine blue (DCMA-8) pigments were also studied in these preliminary validations. The yellow pigment was prepared according to the recommended DCMA-B procedure, while the green and blue pigments were analyzed following the DCMA-A procedure. The preparation of the pigments followed the DCMA procedures except that the preparation was scaled to 1 g of the yellow pigment instead of the recommended 5 g. Blanks, duplicates, and spiked samples were also analyzed with the set of DCMA samples. Sample Analysis All extracts were analyzed by capillary column gas chromatography/electron impact mass spectrometry (CGC/EIMS). Limited mass scanning (LMS) or selected ion monitoring (SIM) mass spectrometry methods were used for extract analysis, depending on the level of PCBs in the sample extracts and the complexity of the matrix. The parameters for analysis via CGC/LMS and CGC/ MS-SIM are presented in Tables 6 through 13. 23 �SECTION 4 METHOD VALIDATION PREPARATION OF ANALYTICAL METHODS Analytical methods were prepared for the analysis of by-product PCBs in: * Commercial products and product wastes (Appendix B). * Air (Appendix C). * Industrial wastewater (Appendix D). The analysis of commercial products and product wastes was covered in one method since the diversity of matrices in both categories dictates the same generalized approach. Air was defined to include stack gases, fugitive emissions, and static (room, other container, or outside) air. Commercial Products and Product Wastes Method The objective was to devise an analytical method suitable for enforcement of the regulation concerning by-product PCBs in commercial products and product wastes. A detailed rationale for selection of the techniques used in the method may be found in a separate report (Erickson and Stanley, 1982). Sample Workup-The general approach taken with sample preparation (collection, preservation, extraction, and cleanup) was to provide a framework within which any reasonable technique could be used. This is the only acceptable approach to a method designed to cover "any" matrix. , The use of 13C-labeled recovery surrogates in conjunction with GC/EIMS was judged to be the most suitable approach (Erickson and Stanley, 1982; Stanley and Erickson, 1982; Roth et al., 1982). Using the recovery surrogates, any losses of PCBs would be detected and could be corrected for in the calculation of the PCB concentration. 24 �When surrogates are not fully incorporated into the matrix, their recovery will not be representative of the analyte PCB recoveries and recovery assessment will not be possible. It is incumbent upon the analyst to recognize this problem and use good scientific judgment with samples that present a potential problem. Nonextractable solid polymers may be an example of a matrix presenting incorporation problems. PCB Determination-As discussed elsewhere (Erickson and Stanley, 1982; Stanley and Erickson, 1982) , GC/EIMS appears to be the only acceptable general technique for determining PCBs in commercial products and product wastes. The use of either capillary or packed column GC is permitted. While strong arguments are presented for both techniques (Stanley and Erickson, 1982), the analytical results should be comparable for both techniques provided proper instrument calibration and operation, analytical, and quality control procedures are followed as described in the analytical methods. Quantitation-The analytical objective of these methods is to determine if the sample contains quantifiable PCBs and, if so, at what concentration. On the assumption that a general knowledge of the congener distribution is important, reporting of the concentration by homolog is proposed in the reporting form. Since a "total PCB" value is also important for summary and comparative purposes, space for this value is also provided on the reporting form. Other reporting formats, including "largest isomer or resolvable peak" or "all peaks greater than a regulatory value," may be easily accommodated using different tabulations and reporting procedures. The PCB concentrations found may be lower than the actual value due to nonquantitative recovery during extraction or cleanup. The measured recoveries of the surrogates may be used to derive a corrected concentration. The analyst must take care that the surrogates are thoroughly incorporated into the matrix prior to extraction, as discussed above. The analyst must also guard against improper corrections because of errors in surrogate quantitation. These errors may arise from background interferences. A more thorough discussion of quantitation options is presented in a previous report (Erickson and Stanley, 1982). Air Method The sample collection, preservation, extraction, and cleanup aspects were taken from the work of Haile and Baladi (1977). The determination, using GC/ EIMS, is identical to that in the commercial products and product wastes method except that recovery surrogates are not used. Wastewater Method The water method is a direct modification of the commercial products and product wastes method. As noted in this method, the cleanup and extraction procedures for EPA Methods 608 (U.S. EPA, 1979b) and 625 (U.S. EPA, 1979a) may be used. It is anticipated that, unless conditions dictate otherwise, most analysts will choose this option. 25 �Quality Control Each method includes a strong quality control (QC) section. Given the complexity of the matrices and complexity of the analyte (209 compounds), the need for QC is evident. The various aspects of the QC section were designed assuming a reasonably large (10 to 100) batch of samples. For small batches of samples, the percentage of effort spent on QC can become sizeable. Alternate Methods The methods presented here are intended to be primary methods capable of generating the best quality data technologically feasible. The development and acceptability of secondary (alternate, equivalent, or screening) methods is not addressed in this report. GAS CHROMATOGRAPHY/MASS SPECTROMETRY OF PCBs Analysis for PCBs requires the use of selected representative standard compounds since all 209 congeners are not available. One of the major disadvantages of many instrumental methods for PCB analysis is the large variance of the instrumental response factors for PCB congeners, both within a homolog and between homologs. These large differences in response factors create problems in selecting representative compounds for quantitation purposes. The response factors of 77 of the possible 209 PCB congeners measured by GC/EIMS are presented in Tables 16 and 17. The data suggests that the EIMS response factor variance among PCB congeners is small relative to other detectors such as the electron capture detector or negative chemical ionization mass spectrometry. Relative Response Factors Quadrupole Mass Spectrometer-The relative response factors (RRF) of the 77 PCB congeners were determined with the Finnigan 4023 quadrupole mass spectrometer as discussed in the experimental section. The RRFs were determined two ways to assess the effects of instrumental variability. The replicate RRF determinations are the average of four replicate analyses for each of the PCB congeners, all determined on a single day to assess the variability of the measurement. The single RRF determinations are single values from an experiment in which all 14 solutions containing all 77 congeners were run on one day to minimize instrumental variability with time. The data are presented in Appendix A. The RRFs vary from approximately 0.2 for decachlorobiphenyl to 4.1 for 2-chlorobiphenyl. Figures 1 and 2 present a visual comparison of average replicate and single RRFs of PCB congeners determined as replicate measurements and as single measurements . 26 �4.5 4.0 Quadrupole Mass Spectrometer 3.5 3.0 ! — 2 o 2.5 cd M-l <1> CO 0 2.0 . CO 0) to CO eu | c I. O »PH 1.0 1 2 7 3 0.5 J_ 3 I 4 5 6 7 Homolog (degree of chlorination) 10 Figure 1. Plot of average response factor versus homolog for 77 PCB congeners. Each average is the mean response per congener, i.e., mean of four replicates with the Finnigan 4023 quadrupole mass spectrometer. This plot indicates the number of data points that overlap for specific isomers. �3.Or— 1 Quadrupole Mass Spectrometer 2.5 2.0 o 0} 1.5 N5 00 1.0 4 2 1 2 1 2 2 1 7 0.5 - _L _1_ _L JL X 3 4 5 6 7 10 Homolog (degree of chlorination) Figure 2. Plot of response factor per isomer versus homolog for 77 PCB congeners, determined on a single day. Each value is representative of single measurements of each congener with the Finnigan 4023 quadrupole mass spectrometer. This plot indicates the number of data points that overlap for specific isomers. �Table 16 is a summary of the RRF data, where the replicate and single measurements are averaged over all measured isomers for a homolog. The relative standard deviation (Table 16) for the replicate measurements reflects the variance of the average RRF for each isomer within a homolog. The absolute area of the internal standard, Congener No. 210, varied by only 4.4% for all solutions during the single day experiment, as compared to 9.9% for the 7 days required to complete the replicate analyses. The relative standard deviations based on the four replicate analyses for each of the PCB congeners, ranged from 0.4 to 9.1%, indicating the reproducibility of the injection for each solution. The average response factors from replicate determinations and single measurements were subjected to a Student's t-test to determine if there were any significant differences in measured response factors. No significant difference was found for the average response factor values for any of the PCB homologs except the heptachlorobiphenyl isomers. A more detailed presentation of the Student's t-test for these values is presented in Table A-2 of Appendix A. A solution of 3,3',4,4'-tetrachlorobiphenyl-de (Congener No. 210) and Solution No. 1 (Table 2) were both analyzed daily. The solution of Isomer No. 210 was used to tune the quadrupole mass spectrometer to the desired working conditions. Solution No. 1 was used to determine fluctuations of response factors from day to day due to differences in instrumental operating parameters. Table 17 presents the data for single day replicate measurements and day-to-day determination of the response factors for the PCB congeners in Solution No. 1. The relative standard deviations calculated for the single day measurements are considerably lower than the relative standard deviations from day-to-day analyses. This is a reflection of the reproducibility on the part of the operator as well as of the stability of the quadrapole mass spectrometer system on a given day. The relative standard deviation calculated for day-to-day analyses is indicative of the variation that might be expected for routine analysis of PCBs. A Student's t-test of the Solution No. 1 data (Table 17) indicated that there are significant differences in response factors from day to day compared to single day measurements for PCB Congener Nos. 1, 11, 29, and 207. A more detailed presentation of this t-test is presented in Table A-3 of Appendix A. Magnetic Sector Mass Spectrometer— The RRFs for the 77 PCB congeners were also determined with a Varian MAT 311A double focusing magnetic sector mass spectrometer. The RRF values were determined by single measurements of all congeners on a single day. The data are presented in Appendix A and summarized in Figure 3. Extrapolation of Response Factor Data to All Congeners-Since all 209 PCB congeners were not available for determination of RRFs, it was necessary to extrapolate the average RRF data to project the range of response factors that might be encountered. This extrapolation was based on the assumption that the number of measured isomers (n) are a representative sample of the entire set of the possible isomers (N). Thus it was assumed that the mean for the measured isomers (n) is an unbiased estimate of the mean for the possible isomers (N). 29 �TABLE 16. AVERAGE RELATIVE RESPONSE FACTORS (RRF) FOR 77 COMMERCIALLY AVAILABLE PCB CONGENERS MEASURED OVER SEVERAL DAYS AS FOUR REPLICATES EACH AND RRF FOR SINGLE MEASUREMENTS OF ALL CONGENERS IN A SINGLE DAY PCB homolog No. of isomers RRF from replicate measurements Relative standard deviation ( ) % RRF from single b measurement Relative standard deviation ( ) % 3 3.331 19.3 2.739 9.3 Dichloro- 10 2.027 22.0 2.048 15.7 Trichloro- 9 1.573 21.7 1.592 18.1 Tetrachloro- 16 0.950 18.4 0.946 20.0 Pentachloro- 12 0.720 16.7 0.725 17.6 Hexachloro- 13 0.513 15.1 0.500 19.1 Heptachloro- 4 0.361 6.6 0.308 8.0 Octachloro- 6 0.253 11,9 0.224 17.3 Nonachloro- 3 0.229 14.7 0.188 16.2 Decachloro- 1 0.213 2.8 0.179 Monochloro- - a Four replicate measurements of the RRF were made for each isomer. For example, the three monochlorobiphenyl isomers were measured four times each. Hence, the RRF and relative standard deviation ( ) were calculated from 12 distinct % values. b A single measurement for each of the 77 PCB congeners was completed in a single day. Hence, the RRF reported is the average of one measured RRF for each isomer within a homolog. For example, the RRF and relative standard deviation ( ) reported for the monochlorobiphenyls were calculated from three distinct % values. 30 �TABLE 17. AVERAGE RELATIVE RESPONSE FACTORS (ERF) FOR PCB CONGENERS IN SOLUTION 1 MEASURED AS REPLICATES ON A SINGLE DAY AND AS SINGLE MEASUREMENTS FOR DAY-TO-DAY BASIS3 Congener no. Single day measurements Relative std. Std. deviation ( ) % RRF deviation £ Day-to-day measurements Std. Relative std. RRF deviation deviation ( ) % 1 0.118 2.905 3.544 0.452 12.767 11 3.073 0.073 2.363 2.733 0.300 10.977 29 2.195 0.048 2.188 2.005 0.171 8.535 47 1.062 0.059 5.591 1.032 0.061 5.876 121 0.948 0.020 2.127 0.955 0.036 3.747 136 0.689 0.016 2.336 0.685 0.046 6.688 181 0.383 0.009 2.379 0.377 0.028 7.347 195 0.263 0.003 1.184 0.270 0.022 8.304 207 0.237 0.008 3.547 0.257 0.030 11.757 209 a 4.073 0.213 0.006 2.837 0.223 0.023 10.352 See Tables 6 and 8 for CGC/EIMS operating conditions. b These values calculated from four replicates. c These values calculated from 11 separate analyses. 31 �2.50 Magnetic Sector Mass Spectrometer 2.00 1.50 to C N3 O p. CO 0) 1.00 - 0.50 - 10 Homolog (degree of chlorination) Figure 3. Plot of response factor per isomer versus homolog for 77 PCB congeners, determined on a single day. Each value is representative of single measurements of each congener with the Varian Fat 311A magnetic sector mass spectrometer. This plot indicates the number of data noints that overlap for specific isomers. �Table 18 presents the upper and lower 95% confidence limits for the measured average RRFs. The extrapolation was necessary for the dichloro- through octachlorobiphenyl homologs. The projected upper and lower limits of the average RRF ranged from 13% for each PCB homolog for trichlorobiphenyls to approximately 6.5% for the dichlorobiphenyls. The projected ranges for the tetrachloro- to octachlorobiphenyls were between these values. Comparison of Magnetic Sector and Quadrupole RRF Data-The two instruments used operate on entirely different principles, so the results may represent the range of RRFs to be expected from these compounds on different instruments. Table 19 presents a summary of the data. As expected, the RRF trends are much different. Since quadrupole spectrometers discriminate at the high masses, the RRFs for high homologs (higer masses) are much lower than corresponding values for the magnetic detector spectrometer. A statistical analysis of the data (Student's t-test presented in Table 4 of Appendix A) confirmed that the average RRFs are significantly different for many of the homologs. However, the relative standard deviations for the average RRF of each homolog are not significantly different. Thus, the extrapolation from a single calibration isomer to all isomers of a homolog should have similar precision for the two instrument types. Relative Retention Times Relative retention times (KRT) were also calculated from the data generated for relative response factor measurements with both the quadrupole and magnetic sector mass spectrometer instruments. All RRTs for each PCB congener were calculated versus 3,3',4,4'-tetrachlorobiphenyl-de- Figure 4 is a plot of the RRT data versus PCB homolog. All data points for the 77 PCB congeners measured with the quadrupole mass spectrometer are presented. This plot also indicates that the relative retention window for the dichloro- to octachlorobiphenyl homologs may be larger than that actually measured if more of the possible congeners were present. Table 20 presents the observed range of RRTs for the 77 PCB congeners and additional congeners, identified only by homolog, in an Aroclor mixture (1016, 1254, 1260). These RRTs were established using a 15-m fused silica DB-5 capillary column. It must be recognized that the RRT windows on other columns may be substantially different. Table 20 also presents a projected RRT window for PCB anaysis. The overlap of the retention windows of each homolog must be considered in establishing an instrumental analysis approach to quantitation of the specific PCB homologs. This consideration has been accounted for in the GC/MS requirements for PCB analysis in Appendices B to D. The relative retention times of the 77 PCB congeners as determined with both the quadrupole and magnetic sector mass spectrometers are presented in tabular form in Appendix A. 33 �TABLE 18. MEASURED AVERAGE RELATIVE RESPONSE FACTOR (RRF) AND CORRESPONDING UPPER AND LOWER 95% CONFIDENCE LIMITS PCB homolog Monochloro- No. of possible isomers (N) Average measured response RRF No. of available isomers (n) Sample std. deviation (S) T Lower a limit T, Upperb limit - 3 3 3.331 0.643 Dichloro- 12 10 2.027 0,447 1.896 2.158 Trichloro- 24 9 1.573 0.341 1.366 1.780 Tetrachloro- 42 16 0.950 0.175 0.877 1.023 Pentachloro- 46 12 0.720 0.120 0.654 0.786 Hexachloro- 42 13 0.513 0.078 0.474 0.552 Heptachloro- 24 4 0.361 0.024 0.326 0.396 Octachloro- 12 6 0.253 0.030 0.231 0.275 Nonachloro- 3 3 0.229 0.034 Decachloro- 1 1 0.213 a Lower 95% limit = RRF -S("'j1* +-c 1 n \ \. N 34 - - �TABLE 19. RELATIVE RESPONSE FACTORS MEASURED VERSUS 3,3',4,4'-TETRACHLORO BIPHENYL-d6 BY ELECTRON IMPACT MASS SPECTROMETRY QUADRUPOLE (FINNIGAN 4023) AND MAGNETIC SECTOR (VARIAN MAT 311A) INSTRUMENTS No. of isomers measured RRF Quadrupole_ Mean RSD" ( ) % Magnetic sector RSDa ( ) % Mean 3 2.739 9.3 2.329 8.5 Dichloro- 10 2.048 15.7 1.663 13.8 Trichloro- 9 1.592 18.1 1.167 21.3 Tetrachloro- 16 0.946 20.0 0.902 14.0 Pentachloro- 12 0.725 17.6 0.780 17.4 Hexachloro- 13 0.500 19.1 0.640 19.4 Heptachloro- 4 0.308 8.0 0.497 12.1 Octachloro- 6 0.224 17.3 0.463 15.3 Nonachloro- 3 0.188 16.2' 0.467 22.5 Decachloro- 1 0.179 - 0.586 PCB homolog Monochloro- a Relative standard deviation. 35 - �H2<-'10 Relative Retention Times of PCB Congeners by Homolog Versus 3. 3'. 4. 4' Tetrachlorobiphenyl-d, C]2H,C19 2M2 7 ° 204 202 2OO C12H2CI8 195 194 185 171 183 181 C12H3CI7 PCB homologs 198 ** 155 C12H4CI6 154 139 153 138 129 136 15t 143 141 137 128 156 97 . „ 121 II6 104 103I009388 101119 115 C12H5CI5 15 545053 Cl2H6CI4 30 C12H7CI3 10 4 C12H8CI2 1 C,2H9CI, 9 7 8 5 14 5247 44 72 -«061 66 6577 2631 33 29 28 21 18 24 lt'^15 2 3 1 1 0.40 0.50 1 0.60 1 1 1 1 0.70 0.80 0.90 1.00 1 1.10 1 1.20 1 1.30 1 1.40 Relative retention time Figure 4. Retention times of 77 PCB congeners relative to 3,3',4,4'-tetrachlorobiphenyl-d6 (RRT of 1.00) The dashed line indicates that not all of the possible isomers of a particular homolog were measured. Relative retention times were determined on a J&W DB-5, 15-m fused silica column in a Finnigan 4023 GC/EIMS system. Temperature program: 110°C for 2 min,. then 10°C/min to 325°C. �TABLE 20. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS VERSUS d6-3,3',4,4'-TETRACHLOROBIPHENYL PCB homo log No. of isomers measured Observed range of RRTsa Calibration solution Congener Observed no. RRT3 Projected range of RRTsD 3 0.40-0.50 1 3 0.43 0.50 0.35-0.55 Dichlorobiphenyl 10 0.52-0.69 7 0.58 0.35-0.80 Trichlorobiphenyl 9 0.62-0.79 30 0.65 0.35-1.10 Tetrachlorobiphenyl 16 0.72-1.01 50 0.75 0.55-1.05 Pentachlorobiphenyl 12 0.82-1.08 97 0.98 0.80-1.10 Hexachlorobiphenyl 13 0.93-1.20 143 1.05 0.90-1.25 Heptachlorobiphenyl 4 1.09-1.31 183 1.15 1.05-1.35 Octachlorobiphenyl 6 1.19-1.36 202 1.19 1.10-1.50 Nonachlorobiphenyl 3 1.31-1.42 207 1.33 1.25-1.50 Decachlorobiphenyl 1 1.44-1.45 209 1.44 1.35-1.50 Monochlorobiphenyl a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were measured versus de-3,3',4,4'-tetrachlorobiphenyl (internal standard) using a 15-m J&W DB-5 fused silica column with a temperature program of 110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an on-column injector. A Finnigan 4023 Incos quadrupole mass spectrometer operating with a scan range of 95-550 Daltons was used to detect each PCB congener. b The projected relative retention windows account for overlap of eluting homologs and take into consideration differences in operating systems and lack of all possible 209 PCB congeners. 37 �Selection of Congeners for a Calibration Standard The data generated from the RRF and RRT measurements were used to select the PCB congeners for an analytical quantitation/calibration standard for GC/EIMS analysis of PCBs. Selection of the standard compounds was based primarily on the ratio of the measured response factor to the average response factor for a particular homolog. The PCBs with KRFs closest to the average values were selected as standard compounds. In addition, the RRT was considered to assure that the selected PCB congeners did not coelute. Two monochlorobiphenyls were selected for the calibration standard because the average RRF and RRT did not clearly coincide with any of the three possible isomers. One isomer (2-chlorobiphenyl) had a substantially different RRF. This isomer was quantitated separately. 4-Chlorobiphenyl was selected as the calibration isomer for the two remaining isomers. Figure 5 is a CGC/EIMS chromatogram of the 11-component PCB calibration standard. The composition of this solution is identified in Tables 4 and 20 along with the observed RRT of each of the 11 congeners. VALIDATION OF SELECTED CLEANUP STEPS As part of the overall method validation, several of the cleanup techniques were validated. A mixture of the 11 calibration standard congeners and three recovery surrogates (the 13C-octachlorobiphenyl was unavailable for these experiments) was diluted in an appropriate solvent and then subjected to the cleanup procedures as described in Appendix B. After the cleanup, the internal standard was added and the volume adjusted. The samples were analyzed by CGC/EIMS using a quadrupole spectrometer operated under the condition listed in Tables 6 through 8. Data were collected in the full scan mode and quantitated using the primary ions listed in Table 10 and the congener pairs listed in Table 13. A blank was run through the procedure alongside the recovery spikes. As expected, no PCBs except the internal standard were observed in the blanks. The results for the 11 calibration congeners were calculated as percentage recovery. Tables 21 through 25 present the uncorrected recoveries, calculated using Equation 12-1 of Appendix B, using the internal standard (Congener No. 210); the actual percentage recoveries of the 13C-labeled recovery surrogates, calculated using Equation 12-2 of Appendix B; and the corrected recoveries of the calibration congeners, calculated using Equation 12-3 of Appendix B. Inspection of Tables 21 to 25 reveals that the accuracy of the corrected recoveries is higher than for the uncorrected recoveries (104% versus 77% average). On the other hand, the precision of the uncorrected recoveries is slightly higher than for the corrected recoveries (11% versus 9% relative standard deviation average). This is the expected trend since the uncorrected recovery relies on two GC/MS measurements (area of the PCB congener peak and area of the internal standard peak) and the corrected recovery relies on those two values and the area of the surrogate peak. Thus, these results indicate that accuracy is improved by recovery correction, at a sacrifice of precision. 38 �T3 4-1 ^-v rH Cfl cn oi cd C 00 0 Ol rJ •U G s v 100.0 U co -' ^£3 O rH rH C C 01 X! D. •H XI 01 XI tX 'H XI 1 i—1 C C Ol Ol XI XI a ex •rH XI Xi ^-s O O 0) M rJ 0 0 CO rH rH X! XI U 0 O 0 4J rH XI U 00 0 CO rJ rJ rJ cx -H ,Q 0 t-i o rH XI O 0 C £ T CM H ^ c rH ^*> c Ol XI ex •H XI o o Ol XI ex •H XI o ^1 o 01 01 H X! rH <- ^j- XI K**I rj Ol XI o o <J- o o X •H }-l x; o cfl a o C oi ex •H Xi o c Ol ex ^ •H XI O r4 0) •U CO 00 o o x; o cfl o Ol p rH )-l O 3 Cfl CD O "--' Ol P " CO CO CO vO LO cfl CM •V c • ^ CM ^ fc XI 33 1 4_) rH rH vD vO cfl X X! O r, s^- o t-l o vO r^ CO ex -H XI O 0) rO •\ CO 0 1 4-> O a r, rH 01 ex •H x> o X! O cfl 01 XI a cfl 4-1 m ex ^ ai m r4 O rH XI <t" ». }-t H C Ol XI •rH •\ r> •. ex ro XI 0 cfl VJ ex »i •\ •H XI CO rH rH o •H Q o H 1 vD 1 •s I. CM ^ <J- " ^ CM 1316 w~ rv i CM CO m CN „ -o- 0> CN 4- CM J I CO l I(V1 i ivi 1178 CM 798 936 , » 1 6K3 c 4J X! CN 618 iH ^» ex iH 01 ,£ rH XI U Cfl r4 4J 01 H 1 •H C75 ^ c x; ex •H Xi o C CD <r sr u C c 01 rH C C 3 o o w S S s-' me 1 o o o'f & "rH 112592. rH rH rH 1101 —' T3 CO LO VO 3 to •H C'JO 13:20 [ ^^. 1030 16:49 1 — 12^0 20 :&) SCA'l Figure 5. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS) chromatogram or the calibration standard solution required for quantitation of PCBs by homolog, This chromatogram includes PCBs representative of each homolog, three ^C-iabeled surrogates, and the deuterated internal standard. The concentration of all components and the CGC/EIl^S parameters are presented in Tables 4, 5, 6 and 9. �TABLE 21. RECOVERY DATA FOR ACID CLEANUP' Congener no. -P- o PCB homolog 1 3 7 30 50 97 143 183 202 207 209 X Standard deviation Relative standard deviation ( ) % Monochlorobiphenyl Monochlorobiphenyl Dichlorobiphenyl Trichlorobiphenyl Tetrachlorobiphenyl Pentachlorobiphenyl Hexa chlo rob ipheny 1 Heptachlorobiphenyl Octachlorobiphenyl Nona chlo rob ipheny 1 De ca chlo rob ipheny 1 211 212 214 X Standard deviation Relative standard deviation ( ) % 13 Ce-nionochlorobiphenyl Ci2~tetrachlorobiphenyl 13 Ci2~o!ecachlorobiphenyl 13 a 0.52 0.50 0.52 0.52 0.76 0.87 0.96 1.30 2.30 2.50 2.10 2.60 5.30 10.20 Spike No. 1 not analyzed. b Total spike level (pg) Corrected via surrogate response. c Not detected. d Large background signal prevented quantitation of the compound. e Not applicable. Spike 2 ( recovery) % Corrected Uncorrected 100.0 83.4 82. £ NQ3 78.0 99.5 81.2 85.9 80.7 83.2 87.3 86.2 7.5 9 70.2 87.1 91.3 82.9 11.2 13 142.4 118.8 117.5 NQ 89.6 114.2 93.2 98.5 88.4 91.1 95.7 104.9 17.7 17 _ — - Blank NDC ND ND ND ND ND ND ND ND ND ND ND ND ND - �TABLE 22. Congener no. RECOVERY DATA FOR FLORISIL COLUMN CLEANUP PCB homo log 1 3 7 30 50 97 143 183 202 207 209 X Standard deviation Relative standard deviation ( ) % Monochlorobiphenyl Monochlorobiphenyl Dichlorobiphenyl Trichlorobiphenyl Tetrachlorobiphenyl Pentachlorobiphenyl Hexa chlo r ob ipheny 1 Heptachlorobiphenyl Octachlorobiphenyl Nonachlorobiphenyl Decachlorobiphenyl 211 212 214 X Standard deviation Relative standard deviation ( ) % 13 Total spike level (pg) 0.52 0.50 0.52 0.52 0.76 0.87 0.96 1.30 2.30 2.50 2.10 Spike 1 ( recovery) % Corrected" Uncorrected 57.9 63.0 66.0 69.4 70.7 73.4 72.6 76.6 77.8 78.1 77.7 71.2 6.7 9 90.6 98.6 103.2 160.5 163.6 169.7 168.1 177.2 102.5 102.9 102.4 130.9 35.8 27 Spike 2 ( recovery) % Corrected Uncorrected 54.9 58.3 60.0 62.3 62.4 66.1 67.0 72.3 72.3 70.5 72.8 65.4 6.2 10 _ C6-monochlorobiphenyl 13 C12-tetrachlorob ipheny 1 13 C12~decachlorobiphenyl a Corrected via surrogate response, b Not detected, c Not applicable. 2.60 5.30 10.20 63.9 43.2 75.9 61.0 16.5 27 - 57.6 47.9 69.6 58.4 10.9 19 95.4 101.2 104.4 130.0 130.3 138.1 140.1 151.0 103.8 101.3 104.5 118.2 19.8 17 - Blank NDb ND ND ND ND ND ND ND ND ND ND c ND ND ND - �TABLE 23. RECOVERY DATA FOR FLORISIL SLURRY CLEANUP Congener no. PCB homolog 1 3 7 30 50 97 143 183 202 207 209 X Standard deviation Relative standard deviation ( ) % Mpnochlorobiphenyl Monochlorobiphenyl Dichlorobiphenyl Trichlorobiphenyl Tetrachlorobiphenyl Pentachlorobiphenyl Hexachlorobiphenyl Hep ta chlorob iphenyl Octachlorobiphenyl Nona chlo r ob iphenyl Decachlorobiphenyl 211 212 214 X Standard deviation Relative standard deviation ( ) % 13 Total spike level (pg) 0.52 0.50 0.52 0.52 0.76 0.87 0.96 1.30 2.30 2.50 2.10 Spike 1 ( recovery) % Corrected" Uncorrected 80.5 81.2 87.5 NQC 90.0 96.0 95.5 95.1 97.2 95.1 96.2 91.4 6.3 7 96.0 96.8 104.4 NQ 91.6 97.6 97.2 96.8 91.2 89.4 90.4 95.1 4.5 5 Spike 2 ( recovery) % Corrected Uncorrected 71.1 72.7 75.0 76.4 80.1 83.5 82.0 79.8 88.8 87.6 83.7 80.1 5.8 7 _ C6-monochlorob iphenyl 13 C12-tetrachlorobiphenyl 13 C12~decachlorobiphenyl 2.60 5.30 10.20 83.9 98.3 106.5 92.5 7.5 8 a Corrected via surrogate response. b Not detected. c Large background signal prevented quantitation of this compound. d Not applicable. - 92.9 94.8 98.1 85.5 89.6 93.5 91.6 89.3 101.0 99.5 95.2 93.7 4.7 5 Blank NDb ND ND ND ND ND ND ND ND ND NDd - _ 76.7 89.4 87.9 84.7 6.9 8 - ND ND ND - �TABLE 24. RECOVERY DATA FOR KOH CLEANUP Congener no. 1 3 7 30 50 97 143 183 202 207 209 X PCB homolog Monochlorobiphenyl Mono chlo rob iphenyl Dichlorobiphenyl Trichlorobiphenyl Tetrachlorob iphenyl Pentachlorob iphenyl Hexachlorob iphenyl Heptachlorob iphenyl Octachlorob iphenyl Nona chl o rob iphenyl Decachlorob iphenyl Total spike level ( j ) |g 0.52 0.50 0.52 0.52 0.76 0.87 0.96 1.30 2.30 2.50 2.10 Standard deviation Relative standard Spike 1 ( recovery) % Uncorrected 60.2 69.0 73.5 75.0 79.7 85.8 84.0 81.2 89.2 88.2 69.9 77.8 9.1 12 Corrected" 82.6 94.6 100.8 83.5 88.7 95.4 93.4 90.3 113.0 111.8 88.6 94.8 10.2 11 Spike 2 ( recovery) % Uncorrected Corrected 67.7 73.6 77.5 77.6 80.7 85.0 85.0 81.3 89.3 87.6 71.9 79.7 6.8 9 90.1 98.0 103.3 89.2 92.9 97.7 97.7 93.5 117.5 115.3 94.6 99.1 9.5 10 Blank NDb ND ND ND ND ND ND ND ND ND NDC - deviation ( ) % _ 211 212 214 X Standard deviation 13 C6-monochlorobiphenyl 13 C12-tetrachlorob iphenyl 13 C12-decachlorob iphenyl Relative standard deviation ( ) % a Corrected via surrogate response, b Not detected, c Not applicable. 2.60 5.30 10.20 72.9 89.9 78.9 80.6 8.6 11 - _ 75.1 87.0 76.0 79.4 6.6 8 - ND ND ND - �TABLE 25. RECOVERY DATA FOR ALUMINA CLEANUP Congener no. PCB homolog 1 3 7 30 50 97 143 183 202 207 209 X Standard deviation Relative standard deviation ( ) % Mono chlo rob iphenyl 211 212 214 X Standard deviation 13 Monochlorobiphenyl Dichlorob iphenyl Trichlorob iphenyl Tetrachlorob iphenyl Pentachlorobiphenyl Hexachlorob iphenyl Heptachlorobiphenyl Octachlorob iphenyl Nona chlorob iphenyl Decachlorob iphenyl Total spike level (|Jg) 0.52 0.50 0.52 0.52 0.76 0.87 0.96 1.30 2.30 2.50 2.10 Spike 1 ( recovery) % Corrected Uncorrected 63.1 60.0 67.9 NQC 67.2 70.4 69.4 75.8 76.8 77.3 74.0 70.2 5.9 8 97.1 92.2 104.8 NQ 97.2 101.9 100.4 109.7 92.2 92.9 88.9 97.8 6.5 7 Spike 2 ( recovery) % Uncorrected 61.1 58.4 66.4 NQ 66.3 68.3 67.5 75.1 75.3 76.8 78.3 70.1 6.9 10 13 Cj2~tetrachlorob iphenyl 13 C12~decachlorobiphenyl Relative standard 2.60 5.30 10.20 64.8 69.1 83.2 72.4 9.6 13 deviation ( ) % a Corrected via surrogate response. b Not detected. c Large background signal prevented quantitation of this compound. d Not applicable. - Blank 101.0 96.2 109.4 NDb ND ND ND ND ND ND ND ND ND NDd NQ 102.2 105.4 104.2 115.8 89.5 91.2 93.0 100.8 8.4 8 - _ _ C6-monochlorob iphenyl Corrected 60.7 64.9 84.2 69.9 12.5 18 - ND ND ND - �The preliminary data presented here contain an apparent anomaly: the low recovery of the 13C-tetrachlorobiphenyl surrogate (Congener No. 212) from the Florisil column cleanup. These two data points contribute substantially to the imprecision of the surrogate recoveries and induce some very high (130 to 177%) corrected recoveries for the tri- through hepta- compounds. The experiment should be repeated. VALIDATION OF THE PRODUCT AND PRODUCT WASTE METHOD WITH INDUSTRIAL SAMPLES Strategy Selected samples, obtained from industrial sources, were subjected to a variety of sample preparations as listed in Table 15 and then analyzed by CGC/EIMS. This section presents the results of this preliminary validation and, where possible, compares our values with those of previous analyses of the same sample. The results for quality control samples are also reported. The most extensively studied matrix was the CMA-A chlorinated benzene waste stream sample. This particular sample was chosen because of the wide distribution of PCB homologs (mono- through decachlorobiphenyls). Sample preparation with this matrix included simple dilution, treatment with sulfuric acid, Florisil, and saponification with ethanolic potassium hydroxide. The CMA-A samples were analyzed in duplicate in two sets of experiments. The 11 PCB congeners used for calibration purposes were spiked into the CMA-A matrix for standard addition experiments. Blind spiked samples and quantitation standards, prepared by the MRI quality control personnel as analytical performance checks, were analyzed along with the other samples. First Sample Set Tables 26 and 27 present the uncorrected and corrected concentrations found for CMA-A samples in preliminary studies of the application of the proposed methods for commercial products and product wastes. Sample 10 was analyzed without surrogates to approximate the analytical procedure used by most other laboratories. As anticipated, the uncorrected values compare well with 20A and 20B, while the corrected values are slightly lower than the values for 10. Both corrected and uncorrected values for the duplicate samples 20A and 20B are in agreement. The values for samples 10, 20A, and 20B average about 400 pg/g. These values are higher than the mean of 280 JJg/g reported in the CMA round robin but are in good agreement with the values (402 |Jg/g) reported by the sample supplier (Appendix E of Pittaway and Horner, 1982). The homolog distribution of our data agrees in general with the CMA data and the data that accompanied the samples. Sample 110 (CMA-E) was determined to contain about 18 (Jg/g PCB (Table 28) mostly as the dichloro homolog. These results are slightly higher than the CMA round robin data, which had a mean reported value of 9 pg/g. The isomer distribution agrees with most of the CMA round robin data (Pittaway and Horner, 1982). 45 �TABLE 26. Congener no. 1 3 7 30 50 97 143 183 202 207 209 UNCORRECTED PCB CONCENTRATIONS (Mg/g) IN CMA-A SAMPLES PCB homolog 10 Dilution, no surrog. 20A Dilution 20B Dilution 1 1 2 3 4 5 6 7 8 9 10 9 19 64 55 60 50 56 60 0 0.9 9.3 11 21 70 52 63 40 48 84 0 0 20 10 19 64 49 55 36 38 68 0 0 20 408 358 96b 108 154 94 97 152 414 Total 211 212 214 1 4 10 a No surrogates added. b NSa NS NS Surrogate recovery (percent). 46 �TABLE 27. CORRECTED PCS CONCENTRATIONS (iig/g) IN CMA-A SAMPLES Congener no. 1 3 7 30 50 97 143 183 202 207 209 PCB homo log 10 Dilution, no surrog. 20A Dilution 20B Dilution 1 1 2 3 4 5 6 7 8 9 10 NSa NS NS NS NS NS NS NS NS NS NS 11 22 73 49 58 37 44 78 0 0 13 11 21 68 50 57 37 39 70 0 0 13 385 366 Total a No surrogates added. 47 �Second Sample Set CMA Product Waste Samples— The corrected and uncorrected concentrations of the PCB homologs for duplicate CMA-A samples from a more extensive study are presented in Tables 29 and 30. Sample 2005 was spiked only with the internal standard so that any interferences corresponding to the 13C-labeled PCBs could be measured. Samples 2010 and 2020 are duplicate samples of CMA-A. The four surrogate compounds were added to approximately 0.1 g of each sample. The mixture was diluted to 1.0 ml and the internal standard added. Sample 2025Q is a sample that was submitted for PCB analysis by the MRI quality control department. This sample was weighed by QC personnel and the final preparation completed as described for the previous samples. The MRI QC coordinator calculated the final concentration for 2025Q from the extract concentration of each PCB homolog and weight of the CMA-A sample recorded in the QC laboratory record book. The surrogate-corrected values reported for samples 2010 through 2025Q are in good agreement with the total PCB concentration and homolog distribution reported in the CMA round robin (Pittaway and Horner, 1982). Tables 31 and 32 present the data from a standard addition experiment with the CMA-A sample matrix. The 11 PCB congener calibration standard was added to three separate aliquots of the CMA-A matrix to give spike levels ranging from approximately 20 to 100 | g of the monochlorobiphenyl and 50 to j 200 (Jg of decachlorobiphenyl. Samples 2030, 2040, and 2050 were prepared in the analytical laboratory. Sample 2060Q was prepared as a blind spike of the CMA-A matrix by MRI quality control personnel. The uncorrected amount found did not increase linearly with the spike level. In fact, at the highest spike level (Sample 2050) the amounts found for each homolog were less than the spike. No explanation is immediately available for this data trend, although the low recoveries of the 13C-octa- and tetrachlorobiphenyl surrogates indicated that the data are at best marginally valid. Tables 33 and 34 present data for CMA-A samples that were subjected to three different cleanup methods (concentrated H2S04, Florisil column chromatography, and saponification with alcoholic KOH). The data from the sulfuric acid cleanup was difficult to interpret because of interferences. As noted previously (Erickson and Stanley, 1982), the acid cleanup results in large losses of lower chlorinated PCB homologs. The poor recoveries of the surrogates shown in Table 33 are clearly outside of the QC criteria in Section 14.2.2 of Appendix B and indicate that the analyses are invalid. These results would not be reported as analyses for compliance with the proposed regulation. All of the blank samples (2001, 1080, 2100, and 2120) were analyzed along with the sample discussed above and found to contain no detectable PCBs. 48 �TABLE 28. UNCORRECTED AND CORRECTED PCS CONCENTRATIONS (pg/g) IN CMA-E SAMPLE (DILUTION PREPARATION) Congener no. 1 3 7 30 50 97 143 183 202 207 209 PCB homolog 110 Uncorrected 110 Corrected 1 1 2 3 4 5 6 7 8 9 10 1.2 1.8 10.5 0 0 0 0.02 0 0.05 0 0.06 1.5 2.4 13.8 0 0 0 0.02 0 0.03 0 0.04 13.4 17.7 b Total 211 212 214 a 76a 103/91C 151 1 4 10 Surrogate recovery (percent). b Not applicable. c Samples run twice on magnetic sector instrument for low and high masses. Congener no. 212 monitored in both runs. 49 �TABLE 29. UNCORRECTED PCB CONCENTRATION ((Jg/g) IN THE CMA-A SAMPLE MATRIX (INTERNAL STANDARD CALCULATION) CMA-A 2005 8/4/82 CMA-A 2010 8/4/82 CMA-A 2020 8/5/82 CMA-A 2025 8/5/82 Monochlorobiphenyl 26 23 37 40 Dichlorobiphenyl 35 28 41 48 Trichlorobiphenyl 17 14 46 50 Tetrachlorobiphenyl 20 31 33 36 Pentachlorobiphenyl 32 29 29 31 Hexachlorobiphenyl 29 23 21 22 Heptachlorobiphenyl 18 12 12 14 PCB homo log CGC/EIMS analysis date Octachlorobiphenyl 5.4 4.1 3.4 4.2 Nonachlorobiphenyl 2.6 2.2 2.0 3.5 Decachlorobiphenyl 12 10 Total PCB 197 176 9.7 11 234 260 Recovery ( ) of Surrogate Compounds % 13 NSa 64 84 89 13 NS 96 96 101 Ci 2 -octachlorobiphenyl NS 73 67 72 C12-decachlorobiphenyl NS 68 69 73 Ce-monochlorobiphenyl C12~tetrachlorobiphenyl 13 13 a NS = no surrogate added. b Final concentration determined from sample weight recorded by QC coordinator. c 302 Daltons used for quantitation. 50 �TABLE 30. CORRECTED PCB CONCENTRATION (|Jg/g) IN THE CMA-A SAMPLE MATRIX CMA-A 2010 8/4/82 CMA-A 2020 8/5/82 CMA-A, 2025Q 8/5/82 Monochlorobiphenyl 37 44 44 Dichlorobiphenyl 44 48 53 Trichlorobiphenyl 15 47 49 Tetrachlorobiphenyl 33 34 34 Pentachlorobiphenyl 30 30 31 Hexa chlo robipheny 1 24 21 22 Heptachlorobiphenyl 16 18 19 PCB homo log CGC/EIMS analysis date Octachlorobiphenyl 5.4 4.9 5.7 Nona chlo rob ipheny 1 3.1 3.0 4.8 Decachlorobiphenyl 15 14 16 Total PCB 223 264 280 a NS = no surrogates added. b Final concentration determined from sample weight recorded by QC coordinator. 51 �TABLE 31. PCB homolog CGC/EIMS analysis date UNCORRECTED PCB CONCENTRATION (|Jg/g) Of SPIKED CHA-A SAMPLES DETERMINED RY THE INTERNAL STANDARD QUANTITATION METHOD CMA-A 2030 Total sample Spike concentration level 8/5/82 CMA-A 2040 Total sample Spike level concentration 8/5/82 CMA-A 2050 Total sample Spike concentration level 8/6/82 CMA-A 2060Q Total sample Spike concentration level 8/6/82 Blind quantilat ion standard Total sample Spike concentration level 8/6/82 Monochlorobiphenyl 60 20 80 49 92 100 100 82 140 184 Dichlorobiphenyl 56 10 58 25 58 51 69 42 53 94 Trichlorobiphenyl 65 10 75 25 39 51 44 42 87 94 Tetrachlorobiphenyl 47 15 55 36 43 75 50 61 110 137 Pentachlorobiphenyl 48 17 58 42 64 86 73 70 140 157 Hexachlorobiphenyl 40 19 48 46 61 95 67 77 160 173 Heptachlorobiphenyl 40 25 58 62 87 130 87 100 340 234 Octachlorobiphenyl 46 45 82 110 100 230 110 180 560 414 Nonachlorobiphenyl 51 49 93 120 130 250 140 200 530 450 Decachlorobiphenyl 60 42 110 100 140 210 140 170 430 369 513 252 717 615 814 1,280 920 2,550 2,306 N5 Total PCB 1,020 Recovery ( ) of surrogate compounds % I3 C6-monochlorobiphenyl 89 79 76 93 88 I3 Ci2-tetrachlorobiphenylb 94 93 84 93 88 Cj2-octachlorobiphenyl 62 56 41 53 78 65 57 48 64 79 13 13 C12-decachlorobiphenyl a Concentration in ng/ml rather than |jg/g since this sample was prepared by dilution of stock solutions of standards by QC personnel, b 302 Daltons used for quantitation. �TABLE 32. CORRECTED PCB CONCENTRATION (pg/g) OF SPIKED CHA-A SAMPLES DETERMINED BY SURROGATE RECOVERY CORRECTION PCB homolog CGC/EIMS analysis date CMA-A 2030 Spike Total sample concentration cone. 8/5/82 CMA-A 2040 Total sample Spike concentration cone. 8/5/82 CHA-A 2050 Spike Total sample concentration cone. 8/6/82 CMA-A 2060 Spike Total sample cone. concentration 8/6/82 Blind quantitation standard 2070Q Spike Total sample concentration cone. 8/6/82 Monochlorobiphenyl 67 20 100 49 120 100 110 82 160 184 Dichlorobiphenyl 63 10 74 25 76 51 74 42 60 94 TrichJorobiphenyl 70 10 80 25 46 51 47 42 99 94 Tetrachlorobiphenyl 50 15 58 36 52 75 53 61 130 137 Pentachlorobipheriyl 51 17 63 42 77 86 78 70 160 157 Hexachlorobiphenyl 43 19 52 46 72 95 72 77 190 173 Heptachlorobiphenyl 64 25 100 62 210 130 160 100 430 234 Octachlorobiphenyl 74 45 150 110 250 230 210 180 720 414 Nonachlorobiphenyl 81 49 170 120 330 250 270 200 680 450 Decachlorobiphenyl 91 42 180 100 280 210 220 170 540 369 Total PCB 650 250 1,030 60 2 1,280 1,290 1,020 3,190 2,310 a 1,510 Concentration in ng/ml rather than pg/g since this sample was a blind quantitation sample. �TABLE 33. PCB CONCENTRATION ((Jg/g) OF CMA-A SAMPLES TREATED WITH DIFFERENT CLEANUP PROCEDURES (INTERNAL STANDARD QUANTITATION) PCB homolog CGC/EIMS analysis date CMA-A 2090 acid cleanup 8/9/82 Monochlorobiphenyl ND3 CMA-A 2110 Florisil cleanup 8/9/82 CMA-A 2130 alcoholic KOH cleanup 8/9/82 4.4 31 Dichlorobiphenyl 4.4 14 44 Trichlorobiphenyl 0.4 31 44 Tetrachlorobiphenyl 25 18 25 Pentachlorobiphenyl 19 17 20 Hexachlorobiphenyl 7.9 5.6 6.3 Heptachlorobiphenyl 5.9 2.2 3.8 Octachlorobiphenyl 2.4 6.0 2.6 Nonachlorobiphenyl 38 2.4 2.6 Decachlorobiphenyl 16 9.5 6.4 Total PCB 119 110 186 % Recovery ( ) of surrogate compounds 13 74 8 145 13 0 0 367 13 115 97 110 13 173 129 64 C6-monochlorobiphenyl Ci2~tetrachlorobiphenyl Ci2'°ctachlorobiphenyl C12~decachlorobiphenyl a ND = not detected. b 302 Daltons used for quantitation. 54 �TABLE 34. PCB CONCENTRATION (|Jg/g) OF CMA-A SAMPLES TREATED WITH VARIOUS CLEANUP PROCEDURES (SURROGATE COMPOUND CORRECTED) CMA-A 2090 acid cleanup 8/9/82 CMA-A 2110 Florisil cleanup 8/9/82 CMA-A 2130 alcoholic KOH cleanup 8/9/82 Monochlorobiphenyl NDa 28 11 Dichlorobiphenyl 30 86 15 PCB homolog CGC/EIMS analysis date Trichlorobiphenyl 0.3 (0.2)b 200 (16) 15 (20) 110 (9.3) 8.4 (11) (8.3) 110 (8.9) 6.8 (9.0) (3.5) 3.5 (2.2) 2.9 (2.9) Tetrachlorobiphenyl 17 (11) Pentachlorobiphenyl 13 Hexachlorobiphenyl 5.3 Heptachlorobiphenyl 2.6 1.2 1.8 Octachlorobiphenyl 1.1 3.1 1.2 1.2 1.2 3.1 4.2 Nona chl o r ob ipheny 1 Decachlorobiphenyl Total PCB 17 3.9 546 (159) 90 (78) 68 (77) a ND = not detected. b 13 C12-tetrachlorobiphenyl was not quantifiable due to interferences. The values reported were calculated using 1*Ce-inonochlorobiphenyl. Values in parenthesis were calculated using 13Ci2~°ctachlorobiphenyl. 55 �DCMA Pigment Samples-Eight DCMA pigment samples were analyzed following the preparation described in the experimental section (Table 15). The results are presented in Table 35. The diarylide yellow pigment (DCMA-1) was analyzed in duplicate and as a blind spike supplied by the MRI quality control department. This sample is reported to contain 3,3'-dichlorobiphenyl at levels of approximately 70 M8/g (Dry Colors Manufacturing Association, 1981). No analyte or surrogate PCBs were detected in the duplicate 1-g samples of the pigment and a known spike of the sample. The lack of detected PCBs indicates a loss of analytes in the sample preparation. The CGC/EIMS analysis of a sample of the yellow pigment spiked by MRI quality control personnel yielded an uncorrected concentration of 76 |Jg/g of 3,3'-dichlorobiphenyl based on the internal standard quantitation and a corrected concentration of 63 Mg/g, based on 120% recovery of the 13C6-4-monochlorobiphenyl surrogate. The level of the 3,3'-dichlorobiphenyl added by the QC personnel was reported to be 60 M8/8- Hence, the total dichlorobiphenyl concentration should have been approximatey 130 |Jg/g (70 |Jg/g endogenous plus 60 (Jg/g added). The phthalocyanine green pigment (DCMA-4) was also analyzed in duplicate following dissolution and fractionation with a Florisil column. This pigment reportedly contains only decachlorobiphenyl at approximately 40 |Jg/g based on the results of the DCMA round robin study (Dry Color Manufacturing Association, 1981). Our analysis of duplicate samples yielded uncorrected concentration levels of 24 and 27 [Jg/g of decachlorobiphenyl by the internal quantitation method. The corrected concentration for both samples was 13 (Jg/g with recovery of the 13Ce-decachlorobiphenyl surrogate at 190 and 210%. Phthalocyanine blue (DCMA-8) was also analyzed as a single sample. Pentachloro- and hexachlorobiphenyls were detected but the concentrations were below the quantitation limits for that particular day. The total PCB concentration of this pigment, as discussed in the results of the DCMA round robin (1981), is reported to be 90 |Jg/g. The DCMA pigment sample analyses did not produce valid results. These data suggest that further development or validation of extraction/cleanup procedure would be necessary to provide acceptable PCB analyses of these samples. All of the blank samples (2001, 2080, and 2100) analyzed along with the DCMA samples were found to contain no detectable PCBs. DISCUSSION The determination of PCBs is a complex problem. The inaccessability of standards for all 209 congeners has traditionally been circumvented by the use of commercial mixtures (e.g., Aroclors) as standards. Quantitation has often been addressed in terms of relating the analyte to an Aroclor standard to give a "total PCB" concentration. Determination of PCBs synthesized as by-products in commercial products or product waste presents three special problems: (a) the analyte does not generally resemble a commercial PCB mixture, so quantitation against Aroclor standards would be incorrect; (b) the matrix often contains high concentrations of other chlorinated organics which are not easily separated during a cleanup procedure and which interfere with the qualitative and quantitative analysis; and (c) the matrix is undefined and can include gases, liquids, or solids of any purity and complexity. 56 �TABLE 35. RECOVERY ( ) OF CARBON-13 LABELED SURROGATE COMPOUNDS FROM DIARYLIDE YELLOW % AND PHTHALOCYANINE BLUE AND GREEN PIGMENTS PCB surrogate DCMA-1 21403 DCMA-1 21503 DCMArl 2160b DCMA-1 2170QC 2175d 2180d DCMA-8 21906 DCMA-8 2200Q 13 ND8 ND ND 120 ND ND ND 12 l3 ND ND ND ND ND ND 94 52 13 ND ND ND 200 120 107 92 71 13 ND ND ND 250 190 210 150 77 C6-Monochlorobiphenyl C12-Tetrachlorobiphenyl C12-Octachlorobiphenyl C12-Decachlorobiphenyl a Samples 2140 and 2150 are duplicates prepared by the DCMA-B method. b Sample 2160 was spiked with 50 |jg/g of 3,3'-dichlorobiphenyl and prepared by the DCMA-B method. c Sample 2170Q was spiked by MRI quality control personnel with 3,3'-dichlorobiphenyl and was prepared by the DCMA-B method. d Samples 2175 and 2180 are duplicates prepared by the DCMA-B method, e Sample 2180 was prepared by the DCMA-A method. f Sample 2200Q was weighed by MRI quality control personnel. preparation by the DCMA-A method. g The four surrogate compounds were added but not detected. An unknown mass of sample was supplied for �In this situation, analytical methods require a different philosophy than the classic approach for a single analyte in a defined matrix where all steps, reagents, and apparatus are specified. The method proposed here leaves many of the analytical steps to the discretion of the analyst while ensuring the reliability of the results with a strong quality control program. Thus, an analyst familiar with general analytical techniques for a product, may readily adapt in-house extraction/cleanup procedures to incidental PCB analysis. Even when the recoveries are not optimized, the 13C-labeled surrogate recoveries will mimic those of the analyte PCBs. As long as the 13C recovery surrogates are thoroughly incorporated, their recoveries can be used to derive corrected analyte PCB concentrations. Several of the method validation analyses presented above, especially Tables 33 and 35, illustrate the importance of the recovery surrogates in QC. The techniques employed are common methods validated for PCB analysis by other laboratories without the 13C-surrogate data. Analyses of this type may have been used by a testing laboratory and erroneous results reported. The complexity of the matrix and the high probability of chlorinated organic interferents precluded the use of GC/ECD. The best available technique is GC/EIMS. During the validation work presented above, the anticipated difficulty of qualitatitve and quantitative data interpretation was confirmed. In addition to the inherent problems resulting from extrapolation from a standard to several analytes, interpretation of the complex peak clusters is a tedious, subjective, and error-prone process. The volume of data for one sample is staggering; for sample 2110, 286 peaks were identified and integrated in the PCB mass chromatograms as shown in Figures 6 through 16. Of these, 58 peaks met the qualitative criteria and were identified as PCBs. Clearly different analysts will obtain different results for those peaks which marginally fit the qualitative criteria. This very high data density relative to other common GC/MS analyses has a much higher potential for error and mistakes. In addition it should be noted that, for many of the samples analyzed in this study, the data interpretation is more time-consuming than the rest of the analytical process. The integration methods are also prone to error. Integration is always conducted interactively with the mass spectrometric data system, either manually or automatically. The selection of baseline criteria, background sensitivity, integration method (valley-to-valley, baseline-to-baseline, etc.), and retention window all affect automatic quantitation. The position of the cursor and integration method affect the manual quantitation results. The day-to-day instrumental variability with quadrupole systems also appears to adversely affect data quality, despite tight calibration specifications. The magnitude of this error soruce should be further documented. The above discussion presents some of our understanding of some of the major problems with analysis for by-product PCBs. Further work will be devoted to characterizing and reducing these problem areas. Even with forseeable improvements in the method, the data for by-product PCBs in many commercial product and product waste samples will exhibit low precision and accuracy. 58 �lie 08/09/82 16:20:60 SAMPLE: SAMPLE 12110 dlA-A FLOMSIL 1/IODIL ItAHCE; G 1.1759 LABEL: H 0. 4.9 OMAN: A 1ULIMJ A. 1.9 DATA: 49011109^5 91 CALI: IHDCAIJWWl 1)1 BASE: U 2fl. SCAIJS 1 TO 1750 3 128450^1. 169.0 Ln vo 1009 15:59 Figure 6. Reconstructed ion chromatogram for 1299 19:00 1409 22:10 1660 25:29 SCAil mm SIM analysis of the CMA-A sample No. 2110. �MASS CHBOIIATOCRAIIS 88/09/82 16:20:00 SAIIPLE: SAIIPLE I2MO CIIA-A FLOftlStL RAflCE: C I.I750 LAHEI.: H 0. 4.0 DATA: 198III09U5 HI CAM: MIDTAUWW H1 l/IODIL 1ULHU OVAII: A 0. 1.0 BASE: 020. 83 SCAIJS 703 TO 900 3 I80.0-, 700 11:95 850 M:27 Figure 7. SIM ion plots for monochlorobiphenyls (188 and 190 Daltons) and the1 monochlorobiphenyl surrogate (194 Daltons) in CMA-A sample No. 2110. 900 SCAJf 14:15 TlliH �MASS aiCOHATOGRAIIS 08/09/82 16:28:60 SANFLE: SAIIPLE 12118 QIA-A FLOBISIL RANGE: C 1.1758 LABEL: H «. 4.9 DATA: 1991II99V5 81 CAII: IIIDCAUWWI HI 1/IODIL 1UI.IHJ QUAII: A 9. 1.9 BASE: U 28. SCAIIS 768 TO 1199 3 222. 1198 SC'J! 17:25 Tim- Figure 8. SIM ion plots for dichlorobiphenyls No. 2110. (222 and 224 Daltons) in Cl'A-A san.nle �HASS CHROMA TOGRAIIS DATA: 4991II09V5 fll 08/69/82 16:28:98 CALI: IIIDCALIttOVI (14 SAMPLE: SAIIPLE 12lie QIA-A FLORISIL I/IODIL IULHU RANGE; G 1.1759 LABEL: H 9. 4.9 QUAII: A 9. 1.9 BASE: U 29. 3 199.8 SCAMS 859 TO 1150 torn 256 1156129. 256.977 ^ 9.5W _8Z5_ Ni 1093 89. 258 1384579. 258.877 ± 8.599 859 13:27 Figure 9. 990 14:15 SIM ion plots for 950 15:92 15:59 1959 1199 16:37 17:25 1150 SCAM 18:12 TKIE trichlorobiphenyls (256 and 258 Daltons) in CVA-A sample No. 2110. �71.3-1 NASS OffiOmTOGRAIIS DATA: 499IIM9V5 II •8/09/82 (6:20:99 HALI: IIIDCAIJWWI SI SAHTLE: SAIIPLE 12110 QIA-A FLOWS! L I/I00IL IULUU BAMCE: G 1.1759 LABEL: II 9. 4.9 OUAII: A 0. 1.9 BASE: U 29. 3 1222 SCAHS 1959 TO 13TO 217296. 299 . 109.9-, 292 . CTi OJ 39.7-1 137728. 298. 298.089 0.590 62.7-1 217344. 304. 304.091 0.509 1350 SCAN 21:22 TIME Figure 10. SIM ion plots for tetrachlorobiphenyls (290 and 292 Daltons), 3 , 3 ' , 4 , 4 ' - t e t r a c h l o r o biphenyl-dg (298 Daltons), and the 13 C 1 2-tetrachlorobiphenyl surrogate (304 Daltons) in O'A-A sample No. 2110. �MASS CUBOIIATOGRAMS DATA: 4901IW9V5 II1669 08/09/82 16:28:89 CALI: IIIDCAI.II09VI 11 SAMPLE: SAMPLE I2II0 CMA-A FLORISIL I/IOD1L 1ULIIU RANGE: G 1.1750 LABEL: II 9, 4.0 WAII: A 9. I.0 BASE: U 29. 3 SCAIB J2W TO f? 15897B. 12.57 326 326.097 * 0.500 i 81.3- 129280. 328 J 328.098 * 0.500 1200 19:06 1250 19:47 1390 20:35 Figure 11. SIM ion plots for pentachlorobiphenyls No. 2110. 1400 22:10 1450 22:57 SCAN 23:15 TIIIE (326 and 328 Daltons) in CMA-A sample �MASS dffiOHATOCRAItS 08/09/82 16:20:08 SAMPLE: SAMPLE 12110 QIA-A FLORISIL I/IOniL RANGE: G IHIAH: A 0. 1.1759 LADEL: II 0. 4.0 1ULIHJ 1.0. SCAMS 1250 TO I500 DATA: 198IIWOV5 81 CALI: IHDCALHOT"! M ;SE: 0 28. 3 15C784. 97.-In 308.108 * 9.599 CT> Ln 106. 154880. 362 302.108 0.S00 1450 22:57 Figure 12. SCAN 23:15 TIHE SIM ion plots of hexachlorobiphenyls ( 6 and 362 Daltons) in CFA-A sample No. 2110. 30 �MASS UIROIIATOGRAIIS DATA: 4991H99V5 II 08/09/82 16:29:89 CAM: HI DCAll »W I 111 SAMPLE: SAIIPLE 12119 C1IA A FLORISIL 1/I0DIL 1ULIHJ RANGE: G 1.1756 LABEL: II 9. 4.9 OWAH: A 9. 1.9 HASP: U 29. 3 99.9 SCAMS 1359 TO 45568. ^91.118 ± 9.599 56384. 396.118 ± 9.599 1359 21:22 Figure 13. 1459 22:57 1559 SCAN 21:32 TIME SIM ion plots of heptachlorobiphenyls (39* and 306 Daltons) in CMA-A sample No. 2110. �MASS dlBffllATOGRAIIS 08/W82 16:20:80 SAMPLE: SAMPLE 12110 QIA-A FLORISIL RAMGE: G 1.1750 LABEL: II 0. 4.0 79.3-1 DATA: 4901H09V5 »l CALIt HIDCAUI0WI HI I/16DIL 1ULIHJ CUAII: A 0. 1.0 BASE: 0 20. ISC SCAMS IfM TO 165!) 3 29728. 428 . 128 •11856. 430.129 0.500 42304. I00.9-! 442 * 1450 22:57 Figure 14. 1509 23:45 1550 24:32 1608 25:20 SIM ion plots of octachlorobiphenyls ( 4 2 8 and 430 Daltons) and the chlorobiphenyl surrogate ( 4 4 2 Daltons) in CMA-A sample No. 2llo. 112.132 0.5W IC50 SCAN :!fi:07 TIME ^ �MASS aiBOHATOGRAIIS DATA: 4961II99V5 II 68/09/82 16:28:60 CALI: IIIDCALIWWI lit SAMPLE: SAMPLE I2H6 CIIA-A FLOIIISIL I/10D1L 1ULIMJ RAMCE: G 1.17561 LABEL: II 6. 4.9 OUAII: A 6. 1.6 BASE: U 29. 3 STAIIS 1559 TO 89.5-1 24256. 1565 1631 164.139 ± 0.599 461 . 109.6- < 27164. oo 466 466.139 * 6.590 1569 24:42 Figure 15. 1580 25:01 1620 25: 39 1616 25:58 SCAN TIME SIM ion plots of nonachlorobiphenyl (464 and 466 Daltons) in CMA-A sample No. 2110, �MASS aiBWIATOGRAIIS 98/89/82 16:28:88 SAHTLEi SAIITLE 12118 QIA-A FLOW SI L BAHGE: G 1.1758 LABEL: N 9. 4.8 99.9-1 1669 DATA: 49eil»9V5 11669 CALI: IHDCALI»WI 01 I/1001L 1ULIHJ QUAH: A 9. 1.8 BASE: U 29. SCAIB 1658 TO 17B!) 3 79232. 4W. 149 * 8.599 498 . 88I92. 499.500 * e.598 43849. 5S9.588 8.588 1709 Figure 16. SIM ion plots of decachlorobiphenyl (498 and 500 Daltons) and the decachlorobiphenyl (510 Daltons) in QIA-A sample No. 2110. 13C i2~ SCAN TIME �SECTION 5 REFERENCES Dry Color Manufacturers Association. 1981. An analytical procedure for the determination of polychlorinated biphenyls in dry phthalocyanine blue, phthalocyanine green, and diarylide yellow pigments. 1117 North 19th Street, Arlington, VA 22209. Erickson MD, Stanley JS. 1982. Midwest Research Institute. Methods of analysis for incidentally generated PCBs literature review and preliminary recommendations. Draft interim report no. 1. Washington, DC: Office of Toxic Substances, U.S. Environmental Protection Agency. Contract 68-01-5915. Haile CL, Baladi E. 1977. Midwest Research Institute. Methods for determining the total polychlorinated biphenyl emissions from incineration and capacitor and transformer filling plants. Washington, DC: U.S. Environmental Protection Agency. Contract 68-02-1780. EPA 600/4-77-048. Pittaway AR, Horner TW. 1982. Heiden, Pittaway Associates. Statistical analysis of data from a round robin experiment on PCB samples. Washington, DC: Chemical Manufacturers Association report. Roth RW, Keys JR, Chien DHT, et al. 1982. Midwest Research Institute. Methods of analysis for incidentally generated PCBs--synthesis of 13C-PCB surrogates. Draft interim report no. 3. Office of Toxic Substances, U.S. Environmental Protection Agency. Contract 68-01-5915. Stanley JS, Erickson MD. 1982. Midwest Research Institute. Peer review and authors' replies to 'methods of analysis for incidentally generated PCBs-literature review and preliminary recommendations.1 Draft interim report no. 2. Washington, DC: Office of Toxic Substances, U.S. Environmental Protection Agency. Contract 68-01-5915. USEPA. 1979a (December 3). U.S. Environmental Protection Agency. neutrals, acids, and pesticides—method 605. 44 FR 69540. Base/ USEPA. 1979b (December 3). U.S. Environmental Protection Agency. Organochlorine pesticides and PCBs—method 608. 44 FR 69501. USEPA. 1982 (June 8). U.S. Environmental Protection Agency. Polychlorinated biphenyls (PCBs); manufacture, processing, distribution, and use in closed and controlled waste manufacturing processes. FR 74 24976. 70 �APPENDIX A SUPPLEMENTARY GC/EIMS DATA ON PCB CONGENERS A-l �The following data support the method validation section for gas chromatography/electron impact mass spectrometry (GC/EIMS) of polychlorinated biphenyls (PCB). Table A-l lists the average relative response factors (RRF) for the 77 commercially available PCB congeners determined as four replicates. Table A-2 presents results of the Student's t-test used to determine the significance of differences for average RRFs for PCB homologs measured on a single day versus multiple days. The data in Table A-2 indicate that only the average RRFs for the heptachlorobiphenyl homolog are significantly different. Table A-3 presents the results of the Student's t-test used to determine the significance of differences for the average RRFs for the PCB homologs determined with the quadrupole and magnetic sector mass spectrometers. All 77 PCB congeners were determined in a single day for each of the instrument studies. This comparison indicates that the average RRF values are significantly different, which was expected. However, the relative standard deviations are not significantly different, indicating that the selection of the calibration standards is appropriate. These conclusions are discussed more fully in the text. Table A-4 presents results of the Student's t-test used to determine significance of differences for the RRFs for the 11 congeners in Solution No. 1, which was analyzed daily. An example of the data generated for multiple analysis of Solution No. 1 is presented in Figures 1 to 23. This information includes a capillary GC/EIMS chromatogram of Solution No. 1, the mass spectra of each component in this solution, and a graphic illustration of the distribution of several measurements of each congener about the average response factor. It should be noted that the standard deviation and relative standard deviation presented in these plots are different from that reported in the text due to calculation of the standard deviation using N weighting rather than the correct N-l weighting. All other standard deviations reported in this document are based on the N-l weighting. The relative retention times of the 77 PCB congeners with respect to 3,3",4,4'-tetrachlorobiphenyl-de determined with the Finnigan 4023 quadrupole and the Varian MAT 311A mass spectrometers are presented in Table A-5. A relative retention time unit of 0.01 (10 sec) is required for resolution of two specific congeners based on the gas chromatography parameters used to generate these numbers. A-2 �TABLE A-l. RELATIVE RESPONSE FACTORS FOR COMMERCIALLY AVAILABLE PCB CONGENERS (QUADRUPOLE) Congener no. Degree of chlorination Average relative response factor Standard deviation Coefficient of variation (%) 1 2 3 1 1 1 4.073 2.951 2.969 0.118 0.056 0.028 2.905 1.894 0.956 4 5 7 8 9 10 11 12 14 15 2 2 2 2 2 2 2 2 2 2 1.232 1.959 2.008 2.049 2.148 1.880 3.073 1.929 2.083 1.909 0.008 0.035 0.027 0.023 0.061 0.031 0.073 0.036 0.098 0.089 0.646 1.803 1.366 1.134 2.846 1.658 2.363 1.877 4.702 4.686 18 21 24 26 28 29 30 31 33 3 3 3 3 3 3 3 3 3 1.104 1.586 1.051 1.714 1.587 2.195 1.526 1.706 1.688 0.012 0.018 0.033 0.013 0.028 0.048 0.067 0.024 0.031 1.089 1.110 3.105 0.731 1.733 2.188 4.418 1.409 1.863 40 44 47 49 50 52 53 54 61 65 66 69 70 72 75 77 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 0.597 0.712 1.062 0.831 0.957 0.732 0.750 0.958 0.975 1.086 1.139 1.058 1.091 0.980 1.185 1.095 0.013 0.007 0.059 0.019 0.025 0.011 0.008 0.013 0.069 0.022 0.068 0.012 0.050 0.048 0.061 0.050 2.152 0.946 5.591 2.245 2.574 1.504 1.006 1.344 7.094 1.994 5.966 1.110 4.548 4.870 5.113 4.595 1 (continued) A-3 �TABLE A-l (continued) Degree of chlorination Average relative response factor Standard deviation Coefficient of variation ( ) % 87 88 93 97 100 101 103 104 115 116 119 121 5 5 5 5 5 5 5 5 5 5 5 5 0.617 0.611 0.574 0.719 0.727 0.653 0.566 0.824 0.853 0.785 0.762 0.948 0.011 0.005 0.010 0.008 0.003 0.004 0.009 0.025 0.061 0.013 0.022 0.020 1.710 0.744 1.677 1.139 0.428 0.538 1.627 3.048 7.146 1.654 2.911 2.127 128 129 136 137 138 139 141 143 151 153 154 154 155 156 6 6 6 6 6 6 6 6 6 6 6 6 6 6 0.499 0.431 0.689 0.533 0.433 0.462 0.419 0.490 0.473 0.549 0.221 0.511 0.587 0.599 0.005 0.004 0.016 0.008 0.008 0.026 0.010 0.005 0.013 0.050 0.001 0.010 0.011 0.044 1.093 0.813 2.336 1.582 1.946 5.686 2.353 0.986 2.826 9.101 0.570 2.039 1.828 7.431 171 181 183 185 7 7 7 7 0.346 0.383 0.380 0.336 0.002 0.009 0.010 0.006 0.640 2.379 2.501 1.729 195 198 200 202 204 8 8 8 8 8 0.263 0.262 0.301 0.250 0.221 0.003 0.008 0.007 0.007 0.007 1.184 2.887 2.392 2.663 3.200 206 207 208 9 9 9 0.193 0.237 0.259 0.003 0.008 0.003 1.723 3.547 1.315 209 10 0.213 0.006 2.837 Congener no. Relative to 3,3',4,4'-tetrachlorobiphenyl-dg. All relative response factors were calculated as the average of four replicate measurements made on the same day. A-4 �TABLE A-2. STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES BETWEEN QUADRUPOLE RESPONSE FACTORS CALCULATED ON THE SAME DAY VERSUS MULTIPLE DAYS PCB homolog MonochloroDichloroTrichloroTetrachloroPentachloroHexachloroHeptachloroOctachloroNonachloroDecachloro- Number of isomers 3 10 9 16 12 13 4 6 3 1 Average RRF from replicate measurements 3.331 2.027 1.573 0.950 0.720 0.513 0.361 0.253 0.229 0.213 Standard deviation 0.643 0.447 0.341 0.175 0.120 0.078 0.024 0.030 0.034 0.006 Average RRF from single , measurement Standard deviation 2.739 2.048 1.592 0.946 0.725 0.500 0.308 0.224 0.188 0.179 0.254 0.322 0.289 0.189 0.127 0.096 0.025 0.039 0.030 c t-Statistic 1.478 -0.119 -0.131 0.0618 -0.1085 0.377 3.119 1.398 1.5,91 Significant at 95% level? No No No No No No Yes No 5 > a Four replicate measurements of the RRF were made for each isomer. For example, the three monochlorobiphenyl isomers were measured four times each. Hence, the average RRF and standard deviation were calculated from 12 distinct values. b A single measurement for each of the 77 PCB congeners was completed in a single day. Hence, the average RRF reported is the average of one measured RRF for each isomer within a homolog. For example, the average RRF and standard deviation reported for the monochlorobiphenyl was calculated from three distinct values. c Single measurement. d Cannot test significance of difference between single measurements. �TABLE A-3. COMPARISON OF THE AVERAGE RELATIVE RESPONSE FACTORS (RRF) DETERMINED WITH QUADRUPOLE (FINNIGAN 4023) AND MAGNETIC SECTOR (VARIAN MAT 311A) MASS SPECTROMETERS'" PCB homolog Number of isomers MonochloroDichloroTrichloroTetrachloroPentachloroHexachloroHeptachloroOctachloroNonachloroDecachloro- 3 10 9 16 12 13 4 6 3 1 Finnigan 4023 quadrupole MS Standard deviation RRF 2.739 2.038 1.592 0.946 0.725 0.500 0.308 0.224 0.188 0.179 0.250 0.32 0.29 0.19 0.13 0.10 0.025 0.04 0.93 Varian MAT 311A magnetic sector MS Standard RRF deviation 2.329 1.663 1.167 0.902 0.780 0.640 0.497 0.463 0.467 0.586 RRFs significantly different at the , 95% confidence level Variances significantly different at the 95% confidence level No Yes Yes No No Yes Yes Yes Yes e No No No No No No No No No 0.199 0.229 0.248 0.130 0.136 0.124 0.060 0.071 °;a05 a The RRF and standard deviation reported in this table for the quadrupole and magnetic sector mass spectrometers were determined as single measurements of all congeners in a single day with each instrument. b Student's two-sided t-test was used to determine significant differences of the RRFs. c An F-test was used to determine significant differences of the standard deviations, where F = (std dev!)2/(std dev2)2 with (n-1, n-1) degrees of freedom. d Single measurement. e Cannot test significance of difference between single measurements. �TABLE A-4. PCB STUDENT'S TWO-SIDED t-TEST TO DETERMINE SIGNIFICANT DIFFERENCES OF THE AVERAGE RELATIVE RESPONSE FACTOR (RRF) FOR SOLUTION NO. 1 FOR REPLICATE ANALYSIS ON A SINGLE DAY VERSUS SINGLE ANALYSES ON MULTIPLE DAYS Replicate analyses on single day no. RRF Standard deviation 1 11 29 47 121 136 181 195 207 209 4.073 3.073 2.195 1.062 0.948 0.689 0.383 0.263 0.237 0.213 0.118 0.073 0.048 0.059 0.020 0.016 0.009 0.003 0.008 0.006 congener Single analyses, on multiple days Standard RRF deviation 3.241 2.538 1.899 1.015 0.959 0 . 683 0.374 0.275 0.269 0.230 0.201 0.161 0.100 0.059 0.043 0.058 0.035 0.028 0.032 0.027 t-Statistic Significant differences of RRF at 95% confidence limit? 7.468 6.204 5.483 1.268 -0.479 0.186 0.662 -1.137 -2.479 -1.599 Yes Yes Yes No No No No No Yes No a The RRF and standard deviations were calculated from four replicate measurements completed in the same day. b The RRF and standard deviatons were calculated from seven single measurements from seven different days. �TABLE A-5. RELATIVE RETENTION TIMES (RRT) OF 77 COMMERCIALLY AVAILABLE PCB CONGENERS MEASURED VERSUS 3,3'4,4'-TETRACHLOROBIPHENYL-d6 DETERMINED WITH MAGNETIC SECTOR (VARIAN MAT 311A) AND QUADRUPOLE (FINNIGAN 4023) MASS SPECTROMETERS RRT RRT PCB congener no. 311A 4023 Monochloro1 2 3 0.403 0.481 0.474 0.425 0.490 0.499 Dichloro4 5 1 8 9 10 11 12 14 15 0.518 0.598 0.559 0.590 0.563 0.521 0.649 0.660 0.616 0.677 0.536 0.606 0.579 0.606 0.577 0.534 0.660 0.671 0.628 0.681 Trichloro18 21 24 26 28 29 30 31 33 0.665 0.762 0.685 0.729 0.745 0.719 0.641 0.741 0.760 0.678 0.767 0.694 0.738 0.753 0.728 0.653 0.752 0.769 Tetrachloro40 44 47 49 50 52 53 54 61 65 66 69 70 72 75 77 0.870 0.838 0.814 0.811 0.746 0.804 0.763 0.720 0.898 0.822 0.905 0.800 0.880 0.853 0.816 1.002 0.875 0.843 0.819 0.817 0.751 0.810 0.773 0.731 0.898 0.826 0.908 0.807 0.904 0.856 0.821 1.003 PCB congener no. 311A 4023 Pentachloro87 88 93 97 100 101 103 104 105 116 119 121 0.979 0.913 0.907 0.976 0.878 0.945 0.870 0.829 0.988 0.985 0.964 0.911 0.978 0.915 0.908 0.979 0.884 0.945 0.874 0.836 0.987 0.986 0.965 0.914 Hexachloro128 129 136 137 138 139 141 143 151 153 154 155 156 1.163 1.128 0.994 1.118 1.108 1.037 1.096 1.050 1.020 1.074 1.002 0.929 1.194 1.156 1.127 0.996 1.115 1.103 1.038 1.093 1.051 1.021 1.073 1.004 0.931 1.188 Heptachloro171 181 183 185 1.189 1.178 1.154 1.166 1.187 1.174 1.148 1.161 Octachloro194 195 198 200 202 204 1.355 1.326 1.275 1.203 1.194 1.209 1.351 1.317 1.265 1.199 1.188 1.203 (continued) A-8 �TABLE A-5 (continued) RRT RRT PCB congener no. 311A 4023 Nonachloro206 207 208 1.414 1.336 1.319 1.399 1.330 1.318 PCB congener no. 311A 4023 Decachloro209 1.453 1.440 A-9 �-Iflfl.O E1C DATA: r-<:91VQ2 1765 SCAIIS CS/01/82 8:33:00 CALI: C.ilOIUI 03 OUT OF SAMPLE: FCB MIXTURE—SOLUTION 01 1UL UIJ COUUS.t -16CO HIV (0-6 79EV .2IIA PCS--I1H 40-2II-335-C/ "OII-COl" R.MKE: C I.I2PO LABEL: t! 0, 4.0 OUAH-. A tt. I 0 J 0 BA<;| BASE: H 20. I TO I209 I TO I2CO 3 1. 328 599016. i-H C (U r^ c J3 •H 43 0) a. fX, vj O 1 : rH o t-i 0 --I rC <~M ^» U •H 43 O C 01 in 1 CN ^ 01 W.«20 -H 43 o 43 O -*" o 0. R59 CM u § PH 1 1 J^l 4J 0) to 1 X a 1 If If) U ^. <fr „ •^ 1 II en CM « •* m •* •• i CM i—i 43 CJ cd CO ^ u n CM * CM CNI | I CM I * || to CM CM 1 en - <r •\ •V ^ ro - ro en -; •s si- 1 ** - to H ^ **ll * ^ ^" O OJ M ** § :• in " to to •* <u cd 4J 43 cd C •£* 43 ft 0> PC 1 43 y cfl O CO - 43 0 M 0 rH 43 CJ 6-Octachloro : ,H -g 43 0 0.912 c ^ [-; a. (U 4= Ul ^ P« H-J tachlorobiph O i—* _r? t.1 C iphenyl-dg a) 2 t-i 0 C i orobiphenyl ^*. lorobiphenyl *rt 1 w ft (i 96 .orobiphenyl i-H t CM C °- CM ' 0.315 i 200 3:20 • i ISO 6-.10 . 1 COO je : eo , '" L "*"I~*~ *• -^ • i 1.102 "—* COO 13:20 _'•£' ItJ.0 1C: IB SCAII TIME Figure A-l. Fused silica capillary gas chromatogram of PCB Solution No. 1 analyzed with electron impact mass spectrometry. Experimental conditions for separation and analysis of the PCBs are presented in the experimental section of report. �UV.'. SVECi'tlUil GV20/32 16:42:00 * PATA: 498IE20W7 (Witt CAM: CALF.20UI 07 1:50 E: rest iimiiue — SOLUTIOII BI IUL iiu -J%8 L-lfV 10 6 70I:V .2IIA Bi!5 -1MI 1 10 -211- 525- 10/ "IJJI-COL J33IAKCED <S 15B 2H OT) ICO 9523:-. 50.0- 126 ^ T ti/n no Figure A-2. 1161 i*U ico 173 101! Electron impact ionization mass spectrum of 2-monochlorobiphenyl. �tt'I-CII'JRI U>:12:W * HATA: :: FOB IIIXTIII::;—soi.Minni ni »« mi IKO iaiv io-6 /ot-:v .M\ imr> j-.u (S I5K 2II OTI U17J DASJ; II/'E: CAM: CAM-JttH /:'j» >/ "n;i toi." C329C. 189.0-1 I I-1 NJ 59.0- 99 U~ IP -^-r-^M-. , T-p-^^-, ll/E (00 Figure A-3. 120 220 Electron impact ionization mass spectrum of 3,5-dichlorobiphenyl, �W5/20/I52 1C:'12:00 ^ I!:'H f.AI.1; r;ALK2ft»l IU SAIIPLH: PCB IIIXTIIRE—f.OLUYIOII 11 IIH. IIU COJ1DS.: -1550 HIV 10-6 /OCV .2IIA DBS-1511 110-211-325-10/ "iKlBillAilCIiD (S 15D 211 01) HAM: H/J-: KW niC: 271072. f 255 100.0 1928( II 16 59.0- U) • 10 93 *>-n ft Z W ,1 1?3 if 1 T .Vii^ l._ .iii Li 169 i IIIwll ll/n Figure A-4. \yt 110 ICO ll iiill 100 1 1. I?G »LiJ.. 280 KG I-L.^ i-l 22t} \l •l ?]S 240 2»iC» Electron impact ionization mass spectrum of 2,4,5-trichlorobiphenyl. �MASS M'ECIRHII 05/2t)/82 IS:4?:W + ?):r.1 (Ml: SA!ir»>;: rcc tiixTiim-—SIU.UTIMI n iw. MM BA'JK II/C: 2 '9 P.IIJ: 3&10J2. 0» C«!l»S.: -1550 HIV Mi> 7»JiV .2IIA 1)115-1511 IIR-2II-325--IO/ "iKJ-ljm.1 211 OT> 2 0 iee.0-i 39369 2- i2 1 e 1 • in r\ 8 I8 255 i: 3 '9 1 ii/n J Jilu^jll k 1eo 1 T ' lk-,,1 1 1. l vlll ^ lll^ l 1 50 l ?1 J,l. » ?96 H"--j |—"* 2Wt .,| I,2?i_T213_rJ r-*" 250 I ' 1 ' 1 ' XY3«3 Figure A-5. Electron impact ionization mass spectrum of 2,2',4,A'-tetrachloroblphenyl. �HAW srcciwiii DMA: 'icuil:2r;'f»/ 07:>0 (J7 85/29/82 I6:12:0fl ^ I V : 00 CAM: CAI.I-WI SAUPLE: riJtt HlXTUnC-—SOLUTIOJI ill IUL J I U UGHIS.: -I550 HIV IO-6 70I-V .2IIA 08'j-Ctll IIO-2II-325-IO/ "«;] OH." BASE Il/Et 2:i:} HIC: 12C3M. EKIIAHCED (s isn 211 OT» 59.9 --r^ 350 Figure A-6. Electron impact ionization mass spectrum of 3,3'4 ,4' -tetrachlorobiphenyl-d^. �IIAS-; «)AiA: I'micyv-)"/ nwi UAI.I: tMf.»»l 05/M/«? 10:12:00 + 11:01 SAiim-t res MiXTunti— r,oi.ufi(iii n sw. IIM ar,»S.: -15W) BIV 10-0 /tfl-Y .2IW IWi Till IIO-2II-323-IO/ (s isn 211 on r.n: 05 "»KJ COI." 3078 180.0 2-36 "7 59.9- 191 99 2 «J 163 I "*"* 19-1 2*> .MUl'^rJ it/'-: Figure A-7. |llL,_,,U J50 li5.lV2.ix 'J«0 Electron impact ionization mass spectrum of 2,3',4,5',6-pentachlorobiphenyl. �IIACS SH3MWUI 85/20/J52 16:12:00 ' 12:06 SAtimi: rcc MIXTURE—SOLUTION ni HJI. im I>ATA: CAM: CAI.K20WI II3 323501. COIffiS.; -I5!i0 EIIY 10-5 70EV .2!l-\ DB5 I'jll 110-211 325- It)/ "OSI-COI." DBIAUCl-n <S ISO 2\\ 811 100.0- 25661. 29fl 50.0- It'li 350 u Figure A-8. Electron impact ionization mass spectrum of 2,2',3,3',6,6'-hexachlorobiphenyl. �IIVJS SrtCTtnfll ItAfA: 1?iUE2T,V»7 IKKJI aV.W/R:> H>:'I2:Ofl ^ M:2I iMl : CAI.E2WI 01 SAllI'LE: r«.B IIIXUli::;— -SDUiriO'l 01 HH. 111! COi!D:i.: -IWU DIY 10 0 TDtV .2JIA llffi I'-ll 1 10-211-325- IO/ "flH-Cor BA';C ll/li: :>V3 niC: 3755J?. nr.i,\i:ci;u is 150 211 OT» IOH.O IGOC1. i? 120 50.0 1 i 1 |i 7 liy 6 i: 0 4 .1 ir 129 IGO 3 1 149 ll I [1 lijW^M K9 IK? I 1" Ti 2W 22» 249 1680 3 1 i i—' oo 50.0- 3'.9 1 289 ,11 JJ-r^L•^*-i-» I111,.. ??*!|,. . 274 >*Wvi 1 ' i il • » • . | i • ILL - i i 3CO | • 300 • •i 1 i ' [ • • * •r ISO Figure A-9. Electron impact ionization mass spectrum of 2,2',3,4,4',5,6-heptachlorobiphenyl. �HAY; si'tornini JIATA: 1!HMI-2Wli/ ()r)7l CA.I.I: CAI.E?C!M 0.1 95/20/82 lti:42:W + Ifi: II SAiirLE: njB uixiimi: — souinoii ai mi. iiu cniros.: -I559 ia iv so r, voi-v .?IIA m^-r.ii n')-2ll-32r.-l'-)/ EII)IAt:r;GD (S 1511 211 OD BASE II/E: I/O KM): 'JJOB11. "OII-COI." i; ') 100.0 I6C7. 50.0- 2 r, i '3 p«125 T 95 I5C lee 120 149 100 183.0-1 1 |R7 159 51 "T8 II I «S ' 209 229 ^ 269 21« IGT. 3! 0 29 1 50.8393 271 I I . ll/fi 209 Figure A-10. I I1 1 \\.,m 388 3ji 1,1.1,,, 320 I 1 •I, i | 310 368 I 1, • • i • | •i i . |i i • • | . i . . ! i i . . | • . ' 308 1«8 120 «0 Electron impact ionization mass spectrum of 2,2',3,3',4,4',5,6-octachlorobiphenyl. �I»ATA: 'IOOIE5AW7 0970 CMI: CAI.K2MM OJ 05/20/82 l«:12:0fl ^ lfi:l'.» SAiffLE: rcis mxiuiw- — soumnii n iw. RASE »/f: Ifti HIO: 123016. lt!J OKWS.t -IfBW DIV 10 li 70EV .2IIA IWK-IriH IIO-2II-325-IO/ "UH--COI." BOIMXED (S I5B 2H OT> 1 G 17661. 1 2 2 50.0- 111 1BO 10 fl 125 1 t 1* l ' f • 101 1 it i I-J-1-.-H J 10 o II •PI 2 1 1 T \L\ 11LjllLjll^llL I 20 N/E 183.9- 1 » IwJ*. iw» III ^?V IllllM^^-JJ llti*^,^ alii . 2^1 220 |i|iii _ 240 I 260 230 41-1 ji 3 2 50.0- 357 290 i l.it, ni . 3dO Figure A-ll. 1 .,332 320 31? 3W jjll. 3CO -,-r-i-i-r- 3(KI 4W> 120 13 1 - 4G8 I, Electron impact ionization mass spectrum of 2,2',3,3',4,4',5,6,6f-nonachlorobiphenyl. �liAG'J SI'ECTBUll r II-MA: '190i 9 >/.WC2 10:12:00 + 17:13 CAI.l: CAM SAiii'LK: ten iiixitirr- soi.urinn ui an. nu ClISiDS.: -1559 HIV 10-6 70EV .2IIA D»r»-I5ll 110-711-325-!0/ Eii!)Ata:nD (s i5n 211 OT> w'JW UIWJ3 '"I f)3 BAJiC H/C: 21 i '• Rir, : 439208. "DJl-COL" 2 1 100.0- 228I6. I7« 2 9 50.0- 107 160 96 II/E 'i 108 262 231 Jl|ljYjl|h. 159 1 2?7 . 3f« 259 - •J 356 50.0- 228I 3 14« 321 ii/n Figure A-12. J-lOJULff^-.J 350 L^^MJEL 1 1, 150 ... 5C3 Electron impact ionization mass spectrum of 2 , 2 ' , 3 , 3 ' , 4 , 4 ' , 5 , 5 ' , 6 , 6 ' - d e c a c h l o r o b i p h e n y l . �ra Ltn.-oi.ii mss: 200 <Rei-.amr:Oi.ii MASS.uiPtD-G TiiTi!Aanc!:iH!miHiYL. isosrei: 15210 J:tT:D-0 nnR.'.Oil.U!:i)-CIFi!BIY[.. ISOIII-i: 0210 r.t*)/«iiT./T.H;.,\:iT.i <AV : i.ocoi i. not ST.IJS-V.- (.1.000 * ST.DEV.0.009 i.e i ts3 N> 0.899 HATE 5/I1/82 Figure A-13. 5/21/02 6/ 3/02 Response factor plotted on a day-to-day basis for the internal standard, 3,3',4,4'-tetrachlorobiphenyl-d6, in Solution No. 1. �RESl'OllSIi L I B : 0 1 . 2 i IIASS: 1«8 (IW.COIIl'iOI. I. MASS: 2!l<!) OlP:2-liai!)ll1IU)l!0-Bin«3IYl..JSCHER fll EEF : I)-G TEHACIILORO-BiriltliYI.. ISOIIEI! 0210 tAKEA/atF.ALEAl/tAIIT./Rb'F.AIIT.) (AV: 'J.51-11 5.600 ... 0.431 ST.TiEV.- n.ni Sl.'il 4. oca 1 NJ u> 3.C80 BATH ^'1-1/02 5/2-1/32 Figure A-14. Relative response factor for 2-monochlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �B:Ol.^i MASS: 222 (KKI-.ailinOI. l> IIAC,S: 2!«5> aiP:3.3--nioii.onn-Binn3ivL. ISWIHR tin B1-F:D-C TKTKACmjmO-ltiniHIYL. 1SOIJER B2IO T./l:m:.AIIT. > (AV: 2.7.}2» .t.AW ST.niv.- t 0.206 X ST.DEY.II-..1G6 3 r n.naa S'tt =• l i . • 2.J18H :; 2.C8!l 2.VC3 i to *• X 2.683 X X X 2.5*3 :< 2.480 •H 2.393 •> oon 5/24/02 6/ 3/32 Figure A-15. Relative response factor for 3,3'-dichlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-dg on a day-to-day basis. �1.111:01.1i IttSS: 256 (RiiF-COllPiOI. li IHSS.- 2!)OI aip:2.i.5-TRiciiinr.o-niriiiiiiYj.. isotiER r.29 REFrH-O TL:inAaiI.Ol;0~BirilFJIVI.. ISOMER II2IO i:nA)/(AiiT./Ki;r.AiiT.) <AV : 2.0051 2.^(10 ST.DLV.0. IR3 £ ST.IEY.n. I38 SMI = i.e 2. x I NJ 1.999 1.899 1.709 DATE 5/11/82 5/2-1/82 6/ 3/B2 Figure A-16. Relative response factor for 2,4,5-trichlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �RESrOKSF 1.111:01.5. IUSS: 202 (MF. COUP: 01. I , IKSS: 29CI niP:2.2M.'r-TFiRAciii.ojH)-iiirii0iYL. isoum: iw REf:D-0 lEll:AaiLi?r.O-Diri!BM.. ISOIinn U2IO (AnP,\/nCI:.A!:i'A)/(AIIT./!iLl;.AIIT.) (AV: 1.032) 1.280 ST.PMV.0.058 Z Sf.DEY.5.596 1.9 >( 1.190 •• i ro * O" X X X X ): O.C33 DATK l'l/82 5/21/02 G/ 3/02 Figure A-17. Relative response factor for 2,2',4,4'-tetrachlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-c^ on a day-to-day basis. �r-rsroiiSK UB.-OI.GI iwss: 326 (i!r.r.coiir oi.ii MASS: a\T-.2.y .1.5'.r.-rciiTAan.oi;o-mriiBiYi.. isomin : 11121 RCF:n-s iF.ii:Aaii.oi:o-iiii'iiaiYi.. isounK ir»io (Ar.t A/Cnp./.l:M>/(AIJT./r;nF./MIT.) (AV: O.'tIO) i. can ; -SU'liV.(1.952 Z ST.C-EV.T 477 Sl'il » 6. £39 i S3 -J DATE 5/11/82 5/21/82 6/ 3/82 Figure A-18. Relative response factor for 2,3',4,5',6-pentachlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �1.18:91.7. IIASS: 3C9 (l:EF.CO!H':fll. I. IIASS: diP^^'.s.s'.e.e'-iiiLVAaiLenn-BmiBiYJ.. isoiirc a i3t> BEF-.D-6 TElfiAaiI.OKO-|im!BIYI.. ISOIIKIt B2IO iiiiA^iAnT./iraF.AiiT.) <AV : o.ocs* O.J'O'.I ST.PEY.- O.OH £ ST.9EY.0.370 i.e X o.ynn * Is) 00 6.609 0.509 DATE 5/1V82 C/ 3/82 Figure A-19. Relative response factor for 2,2',3,',6,6'-hexachlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �RESPONSE i.in : oi.n, iwss: 391 (nnr.raiirrfli.ii MASS: air:2.2\3.i.i'.5.c-iiEi'T'.a!i.or.o-niniFiiYi.. isoiinn i lot EEF:«-c rETR/.ciu.ono-iiiMiBn'i.. iswim: 11210 : (AREA/niH .Ar,FA»/(AIIT./UtF.AIIT.) (AV: O.UO 0.377) ST.I'EV.0.827 x si.LEY.- tt.1?0 V.035 0. MO SMI - 1.0 x U.303 0.380 _• i isj VO I— 0.370 0.360 6.350 0.310 0.330 0.320 DATE 5/11/82 5/21/82 C/ 3/82 Figure A-20. Relative response factor for 2,2',3,4,4',5,6-heptachlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �r.Fsro;isr i.iitroi.o, IIASS 430 (nHP.cniir oi.i, IIASS REP:no Ti;u;.\aiLono-i;iriinj\'i.. isotinn U2io : : aiP:2,2'.3.3 t .1.4'.!;.6-Oi:TAOILORO-BIl l IK.ilV1.. ISWIFR 0 105 tlT. > (AV: 0.2/B> 1'. J/U sr.nr.v.s 0.021 ? Z Sf.DCV.7.972 '<; 0. ^-fi) sK 0. 203 - 1.0 •\c u_«. . .- , . _ _ i 1 0 O v *VM% 0 .270 , | 0.269 * X X 0.259 )C X n -MO — 5/2-1/82 Figure A-21. •• — • ' • II- »- — 1 C/ 3/02 Relative response factor for 2,2',3,3' ,4,4',5,6-octachlorobiphenyl in Solution No. 1 calculated versus 3,3' ,4,4'-tetrachlorobiphenyl-d^ on a day-to-day basis. �nnsronst LIB : OI.IOI HAGS: 101 (RnF.cniir oM air:2.2\3.3vi.i\5,G.<v-i;niiAciiLORiHiiriiniivi..: REF:iM> TErc/xiiioRo-DiFiintm.. isotmn 11210 (Am:4/RE»--.ARirA>/<AMT./r.FF.AHr.J <AV: 0.320 2301 0.257) ST.WiV.O.P29 5; sT.rtv.I I 239 SDil - 1.0 0.209 -v0.2J10 0.2/0 I u> 0.250 0.210 0.239 0.220 DATE Figure A-22. 5/14/82 5/21/82 6/ Relative response factor for 2,2',3,3',4,4',5,6,6'-nonachlorobiphenyl in Solution No.!1 calculated versus 3,3*,4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �BESrOMSE L I B : O I . H i MASS: 498 (C£F.COIir : 0l. li MASS: 2991 air:2.2\3.r.4.1\5.5\G.6'-DECAaH.ORa-BIM!iaiYL. ISOIIHIi Q 209 BEF:D 6 TETIUdlLOBO-BlPliraril, ISOilEII (1210 i!EA>/(AIIT./i;EF.MIT.) <AV: 0.223) 0.270 ST.IlS-V.C.«22 Z Sf.OEV.9.8/7 n ?;T MM •= 1.8 r. • •• 0.210 I U> NJ 0.2JO 0.220 ) 0.210 I x 0.200 0. I'M DATIi 5/M/82 5/2-1/U2 C/ 3/32 Figure A-23. Relative response factor for 2,2' ,3,3',4,4',5,5',6,6'-decachlorobiphenyl in Solution No. 1 calculated versus 3,3',4,4'-tetrachlorobiphenyl-d6 on a day-to-day basis. �APPENDIX B ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN COMMERCIAL PRODUCTS AND PRODUCT WASTES B-l �THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN COMMERCIAL PRODUCTS AND PRODUCT WASTES 1.0 Scope and Application 1.1 This is a gas chromatographic/electron impact mass spectrometric (GC/EIMS) method applicable to the determination of chlorinated biphenyls (PCBs) in commercial products and product wastes. The PCBs present may originate either as synthetic by-products or as contaminants derived from commercial PCB products (e.g., Aroclors). The PCBs may be present as single isomers or complex mixtures and may include all 209 congeners from monochlorobiphenyl through decachlorobiphenyl listed in Table 1. 1.2 The detection and quantitation limits are dependent upon the complexity of the sample matrix and the ability of the analyst to remove interferents and properly maintain the analytical system. The method accuracy and precision will be determined in future studies. 1.3 This method is restricted to use by or under the supervision of analysts experienced in the use of gas chromatography/mass spectrometry (GC/MS) and in the interpretation of gas chromatograms and mass spectra. Prior to sample analysis, each analyst must demonstrate the ability to generate acceptable results with this method by following the procedures described in Section 14.2. 1.4 The validity of the results depends on equivalent recovery of the analyte and 13C PCBs. If the *3C PCBs are not thoroughly incorporated in the matrix, the method is not applicable. 1.5 During the development and testing of this method, certain analytical parameters and equipment designs were found to affect the validity of the analytical results. Proper use of the method requires that such parameters or designs must be used as specified. These items are identified in the text by the word "must." Anyone wishing to deviate from the method in areas so identified must demonstrate that the deviation does not affect the validity of the data. Alternative test procedure approval must be obtained from the Agency. An experienced analyst may make modifications to parameters or equipment identified by the term "recommended." Each time such modifications are made to the method, the analyst must repeat the procedure in Section 14.2. In this case, formal approval is not required, but the documented data from Section 14.2 must be on file as part of the overall quality assurance program. B-2 �TABLE 1. NUMBERING OF PCB CONGENERS3 NO. Structure NO. Henoe(ilaroa<oh«ny1» 1 2 3 2 3 4 D<eh1orob1p(itny1t 4 5 6 7 3 9 10 11 12 13 14 15 }:]; 2)4 !:$' 2,6 3,3' 3,4 1:1' 4,4' Trlehlorablphtnyli 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 13 34 35 36 37 33 39 40 41 42 43 44 45 46 47 48 49 50 51 2.2'.3 2.2', 4 2.2', 5 2.2'. 6 2.3.3' 2.3,4 2.3.4' 2,3.5 2.3.6 2. 3', 4 2, 3'. 5 2, 3'. 6 2.4.4' NO. 52 53 54 5$ 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 105 106 107 108 109 110 2,2*. 5.5' 2,2',5.6* 2.2'.6.6' 2.3,3- .4 2.3,3'. 4' 2,3,3'. 5 2.3,3'.5' 2,3.3'. 6 2.3.4,42.3.4.5 2,3,4.6 2.3,4'. S 2, 3,4', 6 2,3.5,6 2,3' 2,3' 2.3' 2.3* 2.3' 2.3' 2.3' 2.3' in 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 4.4' 4,5 4,5' 4,6 4'. 5 4'. 6 5.5' S',8 2,4,4'. 5 2.4.4', 6 2'. 3, 4. 5 3,3'. 4. 4' 3,3',4,5 3,3'.4.5' 3,3'.5.5' 3,4.4'. 5 Structure 2,3.3' .4' 2.3.3' ,5 2.3.3' '.5 2.3.3' .5' 2.3.3' .6 2.3.3' ',6 2.3.3' ,5' 2.3,3' ,6 2.3.3' ',6 2.3,4.4 .5 2.3.4,4 ,6 2.3.4,5 6 2,3.4', .6 2.3'.*, 2.3'. 4. '.'6 2.3',4, ,5' 2.3". 4, 2'.3.3' 4J5 2'.3.4,4 '.5 2',3.4,. ,5' 2'.3.4.i .6' 3.3' .4.4 '.5 3.3'. 4.. .5' HemchU robtptienyls 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 15$ 156 157 158 159 160 PenUcM orcb< phtny 1 i 2.2'.3.3',4 2.2', 3. 3'. 5 2.2',3.3',6 2,2',3,4. 4' 2.2'. 3. 4. 5 2.2', 3, 4.5' 2,2'. 3,4.6 2,2'.3,4, 6' 2.2'.3.4',5 2,2'. 3,4'. 6 2.2',3,5,5' 2,2'. 3,5,6 2,2'.3.5,6' 2.2', 3, 5'. 6 2.2'. 3. 6.6' 2.2', 3'. 4.5 2,2'. 3' .4. 6 2.2', 4, 4' .5 2,2-.4.4'.6 2.2'. 4, 5,5' 2, 2' .4,5.6' 2.2',4.5',6 2, 2'. 4,6.6' 2.2', 4. 5 2.2' .4, 5' 2.2". 4, 6 2,2',4.6' NO. rtntacli oroMphenylj Tetneft 1 oreo 1 pinny 1 $ 82 83 84 2,4',S 85 2.4'.6 86 2'. 3. 4 87 2'. 3,5 88 3,3', 4 89 3.3',! 90 3,4,4' 91 3.4,5 92 3.4'.S 93 94 TttneM orob< ohnyl » 95 96 97 2.2'. 3.3' 2.2'. 3,4 98 2.2'.3,4' 99 2.2'. 3,5 100 2.2'.3.S' 101 2.2',3.6 102 2.2' .3,6' 103 2.2-.4.4 1 104 2,4.5 2.4.6 structure 2.2',3.. '.4,4' 2.2'. 3.. '.4.5 2.2'. 3. : '.4,5' 2.2'. 3.: '.4.6 2.2'. 3.: ',4.6' 2. 2'. 3,. '.5,5' 2.2' .3.. '.5,6 2.2* .3.: ',5, ' 2.2'. 3.: ',6, ' 2.2' .3,4 ,*'. 2.2'. 3. 4 .4'. ' 2,2' .3. 4 ,*', Z,2'.3,4 ,4' . ' 2.2'. 3.4 .5,5' 2,2'. 3, 4 .5.6 2.2'.3.4 .5,6' 2.2'. 3, 4 .5'. 6 2.2'. 3. 4 .6,6' 2.2', 3.4 '.5.5' 2,2',3.4 '.5,6 2,2'. 3, 4'.5.6' 2,2'. 3, 4 '.5- ,6 2.2'. 3.4 '.6.6' 2.2'.3.S .5' .6 2.2'. 3.S .6,6' 2,2' .4.4 '.5.5' 2,2'. 4.4 '.5.6' 2.2'. 4. 4 ',6,6' 2,3,3'. 4 .4'. 5 2.3, 3'. 4 ,4'.S' 2.3,3'. 4 .4'. 6 2.3,3'. 4 .5.5' 2.3.3'. 4 .5.6 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 2, 2'. 3, 3', 4. 4 ' , 5 2.2' .3. 3 ' . 4. 4 ' , 6 2, 2', 3. 3 ' , 4. 5. 5' 2, 2 ' , 3, 3 ' , 4. 5. 6 2,2'. 3. 3 ' , 4. 5 ' , 6 2, 2'. 3. 3 ' , 4.6. 6' 2, 2', 3. 3 ' , 5, 5 ' . 6 2. 2 ' . 3. 3 ' , 5, 6, 6' 2.2'. 3, 4, 4 ' , 5.5' 2. 2' .3, 4, 4'. 5, 6 2. 2' .3. 4. 4 ' , £ . 6 ' 2,2'.3,4,4',5',6 2, 2 ' , 3, 4, 4 ' . 6, 6' 2.2>.3,4,5.5'.6 2. 2'. 3, 4,5. 6, 6' 2,2', 3. 4 ' . 5. 5 ' . 6 2, 2', 3, 4 ' , 5,6, 6' 2.3, 3'. 4. 4', 5,5' 2, 3, 3', 4,4', 5, 6 2, 3, 3', 4, 4 ' , 5' ,6 2. 3, 3', 4, 5. 5 ' . 6 2.3,3' .4' .5.3'. 6 Ocucnlorotiionenyls 194 195 196 197 198 199 200 201 202 203 204 205 2,2', 3. 3'. .4', 5. 5' 2, 2'. 3, 3 ' , , 4 ' , 5, 6 2, 2 ' , 3, 3 ' . ,4', = , 5 ' 2, 2', 3, 3', ,4' ,5, 6' 2, 2', 3. 3 ' , S . S ' . S 2, 2', 3. 3 ' , ,5.6,6' 2.2', 3. 3 ' . .5'. 6. 6' 2, 2', 3, 3 ' , . S . 5 ' , 5 ' 2.2', 3, 3' .5. 5 ' , 6, 6' 2, 2'. 3, 4 . 4 ' , 5, 5 ' , 6 2.2* .3, 4, 4' ,5. 6.6' 2.3.3',4,4'f5,5',6 206 207 208 2.2'.3,3'.4.4',5,5',6 2, 2'. 3. 3 ' , 4, 4 ' . 5, 5, 6' 2. 2'. 3, 3'. 4, 5. 5 ' , 6, 6' NofueHloro61on«nyls Bteichlorebfomnyl Anil. Clw*., 302. 20-31 (I960). B-3 2.3,3'. 4, 5 ' , 6 2, 3, 3'. 4 ' . 5. 5' 2. 3.3'. 4'. 5, 6 2, 3. 3'. 4'. 5 ' , 6 2,3, 3'. 5, 5 ' . 6 2, 3, 4. 4 ' . 5.6 1 2.3',4,4'.5.5 2.3',4.4',5'.5 3. 3' .4. 4'. 5, 5' HtptieM orab< phtny 1 1 209 **dopt*d froa ttllicftrfttr, K. >nd Ztll, *., FrtstMus 2. Structure Htmehlorooiphtnylt Z.2',3.3'4,4',5,5',6.6' �2.0 Summary 2.1 The process or product must be sampled such that the specimen collected for analysis is representative of the whole. Statistically designed selection of the sampling position, time, or discrete product units should be employed. The sample must be preserved to prevent PCB loss prior to analysis. Customary inventory storage may be adequate for products. For intermediates, process samples, and other non-product specimens, storage at 4°C with optional preservation at low pH is recommended. 2.2 The sample is mechanically homogenized and subsampled if necessary. The sample is then spiked with four 13C PCB surrogates and the surrogates incorporated by further mechanical agitation. 2.3 The surrogate-spiked sample is extracted and cleaned up at the discretion of the analyst. Simple dilution or direct injection is permissible. Possible extraction techniques include liquidliquid partition, thermal desorption, and sorption onto resin columns followed by solvent desorption. Cleanup techniques may include liquid-liquid partition, sulfuric acid cleanup, saponification, adsorption chromatography, gel permeation chromatography, or a combination of cleanup techniques. The sample is diluted or concentrated to a final known volume for instrumental determination. 2.4 The PCB content of the sample extract is determined by capillary (preferred) or packed column gas chromatography/electron impact mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected ion monitoring (SIM), full scan, or limited mass scan (IMS) mode. 2.5 PCBs are identified by comparison of their retention time and mass spectral intensity ratios to those in calibration standards. 2.6 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners, using the response of the 13C surrogate to compensate for losses in workup and determination and instrument variability. 2.7 The PCBs identified by the SIM technique may be confirmed by full scan CGC/EIMS, retention on alternate GC columns, other mass spectrometric techniques, infrared spectrometry, or other techniques, provided that the sensitivity and selectivity of the technique are demonstrated to be comparable or superior to GC/EIMS. 2.8 The analysis time is dependent on the extent of workup employed. The time required for instrumental analysis of a single sample, excluding data reduction and reporting, is about 30 to 45 min. 2.9 Appropriate quality control (QC) procedures are included to assess the performance of the analyst and estimate the quality of the results . These QC procedures include the demonstration of laboratory capability: periodic analyst certification, the use of control B-4 �charts, and the analysis of blanks, replicates, and standard addition samples. A quality assurance (QA) plan must be developed for each laboratory. 2.10 While several options are available throughout this method, the recommended procedure to be followed is: 2.10.1 The sample is collected according to a scheme which permits extrapolation of the sample data to the whole product or product waste. 2.10.2 The sample is preserved to prevent any loss of PCBs or changes in matrix which may adversely affect recovery. 2.10.3 The sample is mechanically homogenized and subsampled if necessary. 2.10.4 The sample is spiked with four 13C PCB surrogates (4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl; 2,2',3,3',5,5*,6,6"-octachlorobiphenyl; and decachlorobiphenyl). 2.10.5 Normally, the sample is extracted, although dilution may also be used. 2.10.6 The extract is cleaned up and concentrated to an appropriate volume. 2.10.7 An aliquot of the extract is analyzed by CGC/EIMS operated in the SIM mode. On-column injections onto a 15-m DB-5 capillary column, programmed (for toluene solutions) from 110° to 325°C at 10°/min after a 2-min hold is used. Helium at 45-cm/sec linear velocity is used as the carrier gas. 2.10.8 PCBs are identified by retention time and mass spectral intensities. 2.10.9 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners. 2.10.10 The total PCBs are obtained by summing the amounts for each homolog found, and the concentration is reported as micrograms per gram. 3.0 Interferences 3.1 Method interferences may be caused by contaminants in solvents, reagents, glassware, and other sample processing hardware, leading to discrete artifacts and/or elevated baselines in the total ion current profiles. All of these materials must be routinely demonstrated to be free from interferences by the analysis of laboratory reagent blanks as described in Section 14.4. B-5 �3.1.1 3.1.2 3.2 4.0 Glassware must be scrupulously cleaned. All glassware is cleaned as soon as possible after use by rinsing with the last solvent used. This should be followed by detergent washing with hot water and rinses with tap water and reagent water. The glassware should then be drained dry and heated in a muffle furnace at 400°C for 15 to 30 min. Some thermally stable materials, such as PCBs, may not be eliminated by this treatment. Solvent rinses with acetone and pesticide quality hexane may be substituted for the muffle furnace heating. Volumetric ware should not be heated in a muffle furnace. After it is dry and cool, glassware should be sealed and stored in a clean environment to prevent any accumulation of dust or other contaminants. It is stored inverted or capped with aluminum foil. The use of high purity reagents and solvents helps to minimize interference problems. Purification of solvents by distillation in all-glass systems may be required. All solvent lots must be checked for purity prior to use. Matrix interferences may be caused by contaminants that are coextracted from the sample. The extent of matrix interferences will vary considerably from source to source, depending upon the nature and diversity of the sources of samples. Safety 4.1 The toxicity or carcinogenicity of each reagent used in this method has not been precisely defined; however, each chemical compound should be treated as a potential health hazard. From this viewpoint, exposure to these chemicals must be reduced to the lowest possible level by whatever means available. The laboratory is responsible for maintaining a current awareness file of OSHA regulations regarding the safe handling of the chemicals specified in this method. A reference file of material data handling sheets should also be made available to all personnel involved in the chemical analysis. 4.2 Polychlorinated biphenyls have been tentatively classified as known or suspected human or mammalian carcinogens. Primary standards of these toxic compounds should be prepared in a hood. Personnel must wear protective equipment, including gloves and safety glasses. Congeners highly substituted at the meta and para positions and unsubstituted at the ortho positions are reported to be the most toxic. Extreme caution should be taken when handling these compounds neat or in concentrated solutions. This class includes 3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotopically labeled). B-6 �4.3 4.4 5.0 Diethyl ether should be monitored regularly to determine the peroxide content. Under no circumstances should diethyl ether be used with a peroxide content in excess of 50 ppm, as an explosion could result. Peroxide test strips manufactured by EM Laboratories (available from Scientific Products Company, Cat. No. P1126-8 and other suppliers) are recommended for this test. Procedures for removal of peroxides from diethyl ether are included in the instructions supplied with the peroxide test kit. Waste disposal must be in accordance with RCRA and applicable state rules. Apparatus and Materials 5.1 Sampling containers - Amber glass bottles, 1-liter or other appropriate volume, fitted with screw caps lined with Teflon. Cleaned foil may be substituted for Teflon if the sample is not corrosive. If amber bottles are not available, samples should be protected from light using foil or a light-tight outer container. The bottle must be washed, rinsed with acetone or methylene chloride, and dried before use to minimize contamination. 5.2 Glassware - All specifications are suggestions only. Catalog numbers are included for illustration only. 5.2.1 Volumetric flasks - Assorted sizes. 5.2.2 Pipets - Assorted sizes, Mohr delivery. 5.2.3 Micro syringes - 10.0 Ml for packed column GC analysis, 1.0 pi for on-column GC analysis. 5.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm ID (Kontes K-420540-9011 or equivalent). 5.2.5 Gel permeation chromatograph - GPC Autoprep 1002 (Analytical Bio Chemistry Laboratories, Inc.) or equivalent. 5.2.6 Kuderna-Danish Evaporative Concentrator Apparatus 5.2.6.1 Concentrator tube - 10 ml, graduated (Kontes K-570050-1025 or equivalent). Calibration must be checked. Ground glass stopper size (519/22 joint) is used to prevent evaporation of solvent. 5.2.6.2 Evaporative flask - 500 ml (Kontes K-57001-0500 or equivalent). Attached to concentrator tube with springs (Kontes K-662750-0012 or equivalent) . 5.2.6.3 Snyder column - Three ball macro (Kontes K-503000-0121 or equivalent). B-7 �5.3 Balance - Analytical, capable of accurately weighing 0.0001 g. 5.4 Gas chromatography/mass spectrometer system. 5.4.1 Gas chroraatograph - An analytical system complete with a temperature programmable gas chromatograph and all required accessories including syringes, analytical columns, and gases. The injection port must be designed for oncolumn injection when using capillary columns or packed columns. Other capillary injection techniques (split, splitless, "Grob," etc.) may be used provided the performance specifications stated in Section 7.1 are met. 5.4.2 Capillary GC column - A 12-20 m long x 0.25 mm ID fused silica column with a 0.25 (Jm thick DB-5 bonded silicone liquid phase (J&W Scientific) is recommended. Alternate liquid phases may include OV-101, SP-2100, Apiezon L, Dexsil 300, or other liquid phases which meet the performance specifications stated in Section 7.1. 5.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column packed with 3% SP-2250 on 100/120 mesh Supelcoport or equivalent is recommended. Other liquid phases which meet the performance specifications stated in Section 7.1 may be substituted. 5.4.4 Mass spectrometer - Must be capable of scanning from 150 to 550 daltons every 1.5 sec or less, collecting at least five spectra per chromatographic peak, utilizing a 70-eV (nominal) electron energy in the electron impact ionization mode and producing a mass spectrum which meets all the criteria in Table 2 when 50 ng of decafluorotriphenyl phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine] is injected through the GC inlet. Any GC-to-MS interface that gives acceptable calibration points at 10 ng per injection for each PCB isomer in the calibration standard and achieves all acceptable performance criteria (Section 10) may be used. Direct coupling of the fused silica column to the MS is recommended. Alternatively, GC-to-MS interfaces constructed of all glass or glasslined materials are recommended. Glass can be deactivated by silanizing with dichlorodimethylsilane. 5.4.5 A computer system that allows the continuous acquisition and storage on machine-readable media of all mass spectra obtained throughout the duration of the chromatographic program must be interfaced to the mass spectrometer. The data system must have the capability of integrating the abundances of the selected ions between specified limits and relating integrated abundances to concentrations using the calibration procedures described in this method. The computer must have software that allows B-8 �TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA Mass Ion abundance criteria 197 198 199 Less than 1% of mass 198 100% relative abundance 5-9% of mass 198 275 10-30% of mass 198 365 Greater than 1% of mass 198 441 Present, but less than mass 443 442 Greater than 40% of mass 198 443 17-23% of mass 442 B-9 �searching any GC/MS data file for ions of a specific mass and plotting such ion abundances versus time or scan number to yield an extracted ion current profile (EICP). Software must also be available that allows integrating the abundance in any EICP between specified time or scan number limits. 6.0 Reagents 6.1 Solvents - All solvents must be pesticide residue analysis grade. New lots should be checked for purity by concentrating an aliquot by at least as much as is used in the procedure. 6.2 Calibration standard congeners - Standards of the PCB congeners listed in Table 3 are available from Ultra Scientific, Hope, Rhode Island; or Analabs, North Haven, Connecticut. 6.3 Calibration standard stock solutions - Primary dilutions of each of the individual PCBs listed in Table 3 are prepared by weighing approximately 1-10 mg of material within 1% precision. The PCB is then dissolved and diluted to 1.0 ml with hexane. The concentration is calculated in mg/ml. The primary dilutions are stored at 4°C in screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. Primary dilutions are stable indefinitely if the seals are maintained. The validity of primary and secondary dilutions must be monitored on a quarterly basis by analyzing four quality control check samples (see Section 14.2). 6.4 Working calibration standards - Working calibration standards are prepared that are similar in PCB composition and concentration to the samples by mixing and diluting the individual standard stock solutions. Example calibration solutions are shown in Table 3. The mixture is diluted to volume with pesticide residue analysis quality hexane. The concentration is calculated in ng/ml as the individual PCBs. Dilutions are stored at 4°C in narrow-mouth, screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. These secondary dilutions can be stored indefinitely if the seals are maintained. These solutions are designated "CSxxx," where the xxx is used to encode the nominal concentration in ng/ml. 6.5 Alternatively, certified stock solutions similar to those listed in Table 3 may be available from a supplier, in lieu of the procedure described in Section 6.4. 6.6 DFTPP standard - A 50-ng/|Jl solution of DFTPP is prepared in acetone or another appropriate solvent. 6.7 Surrogate standard stock solution - The four 13C-labeled PCBs listed in Table 4 may be available from a supplier as a certified solution. This solution may be used as received or diluted further. These solutions are designated "SSxxx," where the xxx is used to encode the nominal concentration in (Jg/ml. B-10 �TABLE 3. Homolog CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)a Congener no. CS100 CS1000 CS050 CS010 1 1 1,040 104 52 10 1 3 1,000 100 50 10 2 7 1,040 104 52 10 3 30 1,040 104 52 10 4 50 1,520 152 76 15 5 97 1,740 174 87 17 6 143 1,920 192 96 19 7 183 2,600 260 130 26 8 202 4,640 464 232 46 9 207 5,060 506 253 51 10 209 4,240 424 212 42 4 255 255 255 255 1 211 (RS) 104 104 104 104 4 212 (RS) 257 257 257 257 8 213 (RS) 407 407 407 407 10 a 210 (IS) 214 (RS) 502 502 502 502 Concentrations given as examples only. B-ll �TABLE 4. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100) CONTAINING 13 C-LABELED PCBsa Congener no. Compound Concentration ((jg/ml) 211 104 212 (13C12)3,3',4,4'-tetrachlorobiphenyl 257 213 (13C12)2,2(,3,3',5,5',6,6'-octachlorobiphenyl 395 214 a (l',2',3',4l,5',6l-13C6)4-chlorobiphenyl (13C12)decachlorobiphenyl 502 Concentrations given as examples only. B-12 �6.8 Internal standard solution - A solution of d6-3,3',4,4'-tetrachlorobiphenyl is prepared at a nominal concentration of 1-10 mg/ml in hexane. The solution is further diluted to give a working standard. 6.9 Solution stability - The calibration standard, surrogate, and DFTPP solutions should be checked frequently for stability. These solutions should be replaced after 6 months, or sooner if comparison with quality control check samples indicates compound degradation or concentration change. 6.10 Quality control check samples will be supplied by the Agency. 7.0 Calibration 7.1 The gas chromatograph must meet the minimum operating parameters shown in Tables 5 and 6, daily. If all criteria are not met, the analyst must adjust conditions and repeat the test until all criteria are met, 7.2 The mass spectrometer must meet the minimum operating parameters shown in Tables 2, 7, and 8, daily. If all criteria are not met, the analyst must retune the spectrometer and repeat the test until all conditions are met. 7.3 The PCB response factors (RF ) must be determined using Equation 7-1 for the analyte homologs? A x M. RF = -£ ~ Eq. 7-1 P A is x Mp where RF = response factor of a given PCB congener A = area of the characteristic ion for the PCB congener ™ peak M = mass of PCB congener injected (nanograms) A. = area of the characteristic ion for the internal standard peak M. = mass of internal standard injected (nanograms) IS Using the same conditions as for RF , the surrogate response factors (RF ) must be determined using Equation 7-2. s A x M. w *= t f ^ where A = area of the characteristic ion for the surrogate peak M s = mass of surrogate injected (nanograms) Other terms are the same as defined in Equation 7-1. B-13 �TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM Recommended Parameter Tolerance Liquid phase Finnigan 9610 15 m x 0.255 mm ID Fused silica DB-5 (J&W) Liquid phase thickness 0.25 |Jm Carrier gas Helium Carrier gas velocity 45 cm/sec Injector Injector temperature On-column (J&W) c Optimum performance Other Optimum performance Injection volume 1.0 plc 70°C (2 min)d 70°-325°C at 10°C/min£ Other Other Other Transfer line temperature None 280°C Glass jet or othe p Optimum6 Tailing factor 0.7-1.5 0.4-3 Peak width 7-10 sec < 15 sec Gas chromatograph Column Other Other Other nonpolar or semipolar < 1 pro Hydrogen Optimum performance p Initial column temperature Column temperature program Separator a Substitutions permitted with any common apparatus or technique provided performance criteria are met. b Measured by injection of air or methane at 270°C oven temperature. c For on-column injection, manufacturer's instructions should be followed regarding injection technique. d With on-column injection, initial temperature equals boiling point of the solvent; in this instance, hexane. e C 12 Cl 1 o elutes at 270°C. Programming above this temperature ensures a clean column and lower background on subsequent runs. f Fused silica columns may be routed directly into the ion source to prevent separator discrimination and losses. g High enough to elute all PCBs, but not high enough to degrade the column if routed through the transfer line. h Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCB congeners in solution CSxxx. i Peak width at 10% height for a single PCB congener is CSxxx. B-14 �TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHRQMATOGRAPHY SYSTEM Tolerance Recommended Parameter Gas chromatograph Finnigan 9610 Other3 Column 180 cm x 0.2 cm ID glass Other Column packing 3% SP-2250 on 100/ 120 mesh Supelcoport Other nonpolar or semipolar Carrier gas Helium Hydrogen Carrier gas flow rate 30 ml/min Optimum performance Injector On-column Other Injector temperature 250°C Optimum Injection volume 1.0 pi S 5 |Jl Initial column temperature 150°C, 4 min Other Column temperature program 150°-260°C 3t 8°/min Other Separator Glass jet Other Transfer line temperature 280°C Optimum Tailing factor 0.7-1.5 0.4-3 Peak width 10-20 sec < 30 sec o a Substitutions permitted if performance criteria are met. b High enough to elute all PCBs. c Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCB congeners in solution CSxxx. d Peak width at 10% height for a single PCB congener in CSxxx. B-15 �TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM Parameter Recommended Tolerance Mass spectrometer Finnigan 4023 Other3 Data system Incos 2400 Other Scan range 95-550 Other Scan time 1 sec Otherb Resolution Unit Optimum performance Ion source temperature 280°C 200°-300°C Electron energy 70 eV Optimum performance Trap current 0.2 mA Optimum performance Multiplier voltage -1,600 V Optimum performance Preamplifier sensitivity 10"6 A/V Set for desired working range a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. B-16 �TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM Parameter Tolerance Recommended Mass spectrometer Finnigan MAT 31 1A Other3 Data system Incos 2400 Other Scan range 98-550 Other Scan mode Exponential Other Cycle time 1.2 sec Otherb Resolution 1,000 > 500 Ion source temperature 280°C 250°-300°C Electron energy 70 eV 70 eV Emission current 1-2 mA Optimum Filament current Optimum Optimum Multiplier -1,600 V Optimum a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. B-17 �If specific congeners are known to be present and if standards are available, selected RF values may be employed. For general samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4), with a similar level of internal standard (de-3,31,4,4'-tetrachlorobiphenyl) added, may be used as the response factor solution. The PCB-surrogate pairs to be used in the RF calculation are listed in Table 9. Generally, only the primary ions of both the analyte and surrogate are used to determine the RF values. If alternate ions are to be used in the quantitation, the RF must be determined using that characteristic ion. The RF value must be determined in a manner to assure ±20% accuracy and precision. For instruments with good day-to-day precision, a running mean (RF) based on seven values determined once each day may be appropriate. Other options include, but are not limited to, triplicate determinations of a single concentration spaced throughout a day or determination of the RF at three different levels to establish a working curve. If replicate RF values differ by greater than ±10% RSD, the system performance should be monitored closely. If the RSD is greater than ±20%, the data set must be considered invalid and the RF redetermined before further analyses are done. 7.4 7.5 8.0 If the GC/EIMS system has not been demonstrated to yield a linear response or if the analyte concentrations are more than two orders of magnitude different from those in the RF solution, a calibration curve must be prepared. If the analyte and RF solution concentrations differ by more than one order of magnitude, a calibration curve should be prepared. A calibration curve should be established with triplicate determinations at three or more concentrations bracketing the analyte levels. The relative retention time (RRT) windows for the 10 homologs and surrogates must be determined. If all congeners are not available, a mixture of available congeners or an Aroclor mixture (e.g., 1016/1254/1260) may be used to estimate the windows. The windows must be set wider than observed if all isomers are not determined. Typical RRT windows for one column are listed in Table 10. The windows may differ substantially if other GC parameters are used. Sample Collection, Handling, and Preservation 8.1 Amber glass sample containers should have Teflon-lined screw caps. With noncorrosive samples, methylene chloride-washed aluminum foil liners may be substituted. The volume and configuration are determined by the amount of sample to be collected and its physical properties. For dry powders, other containers such as heavy-walled polyethylene bags may be appropriate. B-18 �TABLE 9. PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS Analyte Congener no . 1 2,3 Calibration standard Compound 2-C12H9Cl 3- and 4-C12H9Cl 1 C "ID 1 £ OQ f* TT f> "I ^ 1 2 8 2 P IT PI /•A — Cl **U ~ o 1 P IIP! L ^ 2 6 ^ •*- 4 ^ oZ" 1Z / OO 1O T L.^2"5^'-'-5 128-169 170-193 194-205 206-208 209 C12H4C16 C12H3C17 C12H2C18 C12HC19 Ci2CliQ 4 ID" oy *-*l2**7'-'^-3 P TT O T Congener no. I 3 7 30 50 97 143 183 202 207 209 Compound 2 4 2,4 2,4 ,6 2,2 ',4,6 2,2 ',3', 4, 5 2,2 ',3,4,5 ,6' 2,2 ',3', 4, ,5', 6 4' 2»2 ',3,3',5, 5', 6, 6' 2,2 ',3,3',4, 4', 5, 6, 6' C12Clio w a Ballschmiter numbering system, see Table 1. Surrogate Congener no. Compound 211 211 211 212 212 212 212 213 213 213 214 13 C6-4 C6-4 13 C6-4 13 Cl2-3,3' ,4 ,4' 13 C12-3,3' ,4 ,4' 13 Cl2-3,3' ,4 ,4' 13 C12-3,3' ,4 ,4' 13 Cl2-2,2' ,3 ,3' , 5,5', 6, 6' 13 C12-2,2' ,3 ,3' ,5,5' ,6, 6' 13 C12-2,2' ,3 ,3' ,5, 5', 6, 6' 13 C12C110 13 �TABLE 10. PCB homo log RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS VERSUS de-3,31.4,4'-TETRACHLOROBIPHENYL No. of isomers measured Observed range of RRTs3 Calibration solution Congener Observed no. RRT3 Projected range of RRTs 3 0.40-0.50 1 3 0.43 0.50 0.35-0.55 10 0.52-0.69 7 0.58 0.35-0.80 9 0.62-0.79 30 0.65 0.35-1.10 Tetrachloro 16 0.72-1.01 50 0.75 0.55-1.05 Pentachloro 12 0.82-1.08 97 0.98 0.80-1.10 Hexachloro 13 0.93-1.20 143 1.05 0.90-1.25 Heptachloro 4 1.09-1.30 183 1.15 1.05-1.35 Octachloro 6 1.19-1.36 202 1.19 1.10-1.50 Nonachloro 3 1.31-1.42 207 1.33 1.25-1.50 Decachloro 1 1.44-1.45 209 1.44 1.35-1.50 Monochloro Dichloro Trichloro a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using a 15-m J&W DB-5 fused silica column with a temperature program of 110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an oncolumn injector. A Finnigan 4023 Incos quadrupole mass spectrometer operating with a scan range of 95-550 daltons was used to detect each PCB congener. b The projected relative retention windows account for overlap of eluting homologs and take into consideration differences in operating systems and lack of all possible 209 PCB congeners. B-20 �8.2 Sample bottle preparation 8.2.1 8.2.2 Sample bottles are heated to 400°C for 15 to 20 min or rinsed with pesticide grade acetone or hexane and allowed to air dry. 8.2.3 8.3 All sample containers and caps should be washed in detergent solution, rinsed with tap water, and then with distilled water. The bottles and caps are allowed to drain dry in a contaminant-free area. Then the caps are rinsed with pesticide grade hexane and allowed to air dry. The clean bottles are stored inverted or sealed until use. Sample collection 8.3.1 8.3.2 Discrete product units - If the product is small enough that one or more discrete units would be used as the analytical sample, a statistically random sampling approach is recommended. 8.3.3 Liquids or free-flowing solids - If possible, the source is mixed thoroughly before collecting the sample. If mixing is impractical, the sample should be collected from a representative area of the source. If the liquid is flowing through an enclosed system, sampling through a valve should be randomly timed. 8.3.4 8.4 The primary consideration in sample collection is that the sample collected be representative of the whole. Therefore, sampling plans or protocols for each individual producer's situation will have to be developed. The recommendations presented here describe general situations. The number of replicates and sampling frequency also must be planned prior to sampling. Solids - Larger bulk solids which must be subsampled to get a reasonably sized analytical sample must be treated on a case-by-case basis. A representative sample should be obtained by designing a sampling location selection scheme such that all parts of the whole have a finite, known probability of inclusion. Based on such a scheme, the PCS content of the sample can be used to extrapolate to the content of the whole. Sample preservation - Product samples should be stored as the bulk or packaged product inventory would be stored, or in a cool, dry, dark area. Intermediates, process samples, or other non-product specimens should be stored at 4°C. If there is a possibility of microbial degradation, addition of HgSC^ during collection to a pH < 2 is recommended. A test strip is used to monitor pH. Storage times in excess of 4 weeks are not recommended. B-21 �If residual chlorine is present in the sample, it should be quenched with sodium thiosulfate. EPA Methods 330.4 and 330.5 may be used to measure the residual chlorine.1 Field test kits are available for this purpose. 9.0 Sample Preparation Since a wide variety of matrices may be subjected to analysis by this method, the extraction/cleanup procedure cannot be specified. This section describes general guidelines for subsampling, addition of 13C surrogates, dilution, extraction, cleanup, extract concentration, and other sample preparation procedures. 9.1 Sample homogenization and subsampling - The sample is homogenized by shaking, blending, shredding, crushing, or other appropriate mechanical technique. A representative subsample of 100 g or other known mass is then taken. The sample size is dependent upon the anticipated PCB levels and difficulty of the subsequent extraction/ cleanup steps. Note: The precision of the mass determination at this step will be reflected in the overall method precision. Therefore, an analytical balance must be used to assure that the weight is accurate to ±1% or better. 9.2 Surrogate addition - An appropriate volume of surrogate solution SSxxx is pipetted into the sample. The final concentration of the surrogates must be in the working range of the calibration and well above the matrix background. The surrogates are thoroughly incorporated by further mechanical agitation. For nonviscous liquids, shaking for 30 sec should be sufficient. For viscous liquids or free-flowing solids, 10-min tumbling is recommended. In cases where inadequate incorporation may be expected, such as solids, overnight equilibration with agitation is recommended. Note: The volume measurement of the spiking solution is critical to the overall method precision. The analyst must exercise caution that the volume is known to ±J% or better. Where necessary, calibration of the pipet is recommended. 9.3 Sample preparation (extraction/cleanup) - After addition of the surrogates, the sample is further treated at the discretion of the analyst, provided that the GC/EIMS response of the four surrogates meets the criteria listed in Section 7.0. The literature pertaining to these techniques has been reviewed.2 Several possible techniques are presented below for guidance only. The applicability of any of these techniques to a specific sample matrix must be determined by the precision and accuracy of the 13C PCB surrogate recoveries, as discussed in Section 14.2. B-22 �9.3.1 Extraction1 9.3.1.1 Dilution - In some cases, where the PCB concentration is high, a simple volumetric dilution with an appropriate solvent may be sufficient sample preparation. 9.3.1.2 Direct injection - If sample viscosity permits, direct injection with no dilution is permissible. 9.3.1.3 Liquid-liquid extraction - If the matrix is aqueous (or another solvent in which PCBs are only slightly soluble), a liquid-liquid partition may be effective. The solvent, number of extractions, solvent-to-sample ratio, and other parameters are chosen at the analyst's discretion. 9.3.1.4 Sorbent column extraction - PCBs may be isolated from free-flowing liquids onto sorbent columns. The selection of sorbent (XAD, Porapak, carbonpolyurethane foam, etc.) will depend on the nature of the matrix. The available methods have been reviewed.2 9.3.1.5 Thermal desorption - If the matrix is nonvolatile, thermal desorption of the PCBs onto a sorbent column, filter, or cold trap may be an effective extraction/cleanup method. 9.3.2 Cleanup - Several tested cleanup techniques are described below. All but the base cleanup (9.3.2.8) were previously validated for PCBs in transformer fluids.3 Depending upon the complexity of the sample, one or more of the techniques may be required to fractionate the PCBs from interferences. For most cleanups a concentrated (1-5 ml) extract should be used. 9.3.2.1 Acid cleanup 9.3.2.1.1 Place 5 ml of concentrated sulfuric acid into a 40-ml narrow-mouth screwcap bottle. Add the sample extract. Seal the bottle with a Teflon-lined screw cap and shake for 1 min. 9.3.2.1.2 Allow the phases to separate, transfer the sample (upper phase) with three rinses of 1-2 ml solvent to a clean container and concentrate to an appropriate volume. B-23 �9.3.2.1.3 9.3.2.1.4 9.3.2.2 Analyze as described in Section 10.0. If the sample is highly contaminated, a second or third acid cleanup may be employed. Florisil column cleanup 9.3.2.2.1 9.3.2.2.2 Place a 20-g charge of Florisil, activated overnight at 130°C, into a Chromaflex column. Settle the Florisil by tapping the column. Add about 1 cm of anhydrous sodium sulfate to the top of the Florisil. Pre-elute the column with 70-80 ml of hexane. Just before the exposure of the sodium sulfate layer to air, stop the flow. Discard the eluate. 9.3.2.2.3 Add the sample extract to the column. 9.3.2.2.4 Carefully wash down the inner wall of the column with 5 ml of hexane. 9.3.2.2.5 Add 200 ml of 6% ethyl ether/hexane and set the flow to about 5 ml/min. 9.3.2.2.6 Collect 200 ml of eluate in a KudernaDanish flask. All the PCBs should be in this fraction. Concentrate to an appropriate volume. 9.3.2.2.7 9.3.2.3 Variations among batches of Florisil (PR grade or equivalent) may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all PCBs must be verified by the analyst. The weight of Florisil can then be adjusted accordingly. Analyze the sample as described in Section 10.0. Alumina column cleanup 9.3.2.3.1 B-24 Adjust the activity of the alumina (Fisher A450 or equivalent) by heating to 200°C for 2 to 4 hr. When cool, add 3% water (wt:wt) and mix until uniform. Store in a tightly sealed bottle. Allow the deactivated alumina to equilibrate at least 1/2 hr before use. Reactivate weekly. �9.3.2.3.2 Variations between batches of alumina may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all of the PCBs must be verified by the analyst. The weight of alumina can then be adjusted accordingly. 9.3.2.3.3 Place a 50-g charge of alumina into a Chromaflex column. Settle the alumina by tapping. Add about 1 cm of anhydrous sodium sulfate. Preelute the column with 70-80 ml of hexane. Just before exposure of the sodium sulfate layer to air, stop the flow. Discard the eluate. 9.3.2.3.4 Add the sample extract to the column. 9.3.2.3.5 Carefully wash down the inner wall of the column with 5 ml of hexane. 9.3.2.3.6 Add 295 ml of hexane to the column. 9.3.2.3.7 Discard the first 50 ml. 9.3.2.3.8 Collect 250 ml Kuderna-Danish PCBs should be Concentrate to of the hexane in a flask. All of the in this fraction. an appropriate volume. 9.3.2.3.9 Analyze the sample as described in Section 10.0. 9.3.2.4 Silica gel column cleanup 9.3.2.4.1 Activate silica gel (Davison Grade 950 or equivalent) at 135°C overnight. 9.3.2.4.2 Variations between batches of silica gel may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all of the PCBs must be verified by the analyst. The weight of silica gel can then be adjusted accordingly. B-25 �9.3.2.4.3 Place a 25-g charge of activated silica gel into a Chromaflex column. Settle the silica gel by tapping the column. Add about 1 cm of anhydrous sodium sulfate to the top of the silica gel. 9.3.2.4.4 Pre-elute the column with 70-80 ml of hexane. Discard the eluate. Just before exposing the sodium sulfate layer to air, stop the flow. 9.3.2.4.5 Add the sample extract to the column. 9.3.2.4.6 Wash down the inner wall of the column with 5 ml of hexane. 9.3.2.4.7 Elute the PCBs with 195 ml of 10% diethyl ether in hexane (v:v). 9.3.2.4.8 Collect 200 ml Kuderna-Danish PCBs should be Concentrate to of the eluate in a flask. All of the in this fraction. an appropriate volume. 9.3.2.4.9 Analyze the sample as described in Section 10.0. 9.3.2.5 Gel permeation cleanup 9.3.2.5.1 Set up and calibrate the gel permeation chromatograph with an SX-3 column according to the Autoprep instruction manual. Use 15% methylene chloride in cyclohexane (v:v) as the mobile phase. 9.3.2.5.2 Inject 5.0 ml of the sample extract into the instrument. Collect the fraction containing the PCBs (see Autoprep operator's manual) in a Kuderna-Danish flask equipped with a 10-ml ampul. 9.3.2.5.3 Concentrate the PCB fraction to an appropriate volume. 9.3.2.5.4 Analyze the sample as described in Section 10.0. B-26 �9.3.2.6 Acetonitrile partition 9.3.2.6.1 Place the sample extract into a 125-ml separately funnel with enough hexane to bring the final volume to 15 ml. Extract the sample four times by shaking vigorously for 1 min with 30-ml portions of hexane-saturated acetonitrile. 9.3.2.6.2 Combine and transfer the acetonitrile phases to a 1-liter separatory funnel and add 650 ml of distilled water and 40 ml of saturated sodium chloride solution. Mix thoroughly for about 30 sec. Extract with two 100-ml portions of hexane by vigorously shaking about 15 sec. 9.3.2.6.3 Combine the hexane extracts in a 1-liter separatory funnel and wash with two 100-ml portions of distilled water. Discard the water layer and pour the hexane layer through an 8-10 cm anhydrous sodium sulfate column into a 500-ml Kuderna-Danish flask equipped with a 10-ml ampul. Rinse the separatory funnel and column with three 10-ml portions of hexane. 9.3.2.6.4 Concentrate the extracts to an appropriate volume. 9.3.2.6.5 Analyze as described in Section 10.0. 9.3.2.7 Florisil slurry cleanup 9.3.2.7.1 Place the sample extract into a 20-ml narrow-mouth screw-cap container. Add 0.25 g of Florisil (PR grade or equivalent). Seal with a Teflon-lined screw cap and shake for 1 min. 9.3.2.7.2 Allow the Florisil to settle; then decant the treated solution into a second container with rinsing. Concentrate the sample to an appropriate volume. Analyze as described in Section 10.0. B-27 �9.3.2.8 Base cleanup4 9.3.2.8.1 Quantitatively transfer the concentrated extract to a 125-ral extraction flask with the aid of several small portions of solvent. 9.3.2.8.2 Evaporate the extract just to dryness with a gentle stream of dry filtered nitrogen, and add 25 ml of 2.5% alcoholic KOH. 9.3.2.8.3 Add a boiling chip, put a water condenser in place, and allow the solution to reflux on a hot plate for 45 min. 9.8.2.8.4 After cooling, transfer the solution to a 250-ml separatory funnel with 25 ml of distilled water. 9.3.2.8.5 Rinse the extraction flask with 25 ml of hexane and add it to the separatory funnel. 9.3.2.8.6 Stopper the separatory funnel and shake vigorously for at least 1 min. Allow the layers to separate, and transfer the lower aqueous phase to a second separatory funnel. 9.3.2.8.7 Extract the saponification solution with a second 25-ml portion of hexane. After the layers have separated, add the first hexane extract to the second separatory funnel and transfer the aqueous alcohol layer to the original separatory funnel. 9.3.2.8.8 Repeat the extraction with a third 25-ml portion of hexane. Discard the saponification solution, and combine the hexane extracts. 9.3.2.8.9 Concentrate the hexane layer to an appropriate volume, and analyze as described in Section 10.0. B-28 �10.0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination 10.1 Internal standard addition - An appropriate volume of the internal standard solution is pipetted into the sample. The final concentration of the internal standard must be in the working range of the calibration and well above the matrix background. The internal standard is thoroughly incorporated by mechanical agitation. Note: The volumetric measurement of the internal standard solution is critical to the overall method precision. The analyst must exercise caution that the volume is known to be ±1% or better. Where necessary, calibration of the pipet is recommended. 10.2 Tables 2, and 5 through 8 summarize the recommended operating conditions for analysis. Figure 1 presents an example of a chromatogram. 10.3 While the highest available chromatographic resolution is not a necessary objective of this protocol, good chromatographic performance is recommended. With the high resolution of CGC, the probability that the chromatographic peaks consist of single compounds is higher than with PGC. Thus, qualitative and quantitative data reduction should be more reliable. 10.4 After performance of the system has been certified for the day and all instrument conditions set according to Tables 2, and 5 through 8, inject an aliquot of the sample onto the GC column. If the response for any ion, including surrogates and internal standards, exceeds the working range of the system, dilute the sample and reanalyze. If the responses of surrogates, internal standards, or analytes are below the working range, recheck the system performance. If necessary, concentrate the sample and reanalyze . 10.5 Record all data on a digital storage device (magnetic disk, tape, etc.) for qualitative and quantitative data reduction as discussed below. 11.0 Qualitative Identification 11.1 Selected ion monitoring (SIM) or limited mass scan (LMS) data The identification of a compound as a given PCB homolog requires that two criteria be met: 11.1.1 (1) The peak must elute within the retention time window set for that homolog (Section 7.5); and (2) the ratio of two ions obtained by SIM (Table 11) or by LMS (Table 12) must match the natural ratio within ±20%. The analyst must search the higher mass windows, in particular M+70, to prevent misidentification of a PCB fragment ion cluster as the parent. B-29 �-a 4-1 x~* "nj C too 0 cn M QJ l-i QJ K C C/5 1101 100.0 0 TJ rH CO 1 rH 01 rH rH 1 rH rH K*~l rH r*. £*"! ^ C 0) O CO C C c !>•» QJ QJ OJ x; r^ C 0) C QJ XI P. XI P, XI -H rH P. •H Xl o XI O M >> C Ol rH rH xi . x;, pi p -H _fl -H _ /*-* •H -H O O X> XI /—^ O 0) O r-l V-l 4-1 rH CO 60 XI O O O rH H rC XI CJ O O O r-l 1-1 C C CO D 0 r-l 4-1 rH X o o cn S 2 — <r -<r 01 c oj p. H i rH ^r •H f>~> Xi r-iiG -> rH ^ rH > . Xi ro x; p. -H x: xi P* p. O •rt o o CO VJ S-i O rH 4-1 OJ (-I O 1 O rH ^O 4-1 CJ -H l-l H 1 ^D •* O> PL, | m <J P O rH X! O •H O | XI u o c o a x: CO - c CN CN co 117-1 OJ GI3 CN T T 1 I A V. ' 1 " I 1 rH ' £*-» C 0) XI o x ; p. ^-, XI C O rH >. >-l XI -H •H 0) 0 Xl XI 4-1 rC 0) CO rH O rH p>^ CO 0 - 0 rJ O o o O ZZ rH u i v£> v£> VD rH x: 1 -H XI 0 P 0 0 cfl )-i Ml 0 CO P, <-" C OJ 0) H XI 1 P. -H C P . 4-» XI 0 CO rH 4-1 m x; •• CO U CO X Ol SC 1 ^ r-1 cfl O CO D CO OJ Q QJ Q 0 a o ^ *> in v o. m Q) X: ^3•* ^r A pr] 1 ^O co •^ CO •s Ol •> ^ in »> --^ «> ^ ro m v£) *. n •- ^i" n ~^j" * r. CO »\ •CN r> Ol CO CO 01 *^ - l?.0f) oj ii™ A Ol i oi * Ol M78 l;ilc ^ Jl k ii -/20 , _ Off 027 ,'. r-l XI t-> O <T - O Ol <N 1 M 0 1 —1 XI O f-l Xi rH loViw P. m -H XI C QJ •H XI O )_! O 1 O rH 01 >^ 01 XI •-21 OJ *• XI - C rH C w O " C -* O M rH p, rH rJJp. , t 'l O I 112592. ^ O vo co vD ' ' Ki-.-t-j ' 12:111 ,!• >o-.e-:) J— ' v ~ •—~~ " _„ 1 2 J •.;.'.) Till,; Figure 1. Capillary gas chromatography/clectron impact ionization mass spectrometry (CGC/EIMS) chroniatogram or the calibration standard solution required for quantitation of PCBs by homolog. This chroniatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates, and the deuterated internal standard; The concentration of a J 1 components and the CGC/EIMS parameters are presented in Tables 3, 4, 5, and 7. �TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs Ion (relative intensity) Secondary Tertiary Homolog Primary Ci2H9Cl 188 (100) 190 (33) - CiaHgCls 222 (100) 224 (66) 226 (11) C12H7Cl3 256 (100) 258 (99) 260 (33) Ci2H6Cl4 292 (100) 290 (76) 294 (49) Ci2H5Cl5 326 (100) 328 (66) 324 (61) C12H4Cl6 360 (100) 362 (82) 364 (36) CiaHsCl? 394 (100) 396 (98) 398 (54) CisH^Clg 430 (100) 432 (66) 428 (87) Ci2HCl9 464 (100) 466 (76) 462 (76) Ci2Clib 498 (100) 500 (87) 496 (68) Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in Identification of Polychlorinated Biphenyls by Mass Spectrometry," J. Assoc. Offic. Anal. Chem.. 56(1), 188-199 (1973). B-31 �TABLE 12. LIMITED MASS SCANNING (LMS) RANGES FOR PCBs Compound Mass range (m/z) C.ACU 186-190 C12HgCl2 220-226 L 12nyC i-3 254-260 C ^2ngC J-3 288-294 C i2n5Cl5 322-328 Ci2H4Clg 356-364 C12H3C17 386-400 C12H2Clg 426-434 C12HC19 460-468 C12C110 494-504 C12D6C14 294-300 13 192-196 13 300-306 13 438-446 C612C6H9C1 C12H6C14 C12H2C18 13 506-516 C12C110 a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass Spectrometry Method for Identifying and Determining Polychlorinated Biphenyls," J. Chromatogr., 196, 109-119 (1980). B-32 �11.1.2 If one or the other of these criteria is not met, interferences may have affected the results, and a reanalysis using full scan EIMS conditions is recommended. 11.2 Full scan data 11.2.1 The peak must elute within the retention time windows set for that homolog (as described in Section 7.5). 11.2.2 The unknown spectrum must match that of an authentic PCB. The intensity of the three largest ions in the molecular cluster (two largest for monochlorobiphenyls) must match the natural ratio within ±20%. Fragment clusters with proper intensity ratios must also be present. 11.2.3 Alternatively, a spectral search may be used to automatically reduce the data. The criteria for acceptable identification include a high index of similarity. For the Incos 2300, a fit of 750 or greater must be obtained. 11.3 Disputes in interpretation - Where there is reasonable doubt as to the identity of a peak as a PCB, the analyst must either identify the peak as a PCB or proceed to a confirmational analysis (see Section 13.0). 12.0 Quantitative Data Reduction 12.1 Once a chromatographic peak has been identified as a PCB, the compound is quantitated based either on the integrated abundance of the SIM data or EICP for the primary characteristic ion in Tables 11 and 12. If interferences are observed for the primary ion, use the secondary and then tertiary ion for quantitation. If interferences in the parent cluster prevent quantitation, an ion from a fragment cluster (e.g., M-70) may be used. Whichever ion is used, the RF must be determined using that ion. The same criteria should be applied to the surrogate compounds (Table 13). 12.2 Using the appropriate analyte-internal standard pair and response factor (RF ) as determined in Section 7.3, calculate the concentration of^each peak using Equation 12-1. A M. V Concentration (yg/g) = ^ • ~ • ^ • ^ Eq. 12-1 is p e i where A = area of the characteristic ion for the analyte PCB ^ peak A. = area of the characteristic ion for the internal standard peak RF = response factor of a given PCB congener B-33 �TABLE 13. CHARACTERISTIC IONS FOR 13 C-LABELED PCS SURROGATES Primary Ion (relative intensity) Secondary 13 194 (100) 196 (33) 13 304 (100) 306 ( 9 4) 302 (78) 13 442 (100) 444 (65) 440 (89) 13 510 (100) 512 (87) 514 (50) Specific compound C612C6H9C1 C12H6C14 C12H2C18 C12C110 B-34 Tertiary �M. = mass of internal standard injected (micrograms) IS M = mass of sample extracted (grams) V. = volume injected (microliters) V = volume of sample extract (microliters) 12.3 If a peak appears to contain non-PCB interferences, which cannot he circumvented by a secondary or tertiary ion, either: 12.3.1 12.3.2 Perform additional chemical cleanup (Section 9) and then reanalyze the sample; or 12.3.3 12.4 Reanalyze the sample on a different column which separates the PCB and interf erents ; Quantitate the entire peak as PCB. Calculate the recovery of the four 13C surrogates using the appropriate surrogate-internal standard pair and response factor (RF. ) as determined in Section 7.4 using Equation 12-2. A M. Recovery ( ) = ^ • - r • ^ • 100 % |Eq. 12-2 is s s where A S = area of the characteristic ion for the surrogate peak A. = area of the characteristic ion for the internal standard 18 peak RF = response factor for the surrogate compound with respect to the internal standard (Equation 7-2) M. = mass of internal standard injected (nanograms) 3.S MS - mass of surrogate, assuming 100% recovery (nanograms) 12.5 Correct the concentration of each peak using Equation 12-3. This is the final reportable concentration. Corrected concentration (pg/g) = 12.6 . 100 Eq. 12-3 Sum all of the peaks for each homolog, and then sum those to yield the total PCB concentration in the sample. Report all numbers in pg/g. The reporting form in Table 14 may be used. If an alternate reporting format (e.g., concentration per peak) is desired, a different report form may be used. The uncorrected concentrations, percent recovery, and corrected recovery are to be reported. 12.7 Round off all numbers reported to two significant figures. B-35 �TABLE 14. ANALYSIS REPORT INCIDENTAL PCBs IN COMMERCIAL PRODUCTS OR PRODUCT WASTES Sample No. Sample Matrix Sample Source Notebook No. or File Location Volume Extracted Extraction/Cleanup Int. Std. Procedure Mass Added (pg) (Circle one) 298 4-Cl(d6) Surrogates Mass Added (pg) (Circle one) 300 Ratio 194 196 100/33 4-C1 304 306 100/49 8-C1 442 444 100/65 10-C1 510 512 100/87 B-36 Intensity 100/49 1-C1 (continued) Ratio Intensity % Recovery �TABLE 14 (continued) Qualitative Analyte 1° 2° I l° T 1 2° Ratio Theoretical 1-C1 188 190 100/33 2-C1 222 224 100/66 3-C1 256 258 100/99 4-C1 292 290 100/76 5-C1 326 328 100/66 6-C1 360 362 100/82 7-C1 394 396 100/98 8-C1 430 432 100/66 9-C1 464 466 100/76 10-C1 498 500 Quantitative Uncorr. Corr. Ion Cone. Cone. OK? Used RF (|Jg/g) (Hg/g) 100/87 Total M8/8 Uncorr. Reported by: Internal Audit: Name Name EPA Audit: Name Signature/Date Signature/Date Signature/Date Organization Organization Organization B-37 M8/8 Corr. �13.0 Confirmation If there is reason to question the qualitative identification (Section 11.0), the analyst may choose to confirm that a peak is not a PCB. Any technique may be chosen provided that it is validated as having equivalent or superior selectivity and sensitivity to GC/EIMS. Some candidate techniques include alternate GC columns (with EIMS detection), GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory must validate confirmation techniques to show equivalent or superior selectivity between PCBs and interferences and sensitivity (limit of quantitation, LOQ). If a peak is confirmed as being a non-PCB, it may be deleted from the calculation (Section 12). If a peak is confirmed as containing both PCB and non-PCB components, it must be quantitated according to Section 12.3. 14.0 Quality Control 14.1 Each laboratory that uses this method must operate a formal quality control (QC) program. The minimum requirements of this program consist of an initial demonstration of laboratory capability and the analysis of spiked samples as a continuing check on performance. The laboratory must maintain performance records to define the quality of data that are generated. After a date specified by the Agency, ongoing performance checks should be compared with established performance criteria to determine if the results of analyses are within accuracy and precision limits expected of the method. 14.2 The analysts must certify that the precision and accuracy of the analytical results are acceptable by: 14.2.1 14.2.2 14.3 The absolute precision of surrogate recovery, measured as the RSD of the integrated EIMS area (A ) for a set s • of samples^ must be ±10%. The mean recovery (R ) of at least four replicates of a QC check sample to be supplied by the Agency must meet Agency-specified accuracy and precision criteria. This forms the initial data base for establishing control limits (see Section 14.3 below). Control limits - The laboratory must establish control limits using the following equations: Upper control limit (UCL) = RC + 3 RSDc Upper warning limit (UWL) = R + 2 RSD Lower warning limit (LWL) = R - 2 RSD Lower control limit (LCL) = R - 3 RSD B-38 �These may be plotted on control charts. If an analysis of a check sample falls outside the warning limits, the analyst should be alerted that potential problems may need correction. If the results for a check sample fall outside the control limits, the laboratory must take corrective action and recertify the performance (Section 14.2) before proceeding with analyses. The warning and control limits should be continuously updated as more check sample replicates are added to the data base. 14.4 Before processing any samples, the analyst should demonstrate through the analysis of a reagent blank that all glassware and reagent interferences are under control. Each time a set of samples is analyzed or there is a change in reagents, a laboratory reagent blank should be processed as a safeguard against contamination. 14.5 Procedural QC - The various steps of the analytical procedure should have quality control measures. These include but are not limited to: 14.5.1 GC performance - See Section 7.1 for performance criteria. 14.5.2 MS performance - See Section 7.2 for performance criteria. 14.5.3 Qualitative identification - At least 10% of the PCB identifications, as well as any questionable results, should be confirmed by a second mass spectrometrist. 14.5.4 Quantitation - At least 10% of all manual calculations, including peak area calculations, must be checked. After changes in computer quantitation routines, the results should be manually checked. 14.6 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, selected at random, must be run in triplicate to monitor the precision of the analysis. An RSD of ±30% or less must be achieved. If the precision is greater than ±30%, the analyst must be recertified (see Section 14.2). 14.7 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, selected at random, must be analyzed by the standard addition technique. Two aliquots of the sample are analyzed, one "as is" and one spiked (surrogate spiking and equilibration techniques are described in Section 9.2) with a sufficient amount of Solution CSxxx to yield approximately 100 |Jg/g of each compound. The samples are analyzed together and the quantitative results calculated. The recovery of the spiked compounds (calculated by difference) must be 80-120%. If the sample is known to contain specific PCB isomers, these isomers may be substituted for solution CSxxx. If the concentrations of PCBs are known to be high or low, the amount added should be adjusted so that the spiking level is 1.5 to 4 times the measured PCB level in the unspiked sample. B-39 �14.8 Interlaboratory comparison - Interlaboratory comparison studies are planned. Participation requirements, level of performance, and the identity of the coordinating laboratory will be presented in later revisions. 14.9 It is recommended that the participating laboratory adopt additional QC practices for use with this method. The specific practices that are most productive depend upon the needs of the laboratory and the nature of the samples. Field duplicates or triplicates may be analyzed to monitor the precision of the sampling technique. Whenever possible, the laboratory should perform analysis of standard reference materials and participate in relevant performance evaluation studies. 15.0 Quality Assurance Each participating laboratory must develop a quality assurance plan according to EPA guidelines.5 The quality assurance plan must be submitted to the Agency for approval. 16.0 Method Performance The method performance is being evaluated. Limits of quantitation; average intralaboratory recoveries, precision, and accuracy; and interlaboratory recoveries, precision, and accuracy will be presented. 17.0 Documentation and Records Each laboratory is responsible for maintaining full records of the analysis. Laboratory notebooks should be used for handwritten records. GC/MS data must be archived on magnetic tape, disk, or a similar device. Hard copy printouts may be kept in addition if desired. QC records should be maintained separately from sample analysis records. The documentation must describe completely how the analysis was performed. Any variances from the protocol must be noted and fully described. Where the protocol lists options (e.g., sample cleanup), the option used and specifics (solvent volumes, digestion times, etc.) must be stated. B-40 �REFERENCES 1. "Methods 330.4 (Titrimetric, DPD-FAS) and 330.5 (Spectrophotometric, DPD) for Chlorine, Total Residual," Methods for Chemical Analysis of Water and Wastes, U.S. Environmental Protection Agency, Environmental Monitoring and Support Laboratory, Cincinnati, Ohio, March 1979, EPA 600-4/79-020. 2. Erickson, M. D., and J. S. Stanley, "Methods of Analysis for Incidentally Generated PCBs—Literature Review and Preliminary Recommendations," Interim Report No. 1, EPA Contract No. 68-01-5915, Task 51, 1982. 3. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environmental Protection Agency, (1981) EPA-600/4-81-045. 4. American Society for Testing and Materials, "Standard Method for Analysis of Environmental Materials for Polychlorinated Biphenyls," pp. 877-885 in Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania (1980). ANSI/ASTM D 3304 - 77. 5. "Quality Assurance Program Plan for the Office of Toxic Substances," Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C., October 1980. B-41 �APPENDIX C ANALYTICAL METHOD: THE ANALYSIS OF BY-PRODUCTS CHLORINATED BIPHENYLS IN AIR C-l �THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN AIR 1.0 Scope and Application 1.1 This is a gas chromatographic/electron impact mass spectrometric (GC/EIMS) method applicable to the determination of chlorinated biphenyls (PCBs) in air emitted from commercial production through stacks, as fugitive emissions, or static (room, other containers, or outside) air. The PCBs present may originate either as synthetic by-products or as contaminants derived from commercial PCB products (e.g., Aroclors). The PCBs may be present as single isomers or complex mixtures and may include all 209 congeners from monochlorobiphenyl through decachlorobiphenyl listed in Table 1. 1.2 The detection and quantitation limits are dependent upon the volume of sample collected, the complexity of the sample matrix and the ability of the analyst to remove interferents and properly maintain the analytical system. The method accuracy and precision will be determined in future studies. 1.3 This method is restricted to use by or under the supervision of analysts experienced in the use of gas chromatography/mass spectrometry (GC/MS) and in the interpretation of gas chromatograms and mass spectra. Prior to sample analysis, each analyst must demonstrate the ability to generate acceptable results with this method by following the procedures described in Section 14.2. 1.4 The validity of the results depends on equivalent recovery of the analyte and 13C PCBs. If the *3C PCBs are not thoroughly incorporated in the matrix, the method is not applicable. 1.5 During the development and testing of this method, certain analytical parameters and equipment designs were found to affect the validity of the analytical results. Proper use of the method requires that such parameters or designs must be used as specified. These items are identified in the text by the word "must." Anyone wishing to deviate from the method in areas so identified must demonstrate that the deviation does not affect the validity of the data. Alternative test procedure approval must be obtained from the Agency. An experienced analyst may make modifications to parameters or equipment identified by the term "recommended." Each time such modifications are made to the method, the analyst must repeat the procedure in Section 14.2. In this case, formal approval is not required, but the documented data from Sectin 14.2 must be on file as part of the overall quality assurance program. C-2 �No. scnjcturt NO* NMoenlorooloAfWlf 1 2 1eh1oreb1pntnyli 4 5 6 7 8 9 10 11 12 13 14 15 .2' Is,'5 ,6 .3' ,4 3.4' 3.5 4,4' THc)i1erab1plnny1> 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 2,2' .3 2,2',4 2,2', 5 2.2'. 6 2,3,3' 2,3.4 2.3.4' 2.3,5 2,3,6 2. 3', 4 2,3', 5 2, 3', 6 2,4,4' 2.4.5 2.4,6 2.*' .5 2.4', 6 2'. 3,4 2'. 3.5 3.3'. 4 3,3'.5 3.4,4' 3.4.5 3.4', 5 SI 57 SI 59 10 61 62 13 M IS M 67 68 69 70 71 72 73 74 75 71 77 78 79 80 81 2,3' 2.3* 2.3' 2.3' 2,3' 2.3' 2.3' 2.3' 4,4' 4,5 4.5' 4.6 4', 5 4', 6 1 5.S 5',6 2.4.4-.S 2.4.4'.$ 2'. 3 4.5 3,3' 3,3* 3.3' 3.3' 82 83 84 as 2.3,3'. 4,4' 2,3,3', 4,5 2,3,3'. 4'. 5 2.3.3'. 4. 5' 2.3,3'. 4, 6 2.3,3', 4' .6 2.3.3' .5.5' 2.3.3'. 5.6 2.3.3'. 5'. 6 2.3,4, 4'. 5 2.3,4,4'. 6 2.3.4.5.6 2.3.4', 5,6 2.31. 4, 4' .5 2.3'. 4,4'. 6 2,3' ,4,5, 5' 2.3' ,4.5', 6 2'.3,3'.4,5 2 1 .3.4.4' ,5 2'. 3. 4.5. 5' 2'.3,4. 5.6' 3.3',4,4',5 3.3'. 4, 5,5' Htuehloreblphtnyla 4,4' 4,5 4.5' 5.5' 3.4.4' .5 2.2'. 3, 3'. 4 2.2'.3.3',5 2.2'. 3.3'. 6 2.2', 3,4,4' 2.2',3.4,5 2. 2' ,3.4.5' 2,2',3,4.6 2,2'. 3.4,6' 2,2',3,4',5 2,2'. 3, 4'. 6 2.2' ,3.5,5' 2,2'. 3, 5.6 2,2'. 3,5.6' 2.2'. 3. 5'. 8 2.2',3.6.6' 2.2'. 3'. 4. 5 2.2'.3'.4.8 2.2 1 .4.4'.S 2.2',4,4',6 2.2', 4,5.5' 2.2' .4.5.6' 2,2'. 4,5', 6 2.2'.4,«,6' NO. P«nt«cn lorottlphlnyll 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 2.2'.S.S' 2,2'. 5. 6' 2.2'.6.6' 2,3.3'. 4 2.3,3'. 4' 2,3,3', 5 2.3.3'. 5' 2.3,3'. 6 2.3.4.4' 2.3.4.5 2,3.1.6 2.3. 4' .5 2.3.4'. 6 2.3,5,6 Ptntlctll orettl phtny 1 1 88 87 88 39 90 91 92 93 94 95 Tttnrt 1 orob 1 phtfiy 1 » 98 2.2;,3 ' 97 98 99 2i2'l3 100 2,2'. 3 ' 101 2,2'. 3 102 103 2 < 2 ''4 ' 104 2> '.4 NUMBERING OF PCB CONGENERS3 MO. Structurt Tttnen 1 oreH etmiy 1 1 52 S3 54 B 2 3 TABLE 1. structure 128 129 130 137 132 133 134 135 131 137 138 139 140 141 142 143 144 14$ 141 147 148 149 150 151 152 153 154 155 151 157 1S8 159 160 2,2'.3.3'.4.4' 2.2'.3,3' .4.5 2.2',3.3',4.5' 2.2'.3.3'.4,« 2,2' .3.3' ,4.1' 2.2'.3,3'.5,5' 2,2'. 3,3'. 5,6 2.2* .3.3' ,5,1' 2,2' ,3,3'. 6,6' 2.2', 3,4, 4' .5 2,2'. 3, 4. 4 ' . £' 2,2' .3,4, 4 ' . 6 2.21, 3, 4, 4' .6' 2.2' .3, 4, 5, 5' 2,2'. 3.4. 5.6 2. 2' ,3, 4,5.6' 2,2'.3.4.5',6 2,2'.3.4, 6. 6' 2,2'. 3.4'. 5.5' 2.2*. 3,4- ,5,6 2,2' .3.4'. 5, 6' 2.2',3,4' .5', 6 2.2'. 3.4'. 6.6' 2,2',3.5,5',6 2.2', 3.5.6. 6' 2,2' ,4, 4 - , 5,5' 2,2', 4,4', 5.6' 2,2'. 4,4', 6.6' 2.3.3', 4.4- .5 2.3,3' .4, 4 ' . 5' 2.3,3', 4.4', 5 2.3,3' .4.5. 5' 2.3.3', 4.5,6 Il l 112 163 114 115 166 167 161 169 2.3,3' .4. 5'. 6 2.3. 3'. 4 ' . 5. S' 2,3.3' .4'. 5. 6 2,3. 3' .4'. 5'. 6 2.3.3'.5,5',6 2,3' .4,4'. 5,5' 2. 3'. 4. 4' ,5' .6 3.3',4.4'.5.5' Htptiehl orebf phtny 1 1 170 171 172 173 174 175 178 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 2,2'. 3. 3'. 4. 4 ' , 5 2.2' .3.3' ,4, 4 ' , 6 2, 2', 3. 3'. 4, 5. 5' 2, 2'. 3, 3'. 4, 5, 6 2, 2', 3, 3 ' . 4, 5, 6' 2,2'.3.3'.4,5',6 2,2', 3, 3 ' . 4' ,5. 6 2, 2', 3. 3', 5. 5 ' , 6 2. 2' .3, 3' ,5, 6, 6' 2,2',3,4,4',5,6 2, 2', 3, 4, 4 ' ,5,1' 2. 2' .3. 4, 4 ' . 5 ' . 5 2, 2', 3, 4, 4' ,6,6' 2,2', 3. 4,5, 5 ' . 6 2, 2', 3,4, 5, 6, 6' 2,2'.3,4',5,i l .S 2. 3, 3'. 4, 4 ' . 5,5' 2. 3, 3', 4. 4 ' . 5. 6 2.3, 3' .4. 4 ' , S ' , 5 2.3. 3'. 4,5. 5 ' . 1 2,3,3' .4'. 5,5'. 6 OetlcBlorottiphtnyli 194 195 191 197 198 199 200 201 202 203 204 205 2,2' ,3, 3' ,4.4', 5, 5' 2. 2 ' , 3. 3'. 4, 4 ' , 5, 6 2,2', 3. 3 ' , 4, 4 ' , S, 6' 2. 2', 3, 3', 4. 4 ' , 6, 6' 2. 2', 3. 3', 4, 5, 6, 6' 2,2'. 3,3' ,4. 5' .6, 6' 2.2'. 3. 3'. 4. 5, 5'. 6' 2.2' .3.3' ,5. 5 ' . 6 . 6 2.2'.3,4, 4'. 5, 5'. 6 2. 2' ,3.4. 4' .5, 6, 6' 2.3.3'. 4,4'. 5,5'. 6 Monichlorob<Pntnyli 206 207 208 2.2' .3, 3'. 4.4', 5.5', 6 2.2' .3,3* .4, 4' .5,6,1' 2.2'. 3, 3', 4, 5, 5'. 6, 6' OteieBlorottfohtnyl ' 209 •Adopt** fro* Stmcturt Htuchloroblphtnyls tollscMUr, X. ind Zcll, M., FrtMirius Z. Anal.CUM., 302. 20-31 (I960). C-3 2,2',3,3'4.4'.5,5'.6,5' �2.0 Summary 2.1 The air must be sampled such that the specimen collected for analysis is representative of the whole. Statistically designed selection of the sampling position (stack, flue, port, etc.) or time should be employed. Gaseous and particulate PCBs are withdrawn isokinetically from stacks, room air exhausts, process point exhausts, and other flowing gaseous streams using a sampling train.1 The PCBs are collected in the Florisil adsorbent tube and in the impingers in front of the adsorbent. PCBs are sampled from ambient air and other static gaseous sources onto a Florisil adsorbent tube. The sample must be preserved to prevent PCB loss prior to analysis. Storage at 4°C is recommended. 2.2 The Florisil adsorbent is extracted with hexane in a Soxhlet extractor, the aqueous condensate is extracted with hexane and the acetone/hexane impinger rinse is back-extracted with water. All three organic extracts are then combined. Optional cleanup techniques may include sulfuric acid cleanup and Florisil adsorption chromatography. The sample is concentrated to a final known volume for instrumental determination. 2.3 The PCB content of the sample extract is determined by capillary (preferred) or packed column gas chromatography/electron impact mass spectroraetry (CGC/EIMS or PGC/EIMS) operated in the selected ion monitoring (SIM), full scan, or limited mass scan (LMS) mode. 2.4 PCBs are identified by comparison of their retention time and mass spectral intensity ratios to those in calibration standards. 2.5 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners using the internal standard technique. 2.6 The PCBs identified by the SIM technique may be confirmed by full scan CGC/EIMS, retention on alternate GC columns, other mass spectrometric techniques, infrared spectrometry, or other techniques, provided that the sensitivity and selectivity of the technique are demonstrated to be comparable or superior to GC/EIMS. 2.7 The analysis time is dependent on the extent of workup employed. The time required for instrumental analysis of a single sample excluding data reduction and reporting, is about 30 to 45 min. 2.8 Appropriate quality control (QC) procedures are included to assess the performance of the analyst and estimate the quality of the results. These QC procedures include the demonstration of laboratory capability: periodic analyst certification, the use of control charts, and the analysis of blanks, replicates, and standard addition samples. A quality assurance (QA) plan must be developed for each laboratory. C-4 �2.9 While several options are available throughout this method, the recommended procedure for stack gases to be followed is: 2.9.1 2.9.2 The sample is preserved at 4°C to prevent any loss of PCBs or changes in matrix which may adversely affect recovery. 2.9.3 The three sample fractions are extracted and combined. 2.9.4 The extract is cleaned up and concentrated to an appropriate volume. Internal standards are added. 2.9.5 An aliquot of the extract is analyzed by CGC/EIMS operated in the SIM mode. On-column injections onto a 15-m DB-5 capillary column, programmed (for toluene solutions) from 110° to 325°C at 10°/min after a 2 min hold is used. Helium at 45-cm/sec linear velocity is used as the carrier gas. 2.9.6 PCBs are identified by retention time and mass spectral intensities. 2.9.7 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners. 2.9.8 3.0 The sample is collected using a modified Method 5 train1 according to a scheme which permits extrapolation of the sample data to the source being assessed. The total PCBs are obtained by summing the amounts for each homolog found, and the concentration is reported as micrograms per cubic meter. Interferences 3.1 Method interferences may be caused by contaminants, in sample collection media, solvents, reagents, glassware, and other sample processing hardware, leading to discrete artifacts and/or elevated baselines in the total ion current profiles. All of these materials must be routinely demonstrated to be free from interferences by the analysis of laboratory reagent blanks as described in Section 14.4. 3.1.1 Glassware must be scrupulously cleaned. All glassware is cleaned as soon as possible after use by rinsing with the last solvent used. This should be followed by detergent washing with hot water and rinses with tap water and reagent water. The glassware should then be drained dry and heated in a muffle furnace at 400°C for 15 to 30 min. Some thermally stable materials, such as PCBs, may not be eliminated by this treatment. Solvent rinses with acetone and pesticide quality hexane may be substituted C-5 �for the muffle furnace heating. Volumetric ware should not be heated in a muffle furnace. After it is dry and cool, glassware should be sealed and stored in a clean environment to prevent any accumulation of dust or other contaminants. It is stored inverted or capped with aluminum foil. 3.1.2 3.2 4.0 The use of high purity reagents and solvents helps to minimize interference problems. Purification of solvents by distillation in all-glass systems may be required. All solvent lots must be checked for purity prior to use. Matrix interferences may be caused by contaminants that are coextracted from the sorbent material or impingers. The extent of matrix interferences will vary considerably from source to source, depending upon the nature and diversity of the sources of samples. Safety 4.1 The toxicity or carcinogenicity of each reagent used in this method has not been precisely defined; however, each chemical compound should be treated as a potential health hazard. From this viewpoint, exposure to these chemicals must be reduced to the lowest possible level by whatever means available. The laboratory is responsible for maintaining a current awareness file of OSHA regulations regarding the safe handling of the chemical specified in this method. A reference file of material data handling sheets should also be made available to all personnel involved in the chemical analysis. 4.2 Polychlorinated biphenyls have been tentatively classified as known or suspected human or mammalian carcinogens. Primary standards of these toxic compounds should be prepared in a hood. Personnel must wear protective equipment, including gloves and safety glasses. Congeners highly substituted at the meta and para positions and unsubstituted at the ortho positions are reported to be the most toxic. Extrme caution should be taken when handling these compounds neat or in concentrated solution. The class includes 3,3',4'4'-tetrachlorobiphenyl (both natural abundance and isotopically labeled). 4.3 5.0 Waste disposal must be in accordance with RCRA and applicable state rules. Apparatus and Materials All specifications are suggestions only. are included for illustration only. C-6 Catalog numbers and suppliers �5.1 Stack sampling train1 - See Figure 1; a series of four impingers with a solid adsorbent trap between the third and fourth impingers. The train may be constructed by adaptation from a Method 5 train.2 Descriptions of the train components are contained in the following subsections. 5.1.1 Probe nozzle - Stainless steel (316) with sharp, tapered leading edge. The angle of taper shall be £ 30° and the taper shall be on the outside to preserve a constant internal diameter. The probe nozzle shall be of the buttonhook or elbow design, unless otherwise specified by the Agency. The wall thickness of the nozzle shall be less than or equal to that of 20 gauge tubing, i.e., 0.165 cm (0.065 in.) and the distance from the tip of the nozzle to the first bend or point of disturbance shall be at least two times the outside nozzle tubing. Other configurations and construction material may be used with approval from the Agency. 5.1.2 Probe liner - Borosilicate or quartz glass equipped with a connecting fitting that is capable of forming a leakfree, vacuum tight connection without sealing greases; such as Kontes Glass Company "0" ring spherical ground ball joints (model K-671300) or University Research Glassware SVL teflon screw fittings. A stainless steel (316) or water-cooled probe may be used for sampling high temperature gases with approval from the Agency. A probe heating system may be used to prevent moisture condensation in the probe. 5.1.3 Pitot tube - Type S, or equivalent, attached to probe to allow constant monitoring of the stack gas velocity. The face openings of the pitot tube and the probe nozzle shall be adjacent and parallel to each other but not necessarily on the same plane, during sampling. The free space between the nozzle and pitot tube shall be at least 1.9 cm (0.75 in.). The free space shall be set based on a 1.3 cm (0.5 in.) ID nozzle, which is the largest size nozzle used. The pitot tube must also meet the criteria specified in Method 22 and be calibrated according to the procedure in the calibration section of that method. 5.1.4 Differential pressure gauge - Inclined manometer capable of measuring velocity head to within 10% of the minimum measured value. Below a differential pressure of 1.3 mm (0.05 in.) water gauge, micromanometers with sensitivities of 0.013 mm (0.0005 in.) should be used. However, micromanometers are not easily adaptable to field conditions and are not easy to use with pulsating flow. Thus, other methods or devices acceptable to the Agency may be used when conditions warrant. C-7 �Thermometer Florisil Tube Probe (r^. Reverse-Type' Pitot Tube Manometer Tight /—TN Pump Control Box Figure 1. PCB sampling train for stack gases. C-8 Check Valve �5.1.5 Impingers - Four impingers with connecting fittings able to form leak-free, vacuum tight seals without sealant greases when connected together as shown in Figure 1. The first and second impingers are of the GreenburgSmith design. The final two impingers are of the Greenburg-Smith design modified by replacing the tip with a 1.3 cm (1/2 in.) ID glass tube extending to 1.3 cm (1/2 in.) from the bottom of the flask. One or two additional modified Greenburg-Smith impingers may be added to the train between the third impinger and the Florisil tube to accommodate additional water collection when sampling high moisture gases. Throughout the preparation, operation, and sample recovery from the train, these additional impingers should be treated exactly like the third impinger. 5.1.6 5.1.7 Metering system - Vacuum gauge, leak-free pump, thermometers capable of measuring temperature to within ±3°C (y 5°F), dry gas meter with 2% accuracy at the required sampling rate, and related equipment, or equivalent, as required to maintain an isokinetic sampling rate and to determine sample volume. When the metering system is used in conjunction with a pitot tube, the system shall enable checks of isokinetic rates. 5.1.8 5.2 Solid adsorbent tube - Glass with connecting fittings able to form leak-free, vacuum tight seals without sealant greases (Figure 2). Exclusive of connectors, the tube has a 2.2 cm inner diameter, is at least 10 cm long, and has four deep indentations on the inlet end to aid in retaining the adsorbent. Ground glass caps (or equivalent) must be provided to seal the adsorbent-filled tube both prior to and following sampling. Barometer - Mercury, aneroid, or other barometers capable of measuring atmospheric pressure to within 2.5 mm Hg (0.1 in. Hg). In many cases, the barometric reading may be obtained from a nearby weather bureau station, in which case the station value shall be requested and an adjustment for elevation differences shall be applied at a rate of -2.5 mm Hg (0.1 in. Hg) per 30 mm (100 ft) elevation increase. Static air sampling train1 - The sampling train, see Figure 3, consists of a glass-lined probe, an adsorbent tube containing Florisil, and the appropriate valving and flow meter controls for isokinetic sampling as described in Section 5.1. The sampling apparatus in Figure 3 is the same as that in Figure 1 and Section 5.1, except that the Smith-Greenburg impingers and heated probe are not used. If condensation of significant quantities of moisture prior to the solid adsorbent is expected, Section 5.1 of the C-9 �} 28/12 10cm j28/12 Figure 2. Florisil adsorbent tube. C-10 �Probe (to sample from duct) •* Glass- lined Probe Florisil Glass Wool Check Valve Vacuum Line ! Manometer - Integrated | Flow Meter I Figure 3. Air Tight Pump PCB sampling train for static air. C-ll �method should be used. Since probes and adsorbent tubes are not cleaned up in the field, a sufficient number must be provided for sampling and allowance for breakage. 5.3 Sample recovery 5.3.1 5.3.2 Teflon FEP® wash bottle - Two, 500 ml, Nalgene No. 0023A59 or equivalent. 5.3.3 Sample storage containers - Glass bottles, 1 liter, with TFE®-lined screw caps. 5.3.4 Balance - Triple beam, Ohaus Model 7505 or equivalent. 5.3.5 Aluminum foil - Heavy duty. 5.3.6 5.4 Ground glass caps - To cap off adsorbent tube and the other sample exposed portions of the train. Metal can - To recover used silica gel. Analysis 5.4.1 Glass Soxhlet extractors - 40 mm ID complete with 45/50 S condenser, 24/40 S 250 ml round-bottom flask, heating mantle for 250 ml flask, and power transformer. 5.4.2 Teflon FEP wash bottle - Two, 500 ml, Nalgene No. 0023A59 or equivalent. 5.4.3 Separatery funnel - 1,000 ml with TFE® stopcock. 5.4.4 Kuderna-Danish concentrators - 500 ml. 5.4.5 Steam bath. 5.4.6 Separatory funnel - 50 ml with TFE® stopcock. 5.4.7 Volumetric flask - 25.0 ml, glass. 5.4.8 Volumetric flask - 5.0 ml, glass. 5.4.9 Culture tubes - 13 x 100 mm, glass with TFE®-lined screw caps. 5.4.10 Pipette - 5.0 ml glass. 5.4.11 Teflon®-glass syringe - 1 ml, Hamilton 1001 TLL or equivalent with Teflon® needle. 5.4.12 Syringe - 10 (Jl, Hamilton 701N or equivalent. C-12 �5.4.13 Disposable glass pipettes with bulbs - To aid transfer of the extracts. 5.4.14 Gas chromatography/mass spectrometer system. 5.4.14.1 Gas chromatograph - An analytical system complete with a temperature programmable gas chromatograph and all required accessories including syringes, analytical columns, and gases. The injection port must be designed for oncolumn injection when using capillary columns or packed columns. Other capillary injection techniques (split, splitless, "Grob," etc.) may be used provided the performance specifications stated in Section 7.1 are met. 5.4.14.2 Capillary GC column - A 12-20 m long x 0.25 mm ID fused silica column with a 0.25 pm thick DB-5 bonded silicone liquid phase (J&W Scientific) is recommended. Alternate liquid phases may include OV-101, SP-2100, Apiezon L, Dexsil 300, or other liquid phases which meet the performance specifications stated in Section 7.1. 5.4.14.3 Packed GC column - A 180 cm x 0.2 cm ID glass column packed with 3% SP-2250 on 100/120 mesh Supelcoport or equivalent is recommended. Other liquid phases which meet the performance specifications stated in Section 7.1 may be substituted. 5.4.14.4 Mass spectrometer - Must be capable of scanning from 150 to 550 daltons every 1.5 sec or less, collecting at least five spectra per chromatographic peak, utilizing a 70-eV (nominal) electron energy in the electron impact ionizaton mode and producing a mass spectrum which meets all the criteria in Table 2 when 50 ng of decafluorotriphenyl phosphine [DFTPP, bis(perfluorophenyDphenyl phosphine] is injected through the GC inlet. Any GC-to-MS interface that gives acceptable calibration points at 10 ng per injection for each PCB isomer in the calibration standard and achieves all acceptable performance criteria (Section 10) may be used. Direct coupling of the fused silica column to the MS is recommended. Alternatively, GC to MS interfaces constructed of all glass or glasslined materials are recommended. Glass can be deactivated by silanizing with dichlorodimethylsilane. C-13 �TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA Mass Ion abundance criteria 197 198 199 Less than 1% of mass 198 100% relative abundance 5-9% of mass 198 275 10-30% of mass 198 365 Greater than 1% of mass 198 441 442 443 Present, but less than mass 443 Greater than 40% of mass 198 17-23% of mass 442 C-14 �5.4.14.5 A computer system that allows the continuous acquisition and storage on machine-readable media of all mass spectra obtained throughout the duration of the chromatographic program must be interfaced to the mass spectrometer. The data system must have the capability of integrating the abundances of the selected ions between specified limits and relating integrated abundances to concentrations using the calibration procedures described in this method. The computer must have software that allows searching any GC/MS data file for ions of a specific mass and plotting such ion abundances versus time or scan number to yield an extracted ion current profile (EICP). Software must also be available that allows integrating the abundance in any EICP between specified time or scan number limits. 6.0 Reagents 6.1 Sampling 6.1.1 6.1.2 Glass wool - Cleaned by thorough rinsing with hexane, dried in a 110°C oven, and stored in a hexane-washed glass jar with TFE®-lined screw cap. 6.1.3 Water - Deionized, then glass-distilled, and stored in hexane-rinsed glass containers with TFES-lined screw caps. 6.1.4 Silica gel - Indicating type, 6-16 mesh. If previously used, dry at 175°C for 2 hr. New silica gel may be used as received. 6.1.5 6.2 Florisil - Floridin Company, 30/60 mesh, Grade A. The Florisil is cleaned by 8 hr Soxhlet extraction with hexane and then by drying for 8 hr in an oven at 110°C and is activated by heating to 650°C for 2 hr (not to exceed 3 hr) in a muffle furnace. After allowing to cool to near 110°C transfer the clean, active Florisil to a clean, hexane-washed glass jar and seal with a TFE®-lined lid. The Florisil should be stored at 110°C until taken to the field for use. Florisil that has been stored more than 1 month must be reactivated before use. Crushed ice. Solvents - All solvents must be pesticide residue analysis grade. New lots should be checked for purity by concentrating an aliquot by at least as much as is used in the procedure. C-15 �6.3 Calibration standard congeners - Standards of the PCB congeners listed in Table 3 are available from Ultra Scientific, Hope, Rhode Island; or Analabs, North Haven, Connecticut. 6.4 Calibration standard stock solutions - Primary dilutions of each of the individual PCBs listed in Table 3 are prepared by weighing approximately 1-10 mg of material within 1% precision. The PCB is then dissolved and diluted to 1.0 ml with hexane. The concentration is calculated in mg/ml. The primary dilutions are stored at 4°C in screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. Primary dilutions are stable indefinitely if the seals are maintained. The validity of primary and secondary dilutions must be monitored on a quarterly basis by analyzing four quality control check samples (see Section 14.2). 6.5 Working calibration standards - Working calibration standards are prepared that are similar in PCB composition and concentration to the samples by mixing and diluting the individual standard stock solutions. Example calibration solutions are shown in Table 3. The mixture is diluted to volume with pesticide residue analysis quality hexane. The concentration is calculated in ng/ml as the individual PCBs. Dilutions are stored at 4°C in narrow-mouth, screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. These secondary dilutions can be stored indefinitely if the seals are maintained. These solutions are designated "CSxxx," where the xxx is used to encode the nominal concentration in ng/ml. 6.6 Alternatively, certified stock solutions similar to those listed in Table 3 may be available from a supplier, in lieu of the procedures described in Section 6.4. 6.7 DFTPP standard - A 50 ng/(Jl solution of DFTPP is prepared in acetone or another appropriate solvent. 6.8 Internal standard stock solution - The four 13C-labeled PCBs listed in Table 4 may be available from a supplier as a certified solution. This solution may be used as received or diluted further. 6.9 Solution stability - The calibration standard, surrogate and DFTPP solutions should be checked frequently for stability. These solutions should be replaced after 6 months, or sooner if comparison with quality control check samples indicates compound degradation or concentration change. 6.10 Quality control check samples will be supplied by the Agency. C-16 �TABLE 3. CONCENTRATIONS OF CONGENERS IN PCB CALIBRATION STANDARDS (ng/ml)a Homo log Congener no. CS1000 CS100 CS050 CS010 1 1 1,040 104 52 10 1 3 1,000 100 50 10 2 7 1,040 104 52 10 3 30 1,040 104 52 10 4 50 1,520 152 76 15 5 97 1,740 174 87 17 6 143 1,920 192 96 19 7 183 2,600 260 130 26 8 202 4,640 464 232 46 9 207 5,060 506 253 51 10 209 4,240 424 212 42 4 210 (IS) 255 255 255 255 1 211 (RS) 104 104 104 104 4 212 (RS) 257 257 257 257 8 213 (RS) 407 407 407 407 10 214 (RS) 502 502 502 502 a Concentrations given as examples only. C-17 �TABLE 4. COMPOSITION OF INTERNAL STANDARD SPIKING SOLUTION (SS100) CONTAINING 13C-LABELED PCBs3 Congener no. Compound Concentration (pg/ml) 211 (l',2l,3',4',5',6'-13C6)4-chlorobiphenyl 104 212 (13C12)3,3',4,4'-tetrachlorobiphenyl 257 213 (13C12)2,2',3,3',5,5',6,6'-octachlorobiphenyl 395 214 (13C12)decachlorobiphenyl 502 a Concentrations given as examples only. C-16 �7.0 Calibration Maintain a laboratory log of all calibrations. 7.1 Sampling train 7.1.1 Probe nozzle - Using a micrometer, the inside diameter of the nozzle is measured to the nearest 0.025 mm (0.001 in.). Three separate measurements are made using different diameters each time and obtain the average of the measurements. The difference between the high and low numbers must not exceed 0.1 mm (0.004 in.). When nozzles become nicked, dented, or corroded, they must be reshaped, sharpened, and recalibrated before use. Each nozzle must be permanently and uniquely identified. 7.1.2 Pitot tube - The pitot tube must be calibrated according to the procedure outlined in Method 2.2 7.1.3 Dry gas meter and orifice meter - Both meters must be calibrated according to the procedure outlined in APTD0581.3 When diaphragm pumps with bypass valves are used, proper metering system design is checked by calibrating the dry gas meter at an additional flow rate of 0.0057 m3/min (0.2 cfm) with the bypass valve fully opened and then with it fully closed. If there is more than ±2% difference in flow rates when compared to the fully closed position of the bypass valve, the system is not designed properly and must be corrected. 7.1.4 Probe heater calibration - The probe heating system must be calibrated according to the procedure contained in APTD-0581.3 7.1.5 Temperature gauges - Dial and liquid filled bulb thermometers are calibrated against mercury-in-glass thermometers, Thermocouples should be calibrated in constant temperature baths. 7.2 The gas chromatograph must meet the minimum operating parameters shown in Tables 5 and 6, daily. If all of the criteria are not met, the analyst must adjust conditions and repeat the test until all criteria are met. 7.3 The mass spectrometer must meet the minimum operating parameters shown in Tables 2, 7, and 8, daily. If all criteria are not met, the analyst must retune the spectrometer and repeat the test until all conditions are met. C-19 �TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM Parameter Recommended Tolerance Liquid phase Finnigan 9610 15 m x 0.255 mm ID Fused silica DB-5 (J&W) Liquid phase thickness Carrier gas Carrier gas velocity Injector Injector temperature Injection volume Initial column temperature Column temperature program Separator Transfer line temperature 0.25 pm Helium 45 cm/sec r* On-column (J&W) c Optimum performance 1.0 (Jlc 70°C (2 min)d 70°-325°C at 10°C/mine None 280°C Other nonpolar or semipolar < 1 HID Hydrogen Optimum performance Other Optimum performance Other Other Other Glass jet or othe Optimum8 Tailing factorh 0.7-1.5 0.4-3 Peak width1 7-10 sec < 15 sec Gas chromatograph Column a Other Other Substitutions permitted with any common apparatus or technique provided performance criteria are met. b Measured by injection of air or methane at 270°C oven temperature. c For on-column injection, manufacturer's instructions should be followed regarding injection technique. d With on-column injection, initial temperature equals boiling point of the solvent; in this instance, hexane. e C^Clio elutes at 270°C. Programming above this temperature ensures a clean column and lower background on subsequent runs. f Fused silica columns may be routed directly into the ion source to prevent separator discrimination and losses. g High enough to elute all PCBs, but not high enough to degrade the column if routed through the transfer line. h Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCB congeners in solution CSxxx. i Peak width at 10% height for a single PCB congener is CSxxx. C-20 �TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM Gas chromatograph Column Tolerance Recommended Parameter Finnigan 9610 Other3 180 cm x 0.2 cm ID Other glass Column packing 3% SP-2250 on 100/ 120 mesh Supelcoport Other nonpolar or semipolar Carrier gas Helium Hydrogen Carrier gas flow rate 30 ml/min Optimum performance Injector On-column Other Injector temperature 250°C Optimum Injection volume 1.0 pi ^ 5 Ml Initial column temperature 150°C, 4 min Other Column temperature program 150°-260°C at 8°/min Other Separator Glass jet Other Transfer line temperature 280°C Optimum8 Tailing factor0 ,0.7-1.5 0.4-3 10-20 sec < 30 sec Peak widthd a Substitutions permitted if performance criteria are met. b High enough to elute all PCBs. c Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCB congeners in solution CSxxx. d Peak width at 10% height for a single PCB congener is CSxxx. C-21 �TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER Parameter Recommended SYSTEM Tolerance Mass spectrometer Finnigan 4023 Other3 Data system Incos 2400 Other Scan range 95-550 Other Scan time 1 sec Otherb Resolution Unit Optimum performance Ion source temperature 280°C 200°-300°C Electron energy 70 eV Optimum performance Trap current 0.2 mA Optimum performance Multiplier voltage -1,600 V Optimum performance Preamplifier sensitivity 10"6 A/V Set for desired working range a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. C-22 �TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM Parameter Recommended Tolerance Mass spectrometer Finnigan MAT 311A Other8 Data system Incos 2400 Other Scan range 98-550 Other Scan mode Exponential Other Cycle time 1.2 sec Other Resolution 1,000 > 500 Ion source temperature 280°C 250-300° Electron energy0 70 eV 70 eV Emission current 1-2 mA Optimum Filament current Optimum Optimum Multiplier -1,600 V Optimum a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. C-23 �7.4 The PCS response factors (RF ) must be determined using Equation 7-1 for the analyte homologs? A x M. RFp = A. x j -£ ^ Eq. 7-1 H r is Mp where RF = response factor of a given PCB isomer A = area of the characteristic ion for the PCB congener " peak M = mass of PCB congener injected (nanograms) A. = area of the characteristic ion for the internal standard peak M. = mass of internal standard injected (nanograras) 1S If specific congeners are known to be present and if standards are available, selected RF values may be employed. For general samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4) may be used as the response factor solution. The PCB-surrogate pairs to be used in the RF calculation are listed in Table 9. Generally, only the primary ions of both the analyte and surrogate are used to determine the RF values. If alternate ions are to be used in the quantitation, the RF must be determined using that characteristic ion. The RF value must be determined in a manner to assure ±20% accuracy and precision. For instruments with good day-to-day precision, a running mean (RF) based on seven values determined once each day may be appropriate. Other options include, but are not limited to, triplicate determinations of a single concentration spaced throughout a day or determination of the RF at three different levels to establish a working curve. If replicate RF values differ by greater than ±10% RSD, the system performance should be monitored closely. If the RSD is greater than ±20%, the data set must be c6nsidered invalid and the RF redetermined before further analyses are done. 7.5 If the GC/EIMS system has not been demonstrated to yield a linear response or if the analyte concentrations are more than one order of magnitude different from those in the RF solution, a calibration curve must be prepared. If the analyte and RF solution concentrations differ by more than one order of magnitude, a calibration curve should be prepared. A calibration curve should be established with triplicate determinations at three or more concentrations bracketing the analyte levels. C-24 �TABLE 9. PAIRINGS OF ANALYTE. CALIBRATION, AND SURROGATE COMPOUNDS Analyte Congener3 no. Compound 1 2,3 4-15 16-39 40-81 82-127 128-169 170-193 194-205 206-208 o 209 I 2-Ci2HgCl 3- and 4-C12H9Cl Ci2HgCl2 C^HyCls C12H6C14 c 12H5Cls C12HsCl7 Ci2H2Clg Cj^HClg CiaCl10 Calibration standard Congener Compound no. 1 3 7 30 50 97 143 183 202 207 209 2 4 2,4 2,4 ,6 2,2 ',4 ,6 2,2 ',3 ',4,5 2,2 ',3 ,4,5 ,6' 2,2 ',3 ',4,4' ,5', 6 2,2 ',3 ,3',5,5', 6, 6' 3' 2,2 ',3 ,° » 4, 4' ,5,6, 6' C 12Cli 0 NJ a Ballschmiter numbering system, see Table 1. Surrogate Congener no. Compound 211 211 211 212 212 212 212 213 213 213 214 13 C6-4 c "3 e-4 i C6-4 13 C12-3,3' ,4,4' 13 Ci2-3,3' ,4,4' 13 C12-3,3' ,4,4' 13 Ci2-3,3' ,4,4' 13 C12-2,2' ,3,3' ,5,5', 6, 6' 13 q 01 C12-2,2' ,O,O ,5, 5', 6,6' 13 q 01 C12-2,2' ,J,J ,5,5', 6, 6' 13 Ci2Cll0 �7.6 8.0 The relative retention time (RRT) windows for the 10 homologs and surrogates must be determined. If all congeners are not available, a mixture of available congeners or an Aroclor mixture (e.g., 1016/1254/1260) may be used to estimate the windows. The windows must be set wider than observed if all isomers are not determined. Typical RRT windows for one column are listed in Table 10. The windows may differ substantially if other GC parameters are used. Sample Collection, Handling, and Preservation The sampling shall be conducted by competent personnel experienced with this test procedure and cognizant of the constraints of the anaytical techniques for PCBs, particularly contamination problems. 8.1 Stack sampling1 8.1.1 Pretest preparation - All train components shall be maintained and calibrated according to the procedure described in APTD-0581,3 unless otherwise specified herein. This should be done in the laboratory prior to sampling. 8.1.1.1 Cleaning glassware - All glass parts of the train upstream of and including the adsorbent tube and impingers, should be cleaned as described in Section 3.1.1. Special care should be devoted to the removal of residual silicone grease sealants on ground glass connections of used glassware. These grease residues should be removed by soaking several hours in a chromic acid cleaning solution prior to routine cleaning as described above. 8.1.1.2 Solid adsorbent tube - 7.5 g of Florisil activated within the last 30 days and still warm from storage in a 110°C oven, is weighed into the adsorbent tube (prerinsed with hexane) with a glass wool plug in the downstream end. A second glass wool plug is placed in the tube to hold the sorbent in the tube. Both ends of the tube are capped with ground glass caps. These caps should not be removed until the tube is fitted to the train immediately prior to sampling. 8.1.2 Preliminary determinations - The sampling site and the minimum number of sampling points are selected according to Method I2 or as specified by the Agency. The stack pressure, temperature, and the range of velocity heads are determined using Method 22 and moisture content using Approximation Method 42 or its alternatives for the purpose of making isokinetic sampling rate calculations. Estimates may be used. However, final results must be based on actual measurements made during the test. C-26 �TABLE 10. RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS VERSUS d6-3,3'.4.4'-TETRACHLOROBIPHENYL PCB homolog Monochloro No. of isomers measured Observed range of RRTsa Congener no. Observed RRTa Projected range of RRTs 3 0.40-0.50 1 3 0.43 0.50 0.35-0.55 Dichloro 10 0.52-0.69 7 0.58 0.35-0.80 Trichloro 9 0.62-0.79 30 0.65 0.35-0.10 Tetrachloro 16 0.72-1.01 50 0.75 0.55-1.05 Pentachloro 12 0.82-1.08 97 0.98 0.80-1.10 Hexachloro 13 0.93-1.20 143 1.05 0.90-1.25 Heptachloro 4 1.09-1.30 183 1.15 1.05-1.35 Octachloro 6 1.19-1.36 202 1.19 1.10-1.50 Nonachloro 3 1.31-1.42 207 1.33 1.25-1.50 Decachloro 1 1.44-1.45 209 1.44 1.35-1.50 a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were measured versus 3,3',4,4'-tetrachlorobiphenyl-de (internal standard) using a 15-m J&W DB-5 fused silica column with a temperature program of 110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an oncolumn injector. A Finnigan 4023 Incos quadrupole mass spectrometer operating with a scan range of 95-550 daltons was used to detect each PCB congener. b The projected relative retention windows account for overlap of eluting homologs and take into consideration differences in operating systems and lack of all possible 209 PCB congeners. C-27 �The molecular weight of the stack gases is determined using Method 3.2 A nozzle size is selected based on the maximum velocity head so that isokinetic sampling can be maintained at a rate less than 0.75 cfm. It is not necessary to change the nozzle size in order to maintain isokinetic sampling rates. During the run, the nozzle size must not be changed. A suitable probe length is selected such that all traverse points can be sampled. Sampling from opposite sides for large stacks may be considered to reduce the length of probes. A sampling time is selected appropriate for total method sensitivity and the PCB concentration anticipated. Sampling times should generally fall within a range of 2 to 4 hr. A buzzer-timer should be incorporated in the control box (see Figure 1) to alarm the operator to move the probe to the next sampling point. 8.1.3 Preparation of collection train - During preparation and assembly of the sampling train, all train openings must be covered until just prior to assembly or until sampling is about to begin. Immediately prior to assembly, all parts of the train upstream of the adsorbent tube are rinsed with hexane. The probe is marked with heat resistant tape or by some other method at points indicating the proper distance into the stack or duct for each sampling point. 200 ml of water is placed in each of the first two impingers, and the third impinger left empty. CAUTION: Sealant greases must not be used in assembling the train. If the preliminary moisture determination shows that the stack gases are saturated or supersaturated, one or two additional empty impingers should be added to the train between the third impinger and the Florisil tube. See Section 5.1.5. Approximately 200 to'300 g or more, if necessary, of silica gel is placed in the last impinger. Each impinger (stem included) is weighed and the weights recorded to the nearest 0.1 g on the impingers and on the data sheet. Unless otherwise specified by the Agency, a temperature probe is attached to the metal sheath of the sampling probe so that the sensor is at least 2.5 cm behind the nozzle and pitot tube and does not touch any metal. C-28 �The train is assembled as shown in Figure 1. Through all parts of this method use of sealant greases such as stopcock grease to seal ground glass joints must be avoided. Crushed ice is placed around the impingers. 8.1.4 Leak check procedure - After the sampling train has been assembled, the probe heating system(s) is turned on and set (if applicable) to reach a temperature sufficient to avoid condensation in the probe. Time is allowed for the temperature to stabilize. The train is leak checked at the sampling site by plugging the nozzle and pulling a 380 mm Hg (15 in. Hg) vacuum. A leakage rate in excess of 4% of the average sampling rate or 0.0057 m3/min (0.02 cfm) whichever is less, is unacceptable. The following leak check instruction for the sampling train described in APTD-05813 may be helpful. The pump is started with bypass valve fully open and coarse adjust valve completely closed. The coarse adjust valve is partially opened and the bypass valve slowly closed until 380 mm Hg (15 in. Hg) vacuum is reached. The direction of bypass valve must not be reversed. This will cause water to back up into the probe. If 380 mm Hg (15 in. Hg) is exceeded, either the leak check is conducted at this higher vacuum or the leak check is ended as described below and start over. When the leak check is completed, the plug is first slowly removed from the inlet to the probe and the vacuum pump is immediately turned off. This prevents the water in the impingers from being forced backward into the probe. Leak checks, shall be conducted as described above prior to each test run and at the completion of each test run. If leaks are found to be in excess of the acceptable rate, the test will be considered invalid. To reduce lost time due to leakage occurrences, it is recommended that leak checks be conducted between port changes. 8.1.5 Train operation - During the sampling run, an isokinetic sampling rate within 10%, or as specified by the Agency, of true isokinetic shall be maintained. During the run, the nozzle or any other part of the train in front of and including the Florisil tube must not be changed. For each run, the data required on the data sheets must be recorded. An example is shown in Figure 4. The dry gas meter readings are recorded at the beginning and end of each sampling time increment, when changes in flow rates are made, and when sampling is halted. Other data point readings are taken at least once at each sample point during each time increment and whenever significant C-29 �FIELD DATA PLANT. OATE_ SAMPLING LOCATION. SAMPLE TYPE RUN NUMBER OPERATOR PROBE LENGTH AND TYPE. NOZZLE ID. ASSUMED MOISTURE. "„ SAMPLE BOX NUMBER METER BOX NUMBER METER AH p C FACTOR PROBE HEATER SETTING HEATER BOX SETTING REFERENCE Ap_ AMBIENT TEMPERATURE BAROMETRIC PRESSURE . STATIC PRESSURE. (P$)_ FILTER NUMBER ($) SCHEMATIC OF TRAVERSE POINT LAYOUT READ AND RECORD ALL DATA EVERY, MINUTES TRAVERSE POINT NUMBER s ^X CLOCK hTIME LiNG r_ \Aoc K, TIMt.iim N^ ~~ GAS METER READING <Vml. It3 VELOCITY HEAD (APSI. in. H?0 —— __ ORIFICE PRESSURE DIFFERENTIAL (AHI. in. H20l DESIRED STACK TEMPERATURE |TSI.°F ACTUAL n i COMMENTS: Figure 4. EPAlDur) 2K Field data sheet, DRY GAS METER TEMPERATURE INLET (Tm mt."F OUTLET •Tm^.-'F PUMP VACUUM, in. H| SAMPLE BOX TEMPERATURE. °F IMPINGCR TEMPERATURL "F �changes (20% variation in velocity head readings) necessitate additional adjustments in flow rate. The portholes are cleaned prior to the test run to minimize change of sampling deposited material. To begin sampling, the nozzle cap is removed, the probe heater operational and temperature up, and the pitot tube and probe positions are verified (if applicable). The nozzle is positioned at the first traverse point with the tip pointing directly into the gas stream. The pump is started and the flow adjusted to isokinetic conditions. Nomographs are available for sampling trains using type S pitot tubes with 0.85 ± 0.02 coefficients (C ), and when sampling in air or a stack gas with equivalent density (molecular weight, M,, equal to 29 ± 4), which aid in the rapid adjustment of the isokinetic sampling rate without excessive computations. If C and M, are outside the above stated ranges, the nomograph cannot be used unless appropriate steps are taken to compensate for the deviations. When the stack is under significant negative pressure (height of impinger stem), the coarse adjust valve must be closed before inserting the probe into the stack to avoid water backing into the probe. If necessary, the pump may be turned on with the coarse valve closed. When the probe is in position, the openings around the probe and porthole must be blocked off to prevent unrepresentative dilution of the gas stream. The stack cross section is traversed, as required by Method I2 or as specified by the Agency. To minimize chance of extracting deposited material, the probe nozzle should not bump into the stack walls when sampling near the walls or when removing or inserting the probe through the portholes. During the test run, periodic adjustments are made to keep the probe temperature at the proper value. More ice and, if necessary, salt is added to the ice bath to maintain a temperature of less than 20°C (68°F) at the impinger/silica gel outlet, to avoid excessive moisture losses. Also, the level and zero of the manometer should be periodically checked. If the pressure drop across the train becomes high enough to make isokinetic sampling difficult to maintain, the test run should be terminated. Under no circumstances should the train be disassembled during the test run to determine and correct causes of excessive pressure drops. C-31 �At the end of the sample run, the pump is turned off, the probe and nozzle removed from the stack, and the final dry gas meter reading recorded. A leak check is performed, with acceptability of the test run based on the same criteria as in Section 8.1.4. The percent isokinetic is calculated (see calculation section) to determine whether another test run should be made. If there is difficulty in maintaining isokinetic rates due to source conditions, the Agency should be consulted for possible variance on the isokinetic rates. 8.1.6 8.2 Blank train - For each series of test runs, a blank train is set up in a manner identical to that described above, but with the nozzle capped with aluminum foil and the exit end of the last impinger capped with a ground glass cap. The train is allowed to remain assembled for a period equivalent to one test run. The blank sample is recovered as described in Section 8.3. Static air sampling3 - The sampling procedure for static air is identical to that described in Section 8.1 with the following exceptions: (a) impingers and a heatable probe are not required prior to the adsorbent tube; and (b) the PCB concentrations may dictate a longer or shorter sampling time. The selection of sampling time and rate should be based on the approximate levels of PCB residues expected in the sample. The sampling rate should not exceed 14 liter/rain and may typically fall in the range of 5 to 10 liter/rain. Sampling times should be more than 20 min but should not exceed 4 hr. 8.3 Sample recovery - Proper cleanup procedure begins as soon as the probe is removed from the stack at the end of the sampling period. When the probe can be safely handled, all external particulate matter near the tip of the probe nozzle is wiped off. The probe is removed from the train and both ends closed off with aluminum foil. The inlet to the train is capped off with a ground glass cap. The probe and impinger assembly are transfered to the cleanup area. This area should be clean and protected from the wind so that the chances of contaminating or losing the sample will be minimized. The train is inspected prior to and during disassembly and any abnormal conditions noted. The samples are treated as follows: 8.3.1 Adsorbent tube - The Florisil tube is removed from the train and capped with ground glass caps. 8.3.2 Sample Container No. 1 - The first three impingers are removed. The outside of each impinger is wiped off to remove excessive water and other debris. The impingers C-32 �are weighed (stem included), and the weight recorded on a data sheet. The contents are poured directly into Container No. 1. 8.3.3 8.3.4 8.4 Sample Container No. 2 - Each of the first three impingers are rinsed sequentially with 30-ml acetone and then with 30-ml hexane, and the rinses put into Container No. 2. Material deposited in the probe is quantitatively recovered using 100-ml acetone and then 100-ml hexane and these rinses added to Container No. 2. Silica gel container - The last impinger is removed, and the outside wiped to remove excessive water and other debris. It is weighed (stem included), and the weight recorded on the data sheet. The contents are transferred to the used silica gel can. Sample preservation - Samples should be stored in the dark at 4°C. Storage times in excess of 4 weeks are not recommended. 9.0 Sample Preparation1 9.1 Extraction 9.1.1 Adsorbent tube - The entire contents of the adsorbent tube are expelled directly onto a glass wool plug in the sample holder of a Soxhlet extractor. Although no extraction thimble is required, a glass thimble with a coarsefritted bottom may be used. The tube is rinsed with 5-ml acetone and then with 15-ml hexane and these rinses put into the extractor. The extraction apparatus is assembled and the adsorbent extracted with 170-ml hexane for at least 4 hr. The extractor should cycle 10 to 14 times per hour. After allowing the extraction apparatus to cool to ambient temperature, the extract is transferred into a KudernaDanish evaporator. The extract is evaporated to about 5 ml on a steam bath and the evaporator allowed to cool to ambient temperature before disassembly. The extract is transferred to a 50-ml separatory funnel and the funnel set aside. 9.1.2 Sample Container No. 1 - The aqueous sample is transferred to a 1,000-ml separatory funnel. The container is rinsed with 20-ml acetone and then with two 20-ml portions of hexane, adding the rinses to the separatory funnel. The sample is extracted with three 100 ml portions of hexane and the sequential extracts transferred to a Kuderna-Danish evaporator. C-33 �The extract is concentrated to about 5 ml and allowed to cool to ambient temperature before disassembly. The extract is filtered through a micro column of anhydrous sodium sulfate into a 50-ml separatory funnel containing the corresponding Florisil extract from Section 9.1.1. The micro column is prepared by placing a small plug of glass wool in the bottom of the large portion of a disposable pipette and then adding anhydrous sodium sulfate until the tube is about half full. 9.1.3 Sample Container No. 2 - The organic solution is transferred into a 1,000-ml separatory funnel. The container is rinsed with two 20 ml portions of hexane and the rinses added to the separatory funnel. The sample is washed with three 100 ml portions of water. The aqueous layer is discarded and the organic layer transferred to a KudernaDanish evaporator. The extract is concentrated to about 5 ml and allowed to cool to ambient temperature before disassembly. The extract is filtered through a micro column of anhydrous sodium sulfate into the 50-ml separatory funnel containing the corresponding Florisil and impinger extracts (Section 9.1.2). 9.2 Cleanup - Two tested cleanup techniques are described below.4 Depending upon the complexity of the sample, one or both of the techniques may be required to fractionate the PCBs from interferences. If the sample extract is colored, the Florisil column cleanup may be indicated. 9.2.1 Acid cleanup 9.2.1.1 Add 5 ml of concentrated sulfuric acid to the separatory funnel containing the sample extract and shake for 1 min. 9.2.1.2 Allow the phases to separate, transfer the sample (upper phase) with three 1 to 2 ml solvent rinses to Kuderna-Danish evaporator and concentrate to an appropriate volume. 9.2.1.3 Analyze as described in Section 10.0. 9.2.1.4 If the sample is highly contaminated, a second or third acid cleanup may be employed. 9.2.2 Florisil column cleanup 9.2.2.1 Variations among batches of Florisil may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to C-34 �completely elute all of the PCBs must be verified by the analyst. The weight of Florisil can then be adjusted accordingly. 9.2.2.2 Place a 20-g charge of Florisil, activated overnight at 130°C, into a Chroraaflex column. Settle the Florisil by tapping the column. Add about 1 cm of anhydrous sodium sulfate to the top of the Florisil. Pre-elute the column with 70-80 ml of hexane. Just before the exposure of the sodium sulfate layer to air, stop the flow. Discard the eluate. 9.2.2.3 Add the sample extract to the column. Add 225 ml of hexane to the column. Carefully wash down the inner wall of the column with a small amount of the hexane prior to adding the total volume. Discard the first 25 ml. 9.2.2.4 Collect 200 ml of hexane eluate in a KudernaDanish flask. All of the PCBs should be in this fraction. Concentrate to an appropriate volume. 9.2.2.5 Analyze the sample as described in Section 10.0. 10.0 Gas Chromatographic/Electron Impact Mass Spectrometric Determination 10.1 Internal standard addition - Pipet an appropriate volume of internal standard solution SSxxx into the sample. The final concentration of the internal standards must be in the working range of the calibration and well above the matrix background. The internal standards are thoroughly mixed by mechanical agitation. Note: The volume measurement of the spiking solution is critical to the overall method precision. The analyst must exercise caution that the volume is known ±1% or better. Where necessary, calibration of the pipet is recommended. Note: This same solution is used as a surrogate standard solution in the protocols for products/product waste and for water. In this protocol, the 13C-labeled PCBs are spiked after extraction, so are used as internal standards. Alternately, another internal standard solution such as the d63,3',4,4'-tetrachlorobiphenyl used in the product/product waste and water protocols may be used, if acceptable RF precision and accuracy are shown across the homolog range. 10.2 Tables 2, and 5 through 8 summarize the recommended operating conditions for analysis. Figure 5 presents an example of a chromatogram. C-35 �W O) ,H CO CO 00 0) 4-1 l-i 3 C ffl 4-1 c o K 10U.O vO ^O T3 U rH CO 1 rH rH C £*^ C QJ C QJ J^ p., fi Q) J2 p. -H £X ^Q -H -H XI XI ^-s O O Q) M l-l 4J O O O O i H C x; CJ co rJ 1-1 C 3 M 4-1 rH >> QJ QJ 1 o o en a I I ,c ^ )S *—' <f rH >-. r^> C <f H r-f <J •H J>"i *• _O £j ^3" c QJ x: XI -H rH jj") -H XI O IM C QJ P>i p. -H It 1C rH CO rH >% C o) x: P. ca l-i 0) -H l-i a co rH 1-1 >. C QJ 4-1 QJ H x: 1 p. 4-1 O |-i 0 tH QJ H 1 vO f. 0 cO 4J O rH X: O •H Q 1 ^J i X CJ -H M H | 04" 7 "** 01 P. ,--, P . x: 0 ca XI C 0 M O O CO -H o 1 vD 0 vO ^O v£) LO rH 0 X! O - 4-> rH 1-1 O l-i O XI 0 o 1-1 QJ XI rH 1 •H cfl M O x: 1-1 U 3 cfl CO CO a CJ —' QJ Q cfl - <r M - o rH CO X O) ffl rrj 1 vO - oo p. <J *. LO _ QJ rn x: r, LO f. 4-1 O •v ~^r ^ CO LO ** • t < CO CO n *, •- -<r ^ CN ^ CN] «s CM r. 0) Q r, CN ^ " J r, CO *x CO ' , 1208 o f. I3 C ' r. CN CN i ui ||7ll ™" ^ '' 7 | rH 1 2 f. r C7 CM" 1 r T II a o x : 4-1 XI m -<r ^-^ CJ O <f •s QJ PL, | LO - <3~ Ovl 04 fQ -H , n -H — >sO c - ^J" **3" , ; vo 0-1 CIO QJ x: P, rH M C &^ c rH 1 rH p, CO O XI P. •H XI O o ^i o rH x: o o 10: (W o X: •H XI «2_.Mh.. 1 c-.rj -H X3 rO o P. -~ p. n rH XI x; C X! o ,x C QJ rH o I 2 o l-l o rH x: rH i—i x: - 6 5 QJ QJ - CX rH CJ CO C JZ O O HC i t O O o o co rH rH (30 x: x: o 112592 CN rH J>^ 1 rH 1 rH rH >•>,>> ' CO 02? r 1M0 11:20 04 ™ I H (I »;« * "t lono IB: 1D I * 1I ,,,„ i uv: j ->0: CO ^_ 1 - V r-.'f' ^ 23:?'J soA'i Till.: Figure 5. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS) chromatogram or the calibration standard solution required for quantitation of PCBs by homolog. This chromatogram includes PCBs representative of each liomolog, three carbon-13 labeled surrogates, and the deuterated internal standard. The concentration of all components and the CGC/EIMS parameters are presented in Tallies 3, 4, 5, and 7. �10.3 While the highest available chromatographic resolution is not a necessary objective of this protocol, good chromatographic performance is recommended. With the high resolution of CGC, the probability that the chromatographic peaks consist of single compounds is higher than with PGC. Thus, qualitative and quantitative data reduction should be more reliable. 10.4 After performance of the system has been certified for the day and all instrument conditions set according to Tables 2, and 5 through 8, inject an aliquot of the sample onto the GC column. If the response for any ion, including surrogates and internal standard, exceeds the working range of the system, dilute the sample and reanalyze. If the responses of surrogates, internal standard, or analytes are below the working range, recheck the system performance. If necessary, concentrate the sample and reanalyze. 10.5 Record all data on a digital storage device (magnetic disk, tape, etc.) for qualitative and quantitative data reduction as discussed below. 11.0 Qualitative Identification 11.1 Selected ion monitoring (SIM) or limited mass scan (IMS) data The identification of a compound as a given PCS homolog requires that two criteria be met: 11.1.1 (1) The peak must elute within the retention time window set for that homolog (Section 7.6); and (2) the ratio of two ions obtained by SIM (Table 11) or by IMS (Table 12) must match the natural ratio within ±20%. The analyst must search the higher mass windows, in particular M+70, to prevent misidentification of a PCB fragment ion cluster as the parent. 11.1.2 If one or the other of these criteria is not met, interferences may have affected the results and a reanalysis using full scan EIMS conditions is recommended. 11.2 Full scan data 11.2.1 The peak must elute within the retention time windows set for that homolog (as described in Section 7.6). 11.2.2 The unknown spectrum must match that of an authentic PCB. The intensity of the three largest ions in the molecular cluster (two largest for monochlorobiphenyls) must match the natural ratio within ±20%. Frequent clusters with proper intensity ratios must also be present. 11.2.3 Alternatively, a spectral search may be used to automatically reduce the data. The criteria for acceptable C-37 �TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs Ion (relative intensity) Secondary Tertiary Homolog Primary ^12^9^1 188 (100) 190 (33) C^HgC^ 222 (100) 224 (66) 226 (11) Ci2H7Cl3 256 (100) 258 (99) 260 (33) C12H6C14 292 (100) 290 (76) 294 (49) C12H5C15 326 (100) 328 (66) 324 (61) £12^4^16 360 (100) 362 (82) 364 (36) C12H3C17 394 (100) 396 (98) 398 (54) Ci2H2Cl8 430 (100) 432 (66) 428 (87) Cj^HClg 464 (100) 466 (76) 462 (76) C 498 (100) 500 (87) 496 (68) 12 C llO - Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in Identification of Polychlorinated Biphenyls by Mass Spectrometry," J. Assoc. Offic. Anal. Chem., 56(1), 188-199 (1973). C-38 �TABLE 12. LIMITED MASS SCANNING (LMS) RANGES FOR PCBs Compound Mass range (m/z) CX2H.CU 186-190 C12H8C12 220-226 C12H7C13 254-260 C12H6C13 288-294 Ci2H5Cl5 322-328 Ci2H4Cl6 356-364 C12H3C17 386-400 C12H2C18 426-434 C12HC19 460-468 CisCl^ 494-504 C12D6C14 294-300 13 192-196 13 300-306 C612C6H9C1 C12H6C14 13 438-446 -C12C110 506-516 C12H2C18 a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass Spectrometry Method for Identifying and Determining Polychlorinated Biphenyls," J. Chromatogr.t 196, 109-119 (1980). C-39 �identification include a high index of similarity. For the Incos 2300, a fit of 750 or greater must be obtained. 11.3 Disputes in interpretation - Where there is reasonable doubt as to the identity of a peak as a PCB, the analyst must either identify the peak as a PCB or proceed to a confirmational analysis (see Section 13.0). 12.0 Quantitative Data Reduction 12.1 Once a chromatographic peak has been identified as a PCB, the compound is quantitated based either on the integrated abundance of the SIM data or EICP for the primary characteristic ion in Tables 11 and 12. If interferences are observed for the primary ion, use the secondary and then tertiary ion for quantitation. If interferences in the parent cluster prevent quantitation, an ion from a fragment cluster (e.g., M-70) may be used. Whichever ion is used, the RF must be determined using that ion. The same criteria should be applied to the internal standard compounds (Table 13). 12.2 Using the appropriate response factor (RF ) as determined in Section 7.3, calculate the mass of each PCB peak (M ) using Equation P 12-1. A , M = -E • 'M p A.is RF p is Eq. 12-1 H * where A = area of the characteristic ion for the analyte PCB " peak A. = area of the characteristic ion for the internal standard peak RF = response factor of a given PCB congener M. S = mass of internal standard injected (micrograms) ~L 12.3 If a peak appears to contain non-PCB interferences which cannot be circumvented by a secondary or tertiary ion, either: 12.3.1 12.3.2 Perform additional chemical cleanup (Section 9) and then reanalyze the sample; or 12.3.3 12.4 Reanalyze the sample on a different column which separates the PCB and interferents; Quantitate the entire peak as PCB. Sum all of the peaks for each homolog and then sum those to yield the total PCB mass, MT, in the sample. If a concentration-perpeak or concentration-per-homolog reporting format is desired, carry each value through the calculations in an appropriate manner. C-40 �TABLE 13. CHARACTERISTIC IONS FOR Specific compound C612C6H9C1 ^ Primary 13 C-LABELED PCS SURROGATES Ion (relative intensity) Secondary Tertiary 13 194 (100) 196 (33) 13 304 (100) 306 (49) 302 (78) 13 442 (100) 444 (65) 440 (89) 510 (100) 512 (87) 514 (50) C12H6C14 C12H2C18 13 C12C110 C-41 �12.5 Calculation of air sample volume1 12.5.1 Nomenclature M = Mass of PCB represented by a chromatographic peak " micrograms M~ = Total mass of PCBs in sample, micrograms C = Concentration of PCBs in air, micrograms per cubic meter, corrected to standard conditions of 20°C, 760 mm Hg (68°F, 29.92 in. Hg) on dry basis A = Cross-sectional area of nozzle, square meter (square n feet) B = Water vapor in the gas stream, proportion by volume I = Percent of isokinetic sampling MWw = Molecular weight of water, 18 g/g-mole (18 lb/ Ib-mole) P, = Barometric pressure at the sampling site, mm Hg oar /. (in. IT \ Hg) Ps = Absolute stack gas pressure, mm Hg (in. Hg) Pstd, = Standard absolute pressure, 760 mm Hg (29.92 in Hg) R = Ideal gas constant, 0.06236 mm Hg-m3/K-g-raole (21.83 in. Hg-ft3/°R-lb-mole) T = Absolute average dry gas meter temperature °K (°R) TS = Absolute average stack gas temperature °K (°R) = Standard absolute temperature, 293°K (528°R) V, = Total volume of liquid collected in impingers and silica gel, milliliters. Volume of water collected equals the weight increase in grams times 1 ml/g Vm = Volume of gas sample as measured by dry gas meter, dcm (dcf) V , ,x = Volume of gas sample measured by the dry gas meter corrected to standard conditions, dscra (dscf) C-42 �,N = Volume of water vapor in the gas sample corrected to standard conditions, son (scf) V = Total volume of sample, railliliter V = Stack gas velocity, calculated by EPA Method 2, s m/sec (ft/sec) AH = Average pressure differential across the orifice meter, mm H20 (in. H20) Pw = Density of water, 1 g/ml (0.00220 Ib/ral) 6 = Total sampling time, minutes 13.6 = Specific gravity of mercury 60 = Seconds per minute 100 = Conversion to percent 12.5.2 Average dry gas meter temperature and average orifice pressure drop - See data sheet (Figure 4). 12.5.3 Dry gas volume - Correct the sample volume measured by the dry gas meter to standard conditions [20°C, 760 mm Hg (68°F, 29.92 in. Hg)] by using Equation 12-2. ., Vstd) - T „ *std V m - P +M *bar 13.6 _ „„ = P + AH bar 1376" Eq. 12-2 V where K = 0.3855°K/mm Hg for metric units = 17.65 °R/in. Hg for English units 12.5.4 Volume of water vapor P w RTstd Vw(std) = Ic MW P 5,^ = K Ic , , ,, V, ajT- .V, v ' w std where K = 0.00134 m3/ml for metric units = 0.0472 ft3/ml for English units 12.5.5 Eq. ^ 12-3 Moisture content = -J^Iltd)E 1. 2 4 V +V m(std) w(std) If the liquid droplets are present in the gas stream, assume the stream to be saturated and use a psychrometric chart to obtain an approximation of the moisture percentage. B ws C-43 �12.6 Concentration of PCBs in stack gas - Determine the concentration of PCBs in the air according to Equation 12-5 and report in micrograms per cubic meter using Table 14. If an alternate reporting format (e.g., concentration per peak) is desired, a different report form may be used. M C a = K y—-- Eq. 12-5 m(std) where K = 35.31 ft3/m3 12.7 Isokinetic variation 12.7.1 Calculations from raw data. [K V, + (V /T ) (P, ) + AH/13.6)] 1nn Ts l 100 Ic m m bar 10 , T r i = -60s e v P A- Eq- 12' s n 3 where K = 0.00346 mm Hg-m /ml-°K for metric units = 0.00267 in. Hg-ft3/ml-°R for English units 12.7.2 Calculations from intermediate values T V f «. ,, P . . 100 T s m(std) std 1 " T , , s 6n s 60 (1-B J A P std V ws _ ,7 ' /"/ tq ,, _s Vm(std) _ T Ps Vs An 0 (1-B ws) where K = 4.323 for metric units = 0.0944 for English units 12.7.3 Acceptable results - The following range sets the limit on acceptable isokinetic sampling results: If 90% < I < 110%, the results are acceptable. If the results are low in comparison to the standards and I is beyond the acceptable range, the Agency may opt to accept the results. 12.8 Round off all numbers reported to two significant figures. 13.0 Confirmation If there is reason to question the qualitative identification (Section 11.0), the analyst may choose to confirm that a peak is not a PCB. Any technique may be chosen provided that it is validated as having equivalent or superior selectivity and sensitivity to GC/EIMS. Some candidate techniques include alternate GC columns (with EIMS detection), GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory C-44 �TABLE 14. ANALYSIS REPORT INCIDENTAL PCBs IN AIR Sample No. Sample Matrix Sample Source Notebook No. or File Location m3 Volume Collected [V , fcdJ Mass of Internal Stanaara Injected, M. is pg Qualitative Analyte 1° 2° I l° T 2° Ratio Theoretical IS 298 246 100/76 1-C1 188 190 100/33 2-C1 222 224 100/66 3-C1 256 258 100/99 4-C1 292 290 100/76 5-C1 326 328 Quantitative Ion Mass OK? Used RF M (pg) P 1.000 100/66 i 6-C1 360 362 100/82 7-C1 394 396 100/98 8-C1 430 432 100/66 9-C1 464 466 100/76 10-C1 498 500 100/87 Total (MT) Concentration (C.) M£ 3 Mg/m Reported by: Internal Audit: Name Name EPA Audit: Name Signature/Date Signature/Date Signature/Date Organization Organization Organization C-45 �must validate confirmation techniques to show equivalent or superior selectivity between PCBs and interferences and sensitivity (limit of quantitation, LOQ). If a peak is confirmed as being a non-PCB, it may be deleted from the calculation (Section 12). If a peak is confirmed as containing both PCB and non-PCB components, it must be quantitated according to Section 12.3. 14.0 Quality Control 14.1 Each laboratory that uses this method must operate a formal quality control (QC) program. The minimum requirements of this program consist of an initial demonstration of laboratory capability and the analysis of spiked samples as a continuing check on performance. The laboratory must maintain performance records to define the quality of data that are generated. After a date specified by the Agency, ongoing performance checks should be compared with established performance criteria to determine if the results of analyses are within accuracy and precision limits expected of the method. 14.2 The analysts must certify that the precision and accuracy of the analytical results are acceptable by: 14.2.1 14.2.2 14.3 The absolute precision of surrogate recovery, measured as the RSD of the integrated EIMS area (A ) for a set of samples, must be ±10%. The mean recovery (R ) of at least four replicates of a QC check sample to be supplied by the Agency must meet Agency-specified accuracy and precision criteria. This forms the initial data base for establishing control limits (see Section 14.3 below). Control limits - The laboratory must establish control limits using the following equations: Upper control limit (UCL) = R + 3 RSD Upper warning limit (UWL) = R + 2 RSD Lower warning limit (LWL) = RC - 2 RSDc Lower control limit (LCL) = R - 3 RSD These may be plotted on control charts. If an analysis of a check sample falls outside the warning limits, the analyst should be alerted that potential problems may need correction. If the results for a check sample fall outside the control limits, the laboratory must take corrective action and recertify the performance C-46 �(Section 14.2) before proceeding with analyses. The warning and control limits should be continuously updated as more check sample replicates are added to the data base. 14.4 Before processing any samples, the analyst should demonstrate through the analysis of a reagent blank that all glassware and reagent interferences are under control. Each time a set of samples is analyzed or there is a change in reagents, a laboratory reagent blank should be processed as a safeguard against contamination. 14.5 Procedural QC - The various steps of the analytical procedure should have quality control measures. These include but are not limited to: 14.5.1 GC performance - See Section 7.1 for performance criteria. 14.5.2 MS performance - See Section 7.2 for performance criteria. 14.5.3 Qualitative identification - At least 10% of the PCB identifications, as well as any questionable results, should be confirmed by a second mass spectrometrist. 14.5.4 Quantitation - At least 10% of all manual calculations, including peak area calculation, must be checked. After changes in computer quantitation routes, the results should be manually checked. 14.6 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, must be selected at random, sampled, and analyzed in triplicate to monitor the precision of the analysis. An RSD of ±30% or less must be achieved. If the precision is greater than ±30%, the analyst must be recertified (see Section 14.2). 14.7 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, selected at random, must be analyzed by the standard addition technique. Two aliquots of the sample are analyzed, one "as is" and one spiked with a sufficient amount of solution CSxxx to yield approximately 100 pg/ sample of each compound. The spiking compounds are thoroughly incorporated by mechanical agitation. For the liquid impinger contents, shaking for 30 sec should be sufficient. For the Florisil, 10 min tumbling is recommended. For filters where inadequate incorporation may be expected, overnight equilibration with agitation is recommended. Note: The volume measurement of the spiking solution is critical to the overall method precision. The analyst must exercise caution that the volume is known to ±1% or better. Where necessary, calibration of the pipet is recommended. C-47 �The samples are analyzed together and the quantitative results calculated. The recovery of the spiked compounds (calculated by difference) must be 80-120%. If the sample is known to contain specific PCB isomers, these isomers may be substituted for solution CSxxx. If the concentrations of PCBs are known to be high, the amount added should be adjusted so that the spiking level is 1.5 to 4 times the measured PCB level in the unspiked sample. 14.8 Sampling efficiency - The efficiency of PCB collection during sampling should be monitored. This may be achieved by adding a known amount of the 13C surrogate spiking solution (Section 6.4) sufficient to give an analytical signal well above background to the first impinger prior to sampling. The recovery of the four compounds should be > 14.9 Interlaboratory comparison - Interlaboratory comparison studies are planned. Participation requirements, level of performance, and the identity of the coordinating laboratory will be presented in later revisions. 14.10 It is recommended that the participating laboratory adopt additional QC practices for use with this method. The specific practices that are most productive depend upon the needs of the laboratory and the nature of the samples. Field duplicates or triplicates may be analyzed to monitor the precision of the sampling technique. Whenever possible, the laboratory should perform analysis of standard reference materials and participate in relevant performance evaluation studies. 15.0 Quality Assurance Each participating laboratory must develop a quality assurance plan according to EPA guidelines.5 The quality assurance plan must be submitted to the Agency for approval. 16.0 Method Performance The method performance is being evaluated. Limits of quantitation; average intralaboratory recoveries, precision, and accuracy; and interlaboratory recoveries, precision, and accuracy will be presented. 17.0 Documentation and Records Each laboratory is responsible for maintaining full records of the analysis. Laboratory notebooks should be used for handwritten records. GC/MS data must be archived on magnetic tape, disk, or a similar device. Hard copy printouts may be kept in addition if desired. QC records should be maintained separately from sample analysis records. C-48 �The documentation roust describe completely how the analysis was performed. Any variances from the protocol must be noted and fully described. Where the protocol lists options (e.g., sample cleanup), the option used and specifies (solvent volumes, digestion times, etc.) must be stated. C-49 �REFERENCES 1. Haile, C. L., and E. Baladi, "Methods for Determining the Polychlorinated Biphenyl Emissions from Incineration and Capacitor and Transformer Filling Plants," U.S. Environmental Protection Agency, (1977) EPA-600/4-73-048. 2. U.S. Environmental Protection Agency, Federal Register, 42(160), Thursday, August 18, 1977. 3. Martin, R. M., "Construction Details of Isokinetic Source Sampling Equipment," Environmental Protection Agency, Air Pollution Control Office Publication No. APTD-0581. 4. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environmental Protection Agency, (1981) EPA-600/4-81-045. 5. Quality Assurance Program Plan for the Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C., October 1980. C-50 �APPENDIX D ANALYTICAL METHOD; THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN INDUSTRIAL WASTEWATER D-l �THE ANALYSIS OF BY-PRODUCT CHLORINATED BIPHENYLS IN INDUSTRIAL WASTEWATER i.0 Scope and Application 1.1 This is a gas chromatographic/electron impact mass spectrometric (GC/EIMS) method applicable to the determination of chlorinated biphenyls (PCBs) in industrial wastewater. The PCBs present may originate either as synthetic by-products or as contaminants derived from commercial PCB products (e.g., Aroclors). The PCBs may be present as single isomers or complex mixtures and may include all 209 congeners from monochlorobiphenyl through decachlorobiphenyl listed in Table 1. 1.2 The detection and quantitation limits are dependent upon the volume of sample extracted the complexity of the sample matrix and the ability of the analyst to remove interferents and properly maintain the analytical system. The method accuracy and precision will be determined in future studies. 1.3 This method is restricted to use by or under the supervision of analysts experienced in the use of gas chromatography/mass spectrometry (GC/MS) and in the interpretation of gas chromatograms and mass spectra. Prior to sample analysis, each analyst must demonstrate the ability to generate acceptable results with this method by following the procedures described in Section 14.2. 1.4 The validity of the results depends on equivalent recovery of the analyte and 13C PCBs. If the *3C PCBs are not thoroughly incorporated in the matrix, the method is not applicable. 1.5 During the development and testing of this method, certain analytical parameters and equipment designs were found to affect the validity of the analytical results. Proper use of the method requires that such parameters or designs must be used as specified. These items are identified in the text by the word "must." Anyone wishing to deviate from the method in areas so identified must demonstrate that the deviation does not affect the validity of the data. Alternative test procedure approval must be obtained from the Agency. An experienced analyst may make modifications to parameters or equipment identified by the term "recommended." Each time such modifications are made to the method, the analyst must repeat the procedure in Section 14.2. In this case, formal approval is not required, but the documented data from Section 14.2 must be on file as part of the overall quality assurance program. D-2 �TABLE 1. No. Structure NO. Menoeftloraoiphenylt 2 52 3 4 54 55 56 S7 IS » 60 61 (2 63 64 65 66 67 68 01chlorob<BHttiy1i a 9 10 n 12 13 14 IS 2.2' 2.3 2.3' 2.4 2.4' 2.5 2.6 3,3' 3,4 3.4' 3|5 4.4' Trlchloraolphtnyli 16 17 18 19 20 21 22 23 24 2S 26 27 28 29 JO 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 SO SI 2. 2'. 3 2.2'. 4 2. 2'. 5 2, 2', 6 2.3.3' 2,3.4 2.3.4' 2,3.5 2,3.6 2,3', 4 2, 3', 5 2,3', 6 2.4,4' 2.4,5 2.4,6 2, 4'. 5 2. 4' ,5 2', 3,4 2', 3,5 3,3', 4 3.3'.5 3,4,4' 3.4.5 3.4'. 5 NO. 69 70 71 72 73 74 75 76 77 78 79 80 81 10$ 2.3,3 ,4,4' 1 1 1 C 2 .J.J ,4,9 2,3,3 .4'. 5 2.3.3 ,4.5' 2,3.3 ,4.6 2,3.3 .4* ,6 2.3,3 ,5,5' 2,3,3 .5.6 2.3,3 .5', 6 2,3.4 4'.5 2.3,4 4' ,6 2.3,4 5,6 2.3,4', 5, 6 2,3'.4.4-.5 2.3'. 4. 4' .6 2.3'. 4, 5. 5' 2.3". 4,1 6 5', 21 .3. 3 . 4.5 2'. 3.4. 4' .5 21 .3.4.5. 52'.3.4.5,6' 3. 3'. 4. 4' .5 3,3'. 4. 5.5' 170 171 172 173 174 175 176 177 178 179 180 Htmeliloroblptitnylt 2.2'. «,S' 9* « *' 2 •* •*»• 2,2'. 5,62.3. 3'. 4 2.3.3'. 4' 2.3.3' .5 2.3.3-. S' 2.3,J'.< 2.3.4.4' 2.3.4.5 2,3.4,6 2.3. 4' .5 2,3.4-. 6 2.3,5.6 2.3', 4.4' 2.3', 4,5 2.3', 4.5" 2.3'. 4.6 2.3-.4'.5 2.3-,4-.6 .3'.S.5' ,3-.5',6 .4.4'. 5 ,4. 4', 6 '.3 4.5 .3' 4,4.3' 4,5 .3' 4,5' .3' 5.5.4.4-.S 2.2-.3.3'.4 2.2'. 3. 3', 5 2.2'.3.3',6 2. 2'. 3.4,4' 2,2' .3.4.5 2.2- .3, 4, 5' 2.2'. 3, 4,6 2,2'. 3.4,6"' 2.2'. 3, 4'. 5 2.2'. 3. 4'. 6 2.2', 3,5.5' 2.2'. 3, 5.6 2.2-.3.S.6' 2.2'. 3,5', 6 2.2-,3,6.6' 2.2-,3'.4.5 2.2'.3',4,6 2.2'. 4,4', 5 2.21,4.4'.6 2.2'. 4, 5.5' 2,2'. 4.5.5' 2.2-,4.S',6 2,2'. 4.6. S' 10 . 182 P«nt»ctil orebt phtny 1» Irtit IUO 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 Pmticftl arotl pHtny 1 » 82 S3 84 85 86 87 88 39 90 91 92 93 94 T«tricMorob<Bh«nyl» 95 96 97 2.2', 3.3' 2.2'. 3,4 98 2.2'. 3,499 2.2'.3,$ 100 2.2'.3.5' 101 2.2'. 3.6 10Z 2.2'.3,6' 103 2.2'. 4,4104 2,2' ,4, 5 2,2'. 4.J' 2 2 1 4 61 2.2 .4.6 structure Tttnetil arott Ph«nv1 s 1 4 5 6 7 NUMBERING OF PCB CONGENERS3 structure 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 2,2'.3.3-,4.42,2'. 3,3' .4. 2.2'.3.3',4. 2.2' .3,3'. 4. .2', 3.3'. 4, ,2'. 3, 3', 5, .2'.3,3',S. .2'.3.3'.5, ,2'. 3,3' 6, ' ,2'.3,4,4- ( 1 ,2'. 3,4. 4'. .2'. 3. 4.4' , 1 2.2". 3,4.4'. 2.2' .3,4, 5, 5 2,2'. 3,4, 5,6 2.2',3,4.5,6 2.2'. 3. 4,5'. 2.2'. 3,4, 6. 6 2,2' ,3,4'. 5. 2.2' ,3, 4', S. 2.2'.3,4'.5, 2.2', 3,4'. 5' 6 2.2', 3, 4' .5, 2.2-.3,5.5'. 2.2', 3.5.6.6' 2.2' ,4.4'. 5.5' 2.21, 4,4-. 5. S2.2',4.4'.M' 2,3,3'. 4, 4'. 5 2.3. 3' .4,4' .5' 2.3.3', 4, 4'. 6 2.3.3'. 4,5, 5' 2,3,3'. 4,i,« HtueMorobiphtnyl* 161 1X9 IK 163 164 165 166 167 168 169 tallieMttr, 181 183 184 185 186 187 188 189 190 191 192 193 2.3.3'. 4'. 5.6 2,3, 3', 4'. 5 ' , 6 2;3;3'.S.S',6 2.3.4. 4- .5,6 2,3' .4.4'. 5. 5' 2.3'. 4, 4', £'.6 3,3'.4.4'.5,5 1 2,2',3,3',4,4',5 2.2',3,3'.4,4',6 2.2', 3, 3 ' , 4, 5, 5' 2, 2'. 3, 3', 4,5,6 2. 2'. 3, 3', 4, 5,6' 2.2' .3,3' .4. 5'. 6 2,2'. 3.3', 4.6. 6' 2. 2 1 . 3,3' ,4' ,5,6 2. 2', 3,3'. 5, 5'. 6 2,2',3,3',5,6,6' 2.2-.3.4,4'.5.5' 2. 2', 3, 4, 4', 5. 6 2. 2'. 3, 4, 4', 5, 6' 2,2'. 3. 4, 4'. 5'. 6 2.2',3.4,4',j,6 1 2.2' .3. 4,5, 5' ,6 2,2' .3,4,5, 6, 6' 2,2',3,4 1 ,5.5 1 .6 2,2'. 3,4', 5. 6, 6' 2. 3, 3', 4, 4', 5,5' 2,3.3', 4. 4 ' , 5, 6 2. 3. 3' .4. 4' .5'. 6 2, 3, 3'. 4, 5,5' 6 2.3, 3'. 4' .5,5' ,6 Octiehl orobi phtny 1 s 194 195 196 197 198 199 200 201 202 203 204 205 2,2'. 3. 3', 4. 4 ' , 5,5' 2,2'.3,3'.4,4'.5,6 2,2'.3,3'.4.4',5,5' 2, 2', 3. 3'. 4, 4 ' , 6, 6' 2.2'.3.3',4,5.5'.S 2,2'. 3, 3' .4. 5,6, 6'1 2.2'.3.3 1 ,4.S'.6.6 2.2'. 3, 3'. 4. 5, 5 ' , 6' 2.2' .3. 3' ,5, 5'. 6.6' 2,2', 3, 4. 4'. 5, 5 ' . 6 2.2' .3,4,4' .5, 6, 6' 2,3.3' ,4.4'.5.5'.6 NomehlarobfohtnyM 206 207 208 2,2-.3,3'.4.4',5,5 < .S 2.2'.3.3'.4.4'.5,6,6' 2,2',3.3 I ,4,5,5'.6,6' 0«eicMorot>(pn«ityl X. ind Z«11. H., Frwnlui I. Anil. ChM., 302. 20-31 (1980). D-3 2.3. 3', 4, 5', 6 H»otieH1orattfphtnyl » 209 •Adapted 1rm structure 2.2'.3,3'4,4',$,5',6.6' �2.0 Summary 2.1 The wastewater must be sampled such that the specimen collected for analysis is representative of the whole. Statistically designed selection of the sampling position (valve, port, outfall, etc.) or time should be employed. The sample must be preserved to prevent PCB loss prior to analysis. Storage at 4°C with optional preservation at low pH is recommended. 2.2 The sample is mechanically homogenized and subsampled if necessary. The sample is then spiked with four 13C PCB surrogates and the surrogates incorporated by further mechanical agitation. 2.3 The surrogate-spiked sample is extracted and cleaned up at the discretion of the analyst. Possible extraction techniques include liquid-liquid partition and sorption onto resin columns followed by solvent elution. Cleanup techniques may include liquid-liquid partition, sulfuric acid cleanup, saponification, adsorption chromatography, gel permeation chromatography or a combination of cleanup techniques. The sample is diluted or concentrated to a final known volume for instrumental determination. The EPA Method 6081 and 6252 extraction and cleanup procedures may be used. 2.4 The PCB content of the sample extract is determined by capillary (preferred) or packed column gas chromatography/electron impact mass spectrometry (CGC/EIMS or PGC/EIMS) operated in the selected ion monitoring (SIM), full scan, or limited mass scan (IMS) mode. 2.5 PCBs are identified by comparison of their retention time and mass spectral intensity ratios to those in calibration standards. 2.6 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners, using the response of the 13C surrogate to compensate for losses in workup and instrument variability. 2.7 The PCBs identified by the SIM technique may be confirmed by full scan CGC/EIMS, retention on alternate GC columns, other mass spectrometric techniques, infrared spectrometry, or other techniques, provided that the sensitivity and selectivity of the technique is demonstrated to be comparable or superior to GC/EIMS. 2.8 The analysis time is dependent on the extent of workup employed. The time required for instrumental analysis, excluding data reduction and reporting, is about 30 to 45 min. 2.9 Appropriate quality control (QC) procedures are included to assess the performance of the analyst and estimate the quality of the results. These QC procedures include the demonstration of laboratory capability: periodic analyst certification, the use of control charts, and the analysis of blanks, replicates, and standard addition samples. A quality assurance (QA) plan must be developed for each laboratory. D-4 �2.10 While several options are available throughout this method, the recommended procedure to be followed is: 2.10.1 The sample is collected according to a scheme which permits extrapolation of the sample data to the body or containers of water being sampled. 2.10.2 The sample is preserved at low pH and at 4°C to prevent any loss of PCBs or changes in matrix which may adversely affect recovery. 2.10.3 The sample is-mechanically homogenized and subsampled if necessary. 2.10.4 The sample is spiked with four 13C-PCB surrogates (4-chlorobiphenyl; 3,3',4,4'-tetrachlorobiphenyl; 2,2',3,3',5,5',6,6'-octachlorobiphenyl; and decachlorobiphenyl). 2.10.5 The sample is extracted. 2.10.6 The extract is cleaned up and concentrated to an appropriate volume. 2.10.7 An aliquot of the extract is analyzed by CGC/EIMS operated in the SIM mode. On-column injections onto a 15-m DB-5 capillary column, programmed (for toluene solutions) from 110° to 325°C at 10°/min after a 2 min hold is used. Helium at 45-cm/sec linear velocity is used as the carrier gas. 2.10.8 PCBs are identified by retention time and mass spectral intensities. 2.10.9 PCBs are quantitated against the response factors for a mixture of 11 PCB congeners. 2.10.10 The total PCBs are obtained by summing the amounts for each homolog found and the concentration is reported as micrograms per liter. 3.0 Interferences 3.1 Method interferences may be caused by contaminants in solvents, reagents, glassware, and other sample processing hardware, leading to discrete artifacts and/or elevated baselines in the total ion current profiles. All of these materials must be routinely demonstrated to be free from interferences by the analysis of laboratory reagent blanks as described in Section 14.4. D-5 �3.1.1 Glassware must be scrupulously cleaned. All glassware is cleaned as soon as possible after use by rinsing with the last solvent used. This should be followed by detergent washing with hot water and rinses with tap water and reagent water. The glassware should then be drained dry and heated in a muffle furnace at 400°C for 15 to 30 min. Some thermally stable materials, such as PCBs, may not be eliminated by this treatment. Solvent rinses with acetone and pesticide quality hexane may be substituted for the muffle furnace heating. Volumetric ware should not be heated in a muffle furnace. After it is dry and cool, glassware should be sealed and stored in a clean environment to prevent any accumulation of dust or other contaminants. It is stored inverted or capped with aluminum foil. 3.1.2 The use of high purity reagents and solvents helps to minimize interference problems. Purification of solvents by distillation in all-glass systems may be required. All solvent lots must be checked for purity prior to use. 3.2 Matrix interferences may be caused by contaminants that are coextracted from the sample. The extent of matrix interferences will vary considerably from source to source, depending upon the nature and diversity of the sources of samples. 4.1 The toxicity or carcinogenicity of each reagent used in this method has not been precisely defined; however, each chemical compound should be treated as a potential health hazard. From this viewpoint, exposure to these chemicals must be reduced to the lowest possible level by whatever means available. The laboratory is responsible for maintaining a current awareness file of OSHA regulations regarding the safe handling of the chemicals specified in this method. A reference file of material data handling sheets should also be made available to all personnel involved in the chemical analysis. 4.2 Polychlorinated biphenyls have been tentatively classified as known or suspected human or mammalian carcinogens. Primary standards of these toxic compounds should be prepared in a hood. Personnel must wear protective equipment, including gloves and safety glasses. 4.0 Congeners highly substituted at the meta and para positions and unsubstituted at the ortho positions are reported to be the most toxic. Extreme caution should be taken when handling these compounds neat or in concentration solution. The class includes 3,3',4,4'-tetrachlorobiphenyl (both natural abundance and isotopically labeled). D-6 �4.3 4.4 5.0 Diethyl ether should be monitored regularly to determine the peroxide content. Under no circumstances should diethyl ether be used with a peroxide content in excess of 50 ppm as an explosion could result. Peroxide test strips manufactured by EM Laboratories (available from Scientific Products Company, Cat. No. P1126-8 and other suppliers) are recommended for this test. Procedures for removal of peroxides from diethyl ether are included in the instructions supplied with the peroxide test kit. Waste disposal must be in accordance with RCRA and applicable state rules. Apparatus and Materials 5.1 Sampling containers - Amber glass bottles, 1-liter or other appropriate volume, fitted with screw caps lined with Teflon. Cleaned foil may be substituted for Teflon if the sample is not corrosive. If amber bottles are not available, samples should be protected from light using foil or a light-tight outer container. The bottle must be washed, rinsed with acetone or methylene chloride, and dried before use to minimize contamination. 5.2 Glassware - All specifications are suggestions only. Catalog numbers are included for illustration only. 5.2.1 Volumetric flasks - Assorted sizes. 5.2.2 Pipets - Assorted sizes, Mohr delivery. 5.2.3 Micro syringes - 10.0 pi for packed column GC analysis, 1.0 |Jl for on-column CGC analysis. 5.2.4 Chromatographic column - Chromaflex, 400 mm long x 19 mm ID (Kontes K-420540-9011 or equivalent). 5.2.5 Gel permeation chromatograph - GPC Autoprep 1002 (Analytical Bio Chemistry Laboratories, Inc.) or equivalent. 5.2.6 Kuderna-Danish Evaporative Concentrator Apparatus 5.2.6.1 Concentrator tube - 10 ml, graduated (Kontes K-570050-1025 or equivalent). Calibration must be checked. Ground glass stopper size (S19/22 joint) is used to prevent evaporation of solvent. 5.2.6.2 Evaporative flask - 500 ml (Kontes K-57001-0500 or equivalent). Attach to concentrator tube with springs (Kontes K-662750-0012 or equivalent). 5.2.6.3 Snyder column - Three ball macro (Kontes K5030000121 or equivalent). D-7 �5.3 Balance - Analytical, capable of accurately weighing 0.0001 g. 5.4 Gas chromatography/mass spectrometer system. 5.4.1 Gas chromatograph - An analytical system complete with a temperature programmable gas chromatograph and all required accessories including syringes, analytical columns, and gases. The injection port must be designed for oncolumn injection when using capillary columns or packed columns. Other capillary injection techniques (split, splitless, "Grob," etc.) may be used provided the performance specifications stated in Section 7.1 are met. 5.4.2 Capillary GC column - A 12-20 m long x 0.25 mm ID fused silica column with a 0.25 |Jm thick DB-5 bonded silicone liquid phase (J&W Scientific) is recommended. Alternate liquid phases may include OV-101, SP-2100, Apiezon L, Dexsil 300, or other liquid phases which meet the performance specifications stated in Section 7.1. 5.4.3 Packed GC column - A 180 cm x 0.2 cm ID glass column packed with 3% SP-2250 on 100/120 mesh Supelcoport or equivalent is recommended. Other liquid phases which meet the performance specifications stated in Section 7.1 may be substituted. 5.4.4 Mass spectrometer - Must be capable of scanning from 150 to 550 Daltons every 1.5 sec or less, collecting at least five spectra per chromatographic peak, utilizing a 70-eV (nominal) electron energy in the electron impact ionization mode and producing a mass spectrum which meets all the criteria in Table 2 when 50 ng of decafluorotriphenyl phosphine [DFTPP, bis(perfluorophenyl)phenyl phosphine] is injected through the GC inlet. Any GC-to-MS interface that gives acceptable calibration points at 10 ng per injection for each PCB isomer in the calibration standard and achieves all acceptable performance criteria (Section 10) may be used. Direct coupling of the fused silica column to the MS is recommended. Alternatively, GC-toMS interfaces constructed of all glass or glass-lined materials are recommended. Glass can be deactivated by silanizing with dichlorodimethylsilane. 5.4.5 A computer system that allows the continuous acquisition and storage on machine-readable media of all mass spectra obtained throughout the duration of the chromatographic program must be interfaced to the mass spectrometer. The data system must have the capability of integrating the abundances of the selected ions between specified limits and relating integrated abundances to concentrations using the calibration procedures described in this method. The computer must have software that allows D-8 �TABLE 2. DFTPP KEY IONS AND ION ABUNDANCE CRITERIA Mass Ion abundance criteria 197 198 199 Less than 1% of mass 198 100% relative abundance 5-9% of mass 198 275 10-30% of mass 198 365 Greater than 1% of mass 198 441 442 443 Present, but less than mass 443 Greater than 40% of mass 198 17-23% of mass 442 D-9 �searching any GC/MS data file for ions of a specific mass and plotting such ion abundances versus time or scan number to yield an extracted ion current profile (EICP). Software must also be available that allows integrating the abundance in any EICP between specified time or scan number limits. 6.0 Reagents 6.1 Solvents - All solvents must be pesticide residue analysis grade. New lots should be checked for purity by concentrating an aliquot by at least as much as is used in the procedure. 6.2 Stock standard solutions - Standards of the PCB congeners listed in Table 3 are available from Ultra Scientific, Hope, Rhode Island; or Analabs, North Haven, Connecticut. 6.3 Calibration standard stock solutions - Primary dilutions of each of the individual PCBs listed in Table 3 are prepared by weighing approximately 1-10 mg of material within 1% precision. The PCB is then dissolved and diluted to 1.0 ml with hexane. Calculate the concentration in mg/ml. The primary dilutions are stored at 4°C in screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. Primary dilutions are stable indefinitely if the seals are maintained. The validity of primary and secondary dilutions must be monitored on a quarterly basis by analyzing four quality control check samples (see Section 14.2). 6.4 Working calibration standards - Working calibration standards are prepared that are similar in PCB composition and concentration to the samples by mixing and diluting the individual standard stock solutions. Example calibration solutions are shown in Table 3. The mixture is diluted to volume with pesticide residue analysis quality hexane. The concentration is calculated in ng/ml as the individual PCBs. Dilutions are stored at 4°C in narrow-mouth, screw-cap vials with Teflon cap liners. The meniscus is marked on the vial wall to monitor solvent evaporation. These secondary dilutions can be stored indefinitely if the seals are maintained. These solutions are designated "CSxxx," where the xxx is used to encode the nominal concentration in ng/ml. 6.5 Alternatively, certified stock solutions similar to those listed in Table 3 may be available from a supplier, in lieu of the procedures described in Section 6.4. 6.6 DFTPP standard - A 50-ng/pl solution of DFTPP is prepared in acetone or another appropriate solvent. 6.7 Surrogate standard stock solution - The four 13C-labeled PCBs listed in Table 4 may be available from a supplier as a certified solution. This solution may be used as received or diluted further. These solutions are designated "SSxxx," where the xxx is used to encode the nominal concentration in ng/ml. D-10 �TABLE 3. CONCENTRATIONS OF CONGENERS IN PCS CALIBRATION STANDARDS (ng/ml)a Homolog Congener no. CS1000 CS100 CS050 CS010 1 1 1,040 104 52 10 1 3 1,000 100 50 10 2 7 1,040 104 52 10 3 30 1,040 104 52 10 4 50 1,520 152 76 15 5 97 1,740 174 87 17 6 143 1,920 192 96 19 7 183 2,600 260 130 26 8 202 4,640 464 232 46 9 207 5,060 506 253 51 10 209 4,240 424 212 42 4 255 255 255 255 1 211 (RS) 104 104 104 104 4 212 (RS) 257 257 257 257 8 213 (RS) 407 407 407 407 10 a 210 (IS) 214 (RS) 502 502 502 502 Concentrations given as examples only. D-ll �TABLE 4. COMPOSITION OF SURROGATE SPIKING SOLUTION (SS100) CONTAINING 13C-LABELED PCBs3 Congener no. Compound Concentration (Hg/ml) 211 104 212 (13C12)3,3' ,4,4'-tetrachlorobiphenyl 257 213 (13C12)2,2' ,3,3' ,5,5' ,6,6'-octachlorobiphenyl 395 214 a (I1 ,2' ,3' ,4' ,5' ,6'-13C6)4-chlorobiphenyl (13C12)decachlorobiphenyl 502 Concentrations given as examples only. D-12 �6.8 6.9 Solution stability - The calibration standard, surrogate and DFTPP solutions should be checked frequently for stability. These solutions should be replaced after 6 months, or sooner if comparison with quality control check samples indicates compound degradation or concentration change. 6.10 7 .0 Internal standard solution - A solution of de-3,3" ,4,4" -tetrachlorobiphenyl is prepared at a nominal concentration of 1-10 mg/ml in hexane. The solution is further diluted to give a working standard. Quality control check samples will be supplied by the Agency. Calibration 7.1 The gas chroma tograph must meet the minimum operating parameters shown in Tables 5 and 6, daily. If all of the criteria are not met, the analyst must adjust conditions and repeat the test until all criteria are met. 7.2 The mass spectrometer must meet the minimum operating parameters shown in Tables 2, 7, and 8, daily. If all criteria are not met, the analyst must retune the spectrometer and repeat the test until all conditions are met. The PCB response factor (RF ) must be determined using Equat 7-1 for the analyte homologi. where is p RF = response factor of a given PCB isomer A = area of the characteristic ion for the PCB congener P peak M = mass of PCB congener injected (nanograms) A. = area of the characteristic ion for the internal standard peak M. = mass of internal standard injected (nanograms) IS Using the same conditions as for RF , the surrogate response factors (RF ) must be determined usSng Equation 7-2. A x M. IS S where A S = area of the characteristic ion for the surrogate peak MS = mass of surrogate injected (nanograms) Other items are the same as defined in Equation 7-1. D-13 �TABLE 5. OPERATING PARAMETERS FOR CAPILLARY COLUMN GAS CHROMATOGRAPHIC SYSTEM Recommended Parameter Tolerance Gas chromatograph Finnigan 9610 Other Column 15 m x 0.255 mm ID Fused silica Other Liquid phase DB-5 Other nonpolar or semipolar Liquid phase thickness 0.25 urn < 1 |Jm Carrier gas Helium Hydrogen Carrier gas velocity 45 cm/sec Optimum performance Injector On-column (J&W) (J&W) Injector temperature Optimum performance Injection volume Other c 1.0 plc Optimum performance Other d Initial column temperature 70°C (2 min) Column temperature program 70°-325°C at 10°C/min£ Separator None Glass jet or other Transfer line temperature 280°C Optimum** Tailing factor 0.7-1.5 0.4-3 Peak width 7-10 sec < 15 sec Other Other a Substitutions permitted with any common apparatus or technique provided performance criteria are met. b Measured by injection of air or methane at 270°C oven temperature. c For on-column injection, manufacturer's instructions should be followed regarding injection technique. d With on-column injection, initial temperature equals boiling point of the solvent; in this instance, hexane. e C^Clio elutes at 270°C. Programming above this temperature ensures a clean column and lower background on subsequent runs. f Fused silica columns may be routed directly into the ion source to prevent separator discrimination and losses. g High enough to elute all PCBs, but not high enough to degrade the column if routed through the transfer line. h Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCB congeners in solution CSxxx. i Peak width at 10% height for a single PCB congener is CSxxx. D-14 �TABLE 6. OPERATING PARAMETERS FOR PACKED COLUMN GAS CHROMATOGRAPHY SYSTEM Tolerances Recommended Parameter Gas chromatograph Finnigan 9610 Other3 Column 180 cm x 0.2 cm ID glass Other Column packing 3% SP-2250 on 100/ 120 mesh Supelcoport Other nonpolar or semipolar Carrier gas Helium Hydrogen Carrier gas flow rate 30 ml/min Optimum performance Injector On-column Injector temperature 250°C Optimum Injection volume 1.0 pi g 5 pi Initial column temperature 150°C, 4 min Other Column temperature program 150°C-260° at 8°/min Other Separator Glass jet Other Transfer line temperature 280°C Optimum3 Tailing factor 0.7-1.5 0.4-3 Peak widthd 10-20 sec < 30 sec p a Substitutions permitted if performance criteria are met. b High enough to elute all PCBs. c Tailing factor is width of front half of peak at 10% height divided by width of back half of peak at 10% height for single PCS congeners in solution CSxxx. d Peak width at 10% height for a single PCB congener in CSxxx. D-15 �TABLE 7. OPERATING PARAMETERS FOR QUADRUPOLE MASS SPECTROMETER SYSTEM Parameter Recommended Tolerance Mass spectrometer Finnigan 4023 Other3 Data system Incos 2400 Other Scan range 95-550 Other Scan time 1 sec Otherb Resolution Unit Optimum performance Ion source temperature 280°C 200°-300°C Electron energy 70 eV Optimum performance Trap current 0.2 mA Optimum performance Multiplier voltage -1,600 V Optimum performance Preamplifier sensitivity 10"6 A/V Set for desired working range a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. D-16 �TABLE 8. OPERATING PARAMETERS FOR MAGNETIC SECTOR MASS SPECTROMETER SYSTEM Parameter Tolerance Recommended Mass spectrometer Finnigan MAT 311A Other3 Data system Incos 2400 Other Scan range 98-550 Other Scan mode Exponential Other Cycle time 1.2 sec Otherb Resolution 1,000 > 500 Ion source temperature 280°C 250°-300°C Electron energy 70 eV 70 eV Emission current 1-2 mA Optimum Filament current Optimum Optimum Multiplier -1,600 V Optimum a Substitutions permitted if performance criteria are met. b Greater than five data points over a GC peak is a minimum. c Filaments should be shut off 'during solvent elution to improve instrument stability and prolong filament life, especially if no separator is used. D-17 �If specific congeners are known to be present and if standards are available, selected RF values may be employed. For general samples, solutions CSxxx and SSxxx or a mixture (Tables 3 and 4), with a similar level of internal standard (dg-3,3',4,4*-tetrachlorobiphenyl) added, may be used as the response factor solution. The PCB-surrogate pairs to be used in the RF calculation are listed in Table 9. Generally, only the primary ions of both the analyte and surrogate are used to determine the RF values. If alternate ions are to be used in the quantitation, the RF must be determined using that characteristic ion. The RF value must be determined in a manner to assure ±20% accuracy and precision. For instruments with good day-to-day precision, a running mean (RF) based on seven values determined once each day may be appropriate. Other options include, but are not limited to, triplicate determinations of a single concentration spaced throughout a day or determination of the RF at three different levels to establish a working curve. If replicate RF values differ by greater than ±10% RSD, the system performance should be monitored closely. If the RSD is greater than ±20%, the data set must be considered invalid and the RF redetermined before further analyses are done. 7.4 7.5 8.0 If the GC/EIMS system has not been demonstrated to yield a linear response or if the analyte concentrations are more than two orders of magnitude different from those in the RF solution, a calibration curve must be prepared. If the analyte and RF solution concentrations differ by more than one order of magnitude, a calibration curve should be prepared. A calibration curve should be established with triplicate determinations at three or more concentrations bracketing the analyte levels. The relative retention time (RRT) windows for the 10 homologs and surrogates must be determined. If all congeners are not available, a mixture of available congeners or an Aroclor mixture (e.g., 1016/1254/1260) may be used to estimate the windows. The windows must be set wider than observed if all isomers are not determined. Typical RRT windows for one column are listed in Table 10. The windows may differ substantially if other GC parameters are used. Sample Collection, Handling, and Preservation 8.1 Amber glass sample containers should have Teflon-lined screw caps. With noncorrosive samples, methylene chloride-washed aluminum foil liners may be substituted. The volume is determined by the amount of sample to be collected but will usually be 1 liter or 1 qt. The sample size is dependent on the anticipated PCB levels and difficulty of the subsequent extraction/cleanup steps. D-18 �TABLE 9. Analyte Congener no . 1 2,3 PAIRINGS OF ANALYTE, CALIBRATION, AND SURROGATE COMPOUNDS Compound 2-C12H9Cl 3- and 4-C12H9Cl 1 3 /•* TT r>~t Li2ngLl2 -f ' i £. on 1O~ jy r* TJ r*i L. 12^7 *•*-!- 3 An_fti HU O 1 p l2^6*->-'-4 u n L. 82-127 128-169 C12H5C15 C12H4C16 1"7A1OO 1/0-19J r* ur*! Li2h.3Ll7 194-205 206-208 209 C12H2C18 C12HC19 C12C110 Compound Congener no. 2 4 2,4 2,4, 6 2,2' ,4 ,6 2,2' ,3 ',4,5 2,2' ,3 ,4,5 ,6' 2,2' ,3 ',4,4', 5', 6 2,2' ,3 ,3',5, 5', 6, 6' J 2,2' ,3 ,3'» 4, 4', 5, 6, 6' /"* f L12Lli 0 211 211 211 212 212 212 212 213 213 213 214 Congener no. 4— 1 r 1j O 30 50 97 143 183 202 207 209 v£> a Surrogate Calibration standard Ballschmiter numbering system, see Table 1. Compound 13 C6-4 C6-4 13 C6-4 13 Ci2-3,3' ,4 ,4' 13 Ci2-3,3' ,4 ,4' 13 C12-3,3' ,4 ,4' 13 Ci2-3,3' ,4 ,4' 13 ,6,6' C12-2,2' ,3 ,3', 5, 5' 13 C12-2,2' ,3 ,3' , J V J 3 5 , ,6,6' 13 ,6,6' C12-2,2' ,3 ,3', 5, 5' 13 13 c12ci10 �TABLE 10. PCB homolog RELATIVE RETENTION TIME (RRT) RANGES OF PCB HOMOLOGS VERSUS d6-3,3',4.4'-TETRACHLOROBIPHENYL No. of isomers measured Observed range of RRTs3 Calibration solution Congener Observed RRT3 no. Projected range of RRTs 3 0.40-0.50 1 3 0.43 0.50 0.35-0.55 10 0.52-0.69 7 0.58 0.35-0.80 9 0.62-0.79 30 0.65 0.35-1.10 Tetrachloro 16 0.72-1.01 50 0.75 0.55-1.05 Pentachloro 12 0.82-1.08 97 0.98 0.80-1.10 Hexachloro 13 0.93-1.20 143 1.05 0.90-1.25 Heptachloro 4 1.09-1.30 183 1.15 1.05-1.35 Octachloro 6 1.19-1.36 202 1.19 1.10-1.50 Nonachloro 3 1.31-1.42 207 1.33 1.25-1.50 Decachloro 1 1.44-1.45 209 1.44 1.35-1.50 Monochloro Dichloro Trichloro a The RRTs of the 77 congeners and a mixture of Aroclor 1016/1254/1260 were measured versus 3,3',4,4'-tetrachlorobiphenyl-d6 (internal standard) using a 15-m J&W DB-5 fused silica column with a temperature program of 110°C for 2 min, then 10°C/min to 325°C, helium carrier at 45 cm/sec, and an oncolumn injector. A Finnigan 4023 Incos quadrupole mass spectrometer operating with a scan range of 95-550 daltons was used to detect each PCB congener. b The projected relative retention windows account for overlap of eluting homologs and take into consideration differences in operating systems and lack of all possible 209 PCB congeners. D-20 �8.2 Sample bottle preparation 8.2.1 8.2.2 Sample bottles are heated to 400°C for 15 to 20 min or rinsed with pesticide grade acetone or hexane and allowed to air dry. 8.2.3 8.3 All sample bottles and caps should be washed in detergent solution, rinsed with tap water and then with distilled water. The bottles and caps are allowed to drain dry in a contaminant-free area. Then the caps are rinsed with pesticide grade hexane and allow to air dry. The clean bottles are stored inverted or sealed until use. Sample collection 8.3.1 8.3.2 If possible, mix the source thoroughly before collecting the sample. If mixing is impractical, the sample should be collected from a representative area of the source. If the liquid is flowing through an enclosed system, sampling through a valve should be randomly timed. 8.3.3 8.4 The primary consideration in sample collection is that the sample collected be representative of the whole. Therefore, sampling plans or protocols for each individual producer's situation will have to be developed. The recommendations presented here describe general situations. The number of replicates and sampling frequency also must be planned prior to sampling. Fill the bottle with water, add preservative (Section 8.4), cap tightly, and shake well. To prevent the caps from working loose during storage tape the caps on with a water-insoluble tape. Sample preservation - Samples should be stored at 4°C. Since there is a possibility of microbial degradation, addition of H2S04 during collection to a pH < 2 is recommended. A test strip is used to monitor the pH. Storage times in excess of 4 weeks are not recommended. If residual chlorine is present in the sample, it should be quenched with sodium thiosulfate. EPA Methods 330.4 and 330.5 may be used to measure the residual chlorine.3 Field test kits are available for this purpose. 9.0 Sample Preparation 9.1 Sample homogenization and subsampling - The sample is homogenized by shaking, blending, or other appropriate mechanical technique, if necessary. If the density of the sample is not between 0.9 D-21 �and 1.1, the density should be determined and reported. Consideration should be given to treating the sample as a product waste (see separate protocol). Note: 9.2 Surrogate addition - An appropriate volume of surrogate solution SSxxx is pipetted into the sample. The final concentration of the surrogates must be in the working range of the calibration and well above the matrix background. Note: 9.3 The precision of the mass determination at this step will be reflected in the overall method precision. Therefore, an analytical balance must be used to assure that the weight is accurate to ±1% or better. The volume measurement of cal to the overall method exercise caution that the better. Where necessary, recommended. the spiking solution is critiprecision. The analyst must volume is known to ±1% or calibration of the pipet is Sample preparation (extraction/cleanup) - After addition of the surrogates, the sample is further treated at the discretion of the analyst, provided that the GC/EIMS response of the four surrogates meets the criteria listed in Section 7.0. The literature pertaining to these techniques has been reviewed.4 Several possible techniques are presented below for guidance only. The applicability of any of these techniques to a specific sample matrix must be determined by the precision and accuracy of the *3C PCB surrogate recoveries, as discussed in Section 14.2. 9.3.1 Extraction - The entire sample must be transferred to the extraction vessel with PCB-free water washing, if necessary, to transfer all solids. The container is then rinsed with the extraction solvent to recovery any PCBs adhering to' the container wall. The solvent rinses are combined with the extracts from below. Measure the sample volume to the nearest 0.5%. 9.3.1.1 Liquid-liquid extraction - The solvent, number of extractions, solvent-to-sample ratio, and other parameters are chosen at the analyst's discretion. A suggested extraction from water is presented in EPA Methods 60S1 and 625.2 9.3.1.2 Sorbent column extraction - PCBs may be isolated from water onto sorbent columns, although these techniques are not as widely used or thoroughly validated as liquid-liquid extraction. The selection of sorbent (XAD, Porapak, carbonpolyurethane foam, etc.) will depend on the nature of the matrix. The available methods have been reviewed.4 D-22 �9.3.2 Cleanup - Several tested cleanup techniques are described below. All but the base cleanup (9.3.2.8) were previously validated for PCBs in transformer fluids.5 Depending upon the complexity of the sample, one or more of the techniques may be required to fractionate the PCBs from interferences. For most cleanups a concentrated (1-5 ml) extract should be used. 9.3.2.1 Acid cleanup 9.3.2.1.1 Place 5 ml of concentrated sulfuric acid into a 40-ml narrow-mouth screwcap bottle. Add the sample extract. Seal the bottle with a Teflon-lined screw cap and shake for 1 min. 9.3.2.1.2 Allow the phases to separate, transfer the sample (upper phase) with three rinses of 1-2 ml solvent to a clean container and concentrate to an appropriate volume. 9.3.2.1.3 Analyze as described in Section 10.0. 9.3.2.1.4 If the sample is highly contaminated, a second or third acid cleanup may be employed. 9.3.2.2 Florisil column cleanup 9.3.2.2.1 Variations among batches of Florisil (PR grade or equivalent) may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all of the PCBs must be verified by the analyst. The weight of Florisil can then be adjusted accordingly. 9.3.2.2.2 Place a 20-g charge of Florisil, activated overnight at 130°C, into a Chromaflex column. Settle the Florisil by tapping the column. Add about 1 cm of anhydrous sodium sulfate to the top of the Florisil. Pre-elute the column with 70-80 ml of hexane. Just before the exposure of the sodium sulfate layer to air, stop the flow. Discard the eluate. D-23 �9.3.2.2.3 Add the sample extract to the column. 9.3.2.2.4 Carefully wash down the inner wall of the column with 5 ml of the hexane. 9.3.2.2.5 Add 220 ml of hexane to the column. 9.3.2.2.6 Discard the first 25 ml. 9.3.2.2.7 Collect 200 ml Kuderna-Danish PCBs should be Concentrate to of hexane eluate in a flask. All of the in this fraction. an appropriate volume. 9.3.2.2.8 Analyze the sample as described in Section 10.0. 9.3.2.3 Alumina column cleanup 9.3.2.3.1 Adjust the activity of the alumina (Fisher A540 or equivalent) by heating to 200°C for 2 to 4 hr. When cool, add 3% water (wt:wt) and mix until uniform. Store in a tightly sealed bottle. Allow the deactivated alumina to equilibrate at least 1/2 hr before use. Reactivate weekly. 9.3.2.3.2 Variations between batches of alumina may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all of the PCBs must be verified by the analyst. The weight of alumina can then be adjusted accordingly. 9.3.2.3.3 Place a 50-g charge of alumina into a Chromaflex column. Settle the alumina by tapping. Add about 1 cm of anhydrous sodium sulfate. Pre-elute the column with 70-80 ml of hexane. Just before exposure of the sodium sulfate layer to air, stop the flow. Discard the eluate. 9.3.2.3.4 Add the sample extract to the column. 9.3.2.3.5 D-24 Carefully wash down the inner wall of the column with 5 ml volume of hexane. �9.3.2.3.6 Add 295 ml of hexane to the column. 9.3.2.3.7 Discard the first 50 ml. 9.3.2.3.8 Collect 250 ml Kuderna-Danish PCBs should be Concentrate to of the hexane in a flask. All of the in this fraction. an appropriate volume. 9.3.2.3.9 Analyze the sample as described in Section 10.0. 9.3.2.4 Silica gel column cleanup 9.3.2.4.1 Activate silica gel (Davison grade 950 or equivalent) at 135°C overnight. 9.3.2.4.2 Variations between batches of silica gel may affect the elution volume of the various PCBs. For this reason, the volume of solvent required to completely elute all of the PCBs must be verified by the analyst. The weight of silica gel can then be adjusted accordingly. 9.3.2.4.3 Place a 25-g charge of activated silica gel into a Chromaflex column. Settle the silica gel by tapping the column. Add about 1 cm of anhydrous sodium sulfate to the top of the silica gel. 9.3.2.4.4 Pre-elute the column with 70-80 ml of hexane. Discard the eluate. Just before exposing the sodium sulfate layer to air, stop the flow. 9.3.2.4.5 Add the sample extract to the column. 9.3.2.4.6 Wash down the inner wall of the column with 5 ml of hexane. 9.3.2.4.7 Elute the PCBs with 195 ml of 10% diethyl ether in hexane (v:v). 9.3.2.4.8 Collect 200 ml Kuderna-Danish PCBs should be Concentrate to D-25 of the eluate in a flask. All of the in this fraction. an appropriate volume. �9.3.2.4.9 9.3.2.5 Analyze the sample according to Section 10.0. Gel permeation cleanup 9.3.2.5.1 Set up and calibrate the gel permeation chromatograph with an SX-3 column according to the Autoprep instruction manual. Use 15% methylene chloride in cyclohexane (v:v) as the mobile phase. 9.3.2.5.2 Inject 5.0 ml of the sample extract into the instrument. Collect the fraction containing the PCBs (see Autoprep operator's manual) in a Kuderna-Danish flask equipped with a 10-ml ampul. 9.3.2.5.3 Concentrate the PCB fraction to an appropriate volume. 9.3.2.5.4 Analyze as described in Section 10.0. 9.3.2.6 Acetonitrile partition 9.3.2.6.1 Place the sample extract into a 125-ml separatory funnel with enough hexane to bring the final volume to 15 ml. Extract the sample four times by shaking vigorously for 1 min with 30-ml portions of hexane-saturated acetonitrile. 9.3.2.6.2 Combine and transfer the acetonitrile phases to a 1-liter separatory funnel and add 650 ml of distilled water and 40 ml of saturated sodium chloride solution. Mix thoroughly for about 30 sec. Extract with two 100-ml portions of hexane by vigorously shaking about 15 sec. 9.3.2.6.3 Combine the hexane extracts in a 1-liter separatory funnel and wash with two 100-ml portions of distilled water. Discard the water layer and pour the hexane layer through a 8-10 cm anhydrous sodium sulfate column into a 500-ml Kuderna-Danish flask equipped with a 10-ml ampul. Rinse the separatory funnel and column with three 10-ml portions of hexane. D-26 �9.3.2.6.4 Concentrate the extracts to an appropriate volume. 9.3.2.6.5 Analyze as described in Section 10.0. 9.3.2.7 Florisil slurry cleanup 9.3.2.7.1 Place the sample extract into a 20-ml narrow-mouth screw-cap container. Add 0.25 g of Florisil (PR grade or equivalent). Seal with a Teflon-lined screw cap and shake for 1 min. 9.3.2.7.2 Allow the Florisil to settle; then decant the treated solution into a second container with rinsing. Concentrate the sample to an appropriate volume. Analyze as described in Section 10.0. 9.3.2.8 Base cleanup6 9.3.2.8.1 Quantitatively transfer the concentrated extract to a 125-ml extraction flask with the aid of several small portions of solvent. 9.3.2.8.2 Evaporate the extract just to dryness with a gentle stream of dry filtered nitrogen, and add 25 ml of 2.5% alcoholic KOH. 9.3.2.8.3 Add a boiling chip, put a water condenser in place, and allow the solution to reflux on a hot plate for 45 min. 9.8.2.8.4 After cooling, transfer the solution to a 250-ml separatory funnel with 25 ml of distilled water. 9.3.2.8.5 Rinse the extraction flask with 25 ml of hexane and add it to the separatory funnel. 9.3.2.8.6 D-27 Stopper the separatory funnel and shake vigorously for at least 1 min. Allow the layers to separate and transfer the lower aqueous phase to a second separatory funnel. �9.3.2.8.7 Extract the saponification solution with a second 25-ml portion of hexane. After the layers have separated, add the first hexane extract to the second separatory funnel and transfer the aqueous alcohol layer to the original separatory funnel. 9.3.2.8.8 Repeat the extraction with a third 25-ml portion of hexane. Discard the saponification solution, and combine the hexane extracts. 9.3.2.8.9 10.0 Concentrate the hexane layer to an appropriate volume and analyze according to Section 10.0. Gas Chrotnatographic/Electron Impact Mass Spectrometric Determination 10.1 Internal standard addition - An appropriate volume of the internal standard solution is pipetted into the sample. The final concentration of the internal standard must be in the working range of the calibration and well above the matrix background. The internal standard is thoroughly incorporated by mechanical agitation. Note: The volumetric measurement of the internal standard solution is critical to the overall method precision. The analyst must exercise caution that the volume is known to be ±1% or better. Where necessary, calibration of the pipet is recommended. 10.2 Tables 2, and 5 through 8 summarize the recommended operating conditions for analysis. Figure 1 presents an example of a chromatogram. 10.3 While the highest available chromatographic resolution is not a necessary objective of this protocol, good chromatographic performance is recommended. With the high resolution of CGC, the probability that the chromatographic peaks consist of single compounds is higher than with PGC. Thus, qualitative and quantitative data reduction should be more reliable. 10.4 After performance of the system has been certified for the day and all instrument conditions set according to Tables 2, and 5 through 8, inject an aliquot of the sample onto the GC column. If the response for any ion, including surrogates and internal standards, exceeds the working range of the system, dilute the sample and reanalyze. If the responses of surrogates, analyte, or internal standard are below the working range, recheck the system performance. If necessary, concentrate the sample and reanalyze. D-28 �tH CO cd C 0) 4J fi C Ol 0 Xi X* '""•*• i—4 f~! CJ x; x: o CJ CJ O r-l C cd M r-l O C 3 o o to rH 0 01 X! p. •H p. _r\ O M 0 •H Xi o M o rH Xi CJ O o a •rt ;s CN 1 "^ o (~) CN c <r XI O 0) *• Xi P. CO t-l o tH x: o rH iH ho rH -» ^» •* •nH G/ 5 VO 01 H 1 •- ^ C 01 X! P. <r -* 0* 01 r; 4J iH a s^ 1 1 a i C 01 X! P. •H 523 575 1 JL. GS3 10:0!) X! 11 - rH >. 0 x; o rH r^) 4-1 C 01 01 H XI 1 P. -H ^3 » 0 r-l " O " O ^ •H XI rH 01 pr* | Xi ~ o 4-1 01 0 o H 1 vo r-l O rH X! CJ •rH r-l H t Cd vO CO * if) 4-1 K C •s CO -3" 01 CN 1 CNI *. ~3~ CN - CN vO -H Xi O M M 0 rH Xi o t) <T X! P. P> •H X I rH v^- <U r*, P. -H XI 0 CO r-l C C Ol ~* .d rH O £>-. C ) ^ C 01 o x ; iH Xi CJ cfl cd O H2592. CN rH 0 CO iH 1 rH 5*"\ Xi CO O Vi ^0 p. ^ P . -H •H Q) XI 4-1 O cfl M 00 O O Xi 0 O a 1 rH .C rH Xi l-l r-l O 1 •" vo ** O CO CJ 0 3 CO 0) r! vo iO CJ - P. -H XI O M O 4J CJ vO vO rH O ^~- Ol O Ol Q cfl 4j in p. 01 *n ^ _" -* -d~ vO CO CO *v •• m *. «\ co co •• • X V CN ~3~ " CN »v CN *i CN -* co n ^ I /!v/U CNJ iiao I31G •t CN 1 "^ CN 07-1 CO -" 7-W V i—i rH cO r4 613 iii" Ji r-H(V,*< O O 0) M M 4-1 O O cfl tH rH 00 me. G 0) Xi P. •H XI 0 01 P. P. •H -H - M 1 rH C 01 Xi s>* rH ^ CO T3 1 tH rH o VO 0 CO tH 1 ^W -' vO r-l 3 V 198.0- oo 0 CN" RL C'59 13:20 [I 936 llfrl 1 fl | i JLJL_ IOSO 16:10 1178 l | |m J[ .L i l'J{-9 20: CD ' 1 I1';9 23:20 SCA'l Tllh: Figure 1. Capillary gas chromatography/electron impact ionization mass spectrometry (CGC/EIMS) chromatogram or the calibration standard solution required for quantitation of PCBs by homolog. This chromatogram includes PCBs representative of each homolog, three carbon-13 labeled surrogates, and the deuterated internal standard. The concentration of all components and the CGC/EIMS parameters are presented in Tables 3, 4, 5, and 7. �10.5 11.0 Record all data on a digital storage device (magnetic disk, tape, etc.) for qualitative and quantitative data reduction as discussed below. Qualitative Identification 11.1 Selected ion monitoring (SIM) or limited mass scan (IMS) data The identification of a compound as a given PCB homolog requires that two criteria be met: 11.1.1 11.1.2 11.2 (1) The peak must elute within the retention time window set for that homolog (Section 7.5); and (2) the ratio of two ions obtained by SIM (Table 11) or by LMS (Table 12) must match the natural ratio within ±20%. The analyst must search the higher mass windows, in particular M+70, to prevent misidentification of a PCB fragment ion cluster as the parent. If one or the other of these criteria is not met, interferences may have affected the results and a reanalysis using full scan EIMS conditions is recommended. Full scan data 11.2.1 11.2.2 12.0 The unknown spectrum must match that of an authentic PCB. The intensity of the three largest ions in the molecular cluster (two largest for monochlorobiphenyls) must match the natural ratio within ±20%. Fragment clusters with proper intensity ratios must also be present. 11.2.3 11.3 The peak must elute within the retention time windows set for that homolog (as described in Section 7.5). Alternatively, a spectral search may be used to automatically reduce the data. The criteria for acceptable identification include a high index of similarity. For the Incos 2300, a fit of 750 or greater must be obtained. Disputes in interpretation - Where there is reasonable doubt as to the identity of a peak as a PCB, the analyst must either identify the peak as a PCB or proceed to a confirmational analysis (see Section 13.0). Quantitative Data Reduction 12.1 Once a chromatographic peak has been identified as a PCB, the compound is quantitated based either on the integrated abundance of the SIM data or EICP for the primary characteristic ion in Tables 11 and 12. If interferences are observed for the primary ion, D-30 �TABLE 11. CHARACTERISTIC SIM IONS FOR PCBs Ion (relative intensity) Tertiary Secondary Homolog Primary Cj^HgCl 188 (100) 190 (33) - Ci2HgCl2 222 (100) 224 (66) 226 (11) C^HrCls 256 (100) 258 (99) 260 (33) Ci2HeCl4 292 (100) 290 (76) 294 (49) C12H5C15 326 (100) 328 (66) 324 (61) c i2H4Cle 360 (100) 362 (82) 364 (36) C12H3C17 394 (100) 396 (98) 398 (54) Ci^HfcClg 430 (100) 432 (66) 428 (87) Ci2HCl9 464 (100) 466 (76) 462 (76) CiaCljo 498 (100) 500 (87) 496 (68) Source: Rote, J. W., and W. J. Morris, "Use of Isotopic Abundance Ratios in Identification of Polychlorinated Biphenyls by Mass Spectrometry," J. Assoc. Offic. Anal. Chem., 56(1), 188-199 (1973). D-31 �TABLE 12. LIMITED MASS SCANNING (LMS) RANGES FOR PCBs Compound Mass range (m/z) Ci2H9Cli 186-190 Ci2H8Cl2 220-226 Cl2H?Cl3 254-260 Ci2H6Cl3 288-294 C12H5C15 322-328 C12H4C16 356-364 C12H3C17 386-400 Ci2H2Clg 426-434 Cl2HClg 460-468 Ci2Clio 494-504 C12D6C14 294-300 13 192-196 C612CeH9Cl 13 300-306 13 C12H2C18 438-446 13 C12Clio 506-516 C12H6C14 a Adapted from Tindall, G. W., and P. E. Wininger, "Gas Chromatography-Mass Spectrometry Method for Identifying and Determining Polychlorinated Biphenyls," J. Chromatogr., 196, 109-119 (1980). D-32 �use the secondary and then tertiary ion for quantitation. If interferences in the parent cluster prevent quantitation, an ion from a fragment cluster (e.g., M-70) may be used. Whichever ion is used, the RF must be determined using that ion. The same criteria should be applied to the surrogate compounds (Table 13). 12.2 Using the appropriate analyte-internal standard pair and response factor (RF ) as determined in Section 7.3, calculate the concentration ofpeach peak using Equation 12-1. A M. V Eq 12-1 Concentration (Mg/g) = ^ ' RF ' M^ ' \T ' is p e i where A = area of the characteristic ion for the analyte PCB peak A. = area of the characteristic ion for the internal standard peak RF = response factor of a given PCB congener M. = mass of internal standard injected (micrograms) 1S M = mass of sample extracted (grams) V. = volume injected (microliters) V 12.3 = volume of sample extract (microliters) If a peak appears to contain non-PCB interferences which cannot be circumvented by a secondary or tertiary ion, either: 12.3.1 12.3.2 Perform additional chemical cleanup (Section 9) and then reanalyze the sample; or 12.3.3 12.4 Reanalyze the sample on a different column which separates the PCB and interferents; Quantitate the entire peak as PCB. Calculate the recovery of the four 13C surrogates using the appropriate surrogate-internal standard pair and response factor (RF. & as determined in Section 7.4 using Equation 12-2. ) 1 A M. Recovery ( ) = j*- - ^- • ^ • 100 % Eq. 12-2 is s s where A S = area of the characteristic ion for the surrogate peak A. = area . the characteristic ion for the internal standard of IS peak D-33 �TABLE 13. CHARACTERISTIC IONS FOR 13C-LABELED PCS SURROGATES Primary Ion (relative intensity) Secondary 13 194 (100) 196 (33) l3 304 (100) 306 (49) 302 (78) 13 C12H2C18 442 (100) 444 (65) 440 (89) 13 Ci2Clio 510 (100) 512 (87) 514 (50) Specific compound C612C6H9C1 Ci2H6Cl4 D-34 Tertiary �RF = response factor for the surrogate compound with respect to the internal standard (Equation 7-2) M. = mass of internal standard injected (nanograms) is M s = mass of surrogate, assuming 100% recovery (nanograms) 12.5 Correct the concentration of each peak using Equation 12-3. is the final reportable concentration. Corrected concentration (pg/g) = Concentration ug/g .10Q ^re>i */ 12.6 Recovery ( ) % £ This _3 ^ 12 Sum all of the peaks for each homolog, and then sum those to yield the total PCB concentration in the sample. Report all numbers in |Jg/g. The reporting form in Table 14 may be used. If an alternate reporting format (e.g., concentration per peak) is desired, a different report form may be used. The uncorrected concentrations, percent recovery, and corrected recovery are to be reported. 12.7 Round off all numbers reported to two significant figures. 13.0 Confirmation If there is reason to question the qualitative identification (Section 11.0), the analyst may choose to confirm that a peak is not a PCB. Any technique may be chosen provided that it is validated as having equivalent or superior selectivity and sensitivity to GC/EIMS. Some candidate techniques include alternate GC columns (with EIMS detection), GC/CIMS, GC/NCIMS, high resolution EIMS, and MS/MS techniques. Each laboratory must validate confirmation techniques to show equivalent or superior selectivity between PCBs and interferences and sensitivity (limit of quantitation, LOQ). If a peak is confirmed as being a non-PCB, it may be deleted from the calculation (Section 12). If a peak is confirmed as containing both PCB and non-PCB components, it must be quantitated according to Section 12,3. 14.0 Quality Control 14.1 Each laboratory that uses this method must operate a formal quality control (QC) program. The minimum requirements of this program consist of an initial demonstration of laboratory capability and the analysis of spiked samples as a continuing check on performance. The laboratory must maintain performance records to define the quality of data that are generated. After a date specified by the Agency, ongoing performance checks should be compared with established performance criteria to determine if the results of analyses are within accuracy and precision limits expected of the method. D-35 �TABLE 14. ANALYSIS REPORT INCIDENTAL PCBs IN WASTEWATER Sample No. Sample Matrix Sample Source Notebook No. or File Location Volume Extracted Extraction/Cleanup Procedure Int. Std. liter Mass Added (|Jg) (Circle one) 4-Cl(d6) Surrogates 298 Mass Added ()Jg) 300 Intensity 100/49 (Circle one) Ratio 1-C1 194 196 100/33 4-C1 304 306 100/49 8-C1 442 444 100/65 10-C1 510 512 100/87 (continued) D-36 Ratio Intensity % Recovery �TABLE 14 (continued) Qualitative I Analyte 1° 2° l° 1-C1 188 190 100/33 2-C1 222 224 100/66 3-C1 256 258 100/99 4-C1 292 290 100/76 5-C1 326 328 100/66 6-C1 360 362 100/82 7-C1 394 396 100/98 8-C1 430 432 100/66 9-C1 464 466 100/76 10-C1 498 500 Quantitative Uncorr Corr Ion Cone. Cone OK? Used RF (pg/4) (M8/4) 100/87 2° Ratio Theoretical Total MS/4 Uncorr. Reported by: Internal Audit: Name Name EPA Audit: Name Signature/Date Signature/Date Signature/Date Organization Organization Organization D-37 M8/4 Corr. �14.2 The analysts must certify that the precision and accuracy of the analytical results are acceptable by: 14.2.1 14.2.2 14.3 The absolute precision of surrogate recovery, measured as the RSD of the integrated EIMS area (Ag) for a set of samples, must be ±10%. The mean recovery (R ) of at least four replicates of a QC check sample to be supplied by the Agency must meet Agency-specified accuracy and precision criteria. This forms the initial data base for establishing control limits (see Section 14.3 below). Control limits - The laboratory must establish control limits using the following equations: Upper control limit (UCL) = RL. + 3 RSD V_ Upper warning limit (UWL) = R + 2 RSD Lower warning limit (LWL) = R - 2 RSD Lower control limit (LCL) = R - 3 RSD These may be plotted on control charts. If an analysis of a check sample falls outside the warning limits, the analyst should be alerted that potential problems may need correction. If the results for a check sample fall outside the control limits, the laboratory must take corrective action and recertify the performance (Section 14.2) before proceeding with analyses. The warning and control limits should be continuously updated as more check sample replicates are added to the data base. 14.4 Before processing any samples, the analyst should demonstrate through the analysis of a reagent blank that all glassware and reagent interferences are under control. Each time a set of samples is analyzed or there is a change in reagents, a laboratory reagent blank should be processed as a safeguard against contamination. 14.5 Procedural QC - The various steps of the analytical procedure should have quality control measures. These include but are not limited to: 14.5.1 GC performance - See Section 7.1 for performance criteria. 14.5.2 MS performance - See Section 7.2 for performance criteria. D-38 �14.5.3 Qualitative identification - At least 10% of the PCB identifications, as well as any questionable results, should be confirmed by a second mass spectrometrist. 14.5.4 Quantitation - At least 10% of all manual calculations, including peak area calculations, must be checked. After changes in computer quantitation routines, the results should be manually checked. 14.6 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, selected at random, must be run in triplicate to monitor the precision of the analysis. An RSD of ±30% or less must be achieved. If the precision is greater than ±30%, the analyst must be recertified (see Section 14.2). 14.7 A minimum of 10% of all samples, one sample per month or one sample per matrix type, whichever is greater, selected at random, must be analyzed by the standard addition technique. Two aliquots of the sample are analyzed, one "as is" and one spiked (surrogate spiking and equilibration techniques are described in Section 9.2) with a sufficient amount of Solution CSxxx to yield approximately 100 (Jg/liter of each compound). The samples are analyzed together and the quantitative results calculated. The recovery of the spiked compounds (calculated by difference) must be 80-120%. If the sample is known to contain specific PCB isomers, these isomers may be substituted for solution CSxxx. If the concentrations of PCBs are known to be high or low, the amount added should be adjusted so that the spiking level is 1.5 to 4 times the measured PCB level in the unspiked sample. 14.8 Interlaboratory comparison - Interlaboratory comparison studies are planned. Participation requirements, level of performance, and the identity of the coordinating laboratory will be presented in later revisions. 14.9 It is recommended that the participating laboratory adopt additional QC practices for use with this method. The specific practices that are most productive depend upon the needs of the laboratory and the nature of the samples. Field duplicates or triplicates may be analyzed to monitor the precision of the sampling technique. Whenever possible, the laboratory should perform analysis of standard reference materials and participate in relevant performance evaluation studies. 15.0 Quality Assurance Each participating laboratory must develop a quality assurance plan according to EPA guidelines.7 The quality assurance plan must be submitted to the Agency for approval. D-39 �16.0 Method Performance The method performance is being evaluated. Limits of quantitation; average intralaboratory recoveries, precision, and accuracy; and interlaboratory recoveries, precision, and accuracy will be presented. 17.0 Documentation and Records Each laboratory is responsible for maintaining full records of the analysis. Laboratory notebooks should be used for handwritten records. GC/MS data must be archived on magnetic tape, disk, or a similar device. Hard copy printouts may be kept in addition if desired. QC records should be maintained separately from sample analysis records. The documentation must describe completely how the analysis was performed. Any variances from the protocol must be noted and fully described. Where the protocol lists options (e.g., sample cleanup), the option used and specifics (solvent volumes, digestion times, etc.) must be stated. D-40 �REFERENCES 1. Environmental Protection Agency, Organochlorine Pesticides and PCBs— Method 608," Fed. Reg., 44, 69501-69509 (December 3, 1979). 2. Environmental Protection Agency, "Base/Neutrals, Acids, and Pesticides— Method 625," Fed. Reg., 44, 69540-69552 (December 3, 1979), and subsequent revisions. 3. "Methods 330.4 (Titrimetric, DPD-FAS) and 330.5 (Spectrophotometric, DPD) for Chlorine, Total Residual," Methods for Chemical Analysis of Water and Wastes, U.S. Environmental Protection Agency, Environmental Monitoring and Support Laboratory, Cincinnati, Ohio, March 1979, EPA 600-4/79-020. 4. Erickson, M. D., and J. S. Stanley, "Methods of Analysis for Incidentally Generated PCBs Literature Review and Preliminary Recommendations," Interim Report No. 1, EPA Contract No. 68-01-5915, Task 51, 1982. 5. Bellar, T. A., and J. J. Lichtenberg, "The Determination of Polychlorinated Biphenyls in Transformer Fluid and Waste Oils," Prepared for U.S. Environmental Protection Agency (1981). EPA-600/4-81-045. 6. American Society for Testing and Materials, "Standard Method for Analysis of Environmental Materials for Polychlorinated Biphenyls," pp. 877-885, in Annual Book of ASTM Standards, Part 40, Philadelphia, Pennsylvania (1980). ANSI/ASTM D 3304-77. 7. Quality Assurance Program Plan for the Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C., October 1980. D-41 �TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 3. RECIPIENT'S ACCESSION NO. 2. 1. REPORT NO. EPA-560/5-82-006 4. TITLE AND SUBTITLE 5. REPORT DATE Analytical Methods for By-Product PCBs—Initial Validation and Interim Protocols 6. PERFORMING ORGANIZATION CODE October 11, 1982 7. AUTHORIS) Mitchell D. Erickson, John S. Stanley, Gil Radolovich, Kay Turman, Karin Bauer, Jon Onstot, Donna Rose, and Margaret Wickham 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. P R O G R A M E L E M E N T NO. Midwest Research Institute 425 Volker Boulevard Kansas City, MO 64110 11. CONTRACT/GRANT NO. 12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED MRI Project No. 4901-A51 EPA 68-01-5915, Task 51 U.S. Environmental Protection Agency Office of Toxic Substances, Field Studies Branch TS-798 Washington, DC 20460 Interim 4, 4/24-8/31/82 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES The task manager is David P. Redford; the project officer is Frederick W. Kutz. 16. A B S T R A C T This document presents proposed analytical methods for analysis of by-product PCBs in commercial products, product waste streams, wastewaters, and air. The analytical method for commercial products and product waste streams consist of a flexible approach for extraction and cleanup of particular matrices. The 13c-labeled PCB surrogates are added as part of a strong quality assurance program to determine levels of recovery. The wastewater method is based on EPA Methods 608 and 625 with revisions to include use of the 13c-labeled PCB surrogates. The air method is a revision of a proposed EPA method for the collection and analysis of PCBs in air and flue gas emissions. Capillary or packed column gas chromatography/electron impact ionization mass spectrometry is proposed as the primary instrumental method. Response factors and retention times of 77 PCB congeners relative to tetrachlorobiphenyl-d6 are presented in addition to statistical analysis to project validity of the data and extrapolation of relative response factors to all 209 possible congeners. Preliminary studies using the ISClabeled surrogates to validate specific cleanup procedures and to analyze several commercial products and product wastes indicate that the proposed analytical methods are both feasible and practical. KEY WORDS AND DOCUMENT ANALYSIS 17. DESCRIPTORS Polychlorinated biphenyls PCBs Incidentally generated Analytical protocols Air Wastewater Commercial products b.IDENTIFIERS/OPEN ENDED TERMS Commercial waste Capillary column Electron impact EIMS Response factors Relative retenti Surrogates 18. DISTRIBUTION STATEMENT EPA Form 2220-1 (R»v. 4-77) Unclassified PREVIOUS EDITION is OBSOLETE COSATI Field/Group relative response factor n times 19. SECURITY CLASS (This Report) Unclassified 20. SECURITY CLASS (This page) Unlimited c. streams gas chromatography onization mass spectromet 21. NO. OF PAGES 243 22. PRICE �United States Environmental Protection Agency Office of Toxic Substances Washington DC 20460 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 186 Folder The folder containing the original item. 5385 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Erickson, Mitchell D. John S. Stanley Kay Turman Gil Radolovich Karin Bauer Jon Onstot Donna Rose Margaret Wickham Description An account of the resource <strong>Corporate Author: </strong>Midwest Research Institute Date A point or period of time associated with an event in the lifecycle of the resource October 11 1982 Title A name given to the resource Analytical Methods for By-Product PCBs - Preliminary Validation and Interim Methods ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/098a2e011bbd0b7be7ff3574105bb407.pdf 1575af407a801e9ffcf18974c328f9d9 PDF Text Text Item D Number 05331 Author Esposito, M. P. Corporate Author G Not Scanned PEDCO Environmental, Inc. RODOrt/ArtiClOTitlO Dioxins: Volume I. Sources, Exposure, Transport, and Control Journal/Book Title Year Month/Day Color June D Number of Images ° DBSBripton Notes Volume I of a three-volume series. Contract No. 68-03-2577; EPA-600/2-80-156; Also includes letter from David R. Watkins to Alvin L. Young, August 7,1980 and distribution list. Tuesday, March 05, 2002 Page 5331 of 5363 �UNITED STATES ENVIRONMENTAL PROTECTION AGENCY OFFICE OF RESEARCH AND DEVELOPMENT INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY CINCINNATI, OHIO 45268 AUG 7 1980 Major Alvin L. Young Department of the Air Force School of Air and Space Medicine SAM/EK Brooks AFB, Texas 78235 Dear Major Young: Enclosed are three reports which have been published as a series and are intended to serve as a comprehensive reference on the subject of dioxins. Volume I is a state-of-the-art review of dioxin literature. Detailed information is presented on the chemistry, sources, degradation, transport, disposal, and health effects of dioxins. Accounts of public and occupational exposure to dioxins are also included. Volume II details the development of an analytical method for detecting partper-trillion levels of dioxins in industrial wastes. While this report represents the current developments within the EPA (IERL-CI), there are further advancements that have been made by independent laboratories. This report includes a review of the analytical literature on methods of detecting dioxins in various types of environmental samples. Volume III identifies various routes of information of dioxins. The possible presence of dioxins in basic organic chemicals and pesticides is addressed, and production locations for these materials are identified. These reports have been prepared with the hopes of presenting factual information documented by literature references and contractual research. If there are errors that appear in the reports or conclusions that you feel have been drawn incorrectly from the literature, I would appreciate your comments along with suggested corrections. My telephone number is (513) 684-4481. Sincerely, David R. Watkins Organic and Inorganic Chemicals and Products Branch Industrial Pollution Control Division Enclosures (3) �Distribution: Mr. Mr. Mr. Mr. Mr. Dr. Mr. Dr. Dr. Charlie Auer (TS-792) Don Barnes (TS-788) Howard Beard (WH-565) George G. Berlow (Mass Pub. Hea. Kenneth Biglane (WH-548) Marilyn Bracken (TS-793) Thomas J. Buechler (Region VII) Leo Buffa (Ottawa, Canada) Kathleen Cainin (Region VI) Mr. Allen Carpien (A-132) Dr. Ananda Chakrabarty (U. of 111.) Mr. Gangadhar Choudhary (NIOSH) Dr. Richard Cothern (TS-793) Dr. Warren Crummett (Dow Chen.) Dr. Al Cywin (WH-556) Mr. Mike Dalton (Ohio EPA) Mr. Mike Dellarco (TS-791) Mr. Paul DesRosiers (ORD). Mr. J. P. Dickerson (Australia) Rep. Dennis Dollar (MS) Mr. Joe Duckett (Schwartz & Connolly Law Firm) Ms. Barbara Elkus (EN-335) Mr. Bill Fairless (Region VII) Dr. David Firestone (FDA) • Dr. Silvio Garattini (Milan, It.) Dr. M. E. Gibson (FDA) Mr. Harry Gilmer (Region VII) Dr. Michael Gross (U. of Neb.) Dr. Risto Hakulinen (Finland) Mr. Martin P. Halper (TS-793) Ms. Adlene Harrison (Region VI) Dr. Richard Heffelfinger (Battelle) Mr. Charlie Hiremath (RD-689) Dr. Pat Honchar (NIOSH) Dr. Ken Howard (Vertac, Inc.) Dr. R. Huetter (Switzerland) Dr. Otto Hutzinger (Netherlands) Ms. Daphne Kamely (PM-223) Mr Dr Mr Dr Mr Dr Dr Dr Mr Mr Ms Mr Ms Dr Dr Dr Mr Ms Dr Dr Dr Mr Mr Dr Ms Ms Mr Mike Kilpatrick (EN-335) James J. Lichtenberg (EMSL) Raymond Locke (TS-792) David Ludbetter (DOJ) Ed Martin (WH-565) Eugene P. Meier (Las Vegas) W. Lamar Miller (EN-329) R. K. Mitchum (NCTR) Bob Ogg (Region II) Gordon Olson (TS-794) Dorothy Patton (A-132) Charlie Plost (RD-680) Pat Polk (CDC) A. P. Poland (U. of Rochester) Oscar Ramirez (Region IV) Christopher Rappee (Umen, Switz.) Dave Redford (TS-793) Mary Reese (TS-791) Med. G. Reggiani (Hoffman La Roche) Mirja Salkinoja-Salonen (Finland) F. Schaufelberger (Switzerland) Mark Segal (WH-553) John Smith (TS-793 Gerald Southall (State of Arkansas) Deborah Speak (Region IV) Virginia Steiner (WH-565) Dave Sussman (WH-563) Capt Steve TerMaath (U.S.A.F.) Ron Thomas (EPA - Beltsville) Bruce Thomson (Ontario, Canada) Thomas Tiernan (Wright State) Vittorio Treccani (Milan Italy) Vladmir Vasak (Id - Australia) Al Venosa (MERL) Charles Warren (Region II) Reuben Watkins (MS) Don Wilson (lERL-Ci) Mr Russ Wyer (WH-548) Dr Peter Yates (State Pollution Australia) Maj. Alvin Young (U.S.A.F.) Mr Dr Dr Dr Dr Mr Dr Mr Mr �United States Environmental Protection Agency Industrial Environmental Research Laboratory Cincinnati OH 45268 Research and Development vvEPA Dioxins Volume I. Sources, Exposure, Transport, and Control EPA-600/2-80-156 June 1980 �RESEARCH REPORTING SERIES Research reports of the Office of Research and Development, U.S. Environmental Protection Agency, have been grouped into nine series. These nine broad categories were established to facilitate further development and application of environmental technology. Elimination of traditional grouping was consciously planned to foster technology transfer and a maximum interface in related fields. The nine series are: 1. 2. 3. 4. 5. .6. 7. 8. 9. Environmental Health Effects Research Environmental Protection Technology Ecological Research Environmental Monitoring Socioeconomic Environmental Studies Scientific and Technical Assessment Reports (STAR) Interagency Energy-Environment Research and Development "Special" Reports Miscellaneous Reports This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY series. This series describes research performed to develop and demonstrate instrumentation, equipment, and methodology to repair or prevent environmental degradation from point and non-point sources of pollution. This work provides the new or improved technology required for the control and treatment of pollution sources to meet environmental quality standards. This document is available to the public through the National Technical Information Service, Springfield, Virginia 22161. �EPA-600/2-80-156 June 1980 DIOXINS: VOLUME I. SOURCES, EXPOSURE, TRANSPORT, AND CONTROL by M. P. Esposito, H. M. Drake, J. A. Smith, and T. W. Owens PEDCo Environmental, Inc. Cincinnati, Ohio 45246 Contract No. 68-03-2577 Project Officer David R. Watkins Industrial Pollution Control Division Industrial Environmental Research Laboratory Cincinnati, Ohio 45268 INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268 �DISCLAIMER This report has been reviewed by the Industrial Environmental Research Laboratory-Cincinnati (lERL-Ci), U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the U.S. Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. ii �FOREWORD When energy and material resources are extracted, processed, converted, and used, the related pollutional impacts on our environment and even on our health often require that new and increasingly more efficient pollution control methods be used. The Industrial Environmental Research LaboratoryCincinnati (lERL-Ci) assists in developing and demonstrating new and improved methodologies that will meet these needs both efficiently and economically. This report is one of a three-volume series dealing with a group of hazardous chemical compounds known as dioxins. The extreme toxicity of one of these chemicals, 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD), has been a concern of both scientific researchers and the public for many years. The sheer mass of published information that has resulted from this concern has created difficulties in assessing the overall scope of the dioxin problem. In this report series the voluminous data on 2,3,7,8-TCDD and other dioxins are summarized and assembled in a manner that allows comparison of related observations from many sources; thus, the series serves as a comprehensive guide in evaluation of the environmental hazards of dioxins. Volume I is a state-of-the-art review of dioxin literature. Detailed information is presented on the chemistry, sources, degradation, transport, disposal, and health effects of dioxins. Accounts of public and occupational exposure to dioxins are also included. Volume II details the development of a new analytical method for detecting part-per-trillion levels of dioxins in industrial wastes. It also includes a review of the analytical literat jre on methods of detecting dioxins in various types of environmental samples Volume III identifies various routes of formation of dioxins in additio to the classical route of the hydrolysis of chlorophenols. The possible presence of dioxins in basic organic chemicals and pesticides is addressed, and production locations for these materials are identified. For further information, contact Project Officer David R. Watkins, Organic and Inorganic Chemicals Branch, lERL-Ci. Phone (513) 684-4481. David G. Stephan Director Industrial Environmental Research Laboratory Cincinnati 111 �PREFACE This report is Volume I in a series of three reports dealing with a group of hazardous chemical compounds known as dioxins. This volume discusses the occurrence, environmental transport, and toxicity of this class of compounds, and also summarizes the reported incidents of human exposure to them and the techniques available for decontamination and disposal of dioxin-contaminated material. Other volumes of this series examine analytical techniques used to identify the dioxins, the detailed chemistry of dioxin formation, and the commercial products with potential for containing dioxin contaminants. An extensive amount of literature published during the past 25 years has been concerned primarily with one extremely toxic member of this class of compounds, 2,3,7,8-tetrachlorodibenzo-p-dioxin. Often described in both popular and technical literature as "TCDD" or simply "dioxin," this compound is one of the most toxic substances known to science. This report series is concerned not only with this compound, but also with all of its chemical relatives that contain the dioxin nucleus. Throughout these reports, the term "TCDD's" is used to indicate the family of 22 tetrachlorodibenzo-pdioxin isomers, whereas the term "dioxin" is used to indicate any compound with the basic dioxin nucleus. The most toxic isomer among those that have been assessed is specifically designated as "2,3,7,8-TCDD." The objective in the use of these terms is to clarify a point of technical confusion that has occasionally hindered comparison of information from various sources. In particular, early laboratory analyses often reported the presence of "TCDD," which may have been the most-toxic 2,3,7,8-isomer or may have been a mixture of several of the tetrachloro isomers, some of which are relatively nontoxic. Throughout this report series, the specific term 2,3,7,8-TCDD is used when it was the intent of the investigator to refer to this most-toxic isomer. Since early analytical methods could not dependably isolate specific isomers from environmental samples, the generic term "TCDD's" is used when this term appears to be most appropriate in light of present technology. IV �ABSTRACT Concern about the potential contamination of the environment by dibenzo-p-dioxins through the use of certain chemicals and disposal of associated wastes prompted this study. This volume reviews the extensive amount of dioxin literature that has recently become available. Although most published reports deal exclusively with the highly toxic dioxin 2,3,7,8-TCDD, some include information on other dioxins. These latter reports were sought out so that a document covering dioxins as a class of chemical compounds could be prepared. A brief description of what is known about the chemistry of dioxins is presented first. This is followed by a detailed examination of the industrial sources of dioxins. Chemical manufacturing processes likely to give rise to 2,3,7,8-TCDD and other dioxin contaminants are thoroughly discussed. Other sources are also addressed, including incineration processes. Incidents of human exposure to dioxins are reviewed and summarized. Reports on possible routes of degradation and transport of dioxins in air, water, and soil environments are characterized. Current methods of disposal of dioxin-containing materials are described, and possible advanced techniques for ultimate disposal are outlined. Finally, an extensive review of the known health effects of 2,3,7,8-TCDD and other dioxins is presented. This review emphasizes the results of recent toxicological studies of the effects produced by chronic exposures and also the various possible mechanisms of action for these toxicants. This report was submitted in fulfillment of Contract No. 68-03-2577 by PEDCo Environmental, Inc., under the sponsorship of the U.S. Environmental Protection Agency. This report covers the period June 15, 1978 to January 6, 1980, and work was completed as of January 6, 1980. �CONTENTS Foreword Preface Abstract List of Figures List of Tables Acknowledgment List of Abbreviations iii iv y viii x xii xiii 1. Introduction 1 2. Chemistry 2 3. Sources 14 4. Routes of Human Exposure 77 5. Environmental Degradation and Transport 98 6. Disposal and Decontamination 131 7. Health Effects 147 References 200 Index 241 vii �FIGURES Number Page 1 Formation of Dioxins 2 Basic Chlorophenol Reactions 20 3 Direct Chiorination of Phenol 23 4 Flow Chart for 2,4,5-TCP Manufacture 30 5 Flow Chart for Hexachlorophene Manufacture 52 6 Locations of Current and Former Producers of Chlorophenols and Their Derivatives 60 7 Map of Seveso Area Showing Zones of Contamination 79 8 Map of Test Area C-52A, Eg!in Air Force Base Reservation, Florida 112 Diagram of Microagroecosystem Chamber 116 Farms at Which Cow's Milk Samples Were Collected for TCDD Analysis in 1976 (July-August) 125 11 Schematic of Molten Salt Combustion Process 135 12 Schematic of Microwave Plasma System 137 13 Schematic for Ozonation/Ultraviolet Irradiation Apparatus 141 14 Internal View of Pesticide Micropit 146 15 Excretion of 14C Activity By Rats Following A Single Oral Dose of 50 pg/kg (0.14 uCi/kg) 2,3,7,8-TCDD 153 Proposed Mechanism For Induction of AHH and Toxicity By 2,3,7,8-TCDD 156 9 10 16 17 7 Schematic of Rat Liver 13 Days After Administration of 2,3,7,8-TCDD 158 viii �FIGURES (continued) Number 18 Page Drawing of Tissue From Heart of Monkey Fed 2,3,7,8-TCDD; Fixed With Formalin and Stained With Hematoxylin and Eosin 160 19 Drawing of Heart Tissue From Monkey Fed 2,3,7,8-TCDD 161 20 Drawing of Section of Skin of Monkey Fed 2,3,7,8-TCDD 163 21 Drawing of Multinucleated Liver Cell From A Female Rat Given 0.1 ug of 2,3,7,8-TCDD/kg/day For 2 Years 164 22 Drawing of Liver Tissue From Rat Fed 2,3,7,8-TCDD 165 23 Drawing of Normal Membrane Junctions From the Periportal Region of A Test Animal 42 Days After Administration of 200 ug/kg 2,3,7,8-TCDD 166 Drawing of Distorted Periportal Membrane Junction, Showing Loss of Continuity of Plasma Membranes Between Parenchyma! Cells (42 Days After 200 ug/kg 2,3,7,8-TCDD) 167 24 25 Focal Alveolar Hyperplasia Near Terminal Bronchiole Within Lung of Rat Given 2,3,7,8-TCDD At Dosage of 0.1 ug/kg Per Day 26 27 28 168 Lesion Classified Morphologically As Hepatocellular Carcinoma In Liver of Rat Given 0.1 ug of 2,3,7,8TCDD/kg Per Day 185 Lesion Within Lung of Rat Given 0.1 ug of 2,3,7,8-TCDD/kg Per Day 186 Linear Correlation of New South Wales Rate For Neural-Tube Defects With Previous Year's Usage of 2,4,5-T In Australia 197 ix �TABLES Number Page 1 Chlorinated Dioxins 4 2 Physical Properties of Two Chlorinated Dioxins 5 3 Chlorodioxins Reported in Chlorophenols 15 4 Commercial Chlorophenols and Their Producers 18 5 1977 Pentachlorophenol Production Capacity 25 6 Former 2,4,5-TCP Manufacturing Sites 32 7 'Current Basic Producers of 2,4-D and 2,4-DB Acid.s, Esters, and Salts 37 8 Former Basic Producers of 2,4-D and 2,4-DB Acids, and Salts 38 9 Derivatives of 2,4,5-Trichlorophenol and Their Recent (1978) Producers 40 10 Former Producers of 2,4,5-T 44 11 Locations of Current and Former Producers qf and Their Derivatives 61 12 Dioxins in Selected Samples 72 13 Sources of Purified Dioxin Samples for Research 76 14 Dioxins In Commercial Gelatin 88 15 Reported Incidents of Occupational Exposure To B Routine Chemical Manufacturing 92 16 Occupational Exposures To Dioxins Through Aqcldetvts In Chemical Manufacturing Industry 93 17 Industries Using Dioxin-Related Chemicals 95 18 Concentrations of Herbicide Orange and 2,3,7,8*T£D0 In Three : Treated Test Plots ' X 100 �TABLES (continued) Number Page 19 Degradation of 2,3,7,8-TCDD In Soil 101 20 Photodegradation of 2,3,7,8-TCDD 104 21 Photodegradation of DCDD and OCDD 106 22 Concentrations of 2,3,7,8-TCDD at Utah Test Range 4 Years After Herbicide Orange Applications 114 23 Concentrations of 2,3,7,8-TCDD at Eglin Air Force Base 414 Days After Herbicide Orange Application 114 24 TCDD Levels In Wildlife 119 25 TCDD Levels in Milk Samples Collected Near Seveso In JulyAugust 1976 124 26 Soil Application Rates and Replications 128 27 Toxicities of Selected Poisons 148 28 Biological Properties of Dioxins 150 29 Enzyme Induction 150 30 Body Burden of 14C Activity In Six Rats Given A Single Oral Dose of 1.0 ug of [14C]-TCDD/kg 152 31 Toxicities of Organic Pesticides and 2,3,7,8-TCDD 170 32 Acute Toxicities of Dioxins 171 33 Acute Toxicities of 2,3,7,8-TCDD for Various Species 171 34 Summary of Acute Toxicity Effects of 2,3,7,8-TCDD 172 35 Effects of In Vivo 2,3,7,8-TCDD Exposure on Functional Immunological Parameters 177 36 Summary of Neoplastic Alterations Observed In Rats Fed Subacute Levels of 2,3,7,8-TCDD for 78 Weeks 183 37 Mutagenicity of Dioxin Compounds In Salmonella Typhimurium 187 38 Combined Rate of Neural-Tube Defects in New South Wales and Previous-Year Usage of 2,4,5-T In Australia 196 xl �ACKNOWLEDGMENT This report was prepared by PEDCo Environmental, Inc., under the direction of Richard W. Gerstle. M. Pat Esposito was the Project Manager and principal investigator. Contributing authors included H. M. Drake, Jeffrey A. Smith, M.D., and Timothy W. Owens. Additional technical assistance was provided by Terrence W. Briggs, Ph.D., F. Howard Schneider, Ph.D., and A. Christian Worrell of PEDCo, and Dr. Pat Sferra, EPA, lERL-Cincinnati. The chemical figures used throughout this report series were provided by Walk, Haydel & Associates. The hand renderings of photomicrographs in Section 7 of this volume were contributed by Lauren J. Smith. The cooperation of the many organizations and individuals who assisted in the collection of resource material is appreciated. In particular, we acknowledge Battelle Columbus Laboratories, Columbus, Ohio, for their part in the evaluation of disposal and decontamination technology. We also thank Mary Reece and Harvey Warnick (Office of Pesticide Programs, EPA), Charles Auer (Office of Toxic Substances, EPA), and Captain Alvin Young (U.S. Air Force), for their assistance in gathering and clarifying points of information for this document. xii �LIST OF ABBREVIATIONS DBDD's DCDD's Dioxins Hexa-CDD's Hepta-CDD's LD50 MCDD's OCDD PCPx n Penta-CDD's -- -~ "nr" ppb ppm ppt 1 HHU TBDD s TCDD's 2,3,7,8-TCDD TCP Tri-CDD1s dibromodibenzo-p-dioxins dichlorodibenzo-p-dioxins dibenzo-p-dioxins hexachlorodi benzo-p-di oxi ns heptachlorodibenzo-p-dioxins lethal dose to 50% of test group monochlorodibenzo-p-dioxins octachlorodi benzo-p-di oxi n pentachlorophenol pentachlorodibenzo-p-dioxins parts per billion (pg/l or ng/ml) parts per million (mg/1 or (jg/ml) parts per trillion (ng/1 or pg/ml) tetrabromodibenzo-p-dioxins tetrachlorodibenzo-p-dioxins 2,3,7,8-tetrachiorodi benzo-p-di oxi n trichlorophenol trichlorodibenzo-p-dioxins xiii �SECTION 1 INTRODUCTION The growing concern with contamination of the environment by dioxins arises principally from their toxicity and their widespread distribution as contaminants of commercial products. The purpose of this report is to present in a systematic and summary manner what is currently known about dioxins and their effects. Although most published reports deal exclusively with the highly toxic dioxin 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8TCDD), some include information on other dioxins. These latter reports were sought out so that a document covering dioxins as a class of chemical compounds could be prepared. The report first presents an account of the chemistry of dioxins (Section 2), their physical and chemical properties and modes of formation. Section 3 considers the sources of dioxins, focusing on the chemical manufacture of chlorinated phenols and their derivatives. Section 4 provides a brief account of the major known incidents of human exposure to dioxins in the environment. In the aftermath of these incidents, which include both occupational exposures and exposures of the general public, scientists of many disciplines have undertaken extensive and continuing investigations of the fate of dioxins when they are released to the environment. Section 5 reviews the findings of these studies, summarizing the known mechanisms of biodegradation, photodegradation, physical transport, and biological transport. The investigations indicate that the persistence of dioxins poses a serious environmental problem. In attempts to deal with this problem, numerous environmental research and development projects are aimed at developing methods of destroying these toxic contaminants after they have been formed. This work on dioxin disposal methods and decontamination procedures is described in Section 6. Finally, Section 7 reviews the current scientific knowledge of the health effects of dioxins, as indicated in epidemiological and laboratory studies of animal and human subjects who have been exposed to dioxin contamination. It is intended that this review of dioxin contaminants, from their formation through their dispersal into various environmental media and the consequent effects, can provide a point of perspective for those who are concerned with regulatory efforts and with research and development directed toward reducing the hazards of dioxin contamination. �SECTION 2 CHEMISTRY A dipxin is any of a family of compounds known chemically as dibenzopara-dioxins. Each of these compounds has as a nucleus a triple-ring structure consisting of two benzene rings interconnected to each other through a pair of oxygen atoms. The structural formula of the dioxin nucleus and the convention used in numbering the substituent positions are as follows: 9 1 Each of these substituent positions, numbered 1 through 4 and 6 through 9, can hold a chlorine or other halogen atom, an organic radical, or (if no other substituent is indicated in the formula or its chemical name) a hydrogen atom. The only differences in members of the dioxin family are in the nature and position of substituents. Most environmental interest in dioxins and most studies of this family of compounds have centered on chlorinated dioxins, in which the chlorine atom occupies one or more of the eight positions. Theoretically, there are 75 different chlorinated dioxins, each with different physical and chemical properties, differing only in the number of chlorine atoms in each molecule and in their relative locations on the dioxin nucleus. There are, for example, two monochiorodioxins, in which one chlorine atom is attached to the nucleus at either position 1 or position 2. If two or more chlorine atoms are present, additional isomeric forms are possible, in accordance with the following schedule (Buser, Bosshardt, and Rappe 1978): 2 isomers of monochiorodibenzo-p-dioxin (MCDD's) 10 isomers of dichlorodibenzo-p-dioxin (DCDD's) 14 isomers of trichlorodibenzo-p-dioxin (Tri-CDD's) 22 isomers of tetrachlorodibenzo-p-dioxin (TCDD's) 14 isomers of pentachlorodibenzo-p-dioxin (Penta-CDD's) 10 isomers of hexachlorodibenzo-p-dioxin (Hexa-CDD's) �2 isomers of heptachlorodibenzo-p-dioxin (Hepta-CDD's) 1 octachlorodibenzo-p-dioxin (OCDD) Table 1 lists the 75 possible chlorinated dioxins, and also notes the 40 that have been prepared and identified and whose analytical characteristics have been published (Buser, Bosshardt, and Rappe 1978; Buser 1975; Pohland and Yang 1972; Bo1 ton 1978). Five others, as noted in the table, have been identified as distinct compounds but have not been clearly differentiated from each other (Buser, Bosshardt, and Rappe 1978; Buser 1975; Rappe 1978). The interest of health and environmental researchers in dioxins arose1 principally because of the toxicity and distribution of one of these compounds, 2,3,7,8-TCDD, whose structural formula is as follows: This is an unusual organic chemical, symmetrical across both horizontal and vertical axes. It is remarkable for its lack of reactive functional groups and its chemical stability (Poland and Kende 1976). It is an extremely lipophylic molecule, and only sparingly soluble in water and most organic liquids; it is a colorless crystalline solid at room temperature. The physical properties of 2,3,7,8-TCDD are shown in Table 2, along with those of OCDD, another chlorinated dioxin with twofold symmetry (World Health Organization 1977; Crummett and Stehl 1973). No published reports indicate that dioxins are formed biosynthetically by living organisms; these compounds apparently are not constituents of a normal growing environment. The presence of dioxins in fly ash, 2-chlorophenol, 2,4,6-trichlorophenol, and hexachlorobenzene indicates that there may be yet-undiscovered mechanisms that produce these compounds. In a recent study, chlorinated dioxins were created by pyrolysis of chlorobenzenes in the presence of air (Buser 1979b). Dioxins have been made from catechols in condensations with polychlorobenzenes and chloronitrobenzenes (World Health Organization 1977; Gray et al. 1976; March 1968). A pesticide manufacturer has reported the finding of chlorinated dioxins in cigarette smoke and fireplace soot (Dow Chemical Company 1978). Other possible routes of formation are examined in Volume III of this series. One route that has been completely demonstrated by extensive chemical research is the formation of chlorinated dioxins from industrial chemicals, especially from certain "precursor" compounds that lead directly to dioxin formation. In generalized form, this reaction is as follows: �TABLE 1. CHLORINATED DIOXINS 1-chloro a 2-chloro 1,2-dichloro a a a a 1,3-dichloro 1,4-dichloro 1,6-dichloro 1,7-dichloro 1,8-dichloro 1,9-dichloro 2,3-dichloro 2,7-dichloro 2,8-dichloro a a a a 1,2,3-trlchloro a 1,2,4-trichloro a 1,2,6-trichloro 1,2,7-trichloro 1,2,8-trichloro 1,2,9-trichloro 1,3,6-trichloro 1,3,7-trichloro a 1,3,8-trichloro 1,3,9-trichloro 1,2, 3, 4- tetrachl oro a 1,2, 3, 6- tetrachl oro 1,2,3,7-tetrachloro 1,2,3,4,6-pentachloro 1,2,3,4,7-pentachloro 1,2,3,6,7-pentachloro a a 1,2,3,8-tetrachloro a 1,2,3,9-tetrachloro 1,2,4,6-tetrachloro 1,2,4,7-tetrachloro 1,2,4,8-tetrachloro 1,2,4,9-tetrachloro 1,2,6,7-tetrachloro a 1,2,3,6,8-pentachloro 1,2,3,6,9-pentachloro 1,2,3,7,8-pentachloro 1,2,3,7,9-pentachloro 1,2,3,8,9-pentachloro 1,2,4,6,7-pentachloro c 1,2, 6, 8- tetrachl oro 1,2,6,9-tetrachloro a 1,2,4,6,8-pentachloro 1,2,4,6,9-pentachloro 1,2,4,7,8-pentachloro a 1,2,4,7,9-pentachloro 1,2,4,8,9-pentachloro 1, 2,7,8- tetrachloro 1,2,7,9-tetrachloro 1,2, 8, 9- tetrachloro 1,3,6,8-tetrachloro 1,3, 6, 9- tetrachloro 1,3, 7 ,8- tetrachl oro 1,3,7, 9- tetrachl oro a a a a a 1,4, 6, 9- tetrachl oro a 1,4, 7, 8- tetrachl oro 2,3,7,8-tetrachloro a 1,4,6-trichloro 1,4,7-trichloro 2,3,6-trichloro 2,3,7-trichloro a 1,2,3,4,6,7-hexachloro 1,2,3,4,6,8-hexachloro 1,2,3,4,6,9-hexachloro 1,2,3,4,7,8-hexachloro 1,2,3,6,7,8-hexachloro 1,2,3,6,7,9-hexachloro 1,2,3,6,8,9-hexachloro 1,2,3,7,8,9-hexachloro 1,2,4,6,7,9-hexachloro 1,2,4,6,8,9-hexachloro 1,2,3,4,6,7,8-heptachloro 1,2,3,4,6,7,9-heptachloro Octachloro u Identified compounds. One or the other of these compounds has been prepared. . A mixture of these three compounds has been prepared. The Dow Chemical Company has recently reported the synthesis of all 22 TCDD isomers. 4 a c a c a a a a a a b a a b a a a �TABLE 2. PHYSICAL PROPERTIES OF TWO CHLORINATED DIOXINS 2,3,7,8-TCDD Empiric formula Percent by weight C 0 H Cl Molecular weight Melting point, °C Decomposition temperature, °C Solubilities, g/liter o-dichlorobenzene Chlorobenzene Anisole Xylene Benzene Chloroform n-Octanol Methanol Acetone Dioxane Water C12H4C1402 44.7 9.95 1.25 44.1 OCDD Ci2Clg02 31.3 7.0 61.7 322 305 Above 700 459.8 130 Above 700 1.4 0.72 1.83 1.73 3.58 0.57 0.37 0.56 0.048 0.01 0.11 0.38 0.0000002 (0.2 ppb) �2XY This reaction indicates that a compound may be a dioxin precursor if it meets two conditions: 0 The precursor compound must be an ortho-substituted benzene ring in which one of the substituents includes an oxygen atom directly attached to the ring. 0 It must be possible for the two substituents, excluding the oxygen atom, to react with each other to form an independent compound. These conditions are met by many organic compounds, including a class of mass-produced chemicals, the ortho-chlorinated phenols. The hydroxyl group of the phenol supplies the ring-attached oxygen atom. The hydrogen of the hydroxyl group is capable of reacting with chlorine, the other substituent, to form hydrogen chloride, an independent compound. An even more likely precursor is the sodium or potassium salt of an ortho-chlorinated phenol because the coproduct of this condensation is sodium or potassium chloride, either of which is an even more stable inorganic salt. Almost all original dioxin researchers used ortho-chlorinated phenols as precursors. Most often, the reactions were conducted in the presence of sodium or potassium hydroxide, either of which will react spontaneously with the phenol groups to form the phenylate salts. Six chemical reactions, all of which have been performed in laboratory experiments, are shown in Figure 1 (Pohland and Yang 1972; World Health Organization 1977; Crosby, Moilanen, and Wong 1973; Milnes 1971). Not all of these reactions, however, have produced the expected dioxin in high yield, and investigators have detected other dioxins and similar compounds that were not attributable to these simple reactions. Numerous studies have therefore explored the reaction mechanism of dioxin formation and the complex of competing reactions that create other compounds of this type (Buser 1975; Nilsson et al. 1974; Jensen and Renberg 1972; Plimmer 1973; Buser 1978). The basic dioxin reaction actually occurs in two steps. In the condensation of 2,4,5-trichlorophenol, for example, one pair of substituents reacts first to form a phenoxyphenate, or substituted diphenylether, in accordance with the following reaction (Nilsson et al. 1974; Jensen and Renberg 1972; Buser 1978; Moore 1979). �COPPER POWDER CATALYST^ "Cl 0-CHLOROPHENOL POTASSIUM SALT IN WATER ANC POTASSIUM HYDROXIDE UNSUBSTITUTED OIOXIN COPPER POWDER CATALYST^ IN VACUUM SUBLIMATOR 2,4-DICHLDROPHENOL POTASSIUM SALT Cl Cl -ONa CONDITIONS UNREPORTED Cl 2,4,6-TRICHLOROPHENOL SODIUM SALT VARIETY OF CONDITIONS 2,4.5-TRICHLOROPHENOL SODIUM SALT Cl COPPER POWDER CATALYST^ IN VACUUM SUBLIMATOR Cl Cl 2,3,5.6-TETRACHLOROPHENOL POTASSIUM SALT Cl 1,2,4.6,7,9-HEXA-CDD Cl Cl HEATING ONLY •ci ,CI Cl PENTACHLOROPHENOL OCDD Figure 1. Formation of dioxins. �-t- NaCI PREDIOXIN Compounds of this type have been termed "predioxins." They have been identified in waste sludges and commercial products as well as in the products of laboratory experiments (Jensen and Renberg 1972; Arsenault 1976; Jensen and Renberg 1973). There are other competing reactions, however. With some precursor compounds, condensation may occur with a chlorine atom that is not in the ortho position to a hydroxyl group. One study suggests that a meta chlorine will be favored, in accordance with the following reaction (Langer, Brady, and Briggs 1973). ONac, Cl ISOPREDIOXIN The end product has been termed an "iso-predioxin" (Jensen and Renberg 1973). To this iso-predioxin, additional molecules of sodlum-2,4,5" trichlorophenate may attach, creating a polymerized compound of three, four, or more monomers (Langer, Brady, and Briggs 1973; Langer et al. 1973). ONa Cl Investigators have noted similar reactions with para chlorine atoms, which form another type of iso-predioxin. Either of the iso-predioxins may polymerize into longer chains, or they may lead with loss of chlorine to the creation of dibenzofurans (Jensen and Renberg 1972; Langer, Brady, and Briggs 1973; Deinzer et al. 1979; Chemical Engineering 1978). 8 �Cl ONa It is believed that dibenzofurans are also formed by reaction between a chlorophenol and a polychlorobenzene through an intermediate creation of another type of diphenyl ether (Buser 1978). Cl NaCI NaOH NaCt + + H2O Cl Another competing reaction that involves loss of chlorine is the reaction to form dihydroxy chlorinated biphenyls (Jensen and Renberg 1973). Cl The chlorine thus released may react with other rings to form compounds with higher chlorine saturation. Preparation of 2,3,7,8-TCDD was accomplished by treatment of unsubstituted dioxin (World Health Organization 1977). �CI2 -*- Other competing reactions have been described for pentachlorophenol, which has been shown to degenerate, when heated, into hexachlorobenzene and water by a reaction sequence that includes an intermediate decachlorodiphenylether (Plimmer 1973). Cl HCI + HCI Cl Alternatively, the predioxin or the decachlorodiphenylether may lose chlorine through reactions with water to form hexachloro or heptachlorodioxins or to form octa- and nonachlorodiphenylethers. Loss of chlorine may also create octachlorodibenzofuran in accordance with the following reaction (Crosby, Moilanen, and Wong 1973; Jensen and Renberg 1973). + CI2 Cl Cl Cl These competing reactions are predominant only with acidic pentachlorophenol, however. Heating the sodium salt of pentachlorophenol produces OCDD in essentially quantitative yield (World Health Organization 1977). 10 �Except for pentachlorophenol, once a predioxin is formed, there are apparently no competing reactions other than its reversal into the precursor. In one test, when Irgasan DP-300, a predioxin (see Section 3 p. 57), was heated to 980°C, only two classes of compounds were created: dioxins and precursor molecules (Nilsson et al. 1974). The competing reactions clearly indicate why dioxins generally are formed only in trace quantities and why they appear in a complex mixture with polymers and other multi-ring structures, many of which are also toxic. It has been more difficult to explain why dioxins other than the one predicted by theory are also found in these mixtures. In the laboratory, for example, a predioxin for 2,8-DCDD created a small amount of this dioxin when heated; however, the principal dioxin formed was 2,7-DCDD (Boer et al.. 1971). It was originally believed that such unexpected dioxins were created by arbitrary transfers of chlorine that occurred within the energetic predioxin molecules (Boer et al. 1971). More recent work has demonstrated that a long-recognized chemical phenomenon known as the "Smiles rearrangement" is often operational during dioxin creation, in which one of the rings spontaneously reverses into its mirror image at the instant of ring closure (Gray et al. 1976; March 1968). This rearrangement fully explains the reaction shown above, and researchers can now predict with some certainty which dioxins will be formed from specific precursors or predioxins. Even this development has not satisfied all observational evidence, however, especially with the more highly chlorinated dioxins. Some researchers believe that an equilibrium process is at work, in which dioxins slowly lose or gain chlorine atoms to approach the most stable mixture of compounds (Rawls 1979; Miller 1979; Ciaccio 1979). Predioxin formation does not ensure dioxin formation (Jensen and Renberg 1972; Jensen and Renberg 1973). Pentachlorophenol attains equilibrium with its precursor in a reversible reaction but forms large amounts of dioxins only in the presence of an alkali (Langer et al. 1973). Irgasan DP-300 can be chlorinated and otherwise modified chemically without inducing ring closure (Nilsson et al. 1974; Yang and Pohland 1973). "High amounts" of predioxins have been found in commercial products in which no dioxin could be detected. Another study revealed predioxin concentrations as much as 20 times greater than dioxin concentrations (Jensen and Renberg 1972). In still another study, the concentration of hydroxypolychlorodiphenyl ethers (predioxins plus isopredioxins) was more than 50 times the dioxin concentration (Deinzer et al. 1979; Chemical Engineering 1978). Although not specifically noted in published literature, predioxin formation appears 11 �to be more likely than dioxin formation. It is possible that steric or electronic hindrances interfere with the final step of ring closure, and that predioxins may be formed under less-rigorous reaction conditions. Since dioxins usually are formed only in low yields, the minimum conditions leading to their formation are poorly defined. Heat, pressure, catalytic action, and photostimulation have all been shown to encourage the reactions from chlorinated precursors to predioxins and then to dioxins. The temperature required for dioxin formation is variously reported at values from 180°C to 400°C (Milnes 1971; Langer, Brady, and Briggs 1973; Crossland and Shea 1973; Gribble 1974; Buser 1978). As previously noted, sodium pentachlorophenate is converted to essentially pure OCDD at approximately 360°C (Langer et al. 1973). The same series of tests indicated decomposition of several other chlorinated dioxin precursors at temperatures from about 310° to 370°C, with formation of varying quantities of dioxins (Langer et al. 1973). Essentially quantitative formation of many different dioxins from chlorinated catechols and o-chloronitrobenzenes has been achieved at 180°C (Gray et al. 1976; March 1968). Direct combustion of herbicides or impregnated sawdust can create dioxins (Nilsson et al. 1974; Langer, Brady, and Briggs 1973; Stehl and Lamparski 1977; Ahling and Lindskog 1977; Jansson, Sundstrom, and Ahling 1978), especially if there is a deficiency of oxygen (Chem. and Eng. News 1978), but the temperature of formation under these conditions cannot be measured (this phenomenon may be limited to formation of dioxins from pentachlorophenol; reports are indefinite). Apparently no definitive study has determined the temperature of formation of 2,3,7,8-TCDD. Pressure is needed to retain some precursor compounds in the liquid state to permit dioxin formation (Jensen and Renberg 1972). At atmospheric pressure, the boiling point of many precursors is apparently lower than the temperature needed to form dioxins, and therefore the precursors escape from the reaction vessel before decomposition reactions can occur. Irradiation of pentachlorophenol with ultraviolet light has caused the formation of OCDD (World Health Organization 1977; Crosby, Moilanen, and Wong 1973; Plimmer et al. 1973; Crosby and Wong 1976). Irradiation of 2,4-dichlorophenol, however, energized the hydrogen atom at position 6 of one ring and created a predioxin as a principal product, but ring closure apparently did not occur (Plimmer et al. 1973). This experiment also produced a dihydroxy biphenyl, probably through the competing reaction described previously. It has been postulated that although dichloro, trichloro, and tetrachloro dioxins may be formed by irradiation, they do not accumulate because they decompose rapidly by the same mechanism (Crosby, Moilanen, and Wong 1973). As outlined in Section 5, the less chlorinated dioxins are unstable when exposed to ultraviolet light. In laboratory production of dioxins, catalysts have been used to increase reaction rates and reaction yields. Powdered copper, iron or aluminum salts, and free iodine have been used (Pohland and Yang 1972; World Health Organization 1977), and all of these are known to stimulate many 12 �reactions of chlorinated organic compounds (Wertheim 1939). One report indicates that heavy metallic ions may decrease decomposition temperature (Langer et al. 1973). Presence of heavy metals may, however, only encourage competing reactions; the silver salt of pentachlorophenol, for example, decomposes at about 200°C to yield polymerized materials but no dioxins (Langer et al. 1973). Formation of dioxins is an exothermic reaction (Langer et al. 1973) that releases heat as the molecules contract into a more compact arrangement. No published data define the amount of heat created by formation of the various dioxins. Once formed, the dioxin nucleus is quite stable. Laboratory tests have shown that it is not decomposed by heat or oxidation in a 700°C incinerator, but pure compounds are largely decomposed at 800°C (Ton That et al. 1973). A recent report states that the nucleus survives intact through incineration up to 1150°C if it is bound to particulate matter (Rawls 1979; Miller 1979; Ciaccio 1979). Chlorinated dioxins lose chlorine atoms on exposure to sunlight or to some types of gamma radiation, but the basic dioxin structure is largely unaffected (Crosby et al. 1971; Buser, Bosshardt, and Rappe 1978). In comparison with almost any other organic compound, the biological degradation rate of chlorinated dioxins is slow, although measured rates differ widely (Zedda, Cirla, and Sal a 1976; Commoner and Scott 1976b; Matsumura and Benezet 1973; Huetter 1980). 13 �SECTION 3 SOURCES DIOXINS IN COMMERCIAL CHLOROPHENOLS AND THEIR DERIVATIVES Since most reports of dioxins are associated with chlorinated phenolic compounds, this section examines this group of organic materials with respect to their reported dioxin contaminants and their utilization, manufacture, production volumes, and derivatives. Similar information is presented, when available, for hexachlorobenzene, which has been found to contain dioxins, and also for a group of other related commercial chemicals that theoretically could contain dioxin contaminants, although no analyses have been reported. For each chemical, the discussions include the probable processing steps that may promote dioxin formation and also the mechanisms through which dioxins could appear in the associated process wastes or be retained within the chemical products. Chlorophenols Chlorinated phenols are a family of 19 compounds, consisting of a benzene ring to which is attached one hydroxyl group and from one to five chlorine atoms. The positions of the chlorine atoms with respect to the hydroxyl group and to each other provide the opportunity for three monochlorophenols, six each of dichloro- and trichlorophenols, three tetrachlorophenols, and one pentachlorophenol. Many researchers have established the presence of dioxins in these chemicals; Table 3 lists the results of several such studies. Data in this table show that until recently dioxins have not been found in commercially produced mono- or dichlorophenols. The presence of 2,3,7,8-TCDD in low concentration was found in 1979 in a railroad tank car spill of o-chlorophenol. One or more samples of all Chlorophenols with three or more chlorine atoms that have been examined have contained dioxins. TCDD's have been identified not only in the 2,4,5-trichloro isomer but also in the 2,4,6-trichloro isomer. One or more samples of trichlorophenol have contained dioxins with two to eight chlorine substituents. Only dioxins with six to eight chlorine substituents have been found in tetra- and pentachlorophenol. Numerous analyses have confirmed that dioxins with less than six chlorine substituents are not found in pentachlorophenol. Most commercial Chlorophenols are used as raw materials in the synthesis of other organic compounds. Some of the less highly chlorinated phenols are used with formaldehyde to make fire-resistant thermosetting 14 �TABLE 3. CHLORODIOXINS REPORTED IN CHLOROPHENOLS Chlorodioxins (-CDD's), ppraa Chlorophenol sample mono-CDD's DCDD's Honochlorophenol 2-chlorophenol o-chlorophenol Di chlorophenol 2 ,4-di chl orophenol 2,6-dichlorophenol Tn'chlorophenol 2,4,5-trichlorophenol (1969) 2,4,5-trichlorophenol tri-CDD's ND - ND ~ ND NO ND ND ND NO NO ND ND ND ND ND ND TCDD's penta-CDD's hexa-CDD's hepta-CDD's OCDD ND Data source Firestone '72 0.30 (1,3,6,8) 6.2 (2,3,7,8) ND ND - ND - Chemical Week '79 ND NO ND ND ND ND Firestone '72 ND ND ND ND Firestone '72 1.5 ND ND ND - ND ND ND 0.037 ( , , , ) 2378" ND ND ND Firestone '72 Firestone '72 (1970) 2,4,5-trichlorophenol (1970) 2,4,5-trichlorophenol (1970) Na-2,4,5-trichlorophenol ND ND ND ND ND NO ND ND ND ND ND ND ND Firestone '72 ND ND ND ND Firestone '72 ND ND ND ND Firestone '72 (2,3,7,8) ND ND ND ND Firestone '72 0.3 (2,3,7,8) 49 (1,3,6,8) ND ( . ) 05 ND ND ND ND Elvidge '71 ND 0.07 (2,3,7,8) ND (1967) Na-2, 4, 5- trichl orophenol (1969) 2,4,5-trichlorophenol 2 ,4 ,6-tri chl orophenol trichlorophenol Tetrachlorophenol 2,3,4,6-tetrachlorophenol (Dowicide 6) 2,3,4,6-tetrachlorophenol 2,3,4,6-tetrachlorophenol ND 3.72 ( , ) 27 ND - 1.4 ND - ND - - - - - - ND NO ND NO ND NO ND ND ND NO NO - ND - ND - ND ND - 93 (2,3,7) - - 0.5-10 0.5-10 0 .5-10 6 - - 29 4.1 5.1 0.17 ND ND Firestone '72 Wool son et al. '72 Buser '75 Firestone '72 Firestone '72 (1967) 2 , 3 ,4 ,6-tetrachl orophenol tetrachl orophenol (continued) ND (0.5) ND 10-100 ND 10-100 ND 10-100 Firestone '72 Wool son et al. '72 �TABLE 3 (continued) Chlorodioxins (-CDD's), ppma Chlorophenol sample Pentachlorophenol PCP (Dowicide 7) PCP Na-PCP (1967) Na-PCP (1969) PCP (1970) PCP (1970) PCP (1967) PCP ( 9 9 16) PCP (1970) PCP (1970) PCP (1978) Pentachl orophenate PCP formulation PCP (technical grade) PCP (reagent grade) PCP (many samples) PCP's (17) PCP or PCP-Na (7) PCP (Dowicide 7 1970) PCP (Oowicide 7 1970) distilled PCP NaPCP (Dowicide G, 1978) mono-CDD's DCDD's tri -CDD's NO NO ND NO ND ND ND ND - NO ND ND ND ND ND ND ND - ND ND ND ND ND NO ND ND - - - - TCDD's ND (0.5) ND NO ND ND ND ND ND ND ND (0.1) ND ND ND - penta-CDD's hexa-CDD's hepta-CDD's OCDD ND ND ND ND ND ND ND ND - 9 10-100 14 20 39 35 0.17 13 0.91 15 19 33-42 0.02-0.03 9-27 0-23 0.03-10.0 4 1.0 235 100-1000 14.5 11.3 49 23 ND 47 2.1 23 140 •f 870 19-24 0.04-0.09 90-135 0.6-180 125 6.5 - 9-27 ND-2 1-12 250 100-1000 3.8 3.3 15 ND NO ND 5.3 15 432 + 50-3300 7-11 B. 02-0. 03 575-2510 0-3600 5.5-370 2500 15 575-2510 4-173 Data source Buser '75 Wool son '72 Firestone '72 Firestone '72 Firestone '72 Firestone '72 Firestone '72 Firestone '72 Firestone '72 Firestone '72 Dioxin in Industrial Sludges, '78 Jensen and Renberg '72 Jensen and Renberg '72 Villanueva '73 Villanueva '73 PCP - A wood preservative '77 Crummett '75 Buser and Basshardt '76 PCP Ad Hoc Study Rept. 12/78 SAB PCP Ad Hoc Study Rept. 12/78 SAB Johnson et al. '73 Dow Chemical Co, '78 Key to abbreviations and symbols: ND = Not detected (minimum detection level, ppm). Other numbers in parenthesis indicate year Chlorophenol sample was obtained, or specific dioxin detected. ^ - Indicates not analyzed or not reported. Presence of 2,3,7,8-TCDD confirmed but not quantitatively reported. �plastics (Doedens 1964). Those containing three or more chlorine atoms are used directly as pesticide chemicals. 2,4,6-Trichlorophenol is effective as a fungicide, herbicide, and defoliant (Hawley 1971). It was formerly used in large quantities in the leather tanning industry; however, its use in this industry has decreased substantially (U.S. Environmental Protection Agency 1978a), probably as a result of the improved effectiveness and mass production of 2,4,5-trichlorophenol, a substance of sufficient importance to warrant a special section in this report. 2,3,4,6-Tetrachlorophenol is used as a preservative for wood, latex, and leather, and also as an insecticide (Kozak et al. 1979). Pentachlorophenol or its sodium salt is said to be the second most widely used pesticide in the United States. It is effective in the control of certain bacteria, yeasts, slime molds, algae, fungi, plants, insects, and snails. Because of its broad spectrum, pentachlorophenol is used in many ways: As a preservative for wood, wood products, leather, burlap, cordage, starches, dextrins, and glues As an insecticide on masonry for termite control As a fungicide/slimicide in pulp and paper mills, in cooling tower waters, and in evaporation condensers As a preharvest weed defoliant on seed crops As a preservative on beans (for replanting only) As a means of controlling slimes in secondary oil recovery injection water (in the petroleum industry) By far the major use of pentachlorophenol is as a wood preservative. It was once reported to have been used in shampoos; however, this chemical does not now appear to be used as an ingredient in cosmetics or drugs, since it is not listed either in the CTFA Cosmetic Ingredient Dictionary (Cosmetic, Toiletry and Fragrance Association, Inc. 1977), or in the Physicians' Desk Reference (1978). Manufacture-Through either process variations or separation of mixtures by fractional distillation, manufacturers selectively produce chlorophenols with specific numbers and arrangements of chlorine atoms. Table 4 shows that 13 of the 19 possible chlorophenols are currently sold commercially in sufficient volume to be listed in the 1978 Stanford Research Institute Directory of Chemical Producers. Seven of these are made in much higher volume than the other six. The high-volume products are all made by one of two major types of manufacturing processes, referred to herein as the hydrolysis method and the direct chlorination method. 17 �TABLE 4. COMMERCIAL CHLOROPHENOLS AND THEIR PRODUCERS0 Chiorophenol Manufacturer(s) o-Chlorophenol Dow Chemical Company Monsanto Company m-Chlorophenol Eastman Kodak Company Aldrich Chemical Company Specialty Organics, Inc. R.S.A. Corporation p-Chlorophenol Dow Chemical Company Monsanto Company 2,3-Dichlorophenol Specialty Organics, Inc. 2,4-Dichlorophenol Dow Chemical Company Monsanto Company Rhodia, Inc. Vertac, Inc. 2,5-Dichlorophenol Velsicol Chemical Corporation 2,6-Dichlorophenol Aldrich Chemical Company Specialty Organics, Inc. 3,4-Dichlorophenol Aldrich Chemical Company 3,5-Dichlorophenol Aldrich Chemical Company Specialty Organics, Inc. 2,4,5-Trichlorophenol Dow Chemical Company Vertac, Inc. 2,4,6-Trichlorophenol Dow Chemical Company 2,3,4,6-Tetrachlorophenol Dow Chemical Company Pentachlorophenol Dow Chemical Company Vulcan Materials Company Reichold Chemicals a Source: Stanford Research Institute Directory of Chemical Producers, U.S. , 1978. 18 �As mentioned earlier, chlorophenols are benzene rings that contain one hydroxyl group and one or more chlorine atoms'. The basic raw material in the manufacture of chlorophenols is benzene, and the two major manufacturing methods differ primarily in the order in which the substituents are attached to the benzene ring. In the hydrolysis method, chlprophenols are made by replacing one chlorine substituent of a polychlorinated benzene with a hydroxyl group. The hydrolysis method is the only practical method for producing some of the chlorophenols, such as the 2,4,5 isomer; this isomer is apparently the only one currently produced in large quantity by this method (Kozak 1979; Deinzer 1979; Chemical Engineering 1978). In the direct chlorination method, phenol (hydroxybenzene) is reacted with chlorine to form a variety of chlorophenols. Each manufacturing method is more fully described in the paragraphs below. In addition, a detailed description of the manufacture of 2,4,5-trichlorophenol (2,4,5-TCP) is outlined separately. Hydrolysis method—The first step in the hydrolysis method is the direct chlonnation of benzene. Through a series of distillations, rechlorinations, and other chemical treatments, several purified chlorobenzene compounds are obtained that contain from two to six chlorine substituents. Specific chlorophenols are then made by reacting one of the chlorine substituents with caustic, thereby replacing the chlorine atom with a hydroxyl group (see Figure 2). The reaction takes place in a solvent in which both materials are soluble, and the mixture is held at specific conditions of temperature and pressure until the reaction is complete. The product is then recovered from the reaction mixture. The solvent is usually an alcohol (most often methanol), although use of other solvents is possible. A 1957 process patent describes the manufacture of pentachlorophenol from a starting material of hexachlorobenzene (U.S. Patent Office 1957e). Methanol is the solvent, and the reaction takes place at temperatures of 125° to 175°C and pressures of 125 to 360 psi. Reaction time is 0.3 to 3 hours. This method is known to have been used commercially (Arsenault 1976). A variation of this process using ethylene glycol as the solvent also has been used commercially for the production of 2,4,5-trichlorophenol (Commoner and Scott 1976a; Whiteside 1977). A process described in another 1957 patent uses water as the solvent in hydrolysis of dichloro- and trichlorobenzenes (U.S. Patent Office 1957c). Temperature is maintained from 240° to 300°C under alkaline conditions at autogenous pressure. Reaction time varies from 0.5 to 3 hours. By this method, monochlorophenols are produced in yields greater than 70 percent from o-, m-, and p-dichlorobenzene. Metachlorophenol is formed as an impurity from the ortho- and para- starting materials through ring rearrangment mechanisms. Orthochlorophenol, which is the most likely dioxin precursor, is not formed by ring rearrangement but is produced in 86 percent yield from o-dichlorobenzene. Also, hydrolysis of 1,2,4-trichlorobenzene forms a mixture of dichlorophenol isomers in yields up to 95 percent. A 1967 patent describes the use of a combined methanol-water solvent system (U.S. Patent Office 1967b). Temperature is maintained at 170° to 200°C, under above-autogenous pressures. Reaction time is 1 hour or less. 19 �DIRECT CHLORINATION SOME VARIATIONS EMPLOY A CATALYST CI2 SOLVENT UNNECESSARY MIXTURE OF CHLOROPHENOL S PHENOL HYDROLYSIS CATALYST UNNECESSARY ^ SOLVENT REQUIRED POLYCHLORINATED BENZENE Figure 2. SPECIFIC CHLOROPHENOL Basic chlorophenol reactions. 20 �A 1969 patent describes (U.S. Patent Office 1969). permits the reaction to take ysis of hexachiorobenzene to and is complete in about 3 applied commercially. still another solvent, dimethylsulfoxide (DMSO) Use of this solvent in a mixture with water place at atmospheric pressure; caustic hydrolpentachlorophenol occurs at approximately 155°C hours. This process apparently has never been When an alcohol is used as a solvent, the chemical mechanism that occurs involves an initial equilibrium reaction between the alcohol and caustic to form a sodium alkoxide, which is the reagent that actually attacks the chlorobenzene. The compound formed first is the alcohol ether of the chlorophenol. On standing, rearrangement of the compound occurs to form the chlorophenate plus any of several side reaction products (Sidwell 1976). This mechanism is significant because it explains the "aging" step that is a distinct phase in commercial hydrolysis sequences, and it also explains the substantial quantity of byproduct impurities that are derived from the alcohol solvents. In all these processes, the product is recovered through either of two methods. In one, extraction into benzene separates the organic materials from water, salt, and excess caustic. Subsequent vacuum distillation reclaims the benzene for recycle and also separates the chlorophenols into purified fractions. Extraction with benzene (or a similar solvent) is probably the preferred product recovery method for chlorophenols of lower molecular weight, especially the mono- and dichloro- products, since they are more easily distilled than the heavier products. The alternative product recovery method is to filter the reaction mixture, perhaps after partial neutralization or evaporation and subsequent cooling, to reclaim unreacted polychlorobenzenes. The solution is then acidified and filtered again to collect the solid products. This variation is probably best suited to recovery of tri-, tetra-, and pentachlorophenols because these products and their raw materials are solids at room temperature and therefore can be removed more easily in the filtration operations. Chlorophenols can be purified by distillation to separate high-boil ing impurities. Technical feasibility has been reported in three 1974 patents, in which purified pentachlorophenol is recovered in good yield by high vacuum distillation in the presence of chemical stabilizers (U.S. Patent Office 1974a, 1974b, 1974c). Purification of 2,4,5-trichlorophenol by distillation has also been reported (World Health Organization 1977). The high-temperature, high-pressure, and strongly alkaline conditions of the hydrolys.is process are conducive to the formation of dioxin compounds. Although not in present U.S. commercial use, the hydrolysis manufacture of pentachlorophenol was especially favorable for the formation of octachlorodibenzo-p-dioxin (OCDD). As described in more detail later in this section, the commercial hydrolysis method is known to produce 2,3,7,8-TCDD from 1,2,4,5-tetrachlorobenzene. 21 �Direct chlorination method—Direct chlorination begins by the addition of a hydroxyl group to benzene to form hydroxybenzene or phenol. This compound is manufactured in specialized plants, usually through sulfonation, chlorination, or catalytic oxidation of benzene. Dioxins have not been reported as resulting from this portion of the process; this study is therefore concerned only with the second part of the process in which phenol is reacted with chlorine to form various chlorophenols. The reaction of phenol with chlorine actually forms a mixture of chlorinated phenols (see Figure 2), although certain compounds are formed preferentially. Direct chlorination is practical, therefore, only if the desired product is one of the high-yield compounds. Except for low-volume specialty isomers and the high-volume 2,4,5 isomer, all commercial chlorophenols made in this country are those that are formed preferentially by this process (Buser 1978; Kozak 1979; Deinzer 1979; Chemical Engineering 1978). These include mono- and dichlorophenols that are substituted at positions 2 and 4, the symmetrical 2,4,6-trichlorophenol isomer, 2,3,4,6tetrachlorophenol, and pentachlorophenol. Chlorination of phenol can be accomplished in batch reactors, but is best suited to the continuous process shown in simplified form in Figure 3 (U.S. Patent Office 1960; Sittig 1969). Liquid phenol and/or lower chlorinated phenols are passed countercurrently with chlorine gas through a series of reaction vessels. Trace amounts of aluminum chloride catalyst are added, usually as a separate feed into an intermediate vessel. Equipment is sized so that all the chlorine is absorbed by the phenol; the last phenolcontaining vessel is usually built as a scrubbing column to ensure complete chlorine absorption. Gas leaving the scrubber is anhydrous hydrogen chloride, which is either used in other chemical operations or dissolved in water to form substantially pure hydrochloric acid as a byproduct. The chlorophenol compound created in greatest amount by this process is established by the ratio of feed rates of chlorine and phenol. Because all chlorine is consumed, it is fed at rates 1 to 5 times the molecular proportion of phenol, depending on the principal product desired. To prevent excessive oxidation that produces nonphenolic chlorinated organic compounds, temperatures are carefully regulated; the usual temperatures are 130° to 190°C for pentachlorophenol and 170°C for 2,4-dichlorophenol. Pressure is atmospheric, and reaction time is 5 to 15 hours (U.S. Patent Office 1960). The mixture from the first reaction vessel can be vacuum-distilled to separate the various compounds. Unreacted phenol and any undesired lesschlorinated phenols would be recycled. To make some products for which purity standards are rather flexible, very little purification is necessary, and some processes may include no final distillation or other treatment. Also, a chlorinated product may be withdrawn from the scrubber (usually a mixture of 2- and 4-mono- or 2,4-dichlorophenol) and may be either distilled, with portions recycled to the first reactor for further chlorination, or sold as is. 2,4-Dichlorophenol may be further processed to the phenoxy herbicide 2,4-D. 22 �CHLORINE ro CO HYDROGEN CHLORIDE BYPRODUCT PHENOL NONCONTACT HEATING OR COOLING COILS IN EACH VESSEL CHLOROPHENOLS TO PURIFICATION OR SALE ALUMINUM CHLORIDE CATALYST Figure 3. Direct chlorination of phenol. �Supplemental processing steps may be necessary to remove contaminants such as "hexachlorophenol" (hexachlorocyclohexadiene-l,4-one-3), dioxins, and furans from PCP made by this process. Hexachlorophenol may be formed during the process by overchlorination of the reaction mass (U.S. Patent Office 1939). Dioxins may be formed during distillation by the condensation of PCP with itself or with hexachlorophenol (see Table 1 of Volume 3 of this series). Dioxins have been reported in numerous samples of PCP, as shown in Table 3. Although hexa-CDD's, hepta-CDD's, and OCDD are known to be present in commercial PCP, 2,3,7,8-TCDD has never been found (Chemical Regulation Reporter 1978; U.S. Environmental Protection Agency 1978e). All PCP made in the United States is produced by the direct chlorination of phenol; apparently the method involving the hydrolysis of hexachlorobenzene has never been used commercially for PCP production (American Wood Preservers Institute 1977). Dow reportedly changed its production process in 1972 to produce a PCP with lower dioxin content; the other two producers of PCP apparently have not followed Dow's lead (Chemical Regulation Reporter 1978). Details of Dow's process change were not reported. Production-Production figures for di- and tetra- chlorophenols are not available. Although current figures for pentachlorophenol production are also not available, it is estimated from production capacity information (Table 5) that U.S. manufacturers are producing as much as 53 million pounds of PCP annually. Annual U.S. trichlorophenol production is probably also in the range of 50 million pounds (Crosby, Moilanen, and Wong 1973). As Table 4 indicates, chlorophenols are apparently manufactured by at least 11 companies, which represent two diverse groups of chemical producers. Of the 13 commercial chlorophenols, 7 are made by Dow Chemical Company in Midland, Michigan. Except for 2,4,5-trichlorophenol, all of the isomers made by Dow are those formed preferentially through direct chlorination of phenol. Competitive with Dow in the sale of these seven chlorophenols are four other companies: Monsanto Company - Sauget, Illinois Reichold Chemicals, Inc. - Tacoma, Washington Vulcan Materials Company- Wichita, Kansas Rhodia, Inc. - Freeport, Texas All of these companies are engaged for the most part in the mass production of organic chemicals for which market demand is relatively constant. These companies are geared to heavy chemical production, and their products are made to commercial standards of purity and are usually sold at relatively low prices. The other six chlorophenols are made by five companies that generally manufacture fine or specialty chemicals: 24 �TABLE 5. 1977 PENTACHLOROPHENOL PRODUCTION CAPACITY3 Company Production location 1977 Capacity, million of pounds Dow Chemical U.S.A.b Midland, Mich. 17 Monsanto0 Sauget, 111. 26 Reichold Tacoma, Wash. 20 Vulcan Witchita, Kans. 16 Total capacity 79 Source: American Wood Preservers Institute, 1977. These figures presumably do not include production of sodium or . potassium salts of pentachlorophenol. Dow ceased production of the sodium salt of PCP (Dowicide G) in April, 1978 (Dow Chemical Company 1978). Monsanto stopped all PCP production as of January 1, 1978 (Dorman 1978). 25 �Velsicol Chemical Corp. - Beaumont, Texas Eastman Kodak Company - Rochester, New York Aldrich Chemical Co., Inc. - Milwaukee, Wisconsin Specialty Organics, Inc. - Irwindale, California R.S.A. Corporation - Ardsley, New York Products from these manufacturers are often batch-produced under contract with specific industrial customers, sometimes to high standards of purity. They are manufactured in much smaller quantities than those described above, often intermittently, and they are sold at a relatively high price. Often, the products from these companies are used in the manufacture of Pharmaceuticals, photographic chemicals, and similar high-quality chemical materials. Without exception, the chlorophenols made by these companies are those not formed preferentially through direct chlorination of phenol. Any chlorophenol with a chlorine atom at position 2 (ortho to the hydroxyl group) may be a precursor for dioxin formation. Nine of the 11 companies are reported to make at least one chlorophenol of this description. Potential for the occurrence of dioxins is therefore not limited to the manufacture of chlorophenols for pesticide use. It is not known, however, whether the hydrolysis method, which is especially conducive to dioxin formation, is used to make the lower-volume chlorophenols. In many instances, this method probably is not used because the parent polychlorobenzenes needed for raw materials usually cannot be directly synthesized by conventional chlorination techniques. For production of m-chlorophenol in high yields, for example, general chemical references describe a synthesis route that involves chlorination of nitrobenzene, followed by reduction, diazotization, and hydrolysis of the nitrate group (Vinopal, Yamamoto, and Casida 1973). Multistep batch processes of this type are necessary to cause the substituents to attach to the ring at unnatural positions (Kozak 1979). These specialized production methods are not addressed in this report. The primary chemical producers described above are not the only commerical sources of chlorophenols. Other companies purchase chlorophenols from primary producers, combine them with other ingredients, and market the formulated products. Still others deal only in distribution of the chemicals or chemical mixtures. Most often the trade name of the product changes each time it is bought and sold. 26 �2,4,5-Trichlorophenol In 1972, hexa-, hepta- and octachlorodioxins were found at concentrations of 0.5 to 10 ppm in four of six trichlorophenol samples analyzed. Tetrachlorodioxins were not detected (0.5 ppm level of detection). The research report implies that the 2,4,5 isomer of trichlorophenol was being analyzed (Woolson, Thomas, and Ensor 1972). Also in 1972, another study showed dioxins in trichlorophenols (Firestone et al. 1972). Isomers identified in 2,4,5-trichlorophenol (or its sodium salt) at ppm levels were 2,7-di-, 1,3,6,8-tetra-, 2,3,7,8-tetra-, and pentachlorodioxins. High levels of 2,3,7-trichlorodioxin (93 ppm) and 1,3,6,8-tetrachlorodioxin (49 ppm) were found in the 2,4,6 isomers of trichlorophenol. The investigator analyzed for, but could not detect, mono-, hexa-, hepta-, and octachlorodioxins in these trichlorophenol samples. Data from these two studies are included in Table 3. A U.S. EPA position document on 2,4,5-TCP (U.S. Environmental Protection Agency 1978i) was prepared to accompany the August 2, 1978, Federal Register notice of rebuttable presumption against continued registration of 2,4,5-TCP products. The position document gives the following description of the known uses of this chemical: The largest use of 2,4,5-TCP is as a starting material in the manufacture of a series of industrial and agricultural chemicals, the most notable of which is the herbicide 2,4,5-T and its related products including silvex [2-(2,4,5-trichlorophenoxy) propionic acid], ronnel [0,0-dimethyl 0-(2,4,5-trichlorophenyl)-phosphorothioate], and the bactericide hexachlorophene. 2,4,5-TCP and its salts are used in the textile industry to preserve emulsions used in rayon spinning and silk yarns, in the adhesive industry to preseve polyvinyl acetate emulsions, in the leather industry as a hide preservative, and in the automotive industry to preserve rubber gaskets. The sodium salt is used as a preservative in adhesives derived from casein, as a constituent of metal cutting fluids and foundry core washes to prevent breakdown and spoilage, as a bactericide/fungicide in recirculating water in cooling towers, and as an algicide/slimicide in the pulp/paper manufacturing industry. There are some minor uses of 2,4,5-TCP and its salts in disinfectants which are of major importance relative to human exposure. These include use on swimming-pool-related surfaces; household sickroom equipment; food processing plarits and equipment; food contact surfaces; hospital rooms; sickroom equipment; and bathrooms (including shower stalls, urinals, floors, and toilet bowls). It is apparent, therefore, that all the uses of 2,4,5-TCP exploit the poisonous character of the compound and its derivatives. As a pesticide, it is subject to EPA registration in all of its applications except those associated with food processing. 27 �Manufacture-Only trace chlorination of chlorination of these production amounts of 2,4,5-trichlorophenol are created by direct phenol. It can be made in about 50 percent yield by re3,4-dichlorophenol (U.S. Patent Office 1956c). Neither of methods is in commercial use in this country. Domestic commercial production is accomplished through hydrolysis of 1,2,4,5-tetrachlorobenzene, which is a principal isomer produced by rechlorination of o-dichlorobenzene. Conversion of this chemical to the sodium salt of 2,4,5-TCP is a batch reaction with caustic soda. Subsequent neutralization with a mineral acid forms the product. The basic process is a typical application of the hydrolysis method of chlorophenol production described earlier. The reaction sequence is given below: -i- 2NaOH ,2.4.5-TETRACHLOROBENZENE 2.4.5-TRICHLOROPHENOL At least three variations of the basic process have been described in process patents specifically for production of 2,4,5-TCP, differing only in the solvents used and therefore in the conditions needed to drive the reaction to completion. The first patented process (U.S. Patent Office 1950) 28 �uses a solvent of ethylene glycol or propylene glycol at preferred temperatures of 170° to 180°C and pressures up to 20 lb/in.2. A second patent, the most recent, (U.S. Patent Office 1967b), describes the use of methanol as a solvent, with temperatures ranging from 160°2 to 220°C and with pressure less than 350 lb/in.2 (probably 50 to 200 lb/in. ). Both of these alcoholbased processes require 1 to 5 hours to complete. A third patent (U.S. Patent Office 1957b) describes the use of water as the reaction solvent. Use of water necessitates the most severe operating conditions: operating temperatures from 225° to 300°C and pressures from 400 to 1500 lb/in.2. This method permits greater production, since reaction time is reduced to no more than 1.5 hours and in some instances to as little as 6 minutes. In addition to its production efficiency, the water-based process eliminates the side reactions between caustic and the alcohol solvents, which form undesired impurity compounds. The process also improves product yield and eliminates solvent costs. It appears, however, that the high-temperature, high-pressure, and strongly alkaline conditions of the water-based process promote a continuation of the reaction, in which 2,4,5-TCP combines with itself to form 2,3,7,8-TCDD. The patent examples cited above are fairly old, and details of the current 2,4,5-TCP production methods are difficult to obtain. A 1978 EPA report on 2,4,5-TCP briefly describes present-day 2,4,5-TCP manufacture as a reaction of tetrachlorobenzene with caustic in the presence of methanol at 180°C under pressure. Although a final product purification step is described in the most recent patent example (U.S. Patent Office 1967b), the EPA report does not describe it. A more detailed estimate of current production methods is derived from fragmentary descriptions of both U.S. and foreign operations (Sidwell 1976; World Health Organization 1977; Fuller 1977; Whiteside 1977; Fadiman 1979; D. R. Watkins 1980). (One plant from which much of this information was derived ceased production of 2,4,5-TCP in 1979.) Figure 4 is a flow chart prepared from these sources, showing the most likely process details. In this processing scheme, alcohol and caustic are mixed and heated. Tetrachl orobenzene is added, an exothermic reaction begins, and cooling water is turned onto the reactor coils. After all the tetrachlorobenzene has been added, the batch is "aged"; during the aging period, sodium-2,4,5-trichlorophenate (Na-2,4,5.-TCP) is formed. Volatile compounds such as dimethyl ether also are formed during the aging step; these are vented from the reactor, along with small amounts of vaporized methanol. The presence of these flammable vapors presents a fire or explosion hazard, and the reaction vessel is usually enclosed in blastproof walls to minimize physical damage from any accident that may occur during the aging step. On completion of the reaction, the methanol is evaporated, condensed, and recycled. At the same time, water is added to keep the batch contents in solution. In this process, a toluene washing step is conducted to purify the product by removing some of the high-boiling impurities. Toluene condensed from the overhead of an auxiliary still is mixed into the cooled water solu29 �1,2,4,5TETRACHLOROBENZENE SODIUMHYDROXIDE AIR EMISSION ALCOHOL RECYCLE WATER- MIXING AND _ TOLUENE PHASE DISTILLATION SEPARATION TOLUENE"^ IMPURITIES TO CONVERSION PROCESS Na-2,4,5-TCP IN WATER HYDROCHLORIC ACID *-WASTEWATER EMISSION Figure 4. Flow chart for 2,4,5-TCP manufacture. 30 .SOLID WASTE �tion of Na-2,4,5-TCP. The mixture is then allowed to stand quietly so that the water and organic phases can separate into layers. The organic layer, containing impurities, is decanted and returned to the toluene still as feed. The water layer, containing partially purified Na-2,4,5-TCP, can be used directly to manufacture a herbicide derivative. Alternatively, hydrochloric acid can be added to neutralize the mixture. Acidic 2,4,5-TCP precipitates and is separated from the liquid by centrifugation. Many of the impurities created during this process, including 2,3,7,8TCDD, accumulate in the bottom of the toluene still. Still bottoms are removed periodically to be discarded. Toluene still bottoms have been identified as the source of at least one exposure of the public to dioxins, and also as the source of one of the highest concentrations of 2,3,7,8-TCDD (40 ppm) ever discovered in such wastes (Watkins 1979, 1980; Richards 1979a) (Analysis of this waste sample is fully described in Volume II of this report series.) As shown in Figure 4, the acidic 2,4,5-TCP is dried and either packaged for sale or used to manufacture other derivative products. One reference shows one or more stages of purification of the product after it is centrifuged from the water solution (World Health Organization 1977). One stage of high-vacuum distillation is conducted to create what is described as "agricultural grade 2,4,5-TCP." A second stage of distillation removes additional impurities to form "pharmaceutical grade 2,4,5-TCP." It is believed that all U.S. hexachlorophene is made from a distilled grade of this chemical. Process details concerning the only remaining 2,4,5-TCP plant in the United States have not been released. It was reported in 1967 that this plant (Dow Chemical Company, Midland, Michigan) was using the water-based process described in its 1955 patent (Sconce 1959; U.S. Patent Office 1957b), but this probably is not the case today. Another report states that the process is conducted with very careful temperature contro] to prevent the formation of dioxins (Sittig 1974). This source also indicates that still bottoms from the manufacture of 2,4,5-T at this plant are being discarded by incineration; therefore, a distillation is presumably being performed. It is not known whether these still bottoms are from a toluene washing still or from a product still. Production-Dow Chemical Company is apparently the only current producer of both 2,4,5-TCP and Na-2,4,5-TCP. Merck and Company has recently begun producing Na-2,4,5-TCP (SRI 1979). Current records related to the EPA Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) indicate that 42 companies, including Dow, are marketing 94 registered commercial products containing 2,4,5-TCP or its salts (U.S. Environmental Protection Agency 1978i). According to EPA sources, most, if not all, of these companies obtain the basic chemical from Dow (Reece 1978c). Former 2,4,5-TCP manufacturing sites are listed in Table 6 by location and owner. Details of the processes used by these former producers are 31 �TABLE 6. FORMER 2,4,5-TCP MANUFACTURING SITES' Owner Plant location Niagara Falls, New York Hooker Chemicals and Plastics (approximately 45 years ) Jacksonville, Arkansas Reasor-Hill Corp. (1946~61)c Hercules, Inc. (1961-71)c , Transvaal, Inc. (1971-78)°'° Vertac, Inc., Transvaal, (subsidiary} (Nov. 1978March 1979)° Verona, Missouri Northeastern Pharmaceuticals and Chemicals Co. Monmouth Junction, New Jersey Rhodia, Inc. Linden, New Jersey GAP Corp. Chicago, Illinois Nalco Chemical Co. Cleveland, Ohio Diamond Shamrock Corp. Unless otherwise noted, the information in this table was derived from Stanford Research Institute Directory of Chemical Producers, U.S., 1976-1979, and U.S. International Trade Commission Synthetic . Organic Chemicals, U.S. Production and Sales, 1968, 1974, 1976-78. ° Chemical Week 1979a. c Richards 1979a. 32 �not known; however, "toluene still bottoms" were said to be the source that created a dioxin exposure at Verona, Missouri, which indicates that the toluene washing step described above may have been used (see Section 4). The methanol-based process with a toluene washing stage was used by Vertac, Inc. (Watkins 1980). Current U.S. production figures for 2,4,5-TCP and its salts are not available (U.S. Environmental Protection Agency 1978i). In 1970, the estimated level of domestic production for 2,4,5-TCP and its derivatives was 50 million pounds (Crosby, Moilanen, and Wong 1973). In 1974, the reported annual world production of all chlorophenols and their salts was estimated to be 100,000 tons, or 200 million pounds (Nilsson et al. 1974). Chlorophenol Derivatives With Confirmed Dioxin Content The wide utilization of chlorophenols in chemical synthesis makes it virtually impossible to identify all the potential derivatives of this class of compounds. The following paragraphs outline the manufacture of derivatives that, upon analysis, have been reported to contain chlorinated dioxins. The products are all pesticides, which are usually made as only partially purified chemicals and are intended to be distributed rather broadly into the environment. 2,4-0, 2,4-DB, 2,4-DP and 2,4-DEP— The compound 2,4-dichlorophenoxyacetic acid (2,4-D) is a widely used herbicide and a close chemical relative of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) described later in this section. A 50:50 mixture of these two chemicals, known as "Herbicide Orange" (earlier called "Agent Orange"), was used as a defoliant during the Vietnam conflict. The chemical formula of 2,4-D is shown below. OCH2CO2H -Cl The herbicide 2,4-DB is 4-(2.,4-dichlorophenoxy) butyric acid; 2,4-DP is 2-(2,4-dichlorophenoxy) propionic acid; and 2,4-DEP is tris [2 - (2,4dichlorophenoxy) ethyl] phosphite; all are closely related chemically to 2,4-D. 33 �In 1972, Wool son, Thomas, and Ensor found hexachlorodioxin in one sample of 2,4-D at a level between 0.5 and 10 ppm. No other dioxins were observed. Twenty-three other 2,4-D samples, as well as three 2,4-DB and two 2,4-DEP samples were analyzed, but no dioxins were found at a 0.5 ppm limit of detection. Apparently, only tetra-, hexa-, hepta- and octachlorodioxins were sought in these analyses. The samples apparently were not analyzed for dichlorodioxins, which should be more likely to occur. According to the World Health Organization (1977), 2,4-D is widely used as a herbicide for broad!eaf weed control in cereal crops (wheat, corn, grain sorghum, rice, other small grains), sugar cane, and citrus fruits (lemons), and on turf, pastures and noncrop land. Food-related uses account for 58 percent of all 2,4-D used in the United States in 1975. Two manufacturing processes have been described for 2,4-D, only one of which starts with a chlorinated phenol. One process is a direct chlorination of phenoxyacetic acid (U.S. Patent Office 1949). The other process is a reaction between 2,4-dichlorophenol and chloroacetic acid (U.S. Patent Office 1958a). The second process is similar to the 2,4,5-T manufacturing process described in the following section and is also similar to the process used to make 2,4-DB (U.S. Patent Office 1963). Since many companies make 2,4-D and its esters and salts, both production processes may be in use, although it is claimed that chlorination of phenoxyacetic acid produces a higher yield and is a simpler process. In a batch reactor, phenoxyacetic acid is melted by heating it to 100°C. With continuous agitation, chlorine is bubbled through the molten chemical and the temperature is increased slowly to 150°C. A stream of dry air is passed through the reactor to sweep away the hydrogen chloride byproduct. When the calculated amount of chlorine has been added, the resulting mass is cooled, pulverized, and packaged. No solvent is used, no special recovery operation is needed, and product purification is unnecessary. If dioxins are created during this process, the mechanism of their formation is unknown. The second process involves reaction of 2,4-dichlorophenol with chloroacetic acid in a solvent mixture of water and sodium hydroxide. This process is said to be used by at least one large manufacturer (Sittig 1974). Heat is applied to the vessel, and the water is evaporated from the mixture. When the temperature begins to rise, indicating that most of the water has evaporated, heating is stopped and a fresh charge of cold acidified water is added. The product can be filtered from the mixture and dried; thisprocedure would form an impure product. Alternatively, the product can be extracted from the cooled mixture with a water-immiscible solvent and then separated from the solvent by distillation. This latter recovery method would probably create anhydrous organic wastes and therefore is probably in use by at least one company that has been reported to incinerate waste tars from 2,4-D manufacture (Sittig 1974). 34 �This chlorophenol-based process for making 2,4-D could create dioxins because it provides for an alkaline mixture of a dioxin precursor chemical in contact with hot heating surfaces. If the product is only filtered from the reaction mixture, the dioxin contaminants would be captured along with the product. If solvent extraction is employed, part of the dioxin would probably appear in wastes from the process and part would probably be captured with the product. The process for manufacture of 2,4-DB uses 2,4-dichlorophenol and gamma butyrolactone in a solvent mixture of dry butanol and nonane, with sodium hydroxide as a reaction aid. The chemical reactions are shown below: CH2CH2CH2COOH vx O A NaOH 2.4-DICHLOROPHENOL The ingredients are mixed and heated to a temperature of about 165°C for a period that may range from 1 to 24 hours. On completion of the reaction, dilute sulfuric acid is added and 2,4-DB precipitates; the precipitate is centrifuged from the mixtures, dried, and packaged.. Liquids from the centrifuge are allowed to stand quietly and separate into two liquid layers. The water fraction is discarded, and the organic layer is recycled to the subsequent reaction batch. Any water that is brought into the reactor is removed by distillation before the next reaction is started. It is possible that dioxins could be produced in this process by the mixture of 2,4-dichlorophenol with sodium hydroxide being brought into contact with a hot surface. Product recovery methods are such that any dioxins formed would either be removed as solids along with the product or be recycled to the succeeding batch. 35 �Commercial production of 2,4-D in the United States started in 1944 and by the mid-1960's had peaked at 36 million kg (World Health Organization 1977). After the use of Herbicide Orange was discontinued, production dropped. Production in 1974 is estimated to have been 27 million kg (World Health Organization 1977). Production figures for 2,4-DB and 2,4-DEP are not available. The current basic producers of 2,4-D and 2,4-DB acids, esters, and salts as reported by Stanford Research Institute in 1978 are listed in Table 7. Former producers or production sites are listed in Table 8. No current producers of 2,4-DEP are listed in the Stanford Research Institute publication of 1978. Sesone-The chemical name for the pesticide sesone is 2-(2,4-dichlorophenoxy) ethyl sodium sulfate. The only sample known to have been analyzed for dioxins contained 0.5 to 10 ppm hexachlorodioxin (Helling et al. 1973). No tetra-, hepta-, or octachlorodioxins were detected (0.5 ppm detection level). Analysis apparently was not performed for di-, tri-, or pentachlorodioxins. Sesone is made from 2,4-dichlorophenol by boiling it for several hours in a water solution of beta-chloroethyl-sodiurn sulfate and sodium hydroxide. The following are the chemical reactions of the process: OCHjCHjOSOaNa © © CICH2CH2OSO3Na NaOH * In more detail, the straight-chain reactant is made by combining ethylene chlorohydrin and chlorosulfonic acid in a refrigerated water solution (U.S. Patent Office 1958c). After partial neutralization with sodium hydroxide, 2,4-dichlorophenol is added and the mixture is boiled for about 15 hours. According to the patent example, the mixture is probably not purified; it is simply spray-dried to form a usable product. It could be purified by repeated extractions with hot alcohol to separate the sodium sulfate impurity. 36 �TABLE 7. CURRENT BASIC PRODUCERS OF 2.4-D AND 2,4-DB ACIDS, ESTERS, AND SALTS3 Pesticide 2,4-D and esters and salts 2,4-DB and salts Source: Company Production location Dow Chemical Company Fallek-Lankro Corp. Imperial, Inc. North American Phillips Corp., Thompson-Hayward Chemical Co., subsidiary PBI-Gordon Corp. Rhodia, Inc. Riverdale Chemical Co. Union Carbide Corp. Amchem Products, Inc. subsidiary Vertac, Inc. Transvaal, Inc., subsidiary Rhodia, Inc. Union Carbide Corp. Amchem Products, Inc. subsidiary Stanford Research Institute 1978. 37 Midland, Michigan Tuscaloosa, Alabama Shenandoah, Iowa Kansas City, Kansas Kansas City, Kansas Portland, Oregon St. Joseph, Missouri Chicago Heights, Illinois Chicago Heights, Illinois Ambler, Pennsylvania Fremont, California Jacksonville, Arkansas Portland, Oregon Ambler, Pennsylvania �TABLE 8. FORMER BASIC PRODUCERS OF 2,4-D AND 2,4-DB ACIDS, ESTERS, AND SALTS9 Pesticide formerly reported produced 2,4-D acid, esters, and salts 2,4-DB and salts Company Production location Chempar Miller Chemical subsidiary of Alco Standards Rhodia Portland, Oregon Whiteford, Maryland North Kansas City, Kansas St. Paul/Minneapolis, Minnesota Thompson Chemical St. Louis, Missouri Woodbury, subsidiary Orlando, Florida of Comutrix Rhodia North Kansas City, Missouri St. Paul/Minneapolis, Minnesota Source: Dryden et al. 1980 (Volume III of this report series). 38 �The manufacture of sesone meets all of the requirements for promotion of the formation of 2,7-DCDD. Both the raw material and the final product contain a chlorine atom ortho to a ring-connected oxygen atom, and the mixture is heated in the presence of sodium hydroxide. Although overall reaction temperature is only slightly above 100°C, it could be higher at the heating surfaces. The volume of sesone produced annually is not known. Only nine commercial products containing the herbicide are currently registered as pesticides with EPA. DMPA-The chemical name for DMPA is 0-(2,4-dichlorophenyl) 0-methyl isopropylphosphoramidothioate (Merck Index 1978). Some of the relatively higher chlorodioxins (hexa-, hepta- and/or octachlorodioxins) were detected at ppm levels in at least one DMPA sample analyzed in 1972 (Helling et al. 1973). The following is the structure for DMPA. O-P-NHCH(CH3)2 OCH3 DMPA Synthesis of this molecule involves the methanolysis of 0-(2,4-dichlorophenyl) phosphorodichloridothioate, which is made through the phosphoralation of dichlorophenol (U.S. Patent Office 1960; Blair, Kaner, and Kenaga 1963). DMPA is known commercially as Zytron, K-22023, and Dow 1329 (Merck Index 1978). It is useful as an insecticide, especially against hpuseflies (Blair, Kaner, and Kenaga 1963). It is also useful as a herbicide for controlling the growth of undesirable plants (U.S. Patent Office 1963; Merck Index 1978). DMPA is not believed to be produced in large amounts. Currently three companies - Dow Chemical Company, Techne Corp., and Rhodia Chemical Company - have each registered one DMPA pesticide product with EPA (U.S. Environmental Protection Agency 1978f). Trichlorophenpl Derivatives-As mentioned earlier, the largest use of 2,4,5-TCP is as a starting material in the manufacture of several pesticide and bactericide products. Table 9 lists the known 2,4,5-TCP derivatives, their specific uses, and the companies which have recently been reported to produce them. 39 �TABLE 9. DERIVATIVES OF 2,4,5-TRICHLORQPHENOL AND THEIR RECENT (1978) PRODUCERS3 Derivative 2,4,5-T and esters and salts Use Current producers Production location Herbicide for Dow Chemical, U.S.A. Midland, Michigan woody plant control North American Phillips Kansas City, Kansas Corp., Thompson-Hayward Chemical Co. , subsidiary PBI-Gordon Corp. Riverdale Chemical Co. Chicago Heights, Illinois Rhodia, Inc.b Portland, Oregon or St. Joseph, Missouri Union Carbide Corp., Amchem Products, Inc. subsidiary Ambler, Pennsylvania Fremont, California St. Joseph, Missouri Vertac, Inc. Transvaal, Inc. subsidiary Si 1 vex and esters and salts (Fenoprop) Kansas City, Kansas Jacksonville, Arkansas Herbicide for Dow Chemical, U.S.A. Midland, Michigan woody plant control; plant North American Phillips Kansas City, Kansas hormone Corp., Thompson-Hayward Chemical Co., subsidiary Riverdale Chemical Co. Chicago Heights, Illinois Vertac, Inc., Transvaal, Inc., subsidiary Jacksonville, Arkansas Dow Chemical, U.S.A.' Midland, Michigan Ronnel Insecticide (Fenchlorfos) Dow Chemical, U.S.A. Midland, Michigan Hexachlorophene Givaudan Corporation Clifton, New Jersey Erbon Herbicide, weed and grass killer Bactericide . Source: 1978 Directory of Chemical Producers, United States. Rhodia is not listed in the 1978 Directory of Chemical Producers U.S.A., but has been recently cited by the EPA (Blum 1979) and the news media (Wall Street Journal 1979 and Environmental Reporter (1979a) as a manufacturer of 2,4,5-T. In 1979 this company ceased production of 2,4,5-trichlorophenol for j subsequent conversion to 2,4,5-T and silvex. Although erbon is not listed in the 1978 Directory of Chemical Producers, several companies including Dow Chemical have registered erbon pesticide products with EPA. Dow 1s most likely the basic producer of the herbicide. 40 �274,5-T--The chemical name for 2,4,5-T is 2,4,5-trichlorophenoxyacetic acid and it is the most important derivative of 2,4,5-trichlorophenol. It has been a registered pesticide for about 30 years (U.S. Environmental Protection Agency 1978h) and was used primarily as a herbicide for controlling woody plant growth. 2,4,5-T is best known for its combined use with 2,4-D as Herbicide Orange, which was used extensively by the U.S. military as a defoliant during the Vietnam conflict. When the toxicity of this formulation became apparent, the government suspended all further military use of Herbicide Orange, and in 1970 stopped many registered domestic uses including application to lakes, ponds, ditch banks, homesites, recreational areas, and most food crops (World Health Organization 1977). Until 1979, domestic commercial use of 2,4,5-T continued for control of brush and other hardwood in forestry management and on power transmission right-of-ways, rangelands, rice fields, and turfs. Most of these uses have now been suspended (Blum 1979). Parts-per-million quantities of dioxins have been reported in 2,4,5-T since 1970 (World Health Organization 1977). A study (Woolson, Thomas, and Ensor 1972; Kearney et al. 1973b; Helling et al. 1973) of samples manufactured between 1950 and 1970 found 0.5 to 10 ppm TCDD's in 7 of 42 samples tested; another 13 samples contained 10 to 100 ppm TCDD's. Hexa-CDD's were found in 4 of the 42 samples. The limit of detection in this study was reported as 0.5 ppm for each dioxin. Most samples came from a company that no longer produces 2,4,5-T. Elvidge (1971) reported that five of six 2,4,5-T samples contained TCDD's at levels ranging from 0.1 to 0.5 ppm. The dioxin was present in two 2,4,5-T ester samples at 0.2 to 0.3 ppm. TCDD's were also found in two 2,4,5-T ester formulations at 0.1 and 0.2 ppm. The level of detection was 0.05 ppm. Storherr et al. (1971) reported finding 0.1 to 55 ppm TCDD's in seven of eight samples of technical 2,4,5-T. Analysis of 200 samples of Herbicide Orange for TCDD's by the U.S. Air Force showed 0.5 ppm or less in 136 samples and more than 0.5 ppm in the remainder. The highest level was 47 ppm (Kearney et al. 1973). Early in 1976, investigators at Wright State University analyzed 264 samples of U.S. Air Force stocks of Herbicide Orange and found TCDD's at levels ranging from 0.02 to 54 ppm (Tiernan 1975). The level of detection was 0.02 ppm. 2,4,5-T with a TCDD isomer content of less than 0.1 ppm is now commercially available from U.S. producers (U.S. Environmental Protection Agency 1978h). Commercial 2,4,5-T guaranteed to contain less than 0.05 ppm TCDD's is available from foreign producers (World Health Organization 1977). The commercial method of producing 2,4,5-T is briefly described in EPA Position Document .1 (April 1978) on this pesticide (U.S. Environmental Protection Agency i978h). According to this document, 2,4,5-TCP is reacted with chloroacetic acid under alkaline conditions. Subsequent addition of sulfuric acid produces 2,4,5-T (acidic form), which can then be reacted with a variety of alcohols or amines to produce 2,4,5-T esters and amine salts. The chemical reactions are as follows: 41 �Cl ONa Cl Cl ClCH2COONa Na-2.4,5-tCP HCI COOH HCl A more complete description of the 2,4,5-T production process appears in a patent record (U.S. Patent Office 1958a). Sodium 2,4,5-trichlorophenate is most often delivered to the process as a water solution containing excess sodium hydroxide directly from the Na-2,4,5-TCP manufacturing process. Amyl or isoamyl alcohol, or a mixture of these solvents, is added, and heat is applied to remove water as an azeotrope. When all water has been removed, chloroacetic acid is added to initiate the reaction that produces sodium 2,4,5-trichlorophenoxyacetate (Na-2,4,5-T) and sodium chloride. The reaction proceeds under total reflux for about 1.5 hours at 110° to 130°C and atmospheric pressure. An excess of sodium hydroxide is present during the reaction. Water is then fed into the reactor and distillation is resumed, this time to remove the amyl alcohol and replace it with water. At the end of the second distillation, the reaction mixture consists of Na-2,4,5-T dissolved in a sodium chloride brine. The patent example incorporates a purification step that may not be conducted in commercial practice. Near the end of the second distillation, activated carbon may be added to adsorb heavy or colored impurities, which would include dioxins that were present in the Na-2,4,5-TCP feedstock. On completion of the second distillation, the carbon would be filtered from the mixture and discarded. If this step is conducted, the process will generate a waste carbon sludge likely to be contaminated with dioxins. If this step is not conducted, any dioxins present are likely to be carried through the process and appear in the final product. In either variation, the next step is to add acid to neutralize the residual caustic and to form insoluble 2,4,5-T. The product is then filtered or centrifuged from the waste brine, dried, and packaged for sale. The filtrate from this step should contain only soluble sodium chloride and sulfate, excess neutralization acid, and very small quantities of organic matter; it is discarded as a liquid waste. 42 �The patent that describes the manufacture of 2,4,5-T is unusually detailed and indicates that the temperature during the process is never above 140°C, which is lower than the temperature believed to be necessary to create dioxins. Any dioxins that enter with the feed will appear either in the product or in process wastes, but additional dioxins probably are not formed during 2,4,5-T manufacture. Even during abnormal operation or an industrial fire, it would be difficult for the temperature to exceed by far the low boiling point of amyl alcohol, since all operations take place in unpressurized vessels. The highest production of 2,4,5-T occurred between 1960 and 1968, when it peaked at 16 million pounds per year (World Health Organization 1977). Between 1960 and 1970 a total of 106.3 million pounds was produced domestically (Kearney et al. 1973b). Production declined during the 1970's because of restrictions on use of the compound. In 1978 the annual U.S. usage of 2,4,5-T was estimated at only 5 million pounds (American Broadcasting Co. 1978). Because of EPA's March 1979 emergency ban on most of the remaining uses (Blum 1979), current usage is believed to be even less, probably less than 2 million pounds per year. 2,4,5-T may be produced and formulated in several forms as salts and esters of the acid. The low-volatility esters have been used most often. Emulsifiable concentrates of 2,4,5-T salts and esters contain 2 to 6 pounds per gallon of the acid equivalent; oil-soluble concentrates contain 4 to 6 pounds of active ingredient per gallon (U.S. Environmental Protection Agency 1978h). Until 1979, this herbicide was probably produced by the seven companies shown in Table 10. Over a hundred companies were recently marketing more than 400 formulated pesticide products containing 2,4,5-T (U.S. Environmental Protection Agency 1978h). Si 1 vex—Si 1 vex is a family of compounds that act as hormones to plants and can be used as specific herbicides. Formulations containing these materials were used for control of woody plants on uncropped land and for control of weeds on residential lawns until 1979, when sales of most products containing si 1 vex were halted (Blum 1979). Si 1 vex is still being used on noncrop areas, on rangelands and orchards, and on rice and sugar cane (Toxic Materials News 1979b; Chemical Regulation Reporter 1979c). The chemical name for si 1 vex acid is 2-(2,4,5-trichlorophenoxy) propionic acid. It is also known as Fenoprop, 2,4,5-TP, and 2,4,5-TCPPA. acid. Si 1 vex is available either as the acid or as esters and salts of the The low-volatility esters are probably the form most widely used. TCDD's were detected (1.4 ppm) in one of seven si 1 vex samples manufactured between 1965 and 1970 and analyzed in 1972; no other dioxins were detected (Woolson, Thomas, and Ensor 1972; Kearney et al. 1973b). 43 �TABLE 10. FORMER PRODUCERS OF 2,4,5-T (Prior to 1978) Location Company Chempar Portland, Oregon Diamond-Shamrock Cleveland, Ohio Hoffman-Taff, Inc. Springfield, Missouri Hercules, Inc. Wilmington, Delaware Monsanto Co. St. Louis, Missouri Rorer-Amchem Ambler, Pennsylvania Fremont, California St. Joseph, Missouri Jacksonville, Arkansas Wm. T. Thompson Co., Thompson Chemical Div. St. Louis, Missouri Sources: SRI Directory of Chemical Producers, United States, 1976 and 1977. United States Tariff Commission/United States International Trade Commission. Synthetic Organic Chemicals, United States Production and Sales, 1968, 1974, 1976, and 1977. 44 �The following are recent producers of silvex as listed in the 1978 Stanford Research Institute Directory of Chemical Producers: Dow Chemical U.S.A. - Midland, Michigan North American Phillips, Thompson Hayward Chemical, subsidiary - Kansas City, Kansas Riverdale Chemical - Chicago Heights, Illinois Vertac, Inc., Transvaal, Inc., subsidiary Jacksonville, Arkansas Hercules, Inc., of Wilmington, Delaware, is a former producer (U.S. Tariff Commission 1968). The 1978 EPA pesticide files indicate that more than 300 products or formulations containing silvex are registered (U.S. Environmental Protection Agency 1978f). Silvex manufacture is more complex than that of other 2,4,5-TCP derivatives. The compounds sold commercially are usually complex esters, made from a specialized alcohol and silvex acid. The final manufacture of the ester is well documented in a process patent (U.S. Patent Office 1956a), as is the manufacture of the specialized alcohol. No definitive information has been found, however, on manufacture of the silvex acid, probably because compounds of this type can be manufactured by a long-established chemical reaction that is used in many categories of the organic chemical industry (J. Am. Chem. Soc. 1960). Silvex acid would be the source of any dioxins in commercial silvex products. The figure below illustrates the most likely chemical reaction that would form the silvex acid and also shows the subsequent esterification, as described in the patent. COOCH3 CH3-CH OH 9 CH3CHCOOCH3 Cl 2,4,5-TCP AQUEOUS ACID OC.H, HOCH2CH2CH OC3H7 H2SO4 SILVEX ACID SILVEX ESTER 45 �In the first step, 2,4,5-TCP Is probably brought into reaction with the methyl ester of 2-chloropropionic acid, with methanol as the solvent and sodium methoxide as a reaction aid. This reaction would occur approximately at the boiling temperature of methanol, which is 65°C. The resulting compound would probably be separated from the reaction mixture by treatment with acidified water followed by extraction with a chlorinated hydrocarbon. The addition of more acidified water to the extractant and a subsequent evaporation at a temperature near 100°C would hydrolyze the intermediate compound and also would drive off the chlorinated hydrocarbon for recycle and the methanol byproduct to be reclaimed for other uses. The resulting compound is 2-(2,4,5-trichlorophenoxy)-propionic acid, which is known to be a reactant in the subsequent processing (U.S. Patent Office 1956a). Other methods could be used to prepare this intermediate acid, but none of them would utilize high temperatures or unusual solvents. The use of a strongly alkaline hydrolysis step, rather than an acidic medium, is possible. In any method, the last step is probably another solvent extraction using 1,2-dichloroethane to prepare the mixture for the next operation. Silvex acid can be converted to various esters by using selected ether alcohols. The esterification steps are identical except for variations in the alcohol raw material. In a solvent of 1,2-dichloroethane, with concentrated sulfuric acid as a reaction aid, the intermediate acid is mixed with an ether alcohol. In the example shown on the previous page, butoxypropoxypropanol is used. The mixture is held at about 95°C for about 7 hours. During this period, the water formed in the reaction is removed by passing the reflux condensate through a decanter. At the end of the reaction, the product is present as an insoluble precipitate, which is filtered from the mixture, washed with sodium carbonate solution, and vacuum-dried at about 90°C. Although complete data are unavailable, no information indicates that temperatures greater than 100°C would occur at any step in the manufacture of acidic silvex or its esters. It is therefore unlikely that dioxin compounds would be created as side reaction products. Absence of detailed information makes it impossible to establish whether dioxin contamination would carry through from the 2,4,5-TCP raw material into the final product. Theoretical considerations do not permit an estimation of the degree of purification required by the various intermediate compounds. Probably, as noted above, at least two solvent extraction operations are used to separate the principal processing materials from water solutions. Since TCDD's are very slightly soluble in chlorinated organic solvents, some could be carried through these operations, but most should be rejected. Erbon--Very little information is available on erbon, which is derived from 2,4,5-trichlorophenol. Analysis of one erbon sample produced in 1970 indicated more than 10 ppm octachlorodioxin (Woolson, Thomas, and Ensor 1972). Tetra-, penta-, hexa-, and heptachlorodioxins were not detected (0.5 ppm limit of detection). 46 �In 1978, nine companies had registered 17 products containing erbon (U.S. Environmental Protection Agency 1978). Dow is probably the only producer of the basic chemical. The other companies are most likely formulators who obtain their basic erbon ingredient from Dow. The volume of erbon produced annually is not known. This herbicide is an ester based on 2,4,5-TCP. Although the initial manufacturing step is not reported, the first intermediate is almost identical to that used to make sesin. General organic chemical references indicate that it is probably made by an initial reaction of 2,4,5-TCP with ethylene chlorohydrin (March 1968). Water is the most likely solvent, made strongly alkaline with sodium hydroxide, and the intermediate probably precipitates on addition of acid and is filtered from the solution and dried. A process patent (U.S. Patent Office 1956b) discloses that the second reaction step is a combination of the intermediate with 2,2-dichloropropionic acid in a solution of ethylene dichloride (1,2-dichloroethane), with addition of a small amount of concentrated sulfuric acid to remove the water formed in the reaction. These chemical reactions are shown by the following sequence drawing: OCH2CH2OH CICH2CH2OH Cl NaOH Cl CH3CCI2COOH H2S04 O I! OCH2CH2O-C-CCI2CH3 47 �The resulting reaction mixture is partially purified by washing with water and is then fractionally distilled under vacuum to recover ethylene dichloride for recycle and possibly to separate the product from any impurities. The first step of the reaction is the one that could possibly form dioxins. Both the raw material and the resulting intermediate contain a chlorine atom ortho to a ring-connected oxygen atom, and the mixture is heated with sodium hydroxide. Temperatures are not high, however, since water is probably the solvent used and this simple reaction ordinarily does not require application of pressure. Dioxin formation could occur at the surface of steam coils if high-pressure steam is used for distillation. Apparently no operation other than the final distillation would remove any dioxin contamination from this material. Since the most likely impurities would be more volatile than the final ester, even the distillation may not serve to isolate dioxins into a waste stream. Most dioxins either formed by the process or present in the raw material would probably be collected with the final product. Ronne1--The chemical name of ronnel is 0,0-dimethyl 0-(2,4,5-trichlorophenyl) phosphoroate. This insecticide is also known by such names as fenchlorfos, Trolene, Etrolene, Nankor, Korlan, Viozene, and Ectoral (Merck Index 1978). Ronnel is effective in the control of roaches, flies, screw worms, and cattle grubs (Merck Index 1978). In 1972, highly chlorinated dioxins were detected at ppm levels in an unknown number of ronnel samples (Woolson, Thomas, and Ensor 1972). The manufacture of ronnel is a two-step process (U.S. Patent Office 1952) in which Na-2,4,5-TCP is reacted first with thiophosphoryl chloride, then with sodium methoxide. The chemical reactions are shown below: ONa NaOCH3 PSCI3 In the first step, dry Na-2,4,5-TCP is added to an excess of thiophosphoryl chloride (2 to 4 times the theoretical amount) and heated 48 �slightly, perhaps to 80°C. Sodium chloride is formed as an insoluble precipitate; it is filtered from the mixture and discarded. The clear filtrate is vacuum-distilled to recover the excess thiophosphoryl chloride for recycle and to fractionally separate the intermediate from side reaction impurities. In a separate reaction vessel, metallic sodium is mixed with methanol. Hydrogen gas is liberated, creating a methanolic solution of sodium methoxide. This solution is mixed slowly with the purified intermediate while the mixture is maintained at approximately room temperature with noncontact cooling water. When measured amounts of both reactants have been combined, the mixture is held for a period of time to ensure completion of the reaction. A nonreactive organic solvent is then used to extract the product from a mixture of methanol, excess sodium methoxide, and byproduct sodium chloride. Suitable extraction solvents are carbon tetrachloride, methylene dichloride, and diethyl ether. The extraction solvent is decanted from the mixture, washed with water solutions of sodium hydroxide, and fractionally vacuum-distilled to separate the extraction solvent for recycle and to separate ronnel from side reaction byproducts. Throughout this process, the temperature probably does not exceed 150°C'. The highest temperature probably occurs in the base of the final distillation column. In theory, additional dioxins are not likely to be created by this process because of the absence of high temperature and pressure, although all other conditions meet the requirements for formation of 2,3,7,8-TCDD. It appears even less likely that dioxins originally present in the Na-2,4,5-TCP raw material would be carried through into the product. If all the steps outlined above are properly conducted, some of the operations might isolate dioxins into waste streams. The solubility of dioxins in thiophosphoryl chloride is unknown; if they are insoluble, they would be removed with the first filtration. Because the solubility of dioxins in chlorinated methanes is very slight (0.37 g/liter for TCDD in chloroform), much of the dioxin present would not be captured by the extraction solvent and would be carried away with the methanol reaction solvent. Distillations afford two other opportunities to isolate dioxin contaminants into waste organic fractions. Although the probability of dioxins carrying through into the final product appears slight, definitive information is not recorded. Ronnel is reportedly produced by only one company - Dow Chemical Co., Midland, Michigan (Stanford Research Institute 1978). Annual production volume is not known. It is found in over 300 pesticide formulations registered by more than 100 companies. 49 �Chlorophenol Derivatives With Unconfirmed Pioxln Content This subsection deals with several other chlorophenol derivatives that may contain dioxins. The compounds discussed include those that have been analyzed for dioxin content with negative results and also those for which analytical data have not been reported. Hexachlorophene-Hexachlorophene is known chemically as either bis-(3,4,6,-trichloro2-hydroxyphenyl) methane, or 2,2'-methylene-bis (3,4,6-trichlorophenol). It is also known commercially as G-ll (Cosmetic, Toiletry, and Fragrance Association, Inc. 1977). Hexachlorophene is an effective bactericide and fungicide. Prior to 1972 it was widely advertised and distributed as an active constituent of popular skin cleansers, soaps, shampoos, deodorants, creams, and toothpastes (Wade 1971; U.S. Dept. HEW 1978). Although its use has been considerably restricted by the Food and Drug Administration, it still may be used as a preservative for cosmetics and over-the-counter drugs; the concentration is restricted to 0.1 percent in these products. Skin cleansers containing higher levels may also be sold but only as ethical Pharmaceuticals, available by medical prescriptions (U.S. Code of Federal Regulations Title 21 1978). As an agricultural pesticide, hexachlorophene is a constituent of formulations used on three vegetables and on some ornamental plants for control of mildew and bacterial spot. It is also used in limited industrial and household applications as a disinfectant. The grade of hexachlorophene produced today is reported to contain less than 15 ug/kg (<15 ppb) 2,3,7,8-TCDD (World Health Organization 1977). In a 1972 analysis, dioxins could not be detected in hexachlorophene at a detection limit of 0.5 mg/kg (0.5 ppm) (Helling et al. 1973). Four process patents have been issued on manufacture of hexachlorophene, and all are variations of the following chemical reaction: H+ H2C=O Cl 2,4,5-TCP Cl HEXACHLOROPHENE 50 �Hexachlorophene Is formed by reacting one molecule of formaldehyde with two molecules of 2,4,5-TCP at elevated temperatures in the presence of an acid catalyst (Moye 1972). The patented processes differ in temperature, reaction time, order of reagent additions, reaction solvents, and other physical parameters. In the first process, patented in 1941, methanol is the solvent and large amounts of concentrated sulfuric acid are used to bind the water that is formed as a reaction byproduct; the process takes place at 0° to 5°C over a 24-hour period (U.S. Patent Office 1941). A second patent issued in 1948 discloses that the methanol solvent is eliminated and the reaction is conducted with paraformaldehyde at an elevated temperature (135°C) over a 30-minute period (U.S. Patent Office 1948). A 1957 patent reintroduces a solvent, which is one of several chlorinated hydrocarbons (U.S. Patent Office 1957d). Temperature is 50° to 100°C, and reaction time is 2 to 3 hours. Oleum (sulfuric acid plus S03) is used as the catalyst and concentrated sulfuric acid is recovered as the byproduct. Finally, a 1971 patent revises the order of reagent addition and also emphasizes the chemical reaction mechanism (U.S. Patent Office 1971). This last-mentioned process is probably the one in present use; its processing sequence is shown in Figure 5. Patent information indicates that older manufacturing methods probably reclaimed the product from the reaction mixture by neutralizing the sulfuric acid with sodium hydroxide, which would have created a rather large amount of brine waste. In modern processes, conditions are probably maintained so that the residual sulfuric acid separates as a distinct liquid layer when agitation of the mixture is stopped after completion of the reaction. This acid, which contains the water formed during the reaction, is decanted from the mixture; it is strong enough to be used elsewhere in the plant complex, although it probably cannot be used in subsequent hexachlorophene batches. In the patent examples, the organic layer that remains after the acid is removed is mixed with activated carbon, which is then filtered from solution. The purpose of this treatment is to remove colored impurities. The clear filtrate is then chilled to approximately 0°C; crystals of hexachlorophene precipitate and are filtered from solution, dried, and packaged. The filtrate, which would contain some hexachlorophene, is probably directly recycled for use in succeeding batches. There is no indication that dioxins would be formed during the production of hexachlorophene, since highly acidic conditions are maintained throughout the process and temperatures are well below those known to be needed for dioxin reactions (Kimbrough 1974). If dioxins are found in hexachlorophene, the most likely explanation for their presence is that contamination in the 2,4,5-TCP raw material is carried through into the final product. In a situation identical to that of the 2,4,5-T process, the patent descriptions show the possiblity of activated carbon adsorption, which could cause accumulation of dioxins into an extremely hazardous waste. If carbon adsorption is not used in commercial practice or if it is not totally effective, any dioxins in the raw material will either appear in the 51 �2,4.5TRICHLOROPHENOL SULFURIC ACIDAND SO- fr '' CHLORINATED HYDROCARBON FORMALDEHYDE ~ ~ REACTION DECANTATION SULFURIC ACID BYPRODUCT ACTIVATED CARBON CARBON TREATMENT I CENTRIFUGATION WASTE SLUDGE RECYCLE RECYCLE Figure 5. Flow chart for hexachlorophene manufacture. 52 �hexachlorophene product or be recycled to succeeding batches. Although dioxins are not known to be soluble in sulfuric acid, they might be carried out of the process with the acid byproduct; if this were the case, dioxins could then appear in other products of the plant in which the sulfuric acid is utilized. Givaudan Corporation in Clifton, New Jersey, is apparently the only active U.S. producer of hexachlorophene. Until 1976, the 2,4,5-TCP for hexachlorophene manufacture was produced by Givaudan's ICMESA plant in Seveso, Italy, and shipped to New Jersey for conversion. In 1976, Wright State University analyzed two representative samples of this trichlorophenol and found 1.8 and 1.9 ppb TCDD's (Tiernan 1976). An accident in 1976 closed the ICMESA plant and eliminated Givaudan 1 s primary supply of 2,4,5-TCP. (For further details of the ICMESA incident see Section 4, p. 77.) It is now believed that all the 2,4,5-TCP for hexachlorophene manufacture is supplied by Dow Chemical Company and that Givaudan specifies an extremely low dioxin content. In 1978, five waste samples from the Clifton plant were analyzed for chlorinated dioxins. None were found at a 0.1 ppm level of detection (U.S. Environmental Protection Agency 1978d). Subsequent analysis of three of these samples found no TCDD's at 0.1 or less ppb (see Volume II of this series). .About 400 commercial products containing hexachlorophene have been marketed recently in pesticide, drug, cosmetic, and other germicidal formulations. The annual production volume of the germicide is not reported. Bithionol-Bithionol (2,2'-thio-bis[4,6-dichlorophenol]) is an antimicrobial agent that was approved at one time for drug use by the U.S. Food and Drug Administration. This approval was withdrawn in October 1967 because the chemical was found to produce photosensitivity among users (Kimbrough 1974; Merck Index 1978). The U.S. EPA continues to approve its use as a pesticide in three animal shampoo formulations. These formulated bithionol products may no longer be actively marketed, however, because the single basic source of this chemical (Sterling Drug's Hilton Davis Chemical Co.) apparently no longer produces it (Chem Sources 1975; Stanford Research Institute 1978). The manufacture of bithionol is a one-step reaction between 2,4-dichlorophenpl and sulfur dichloride (U.S. Patent Office 1962; U.S. Patent Office 1958b). Carbon tetrachloride is used as the solvent, and a small amount of aluminum chloride serves as the catalyst. Bithionol is formed in a reaction at about 50°C; batch time is about 2 hours. The chemical reaction is shown belov,. AICI3 ci BITHIONOL 53 �Two methods of product recovery are outlined in one process patent (U.S. Patent Office 1958b). In one method, water is added and impure bithionol precipitates. To form a crude product, it is necessary only to filter the solids from the mixture and wash them several times in water and cold carbon tetrachloride. They are then dried and packaged. Alternatively, to recover a purified product, water is added and the mixture is distilled to remove the carbon tetrachloride for recycle. Bithionol collects as an organic sediment, which is separated from the water solution by decantation, washed with water and sodium bicarbonate, vacuumdried, redissolved in hot chlorobenzene, filtered, chilled to precipitate bithionol, and again filtered. A separate patent outlines a procedure for forming metallic salts of bithionol, which are compounds that permanently impregnate cotton fabrics with disinfectant properties (U.S. Patent Office 1962). The process uses sodium hydroxide and various metallic salts in room-temperature reactions, with water as the solvent. This manufacturing operation apparently provides no potential for production of dioxins by the known process of dioxin formation. In the manufacture of crude bithionol, there is no opportunity to reject any dioxins that may be present in the 2,4-dichlorophenol raw material. They would be carried through into the final product. If bithionol is purified by the process outlined above, one filtration operation would remove compounds that are insoluble in hot chlorobenzene. Some dioxins, however, are slightly soluble in this solvent and thus might persist even in purified bithionol or its salts. Sesin-Sesin is an ester based on 2,4-dichlorophenol. The manufacture is similar to that of erbon, a 2,4,5-TCP-based herbicide described earlier. Although details of the first process step have not been reported, general organic chemical references indicate that sesin manufacture probably begins by a reaction between 2,4-dichlorophenol and ethylene chlorohydrin, as shown in the reaction sequence on the following page (March 1968). Water is the most likely solvent, made strongly alkaline with sodium hydroxide, and the intermediate probably precipitates on addition of acid and is filtered from solution and dried. A process patent discloses that the second reaction step is a combination of the intermediate with benzoic acid (U.S. Patent Office 1956d). Xylene is the solvent, and a small amount of sulfuric acid is used to remove the water formed in the reaction. The resulting reaction mixture is neutralized with sodium carbonate and is then fractionally distilled under vacuum to recover the xylene for recycle and possibly to separate the product from any impurities. 54 �OCH2CH2OH OH CICH2CH2OH NaOH Cl C02H 2,4-DICHLOROPHENOL H2SO4 « // OCH2CH2O-C--</ SESIN The first step of the reaction is the one that could possibly form dioxins. Both the raw material and the resulting intermediate contain a chlorine atom ortho to a ring-connected oxygen atom, and the mixture is heated with sodium hydroxide. High temperature is not present, however. Since water is probably the solvent, this simple reaction would not ordi, „ _ , , narily require application of pressure. Dioxin formation could occur at the surface of steam coils if high-pressure steam is used for distillation. Apparently no operation other than the final distillation would remove any dioxin contamination from this material. Even this distillation may not isolate dioxins into a waste stream. Most dioxins either formed by the process or present in the raw material would probably be collected with the final product. Triclofenol Piperazine— A pharmaceutical compound can be made from commercial 2,4,5-trichlorophenol for use as an anthelmintic (deworming medication) (U.S.) Patent Office 1961a; Short and Elslager 1962). The research and animal tests of this drug were conducted prior to 1962 with unpurified commercial-grade 2,4,5-TCP. The drug was made by dissolving the chlorophenol in warm benzene and adding 55 �a measured quantity of piperazine. The resulting solution was filtered to remove insoluble matter, diluted with petroleum ether, and chilled. Crystals of the drug precipitated and were filtered from the mixture, washed with petroleum ether, dried, and packaged in gelatin capsules. If this drug is being manufactured, the volumes are very low because it is not listed in most pharmaceutical trade references. Manufacture would probably be by the same process used in the laboratory, probably in very small batches, and with equipment not much larger than standard laboratory apparatus. Any dioxins present in the TCP raw material are probably discharged in plant wastes rather than being concentrated into the pharmaceutical. Most of the dioxin probably is filtered from the benzene solution as part of the insoluble matter. Since some dioxins are slightly soluble in both benzene and petroleum ether, a portion might remain in solution and be transferred to solvent recovery distillation columns. The remaining dioxin would be discarded as part of an anhydrous tar from the base of these columns. The pharmaceutical industry usually incinerates both solid organic residues and solvent recovery tars. Dicamba-The herbicide dicamba is a derivative of salicylic cally as 3,6-dichloro-2-methoxybenzoic acid. In 1972, samples indicated no tetra-, hexa-, or hepta-CDD's at a 0.5 ppm (Woolson, Thomas, and Ensor 1972). The presence retically possible, however. acid known chemianalysis of eight detection level of of DCDD's is theo- Dicamba is made by acylation of 3,6-dichlorosalicylic acid, which in turn is made from 2,5-dichlorophenol. The chemical reactions are shown below. O NaOH _ J" OCH3 'CICH3OSO3CH3H -J NaOH 2,5-DICHLOROPHENOl OICMBt The first step is known as the Kolbe-Schmitt reaction and is also used to make unsubstituted salicylic acid from unsubstituted phenol in addition to haloginated derivatives (U.S. Patent Office 1955a). Operating temperature is probably below 200°C, and operating pressure is probably greater than 8 atmospheres. The chlorinated salicylic acid is mixed into water and sodium hydroxide and treated with dimethyl sulfate (U.S. Patent Office 1967a). The reaction is conducted initially with refrigeration to retard the otherwise violent reaction; the mixture is then heated for a few hours at reflux temperature (slightly above 100°C). 56 �On completion of the reaction, the mixture is acidified with hydrochloric acid. Dicamba precipitates and is filtered from the mixture, rinsed with water, and dried. Recrystallization from an organic solvent such as ether is possible, but probably is not conducted in commercial practice. Except for high temperature, all conditions necessary for formation of chlorinated dioxins are present. It is likely that at high temperature dicamba would lose carbon dioxide in a reversal of the initial manufacturing reaction, and any dioxins formed would not contain carboxyl groups. Dicamba is reported to be made by Velsicol Chemical Corporation in Beaumont, Texas, under the trade name Banvel (Stanford Research Institute 1978). It is commercially available in many formulated pesticide products. Other Chlorophenol Derivatives-Compounds other than the products listed above are also potential dioxin sources, but are made and used in smaller volumes. A compound with the trade name of Irgasan B5200 is used as a bacteriostat and a preservative. Often described by the generic abbreviation TCS, it is an acid amide derivative of a chlorinated salicylic acid, made by first reacting 2,4-dichlorophenol with sodium hydroxide and carbon dioxide at high pressure, then reacting the resulting intermediate with 3,4dichloroaniline (U.S. Patent Office 1955a). The germicide Irgasan DP 300 is a predioxin that was once sold in this country by Ciba-Geigy Corporation. As outlined in Section 2, it was used in some of the research of chlorinated dioxin chemistry, and dioxins were formed readily on heating of this compound. Its chemical formula is as follows: CIHO This compound is a derivative of 2,4-dichlorophenol, although the process of manufacture has not been reported. The formulation called Dowlap was once used in the Great Lakes to control the sea .lamprey, an eel-like fish. The active ingredient of the formulation was 3,4,6-trichloro-2-nitrophenol, whose chemical formula is as follows: Cl 57 �This compound was made by direct nitration of 2,4,5-trichlorophenol using concentrated nitric acid in a solvent of glacial acetic acid (Merck Index 1978). A dye assistant chemical for use with polyester fibers was once made with the trade name Tyrene (Merck Index 1978). Its chemical name is 2,4,6trichloroanisole or 2,4,6-trichloromethoxybenzene, with a structural formula as follows: It was probably made by acylation of 2,4,6-trichlorophenol with dimethyl sulfate. Dioxins in Chlorophenol Production Wastes Although the dioxin content of many products containing chlorophenols or their derivatives has been reported in the literature, little information is available on dioxins in the industrial wastes created by chlorophenol manufacture. One unpublished report (U.S. Environmental Protection Agency 1978d) describes analysis for dioxins in 20 samples of liquid wastes from plants manufacturing trichlorophenol, pentachlorophenol, and hexachlorophene. The limit of detection was 0.1 ppm. No TCDD's were detected in any of the samples. Hexa-, hepta-, and octachlorodioxins were found in the pentachlorophenol wastes. The report does not indicate clearly whether any of the higher chlorodioxins were found in the hexachlorophene wastes. Considerations of solubility and volatility suggest that large concentrations of dioxins will be found in the stillbottom wastes from 2,4,5-TCP manufacture. Direct analytical evidence to this effect, though limited, is affirmative. Waste oils identified as early 1970 still residues from a former 2,4,5-TCP manufacturing plant in Verona, Missouri, have been analyzed and reported to contain ppm quantities of 2,3,7,8-TCDD (Johnson 1971; Commoner and Scott 1976a). A toluene still bottom waste taken from Transvaal's plant in Jacksonville, Arkansas, has recently been found to contain 40 ppm of TCDD's (Watkins 1979; also see Volume II of this series). The effect of biological treatment on removal of dioxins from liquid industrial wastes is not known. In 1978, the Dow Chemical Company reported that no 2,3,7,8-TCDD could be detected in 13 of 14 grab and composite samples from the secondary and tertiary outfall of its manufacturing plant, which produces large quantities of 2,4,5-TCP, 2,4,5-T, and other chlorophenolic compounds; one sample was questionable. The reported level of detection ranged from 1 to 8 ppt. No information is given on the dioxin content of the untreated waste stream or on the treatment methods. 58 �Apparently it has been common practice for chemical manufacturers to dispose of dioxin-contaminated wastes or other toxic chemical wastes by landfill. Either liquid or solid forms of the wastes are placed in drums and stored or buried. Dioxin wastes disposed of in this manner would undoubtedly be quite concentrated and potentially very dangerous. Recently ppt to ppb levels of TCDD's were reported in environmental samples from two landfills in Niagara Falls, New York (Chemical Week 1979). Hooker Chemical reportedly has dumped a total of 3700 tons of 2,4,5-trichlorophenol wastes over the past 45 years in these two dumps (Hyde Park and Love Canal) and in one other disposal site on the company's Niagara Falls property. The report estimated that the wastes buried in these landfills could contain over 100 pounds of TCDD's. At the Transvaal pesticide plant in Jacksonville, Arkansas, more than 3000 barrels of dioxin-contaminated wastes are stored on the plant property (Fadiman 1979). The total quantity of TCDD present in the wastes has not been estimated. No other known information describes the quantities of dioxins that might be buried elsewhere in the United States. In an effort to identify areas where landfills are most likely to contain large dioxin wastes, Figure 6 illustrates the locations where chlorophenols and their derivatives are now or were formerly produced. A list of these locations is presented in Table 11; note that this list does not include locations of the many companies that are believed only to formulate or otherwise merchandise the chlorophenols or their derivatives. A detailed discussion of the methods used for disposal of dioxins is presented in Section 6. Additional information related to the environmental effects of dioxin disposal is presented in the subsection on Water Transport in Section 5. HEXACHLOROBENZENE In 1974, a technical paper reported the presence of OCDD in samples of commercial hexachlorobenzene (Villaneuva et al. 1974). Three samples were analyzed, two of which contained OCDD in concentrations of 0.05 and 211.9 ppm. All three contained octachlorodibenzofuran (OCDF) in concentrations of 0.34, 2.33, and 58.3 ppm. One sample contained a trace amount of heptachlorodibenzofuran. It was established that the principal impurity in these samples was pentachlorobenzene in amounts ranging from 0.02 percent to 8.1 percent. When the samples were examined qualitatively, 11 other impurities having polychlorinated ring-type structures were identified: Octachlorobiphenyl Decachlorobiphenyl 1-Pentachlorophenyl-l,2,3-dichloroethylene Decachlorobiphenyl 59 �( nMtrMoAKo i A"™ " ~"''~™"' r™ • "~' ....j:-^.-. -"I T r*™*™ ~•«••«( '•-—L.r~\ 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. PHILADELPHIA, PA. SAN MATEO, CAL. PORTLAND, OREG. CLEVELAND, OHIO MIDLAND, MICH. TUSCALOOSA, ALA. LINDEN, N.O. CLIFTON, N.J. NAPERVILLE, ILL. JACKSONVILLE, ARK. SPRINGFIELD, MO. NIAGARA FALLS, N.Y. DOVER, OHIO SHENANDOAH, IOWA RAHWAY, N.J. WHITEFORD, MD. Figure 6. SAUGET, ILL. CHICAGO, ILL. KANSAS CITY, KANS. VERONA, MO. TACOMA, WASH. ST. PAUL, MINN. ST. JOSEPH, MO. CHICAGO HEIGHTS, ILL. NITRO, W. VA. AMBLER, PA. FREMONT, CAL. PORT NECHES, TEX. ST. LOUIS, MO. WICHITA, KANS. ORLANDO, FLA. Locations of current and former producers of chlorophenols and their derivatives. 60 �TABLE 11. LOCATIONS OF CURRENT AND FORMER PRODUCERS OF CHLOROPHENOLS AND THEIR DERIVATIVES9 Producer Location Chemical Type Alco Chemical Corp. Philadelphia, Pennsylvania 2,4-D J. H. Baxter and Company San Mateo, California PCP Chempar Portland, Oregon 2,4,5-T 2,4-D Diamond Shamrock Corp. Cleveland, Ohio 2,4,5-TCP 2,4,5-T 2,4-D Dow Chemical, U.S.A. Midland, Michigan 2,4,5-TCP 2,4,6-TCP 2,3,4,6-Tetrachlorophenol 2,4-D 2,4,5-T Si 1 vex Ronnel Erbon DMPA Fallek-Lankro Corp Tuscaloosa, Alabama 2,4-D GAF Linden, New Jersey 2,4-D Givaudan Corporation Chemicals Division Clifton, New Jersey Hexachlorophene Guth Corp. Naperville, Illinois 2,4-D Hercules, Inc. Jacksonville, Arkansas 2,4-D Si 1 vex 2,4,5-TCP Hoffman-Taft, Inc. Springfield, Missouri 2,4,5-T Hooker Chemical Corp. Occidental Petroleum Corp., subsidiary Niagara Falls, New York 2,4,5-TCP (continued) 61 �TABLE 11 (continued) Producer Location Chemical Type ICC Indus., Inc. Dover Chem. Corp., subsidiary Dover, Ohio PCP Imperial, Inc. Shenandoah, Iowa 2,4-D Merck and Co., Inc. Rahway, New Jersey 2,4,5-TCP Miller Chemicals Alco Steel subsidiary Whiteford, Maryland 2,4-D Monsanto Company Monsanto Industrial Chemicals Company Sauget, Illinois PCP 2,4,5-T 2,4-D Nalco Chemical Co. Chicago, Illinois PCP 2,4,5-TCP North American Phillips Corp., Thompson-Hayward Chemical Co., subsidiary Kansas City, Kansas 2,4-D 2,4,5-T Silvex North Eastern Pharmaceuticals Verona, Missouri 2,4,5-TCP PBI-Gordon Corporation0 Kansas City, Kansas 2,4-D 2,4,5-T Private Brands, Inc.c Kansas City, Kansas 2,4-D 2,4,5-T Reichhold Chemicals, Inc. Tacoma, Washington PCP Rhodia, Inc. Agricultural Division Portland, Oregon St. Paul, Minnesota St. Joseph, Missouri 2,4-D 2,4-DB Riverdale Chemical Co. Chicago Heights, Illinois 2,4-D 2,4,5-T Silvex Roberts Chemicals, Inc. Nitro, West Virginia 2,4,6-TCP Rorer-Amchem , Ambler, Pennsylvania u p* .-__•_ ~ i. ^_.*i£^;.ihu___ Amchem Products, Inc., Div.U Fremont, California I St. Joseph, Missouri (continued) 62 2,4,5-T 2,4-D �TABLE 11 (continued) Producer Location Chemical Type Sanford Chemicals Port Neches, Texas 2,3,4,6-Tetrachlorophenol PCP Thompson Chemicals St. Louis, Missouri 2,4,5-T 2,4-D Union Carbide Corp. Agricultural Products Division Amchem Products, Inc., subsidiary Ambler, Pennsylvania Fremont, California St. Joseph, Missouri 2,4-D 2,4,5-T Vertac, Inc. Transvaal,, Inc. , subsidiary Jacksonville, Arkansas 2,4,5-TCP 2,4-D 2,4,5-T Si 1 vex Vulcan Materials Co. Chemicals Division Wichita, Kansas PCP Woodbury Comutrix subsidiary Orlando, Florida 2,4-D b c d Sources: SRI Directory of Chemical Producers, United States. 1976, 1977, 1978, and 1979. U.S. Tariff Commission. Synthetic Organic Chemicals, United States Production and Sales, 1968. U.S. International Trade Commission. Synthetic Organic Chemicals, United States Production and Sales. 1974, 1976, 1977, and 1978. Hercules, Inc., was a former owner of the Jacksonville, Arkansas,, facility now owned by Vertac, Inc. Private Brands, Inc., is believed to be a former owner of the Kansas City, Kansas, facility now owned by PBI-Gordon Corp; Former"Rorer-Amchem facilities in Ambler, Pennsylvania; Fremont, California; and St. Joseph, Missouri, are now owned by Union Carbide Corp. 63 �Octachloroblphenylene Octachoro-1,1-bicyclopentadienylidene Hexachlorocyclopentadiene Nonachlorobiphenyl DecachlorobiphenyT Pentachioroiodobenzene Heptachloropilium It is significant that this list includes no phenolic compounds and no predioxins or iso-predioxins. In fact, the only compounds in these samples that contain oxygen are dioxins and dibenzofurans. Uses Hexachlorobenzene is a registered pesticide formerly used to control a fungus infection of wheat. It is also a waste byproduct from manufacturing plants that produce chlorinated hydrocarbon solvents and pesticides (Villaneuva 1974; U.S. Environmental Protection Agency 1978g). It can be used as a raw material in the manufacture of pentachlorophenol, but is not so used in this country. Hexachlorobenzene is not the same compound as benzene hexachloride. The empiric formula of hexachlorobenzene (HCB) is C6C16, and its structure is that of benzene in which all of the hydrogen atoms have been replaced with chlorine. Benzene hexachloride (BHC) is the common name of hexachlorocyclohexane. Its empiric formula is C6H6C16, and its structure results from direct addition of chlorine to benzene rather than from replacement of hydrogen. One stereoisomer of BHC, the gamma form, is a powerful insecticide, and its use in this country is severely restricted. It is still made, however, because BHC is an intermediate in the most common synthesis method of producing HCB. Manufacture In the manufacture of HCB, the first step is a photochlorination, in which chlorine gas is bubbled through benzene (Wertheim 1939; U.S. Patent Office 1955b). This occurs in specialized reaction vessel fitted with a strong source of ultraviolet light. In a low-temperature reaction, the light catalyzes the conversion of approximately 25 percent of the benzene into a mixture of BHC isomers. This mixture is "crude" BHC, consisting of about 65 percent of the alpha isomer, 10 percent beta, 13 percent gamma, 8 percent delta, and 4 percent epsilon. It is separated by distilling off most of the excess benzene for recycle and then filtering the BHC crystals from the mixture. All stereoisomers of BHC are equally suitable for the manufacture of HCB. The continuation of the process consists of mixing BHC with chlorosulfonic acid or sulfuryl chloride and holding the mixture at approximately 200°C for several hours (U.S. Patent Office 1957a). This step removes the 64 �hydrogen from BHC and thereby restores the unsaturated benzene ring. When the mixture is cooled, HCB precipitates and is separated by filtration, rinsed with water, dried, and packaged. The following shows the overall chemical reactions of the process. Cl CI2 H U.V. LIGHT »- BENZENE HEXACHLOROBEHZENE Decriptions of these process steps provide no indication that dioxins are formed. The raw materials are benzene, chlorine, and chlorosulfonic acid, none of which are likely sources of dioxins. The only reactant that could contribute the oxygen needed to complete the dioxin ring is chlorosulfonic acid, but in this compound the oxygen is tightly bound in a linkage with sulfur. There is, however, a supplemental process that contributes other chemicals that may lead to dioxin formation. This extra step may be conducted at some plants, or may have been conducted in earlier years. If a market exists for the sale of gamma-BHC as an insecticide, this material is extracted from the mixture of crude BHC and benzene after most of the excess benzene has been distilled off for recycle. To this concentrated solution, water is added, along with other chemicals. The objective is to form an emulsion that will entrain part of the BHC. The solution is then filtered; the emulsion passes through the filter, while the solids that were not emulsified are captured. Since gamma-BHC accumulates preferentially in the emulsion, the solids from this first filtration are used for HCB manufacture and the filtrate is treated with salt to break the emulsion and then refiltered. The second crop of solids contains up to three times as much gamma-BHC as the crude product and is dried and sold separately (U.S. Patent Office 1955b). As indicated by the process patent, chemicals added during this supplemental step include a wide range of organic detergents and solvents, but 65 �none of those listed are phenolic or have been shown to create dioxins. Detergents of the anionic type are preferred, especially salts of sulfonated succinic esters, although any of the common surface-active agents are suitable. Supplemental solvents may not be employed, since benzene alone is said to be preferred, but other suitable solvents include dioxanes, any of the aliphatic substituted benzenes, any of the common chlorinated paraffin hydrocarbons, kerosenes, and ethyl ether. Dioxane is the one compound listed that might contribute to dioxin formation, although the reaction is not reported in published literature. Production Current information on the volume or production of hexachlorobenzene is uncertain. Annual production estimates range from 420,000 to 700,000 pounds. Stauffer was the only reported domestic producer in 1974; Dover Chemical Company of Niagara Falls, New York, was the only reported producer in 1977 (U.S. Environmental Protection Agency 1978g). Dioxins have not been reported in any other chlorobenzene compounds. OTHER PHENOLIC COMPOUNDS WITH DIOXIN-FORMING POTENTIAL Several compounds with a phenol nucleus that do not contain chlorine are now being manufactured or were manufactured at one time. Four such compounds or classes of compounds are examined for their dioxin-forming potential in this section. These and others are described more fully in Volume III of this report series. Brominated phenols Three brominated phenolic compounds were once manufactured, and may still be. Because brominated dioxins have been made in laboratory experiments, they may be created during the manufacture of these compounds. Published data describe the production method for tetra-bromo-cresol, which is made by direct bromination of o-cresol in a solvent of carbon tetrachloride with aluminum and iron powders as catalysts (U.S. Patent Office 1943). The following reaction is conducted at room temperature, and it requires about 24 hours to complete a batch. Bra O-CRESOL TETRABROMO-0-CRESOL 66 �When the reaction is complete, the mixture is heated to about 80°C to drive off the carbon tetrachloride solvent and excess bromine. The residue is mixed with dilute hydrochloric acid to form a slurry, which is then filtered. The resulting solids are washed with water, dried, and packaged. Yield is about 95 percent. It is possible to recrystallize this product to separate nonphenolic impurities by dissolving the crude product in sodium hydroxide solution, filtering out insolubles, neutralizing the mixture with hydrochloric acid, and refiltering. This step may or may not be conducted in commercial practice. Two other brominated phenolic compounds are believed to be made by essentially the same process. Structural formulas are as follows: Br 2,4,6-TRIBROMOPHENOL 2,4,6-TRIBROMO-M-CRESOL Almost all brominations of organic compounds are low-temperature processes because bromine is readily vaporized and would be driven from the reaction vessels at high temperatures. A metallic catalyst is needed to activate the diatomic liquid bromine, and a volatile solvent is usually employed to maintain all reactants in the liquid state. Because the temperature during manufacture of these compounds does not usually exceed 80°C except at the surface of heating coils, dioxin formation would not be expected. If dioxin contamination enters with the raw materials, brominated dioxins likely would appear in the crude product. If the product is recrystallized, the dioxins could be constituents of a waste sludge. The literature mentions dioxins that are both brominated and methylated (See Table 3 of Volume III of this series). By the known process of dioxin formation, 2,4,6-tribromo-m-cresol would be expected to form several dimethyltetrabromodioxins, and other cresols would also, in theory, form dimethyl dioxins. 0-Nitrophenol There is no direct utilization of o-nitrophenol as a completed chemical. It is a chemical synthesis intermediate, although it has fewer uses than p-nitrophenol. 67 �The manufacture of o-nitrophenol is a hydrolysis of o-nitrochlorobenzene with sodium hydroxide in a process essentially identical to the hydrolysis method of chlorophenol production. follows: The chemical reaction is as + NaCI NaOH Although the operating conditions of this reaction are not known, conditions of temperature are probably mild. In nitrochlorobenzenes, the chlorine atom is weakly attached, especially when the substituents are in the ortho position. The chlorine atom behaves like that of an alkyl halide and is readily replaced. In contrast, the nitro group is very strongly attached and its replacement is difficult (Wertheim 1939). Unsubstituted dioxin would be created if a further reaction did occur to remove the nitrate group by the following theoretical reaction: NaOH 2NaNO 2 This reaction is possible, and o-nitrophenol may be a source of dioxin contamination. See also Volume III of the report series. This Illinois. compound is manufactured by the Monsanto Company, Sauget, Salicylic Acid Salicylic acid is an important chemical synthesis intermediate used to make dyes, flavoring chemicals, and pharmaceutical compounds such as aspirin. Unsubstituted dioxin may be present, but has not been reported. Salicylic acid is made by a high-pressure reaction between phenol and carbon dioxide in the presence of sodium hydroxide; this reaction is known as the Kolbe-Schmitt reaction. 68 �+ CO2 NaOH COOH Operating temperature is about 150°C. Higher temperatures are avoided to prevent a side reaction that forms p-hydroxybenzoic acid. This process includes some of the conditions needed to produce unsubstituted dioxin, but not all of them. The hypothesis of possible dioxin formation is strengthened, however, by observations of products created by thermal decomposition of salicylic acid. When heated strongly, it decomposes primarily into phenol and carbon dioxide, but also into smaller amounts of phenyl salicylate, which in turn condenses to xanthone: SUICYIJITE XANTHONE Since the ortho carbon is held weakly in the salicylic acid molecule, and since the triple-ring xanthone structure has been identified, the formation of dioxins may also be possible, especially if oxygen is present. Both salicylic acid and xanthone are widely distributed in nature. Salicylic esters are responsible for some plant fragrances, and xanthone is a yellow pigment in flowers. 69 �Salicylic acid is manufactured by four companies in this country: Dow Chemical Company - Midland, Michigan Monsanto Company - St. Louis, Missouri Hilton-Davis Chemical Company - Cincinnati, Ohio Tenneco Chemicals, Inc. - Garfield, New Jersey The combined capacity of these four plants is 24 million kilograms annually. Aminophenols The o-aminophenols could conceivably form dioxins through condensation with loss of ammonia. These are not high-volume chemicals and are not known to be made with halogen substituents. A class of related compounds is used in much larger quantity; these are the derivatives of o-anisidine (methoxyaminobenzene), which in several forms are important dye intermediate chemicals. These might condense in appropriate environments into the dioxin structure through loss of methyl amine. The environments would probably be acidic: 2 NH2CH3 Although this reaction is possible, it is unlikely because the amine group is tightly bound to the benzene ring. Aminophenols or similar compounds are not likely sources of dioxin contamination. DIOXINS IN PARTICIPATE AIR EMISSIONS FROM COMBUSTION SOURCES Several reports describe the occurrence of dioxins in fly ash and flue gases from municipal incinerators and industrial heating facilities. In 1977, analysis of .samples of fly ash from three municipal incinerators in the Netherlands showed 17 different dioxins (5 TCDD's, 5 penta-CDD's, 4 hexa-CDD's, 2 hepta-CDD's, and OCDD) (Olie, Vermeulen, and Hutzinger 1977). Although the specific number of isomers was not stated, the same dioxins were found in flue gas from one of the incinerators. In addition, large amounts of di-, tri- and tetrachlorophenols were found in flue gases, and high levels of chlorobenzenes, especially hexachlorobenzene, were detected in all fly ash samples. 70 �Another team of investigators reported finding the same dioxins in Switzerland (Buser and Bosshardt 1978). This'study quantitatively determined that the total amount of polychlorinated dibenzo-p-dioxins in the fly ash from a Swiss municipal incinerator and industrial heating facility were 0.2 ppm and 0.6 ppm, respectively. High-resolution gas chromatography was then used to identify 33 specific dioxin isomers found in the fly ash samples. The dioxin isomers known to be most toxic, which are 2,3,7,8-TCDD, 1,2,3,7,8-penta-CDD, 1,2,3,6,7,8- and 1,2,3,7,8,9-hexa-CDD, were only minor constituents of the total dioxins found. Later in 1978, researchers at Dow Chemical Co. reported finding ppb levels of chlorinated dioxins in particulate matter from air emissions of two industrial refuse incinerators, a fossil-fueled powerhouse, and other combustion sources such as gasoline and diesel autos and trucks, two fireplaces, a charcoal grill, and cigarettes. See Table 12. All of these sources are believed to be located on or near the Dow facilities in Midland, Michigan. Tetra-, hexa-, hepta- and octachlorodioxins were found. Concentrations of 2,3,7,8-TCDD were minor relative to concentrations of other dioxins. Dow concluded from the study that their Midland facility was not a measureable source of the dioxins found in fish from nearby rivers, and that, in fact, chlorinated dibenzo-p-dioxins may be ubiquitous in combustion processes. A preliminary data analysis by the EPA does not entirely agree with Dow's conclusions. EPA continues to believe that Dow's Midland plant is the major and possibly the only source of the dioxins contaminating fish in nearby rivers. EPA has asked Dow for further clarification of the company's findings and analytical methods (Merenda 1979). In contrast to the Dow finding of 38 ppb TCDD's in powerhouse emissions, Kimble and Gross (1980) report finding no TCDD's in fly ash from a typical commercial coal-fired power plant in California; the detection limit was 1.2 ppt. In 1980 Wright State University chemists analyzed emissions from a U.S. municipal incinerator for chlorodioxins (Tiernan and Taylor 1980). TCDD's were detected in all seven samples. Isomer-specific analyses indicate that 2,3,7,8-TCDD is a minor product, and evidence was obtained for the presence of 1,3,6,8-, 1,3,7,9-, 1,3,7,8-, 1,3,6,7-, and at least 6 other TCDD isomers. The formation of dioxins from the thermal decomposition of chlorophenols and their salts (chlorophenates) is well documented. In 1971, Milne reported finding no evidence of formation of lower chlorinated dioxins from the thermal decomposition of dichlorophenols; all six dichlorophenol isomers were studied. However, Aniline (1973) found that pyrolysis of 2,3,4,6-tetrachlorophenate produced two hexa-CDD isomers. Later, Stehl et al. (1973) found that burning paper treated with sodium pentachlorophenate produced OCDD but burning either wood or paper treated with pentachlorophenol did not produce the dioxin. In 1975, a series of pyrolysis experiments was conducted with 2,3,4-, 2,3,5-, 2,4,5- and 2,3,6-tri, 2,3,4,5-, 2,3,5,6- and 2,3,4,6-tetra, and pentachlorophenates to obtain samples of many tetra-, penta-, hexa- and octa-CDD's for study (Buser 1975). In 1977, 71 �TABLE 12. DIOXINS IN SELECTED SAMPLES' (ppb except as noted) TCDD's Source 2,3,7,8-TCDD Other TCDD isomers Soil inside plant 0.3-100 0.8-18 7-280 70-3200 Dust samples from Dow Research Building 0.7-2.6 0.5-2.3 9-35 140-1200 650-7500 0.09-0.4 0.3-3.9 0.4-31 0.02-0.14 0.10-3.3 0.35-22 Soil and dust from Midland and metro areas 0.03-0.04 1 0.005-0.03 Soil and dust from major metro area Hexa-CDD's Hepta-CDD's OCDD 490-20,500 PO Soil and dust from urban area none none 0.03-1.2 0.035-1.6 0.05-2.0 Soil and dust from rural area none none none 0.02-0.05 0.10-0.35 Dow stationary tar incinerator parti culates none none 1-20 27-160 190-440 Dow rotary kiln incinerate r w/supplementary fuel none none 1.4-5.0 4-110 9-950 Dow rotary kiln incinerate r w/o supplementary fuel Dow powerhouse fired with fuel oil /coal (continued) 110-8200 none 1800-12,000 1300-65,000 38 2 2000-510,000 4 3000-81 O.OOC 24 �TABLE 12 (continued) Other TCDD isomers Source Automobiles catalytic - carbon catalytic - rust noncatalytic diesel trucks none 0.4 none 3.0 Fireplaces (scrapings) 0.1 Cigarettes (tars) none none Charcoal -grilled steaks CO 2,3,7,8-TCDD none none Residential electrostatic precipitator 0.6 0.1 4.0 4.0 20.0 0.27 0.40 Hexa-CDD ' s Hepta-CDD's OCDD 0.5-2.0 0.7 none 4-37 2-14 3 3 35-110 8-72 28 10-16 190-280 0.23-3.4 0.67-16 0.89-25 8.5-9.0 18-50 none 3-7 5-29 34 430 1300 1200 4.2-8.0 Parti culates from rotary kiln scrubber water w/supplemental fuel 4 6 200 970 w/o supplemental fuel 2500 3400 26,000 42,000 0.005 0.024 0.026 12-25 56-119 Filtered scrubber water 0.0028 Cooling tower residues 1.6-6.0 10 1-4 N.A.b Sewer waters before treatment (concentration-ppt) f* Source: Dow Chemical, U.S.A. 1978. N.A. = not applicable. N.A.b 3-1500 �2,3,7,8-TCDD was found as a combustion product of many 2,4,5-trichlorophenoxy compounds, but the amount of this dioxin was very small relative to the amount of the 2,4,5-trichlorophenoxy compound that was burned (Stehl and Lamparski 1977). Results of the study showed that only 1.2 x 10-5 to 5 x 10-5 percent by weight of the 2,4,5-trichlorophenoxy species was converted to 2,3,7,8-TCDD by combustion. The origin of the dioxins in airborne particulates from combustion is not yet clarified. Rappe et al. (1978) suggest that polychlorinated dibenzo-p-dioxins can be formed during combustion by dimerization of chlorophenates, by dechlorination of more chlorinated polychlorinated dibenzo-pdioxins, and by cyclization of predioxins. Dow Chemical Company (1978) suggests that because of the complex nature of the materials being burned and the complex chemistries of fire, the formation of chlorinated dioxins occurs in all combustion processes, i.e., that the formation is not necessarily limited to combustion in the presence of chlorophenates or chlorophenols. The presence of biosynthesized compounds with characteristics of dioxin precursors may give some credence to this contention. An alternative explanation for the observed presence of dioxins in the fly ash of refuse incinerators is that the dioxins enter intact as contaminants of the wastes being burned. For example, silvex-treated grass clippings, sawdust or other wastes from PCP-treated wood (landscape timber, railroad ties), and "empty" PCP, silvex, or other pesticide containers from home or industrial use might be direct sources of the dioxins detected in municipal incinerator fly ash. If this were the case, seasonal variations in fly ash dioxin content would be expected, with larger amounts in spring and summer. DIOXINS IN PLASTIC In 1965, it was reported that dioxin is an impurity in the preparation of polyphenylene ethers (Cox, Wright, and Wright 1965). No reports further substantiating this finding are known. "PPO" is a trademark of General Electric Company for a polyphenylene thermoplastic derived from 2,6-dimethylphenol (Hawley 1971). The dioxin configuration one would expect from condensation of the dimethylphenol is as follows: OH 2 CH4 2,6-OIMETHUPHENOk 1,6-DIMETHVLDIBENZO-P-DIOXIN 74 �Because PRO is highly resistant to acids, bases, detergents, and hydrolysis it may be used in hospital and laboratory equipment, and in pump housings, impellers, pipes, valves, and fittings in the chemical and food industries. DIOXINS PRODUCED FOR RESEARCH PURPOSES Many investigators have reported the sources of purified dioxin standards used in their studies. Some of these dioxin sources and the names of the dioxins provided are listed in Table 13. 75 �TABLE 13. SOURCES OF PURIFIED DIOXIN SAMPLES FOR RESEARCH Source Dioxin provided Reference Givaudan Ltd. Dubendorf, Switzerland 2-mono-CDD 2,3-di-CDD 2,7-di-CDD 2,8-di-CDD 1,2,4-tri-CDD 1,3,7-tri-CDD 2,3,7-tri-CDD Buser (1978) Or. K. Anderson University of Umea, Sweden 1,2,3,4-tetra-CDD Buser (1978) Dr. C. A. Nilsson University of Umea, Sweden 1,2,3,8-tetra-CDD 1,2,3,7-tetra-CDD Buser (1978) Stickstoffwerke Linz, Austria 2,3,7,8-tetra-CDD Buser (1978) Dr. David Firestone Food and Drug Administration Washington, D.C., U.S.A. 1,2,3,7,8-penta-CDD 1,2,4,7,8-penta-CDD 1,2,3,6,7,8-hexa-CDD 1,2,3,7,8,9-hexa-CDD Buser (1978) Dow Chemical, U.S.A. Midland, Michigan Unspecified dioxin standards Villanueva (1973) ITT Research Institute Chicago, Illinois, U.S.A. 1,2,4,6,7,9-hexa-CDD 1,2,3,6,7,9-hexa-CDD 1,2,3,6,7,8-hexa-CDD 1,2,3,7,8,9-hexa-CDD 1,2,3,4,6,7,9-hepta-CDD 1,2,3,4,6,7,8-hepta-CDD Firestone (1977) A. E. Pohland Food and Drug Administration Washington, D.C., U.S.A. 2,3,7,8-TCDD OCDD Firestone (1977) A. Poland McArdle Laboratory for Cancer Research University of Wisconsin Madison, Wisconsin, U.S.A. 14 C-TCDD O'Keefe et al. (1978) Dow Chemical, U.S.A. Midland, Michigan hexa-CDD hepta-CDD octa-CDD C. D. Pfeiffer (1978) 76 �SECTION 4 SOURCES OF HUMAN EXPOSURE The toxicity of some dioxins, especially 2,3,7,8-TCDD, has been demonstrated in a number of incidents of human exposure. The most serious incidents, including one man-made disaster, have affected the general public; these incidents have resulted from industrial accidents, improper disposition of industrial wastes, and a variety of other exposure routes. In addition to exposures of the general public, human contact with dioxins has occurred in chemical manufacturing plants and in other locations because of the occupational handling of these materials. This report section summarizes both the reported incidents of human exposure to dioxins and the potential exposure routes. PUBLIC EXPOSURE Industrial Accidents The clearest demonstration of dioxin toxicity was a disastrous incident that occurred on July 10, 1976, in Meda, Italy, at a plant producing 2,4,5-TCP for the manufacture of hexachlorophene. The plant was operated by the Industrie Chemiche Meda Societa, Anonima, (ICMESA), an Italian firm owned by the Swiss company Givaudan, which in turn is owned by HoffmanLaRoche, a Swiss pharmaceutical manufacturer. The incident often is described inappropriately as an explosion. A safety disc on an overpressured 2,4,5-TCP reactor ruptured, and a safety valve opened, releasing the reactor contents directly to the atmosphere (Homberger et al. 1979; Peterson 1978). The quantity of TCDD's released has been estimated to be from 300 g to 130 kg (despite extensive study, there is still no agreement as to the most likely amount) (Bonaccorsi, Panel!i, and Tognoni 1978; Carreri 1978). The incident occurred late on a Saturday afternoon. It resulted from the closing of a valve that supplied cooling water to the reactor jacket. In the manufacturing process, caustic soda had been used .to hydrolyze 1,2,4,5-tetrachlofobenzene in a solvent of ethylene glycol. After the mixture was heated, cooling water was turned onto the jacket and should have remained on until the reaction was complete. A decision had been made to postpone the next operation, a distillation to remove ethylene glycol, until the following Monday. During the standby shutdown procedures the cooling water valve apparently was closed inadvertently. Since the reaction was incomplete, temperature and pressure continued to increase until the .limiting pressure of the safety devices was reached. When the release 77 �occurred, the regular operators were not in the plant. Five minutes after the release started, someone opened the cooling water valve and the influx of cooling water began to slow down the reaction. Within 15 minutes, release of chemicals to the atmosphere had stopped. A slight breeze carried the toxic cloud over parts of 11 towns and villages, as condensed chemicals fell from the cloud like snow. The town most affected was Seveso, whose corporate limits adjoin the plant grounds. No emergency action was taken by plant personnel or local authorities, although several people reported to hospitals with chemical burns. Not until the next day, Sunday, was the mayor of Seveso notified of the accident, and officials of other affected towns were not told until Monday. The plant resumed normal operations Monday morning. No official emergency decree was issued until 5 days after the accident, and the possible presence of 2,3,7,8-TCDD was not announced to the local population until after 8 days (Carreri 1978). By then, hundreds of animals had sickened and died, and people with chloracne, principally children, were being hospitalized. The plant workers went out on strike, finally closing the plant. Since ICMESA had no suitable laboratory, samples of the contamination had to be sent to Switzerland for analysis; not until 10 days after the accident did Givaudan and Hoffman-LaRoche confirm that the contamination was 2,3,7,8-TCDD. Only then were organized steps taken to assess the damage and to safeguard the health of the people who had been exposed (Reggiani 1977; Peterson 1978; Bonaccorsi, Panel!i, and Tognoni 1978; Carreri 1978). It was discovered that most of the dioxin had fallen in a narrow strip extending for about 5 km to the southeast from the plant (see Figure 7). The most heavily contaminated area of 267 acres was designated Zone A, and was further divided into seven numbered subzones corresponding to the relative degrees of contamination. The population of Zone A was evacuated. A less contaminated area of 665 acres was designated Zone B; official evacuation of this zone was not ordered. A much larger area was designated Zone R (Respect or Risk), in which dioxin contamination was judged to be too slight to be harmful, Chloracne began to appear about 2 days after the accident. Within 6 days, 12 children were hospitalized; within 8 days, there were 14 (Parks 1978). Those first affected were the most seriously affected, and some were still undergoing treatment 3 years after the incident (Revzin 1979). A screening of more than 32,000 children of school age in the Seveso region resulted in the discovery of 187 cases of chloracne (Hay 1978b). Officially, there were 135 confirmed cases within the first year, with "new" waves of the skin disease appearing 18 and 24 months after the accident (Bonaccorsi, Fanneli, and Tognoni 1978). Hoffman-LaRoche reported that most chloracne was of "mild severity and quick recovery" and that there was no increase in the susceptibility of the children to infectious disease (Reggiani 1979a). Only a small percentage of those affected were adults, but a thorough medical survey of the adult population was never conducted. Since 2,3,7,8-TCDD had been shown to cause birth defects and spontaneous abortions in laboratory animals, the incidence of birth problems in 78 �ICMESA N Figure 7. Map of Seveso area showing zones of contamination (A and B) and zone of respect (R). (Source: Adapted from Panel!i, et al. 1980) 79 �the affected population was studied. At present, the resulting data are inconclusive and controversial, in part because of poor statistical data from prior years (Toxic Materials News 1979c). Through May 1977, the spontaneous abortion rate for the entire Lombardy region of Italy, which includes the Seveso area, was lower than the worldwide frequency (15 percent versus 20 to 25 percent) (Reggiani 1977). A private organization, however, reported that 146 malformed infants were born during 1978 in the Seveso area, almost 3 times the number reported officially (Chemical Week 1979b; Revzin 1979). Four years after the ICMESA incident, the people of Seveso are resuming an almost normal life. Hoffman-LaRoche has bought some of the heavily contaminated properties near the plant and has enclosed them and the plant within a tall plastic fence. Contaminated debris and soil from other locations, including the carcasses of 35,000 animals that died or were slaughtered (Parks 1978) have been dumped in the enclosure, and this area is now believed to contain 80 percent of all the dioxin that was released (Chemical Week 1979h). Some nearby houses have been decontaminated by removing the tile roofs, vacuuming and scrubbing the walls with detergents and solvents, and clearing the grounds around them (Parks 1978). All the former residents have been allowed to return to their homes. Having decided the danger is over, many no longer practice any safety precautions (Revzin 1979). None of the many proposals for decontaminating the plant property has satisfied everyone; the situation not only poses a massive technical problem, but is clouded with legal and political difficulties. The Seveso incident has been called an environmental calamity (Parks 1978), and the release of dioxins has been compared to an escape of nuclear radiation in its potential for disaster (Revzin 1979). The effects of the 20-minute release on July 10, 1976, are still continuing and will not be known for years, perhaps not for generations (Bonaccorsi, Panel!i, and Tognoni 1978). Although no human deaths have resulted from the incident thus far, in the light of present toxicological knowledge, late effects can be expected (Peterson 1978). Operations at the ICMESA plant have not resumed since the 1976 accident (Watkins 1979b). Contaminated Industrial Wastes Manufacture of organic chemicals creates wastes, some of which may contain dioxins. In one recorded' incident a chemical plant waste known to contain a dioxin has been clearly responsible for illness of a person not associated with chemical handling operations (Beale et al. 1977). Other instances have been recorded and continue to be discovered in which dioxins have been or are being discarded with wastes in a manner that brings into contact with the general public. This report section lists the known examples of dioxin contamination of public land, air, and water from disposal of industrial wastes. All are associated with present or former producers of 2,4,5-TCP. 80 �Contained or Landfilled Wastes— The most concentrated waste sources of dioxins are the anhydrous liquids, tars, and slurries, which 2,4,5-TCP manufacturers may discard by burying them in the ground or by storing them in drums. These materials are handled both by personnel of the manufacturing company and by contractors responsible to the manufacturer. The most notable incident of nonoccupational exposure to dioxincontaminated wastes of this type involved the spraying of waste oils containing TCDD's on horse arenas and a private road in east central Missouri in 1971 (Shea and Lindler 1975; Environmental Protection Agency 1975b; Commoner and Scott 1976a; World Health Organization 1977; Kimbrough et al. 1977). The wastes were traced to a plant of the North Eastern Pharmaceutical Co. (NEPACCO) in Verona, Missouri, which manufactured 2,4,5-TCP at that time. The residues of a distillation phase of the process were stored above ground in a 7500-gallon tank. Periodically, NEPACCO would contract with someone to dispose of the wastes. Between February and October of 1971, the Bliss Salvage Oil Company held this contract and during these 8 months hauled away 16,000 gallons. Presumably, most was incinerated. In May and June, however, waste oils mixed with these distillation residues were sprayed to control dusts on four horse arenas and a road on a farm owned by the operator of the oil salvage company. 'Unexplained deaths of animals occurred for almost 2 years. By December 1973, over 60 horses had died in the arenas and over 40 had become ill (Commoner and Scott 1976; Kimbrough et al. 1977). Many cats, dogs, rodents, birds, and insects had also died. Seven people developed various disorders as a result of exposure. A six-year-old girl who played regularly on an arena floor was most seriously affected; she was treated for inflammation of the kidneys and hemorrhaging of the bladder, along with other symptoms (Beale et al. 1977). She lost 50 percent of her body weight over the course of the illness, but has since recovered. Finally, the most heavily contaminated soil was removed from the arenas and replaced. This apparently solved the problem, as no further incidents have been reported. The soil, probably still containing dioxins, is now buried in a landfill and under a concrete highway that was being built at the time (Commoner and Scott 1976a). In Australia, Union Carbide of Australia Limited (UCAL), previously a manufacturer of 2,4,5-TCP and 2,4,5-T, disposed of dioxin-contaminated wastes by landfill ing during the years between 1949 and 1971 (Chemical Week 1978b; Dickson 1978). At the time these wastes were buried, landfilling was the most acceptable method of disposal. It has been estimated that 16 to 30 kilograms of dioxins may be present in the buried wastes (Chemical Week 1978b; Dickson 1978; Chemical Week 1978c). In 1969, when dioxin contaminants in 2,4,5-trichlorophenol were being publicized, UCAL began removing the dioxins by adsorption onto activated carbon. The dioxin-contaminated carbon, now stored in steel drums, presents a disposal problem (Dickson 1978). 81 �Dioxins have been found in two chemical landfills in Niagara Falls, New York. One of these, the Love Canal, is now the site of a residential community, including a school. The landfill previously was used by the Hooker Chemical Company for burying chemical wastes, including those from the manufacture of 2,4,5-TCP. A rising water table has brought the chemicals to the surface (Chem. and Eng. News 1978). Approximately 80 different chemicals have been identified, including a number of known carcinogens (Cincinnati Enquirer 1978a). Recently it was reported that TCDD's were found at the site (Chemical Week 1979a; Wright State University 1979a, 1979b). About 30 tons of 2,4,5-TCP wastes are buried in the Love Canal. Hyde Park, a larger toxic landfill used by Hooker, also has yielded positive analyses. Environmental evaluations of three plants located near the landfill found TCDD's in dust from these plants and in water samples taken from sediments in a nearby creek (Chemical Regulation Reporter 1980). One of the largest accumulated quantities of dioxin-contaminated anhydrous wastes now known is a cache of approximately 3000 drums of chemicals found in 1979 at the Vertac plant in Jacksonville, Arkansas (Fadiman 1979). The proper procedure for final disposition of this material, which may contain as much as 40 ppm or more TCDD's, has not been determined. (See Volume II of this report series.) Incinerated Wastes-A number of present and previous producers of 2,4,5-TCP and 2,4,5-T disposed of wastes by incineration. This method is used by the Dow Chemical Company and was once used by the ICMESA plant and by NEPACCO, which discarded its wastes through a contract incineration company. A recent report has raised a significant question as to whether past or present incineration methods destroy all dioxins. Dow reported in 1978 that fly ash from both stationary tar and rotary kiln incinerators contains low concentrations of dioxins, even that from incinerators designed to burn chemical wastes (Dow Chemical Company 1978). TCDD's bound to particulate matter are largely unaffected by even high-temperature incineration (Rawls 1979; Ciaccio 1979; Miller 1979). It has been suggested that incineration of dioxin-contaminated chemical wastes is primarily responsible for the observed presence of TCDD's in and around the Dow plant in Midland, Michigan (Merenda 1979; Ciaccio 1979).* If this is shown to be the case, pollution of the atmosphere from chemical incinerators may be an important route in the exposure of the public to dioxin chemicals. Miller (1979) has suggested that a worldwide background of atmospheric dioxin contamination may exist as a result of the incineration by the U.S. Air Force of 10,400 metric tons of Herbicide Orange containing up to 47 ppm TCDD's (see Ackerman et al. 1978). This operation took place in the Pacific in 1977. Although there are no data that confirm the presence of widespread atmospheric pollution from this source, TCDD's were detected in some stack emission samples (Tiernan et al. 1979). Dow believes that the observed presence of TCDD's and other dioxins in Midland and other metropolitan areas is due not only to chemical incinerators but to various other combustion sources such as powerhouses, diesel engines, charcoal grills, etc. (Dow Chemical Co. 1978; Rawls 1979). 82 �Discharged Water Wastes-Dioxin concentrations that exceed theoretical solubility limits (Crummett and Stehl 1973) may occur in industrial wastewaters because of 1) the presence of other organic materials in the wastewater that would tend to increase the solubility of the dioxin, and/or 2) the presence of suspended solids to which the dioxins are adsorbed. In either event, it is possible that low levels of dioxins may be carried routinely into the environment by industrial effluents, especially those associated with the production of chlorophenols. Little published information addresses the question of dioxins in such industrial water effluents. A 1978 report from Dow Chemical Co. contends that their effluent discharges were not responsible for the dioxins found in a number of Tittabawassee River fish, collected downstream from the Dow discharge. The report states that dioxins are formed during any combustion process and therefore may be found everywhere in the environment. No dioxins were detected, however, in fish collected above the Dow effluent outfall. Other data presented in the Dow report indicate that particulates in scrubber water contained 46 ppb TCDD's, 200 ppb hexa-CDD's, 970 ppb heptaCDD's, and 120 ppb OCDD. The water was used to scrub the gas stream from a rotary kiln incinerator fired with a supplemental fuel to burn chemical wastes. Disposition of the overflow from the scrubber is unknown; however, it is unlikely that any water treatment system can consistently remove 100 percent of a low-level constituent such as TCDD's, especially if a portion of the TCDD's are adsorbed to particulate matter. In 1976, analysis of effluent water from the Vertac plant in Jacksonville, Arkansas, showed 0.2 to 0.6 parts per billion of 2,3,7,8-TCDD (Sidwell 1976a). In contrast, analysis of effluent from the city stabilization ponds, to which the plant effluent was sent, showed no 2,3,7,8-TCDD (Sidwell 1976b). Because no detection limits were reported, the presence of 2,3,7,8-TCDD in low concentration in the stabilization pond effluent remained a possibility. There was also a question of the validity of the analytical method used in the latter examination. Chemists at Wright State University have recently reported on the analysis of one hundred process and environmental samples taken by the U.S. EPA from the Vertac site and surrounding area (Tiernan et al. 1980). TCDD's were detected in many of the samples at ppt to ppb levels. Composite samples of soil and water from the city sewage treatment plant lagoon contained 8 ppb TCDD's Bottom core samples from the Vertac cooling pond contained 2 to 102 ppb TCDD's; however, no TCDD's were detected in the cooling pond discharge sample (detection limit of 0.05 ppb). Similarly, liquid discharge samples (2) from the equilization basin contained no detectable TCDD's (detection limit 0.010 ppb), even though a bottom mud sample from the basin contained about 400 ppb TCDD's. Treatment of wastes at PCP production plants and wood treatment plants is usually accomplished by oxidation ponds, lagoons, or spray irrigation. The efficiency of these treatment schemes has not yet been evaluated where 83 �dioxins are concerned. There is evidence, however, that water-mediated evaporation is at least partly responsible for the removal of chlorophenols (and also possibly dioxins) from oxidation ponds (Salkinoya-Salonen 1979b). Insufficient treatment could result in contamination of waterways and thus in potential public exposure. Transportation Accidents In January 1979, the derailment of a tank car of orthochlorophenol in Sturgeon, Missouri, resulted in symptoms of chloracne in a cleanup worker. Analysis of the tank car contents showed less than 0.1 percent trichlorophenol contamination and also 37 parts per billion TCDD's. Subsequent analyses by the EPA confirmed that the dioxin contamination was 2,3,7,8-TCDD (Chemical Week 1979d and 1979e; Poole 1979). Further details of the incident have not been released because of extensive legal actions now pending involving the residents of the town and employees of the manufacturing, transportation, and contract clean-up companies. Although the incident at Sturgeon is the only one reported in which dioxins were identified, it is especially significant because of the nature of the chemical involved. The manufacture of orthochlorophenol offers no direct chemical pathway to the side reactions that form 2,3,7,8-TCDD. Nevertheless, contamination with this most-toxic dioxin was present. Product distillation is at least a hypothetical origin. Continuing examinations of the source of the 2,3,7,8-TCDD are indicated and are being conducted. Herbicide Applications For many years, herbicides made from dioxin-contaminated 2,4,5-TCP were widely distributed into the environment. Since the herbicides were less toxic to grasses, canes, and established trees than to broadleaf weeds and undergrowth plants, they found wide application wherever the objective was to stimulate growth of the more resistant plants. The applications included residential lawns; right-of-ways for power lines, railroads, and highways;, forest lands intended for future lumbering; pasturelands; and food crops such as rice and sugar cane. Regulatory and environmental actions have now halted most of these uses of chemicals that may contain dioxins, but a number of public health incidents have been associated with herbicide applications. In Oregon, application of 2,4,5-T and silvex by timber companies and the government to forest areas has brought charges of increased incidences of miscarriage by women living near the sprayed areas (American Broadcasting Company 1978; WGBH Educational Foundation 1979). It is claimed that among 8 of the women, 11 miscarriages occurred within 1 month after herbicide applications. EPA investigated these charges and found sufficient evidence of danger of the public health in sprayed areas to place an emergency ban on continued use of 2,4,5-T and silvex in these and other areas (Blum 1979). Other incidents in Oregon involved several people who complained of illness after herbicide sprayings (WGBH 1979). Abortions among cows and deer, and the deaths of fish, quail, and grouse were also reported to be associated with the sprayings (WGBH 1979). An allergist specializing in environmental 84 �medicine reported that the complaints of diarrhea and recurrent boils among the exposed people could have been caused by a dioxin contaminant in the herbicides (Anderson 1978). In northeastern Minnesota, a family reported that offspring of pigs, chickens, and rabbits that had fed in areas sprayed by a U.S. Forest Service helicopter were born deformed, or later developed deformities (ABC News 1978; Anderson 1978; Cincinnati Enquirer 1978c). For over 5 months after the spraying, the family complained of intense bellyaches, headaches, fever, nausea, diarrhea, and convulsions. An analysis of the family's water supply by the Minnesota health authorities revealed traces of a herbicide that contained 2,4-D, and 2,4,5-T. The presence of dioxins was not reported. Another source of concern is the possible effects of the massive applications of Herbicide Orange in Vietnam. Reports from some researchers indicate that numerous deformities have been found in children 6 to 14 years old (Young et al. 1978). Some reports also state that spontaneous abortions among women in sprayed areas were not uncommon, and that some people died as a result of the spraying. It has been estimated that at least 25,000 children in South Vietnam could be assumed to have acquired hereditary defects from this cause (Young et al. 1978). Others claim that these reports are virtually impossible to validate. The National Academy of Sciences concluded from their studies that there was no consistent correlation between exposure to herbicides and birth defects (Young et al. 1978). In 1969, citizens of Globe, Arizona, complained of human and animal illnesses after the U.S. Forest Service had applied 3680 pounds of silvex and 120 pounds of 2,4,5-T to the nearby Kellner Canyon and Russell Gulch (Young et al. 1978). After investigation by the Office of Science and Education and by the U.S. Department of Agriculture, it was concluded that there were no significant effects on birds and wildlife, there was no indication of illnesses in livestock greater than in other regions, and human illnesses were those that commonly occur in the normal population, except for one individual who developed skin rash and eye irritation from cleaning out an empty herbicide drum. In Swedish Lapland, two infants with congenital malformations were born to women who had been exposed to phenoxy herbicides (Young et al. 1978). Medical scientists could find no evidence to substantiate any conclusion beyond a coincidental occurrence of the birth defects and the herbicide spraying. In New Zealand, two women who had been exposed to 2,4,5-T during their pregnancies gave .birth to deformed babies (Young et al. 1978). In one case 2,4,5-T was ruled out as the cause because although the mother had been exposured to the herbicide during pregnancy, the exposure had occurred after the time in the pregnancy when the deformity is known to usually occur. No conclusions were reached on the other case. Also in New Zealand, it was reported that deformities in infants occurred in three areas of the country and that 2,4,5-T was suspect (Young et al. 1978). After an investigation, it was concluded that there was no evidence to implicate 2,4,5-T as the cause of the deformities. 85 �In Australia, skin rashes, respiratory problems, and higher incidences of birth defects and infant mortality may be associated with 2,4,5-T sprayings and dioxin contaminants (Chemical Week 1978d). Although no published reports deal with the subject, large segments of the suburban U.S. population are seasonally exposed to 2,4-D spray applications to lawns for weed control. Until 1979, silvex was also a common constituent of many of these formulations. There is no published information relating to the use of 2,4,5-T in rice fields. Rice is grown in Arkansas, Louisiana, and Texas and possibly also in Mississippi, usually in localized areas that include facilities for flooding of the fields (a requirement in rice culture). Dioxins, including TCDD's could be accumulating in the soil of these fields or in runoff channels. This appears to be a principal area of missing information with respect to continued use of these herbicides. Foods A number of human food sources have been found to be contaminated with TCDD's. Three different research teams have reported finding dioxins in the fat of cattle that had grazed on pasture treated with 2,4,5-T (Meselson, O'Keefe, and Baughman 1978; Kocher et al. 1978; Solch et al. 1978, 1980). Levels reported ranged from 4 to 15 ppt and 12 to 70 ppt, and 10 to 54 ppt, respectively. In contrast, however, samples from cattle fed ronnel contaminated with TCDD's showed no dioxins at a detection limit of 10 ppt (Shadoff 1977). TCDD's have been found at levels ranging from 14 to 1020 ppt in fish and crustaceans collected in South Vietnam (Baughman and Meselson 1973). Panel!i et al. (1980b) and Cocucci et al. (1979) found TCDD's in locally grown garden vegetables, fruit, and dairy milk supplies following the ICMESA accident in Italy in 1976. An investigator analyzed human milk samples collected in 1970 during the herbicide operations in South Vietnam, and found that they were contaminated with 40 to 50 ppt TCDD's (Baughman 1974). He reported that the mothers could have been contaminated either by direct exposure or by ingestion of contaminated foods. About 1 ppt TCDD's has been reported in breast milk from U.S. mothers living near pasture land (Meselson, O'Keefe, and Baughman 1978); however, a subsequent study of 103 samples of breast milk from mothers living in sprayed areas revealed no TCDD's at a detection limit of 1 to 4 ppt (Chemical Regulation Reporter 1980b). In 1973, TCDD's were detected in several U.S. commercial fatty acids (Firestone 1973). Other chlorinated dioxins have also been detected in foods. Tiernan and Taylor (1978) found hexa-, hepta-, and/or OCDD in 19 of 189 USDA beef fat samples at levels in excess of 0.1 ppb. Firestone reported finding hexa-CDD's, hepta-CDD's, and OCDD in gelatin samples obtained from supermarkets and in bulk gelatin (Firestone 1977). Gelatin is a byproduct of the leather tanning industry, which routinely used PCP and TCP as preservatives (U.S. Environmental Protection Agency 1978b). Total United States consumption of gelatin is estimated at 32 million kilograms per year, of which 20 percent is imported. In this study, dioxins occurred in 14 of 15 commercial gelatin samples at levels ranging from 0.1 86 �to 28 ppb total dioxins. Pentachlorophenol was also identified in most samples. 2,3,7,8-TCDD was not detected in any sample. These data are presented in Table 14. Analysis by Dow Chemical Company of fish from the Tittabawassee River, which receives the effluent from their Midland complex, revealed the presence of TCDD's, Hexa-CDD's, and OCDD in trace quantities (Dow Chemical Company 1978). Catfish from the Saginaw Bay contained 0.024 ppb TCDD. Michigan health authorities have found TCDD's in fish from the Flint, Cass, and Shiawassee Rivers. The Food and Drug Administration has recommended that Michigan set a maximum residue level for dioxins in fish at 100 parts per trillion (Toxic Materials News 1979e). TCDD's have been recently detected in leather meal, although in unquantified amounts (U.S. Environmental Protection Agency 1978b). Like gelatin, leather meal is a byproduct of the leather tanning industry. It is reported that the FDA permits up to 1 percent leather meal in swine food diets, but this level is believed to be too restrictive to be economically advantageous. Poultry feeding tests have indicated that 6 percent leather meal in the diet could be economically advantageous if the leather meal were free of dioxins. EPA recently withdrew an application to FDA for approval of the inclusion of leather meal in poultry feed because of the discovery of TCDD's in the meal. There is no published information relating to the residual level of TCDD's on harvested rice crops that have been treated with the herbicide 2,4,5-T. Pentachlorophenol has been found in dairy products, grains, cereals, root vegetables, fruits, and sugars (U.S. Environmental Protection Agency 1978e). Water Supplies Another apparent gap in information concerns drinking water. There are no published reports of studies that searched specifically for dioxins in surface or well waters used for drinking water supplies. A report from the National Academy of Sciences (1977) indicates that there are no reports of dioxins in drinking water, but does not indicate clearly whether dioxins have not been detected, or whether no research has been conducted. Dr. James Allen of the University of Wisconsin reported in 1978 that dioxins have been detected in Great Lakes waters, but apparently no data to this effect have been published. In 1978, Dow Chemical Co. reported that their analysts were unable to detect 2,3,7,8-TCDD in two surface water samples taken from the Tittabawassee River near Dow's Midland plant. The detection limit cited was 0.001 ppb. It is possible that even if toxic chlorodioxins are not present in surface waters, they might be formed at low levels during purification of public water supplies. Early research with unsubstituted dioxins showed that chlorinated dioxins could be formed from the unsubstituted dioxin by 87 �TABLE 14. DIOXINS IN COMMERCIAL GELATIN PCP, Sample No. 1 2 3 4 5 6 7 8 00 00 9 10 11 12 13 14 15 Source: Sample identity ppm Bulk domestic pork skin gelatin Bulk domestic pork skin gelatin 1975 Consumer package (Texas) 1975 Consumer package (Texas) 1977 Consumer package (Washington, D.C.) 1977 Consumer package (Washington, D.C.) 1977 Consumer package (Washington, D.C.) Imported bulk gelatin (Columbia, South America) Imported bulk gelatin-A (Mexico) Imported bulk gelat"in-A (Mexico) Imported bulk gelatin-A (Mexico) Imported bulk gelatin-B (Mexico) Commercial blend (67% domestic pork skin gelatin, 33% Hexican-A) Commercial blend (65% domestic pork skin gelatin, 35% Hexican-A) Commerical blend (91% domestic pork skin gelatin, 9% Mexican-A) 0.0 0.0 3.8 6.4 1,2,4,6,7,9 HCDD 1,2,3,6,7,9 HCDD Dioxins, ppb 1,2,3,6,7,8 1,2,3,7,8,9 HCDD HCDD 1,2,3,4,6,7,9 HpCDD 1,2,3,4,6,7,8 HpCDD 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.02 0.05 0.4 0.00 0.02,0.03 OCDD Total Dioxins 0.02 0.1 0.0 0.2 0.4 0.1 0.1 0.0 0.6 1.0 0.2 0.2 0.16 0.2 0.8 0.09 0.8 0.8 0.6 3.6 0.00 0.00 0.2 0.2 0.6 0.9 0.3,0.3 0.4,0.6 0.05,0.02 3.8,3.9 4.6,5.3 20,16 30,26 0.00 0.00 N.A. C 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.04 0.00 0.01 0.00 Q.OO 0.00 0.00 0.00 0.00 N.A. 0.03 0.2 0.03 N.A. 0.1 0.7 0:01 0.00 3.5 0.2 0.2 7.5 0.02,0.02 0.1,0.1 0.3,0.2 0.05,0.09 2.5,2.7 2.8,2.9 20,17 25,23 8.3 0.02,0.02 0.2,0.4 0.6,0.8 0.07,0.2 3.5,4.0 3.6,5.0 21,18 29,28 0.3 0.00,0.00 0.00,0.00 0.00,0.00 0.00,0.00 0.02,0.02 0.02,0.02 0.1,0.1 0.1,0.1 2.2 0.01,0.01 0.06,0.08 0.2,0.3 0.02,0.09 0.9,0.9 1.2,1.2 4.8,4.3 7.0,6.9 3.1 0.01,0.01 0.05,0.08 0.1,0.2 0.02,0.07 0.6,0.5 0.6,0.8 2.9,1.9 3.8,3.6 1.0 0.01,0.01 0.02,0.03 0.04,0.09 0.01,0.02 0.2,0.3 0.3,0.4 1.4,1.1 2.0,2.0 Firestone 1977. b Limits of quantitation were about 0.006, 0.012 , and 0.018 ppb for the HCDD's, HpCDD's, and OCDD respectively, using electron-capture gas-liquid chromatography. c N.A. = Not analyzed. �direct chlorination (Oilman and Dietrich 1957). Although no tests of this possibility have been reported, any dioxin entering a municipal drinking water system may become chlorinated during routine chlorine disinfection processes, and thus its toxicity could be greatly increased. Combustion Residues The presence of dioxins in fly ash from municipal incinerators is described in Section 3. Tests by Dow Chemical Company that found dioxins in fireplace soot and other combustion processes are also described elsewhere in the report. Here it is emphasized that these observations identify another source of exposure of the public to dioxins. To date, the available data are insufficient to allow definition of the relative importance of nonpesticide combustion as a contributor to dioxin pollution of the environment. Miscellaneous Pesticide Uses In addition to their principal uses as a raw material and an agricultural pesticide, 2,4,5-TCP and other chlorophenols that may contain dioxins are brought into contact with the public in other ways. One such use is in disinfectants (U.S. Environmental Protection Agency 1978i). These are used on surfaces of swimming pools, household and hospital sickroom equipment, food processing plants and equipment, and hospital rooms, as well as on surfaces that contact food. They are also used in bathrooms and restrooms, on shower stalls, urinals, floors, and toilet bowls. Another minor use is as a constituent of metal cutting fluids. It is not known whether any of these cutting fluids are sold commercially. Commercial products containing pentachlorophenol are readily available to the public. Examples of such products are paints containing PCP as a fungicide or preservative, and formulations for wood preserving. The latter typically contain about 4 percent PCP. Exposure of the users of PCP products is most likely to occur during use. In one reported case, however, a woman became weak and lost 20 pounds over a 3-month period that followed the application of paint containing PCP to interior paneling. Chronic inhalation of the PCP vapors from the walls was said to be the cause (U.S. Environmental Protection Agency 1978e). Dermal absorption of sodium pentachlorophenate (Na-PCP) resulted in the illness of nine newborn infants and the subsequent death of two (U.S. Environmental Protection Agency 1978e). This exposure occurred in a hospital after clothing and linens were accidentally washed with Na-PCP. Analysis of clothing and bed linens showed PCP residues ranging from 2.64 to 195.0 mg/100 g. Analysis for dioxins was not reported. Since many wood products are treated with PCP, exposure could occur by excessive handling or contact. Items such as telephone posts, fence posts, and similar products, readily accessible to the public, could present health hazards if subsequently handled. 89 �Hexachlorophene Exposures Until 1972 hexachlorophene was widely used as a bacteriostatic agent in many commercially available products. Hexachlorophene is made from 2,4,5-TCP, a known dioxin source. In September 1972 the FDA began requiring new drug applications for all drugs containing 0.75 percent or more hexachlorophene and also required that these drugs be made available only by prescription. Products containing 0.1 percent hexachlorophene as a preservative are not subject to the prescription requirement and are still marketed commercially. Hexachlorophene for use in drug and cosmetic products is apparently made from purified 2,4,5-trichlorophenol. The dioxin content of currently marketed hexachlorophene is believed to be less than 15 ug/kg (15 ppb) (World Health Organization 1977). There apparently are no published references that report positive analyses of dioxins in hexachlorophene. Sickness and death resulting from exposure to hexachlorophene have been reported, occurring primarily among children and infants (Kimbrough 1976; U.S. National Institute of Environmental Health Sciences 1978). It is not known whether dioxin contaminants are responsible. In one incident, four children died following exposure to a detergent containing 3 percent hexachlorophene (Kimbrough 1976). In 1972, 41 infants and children died and a much larger number became ill after being exposed to baby powder to which excessive quantities of hexachlorophene had been added accidentally (Kimbrough 1976). The hexachlorophene concentration in the baby powder was 6 percent. A Swedish study concerned children born to mothers who were nurses in hospitals and who had been exposed to hexachlorophene soap in early pregnancy; among 65 children, 11 malformations were found, 5 of which were severe (U.S. National Institute of Environmental Health Sciences 1978). Out of 68 children born to unexposed mothers, only one slight malformation was observed. OCCUPATIONAL EXPOSURE Except for the 1976 disaster at Seveso, most clearly recognized human injuries associated with dioxins have been suffered by persons who came into contact with the chemicals as a result of their occupation. The most directly affected probably would be workers in plants of the chemical manufacturing industry where the dioxins are created. Other industries and activities, however, also use dioxin-contaminated chemical products and thus represent another source of worker exposure (for purposes of this report, the exposure of Vietnam military personnel to dioxins is considered occupational). Still other occupational exposures result from work in analytical or research laboratories and from handling of chemical wastes. This report section describes the reported incidents and the potential for human exposure due to occupational activities. A large-scale study of occupational exposure to dioxins is now underway by the National Institute for Occupational Safety and Health (NIOSH). With 90 �cooperation from the chemical industry, major unions, and the Department of Defense, NIOSH is compiling a registry of the population of chemical workers in the United States who have had documented exposure to 2,3,7,8-TCDD, either in the manufacture of herbicides or in industrial accidents. Once this registry has been developed, NIOSH plans to evaluate trends in mortality of the exposed workers and, if the data permit, will consider conducting studies of morbidity and reproductive effects (Robbins 1979). The NIOSH program will augment similar studies in progress in connection with present and former workers exposed to dioxins in Jacksonville, Arkansas, and Nitro, West Virginia (Occupational Safety and Health Reporter 1979). Chemical Manufacturing Industry More than 200 dioxin-related industrial accidents occurred around the world during the 30 years prior to 1979 (American Industrial Hygiene Association Journal 1980). The following paragraphs represent only a sampling of these incidents, most of which involve the manufacture of 2,4,5-TCP. Table 15 summarizes some of the other incidents not described in detail. Table 16 is a sampling of the incidents involving plant accidents. The earliest major incident was an explosion in 1949 at a plant of the Monsanto Company in Nitro, West Virginia. This plant operated from 1948 to 1969, and the explosion was reported to have affected 228 people (Whiteside 1977; Young et al. 1978). The symptoms included melanosis, muscular aches, nervousness, and intolerance to cold, in addition to chloracne. A current occupational study of the long-term effects of dioxin exposure is being conducted of 121 people who were working in the plant at the time, including all of those who developed chloracne as a result of the accident. Preliminary study reports indicate no excess deaths from cancer or cardiovascular disease among these workers (American Industrial Hygiene Association Journal 1980). In 1953, an explosion occurred in Germany at the factory of Badischer Anilin and Soda-Fabrik, which was producing 2,4,5-TCP by hydrolysis of 1,2,4,5-tetrachlorobenzene with sodium hydroxide in a solvent of methanol (Goldmann 1972). Following the explosion the safety valves released vapors, which filled all reactor rooms on all four floors of the plant. After a few minutes, vapors that had not been withdrawn with exhaust fans had condensed as solids on the apparatus, walls, windows, and doors. Chloracne developed in 42 people, 21 of whom also developed disorders of the central nervous system or internal organs. In1 addition, 5 years after the explosion a worker replacing a gasket on one of the reactors developed several disorders a few days later; one year later the worker died. An explosion at the TCP-producing factory of the Coalite and Chemicals Products at Derbyshire, U.K., resulted in 79 workers contracting chloracne (May 1973). Six months after an explosion in the Netherlands at the Philips-Duphar plant, which was producing 2,4,5-TCP, 9 of 18 men working on decontaminating the plant contracted chloracne (World Health Organization 1977). 91 �TABLE 15. REPORTED INCIDENTS OF OCCUPATIONAL EXPOSURE TO DIOXINS DURING ROUTINE CHEMICAL MANUFACTURING3 Year 1949 1952 1952-53 1954 1956 1956 1960 1964 1964 1965-69 1970 1972 1973 1974 1975 Country West Germany West Germany West Germany West Germany United States United States United States U.S.S.R. United States Czechoslovakia Japan U.S.S.R. Austria West Germany United States Chemical produced Manufacturer/plant location N.A.b/Nordrhein, Westfallen N.A./N.A. Boehringer/N.A. Boehringer, Ingelheim/Hamburg Diamond Alkali/Newark, New Jersey Hooker/N.A.c Diamond Shamrock/N.A.c N.A./N.A. Dow Chemical /Midi and, Michigan Spolana/N.A. N.A./N.A. N.A./N.A. Linz Nitrogen Works/N.A. Bayer/Uerdi ngen Thompson Hayward/Kansas City, Mo. PCPJCP TCP TCP TCP; 2,4,5-T 2,4-D; 2,4,5-T TCP TCP 2,4,5-T 2,4,5-T TCP PCP; 2,4,5-T TCP 2,4,5-T 2,4,5-T TCP . Adapted from Young et al. 1978. ° N.A. - Not available. Not known whether occupational exposure was involved in the incident. Number of persons exposed 17 60 37 31 29 N.A. N.A. 128 60 78 25 1 50 5 N.A. �TABLE 16. OCCUPATIONAL EXPOSURES TO DIOXINS THROUGH ACCIDENTS IN a THE CHEMICAL MANUFACTURING INDUSTRY Manufacturer/1 ocati on Product involved Year Country 1949 United States Monsanto/Nitro, West Virginia TCP 1953 West Germany BSAF/Ludwigshafer 1956 France Rhone Poulene/Grenoble TCP 2,4,5-T TCP 1962 Italy 1963 Netherlands 1966 Number of workers affected 228 55 17 TCP 5 Phi 1 ips-Duphar/ Amsterdam TCP 50 France Rhone Poulene/Grenoble TCP 21 1968 United Kingdom TCP 79 1976 Italy Coalite and Chemicals Products/ Bolsover, Derbyshire ICMESA/Meda TCP 134b CO . Adapted from Young et al. 1978. These were not workers but local residents (124 children and 10 adults); no workers were reported affected. �During the Seveso incident, the public was more seriously affected, but the plant workers were also exposed to dioxins. Reports are fragmentary and sometimes conflicting. A company-sponsored report says that of the 10 workers in the plant at the time of the accident, none, not even those who came in direct contact with the reactor, showed signs of exposure; further, a year later, none of the plant workers showed any signs of disease associated with dioxin toxicity (Reggiani 1977). Another report states that one volunteer worker, after helping to clean out the material that remained in the reactor after the accident, developed severe chloracne (Parks 1978). Another report states that among 170 workers exposed to the contamination, 12 developed chloracne, 29 developed liver disease, 17 developed high blood pressure, and 20 others suffered from other various disorders (Zedda, Cirla, and Sal a 1976). Finally, another report states that 64.5 percent of 141 former workers suffer from liver problems and others suffer from a variety of other complaints; 79 of 160 workers involved in the cleanup campaign show chromosomal abnormalities (Chemical Week 1978a). Workers at the Vertac plant in Jacksonville, Arkansas, apparently have been affected by exposure to dioxins, even though no catastrophic event occurred during the many years the plant produced 2,4,5-TCP. Graphic accounts of chloracne attacks in plant workers appeared in an investigative article published in a popular U.S. magazine (Fadiman 1979). In June 1979, Arkansas health officials found signs of chloracne in 13 of the 74 current Vertac employees (Richards 1979c). In July 1979, a task force of medical experts began an intensive examination of about 150 present and former employees; no definitive conclusions have been reported. Although not necessarily employees of chemical manufacturers, some workers undergo occupational exposure to dioxins in the handling or transportation of bulk chemicals outside of the plant. In one reported incident after the railway derailment in Sturgeon, Missouri, low levels of 2,3,7,8-TCDD were found in the blood of two of the cleanup workers (Chemical Week 1979d, 1979e, and 1979i; Poole 1979; Taylor and Tiernan 1979). These were employees of a firm hired by the railroad to clean up the spill. In a similar incident in Sweden, railroad workers were exposed to 2,4-D and 2,4,5-T. A medical study concluded that these herbicides showed a possible tumor-inducing effect (Young et al. 1978). The presence of dioxins apparently was not considered in this study. Use of Chemical Products When makers of dioxin-contaminated products sell these products to other industries or organizations, the personnel of these secondary users are subject to occupational exposure to dioxins. Table 17 lists several related industries that process or handle chemical products with a potential dioxin content. It is estimated that 80 percent of all pentachlorophenol produced is used in wood-treating operations (Arsenault 1976; American Wood Preservers 94 �TABLE 17. INDUSTRIES USING DIOXIN-RELATED CHEMICALS Industry Chemical(s) Process application Texti1es TCP Process water fungicide Leather tanning TCP, PCP Process water fungicides Wood preserving PCP Active ingredient in dip vat/pressure treatment Pulp and paper TCP, PCP Process water slimicide, fungicide Pesticide formulators and applicators 2,4,5-T 2,4-D si 1 vex ronnel erbon hexachlorophene Active Active Active Active Active Active Automotive TCP Metal cutting fluids, foundry core washes Miscellaneous industries TCP Slimicide in cooling tower waters Household and industrial cleaning products TCP, Active ingredient disinfectant Bui 1di ng/constructi on PCP Termite control Drug and cosmetics hexachlorophene Product preservative or active ingredient Paint TCP, PCP Preservati ve/mi1dewci de Farming (cattle) 2,4,5,-T, 2,4-D Range!and weed control Railroad, telephone, (construction and maintenance) 2,4,5-T si 1 vex 2,4-D Weed control on rights-of-way ID C7I ingredient ingredient ingredient ingredient ingredient ingredient formulated formulated formulated formulated formulated formulated or or or or or or sprayed sprayed sprayed sprayed sprayed sprayed hexachlorophene �Institute 1977; U.S. Environmental Protection Agency 1978e). Exposure in this secondary industry may occur during the mixing of the PCP crystals and solvent (American Wood Preservers Institute 1977). Many of the larger wood-treating operations now use automatic closed mixing systems, which limit the chances for worker exposure. Chloracne symptoms have developed, however, in workers in one wood-treating plant; the exposures resulted from manual opening and dumping of bagged PCP (U.S. Dept. HEW 1975). Workers also may be exposed to PCP by handling of wood after treatment. Other uses for pentachlorophenol and its sodium salt are in cooling tower water treatments, in pulp and paper mills, and in tanneries (U.S. Environmental Protection Agency 1978e). Potential for worker exposure therefore exists in these industries. Cooling tower waters from one 2,4,5-TCP facility have recently been found to contain ppb levels of TCDD's (see Volume II of this series). People involved in the application of herbicides manufactured from or formulated with 2,4,5-TCP and derivatives may be exposed to dioxin contaminants. These include workers involved in aerial applications and those employed by commercial lawn-care companies who apply phenoxy herbicides manually. Exposures to Herbicide Orange-Thousands of military personnel were exposed during the Vietnam conflict to Herbicide Orange; these exposures are currently the topic of considerable litigation and are not outlined in detail in this report. The General Accounting Office (GAO) notes that 4800 veterans have asked for treatment for exposure to Herbicide Orange (Toxic Materials News 1979d), and the suits are being brought against former manufacturers, reported to include Dow Chemical, Hercules, Diamond-Shamrock, Monsanto, Northwest Industries, and North American Philips (Chemical Week 1979c). Summaries of the situation were published in Science (Hoiden 1979) and by the New York Times (Severe 1979). Chemical Laboratories In 1957, a research worker in a laboratory synthesized the 2,3,7,8tetrabromo dioxin. That same year, another researcher first synthesized 2,3,7,8-TCDD (about 20 grams) by chlorination of unsubstituted dioxin. In both cases, on completion of these achievements, the researcher was hospitalized (Rappe 1978). The chemical laboratory continues to be a potential source of human exposure to dioxins. One case is reported involving three scientists in the United Kingdom (May 1973). Although it was believed that adequate precautions had been taken, all three were afflicted with various disorders. Two of the scientists had been working on the synthesis of dioxin standards. They had performed the synthesis under a fume hood and had worn overalls and dispos- 96 �able plastic gloves. Both persons developed chloracne in addition to other symptoms. The third scientist, who had been working with dilute dioxin standards, had taken similar protective measures. He did not develop chloracne but he exhibited other symptoms, including hirsutism and excess cholesterol in the blood. In 1978, Dow Chemical Company reported that an employee contracted chloracne after disposing of laboratory wastes contaminated with dioxins. He reportedly had not followed standard safety procedures. Dow has developed a set of elaborate laboratory safety rules to be used when working with dioxins. Similarly, stringent procedures are exercised by independent laboratories who analyze samples containing dioxins. The Brehm Laboratory of Wright State University, Dayton, Ohio, includes a specially equipped laboratory with restricted access, specially trained personnel, and tight internal quality control based on mandatory routine wipe tests. All personnel use disposable gowns, gloves, and shoe covers. "Cradle-to-grave" control is exercised for all reagents, wash water, disposable clothing, towels, and all other materials used or consumed in the laboratory; nothing enters the sewer or is discarded as common trash. Everything enters sealable transportation barrels to be discarded in an environmentally acceptable manner. Gas chromatographs are vented through charcoal filter cartridges, which are routinely discarded into the barrels. Any dusty samples are handled in a special filtered glove box with total control of all dust and unused sample material. This laboratory has experienced no incidents of dioxin poisoning (Taylor 1980). Waste Handling Another possible route of exposure to workers is the handling of production wastes generated from manufacturing and formulation processes. Not only the employees of the company that generates dioxin-containing wastes can be affected by these wastes, but also those who work for a contract waste disposal firm. The incident at Verona, Missouri, indicates that the waste disposal company owner and/or his employees did not recognize the dangers of wastes with potential dioxin content. The synthesis of pentachlorophenol and its use in wood treatment also generate waste products. A current study sponsored by the EPA Office of Solid Wastes includes an analysis of sludge samples from various locations within three industrial plants that produce either trichlorophenol, pentachlorophenol, or hexachlorophene (U.S. Environmental Protection Agency 1978d,). Also being sampled is a wood-preservation operation in which pentachlorophenol is used. Initial results have shown low-ppm concentrations of hexa-CDD's, hepta-CDD's, and OCDD in sludges resulting from PCP production. Concentrations of the dioxins are not specified, but it is stated that the levels are below those designated as toxic in the published literature. Also, 0.06 ppm OCDD and low levels (not quantified) of hexa-CDD's and hepta-CDD's were found in the soil in the vicinity of the product storage area. 97 �SECTION 5 ENVIRONMENTAL DEGRADATION AND TRANSPORT This section addresses the fate of dioxins once they are released to the environment. Subsections on biodegradation and photodegradation deal with recent literature relating to biochemical and physical actions of the environment as they affect the integrity of the dioxin structure. Subsections on physical and biological transport deal with the movement of dioxins in soil, water, and air and with the uptake of dioxins by plants and their fate in animals at various trophic levels. BIODEGRADATION In assessment of the persistence of a substance in the environment, the susceptibility of that substance to biodegradation* is a primary concern. Several studies on the biodegradabilityt of dioxins are described in the literature. The investigations show that dioxins exhibit relatively strong resistance to biodegradation, though they may not necessarily be totally recalcitrant. Most of the work has focused on 2,3,7,8-TCDD because of its extreme toxicity. This dioxin has been studied in both aqueous and soil environments, and results have been somewhat equivocal. Only one study (Kearney et al. 1973) has examined the biodegradability of another dioxin, 2,7-DCDD. Data from this study indicate that this dioxin can be at least partially degraded in soils. Several dioxin biodegradation studies are described in the following paragraphs. Approximately 100 strains of microbes that had previously shown the ability to degrade persistent pesticides were tested for their ability to degrade 2,3,7,8-TCDD. After incubation, extracts from microorganisms were prepared and analyzed for metabolites by thin-layer chromatography. Of the strains tested, five showed some ability to degrade the dioxin. Biodegradation: the molecular degradation of an organic substance resulting from the complex actions of living organisms. A substance is said to be biodegraded to an environmentally acceptable extent when environmentally undesirable properties are lost. Loss of some characteristic function or property of a substance by biodegradation may be referred to as biological transformation. (CEFIC 1978) Biodegradability: the ability of an organic substance to undergo biodegradation. 98 �Ward and Matsumura studied the biodegradation of 14C-labelled 2,3,7,8TCDD in Wisconsin lake waters and sediments and reported in 1977 that the dioxin may be genuinely metabolized in aqueous systems, but that the rate is very low. They concluded that there is an optimum time for microbial degradation, probably 1 month, and that during this period available 2,3,7,8-TCDD is degraded while the nonavailable fraction is bound to the water sediments. The limited degradation of 2,3,7,8-TCDD is favored by the presence of sediment, microbial activity, and/or organic matter in the aqueous phase. The observed half-life of 2,3,7,8-TCDD in sedimentcontaining lake waters was 550 to 590 days; the half life in waters without sediment was longer. Kearney and coworkers studied two types of soil, which were incubated with 2,3,7,8-TCDD at concentrations of 1, 10, and 100 ppm and with 14Clabeled 2,3,7,8-TCDD at concentrations of 1.78, 3.56, and 17.8 ppm (Kearney et al. 1973a). The two soils were also inoculated with 14C-labeled 2,7-DCDD at concentrations of 0.7, 1.4, and 7.0 ppm. The soil types were Hagerstown silt clay loam, which is relatively high in organic matter and microbial activity, and Lakeland loamy sand, which is low in organic matter and microbial activity. Over a 9- to 10-month period, the soil samples were monitored weekly for evolution of gaseous 14C02 as an indication of microbial degradation of the labeled dioxins. Very little C02 was liberated from soils containing either labeled or unlabeled 2,3,7,8-TCDD. In most cases 75 to 85 percent of the dioxin was recovered from both soil types up to 160 days after addition. No metabolites were found in TCDD-treated soil 14 after 1 year. About 5 percent of the 14 C-2,7-DCDD had degraded to liberate C02 after 10 weeks. Concentrations of 14C-2,7-DCDD in the soil had a slight effect on 14C02 evolution. It was postulated that the decrease in C02 liberation at the highest level may have resulted from the toxicity of the DCDD isomer to the microbes at this concentration. Evolution of 14C02 was significantly higher in the Lakeland soil than in the Hagerstown soil. Analysis of DCDD-treated soil extracts also revealed the presence of metabolites, but the major metabolite could not be identified. In the same study, incubation of a clay loam (with relatively low organic matter) to which 1414 C-2,3,7,8-TCDD had been applied led to liberation of a "very small amount of C02" after 2 weeks. The U. S. Air Force studied test plots in Utah, Kansas, and Florida to determine the soil degradation rate of 2,3,7,8-TCDD under field conditions (Young et al. 1976). The three test plots were considered representative of various climatic conditions and soil types. Herbicide Orange containing 3700 ppb 2,3,7,8-TCDD was applied to all three plots at a rate of 4480 kg/hectare. Initial soil concentrations of the dioxin were not reported for any of the sites. Composite samples from the upper 15 cm of each soil were taken from time to time after the initial herbicide application, and analyzed for both the herbicide and 2,3,7,8-TCDD. Results are presented in Table 18. 99 �TABLE 18. CONCENTRATIONS OF HERBICIDE ORANGE AND 2,3,7,8-TCDD IN THREE TREATED TEST PLOTS3 Test plot Utah Kansas Florida Total . herbicide, Days after application ppm 2,3,7,8-TCDD, ppb 15.0 7.3 5.6 3.2 2.5 282 637 780 1000 1150 8490 4000 2260 2370 960 8 77 189 362 600 659 1950 1070 490 210 40 <1 c 0.255 c c 5 414 513 707 834 1293 4897 1866 824 508 438 <10 0.375 0.250 0.075 0.046 c c . Plots treated with 4480 kg herbicide per hectare. Composite sample from upper 0 to 15 cm layer of soil. Not analyzed. 100 c 0.042 �From these data and other leaching data, the Air Force concluded that the disappearance of 2,3,7,8-TCDD was most likely due to degradation by soil microbes, because dioxin concentrations in the 15- to 30-cm layer indicated that leaching was insignificant. The Air Force report further stated that dioxin degradation was most rapid in the Kansas soil (Ulysses silt loam), followed by the Florida soil (Lakeland Sandy loam), and finally the Utah soil (Lacustine clay loam), but that variations in soil and climate had little overall influence on dioxin persistence. It was also reported that the initial breakdown rate was rapid, but decreased substantially over the test period. On the basis of this observation the investigators speculated that microbial enzymes responsible for herbicide metabolism and possibly dioxin metabolism are inducible. In an evaluation of the Air Force studies, Commoner and Scott (1976) came to different conclusions. After constructing semilogarithmic plots of dioxin concentrations in soil against days after incorporation of the dioxin, they concluded: (1) that there was no evidence that the rate of degradation changed with time; and (2) that degradation appeared to be more rapid in the Florida soil than in the Kansas soil (opposite of the Air Force conclusion). In another Air Force study with dioxin-contaminated soil the effects of nutrients and mixing on 2,3,7,8-TCDD degradation were assessed (Bartleson, Harrison, and Morgan 1975). Pots containing either test soils or control soils were placed outdoors and in a greenhouse. The soils were analyzed after 9 and 23 weeks. Soils tested in the greenhouse were moistened with a nutrient solution. The results are presented in Table 19. TABLE 19. DEGRADATION OF 2,3,7,8-TCDD IN SOIL a (parts per trillion 2,3,7,8-TCDD) Length of exposure, weeks 0 Controls 9 23 1100 1000 520 530 640 810 460 530 1100 - 1300 Outdoor exposure Tilled (top layer) Until led Greenhouse Tilled (top layer) Untilled Source: Bartleson, Harrison, and Morgan 1975. 101 �The investigators concluded that the accelerated rate of degradation observed in soil from the pots in the greenhouse during the first 9-week period was probably due to increased microbial populations resulting from initial soil aeration and increased soil temperatures in the pots. Reduction in the rate of breakdown after 9 weeks may have been caused by leaching or entrapment of dioxin in the bottom soil layer, which had not been mixed. It was also proposed, however, that the nutrient solution together with light or aeration caused either a direct chemical breakdown of 2,3,7,8-TCDD in the soil or an increase in microbial populations that accelerated breakdown. Because green algae were observed on the surface of the greenhouse pots between till ings, it was also postulated that the algae were partly responsible for the degradation. This study was also evaluated by Commoner and Scott (1976), who concluded that mixing, nutrients, and increased exposure to sunlight did not significantly enhance degradation of 2,3,7,8-TCDD in soil. Pocchiari ( 9 8 attempted to stimulate the microbial degradation of 17) 2,3,7,8-TCDD in samples of Seveso soil contaminated with the dioxin from the 1976 ICMESA accident. The dioxin-contaminated soil samples were either inoculated with promising microorganisms (according to the previously described results of Matsumura and Benezet in 1973) or enriched by the addition of organic nutrients. No positive degradation effects have been found. Investigators from the Microbiological Institute in Zurick, Switzerland, have found that microbes cannot contribute quickly or efficiently to the decontamination of soil-bound 2,3,7,8-TCDD, although they might contribute slowly (Huetter 1980). The latter point is supported by the observation of two polar bands in thin layer chromatographs of some microbial incubations. Huetter and coworkers also have observed that when 2,3,7,8-TCDD is incubated with soil for a prolonged period of time, it is not as extractable as when it is freshly added to the soil, indicating that recoverability of the dioxin becomes increasingly more difficult with time. This information raises questions about the accuracy of work done by others in the past to measure the soil half-life of 2,3,7,8-TCDD. Preliminary findings of studies under way in Finland indicate that 2,3,7,8-TCDD may be slowly biodegraded by anaerobic microorganisms in an organic matrix used for secondary treatment of chlorophenolic wastewaters from paper pulping operations (Salkinoya-Salonen 1979). Klecka and Gibson (1979) have recently reported that unsubstituted dibenzo-p-dioxin can be readily metabolized by a mutant strain of Pseudomonas (sp. N.C.I.B. 9816 strain II) when an alternative source of carbon such as salicylate is available. The dioxin molecule was metabolized first to cis-l,2-dihydroxy-l,2,dihydrodibenzo[l,4]dioxan (I), which was subsequently dehydrated to yield 2-hydroxydibenzo[l,4]dioxan (II) as the major metabolite. The authors reported finding no organisms capable of utilizing dibenzo-p-dioxin as a sole carbon source. 102 ii �PHOTODEGRADATION Photodegradation is the process of breaking chemical bonds with light. The process, also known as photolysis, involves the breakdown of a chemical by light energy, usually in a specific wavelength range. In photodegradati on of dioxins the ultraviolet wavelengths of light have been-shown to be the most effective. In most photolysis studies, scientists are interested in determining one or more of the following parameters: 1. Photolysis reaction rates 2. Photolysis reaction products 3. Wavelength(s) required for photolysis 4. Other specific conditions required for photolysis The photolysis of chlorinated aromatic compounds usually involves loss of a chlorine molecule to a free radical, or loss through nucleophilic displacement if a solvent or substrate molecule is present. These mechanisms may be influenced by the presence of other reagents or the nature of the reaction medium. Photolysis studies have clearly shown that dioxins may be photolytically degraded in the environment by natural sunlight. The extent to which this mechanism actually removes or degrades dioxins in the "real world" environment is difficult to assess, but of all the possible natural removal mechanisms, photolysis appears to be the most significant. It should be noted that photolysis apparently results in the removal of one or more chlorine atoms from the dioxin molecule. Removal of chlorine from 2,3,7,8-TCDD may make it less toxic, but the basic dioxin structure remains. When penta-CDD is photodegraded, it may go to a TCDD isomer. (For further discussion see pp. 138-139 of Section 6.) Several dioxin photodegradation studies are discussed in the paragraphs that follow. Major findings from these studies are summarized in Tables 20 and 21. Crosby et al. (1971) studied photolysis rates of 2,3,7,8-TCDD, 2,7-DCDD, and OCDD dissolved in methanol. Samples were irradiated with natural sunlight or artificial sunlight with a light intensity of 100 MW/cm2 at the absorption maximum of 2,3;7,8-TCDD (307 nm). Irradiation of a single solution of 2,3,7,8-TCDD in methanol for 24 hours in natural sunlight resulted in complete photolysis to less chlorinated dioxin isomers. The degradation of 2,7-DCDD was at least initially more rapid than that of 2,3,7,8-TCDD. After 6 hours of irradiation in artificial ultraviolet light, about 30 percent of the 2,7-DCDD remained unreacted whereas almost 50 percent of the 2,3,7,8-TCDD remained unreacted. The amount of 2,7-DCDD remaining after 24 hours was not reported. The OCDD was photolyzed much more slowly than the TCDD or DCDD isomers; after 24 hours, over 80 percent 103 �TABLE 20. Physical conditions Light source TCDD in methanol Artificial 2 (100 |jw/cm ) TCDD in nethanol TCDD (crystalline) in water PHOTODEGRADATION OF 2,3,7,8-TCDD Length of exposure Amount degraded, % Reaction products 24 h 100 Natural sunlight 7h 100 NRa Crosby et al . 1971 Artificial (sunlamp) NR 0 NAb Crosby et al. 1973 96 h 0 >0 NR Plimmer et al. 1973 0 NR Crosby et al . 1971 50 NR NR Stehl et al. 1973 Stehl et al. 1973 TCDD on soil TCDD in benzene/water/ surfactant Artificial (-sunlamp) TCDD crystals on glass plate Natural sunlight 14 days TCDD in isooctane and 1-octanol Artificial (G.E. RS sunlamp) 40 min 24 h 100 TCDD in Herbicide Orange, on glass Natural sunlight 6 h TCDD in commercial Esteron herbicide, on glass Natural sunlight 6 h 70 TCDD in Esteron base, on glass Natural sunlight 2h 90 Trichlorodibenzo-p-dioxin, Dichlorobenzo-p-dioxin 60 * NR = Not reported. NA = Not applicable. 0 (continued) Reference NR Crosby et al . 1971 Crosby and Wong 1977 NR Crosby and Wong 1977 Crosby and Wong 1977 �TABLE 20 (continued) Physical conditions Length of exposure Light source Amount degraded, % Reaction products Reference TCDD in Herbicide Orange, on plant leaves 6 h 6h 100 70 Crosby and Wong 1977 TCDD in Herbic.ide Orange, on soil Sunlight 6 h 10 Crosby and Wong 1977 TCDD on silica gel Artificial A. >290 ran 7 days 92 NRa Gabefuigi 1977 TCDD on silica gel Artificial X = 230 nm 7 days 98 NR Gabefuigi 1977 TCDD in Seveso soil with ethyl oleatexylene mixture o tn Sunlight Sunlight artificial (Phillips MLU 300 W) 7 days >90 NR Bertoni 1978 3 days 100 TCDD in 1-hexadecylpyridinium chloride (CPC) Artificial 4h >90 NR Botre et al . 1978 TCDD in sodium dodecyl sulfate (SDS) Artificial 4h 8h =50 =100 NR NR Botre et al. 1978 TCDD in methanol Artificial 4h 8h =50 =75 NR NR Botre et al . 1978 9 days >90 NR Crosby 1978 TCDD in Seveso soil/ ^' tural treated with aqueous sunlight olive oil solution or olive oil/cyclohexanone TCDD in emulsif iable si 1 vex f omul at ion Natural sunlight =8 days 50 NR Nash and Bealle 1978 TCDD in granular si 1 vex formulation Natural sunlight • a3. 5 days 50 NRa Nash and Bealle 1978 a 0 NR = Not reported. NA = Not applicable. �TABLE 21. PHOTODEGRADATION OF DCDD AND OCDD Light source Length of exposure OCOD in methanol Artificial UV light 100 uw/cm2 24 h >20 OCDD on filter paper Artificial sunlight NRa More rapid in natural sunlight than artificial UV light NR Arsenault 1976 Physical conditions Natural sunlight Amount degraded, % Reaction products Series of chlorinated dioxins of decreasing chlorine content Reference Crosby et al. 1971 OCDD in oil (mineral or petroleum) Natural sunlight 16 h 66 NR Arsenault 1976 OCDD - no oil Natural 16 h 20 NR Arsenault 1976 OCDD/benzene- hexane Mercury UV lamp 4h 70 Hexa-CDD, hepta-CDO, penta-CDD OCDD/benzene-hexane Mercury UV lamp 24 h 90 Hexa-CDD, hepta-CDD, Buser 1976 penta-CDD, TCDD (trace) OCDD in isooctane Artificial UV light 18 h 20 NR Stehl et al. 1973 OCDD in 1-octanol Artificial UV light 20 h 6 NR Stehl et al. 1973 DCDD in methanol Artificial UV light =6 h =70 NR Crosby et al. 1971 DCDD in isooctane and 1-octanol Artificial UV light 40 min 50 NR Stehl et al. 1973 NR = Not reported. Buser 1976 �of the initial OCDD (2.2. mg/liter) remained unreacted. Analysis of reaction products indicated chlorinated dioxins of reduced chlorine content. In another study the degradation of OCDD on filter paper was reported as being more rapid in natural sunlight than in artificial ultraviolet light (Arsenault 1976). Degradation of OCDD also proceeded more rapidly in the presence of mineral oil or a petroleum oil solvent than in the absence of oil. When OCDD in oil was exposed to natural sunlight, 66 percent was decomposed in as little as 16 hours. When exposed in the absence of oil, only 20 percent was decomposed within 16 hours. No TCDD's were found in the decomposition products. The same report describes a study of the rate of OCDD degradation on the surfaces of wooden poles treated with PCP-petroleum and Cellon. Preliminary results show that the OCDD is rapidly degraded. Breakdown products are not reported. In tests involving exposure of a crystalline water suspension of 2,3,7,8-TCDD to a sun!amp, the insolubility of the dioxin caused difficulties. Irradiation apparently had no effect on the water suspension. A crystalline state may prohibit the loss of chlorine or obstraction of hydrogen atoms from each other (Plimmer 1978a). When a benzene solution of 2,3,7,8-TCDD was added to water stabilized with a surfactant and irradiated with a sun!amp, the dioxin content was reduced (Plimmer et al. 1973). In another study when 2,3,7,8-TCDD was applied to dry or moist soil, irradiation caused no change after 96 hours. Similar results were obtained by applying this substance to a glass plate and irradiating up to 14 days (Crosby et al. 1971). Buser (1976) irradiated samples of a solution of OCDD in benzene-hexane for 1 to 24 hours with a mercury ultraviolet lamp. After 4 hours of exposure, 30 percent of the OCDD remained unchanged; the major reaction products were hexa- and hepta-CDD's and trace amounts of penta-CDD's. After 24 hours of irradiation, the hexa- and hepta-CDD's still constituted the major reaction products, with significant amounts of penta-CDD's and trace amounts of TCDD's. Only 10 percent of the initial OCDD remained unchanged. It was concluded that since some commercial products contain up to several hundred ppm of the octa- and hepta-CDD's, photolytic formation of more toxic polychlorinated dioxins could have environmental significance. Exposure of TCDD's and DCDD's in iso-octane and 1-octanol to artificial sunlight (General Electric RS sunlamp) showed that both substances had half-lives of about 40 minutes in each solvent (Stehl et al. 1973). Analysis of the mixtures after 24 hours of irradiation showed no 2,3,7,8-TCDD at a detection limit of 0.5 ppm. A bioassay of rabbit ear skin tissue to which the photolysis products had been applied revealed no chloracnegenic activity. 107 �When a solution of OCDD and iso-octane was exposed to artificial sunlight, about 80 percent of the OCDD remained unreacted after 18 hours. With a solution of OCCD and 1-octanol, about 94 percent of the OCDD remained unreacted after 20 hours (Stehl et al. 1973). In a series of tests, thin layers of Herbicide Orange containing 15 ppm 2,3,7,8-TCDD were exposed to summer sunlight in glass petri dishes (Crosby and Wong 1977). After 6 hours, just over 40 percent of the dioxin remained. A commercial herbicide composed of butyl esters of 2,4-D and 2,4,5-T and containing 10 ppm 2,3,7,8-TCDD was exposed in the same manner; after 6 hours only about 30 percent of the initial dioxin remained. A commercial mixture containing no herbicides, but with 10 ppm 2,3,7,8-TCDD was also exposed to sunlight on glass petri dishes. The original dioxin concentration was reduced by about 90 percent after 2 hours. Herbicide Orange was applied in droplets to excised rubber plant leaves and to the surface of Sacramento loam soil; the samples were then exposed to sunlight. At an application rate of 6.7 mg/cm2 of leaf surface no TCDD's were detected on the leaves after 6 hours. At a lower application rate of 1.3 mg/cm2, however, about 30 percent of the TCDD's remained after 6 2hours. It was also reported that upon application to the soil (10 mg/cm ) approximately 90 percent of the dioxin remained after 6 hours. The authors attributed the lesser degree of photolysis of 2,3,7,8-TCDD on the soil partly to shading of lower layers by soil particles. Investigators in this study concluded that there are three requirements for dioxin photolysis: 1. Dissolution in a light-transmitting film. 2. Presence of an organic hydrogen donor. 3. Ultraviolet light. In another study, 2,3,7,8-TCDD deposited on silica gel was irradiated with light having a wavelength greater than 290 nm. The original concentration of the dioxin was reduced by 92 percent after 7 days. When irradiation was done with light of shorter wavelength (>230 nm), the dioxin concentration was reduced by 98 percent after 7 days. It was concluded that cleavage of 2,3,7,8-TCDD was possible without a proton donor if the intensity of the sun at ground level was great enough to supply the required irradiation (Gebefuigi, Baumann, and Korte 1977). In a study reported by Bertoni et al. (1978) about 150 ml/m2 of an ethyloleate-xy1ene mixture was sprayed on a 1-cm-deep sample of Seveso soil contaminated with 2,3,7,8-TCDD. More than 90 percent of the 2,3,7,8-TCDD was destroyed after 7 days of sunlight exposure. When a dioxin sample was placed in a room sprayed with the ethyloleate-xy1ene mixture, disappearance of the dioxin was almost complete after 3 days exposure under a Phillips MLU 300 W lamp. The xylene was used to reduce viscosity, although ethyloleate was just as effective when used alone. The more rapid photolysis in the room was attributed mainly to the smooth walls of the room receiving the full intensity of the radiation, including the wavelength of light that was absorbed most readily by dioxins. 108 �The smooth gradual decrease of dioxin concentration in the 1-cm-deep soil samples was unexpected because ultraviolet light does not penetrate soil. It was hypothesized that dioxin decomposition below the soil surface could result either from a diffusion mechanism in the oleate medium or from photolytic reactions occurring through long-lived free radicals. The solubility and photodecomposition of 2,3,7,8-TCDD in cationic, anionic, and nonionic surfactants was studied by use of both pure dioxin samples and contaminated materials obtained from the Seveso area (Botre, Memoli, and Alhaique 1979). To test the effectiveness of the solubilizing agents, homogeneous soil samples were treated twice with surfactant and then three times with the same volume of water to remove the surfactant. Extracts from the residual soil were then obtained with benzene and methanol, and the extracts were analyzed for 2,3,7,8-TCDD. Untreated contaminated soil samples were used for standards. In the pure dioxin solubilization study, 4 ml of surfactant was used to treat the residues. Methanol was used as the reference solvent. The surfactants used were sodium dodecyl sulfate (SDS), an anionic surfactant, 1-hexadecylpyridinium sorbitan monooleate (Tween 80), hexadecyltrimethylammoniurn bromide, and 1-hexadecylpyridinium chloride (CPC). Results showed that CPC was the best solubilizing agent for contaminated soil taken from the Seveso area, whereas in the pure dioxin experiment the differences were slight. Photodecomposition experiments performed using 2,3,7,8-TCDD dissolved in surfactants and in methanol also revealed CPC as the superior medium. Irradiation with an ultraviolet lamp for 4 hours destroyed about 90 percent of the dioxin in the CPC solution. Only 50 percent of the dioxin in the SDS solution was destroyed after 4 hours of irradiation, although almost 100 percent disappeared after 8 hours. Over 25 percent of the dioxin in methanol remained after 8 hours. In a small-scale study in Seveso, olive oil was used in either a 40 percent aqueous emulsion or an 80 percent cyclohexanone solution and applied on a heavily contaminated area of grassland. These solutions supplied a hydrogen donor in an effort to facilitate photodegradation of the dioxin present. It was reported that after 9 days 80 to 90 percent of the 2,3,7,8-TCDD was destroyed, whereas concentrations in controls remained virtually unchanged (Wipf et al. 1978; Crosby 1978). In a study of the fate of 2,3,7,8-TCDD in an aquatic environment, samples of lake sediment and water containing 14C-labeled 2,3,7,8-TCDD were incubated in glass vials under light and dark conditions for 39 days (Matsumura and Ward 1976). Results indicated no significant photolytic destruction of the dioxin. Whether artificial or natural light was used is not mentioned. The fate of 2,3,7,8-TCDD in emulsifiable and granular silvex formulations was studied after application to microagroecosystems and outdoor field plots (Nash and Beall 1978). (Experimental conditions of this study are described more completely in the subsection on physical transport.) It was observed that upon volatilization, the dioxin in both the emulsifi- 109 �able and granular formulations was photolyzed not only in direct sunlight but also in shaded areas outdoors and in filtered sunlight passing through the glass of the microagroecosystem chambers. The mean half-life of the dioxin in the emulsifiable concentrate was approximately 7.65 days; the half-life in the granular formulation was 13.5 days. The half-life of the dioxin in the emulsifiable formulation on grass in a microagroecosystem ranged from 5 to 7.5 days. Crosby and Wong reported in 1973 that the major photodecomposition products of 2,4,5-T are 2,4,5-TCP, 2-hydroxy-4,5-dichlorphenoxyacetic acid, 4,6-dichlororesorcinol, 4-chlororesorcinol, and 2,5-dichlorophenol; 2,3,7,8TCDD was not detected as a photolysis product. PHYSICAL TRANSPORT This section describes studies of the movement of dioxins in or into soil, water, and air. Because of episodes involving actual contamination, such movement has become a critical issue. The transport of a chemical in the environment depends greatly upon the properties of the chemical: Is it soluble in water? Is is volatile? Does it cling to soils readily? With the answers to these questions, it is possible to at least postulate reasonably where these chemicals might be found following release into the environment and by what means human or animal receptors are most likely to be affected. Transport in Soil Many studies have addressed the mobility of dioxins, especially 2,3,7,8-TCDD, in soils. Generally it has been found that dioxins are more tightly bound to soils having relatively higher organic content. Dioxins applied to the surface of such soils generally remain in the upper 6 to 12 inches. They migrate more deeply into more sandy soils, to depths of 3 feet or more. In areas of heavy rainfall, not only is vertical migration enhanced but lateral displacement also occurs by soil erosion with runoff and/or flooding. Dioxins may appear in normal water leachate from soils that have received several dioxin applications. Kearney et al. (1973b) studied the mobility of 2,7-DCDD and 2,3,7,8TCDD in five different types of soil. They observed that the mobility of both dioxins decreased with increasing organic content-of the soil. Based on this observation and the finding that these dioxins were relatively immobile in the soils tested, the conclusion was that these dioxins would pose no threat to groundwater supplies because they would not be mobilized deep into soils by rainfall or irrigation. Similar conclusions were reached by Matsumura and Benezet (1973), who showed that mobility of 2,3,7,8-TCDD is relatively slow, much slower than that of DDT. It was concluded that any movement of 2,3,7,8-TCDD in the soil environment would be by horizontal transfer of soil and dust particles or by biological transfer (other than by plants). 110 �During the 8-year period from 1962 to 1970, the U.S. Air Force sprayed 170,000 pounds of 2,4-D, and 161,000 pounds of 2,4,5-T, in two herbicide formulations (Herbicide Orange and Herbicide Purple) over a test area 1 mile square at the Eg!in Air Force Base in Florida (Commoner and Scott 1976). A map of this area is shown in Figure 8. Originally, the applications were done for the purpose of testing spray equipment to be used in Vietnam (Young 1974). The exact concentration of 2,3,7,8-TCDD in the herbicides used for the spraying tests is not known, but is estimated to have ranged from 1 to 47 ppm. The test site has since been analyzed for dioxin residues. In 1970 a 36-inch-deep soil core was taken from a portion of the test area that had received approximately 947 pounds per acre of the 2,4-D, 2,4,5-T Herbicide Orange mixture (Woolson and Ensor 1973). At the limits of detection (0.1 to 0.4 ppb), no 2,3,7,8-TCDD was found at any depth. Several explanations were presented for the absence of dioxin: (1) the 2,4,5-T applied contained less than 2 ppm of 2,3,7,8-TCDD, a concentration undetectable in the soil by the analytical method used; (2) the dioxin had migrated to a depth below 36 inches because of the sandy nature of the soil and the high incidence of rainfall in the area; (3) wind erosion had displaced the dioxin; and (4) biological and/or photochemical decomposition had occurred. In 1973, four soil samples were taken from the same test area and analyzed at low levels for 2,3,7,8-TCDD (Young 1974). The samples contained the dioxin in approximate concentrations of 10, 11, 30, and 710 ppt, and these concentrations were confined to the upper 6 inches of the soil layer. From March 1974 to February 1975 the Air Force performed another study at the Eg!in Air Force Base (Bartleson, Harrison, and Morgan 1975). Two test areas were studied, and also an area where the herbicides had been stored and loaded onto planes. The original 1-mile-square area sampled in 1971 and 1973 contained dioxin in concentrations up to 470 ppt. A second test area, designated Grid 1, contained concentrations of 2,3,7,8-TCDD as high as 1500 ppt. The highest dioxin concentrations were generally found in low-lying areas, and the lowest concentrations usually were in areas of loose sand; these findings indicate that the horizontal translocation had probably occurred through water runoff and wind and water erosion. The storage and loading area contained up to 170,000 ppt of 2,3,7,8TCDD. This area was elevated relative to a nearby pond. Limited sampling of the pond silt revealed a maximum concentration of 85 ppt, and 11 ppt was found in the pond drainage stream. These findings also indicated horizontal translocation of the dioxin, probably as a result of soil erosion. A core sample of soil taken from Grid 1 in 1974 showed the following concentrations of 2,3,7,8-TCDD: Depth, in. Concentration, ppt 0-1 1-2 2-4 4-6 150 160 700 44 111 �LEGEND CONTRAVES INACTIVE ASKANIA © SPOTTING TOWER ® CONTROL BLDG. PAVED ROAD CLAY ROAD SAND ROAD O TOWER INTERANGE BOUNDARY LINE RANGE GATE AND BARBED WIRE FENCE INSTRUMENTED 1 SQUAREMILE TEST GRID 61i»«.«eD«>'I><>- 8 3 Si ** Figure 8. Map of Test Area C-52A, Eglin Air Force Base Reservation, Florida (Source: Young, Thai ken, and Ward 1975). �These data indicate some vertical movement of 2,3,7,8~TCDD, probably as a result of water percolation through the soil. In another test, application of 0.448 kg/m2 of Herbicide Orange to a test site in Utah resulted in the following concentrations of 2,3,7,8-TCDD 282 days after application: Sample depth, in. Control Concentration, ppt 0-6 <10 0-6 15,000 6-12 3,000 12-18 90 18-24 120 In 1978, additional measurements at the Utah test site were reported (Young et al. 1978). Table 22 presents analytical results of plot sampling 4 years after, application of Herbicide Orange at various rates. Table 23 gives results of a similar test performed at Eglin Air Force Base in Florida. In the tests reported in Tables 22 and 23, samples were taken by means of a soil auger. Subsequent tests revealed that dioxin-containing soil was being carried downward as a result of the auger sampling technique and that the concentrations of 2,3,7,8-TCDD below 6 inches were not detectable. Followup studies of the residual levels of 2,3,7,8-TCDD in three loading areas of Eglin Air Force Base were conducted during the period from January 1976 to December 1978 (Harrison, Miller, and Crews 1979). Two of the loading areas were relatively free of contamination. The third (described earlier on p. Ill) had surface soil concentrations of TCDD's as high as 275 ppb. TCDD's were found at 1 meter depths at concentrations one-third the surface amount. The accident at Seveso in July 1976 released quantities of 2,3,7,8-TCDD estimated to range from 300 g to 130 kg over an area of approximately 250 acres (Carreri 1978). Because the Seveso soil is drained by an underlying gravel layer, much concern has arisen over the possibility of groundwater contamination. Early soil migration studies in some of the most contaminated areas at Seveso showed that the dioxin penetrated to a depth of 10 to 12 in. Later studies reported by Bolton (1978) found 2,3,7,8-TCDD at soil depths greater than 30 in. An observed 70 percent decrease in 2,3,7,8-TCDD soil concentration over a period of several months may support the suggestion that the dioxin can be mobilized laterally as well as vertically from soils during heavy rainfall or flooding (Commoner 1977). Following the incident at Verona, Missouri, when oil contaminated with 2,3,7,8-TCDD was sprayed on a horse arena to control dust, the top 12 in. of 113 �TABLE 22. CONCENTRATIONS OF 2,3,7,8-TCDD AT UTAH TEST RANGE 4 YEARS AFTER HERBICIDE ORANGE APPLICATIONS1 (ppt) Rate of Herbicide Orange application, Ib/acre Soil depth, in. 1000 2000 4000 0-6 6-12 12-18 650 11 NAb 1600 90 NA 6600 200 14 ? Source: Young et al. 1978. NA = Not analyzed. TABLE 23. CONCENTRATIONS OF 2,3,7,8-TCDD AT EGLIN AIR 3 FORCE BASE 414 DAYS AFTER HERBICIDE ORANGE APPLICATION Soil depth, in. Herbicide Orange, ppm 0-6 6-12 12-18 18-24 24-30 30-36 1866 263 290 95 160 20 f* Source: Young et al. 1976. Detection limit. 114 2,3,7,8-TCDD concentration in soil, ppt 250 50 <25b <25b <25b <25b �soil was removed and replaced with fresh soil. After removal and replacement of the soil, no further episodes occurred involving sickness or death of human beings or animals. Investigators concluded that this supported the notion that the vertical mobility of TCDD's is limited (Commoner and Scott 1976). Nash and Beall (1978) report studies of the fate of 2,3,7,8-TCDD by use of microagroecosystems and outdoor field plots. A diagram of the microagroecosystem is shown in Figure 9. Two commercially available silvex formulations, one granular and one emulsifiable, were tested. The test and control formulations were applied three times to turf in five microagroecosystems and once to turf on the outdoor plots. Throughout the test period a sprinkler system applied water to the soils to simulate rainfall. The 2,3,7,8-TCDD used in the study was labeled with radioactive hydrogen or 3H. Throughout the study the labeled dioxin (or breakdown product) was tracked by extremely sensitive radiochemical assay methods. The presence of the dioxin molecule in samples was confirmed by gas-liquid chromatography. In the first two applications (on days 0 and 35) the concentration of 2,3,7,8-TCDD in the silvex was 44 ppb. In the third application (on day 77) the silvex formulations contained 7500 ppb (7.5 ppm) 2,3,7,8-TCDD. Soil, water, air, grass, and earthworms were analyzed for 2,3,7,8-TCDD at various times following each of the herbicide applications. Soil analyses showed that most (-80 percent) of the applied 2,3,7,8-TCDD remained in the top 2 cm of the soil. Trace levels at depths of 8 to 15 cm indicated some vertical movement of the dioxin in the soil. Analysis of water leachate samples from the silvex-treated microagroecosystems following the first two herbicide applications showed no detectable 2,3,7,8-TCDD (limits of detection were 10"16 g/g*). The dioxin was detected later, however, following the third herbicide application, and maximum concentrations of 0.05 to 0.06 ppb were found in the leachate samples taken 7 weeks after that third application. In an ongoing study at Rutgers University 54 soil core samples (6 in depth) have been taken from samples of turf and sod from areas in United States having histories of silvex and/or 2,4-D applications. The will analyze the samples for dioxins or herbicide residues. Results are yet available (Hanna and Goldberg, n.d.). in. the EPA not Transport in Water Contamination of streams and lakes by 2,3,7,8-TCDD has also been of concern, especially because of th.e spraying of 2,4,5-T on forests to control underbrush. Possible routes of water contamination from spraying are direct 10 16 g/g may also be expressed as 0.1 fg/g (0.1 femtogram per gram). It is equivalent to 0.0001 ppt. 115 �PLATE GLASS 1 cm) IIULET FILTER / I HOLDER 0 ° REMOVABLE ACCESS PANELS :YLIC PliASTIC(0.7 cm) OUTLET FILTER HOLDER Figure 9, Diagram of microagroecosystem chamber. 116 �application, drift of the spray, and overland transport after heavy rains. The latter, however, seldom occurs on forest lands because the infiltration capacity of forest floors is usually much greater than precipitation rates (Miller, Norris, and Hawkes 1973). The transport of dioxin-contaminated soil into lakes or streams by erosion constitutes another possible route of contamination. This is evidenced by the detection of 2,3,7,8-TCDD in water samples from a Florida pond adjacent to a highly contaminated land area (Bartleson, Harrison, and Morgan 1975). Additionally, several laboratory studies have shown that lakes or rivers could become contaminated with minute quantities (ppt) of 2,3,7,8-TCDD and possibly other dioxins through leaching from contaminated sediments. In a study reported by Isensee and Jones (1975), 2,3,7,8-TCDD was adsorbed to soils, which were then placed in aquariums filled with water and various aquatic organisms. Concentrations of the dioxin in the water ranged from 0.05 to 1330 ppt. These values corresponded to initial concentrations of 2,3,7,8-TCDD in the soil ranging from 0.001 to 7.45 ppm. The investigators concluded that dioxin adsorbed to soil as a result of normal application of 2,4,5-T would lead to significant concentrations of 2,3,7,8-TCDD in water only if the dioxin-laden soil was washed into a small pond or other small body of water. Other investigations have shown similar results. Using radiolabeled 2,3,7,8-TCDD, Matsumura and Ward (1976) showed that, after separation from lake bottom sediment, water contained 0.3 to 9 percent of the original dioxin concentration added to the sediment. Results of another test indicated that a total of about 0.3 percent of the applied dioxin concentration passed through sand with water eluate (Matsumura and Benezet 1973). In some cases, the observed concentration of TCDD's in the water was greater than its water solubility (0.2 ppb). The 1976 report suggests that some of the radioactivity apparent in the aqueous phase was probably due to a combination of lack of dioxin degradation, presence of 2,3,7,8-TCDD metabolites, and binding or adsorption of TCDD's onto organic matter or sediment particles suspended in the water. In another study, application of 14C-TCDD to a silt loam soil at concentrations of 0.1 ppm led to 14C-TCDD concentrations in the water ranging from 2.4 to 4.2 ppt over a period of 32 days (Yockim, Isensee, and Jones 1978). The findings of such investigations are consistent with recent reports that TCDD's are migrating to nearby water bodies from industrial chlorophenol wastes buried or stored in various landfills. At Niagara Falls, New York, for example, 1.5 ppb TCDD's have been detected at an onsite lagoon at the Hyde Park dump where 3300 tons of 2,4,5-TCP wastes are buried (Chemical Week 1979a; Wright State University 1979a, b). Sediment from a creek adjacent to the Hyde Park fill (also in the Niagara Falls area) is also contaminated with ppb levels of the dioxin (Chemical Week 1979a, 1979d). In Jacksonville, Arkansas, there is -growing evidence that TCDD's have migrated from process waste containers in the landfill of a former 2,4,5-T production 117 �site. The dioxins have been found both in a large pool of surface water on the site (at 500 ppb) and downstream of the facility in the local sewage treatment plant, in bayou bottom sediments, and in the flesh of mussels and fish (Richards 1979; Fadiman 1979; Cincinnati Enquirer 1979; Tiernan et al. 1980). TCDD's apparently are also being leached into surface and groundwaters from an 880-acre dump site of the Hooker Chemical Company at Montague, Michigan (Chemical Week 1979c; Chemical Regulation Reporter 1979b). Dioxins were found at the site at levels approaching 800 ppt. Transport in Air One study has been identified in which levels of _2,3,7,8-TCDD in air have been measured (Nash and Beall 1978). Femtogram (10 1S g) quantities of the dioxin were detected in the air after granular and emu!sifiable si 1 vex formulations containing radiolabeled 2,3,7,8-TCDD had been applied to microagroecosystems. Air concentrations of the dioxin decreased appreciably with time following application. The data appear to confirm that TCDD has a very low vapor pressure and that loss due to volatilization is extremely low, especially when low levels of 2,3,7,8-TCDD are involved and granular formulations containing the dioxin are used. Results of other investigations indicate that water-mediated tion of TCDD's may take place (Matsumura and Ward 1976). evapora- Transport of dioxins by way of airborne particulates has recently received much attention. Several studies have shown the presence of dioxins in fly ash from municipal incinerators (Nilsson et al. 1974; Olie, Vermuelen, and Hutzinger 1977; Buser and Rappe 1978; Dow Chemical Co. 1978; Tiernan and Taylor 1980). A recent report of Dow Chemical (1978) contends that particulates from various combustion sources may contain dioxins and that these dioxin-laden particulates are a significant source of dioxins in the environment. More details on these studies are presented in Section 3. It has also been recently reported that dioxins from buried chlorophenol wastes are being mobilized by means of airborne dust particles (Chemical Regulation Reporter 1980a). BIOLOGICAL TRANSPORT This section discusses the potential for dioxins to accumulate and to become concentrated and magnified in biological tissues. In the past, pesticides (most notably DDT) have been found to accumulate in organisms at almost every trophic level. In some organisms these chemicals have been concentrated in the tissues. When an animal in a higher trophic level feeds on organisms that accumulate these chemicals, the animal receives several "doses" of the chemical, resulting in what is termed biomagnification. If this process proceeds to higher levels in the food chain, the chemicals may become concentrated hundreds or thousands of times, with possibly disasterous consequences. The ability for a chemical to accumulate and to become concentrated or participate in biomagnification depends primarily on its availability to 118 �organisms, its affinity for biological tissues, and its resistance to breakdown and degradation in the organism. Bioaccumulation, Bioconcentration, and Biomagm'fication in Animals The biological activity of dioxins with respect to accumulation, concentration, and magnification has been addressed by several researchers. Briefly, bioaccumulation is the uptake and retention of a pollutant by an organism. The pollutant is said to be bioconcentrated when it has accumulated in biological segments of the environment. The increase of pollutant concentrations in the tissues of organisms at successively higher trophic levels is biomagnification. Several investigators (Fanelli et al. 1979, 1980; Frigerio 1978) have studied the levels of TCDD's in animals captured in the dioxin-contaminated area near Seveso, Italy. Data shown in Table 24 indicate that TCDD's accumulate in environmentally exposed wildlife. All field mice were found to contain TCDD's at whole-body concentrations ranging from 0.07 to 49 ppb (mean value 4.5 ppb). The mice were collected from an area where the soil contamination (upper 7 cm) varied from 0.01 to 12 ppb (mean value 3.5 ppb). These data are in agreement with Air Force studies by Young et al. (described below), which indicate that rodents living on dioxin-contaminated land concentrate TCDD's in their bodies only to the same order of magnitude as the soil itself; biomagnification does not occur. Several rabbits and one snake have been found to concentrate TCDD's in the liver. The snake also had accumulated a very high level of TCDD's in the adipose (fat) tissue. Liver samples from domestic birds were analyzed for TCDD's with negative results. TABLE 24. TCDD LEVELS IN WILDLIFE3 Animal Field mouse No. of samples analyzed 14 TCDD level, ng/g (ppb) Range Average Tissue Positive Whole body 14/14 4.5 0.07-49 2.7-13 Hare 5 Liver 3/5 7.7 Toad 1 Whole body 1/1 0.2 Snake 1 Liver, adipose tissue 1/1 2.7 16 Earthworm 2b Whole body 1/2 12 ? Source: Fanelli et al. 1980. Each sample represents a 5-g pool of earthworms. Earlier studies by the Air Force evaluated alternative methods for disposal of an excess of 2.3 million gallons of Herbicide Orange left from 119 �the defoliation program in Southeast Asia. The studies took place at the test site at Eglin Air Force Base in Florida (Figure 8) and at test areas in Utah and Kansas. In June and October of 1973, samples of liver and fat tissue of rats and mice collected from grids on a 3-mile-square test area (TA C-52A) at Eglin Air Force Base were analyzed for the presence of TCDD's (Young 1974). The samples contained concentrations of TCDD's ranging from 210 to 542 ppt. Tissue of control animals contained less than 20 ppt TCDD's. Because most of the concentrations of TCDD's in the group of animals tested were higher than those found in the soil, it was suggested that biomagnification might have occurred; however, because the animals studied failed to show teratogenic or pathologic abnormalities, the presence of a substance similar to TCDD's but with a lower biologic activity was postulated. Another Air Force report gives results of additional studies conducted at Eglin Air Force TA C-52A (Young, Thai ken, and Ward 1975). In an effort to test the possible correlation between levels of TCDD's in the livers of beach mice and in soil, experiments were conducted to determine the possible exposure routes. Because contamination by TCDD's could be detected only in the top 6 inches of soil, it was thought that a food source might be responsible for the presence of the dioxin in animal tissue. Analysis of seeds (a food source for beach mice) collected in the area revealed no TCDD's (at 1 ppt detection level); therefore, another route of contamination was suggested. Since the beach mouse spends as much as 50 percent of its time grooming, investigators postulated that the soil adhering to the fur of the mice as they move to and from their burrows was being ingested. As a test of this hypothesis, a dozen beach mice were dusted 10 times over a 28-day period with alumina gel containing TCDD's. Analysis of pooled samples of liver tissue from controls indicated concentrations of TCDD's of less than 8 ppt (detection limit), whereas concentrations in samples of tissue from the dusted mice reached 125 ppt. Further analysis was done on samples of liver tissue from beach mice collected from Grid 1 of TA C-52A. A composite sample of male and female liver tissue contained TCDD's at levels of 520 ppt, and a composite sample of male tissue contained 1300 ppt. In contrast, the liver tissue of mice collected from control field sites contained TCDD's in concentrations ranging from 20 ppt (male and female composite) to 83 ppt (female composite). Air Force researchers concluded that although bioaccumulation was evident, there were no data to support biomagnification because the levels of TCDD's in the liver tissue of beach mice were in general no greater than levels found in the soil on Grid 1 (ranging from <10 to 1500 ppt). In evaluation of this Air Force study Commoner and Scott (1976) again reached a different conclusion. Because dioxin concentrations in the pooled liver samples represented an average value for the mice, they believed that this value should be compared with the average value for TCDD's in the soil of Grid 1, which was 339 ppt. They concluded that biomagnification was evidenced by the significantly higher levels of TCDD's in mouse liver than in soil. 120 �Analysis for TCDD's in the six-lined racerunner, a lizard found in the area, showed concentrations of 360 ppt in a pooled sample of viscera tissue and 370 ppt in a pooled sample of tissue from the trunks of specimens captured in TA C-52A. Specimens captured at a control site showed concentrations of TCDD's less than 50 ppt (detection limit). Early studies of aquatic specimens obtained from ponds and streams associated with TA C-52A showed no TCDD's at a detection limit of less than 10 ppt (Young 1974). In further studies, however, three fish species showed detectable (ppt) levels of TCDD's (Young, Thai kin, and Ward 1975). Pooled samples of skin, gonads, muscle, and gut from a species of bluegill, Lepomis puntatus, contained 4, 18, 4, and 85 ppt TCDD's, respectively. All of these specimens were obtained from the Grid 1 pond on TA C-52A, where bluegill was at the top of the food chain. Two other fish species, Notropis Lypselopterus (sailfin shiner) and Gambusia affinis (mosquito fish), also showed 12 ppt of TCDD's. These specimens were collected from Trout Creek, a stream draining Grid 1. (Mosquito fish samples consisted of bodies minus heads, tails, and viscera, whereas shiner samples consisted of gut). Inspection of gut contents of Lepomis specimens from Trout Creek showed that the food source of this fish consisted mostly of terrestrial insects. The source of the TCDD's was not identified, however. In another Air Force study, tests were done on 22 biological samples from TA C-52A and 6 samples (all fish) from the pond at the hardstand-7 loading area designated as HS-7 (Bartleson, Harrison, and Morgan 1975). A composite of whole bodies of 20 mosquito fish Gambusia collected from the HS-7 pond and 600 feet downstream showed a concentration of 150 ppt TCDD's. Liver samples from six small sunfish from the HS-7 pond also showed 150 ppt TCDD's, whereas samples of the livers and fat of 12 medium-sized sunfish from the HS-7 pond showed concentrations of 0.74 ppb. Because the solubility of 2,3,7,8-TCDD in water is far below these levels (0.2 ppb), the data seem to indicate biomagnification in addition to bioaccumulation. The stream that drains the HS-7 pond flows north into a larger pond known as Beaver Pond. Composite samples of four whole large fish from Beaver Pond showed a concentration of 14 ppt TCDD's. The livers of 25 large fish and fillets of 8 large fish from Beaver Pond showed no TCDD's at a detection limit of 5 ppt. A followup study conducted from 1976 to 1978 showed that TCDD's were present in turtle fat and beach mouse liver and skin (Harrison, Miller and Crews 1979). In the same study, samples obtained from deer, meadowlark, dove, opossum, rabbit, grasshopper, six-lined racerunner, sparrow, and miscellaneous insects from TA C-52A were analyzed for TCDD's. TCDD's were detected in the livers and stomach contents of all of the birds. One composite sample of meadowlark livers contained 1020 ppt TCDD's, the highest level found in all samples. No TCDD's were detected in samples from deer, opossum, or grasshopper. The sample from miscellaneous insects contained 40 ppt TCDD's, and the composite sample from racerunners, 430 ppt TCDD. The authors concluded that this study demonstrated bioaccumulation. The data also indicate that biomagnification may have occurred. Commoner and Scott (1976b) point out that the average concentration of TCDD's in soil from TA C-52A was 46 ppt. 121 �It should also be noted that the composite insect sample most likely included insects that are eaten by the birds. In all cases the concentration of TCDD's in animal liver samples was greater than that in the insect sample, an indication of the possibility of biomagnification. Because none of the Air Force studies analyzed for TCDD's in a series of trophic levels, biomagnification was not clearly demonstrated. Woolson and Ensor (1972) analyzed tissues from 19 bald eagles collected in various regions of the country in an effort to determine whether dioxins were present at the top of a food chain. At a detection limit of 50 ppb, no dioxins were found. Another study failed to show dioxin contamination in tissues of Maine fish and birds (Zitco, Hutzinger, and Choi 1972). In a similar study 45 herring gull eggs and pooled samples of sea lion blubber and liver were analyzed for dioxins and various other substances (Bowes et al. 1973). Analysis by gas chromatography with electron capture and high-resolution mass spectrophotometry revealed no dioxins. Fish and crustaceans collected in 1970 from South Vietnam were analyzed for TCDD's in an effort to determine whether the spraying of Herbicide Orange had led to accumulation of TCDD's in the environment (Baughman and Meselson 1973). Samples of carp, catfish, river prawn, croaker, and prawn were collected from interior rivers and along the seacoast of South Vietnam and were immediately frozen in solid C02. Butterfish collected at Cape Cod, Massachusetts, were analyzed as controls. Samples of fish from the Dong Nai river (catfish and carp) showed the highest levels of TCDD's, ranging from 320 to 1020 ppt. Samples of catfish and river prawn from the Saigon River showed levels ranging from 34 to 89 ppt. Samples of croaker and prawn collected along the seacoast showed levels of 14 and 110 ppm of TCDD's, whereas in samples of butterfish from Cape Cod the mean concentration of TCDD's was under 3 ppt (detection limit). The authors concluded that TCDD's had possibly accumulated to significant environmental levels in some food chains in South Vietnam. Other investigators have studied the accumulation of TCDD's in mountain beavers after normal application of a butyl ester of 2,4-D and 2,4,5-T to brushfields in western Oregon (Newton and Snyder 1978). They reported that the home range of the mountain beavers was small and that among all animals collected inside the treatment areas the home ranges centered at least 300 feet from the edge of the treatment area. Thus their food supplies, consisting primarily of sword fern, vine maple, and salmonberry, had definitely been exposed to the herbicide. Analysis of 11 livers from the beavers showed no TCDD's in 10 of the samples at detection limits of 3 to 17 ppt. One sample was questionable; the concentration was calculated at 3 ppt TCDD's. Investigators in another study analyzed milk from cows that grazed on pasture and drank from ponds that had received applications of 2,4,5-T (Getzendance, Mahle, and Higgins 1977). Sample collection ranged from 5 days to 48 months after application; 14 samples were collected within 1 year 122 �after application. Application rates ranged from 1 to 3 pounds per acre. Milk purchased from a supermarket was used as the control. The control samples contained levels of TCDD's ranging from nondetectable to 1 ppt. No milk samples from cows grazing on treated pasture contained levels of TCDD's above 1 ppt. In a similar study, milk samples were collected throughout the Seveso area just after the ICMESA accident occurred (Panel!i et al. 1980). The samples were analyzed for TCDD's by GC-MS methods. Results are given in Table 25. Figure 10 shows the sites where the milk samples were collected. Dioxin levels were highest in samples from farms close to the ICMESA plant. The high levels of TCDD's found in the milk samples strongly suggest that human exposure via oral intake must have occurred after the accident through consumption of dairy products. A milk monitoring program that has been sampling milk from outside Zone R since 1978 no longer detects TCDD's in any of the samples. Three research teams have analyzed fat from cattle that had grazed on land where 2,4,5-T herbicides were applied. In one study, five of eight samples collected from the Texas A&M University Range Science Department in Mertzon, Texas, showed the possible presence of TCDD's at low ppt levels when analyzed by gas chromatography/low-resolution mass spectrometry (Kocher et al. 1978). Apparent TCDD concentrations ranged from 4 to 15 ppt at detection limits ranging from 3 to 6 ppt. In the second study, 11 of 14 samples analyzed contained TCDD's (Meselson, O'Keefe, and Baughman 1978). The four highest levels reported were 12, 20, 24, and 70 ppt TCDD. In the third study, Solch et al. (1978, 1980) detected TCDD's in 13 of 102 samples of beef fat at levels ranging from 10 to 54ppt. Shadoff and coworkers could find no evidence that TCDD's are bioconcentrated in the fat of cattle (Shadoff et al. 1977). The animals were fed ronnel insecticide contaminated with trace amounts of TCDD's for 147 days. Sample cleanup was extensive to permit low-level detection of the dioxin. Analysis was by combined gas chromatography/mass spectrometry (both high and low resolution). No TCDD's were detected at a lower detection limit of 5 to 10 ppt. Samples of human milk obtained from women living in areas where 2,4,5-T is used have also been analyzed for dioxins. In one study, four of eight samples were reported to contain about 1 ppt TCDD's (Meselson, O'Keefe, and Baughman 1978). In a subsequent study, no evidence of 2,3,7,8-TCDD contamination was found in 103 samples of human milk collected in western states (Chemical Regulation Reporter 1980). The lower level of detection in the latter study ranged from 1 to '4 ppt. Model ecosystems have been developed in aquariums to study the bioaccumulation and concentration of several pesticides including TCDD's (Matsumura and Benezet 1973). Concentration factors for TCDD's calculated from these studies were: Daphnia: 2198 Mosquito larvae: 2846 Ostracoda: Northernbrook silverside fish: 54 107 123 �TABLE 25. TCDD LEVELS IN MILK SAMPLES COLLECTED NEAB SEVESO a m JULY-AUGUST 1976 Map . number Date of collection TCDD concentration, ng/ liter (ppt) 1 7/28 76 2 7/28 8/2 8/10 7919 5128 2483 3 7/28 8/2 8/10 469 1593 496 4 8/10 1000 5 7/29 116 6 7/29 59 7 8/3 8 8/3 94 9 7/27 8/3 180 75 10 8/5 <40 i Source: Panel!i et al 1980. Locations shown in Figure 10 124 80 �ICMESA IN Figure 10. Location of farms near Seyeso at which cow's milk samples were collected for TCDD analysis in 1976 (July-August). (Source: Panelli et al. 1980) 125 �The authors concluded that the biological and physical characteristics of organisms played an important role in the bioaccumulation and concentration of TCDD's and the other pesticides studied. They also indicated that because of the low solubility of TCDD's in water and liquids and their low partition coefficient in liquids, TCDD's are not likely to accumulate in biological systems as readily as DDT. Another aquatic study involved a recirculating static model ecosystem in which fish were separated from all the other organisms (algae, snails, daphnia) by a screened partition (Yockim, Isensee, and Jones 1978). In this study 14C-TCDD was added to 400 g of Metapeake silt loam clay to yield TCDD's at a concentration of 0.1 ppm. Treated soils were placed in the large chambers of the ecosystem tanks and flooded with 16 liters of water. One day after the water addition, all organisms except the catfish were added. Samples of organisms and water were collected on days 1, 3, 7, 15, and 32. On day 15 a second group of 15 mosquito fish was added. On day 32 all organisms remaining were collected and analyzed. Also on day 32, nine channel catfish were added to the large chambers of the tanks containing the soil. Catfish were collected 1, 3, 7, and 15 days later. Of the two collected on each day, one was sacrificed for analysis and one was placed in untreated water. Bioaccumulation ratios (tissue concentration of TCDD's divided by water concentration) for the algae ranged from 6 to 2083, the maximum exhibited after 7 days. Bioaccumulations ratios for the snails ranged from 735 to 3731, with the maximum at 15 days. The ratios in daphnia ranged from 1762 to 7125, with the maximum at 7 days. The accumulation ratios in the mosquito fish ranged from 676 at day 1 to 4875 after 7 days. All mosquito fish were dead after 15 days, and their tissues showed an average of 72 ppb TCDD's. No bioaccumulation ratios were calculated for the catfish, but levels of TCDD's in the tissues ranged from 0.9 ppt after day 1 to 5.9 ppt after day 15. By day 32 of exposure all catfish had died. The average concentration of TCDD's in the tissue at this time was 4.4 ppb. It was concluded that under normal use of 2,4,5-T, concentration of TCDD's in sediments of natural water bodies would probably be 104 to 106 times lower than the concentration used in this experiment, and although the TCDD's could be a potential environmental hazard, the magnitude of the hazard would depend on biological availability and persistence in the aquatic ecosystem under conditions of normal use. In previously mentioned studies with microagroecosystems, earthworms contained 0.2 and 0.3 ppt 2,3,7,8-TCDD and/or breakdown products of TCDD's following two silvex applications to soil (Nash and Beall 1978). The silvex contained 44 ppb TCDD's. Another study not yet completed concerns the possible accumulation of dioxins in vegetation and earthworms in turf and sod of areas having a history of silvex and/or 2,4-D applications (Hanna and Goldberg, n.d.). 126 �Isensee and Jones (1975) performed three experiments using algae, duckweed, snails, mosquito fish, daphnia, channel catfish and other organisms. Radio!abeled dioxin (14C-TCDD) was adsorbed to two types of soil, which were then placed in glass aquariums and covered with water. One day later daphnia, algae, snails, and various diatoms, protozoa, and rotifers were added. In one experiment duckweed plants were also added on the second day. After 30 days, some daphnia were analyzed and two mosquito fish were added to each tank. Three days later, all organisms were harvested; in Experiments II and III, two fingerling channel catfish were added to each tank and exposed for 6 days. At the conclusion of each experiment the concentrations of 14C-TCDD in the water and in the organisms were determined and the concentration factors calculated. Table 26 summarizes soil application rates in each experiment and type of soil used. At soil concentrations as low as 0.1 ppb, 14C-TCDD was leached into the water and accumulated in the organisms. Bioaccumulation factors at this soil concentration and a water concentration of 0.05 ppt were 2,000 for algae, 4,000 for duckweed, 24,000 for snails, 48,000 for daphnia, 24,000 for mosquito fish, and 2,000 for catfish, corresponding to concentrations of 0.1, 0.2, 1.2, 2.4, and 0.1 ppb of 14C-TCDD in the tissues. Although some biomagnification was evident, results were highly variable. The differences in bioaccumulation factors found in this study relative to those of Yockim et al. (1978) were attributed to system design, differences in the organisms, and the fact that bioaccumulation factors in the other study were based on fresh weight whereas those in this study were based on dry weight. The authors conclude that since some bioaccumulation ratios were relatively high (as compared with those observed with other pesticides), especially in daphnia and mosquito fish, the potential of TCDD's to accumulate in the environment is considerable. They further project, however, that at suggested application rates of 2,4,5-T, concentrations of TCDD's in the soil would probably not result in accumulation in biological systems unless erosion or runoff from recently sprayed areas is discharged to a small body of water (e.g., a pond). Dow Chemical Company, a producer of pentachlorophenol and the major producer of 2,4,5-trichlorophenol, reported in 1978 on a series of studies to determine whether dioxins are present in the Tittabawassee River, into which Dow discharges treated wastes. In one study, rainbow trout were placed in cages at various locations above and below the Dow Midland plant, in a tertiary effluent stream and in clear well water. Five of six fish placed in the tertiary effluent stream showed levels of TCDD's ranging from 0.2 to 0.05 ppb. Analysis of whole fish exposed for 30 days at a point 6 miles downstream of the effluent'discharge showed concentrations of 0.01 and 0.02 ppb TCDD's. Analysis of whole fish from the tertiary effluent showed levels ranging from 0.05 to 0.07 ppb. In a laboratory experiment with 14C-2,3,7,8-TCDD, Dow (1978) determined that the bioconcentration factor in rainbow trout was about 6600. Dow also analyzed native catfish taken randomly from various locations in the Tittabawassee River and tributaries. The analyses showed levels of TCDD's 127 �TABLE 26. SOIL APPLICATION RATES AND REPLICATIONS1 Total 14C-TCDD Final concentrations added per tank, Type 14 soil and amount of of 14C-TCDD No. of of C-TCDD added, g in soil , ppm replicates M9 149 0 Experiment I L-20 L-20 7.45 0 3 1 63 63 63 63 0 Experiment II L-20 L-20 + M-100 L-20 + M-200 L-20 + M-400 L-20 3.17 0.53 0.29 0.15 0 2 2 2 2 2 0.1 0.01 0.001 0.0001 0 2 2 2 2 2 Experiment III 10 1 0.1 0.01 0 M-100 M-100 M-100 M-100 M-100 Isensee and Jones 1975. L = Lakeland sandy loam, M = Metapeake silt loam. In Experiment II, L was first treated with 14C-TCDD, then dry-mixed with M in treatment tanks. c Soil concentrations based on total quantity of soil in tanks. 128 �ranging from 0.07 to 0.23 ppb, levels of OCDD from 0.04 to 0.15 ppb, and one sample with 0.09 ppb of hexa-CDD. Highest levels of TCDD's and OCDD were found in fish collected from the Tittabawassee at points approximately 1 to 2 miles downstream from Dow. Dow noted that caustic digestion used in sample preparation may have degraded octa-, and hexachlorodioxins. No other fish analyzed contained detectable levels of TCDD's (Dow Chemical Company 1978). Subsequent to the Dow studies, the U.S. EPA collected and analyzed fish samples from the Tittabawassee, Grand, and Saginaw Rivers in Michigan (Harless 1980). TCDD's were found in 26 of 35 samples (74 percent) at levels ranging from 4 to 690 ppt. Catfish and carp contained the highest concentrations, while perch and bass had the lowest. Accumulation in Plants Because dioxins are sometimes used in herbicides applied on and near areas where food plants may be growing, it is important to determine whether the dioxins may be incorporated into the plants. Thus far few studies have been done to determine whether dioxins might accumulate in plants. In the few studies that have considered this question, results seem to indicate that very small amounts, if any, are accumulated in plants. Kearney et al. (1973a) studied the uptake of DCDD's and TCDD's from soil by soybeans and oats. Soil applications of 14C-DCDD (0.10 ppm) and 14 C-TCDD (0.06 ppm) were made, and a maximum of 0.15 percent of the dioxins was detected in the above-ground portion of the oats and soybeans. No dioxins were found in the grains harvested at maturity. Application of a solution of Tween 80 (a surfactant) and TCDD's or DCDD's to the leaves of young oat and soybean plants showed no translocation to other plant parts after 21 days. Studies of the absorption and transportation of TCDD's by plants in the contaminated area near Seveso have been reported (Cocucci et al. 1979). Samples of fruits, new leaves, and in some cases twigs and cork were taken from various types of fruit trees a year after the dioxin contamination occurred. TCDD's were found in all samples at |jg/kg levels. Concentrations in the leaves were 3 to 5 times higher than in the fruits, which had the lowest concentrations. Levels in the cork samples were generally higher than in the leaves, but not as high as in the twigs. The findings show that the dioxin is translocated from the soil by plants in newly formed organs and suggest that the lower concentrations in fruits and leaves may be due to some form of elimination such as transpiration or ultraviolet photodegradation. The latter possibility would agree with the photolysis results reported by Crosby and Wong in 1977. Cocucci and coworkers also examined specimens of garden plants such as the carrot, potato, onion, and narcissus. Again, ug/kg levels of TCDD's were found. In all plants, the new aerial portions appeared to contain less dioxin than the underground portions. Concentrations of TCDD's differed in the inner and outer portions of potato tubers and carrot taproots; the .variation was attributed to the prevalence of conductive tissues in these 129 �plant parts. The authors again suggested that the relatively low concentrations in the aerial parts of these garden plants was due to an elimination process such as transpiration or photodegradation, or possibly to metabolism of the dioxin by the plants. The elimination hypothesis was supported by the further observation that when contaminated plants were transplanted in unpolluted soil, the dioxin content disappeared. Young et al. (1976) used specially designed growth boxes to study the uptake of 14C-TCDD by Sorghum vulgave plants. After placing Herbicide Orange containing 14 ppm 14C-TCDD under the soil in the growth boxes, 100 plants were grown for 64 days. After 64 days the plants were harvested, extracted with hexane, and analyzed for 14C-TCDD. Some plant samples were also analyzed for 14C-TCDD before hexane extraction by combustion and collection of the C02. Analysis before extraction showed a concentration of about 430 ppt 14C-TCDD in the plant tissue. After hexane extraction, the concentration of 14C-TCDD in the plant tissue was reported as being 14 not significantly reduced. Young et al. concluded that the relatively high C activity in the plant tissue could have been due to the presence of (1) nonhexane-soluble TCDD, (2) a soil biodegradation product of TCDD's that was taken up, (3) a metabolic breakdown product of TCDD's found after plant uptake of the TCDD's, or (4) a contaminant in the original 14C-TCDD stock solution that was taken up by the plant. As mentioned elsewhere, concentration of 14C-TCDD in algae and duckweed has been observed. Bioaccumulation factors were 2000 and 4000, respectively (Isensee and Jones 1975). 130 �SECTION 6 DISPOSAL AND DECONTAMINATION GENERAL CONSIDERATIONS One of the principal unsolved problems that has followed the discovery of dioxins is development of methods for destroying them once they are produced. Many investigators have studied various methods for disposing of commercial chemicals and production wastes that contain these compounds, and further research is needed. Even more important is the need for methods of destroying dioxins after they are released into the environment. Simple out-of-sight storage has been used on several occasions to dispose of dioxin-contaminated soils and equipment following industrial accidents from the manufacture of 2,4,5-TCP. Soil contaminated by the application of dioxin-containing wastes at Verona, Missouri, was used as fill under a new concrete highway and was also placed in a sanitary landfill. Some was also used as fill at two residential sites, but was later removed and placed elsewhere (Commoner 1976a). The soil contaminated by the accident at Seveso, Italy, was partially removed from moderately contaminated areas and added to the more heavily contaminated areas, which will remain uninhabitable for an indefinite period of time (Reggiana 1977). Following an explosion at Coalite and Chemical Products, Ltd., in England, portions of the plant equipment were buried in an abandoned coal mine (May 1973). Portions of the Phillips Duphar plant in the Netherlands, following its explosion, were encased in concrete and dumped into the ocean (Hay 1976a). The quantities of TCDD-containing wastes from the normal manufacture of 2,4,5-TCP that have been buried at various sites in the United States are not well documented, although some published figures are available. One company at Verona, Missouri, reportedly disposed of 16,000 gallons of 2,4,5-TCP distillation residues over an 8-month period (Shea and Lindler 1975). A New York company reportedly disposed of 3700 tons of 2,4,5-TCP production wastes at three dumps in the Niagara Falls area over a 45-year period (Chemical Week 1979a). It is estimated that the 3700 tons of waste produced by this company could contain 100 pounds of TCDD (Chemical Week 1979a). An Arkansas facility has been producing 2,4,5-TCP and related products since 1957 and possibly earlier (Sidwell 1976a). Reports indicate that 3000 to 3500 barrels of TCP wastes are buried or stored on the manufacturing site (Fadiman 1979; Cincinnati Enquirer 1979). Many of these barrels are now leaking and contaminating nearby water bodies (Richards 1979a; Tiernan et al. 1980). 131 �Continuation of land disposal is still being proposed as at least a temporary measure, however. Other proposals include chemical fixation, deep well disposal, burial in salt mines, and inclusion of these chemicals with nuclear fission byproducts in secured cavities. Although these practices postpone the need for solving the problems of disposal and decontamination, they offer no permanent solutions. Techniques that may be used to decompose dioxins and thereby remove them permanently from the environment are discussed in this section. The most extensively tested method is incineration, which entails a high-temperature oxidation of the dioxin molecules. Physical methods have also been proposed for some applications; these include the use of solvents or adsorbents to concentrate dioxins into smaller volumes for final disposal by incineration or other methods, and also physical methods of detoxification including exposure to ultraviolet light or gamma radiation. Proposed chemical techniques include the use of ozone or special chloroiodide compounds. Biological degradation techniques are also being considered. INCINERATION DISPOSAL METHODS Conventional Incineration Conventional incineration has reached a high level of development for disposal of pesticides and other highly toxic, hazardous materials (Wilkenson, Kelso, and Hopkins 1978; Ferguson et al. 1975; Ottinger 1973; Scurlock et al. 1975; U.S. EPA 1977a; U.S. EPA 1975a; Duvall and Rubey 1976). It is often preferred over other disposal alternatives (Lawless, Ferguson, and Meiners 1975; Kennedy, Stojanovic, and Shuman 1969), and has been used extensively (Ackerman et al. 1978). Incineration as defined here does not include open, uncontrolled burning, but denotes the use of special furnaces equipped with means for accurate regulation of furnace temperature, supplemental fuel usage, and excess air ratios. Industrial incinerators are also equipped with some form of emission control, often a water scrubber. Incinerator off-gas usually contains only low concentrations of carbon particulates, but does contain chlorine and hydrogen chloride if chlorinated organic chemicals are being burned. Incinerator operating conditions currently considered adequate for complete destruction of 2,3,7,8-TCDD and most other chlorinated organics are a temperature of at least 1000°C (1932°F) with a dwell time of at least 2 seconds (Tenzer et al.; Wilkenson et al. 1978). Laboratory tests have demonstrated that with a dwell time of 21 seconds, only half of the 2,3,7,8-TCDD in a sample decomposes at 700°C, whereas 99.5 percent decomposes at 800°C (Ton That et al. 1973). These data were obtained with a quartz tube apparatus. Using differential thermal analysis two other experimenters have observed that complete destruction occurs between 800° and 1000°C (Kearney et al. 1973b), which agrees with the work of Langer et al. 132 �(1973). All of these studies have been conducted with relatively pure samples of dioxins. For incineration of impure mixtures, temperatures above 800°C are especially important because at lower temperatures (300° to 500°C) more TCOD may be formed from precursor material (Rappe 1978). Incineration is now used to dispose of wastes from pesticide manufacture at the Midland, Michigan, facility of Dow Chemical Company. Stationary and rotary kiln incinerators used at this location can handle almost any solid, semisolid, or liquid waste. Dow has emphasized in a 1978 report to the EPA that complete destruction of dioxins is difficult, in that reducing the concentration of a substance from 1 ppm to the equivalent of 1 ppb necessitates an overall efficiency of 99.9 percent, which is not possible with conventional high-capacity incinerators. The most extensive incineration of a waste chemical containing dioxins was the destruction of 10,400 metric tons (more 'than 2 million gallons) of Herbicide Orange left over from military defoliation operations in southeast Asia (Ackerman et al. 1978). This substance was decomposed in two large incinerators mounted on the Vulcanus, a chemical tanker ship operated by a company from the Netherlands. Burning took place in the mid-Pacific ocean. In three separate trips, the herbicide was emptied from steel storage drums to railroad tank cars to the cargo holds of the tanker (the drums were rinsed with diesel fuel, which was added to the herbicide). The ship was then' moved to the burn location and the mixture was incinerated at an average flame temperature of 1500°C with an incinerator residence time of 1 second. Flow of combustion air was regulated to maintain a minimum of 3 percent oxygen in the stack gases. Combustion efficiency was about 99.9 percent. Stack effluents were sampled and analyzed routinely, with a minimum detection limit of 0.047 ng/ml (ppb). Only one set of samples contained measurable amounts of 2,3,7,8-TCDD (Tiernan et al. 1979). No analyses were performed for any other chemical constituents or decomposition products. This operation also resulted in more than 40,000 steel drums that were still slightly contaminated with Herbicide Orange. These drums were to have been crushed mechanically, then shipped to a steel mill to be melted as steel scrap at a temperature of about 2900°C (Whiteside 1977). No available reports confirm the completion of this procedure. Portions of the ship 2used in the incineration operation were also contaminated with 86 ng/m of Herbicide Orange. Subsequent decontamination reduced the concentration by as much as 96 percent (Erk, Taylor and Tiernan 1979). The decontamination procedure and the fate of the residue are not known (Chemical Week 1978d). A high-temperature liquid and solid incinerator is being constructed as a mobile unit under an EPA contract (Brugger 1978). Its purpose is to decompose hazardous chemicals such as dioxins, and it is expected to be used to incinerate the dioxin-contaminated sludge now being stored in Verona, Missouri. It may also be used to burn some dioxin-contaminated activated carbon remaining from initial efforts by the U.S. Air Force to remove dioxins from Herbicide Orange by adsorption. This mobile incineration unit is to be equipped with an afterburner and a scrubber for the exhaust gases. 133 �It will be able to handle the combustion equivalent of 75 gallons per hour of fuel oils and a solids equivalent of 3.5 tons per hour of dry sand. In another project a private partnership plans to convert a tanker for ocean incineration of toxic wastes including 2,4,5-TCP wastes. The ship will be equipped with three 25-ton/h incinerators capable of burning a 10,000-ton load of waste on a week's cruise. EPA will monitor the test burns during initial operations (Chemical Week 1979g). Incineration has been suggested for decontamination of the soil and other materials at Seveso, Italy (Commoner 1977; Pocchiari 1978), but local political pressure has killed the idea (Revzin 1979; Chemical Week 1979h). A giant incinerator was to have been built that would have held each furnace charge at 800 to 1000°C for 30 to 40 minutes. Estimates of the amounts of soil to be processed range from 150,000 to 300,000 megagrams. In addition there are huge quantities of contaminated furniture and decaying plants and small animals (about 87,000 in number), which are presently quarantined, awaiting final disposal. Authorities have refused to allow the incinerator to be built because the burning of such massive amounts of dioxin-contaminated debris would take years. Furthermore, the residents and authorities fear that the presence of such an incinerator would result in Seveso becoming the industrial waste dumping ground for all of Italy. Advanced Incineration Techniques Two advanced incineration techniques have been studied for the decomposition of toxic substances. Molten salt combustion consists of burning a contaminated chemical with air below the surface of a liquified inorganic material. Microwave plasma destruction, although not a true combustion process, converts a mixture of contaminated chemical and oxygen into elemental oxides through the action of microwave radiation. Molten Salt Combustion-The technology of molten salt combustion has been developed over the past 20 years by Atomics International Division of Rockwell International Corporation (Wilkinson, Kelso, and Hopkins 1978). It has potential application to the destruction of pesticides and hazardous wastes. A schematic of the process is given in Figure 11. A difficulty with developing this system for full-scale practice may be in locating suitable materials of construction. The molten salt is sodium or potassium carbonate containing 10 percent by weight of sodium sulfate. It is maintained at 800° to 1000°C by application of heating or cooling as needed. When the molten salt is applied to chlorinated hydrocarbon wastes, the carbon and hydrogen in the waste are oxidized to C02 and steam, while the chlorine content is changed into sodium chloride. Tests have demonstrated that this bench-scale combustor can achieve virtually complete decomposition (more than 99 percent) of chlorinated hydrocarbons, 2,4-D, chlordane, chloroform, and trichloroethane. The 2,4-D tested was part of an actual waste that contained 30 to 50 percent 2,4-D and 50 to 70 percent bis-ester and dichlorophenol tars. The waste was diluted with ethanol and burned at 830°C. This combustion test destroyed 99.98 percent of the organic materials. 134 �i STACK OFF-GAS CLEANUP , H20, WASTE 1 SALT RECYCLE WASTE TREATMENT AND FEED CO on AIR' WASTE AND AIR .. 1 MOLTEN SALT FURNACE SPENT MELT REPROCESSING OPTION SPENT MELT DISPOSAL T ASH Figure 11. Schematic of molten salt combustion process. (Source: Wilkinson, Kelso, and Hopkins 1978, as adapted from Atomics International 1975.) �Microwave Plasma Destruction-Microwave plasma refers to a partially ionized gas produced by microwave-induced electron reactions with neutral gas molecules (Bailen and Hertzler 1976; Bailen 1978). The ionized gas or plasma is derived from the carrier gas which transports the molecules into the plasma zone (Oberacker and Lees 1977). When oxygen is used as the reactant gas in the plasma, highly reactive atomic oxygen is produced which then rapidly oxidizes organic compounds introduced into the system discharge (Bailen 1978). A laboratory-scale microwave plasma reactor with capacity of 1 to 5 g/h, and a pilot-scale reactor with capacity of 430 to 3,200 g/h have been tested by the Lockheed Palo Alto Research Laboratory under a contract from EPA (Bailen and Hertzler 1976). A schematic diagram of these units is shown in Figure 12. Tests have been conducted with a variety of toxic materials, including two commercial PCB's, Aroclor 1242, and Aroclor 1254. The laboratory-scale reactor converted 99.9 percent of the PCB's into carbon monoxide, carbon dioxide, water, phosgene, and chlorine oxides. The pilot-scale reactor converted at least 99 percent of most materials tested into smaller molecules. One test, however, did not achieve complete destruction and left a black, tarry substance that still contained PCB's. The pilot reactor was also used in tests with supported formulation of kepone charged to the reactor material, a 10 percent slurry in water, and a 20 percent Conversion of at least 99 percent of each charge matrial hydrogen chlorine was achieved in all tests. a commercial clayas compressed solid slurry in methanol. to basic oxides and Microwave plasma decomposition has also been used to detoxify U.S. Navy red dye (Bailen 1978). Specific application of this technique to dioxins is not reported, although it has been considered for detoxification of dioxincontaminated wastes stored in Missouri (Bailen 1977). PHYSICAL METHODS Concentration One approach to disposal or decontamination of toxic substances is by use of techniques that selectively remove toxic constituents from mixtures. Such techniques would reduce the volume of material that must be treated and would offer potential for salvage of useful materials. To date, however, such techniques have presented serious problems because they have been used to concentrate dioxins even with no available means or facilities for disposal of the concentrate. In at least two instances, quantities of activated carbon heavily contaminated with dioxins are being stored because disposal methods are not available. In this country, extensive pilot-plant studies of carbon adsorption were conducted before the Air Force decided to incinerate Herbicide Orange (Whiteside 1977; Young et al. 1978). Although the reprocessing 136 �PESTICIDE DROPPING FUNNEL TUNING UNIT 1 PLASMA REACTOR TUBE MICROWAVE POWER SOURCE °2 SUPPLY FLOW METERS ALTERNATE GAS SUPPLY 3-WAY STOPCOCK MICROWAVE 7^ APPLICATOR MANOMETER TUNING UNIT MASS SPECTROMETER COLD TRAP THROTTLE VALVE MICROWAVE POWER SOURCE RECEIVER^ "—' COLD TRAP Figure 12. Schematic of microwave plasma system (Source: Wilkinson, Kelso, and Hopkins 1978, as adapted from Bail en and Hertzler 1976). 137 �method was technically and environmentally feasible, it was not possible to demonstrate an acceptable method for safely disposing of the dioxin-laden carbon. The contaminated carbon is now stored on an island in the Pacific. Similarly, Union Carbide of Australia created quantities of dioxin-contaminated carbon in efforts to detoxify 2,4,5-TCP after they became aware of the 2,3,7,8-TCDD problem in 1969 (Chemical Week 1978b; Dickson 1978). This carbon is still stored in steel drums in that country. Although data are unavailable, activated carbon apparently can adsorb dioxins selectively from chemical mixtures, but the carbon cannot be regenerated. Even after long periods of contact, solvent extraction will not desorb a major portion of the adsorbate. One study evaluated the desorption of phenol from activated carbon with 10 different solvents (Modell, deFilippi, and Krukonis 1978). After 2 hours of continuous extraction, the most effective solvent desorbed only 28 percent of the phenol. A newly proposed technology for regeneration of activated carbon is the use of supercritical fluids (fluids in the region of their critical temperatures and pressures), and in particular supercritical carbon dioxide (Model!, deFilippi, and Krukonis 1978). With one type of activated carbon (Filtrasorb 300, Calgon Corp.), 100 percent desorption was obtained within 3 hours. After the first regeneration, however, adsorption capacity of the carbon is only 50 to 85 percent. It is believed that the initial treatment causes formation of carboxyl, hydroxyl, and carbonyl groups on the surface of the carbon and that their chemical interaction with the carbon may lead to irreversible adsorption. In general, carbon adsorption techniques have not been proven effective for toxics disposal, even if the carbon is to be destroyed by incineration or other methods. After being contaminated with heavy organic chemicals, activated carbon must usually be dried and pulverized prior to incineration to ensure complete destruction. These additional handling steps provide the possibility of fugitive losses. Bailen and Littauer (1978) are presently investigating the possibility of using microwaves to regenerate spent activated carbon. It is not known whether activated carbon containing dioxins will be evaluated in the study. Solvent extractions of soil have been shown to be effective in analytical determinations of TCDD's (Teirnan et al. 1980). It has been suggested that solvents such as hexane could be used to extract dioxins from soil by use of equipment similar to that used to extract oil from olive seeds (Commoner 1977). It is not known whether this concentration process has been tested. The use of steam distillation has also been suggested as a means of concentrating dioxins, but no details are available. Photolysis The use of light to degrade halogenated aromatic compounds is well established in published literature (Mitchell 1961; Plimmer 1972, 1978a; Rosen 1971; Watkins 1974; Wilkinson, Kelso, and Hopkins 1978). Regarding degradation of dioxins, most studies have been concerned with the effect of sunlight on dioxins released into the environment, as outlined in Section 5. 138 �Application of the same principle to detoxify dioxins with artificial light could lead to a means of decontaminating chemical mixtures. The Velsicol Chemical Corporation has proposed such a photolytic system as an alternative method for disposal of Herbicide Orange (Crosby 1978a, 1978b; Lira 1978). The herbicide mixture would first be hydrolyzed with caustic and converted into butyl alcohol, water, and salts of 2,4-D and 2,4,5-T. Additional butyl alcohol would then be used to extract the dioxins. The butyl alcohol and dioxins would be separated from the phenolic salts and water by decantation, and the organic layer would be irradiated with ultraviolet light. Irradiation would be accomplished in a special reaction apparatus, in which thin films of the liquid are exposed to light from quartz tubes. Although preliminary tests did succeed in destroying. 2,3,7,8-TCDD, the process had not been pursued because the toxicity of the resulting decomposition products was unknown and the butyl alcohol would have to be disposed of by incineration or other methods. Further tests of this principle were discontinued. No other studies of large-scale decomposition of dioxins by use of artificial light have been reported. Laboratory studies have shown, however, that light does not destroy the structure of dioxins. Under appropriate conditions, light converts the more toxic dioxins to less toxic forms by removing halogen substituents (Crosby 1971). Any applications of this principle will therefore be limited to decontamination and partial degradation, rather than to complete disposal. Radiolysis Radiolysis, an extension of the photolytic method, has been studied experimentally. Gamma rays having properties similar to light have been shown to partially degrade dioxins. As with ultraviolet light, these rays may not totally destroy the dioxin structure, but only remove substituent halogens. In the most recent series of tests, investigators dissolved 2,3,7,8-TCDD in either ethanol, acetone, or dioxane6 at a concentration of 100 ng/ml (ppm) and irradiated the solutions at 10 rads/h (Chemical Week 1977; Panel!i et al. 1978). They found that 97 percent of the dioxin was degraded after 30 hours, when ethanol was the solvent. Degradation was somewhat slower in the other solvents. All irradiated samples showed the presence of tri-CDD and DCCD. In 1976, Buser dissolved OCDD in benzene and hexane at a concentration of 25 g/liter and exposed it to gamma radiation. After 4 hours, 80 percent of the OCDD was converted into dioxins with five, six, or seven chlorine substituents. Further degradation did not occur. Other researchers completed an extended series of tests using gamma radiation of the ionizing type to destroy pesticides (Craft, Kimbrough, and Brown 1975). Significant destruction of single representative compounds such as pentachlorophenol, 2,4,5-T, and 2,4-D was obtained, but no change 139 �in PCB's or mixtures of compounds such as Herbicide Orange could be detected. This test series led to the conclusion that because of the inefficiency of radiation in destroying mixtures of pesticides and dioxins, costs would be prohibitive for routine use of this method in waste treatment. CHEMICAL METHODS Several chemical techniques have been proposed for the destruction of toxic dioxins. Vertac, Inc., reportedly developed a process for safely destroying its dioxin-containing wastes, but no details are available (Environment Reporter 1979b). Of the five methods outlined in the following paragraphs, only the first two have been tested specifically with dioxins. Ozone Treatment (Ozonolysis) The use of ozone is common in chemical waste treatment applications, especially in decomposition of cyanides. It has been used most often in laboratory applications for decomposition of large organic molecules (Wilkinson, Kelso, and Hopkins 1978). In a recent test, ozone was bubbled through a suspension of 2,3,7,8-TCDD in water and carbon tetrachloride. It was reported that after 50 hours, 97 percent of the 2,3,7,8-TCDD had degraded. In this process, the dioxin apparently is suspended as an aerosol combined with carbon tetrachloride, which facilitates ozone attack (Cavolloni and Zecca 1977). Another modification of ozone treatment has been developed by Houston Research, Inc. (Wilkinson, Kelso, and Hopkins 1978; Mauk, Prengle, and Payne 1976). Tests with dioxins, however, have not been reported. In this technique, treatment with ozone is combined with ultraviolet irradiation. The light activates organic molecules to a highly energetic state, thereby rendering them more susceptible to ozone attack. When this technique was applied to pentachlorophenol and DDT, these compounds were decomposed into carbon dioxide, water, and hydrochloric acid. A schematic diagram of the apparatus is shown in Figure 13. Two bench-scale reactors of 10- and 21-liter capacity have been constructed (Mauk, Prengle, and Payne 1976). Although these examples indicate that ozone treatment may be effective for use in dioxin disposal or decontamination, the use of ozone must be combined with some other mechanism that will activate the dioxin and promote the attack of ozone. Chloroiodide Degradation In a recently described method, 2,3,7,8-TCDD in contaminated soil is degraded by use of a class of compounds derived from quaternary ammonium salt surfactants and referred to as chloriodides (Botre, Memoli, and Alhaique 1979). The compounds are formulated in mi cellar solutions with surfactants that increase the water solubility of the substances. The two 140 �MIXER EXHAUST GAS UV LIGHT TEMPERATURE | CONTROL L SPARGED BATCH REACTOR J pH MONITORING ' AND SAMPLING IMPELLER ti > OZONE GENERATOR . 2% Kl VENT SOLUTION POWER OXYGEN OR AIR Figure 13. Schematic for ozonation/ultraviolet irradiation apparatus (Source: Wilkinson, Kelso, and Hopkins 1978, as adapted from Mauk, Prengle, and Payne 1976), 141 �derivatives showing the most degradation potential are alkyldimethyl benzylammonium (benzalkonium) chloroiodide and 1-hexadecylpyridinium (cetylpyridinium). When 2,3,7,8-TCDD in benzene was vacuum evaporated and the residue treated with a cationic surfactant aqueous solution containing benzalkonium chloroiodide, 71 percent of the 2,3,7,8-TCDD decomposed. When cetylpyridinium chloroiodide in cetylpyridinium chloride was used, 92 percent of the 2,3,7,8-TCDD was decomposed. These experiments were performed in absence of light to prevent photolytic degradation. In a test with soil from Seveso contaminated with 2,3,7,8-TCDD, only about 14 percent was degraded within 24 hours following treatment with benzalkonium chloride. When benzalkonium chloroiodide was added, an additional 38 percent of the 2,3,7,8-TCDD was degraded. Total degradation during this test was 52 percent. Wet Air Oxidation Wet air oxidation is an accelerated oxidation process performed at high pressure and temperature. Oxidation takes place in an autoclave in which a charge of water and organic material is heated to 150° to 350°C while being pressurized with air to 40 to 140 atmospheres. Three commercial processes of this type are known as the Zimpro, Wetox, and Lockheed processes. They are used for rapid decomposition of sewage sludge, munitions waste, and sulfite liquor from pulp and paper mills. It has been proposed to evaluate the Wetox system for disposal of priority pollutants and other hazardous chemicals (Wertzman n.d.). This might also be an alternative method for disposal of dioxin and dioxin contaminated materials, but no tests have yet been reported. Chlorinolysis and Chlorolysis Although chlorinolysis and chlorolysis were developed primarily to produce chlorinated products from nonchlorinated or less chlorinated organics, some attention has been focused on their use in waste treatment (Shiver 1976). Chlorinolysis is used primarily to convert hydrocarbons containing one to three carbon atoms into perch!oroethylene, trichloroethylene, and carbon tetrachloride (Diamond Alkali Company 1950; U.S. Patent Office 1972). As most often practiced, the process continuously reacts chlorine with ethylene or ethylene dichloride in a fluid bed catalyst reactor. The process usually creates small amounts of hexachlorobenzene, hexachloroethane, hexachlorobutadiene, tetrachloroethane, and pentachloroethane as side-reaction products. Chlorolysis, an associated process, is sometimes used to convert the side-reaction products from chlorinolysis into carbon tetrachloride; it can also be used with benzene or its derivatives or with mixtures of chlorinated aromatic or aliphatic compounds. Chlorolysis is a two-stage process in which gaseous feed materials are reacted with chlorine at pressures of 200 to 700 atmospheres and temperatures up to 800°C. No catalyst is used. 142 �In cooperation with the U.S. Department of Agriculture, the Diamond Shamrock Company conducted pilot-plant studies to test the stability of 2,3,7,8-TCDD under the severe reaction conditions of chlorolysis (Kearney et al. 1973). Although the results of these studies are not known, the techniques may be applicable to disposal of certain dioxin-contaminated chemicals and might yield marketable products from otherwise waste chemicals. Catalytic Dechlorination Catalytic dechlorination is a simple chemical process in which the action of a catalyst reductively dechlorinates an organic compound. The usual catalyst is nickel borohydride, which is prepared in a reaction vessel by mixing sodium borohydride and nickel chloride in a solvent of alcohol. When this solution is mixed with a chlorinated organic chemical, the chlorine atoms are removed from the molecules and hydrogen atoms are substituted (Cooper and Dennis 1978; Dennis 1972; Dennis and Cooper 1975, 1976, 1977; Wilkinson, Kelso, and Hopkins 1978). Laboratory tests have been conducted with this process to detoxify several commercial pesticides, including DDT's, heptachlor, chlordane, and lindane. Tests with chlorinated dioxins have not been reported. The process does not completely dechlorinate most organic chemicals and would not break down the basic dioxin structure. The reaction occurs rapidly, however, and at room temperature; for these reasons, the process may be of value in decontamination operations or in detoxifying small volumes of toxic dioxins. Other processes have been used to dechlorinate aromatic compounds, including conventional catalytic hydrogenation with metallic catalysts and hydrogen gas (Dennis and Cooper 1975). In a small-scale laboratory experiment with a catalyst of palladium on charcoal, about 60 percent of a charge of 1,6-DCDD was reduced to unsubstituted dioxin in 1 hour at room temperature and less than 1 atmosphere pressure. BIOLOGICAL TREATMENT One of the least expensive techniques for breaking down large organic molecules, and often one of the most effective, is to subject the molecules to the action of microorganisms. Although toxic chemicals are usually degraded slowly in uncontrolled exposure to the environment, more complete and more rapid breakdown can be achieved by controlling the microorganism species and providing specialized environments. Numerous studies have examined the susceptibility of dioxins, particularly 2,3,7,8-TCDD, to microbial decomposition. Most of the studies have concerned decomposition in the uncontrolled environment, as described in Section 4. Much less attention has been directed to the controlled use of microorganisms. The following paragraphs describe available data on two aspects of the microbial decomposition of dioxins: soil conditioning and biochemical wastewater treatment. A specialized treatment system for toxic wastes is also discussed. 143 �Soil Conditioning The large area of dioxin-contaminated soil surrounding Seveso, Italy, has stimulated studies of degradation of dioxins by soil microorganisms. Available data indicate that 2,3,7,8-TCDD is resistant to this method of decontamination, although under optimum conditions some slow degradation occurs. Rates of uncontrolled degradation have been variously measured in two studies. The U.S. Air Force reported the half-life of 2,3,7,8-TCDD at 225 and 275 days (Young et al. 1976). In a separate analysis of the same test data, Commoner (1976b) obtained a half-life of 190 to 330 days. In Seveso, however, Bolton (1978) reported finding no reduction in dioxin levels in the most heavily contaminated zone, and in the less contaminated zone reduction after 400 days was only 25 percent. Researchers in Zurich, Switzerland, have found that soil-bound 2,3,7,8-TCDD becomes increasingly difficult to recover quantitatively with time (Huetter 1980). This observation may explain the decreasing recoveries of 2,3,7,8-TCDD in soil degradation studies by the U.S. Air Force and others in which the "disappearance" of 2,3,7,8-TCDD with time was interpreted as evidence of biodegradation. Half-lives for 2,3,7,8-TCDD calculated from these studies may not accurately reflect the true persistance of this dioxin in the soil environment. One proposal for modifying the Seveso soil environment is to use charcoal or activated carbon to hold the dioxins in the soil, then to spread manure on the treated soil to increase the rate of bacterial growth (Young 1976). U.S. Air Force studies have shown, however, that although treatment of this sort increases the number and activity of soil microorganisms, the rate of dioxin degradation is reduced. Apparently, adsorption on charcoal causes the dioxin to be less available to the bacteria. No other proposals to modify the open soil environment have been advanced. Attempts have been made to inoculate Seveso soil with selected bacteria that might facilitate the breakdown of dioxins. Although initial results appeared promising, subsequent data indicated that the method had not been effective (Commoner 1977). The inoculated species either died out or became mutated to a strain that rejected the dioxins. In a similar laboratory study of 100 microbial strains that had shown ability to degrade pesticides, only 5 showed any ability to degrade 2,3,7,8-TCDD (Matsumura and Benezet 1973). Wastewater Treatment Systems Very little is known concerning the ability of biological or biological/chemical wastewater treatment to remove dioxins. Dow Chemical Company operates a tertiary treatment system to treat wastewater from its Midland, Michigan, pesticide manufacturing plant (Dow Chemical Co. 1978). A 2-year program of analysis of grab and composite samples taken from the tertiary effluent stream revealed only one with a 144 �detectable amount (0.008 ppb) of TCDD's. In further investigations, six caged fish were placed in the tertiary pond effluent; subsequent analyses showed, in five of the six fish, concentrations of TCDD's ranging from 0.02 to 0.05 ppb in the edible portions and from 0.05 to 0.07 ppb in the whole bodies. These findings, when compared with data on control fish containing no detectable levels of TCDD's, clearly indicate the presence of TCDD's in the tertiary pond effluent. Data obtained in 1976 from Transvaal, Inc., showed no TCDD's in effluent from the city stabilization ponds, to which Transvaal sends all or part of its plant wastewater effluent (Sidwell 1976b). A sample from the Transvaal plant effluent, however, showed 0.2 to 0.6 ppb of this dioxin. Other than pH adjustment with lime, the effluent apparently undergoes no pretreatment. As previously discussed (p. 83) more recent studies of this site have been reported (Tiernan et al. 1980). In a third study sludge was sampled at the outlet of a lagoon holding effluent from a pentachlorophenol manufacturing plant. The sludge was analyzed for TCDD's, but none was found (U.S. Environmental Protection Agency 1978d). Since this dioxin has never been found as a decomposition product of pentachlorophenol, the negative analysis would be expected. The sludge was not analyzed for hexa-CDD's, hepta-CDD's, or OCDD, the dioxins normally associated with PCP manufacture. Researchers in Finland have patented a process for purifying wastewaters containing chlorinated aromatics in a biofilter (Salkinoja-Salonen 1979a). The filter consists of a layer of wood bark that contains a strain of bacteria able to degrade the organic compounds (Salkinoja-Salonen 1979 a,b). These bacterial strains were isolated by taking samples of bacteriferous water, mud, or bark residue from water bodies polluted by chlorinated and unchlorinated phenols and aromatic carboxylic acids, then feeding pollutants to the bacterial populations collected. Work is under way to prove the effectiveness of the filter in treating dioxins; its efficacy in treating aromatics such as tri- and tetrachlorophenols has been demonstrated. Micropit Disposal A detailed study of biological degradation of pesticides is being conducted by Iowa State University (Rogers and Allen 1978). The apparatus used in the study, shown in Figure 14, consists of a partially buried polyethylene garbage can filled with layers of rock and soil, and flooded with water. The study, sponsored by the U.S. EPA, deals with a variety of pesticides at various concentrations, and with the effects of nutrient additives and aeration. Two organochloride compounds are included among the pesticides being examined, but it is not clear whether the test includes dioxins. Test data are not available. 145 �GALVANIZED BASKET GROUND LEVEL 29.2 cm (11.5 in.) 20.3 cm (8 in.) 35.6 cm (14 in.) GALVANIZED METAL SLEEVE 212 LITER POLYETHLENE BARREL PERFORATED CLAY TILE 55.2 cm (1.8 ft) Figure 14. Internal view of pesticide micropit (Source: Rogers and Allen 1978). 146 �SECTION 7 HEALTH EFFECTS INTRODUCTION On a molecular basis 2,3,7,8-TCDD is perhaps the most poisonous synthetic chemical. As shown in Table 27, only bacterial exotoxins are more potent poisons. Not only is this TCCD isomer extremely poisonous but it also has extremely high potential for producing adverse effects under conditions of chronic exposure. Human exposure to 2,3,7,8-TCDD has induced chloracne (an often disfiguring and persistent dermatologic disorder), polyneuropathy (multiple lesions of peripheral nerves), nystagmus (involuntary rapid movement of the eyeball), and liver dysfunction as manifested by hepatomegay (increase in liver size) and enzyme elevations (Pocchiari, Silano, and Zampieri 1979). In animals, this compound has been shown to be teratogenic, embryotoxic, carcinogenic, and cocarcinogenic (Neubert and Dill man 1972; Courtney 1976; Kociba et al. 1978; and Kouri et al. 1978). It has been established that under certain conditions 2,3,7,8-TCCD can enter the human body from a 2,4,5-T-treated food chain and can accumulate in the fatty tissues and secretions, including milk (Galston 1979). The available data indicate significant risks associated with the use of dioxincontaminated herbicides. Based upon the work of Van Miller et al., estimates done by accepted risk assessment procedures indicate that daily human exposure to 0.01 ug (10 ng) of 2,3,7,8-TCDD is the dosage expected to result in "incipient carcinogenicity." Additionally, daily human exposure to 4 ug 2,3,7,8-TCDD would be expected to result in a shortened lifespan, and daily exposure to 290 ug would likely result in acute toxicity (Galston 1979). Although 2,3,7,8-TCDD is considered to be the most toxic dioxin, others are also cause for concern. Kende and Wade (1973) have established certain chemical structural requirements that must be met for a dioxin to be toxic: Halogen substituents at positions 2,3, and 7 are minimum structural requirements. Bromine as a substituent is more active toxicologically than chlorine, which is more active than fluorine. At least one hydrogen atom must remain on the dibenzo-para-dioxin nucleus. 147 �TABLE 27. TOXICITIES OF SELECTED POISONS' Substance Molecular weight Minimum lethal dose, moles/kg Botulinum toxin A 9 x TO5 3.3 x 10"17 Tetanus toxin 1 x 10s 1 x 10"15 7.2 x 104 4.2 x 10"12 2,3,7,8-TCDDb 322 3.1 x TO"9 S ax i toxin 372 2.4 x 10"8 Tetrodotoxin 319 2.5 x 10"8 Bufotoxin0 757 5.2 x 10"7 Curare 696 7.2 x 10"7 Strychnine 334 1.5 x 10"6 Muscarin0 210 5.2 x 10"6 Di isopropyl f 1 uorophosphate 184 1.6 x 10"5 49 2.0 x 10"4 Diphtheria toxin Sodium cyanide Source: Poland and Kende 1976. These data were compiled by Mosher et al., and the values indicate only relative toxicity. It should be noted that the values deal with different species, routes of administration, survival times, and in one case the mean lethal dose rather than the minimum lethal dose. Except where noted, administration was by the intraperitoneal route u in mice. LD50* upon oral administration in the guinea pig. Intravenous injection in the cat. _ LD50 - the dosage lethal to half of a group of test animals. 148 �Another finding is that the ability for a dioxin to induce* various enzymes correlates with its toxicity, as illustrated in Tables 28 and 29. As these tables show, 2,3,7,8-TBDD and Hexa-CDD are the only dibenzo-paradioxin derivatives nearly comparable to 2,3,7,8-TCDD in acute toxicity or ability to produce chloracne. These two compounds are also comparable to 2,3,7,8-TCDD in induction of aryl hydrocarbon hydroxylase (AHH). The compounds OCDD and 2,7-DCDD are mildly toxic, with minimal ability to induce AHH. Thus bioassays of unknown dioxin isomers based upon enzyme induction hold promise for predicting biological activity and toxicity. METABOLISM In guinea pigs, 2,3,7,8-TCDD is moderately well absorbed from the gastrointestinal tract and has a plasma half-life of about 1 month (Nolan et al. 1979). Although dibenzo-para-dioxin is rapidly converted by the microsome-NADPH system into polar metabolites, this system has little effect upon 2,3,7,8-TCDD (Vinopal and Casida 1973). A large proportion of administered 2,3,7,8-TCDD persists in unmetabolized form in the liver, partially concentrated in the microsomal fraction in all species studied. This finding implies that the unmetabolized compound, rather than a metabolite, is responsible for its toxic effects in mammals. A recent study has shown that 2,3,7,8-TCDD is slowly excreted via the biliary tract in the form bf glucuronide and other more polar metabolites (Ramsey 1979). The same study indicated that enterohepatic recirculation of the compound was not extensive. Studies have indicated that its toxicity is not mediated by: Inhibition of mitosis (cell division) in mammalian cells Alteration of glucocorticoid metabolism Alteration of thyroid hormone function Increasing serum levels of ammonia Inhibition of the synthesis of flavin enzymes or The effect of superoxide anion via DT-diaphorase stimulation (Beatty 1977). Another aspect of 2,3,7,8-TCDD metabolism is its interaction with iron metabolism. Rats given 1.7 [jg of the substance intragastrically have shown a 2-fold increase in the serosal transfer of iron, whereas no effect was observed on the mucosal iron uptake (Mam's 1977). Sweeny (1979) has shown, however, that iron deficiency protects mice from many of the toxic effects of 2,3,7,8-TCDD. In the latter study, animals rendered iron-deficient were protected from elevated porphyrin levels (including the consequent skin * An induced enzyme is one that is synthesized only in response to the presence of a certain substrate or substrates. 149 �TABLE 28. BIOLOGICAL PROPERTIES OF DIOXINSC LD50 (rat), mg/kg Compounds Chloracne aptitude Hexa-CDD (mixture) OCDD +++ 0 + +++ 0.04 Unsubstituted dioxin >1000 2,3,7,8-TCDD 2,7-DCDD 2,3-DCDD 2,3,7-tri-CDD 2,3,7-tri-BDD 1,2,3,4-TCDD 1,3,6,8-TCDD 2,3,7,8-TBDD Teratogenic Embryo toxic effect effect 0 0 0 -2000 >1000 >1000 >1000 >1000 >100 < 1 -100 -2000 +++ 0 0 0 0 0 ++ + 0 0 +++ + 0 0 + 0 ++ + Source: Saint-Ruf 1978. TABLE 29. ENZYME INDUCTION' Compounds ALASb (chick embryo) AHHC (chick embryo) Zoxazolamine hydroxylase (rat) +++ 1 +++ 0 0 0 0 0 2,3,7,8-TCDD Unsubstituted dioxin 2,3-DCDD 2,7-DCDD 2,8-DCDD 1,3-DCDD 2,3,7-tri-CDD 2,3,7-tri-BDD 1,2,3,4-TCDD 1,3,6,8-TCDD 2,3,7,8-TBDD 0 0 0 0.02 0.6 0 0.2 1.0 0.8 0 0 ++ ++ 0 + Hexa-CDD OCDD Source: Saint-Ruf 1978. Amino-levulinic Acid Synthetase. Aryl Hydrocarbon Hydroxylase. 150 +++ 0 �disease that resembles human porphyria cutanea tarda) and liver damage. Since mixed function oxidase enzymes were elevated in the iron-deficient mice, the authors speculated that depleted stores of iron in tissue were responsible for the observed amelioration of toxicity. The results of these studies have significant implications for toxicity in humans. Persons with high dietary iron intake would be expected to be more susceptible to 2,3,7,8-TCDD toxicity than persons with marginal iron intakes. Similarly, females might be less susceptible to its toxicity than males because they usually store less iron in the body. Finally, phlebotomy may prove clinically useful in treating 2,3,7,8-TCDD intoxication. Pharmacokinetics and Tissue Distribution Two studies have extensively examined the pharmacokinetics of 2,3,7,8TCDD (Piper, Rose, and Gehring 1973; Rose et al. 1976). Rose demonstrated that elimination of this dioxin followed first-order kinetics, and he fit the 14 data to the one-compartment open model. Table 30 shows the body burden of C-2,3,7,8-TCDD in rats given a 14 single oral dose of 1.0 ug/kg; the average fractional oral absorption of C-2,3,7,8-TCDD was approximately 84 percent, and the elimination half-life averaged 31 days. Piper's earlier study also found that after the first 2 days following oral dosages of rats, elimination followed first-order kinetics. The results of this study, however, which are summarized in Figure 15, show that only about 70 percent of ingested 2,3,7,8-TCDD was absorbed and the elimination half-life was only about 17 days. Over a 21-day period, a total of 53 percent of the ingested dose was excreted in the feces, while about 13 percent and 3 percent were excreted in the urine and expired air, respectively. Tissue distribution of ingested 2,3,7,8-TCDD has been examined in many species, including rats, guinea pigs, and monkeys (Piper, Rose and Gehring 1973; Rose et al. 1976; Gasiewicz and Neal 1978; Van Miller, Marlar, and Allen 1976). Rose et al. established that the accumulation of 14C-2,3,7,8TCDD in rat liver follows apparent first-order kinetics. In this study, the accumulation of 2,3,7,8-TCDD in rat liver could be simulated by the following equation: C t = Css (1-e"kt) where C = the concentration of 14C activity in the liver at time t Cs = the concentration of 14C activity in the liver at steady state l = elimination rate constant from the liver Values of C 5 equal to 0.25 (jg equivalent 2,3,7,8-TCDD per gram of liver per ug dose, and k equal to 0.026 days' were obtained by fitting experimental data. In this study, the concentration of the dioxin in rat liver was 5 times greater than that in fat, while concentrations in kidney, thymus, and spleen were l/12th to l/50th of those in the liver. Rose et al. (1976) also assumed that first-order elimination kinetics applied to accumulation of 2,3,7,8-TCDD in rat fat, and they calculated values of C and k equal _£o 0.058 ijg equivalent TCDD per gram of fat per ug dose and 0.029 day , respectively. 151 �TABLE 30. 14C BODY BURDEN ACTIVITY IN SIX RATS GIVEN A SINGLE ORAL DOSE OF 1.0 ug OF 14C-2,3,7,8-TCDD/kga Sex k, days f _i tjs> days Male 0.66 0.026 ±0.001b 27 Male 0.77 0.018 ±0.001 39 Male 0.91 0.021 ±0.000 33 Female 0.93 0.022 ±0.001 32 Female 0.87 0.019 ±0.001 36 Female 0.91 0.033 ±0.002 21 0.84 ±0.11 0.023 ±0.006 31 ±6 Mean ± SD Source: Rose et al. 1976. Rose gives the following equation: Body burden = f (dose)e-kt where f is the fraction of the dose absorbed; k, the elimination rate constant; t, , the body burden half-life. Confidence limits 95%. 152 �PERCENTAGE OF'DOSE EXERETED ro o (O 03 O —' c/i o -s o CD -a —i, -a -s -s cn a. m O X </> o ro -s fD O r+ cn o (a cn CO o v> fD O) 3 Q. O3 O OO O CD CO IQ O) IQ ro c+ CO w C/> CO » -h •-J O 00 —' I O —is o ->• O 3 OIQ I ro �In a study of male guinea pigs, Gasiewicz and Neal (1978) found the highest levels of radioactivity (percent of original dose per gram of tissue) on day 1 after injection in the adipose tissue (2.36 percent), adrenals (1.36 percent), liver (1.13 percent), spleen (0.70 percent), intestine (0.92 percent), and skin (0.48 percent). On day 15 of this study, the level of 14C-2,3,7,8-TCDD in the liver had increased to 3.23 percent/g; increases were also noted in the adrenals, kidneys, and lungs, and general decreases were seen only in adipose tissues and skin. Van Miller et al. (1975) found that 40 percent of the radioactivity of an administred dose of labeled 2,3,7,8-TCDD was concentrated in rat liver, whereas less than 10 percent was concentrated in monkey livers. In this study, high concentrations of the radioactivity were found in the skin, muscle, and fat of monkeys. Thus, there appear to be significant differences in the tissue distribution of 2,3,7,8-TCDD among various animal species. One study examined the tissue distribution and excretion of labeled OCDD in the rat (Norback 1975). A radioactive analog of OCDD at a daily dosage of about 12.4 mg/kg was administered for 21 days. Over 90 percent of the OCDD administered was recovered in the feces as unabsorbed material. The major route of elimination of absorbed OCDD in the rat was the urinary system, and the rate corresponded to a biological half-life of about 3 weeks. After 21 days of administration, approximately 50 percent of the body burden of OCDD was found in the liver; over 95 percent of the radioactivity in the liver was associated with the microsomes and was equally distributed within the rough and smooth fractions. The radioactivity in adipose tissue was about 25 percent of that in the liver. Significant levels of radioactivity were also found in the kidneys, breast, testes, skeletal muscle, skin, and serum. Enzyme Effects Several investigations show that 2,3,7,8-TCDD has a dramatic influence upon various enzyme systems in many species including man. The most notable were the mixed-function oxygenases. For example, 2,3,7,8-TCDD is approximately 30,000 times more potent than 3-methylcholanthrene in inducing activity of the enzyme aryl hydrocarbon hydroxylase (AHH) in rat liver (Poland and Glover 1974). This dioxin is also a potent inducer of 6-amino-levulinic acid synthetase in the liver of chick embryo (Poland 1973). These properties of 2,3,7,8-TCDD have a considerable influence upon its toxicity. For instance, its ability to act as a cocarcinogen or to produce porphyria cutanea tarda depends upon alteration of enzymatic systems. Before the effects on enzymatic systems are catalogued, an examination of the mechanism of its effects on the cytochrome P-450-mediated monooxygenase enzyme system may prove informative. This enzyme system handles much of the influx of "foreign" chemicals and appears to rival the immune system in complexity (Fox 1979). A well-characterized subset of the P-450-mediated enzymes is a group of cytochromes whose induction is regulated by one of a small number of genes. Fox (1979) has termed this genetic system the Ah complex (for aromatic hydrocarbon responsiveness). Work with 2,3,7,8-TCDD has demonstrated that 154 �the Ah locus must involve a minimum of three gene products at each of two nonlinked loci, plus a structural gene for cytochrome Pi-450 (P-448) as well. Other investigators have demonstrated that cytosolic binding sites for 2,3,7,8-TCDD enhance AHH activity by de novo* protein synthesis of apocytochrome P-448, and that these binding sites are not necessarily associated with AHH inducibility regulated by the Ah locus (Guenthner and Nebert 1977; Kitchin and Woods 1978). It has been postulated that the ratelimiting factor in AHH induction is protein synthesis of apocytochrome P-448 (Kitchin and Woods 1978). Fox (1979) suggests that 2,3,7,8-TCDD may act in a manner similar to steroid hormones. He postulates that the dioxin may ride its receptor into a cell's nucleus, where it turns on specific Ah genes. Activation of these genes would then lead to the requisite protein synthesis for AHH induction. Figure 16 summarizes the mechanism of AHH induction proposed for 2,3,7,8-TCDD and possibly the mechanism by which this substance produces other toxic effects. As the figure shows, 2,3,7,8-TCDD moves into a cell and binds to a specific cytosolic receptor. The receptor-dioxin complex then moves into a cell's nucleus, where it "turns on" the synthesis of specific messenger RNAs, which direct the synthesis of cytochrome Pi-450. Other 2,3,7,8-TCDD molecules can then react with newly formed cytochrome Pj-450, possibly to produce reactive intermediates. These metabolites may be excreted as innocuous products, may afflict specific critical target cells in other organs, or may act as carcinogens or cocarcinogens. Several studies show that 2,3,7,8-TCDD induces many enzyme systems and suppresses others. Studies with rats indicate that females are more susceptible than males to enzyme alteration by the dioxin (Lucier et al. 1973). Further, 2,3,7,8-TCDD induces the following enzymes in addition to AHH, 6-amino-levulinic acid synthetase, and the cytochrome P-450-containing monooxygenases, mentioned earlier: UDP glucuronyl transferase (Lucier 1975); Aldehyde dehydrogenase (Roper 1976); Glutathione transferase B (Kirsch 1975); DT-diaphorase (Beatty and Neal 1976); Benzopyrene hydroxylase (Lucier 1979); Glutathione S-transferase (Manis 1979); Ethoxycoumarin deethylase (Parkki and Aitio 1978). Marselos et al. (1978) found that 2,3,7,8-TCDD decreases activity of the following enzymes: primary or of recent onset 155 �NUCLEUS LIVING CELL UNKNOWN SITE IN NUCLEUS RECEPTOR IN CELL INDUCER-RECEPTOR COMPLEX MOVES INTO NUCLEUS -450 GOES INTO MEMBRANE REACTS WITH POLLUTANTS 2,3,7,8-TCDD MESSAGE NOW "RECEIVED-RESPONSE IS SYNTHESIS OF SPECIFIC MESSENGER RNA'S CJl INNOCUOUS PRODUCTS EXCRETED mRNA'S DIRECT SYNTHESIS OF SPECIFIC PROTEINS (CYTOCHROME Pi-450) REACTIVE INTERMEDIATE ] 4 CRITICAL TARGET IN OTHER CELLS UNKNOWN CRITICAL TARGET m _j REACTIVE INTERMEDIATE BINDS CRITICAL TARGET DRUG TOXICITY OR INITIATION OF CANCER Figure 16. Proposed mechanism for induction of AHH and toxicity by 2,3,7,8-TCDD (adapted from Fox 1979). �UDP-glucuronic acid pyrophosphatase; D-glucuronolactone dehydrogenase; L-gluconate dehydrogenase The following enzymes have shown no effects upon exposure to 2,3,7,8" TCDD: NADPH cytochrome (Lucier et al. 1973); B-glucuronidase (Lucier et al. 1973); UDP-glucose dehydrogenase (Marselos et al. 1978); Epoxide hydrase (Parkki and Aitio 1978); Glycine N-acetyl transferase (Parkki and Aitio 1978). As these lists indicate, the effects of 2,3,7,8-TCDD on more than a dozen enzyme systems have been studied extensively. Effects on Lipids 2,3,7,8-TCDD has dramatically altered the lipid profiles in laboratory animals and man. One study examined the effects of both sublethal and lethal doses upon the lipid metabolism of the Fischer rat (Albro 1978). A sublethal dose of 2,3,7,8-TCDD caused a temporary increase in triglyceride and free fatty acid levels, with a persistent decrease in levels of sterol esters. Lethal doses resulted in fatty livers and large increases in serum cholesterol esters and free fatty acids, with little change in triglyceride levels. These changes appeared to be due in part to damage sustained by lysosomes. A decrease in acid lipase activity observed in the study also supports the hypothesis that the 2,3,7,8-TCDD-induced myeloid bodies (see Figure 17) were derived from damaged lysosomes and probably accounted for the increased levels of cholesterol esters in animal livers. A mechanism by which 2,3,7,8-TCDD may exert its toxic effects is suggested by the observed rapid, dose-dependent increase in lipofuscin pigments.* Lipid peroxidation, which precedes the formation of polymeric lipofuscins, is known to seriously damage membranous subcellular organelles, including lysosomes. Studies of workers occupationally exposed to 2,3,7,8-TCDD have shown lipid abnormalities (Walker and Martin 1979; Poland et al. 1971). In Poland's study, 7 of 71 persons (10 percent) occupationally exposed to the dioxin in a plant manufacturing 2,4-D and 2,4,5-T showed elevated serum cholesterol levels (greater than 294 mg/100 ml). Walker's more recent study of eight dioxin-exposed workers with chloracne showed significant abnormalities in lipid metabolism and liver function. In this study, the levels of triglycerides and y-glutamyl transpeptidase (GGT)f were elevated in five men and were * Bronze colored (wear and tear) pigments. f Liver enzyme. 157 �LIPID DROPLET ,rs •*''' •**.. Figure 17. Schematic of rat liver 13 days after administration of 2,3,7,8-TCDD (50 yg/kg). Note concentric membrane array surrounding lipid droplet. X20502. (Source: Redrawn from Albro 1978) 158 �normal in the other three. In all of the dioxin-exposed workers with chloracne, however, the levels of high-density 1'ipoprotein (HDL) cholesterol were below the method mean, total cholesterol levels were above the method mean, and ratios of total to HDL cholesterol were consistent with a higherthan-average risk of ischemic (oxygen insufficiency) vascular disease. Two of the men in the study had experienced previous myocardial infarction (heart attack), and one had experienced possible transient ischemic attacks (TIA's) (reversible cerebrovascular insufficiency). In any event, the lipid abnormalities resulting from 2,3,7,8-TCDD exposure may be a significant risk factor for ischemic vascular disease. GROSS AND HISTOPATHOLOGIES The gross (macroscopic) and histopathologies (microscopic) of dioxinexposed chickens, rats, and monkeys have been examined extensively (Gupta et al. 1973; Norback and Allen 1973; Allen 1967; Allen et al. 1975; Greig and Osborne 1978). The chicken develops extreme morbidity and mortality at dietary concentrations of 2,3,7,8-TCDD that are only mildly toxic to rats, whereas response in the monkey is intermediate (Norback and Allen 1973). At post-mortem examination, the most striking finding in dioxin-exposed animals is usually substantial loss of body fat. Two types of lesions have been reported in all species studied: (1) involution of the thymus; and (2) testicular alterations, including atrophy, necrosis, and abnormal spermatocyte development. One lesion, hypertrophic gastritis, has been observed only in primates. This lesion is characterized by marked hypertrophy of the gastric (stomach) mucosa, which occurs in the fundic and pyloric regions combined with small gastric ulcers penetrating the mucosa (Allen 1967). In experiments with Macaca mulatta monkeys exposed to dioxins, (Allen 1967; Allen et al. 1975; Norback and Allen 1973) researchers found reduced hematopoiesis (formation of blood cells) and spermatogenesis, degeneration of the blood vessels, focal necrosis of the liver, and gastric ulcers. Under gross observation, experimental monkeys exhibited obvious dilatation of the heart, especially on the right side. Under microscopic examination, the cardiac muscle fibers were distinctly separated by fluid, and individual muscle cells were hypertrophic, with enlarged, distorted, and hyperchromic nuclei (see Figures 18 and 19). Although the lungs of the animals were not altered appreciably, isolated areas of atelectasis (small areas of collapse), congestion, edema, and fibrosis were observed. Livers from the monkeys were small, firm, and moderately yellow, with many enlarged, multinucleated parenchyma! cells. Necrosis of parenchyma! liver cells occurred in the centrilobular zone, and some areas of fibrosis occurred in the periportal area. Spleens from the animals were small; the germinal centers were surrounded by only scattered lymphocytes, and the blood sinuses were practically devoid of cells. The seminiferous tubules of the testes had abundant spermatogonia and sertoli cells; only a few primary spermatocytes were present, however, and no spermatids or mature spermatozoa were observed. Gastrointestinal changes have been described earlier. 159 �Figure 18. Drawing of tissue from heart of monkey fed 2,3,7,8-TCDD; tissue fixed with formalin and stained with hematoxylin and eosin. Muscle cells are hypertrophic with enlarged and distorted nuclei. X115. (Source: Redrawn from Norback and Allen 1973) 160 �MITOCHONDRION MYOFIBRILS Figure 19. Drawing of heart tissue from monkey fed 2,3,7,8-TCDD. Myofibrils of dilated cardiac fibers are separated, and the mitochondria are moderately swollen. Tissue fixed with Veronal acetate-buffered osmium tetroxide solution and stained with uranyl acetate. X9700. (Source: Redrawn from Norback and Allen 1973) 161 �Mesenteric (abdominal) lymph nodes of the monkeys were light tan and edematous, microscopically resembling the splenic disarray of cellular architecture. Grossly, the bone marrow resembled coagulated plasma. Microscopically, only a few hematopoietic cells were seen in the marrow; these were equally divided between members of the myeloid (white blood cell line) and erythroid (red blood cell line) series. Changes in the skeletal muscle resembled those of cardiac muscle. Skin from the experimental animals was dry and flaky; loss of eyelashes with facial edema and petechiae (small hemorrhages) were commonly observed. Microscopic changes in the skin are illustrated in Figure 20. Along with facial edema, anasarca (widespread edema of abdomen and extremities) was commonly observed. The rat also has been studied extensively (Gupta et al. 1973; Norback and Allen 1973; Kociba et al. 1978; Greig and Osborne 1978). Gross pathological observation indicated that rats died with jaundiced ears, subcutaneous tissues, and visceral organs. Uterine size was decreased, and there was a generalized loss of subcutaneous and abdominal fat. The liver and spleen were small, and the liver was friable and dark tan. All thymuses were markedly atrophied, and hemorrhages were present in the gastrointestinal tract and meninges. Microscopic observation showed a relative depletion of lymphoid cells in the spleen and lymph nodes, and markedly smaller thymic lobules with no demarcation between the cortex and medulla. Rats given large doses of 2,3,7,8-TCDD showed marked changes in liver cellular morphology and architecture, as illustrated in Figures 21 through 24. Hepatocytes were round and large, and the hepatic cords were disorganized. Increased mitoses were seen in the liver parenchyma (mass of cells), and some areas contained hepatocytes with seven to ten nuclei (see Figure 21). Individual hepatocytes showed proliferation of smooth endoplasmic reticulum and often distorted cell membranes. Also, the number of lipid droplets are increased. Atretic (degenerative and distorted) changes were noted in the ovarian follicles, and mucosol folds and glandular structures in the uterus were atrophied. Epithelial cells of the renal tubules were foamy and vacuolated with numerous hyaline droplets. Moderate to marked degenerative changes were noted in the epithelial cells of the thyroid follicles, and there were papillary projections into the lumen of the follicles. Focal hyperplasia (increased cell number) was noted in the terminal bronchioles of the lung (Figure 25). Congestion and elongation of the intestinal v i l l i also were noted. Pathology of chickens exposed to dioxins is similar to that observed in other animals (Norback and Allen 1973). Chickens succumbed very rapidly, with hydropericardium (fluid in sac surrounding heart), hydrothorax (fluid in chest cavity surrounding lungs), and ascites. They also developed liver necrosis, hypoplastic testes, altered capillary permeability, and decreased hematopoiesis. Gupta et al. (1973) report pathologic findings in guinea pigs and mice exposed to 2,3,7,8-TCDD. In guinea pigs, mitotic figures and loss of lipid vacuoles were observed in the zona fasiculata, along with atrophy of the 162 �LARGE KERATIN CYST HAIR FOLLICLE • • *a S / ••--,•• SAvWtf ••• *-V SMALL KERATIN CYST Figure 20. Drawing of section of skin of monkey fed 2,3,7,8-TCDD. Note the presence of keratin cysts and the lack of a hair shaft in the hair follicle. Tissue fixed with formalin and stained with hematoxylin and eosin. X15. (Source: Redrawn from Norback and Allen 1973) 163 �NUCLEI Figure 21. Drawing.of part of a multinucleated liver cell from a female rat given 0.1 yg of 2,3,7,8-TCDD/kg/day for 2 years. Uranyl acetate-lead citrate stain. XI620. (Source: Redrawn from Kociba, et al.;1978) 164 �ROUGH ENDOPLASMIC RETICULUM NUCLEUS CELL MEMBRANE CELL MEMBRANE MITOCHONDRION LIPID DROPLET VESICLE SMOOTH ENDOPLASMIC RETICULUM LIPID DROPLETS Figure 22. Drawing of liver tissue from rat fed 2,3,7,8-TCDD. Tissue'sample fixed in Verona! acetate-buffered osmium tetroxide solution and stained with uranyl acetate. X20400. (Source: Redrawn from Norback and Allen 1973) 165 �NORMAL MEMBRANES Figure 23. Drawing of normal membrane junctions from the periportal region of a test animal 42 days after administration of 200 yg/kg 2,3,7,8-TCDD. Uranyl acetate and lead citrate stain. X16000. (Source: Redrawn from Greig and Osborne 1978) 166 �Figure 24. Drawing of distorted periportal membrane junction, showing loss of continuity of plasma membranes between parenchymal cells (42 days after 200 yg/kg 2,3,7,8-TCDD); small blebs of normal membrane remain. Uranyl acetate and lead citrate stain. X42500. (Source: Redrawn from Greig and Osborne 1978) 167 �Figure 25. Focal alveolar hyperplasia near terminal bronchiole within lung of rat given 2,3,7,8-TCDD at dosage of 0.1 yg/kg per day. H&E Stain. XI00. (Source: Redrawn from Kociba 168 et al. 1978) �zona glomerulosa of the adrenals. Guinea pigs also had widespread hemorrhages in the subserosal region of the gastrointestinal tract, bladder, lymph nodes, and adrenals. Pathologic findings observed in mice are similar to those noted in other animals. ACUTE TOXICITY The acute and subacute toxic potential of 2,3,7,8-TCDD in animals relative to some other chlorodioxins and pesticides is illustrated in Tables 31 and 32. As the tables indicate, 2,3,7,8-TCDD is a highly toxic material, several orders of magnitude more potent than many pesticides. Some consider it to be the most toxic small molecule made by man (Poland and Kende 1976). Comparative Lethal Doses Table 33 lists the LD50 values for various substituted dibenzo-paradioxins. The 2,3,7,8-TCDD isomer is 3 to 100 times more potent than the other tetrachlorinated isomers (Dow 1978). In comparison with 2,3,7,8-TCDD, the 1,3,6,8- and 1,3,7,9-tetrachlorinated isomers have little biological activity (Rappe 1978). Both octachlorodioxin and the unsubstituted dioxin are relatively nontoxic. Dioxin structure-activity relationships are discussed in a later subsection. The LD50 values for 2,3,7,8-TCDD in rats, guinea pigs, and rabbits are presented in Table 33. The male guinea pig appears to be the most sensitive, having an LD50 of 0.0006 mg/kg (0.6 ug/kg). The LD50 values in monkeys exposed to a single oral dose range from 50 to 70 ug/kg body weight (McConnell, Moore, and Dalgard 1978). In mice, the LD50 is 0.2837 mg/kg body weight (McConnell et al. 1978). Target organs for the acute toxic effects of TCDD in commonly studied laboratory animals are listed in Table 34. All species of animals studied by Moore et al. (1976) showed severe thymus involution and testicular degeneration. Reduction in the white pulp of the spleen combined with bone marrow hypoplasia (decreased cell number) were other common effects. Mice exhibited the greatest degree of liver toxicity, and female monkeys showed the most skin lesions and bile duct hyperplasia. Ascites was common in monkeys, but was more prominent in mice. Hyperplasia of the renal pelvis and urinary bladder was common in guinea pigs. Gastrointestinal hemorrhages were common in both mice and guinea pigs. Aquatic Toxicity No data are available concerning the acute toxicity of 2,3,7,8-TCDD on saltwater organisms, and there are only scant data relative to freshwater aquatic life (U.S. EPA 1978c). Exposures of fish and invertebrate species to the dioxin in water and food and by intraperitoneal injection have demonstrated a variety of adverse effects at very low concentrations. Model 169 �TABLE 31. TOXICITIES OF ORGANIC PESTICIDES AND 2,3,7,8-TCDDa Maximum dose producing no observed adverse effect, mg/kg per day Compound 2,3,7,8-TCDD Disolfoton and Phorate Diazinon Parathion and Methyl parathion Aldicarb Malathion Silvex (2,4,5-TP) Hexachlorobenzene Hexachlorophene Toxaphene MPCA Pentachlorophenol Butachlor Methoxychlor 2,4,5-T Bromacil 2,4-D Ortho- and Paradichlorobenzene Atrazine Captan Arachlor Methyl methacralate Di-n-butyl phthalate Styrene Source: National Academy of Science 1977. 170 _5 10 0.01 0.02 0.043 0.1 0.2 0.75 1 1 1,,25 1 ,25 3 10 10 10 12. 5 12.,5 13.,4 21 5 50 100 100 110 133 �TABLE 32. ACUTE TOXICITIES OF.DIOXINSC Substitutions with chlorine LD50 (Ma/kg)b Guinea pigs Mice Nonec >50 x 103 (i.p.)e 2,8 >300,000 29,444 0.6-2 3.1 1,125 72.5 70-100 60-100 , >600;7180a 2,3,7 2,3,7,8 1,2,3,7,8 1,2,4,7,8 1,2,3,4,7,8 1,2,3,6,7,8 1,2,3,7,8,9 1,2,3,4,6,7,8 rC 1,2,3,4,6,7,8,9 l-N02-3,7,8 l-NH2-3,7,8 l-N02-2, 3,7,8 l-NH2-2,3,7,8 >30,000 >30,000 47.5 194.2 >3,000 283.7 337.5 >5,000 825 1,250 >1,440 >4 x 106 >2,000 >4,800 ? Unless otherwise noted, taken from McConnell et al. 1978. ° All values are for oral doses unless noted; test period is 30 days. World Health Organization, IARC Monographs on the Evaluation of the d Carcinogenic Risk of Chemicals to Man. 15:69-70, August 1977. EPA-RPAR on Pentachlorophenol. Federal Register 43(202):48454, October 18, 1978. Interperitoneal. TABLE 33. ACUTE TOXICITIES OF 2,3,7,8-TCDD FOR VARIOUS SPECIES3 Species Route of exposure Sex Dosage, LD50 mg/kg Rat Male Female Oral Oral 0.022 0.045 Guinea pig Male Female Oral Oral 0.0006 0.0021 Rabbit Female and male Female and male Female and male a Oral Dermal Interperitoneal Source: Schwetz et al. 1973. 171 0.115 0.272 >0.252 �TABLE 34. SUMMARY OF ACUTE TOXICITY EFFECTS OF 2,3,7,8-TCDDc Mice Guinea pigs Monkeys (female) Thymus involution Spleen reduction (white pulp) Bone marrow hypoplasia Liver, megalocytosis/ degeneration Bile duct hyperplasia Testicular degeneration N/A Renal pelvis hyperplasia Urinary bladder hyperplasia Adrenal cortical atrophy (Zona Glomerulus) Hemorrhage: Intestinal Adrenal Ascites Cutaneous lesions Source: Moore et al. 1976. Key as follows: - no effects + mildly affected ++ moderately affected severely affected 172 �ecosystem studies have demonstrated bioconcentration factors for 2,3,7,8TCDD of 3,600 and 26,000 over a period of 3 to 31 days (Isensee and Jones 1975). Exposure of coho salmon to an aqueous concentration of 0.000056 ug/liter under static conditions for 96 hours resulted in 12 percent mortality, whereas mortality of control fish was 2 percent (Miller, Morris, and Hawks 1973). In the same study, all coho salmon exposed to 0.056 (jg/liter for 24 hours were dead within 40 days. Isensee (1978) reports that 3 ppt of 2,3,7,8-TCDD is acutely toxic to mosquito fish. Structure-Activity Relationships The general structure-activity relationships of dibenzo-para-dioxins are presented earlier in this Section. Briefly, at least one hydrogen atom and a minumum of three laterally placed halogen atoms must be present in the dioxin structure for it to be toxic (Kende and Wade 1973). Studies have shown that a dioxin 1 s ability for enzymatic induction correlates well with its toxic potential and thus its structure. In one study, age- and sex-related differences in hepatic mixed-function oxidase activity in rats apparently were inversely correlated with the 20-day LD50 of 2,3,7,8-TCDD (Beatty et al. 1978). The study also examined the effects of administering inducers and inhibitors of the hepatic mixed-function oxidase enzyme systems on the 20-day LD50 of 2,3,7,8-TCDD in rats. In all cases, there was an inverse relationship. CHRONIC TOXICITY Although chloracne is a common indicator of 2,3,7,8-TCDD exposure in humans and some animals, chronic exposure to this dioxin can affect nearly every organ system. In addition to chloracne, another dermatologic manifestation of exposure is porphyria cutanea tarda (PCT), a photosensitive dermatosis caused by altered porphyrin metabolism. Hepatic (liver) toxicity resulting from prolonged exposure to 2,3,7,8-TCDD is common in animal models and has been observed in human workers after industrial exposures. In animal models, the dioxin has caused damage to renal (kidney) tubular epithelium and caused alteration in levels of serum gonadatropin (pituitary hormones influencing reproductive organs). A profound deficit in cellmediated immunity is produced in experimental animals exposed to 2,3,7,8TCDD in the perinatal period. Along with thymic atrophy, exposure to 2,3,7,8-TCDD leads to a depletion of cells in the spleen, lymph nodes, and bone marrow. Hypertrophic gastritis has been observed frequently in exposed monkeys. Alterations in lipid metabolism produced by 2,3,7,8-TCDD exposure may greatly increase the risk of atherogenesis in occupationally exposed workers. Neuropsychiatric symptoms including neurasthenia (depressive syndrome with vegetative symptoms) and peripheral neuropathies have been attributed to 2,3,7,8-TCDD exposure. These various aspects of chronic toxicity are discussed in the following subsections. 173 �Dermatologic Effects Dermatologic diseases are perhaps the most sensitive indicators of 2,3,7,8-TCDD exposure and toxicity in humans. Although chloracne is the most frequently observed dermatosis, PCT has been observed in as many as 10 percent of a group of occupationally exposed workers (Purkyne et al. 1974). Chloracne— Chloracne, which is characterized by comedones, keratin cysts, pustules, papules, and abscesses, is a classical sign of 2,3,7,8-TCDD exposure in humans (U.S. NIEHS IARC 1978). Chloracne can be caused by ingestion, inhalation, or skin contact with chlorodibenzodioxins, and the disease may clear in a few months or persist for as long as 15 years (Crow 1978). All chlorodibenzodioxins that are acnegenic are also systemic toxins, but the external dose needed to produce chloracne is far lower than that needed to cause systemic toxicity (Crow 1978). Chloracne, which can be an extremely refractory form of occupational acne, was first described by Von Bettman in 1897 (Taylor 1974). The symptoms may appear weeks or months after the initial exposure to chlorodibenzodioxins. Rabbits can be used to test the acnegenicity of a chlorodibenzodioxin, because these compounds are active skin irritants and induce acneform lesions when applied to the skin of rabbit ears (Kimmig and Schulz 1957). Kimmig and Schulz (1957) provided a detailed description of the clinical manifestations of chloracne that developed in 31 workers in a German plant producing 2,4,5-T in 1954. In heavily exposed workers, dermatitis of the face accompanied by erythma and swelling was first observed. As these symptoms faded, acneform lesions appeared on the face and later on other parts of the body. In most workers, the initial manifestations of chloracne were patches of open comedones (blackheads) followed by pustules in the zygomatic region (cheeks) of the face. Upon initial examination, the observed skin changes included many blackheads, pinhead to pea-size closed comedones (whiteheads), associated follicular hyperkeratosis, inflamed pimples, pustules, and large boils. The face, ears, throat, and neck were affected in all cases; in severe cases, lesions were encountered on the breast, back, epigastrium (skin of upper abdomen), genitals, and extensor surfaces of the arms and thighs. Porphyria Cutanea Tarda (PCT)-Porphyria cutanea tarda (PCT) is a skin condition that usually occurs as a photosensitive dermatosis and is characterized by development of vesiculobullous (blistering) lesions over exposed areas (Benedetto and Taylor 1978). The dermatosis is precipitated by minor trauma, and may result in areas of healed bullae, crusts, scars, and milia. Hyperpigmentation, hypertrichosis (excessive growth of hair), and schlerodermoid (tightening of skin over the fingers) changes can also occur, along with dark red urine (Benedetto and Taylor 1978). Animal studies have shown that 2,3,7,8-TCDD is the porphyrinogenic compound formed during the manufacture of 2,4,5-T. Jones and Sweeney (1977) have shown that uroporphyrinogen decarboxylase (UD) levels can be depressed in rats given 2,3,7,8-TCDD. Their results indicate that the dioxin depresses UD levels sufficiently to produce the biochemical 174 �disturbance of PCT. Sweeney (1979) notes that iron-deficient mice are protected from porphyria produced by 2,3,7,8-TCDD exposure. Hepatic Effects The hepatotoxicity of 2,3,7,8-TCDD appears to be dose-dependent, and the severity of any changes produced varies among species (Gupta 1973). In rats and rabbits, hepatic necrosis produced by this compound is probably a contributing cause of death, whereas hepatic necrosis and liver insufficiency are less extensive in mice and are minimal relative to these disorders observed in guinea pigs and monkeys (U.S. NIEHS IARC 1978). Van Miller et al. (1977) noted liver necrosis and bile duct hyperplasia in a group of rats fed 1.0, 0.6, and 0.05 ppm 2,3,7,8-TCDD for 65 weeks. In a 13-week toxicity study in which the dioxin was administered orally to rats, doses of 1.0 ug/kg per day increased the levels of serum bilirubin and alkaline phosphatase and caused pathologic changes in the liver; doses of 0.1 (jg/kg per day caused a slight degree of liver degeneration (Kociba et al. 1976). The histopathologic changes in rat liver resulting from 2,3,7,8-TCDD exposure were described earlier. Renal Effects Several recent studies have examined the effects of 2,3,7,8-TCDD upon renal function in the rat (Anaizi et al. 1978; Hook et al. 1978). Anaizi et al. studied the steady-state secretion rate of phenosulfonphthalein (PSP) in rats pretreated with 10 ug/kg of 2,3,7,8-TCDD 5 to 7 days prior to in vivo measurements. The results were as follows: A significant increase in the tubular secretion rate of PSP occurred at low plasma levels of PCP. There was no increase in the maximum secretory (Tm-PSP). capacity for PSP A significant change in the glomerular filtration rate from 1.17 to 0.90 ml/min per gram of wet kidney weight was observed in treated rats without a change in the mean arterial pressure. Anaizi et al. inferred from this study that glomerular structures in rats are highly sensitive to 2,3,7,8-TCDD. Hook et al. (1978) examined renal accumulation of p-aminohippurate (PAH) and N-methyl-nicotinamide (NMN) in rats given 10, 25, or 50 pg/kg 2,3,7,8-TCDD. In the 10 ug/kg dose group, only NMN accumulation was slightly decreased at 7 days. At 25 ug/kg, the capacity of renal tissue to transport both PAH and NMN was reduced 7 days after exposure. The GFR and effective renal plasma flow were decreased in rats after doses of 25 or 50 ug/kg. Volume expansion did not alter this relationship in the study. Thus these two independent studies confirmed the ability of 2,3,7,8-TCDD to decrease renal function in the rat. 175 �Endocrine Effects It has been known for some time that 2,3,7,8-TCDD exposure in man is associated with hormonal imbalances that lead to acne, hirsutism, and loss of libido. Recently it has been shown that 2,3,7,8-TCDD can also have a dramatic effect upon hormones involved in reproduction. A recent study has indicated a suppressive effect upon testicular microsomal cytochrome P-450 content in guinea pigs (Piper 1979). Another study has shown that 2,3,7,8TCDD increases serum thyroid stimulating hormone in humans 4- to 5-fold, and preliminary observations indicate that serum levels of prolactin and follicle stimulating hormone are affected in rats following treatment with the dioxin (Gustafsson and Ingelman-Sundberg 1979). Testosterone hydroxylation in the 2p- and ISorpositions has been reduced by 50 percent in rats receiving less than 1 M9/kg of 2,3,7,8-TCDD orally (Hook et al. 1975). Similarly, exposures of female rats have shown 3- to 5-fold increases in the following enzyme activities (Gustafsson and Ingelman-Sundberg 1979): 1. 7a and 6p-hydroxylases active on 4-androstene-3,17-dione; 2. 7ot and 2a hydroxylases active on 5orandrostane-3a, 17p-diol; and 3. 16a and 6p-hydroxylases active on 4-pregnene-3,10-dione. One recent study examined hormonal alterations in female rhesus monkeys fed a diet containing 500 ppt of 2,3,7,8-TCDD per day for 9 months (Barsotti, Abrahamson, and Allen 1979). Steroid analysis at 6 months showed alterations in five of seven animals treated. Progesterone levels in three animals decreased to 72.4 percent, 51.9 percent, and 47.3 percent of their pretreatment values. During the same interval, estradiol levels in two of these animals also decreased to 50.4 percent and 43.2 percent of the control values. The remaining two animals with abnormalities showed anovulatory patterns for both steroids. Estradiol never rose above 30 pg/ml of serum and progesterone remained below 400 pg/ml of- serum throughout the menstrual cycles. After these analyses, all animals were bred. All of the control animals conceived and gave birth to healthy infants. The two dioxin-treated animals in which estradiol and progesterone levels had remained normal did conceive, but one animal aborted the conceptus. Several other treated monkeys conceived, but all subsequently aborted. The one dioxin-treated animal that carried a fetus to term delivered a normal, healthy infant. After nine months, the only monkey that had showed hormonal alterations and survived was placed back on the control diet and subsequently delivered a normal, healthy infant. Immunologic Effects Exposure to 2,3,7,8-TCDD has caused thymus atrophy in all mammalian species studied. As illustrated in Table 35, impairment of cellular immunity has been a constant finding in studies of the effects of this dioxin on the immune system of animals. Thymus (T-)-dependent lymphocytes are most affected by the exposure; however, T-helper-cells are less compromised than other types of T-cells (Faith and Luster 1977). 176 �TABLE 35. EFFECTS OF IN VIVO 2,3,7,8-TCDD EXPOSURE ON FUNCTIONAL IMMUNOLOGICAL PARAMETERS' Species Parameter Guinea pig Rat Rat Mouse Rat, mouse Rat Delayed type hypersensitivity Delayed type hypersensitivity Graft versus host activity Graft versus host activity Rejection of skin allografts Lymphocyte transformation by PHA and Con A Mouse Guinea pig Rat Lymphocyte transformation by PHA Antibody response to tetanus toxoid Antibody response to bovine ^" Effect Reference Vos et al. 1973 Moore and Faith 1976; Vos et al. 1973 Vos and Moore 1974 Vos and Moore 1974; Vos et al. 1973 Vos and Moore 1974 Vos and Moore 1974; Moore and Faith 1976 +C/-e _c,f/+c,g _d,f/_d,g Vos and Moore 1974 Vos et al. 1973 Moore and Faith 1976 u Source: Vos et al. 1978. Denotes the suppressive effect on immunological parameters +, slight; ++, moderate effect; -, no effect. . Treatment of young animals. Treatment during the perinatal period. .- Treatment of adult animals. Primary antibody response. 9 Secondary antibody response. �Suppression of cell-mediated immunity appears to be age-related in the mouse and rat; perinatal exposure causes the greatest effect (Luster et al. 1978). It is important to recognize that TCDD can produce immunosuppressive effects at exposure levels too low to produce clinical or pathological changes (Thigpen et al. 1975). Many studies have examined the effects of exposure to 2,3,7,8-TCDD on impairment of cell-mediated immunity. Several studies have examined the effects of either postnatal or both pre- and postnatal exposure of rat pups by maternal dosing (Faith and Luster 1977; Luster et al. 1978). Results indicated that cell-mediated immune functions were depressed up to 133 days of age in both groups but less severely in animals exposed only postnatally. In addition, the ratio of thymus to body weight was depressed up to 145 days of age in prenatal ly exposed rats, but the ratio was suppressed only up to 39 days of age in the postnatally exposed group. These studies established that depression of T-cell function is selective in that helper T-cell function was spared. Vos and Moore (1974) demonstrated that cell-mediated immunity in 1-month old rats was depressed only when toxic doses of 2,3,7,8-TCDD were administered. In vitro testing has demonstrated that DMA, RNA, and protein synthesis in splenic lymphocytes is severely inhibited when mouse spleens are only briefly exposed to 10 "7 millimolar solutions of 2,3,7,8-TCDD (Luster 1979a). Multiple studies have examined the effects of 2,3,7,8-TCDD exposure upon in vivo susceptibility to pathogenic organisms. Thigpen et al. (1975) administered sublethal levels of the dioxin to mice and then subjected them to challenges with Salmonella bern and Herpesvirus suis. At dose schedules of 1 ug/kg weekly for 4 weeks, Salmonella infection led to significant increases in mortality and reduction of time from infection to death. The dioxin exposure had no apparent effect upon the outcome of infection with Herpesvirus suis. Other researchers found that mouse pups from mothers fed up to 5 ppb of 2,3,7,8-TCDD withstood a live Listen'a challenge as well as did the controls; however, maternal feeding at 2,3,7,8-TCDD levels as low as 1 ppb rendered offspring more sensitive to challenge with endotoxin (cell walls of gram negative bacteria) (Thomas and Hinsdill 1979). Nonspecific killing and phagocytosis* of Listeria monocytogenes in mice were not influenced by administration of 2,3,7,8-TCDD (Vos et al. 1978). In the same study, treatment with the dioxin did not affect macrophage reduction of nitro-bluetetrazoliurn, and the authors speculated that endotoxin sensitivity in treated animals is not the result of altered phagocytic function of macrophages. Similarly, challenge with pathogenic streptococcus in aerosol form led to similar mortality rates among treated mice and controls (Campbell 1979). Humoral immunity and B-lymphocyte function are resistant to the toxic effects of 2,3,7,8-TCDD. Faith and Luster (1977) found that humoral immune responses to bovine gamma globulin were not suppressed in rats treated with the dioxin. Luster (1979b) then demonstrated that T-lymphocytes are much The process by which cells engulf and destroy foreign material. 178 �more susceptible to dioxin-induced immune-suppression than B-lymphocytes with mitogens specific for lymphocyte subpopulations. By measuring the antibody response against tetanus toxoid in guinea pigs, Vos et al. (1973) showed only a slight decrease in humoral immunity in 2,3,7,8-TCDD-treated animals. Thomas and Hindsill (1979) demonstrated normal primary and secondary antibody responses in treated mice. Hematologic Effects One of the major target organs for TCDD toxicity is the hematopoietic system. Although many species have been studied, anemia has been observed only in rhesus monkeys (Allen 1967). This anemia was of an aplastic type (characterized by lack of cells in bone marrow) and was accompanied by atrophic bone marrow. The only abnormalities of the hematopoietic system noted in 2,3,7,8-TCDD-treated rats have been thrombocytopenia (increased numbers of platlets) and terminal elevated packed red cell volumes secondary to hemoconcentration (Weissberg and Zinkl 1973). In this study, the platelet counts of treated rats were significantly reduced and their bone marrows contained normal numbers of megakaryocytes. Zinkl et al. (1973) studied the hematologic effects of exposing guinea pigs and mice to TCDD. The leukocyte and lymphocyte counts in mice given a single oral dose of as little as 1.0 ug/kg TCDD were significantly lower after 1 week. A similar relationship was observed in guinea pigs treated with tetanus toxoid or Mycobacterium tuberculosis. In mice, the lymphopenia (decreased numbers of lymphocytes) was reversed 5 weeks after exposure to the dioxin. Gastrointestinal Effects Two studies have explored the effect of dibenzo-para-dioxins upon intestinal absorption of nutrients. Ball and Chhabra (1977) used in vitro everted sac and in situ closed loop techniques to study the effect of a toxic dose of 2,3,7,8-TCDD (100 ug/kg po) on adult male rats. Glucose uptake declined during the first few hours following dosage, rose above controls between one and two weeks, and declined again after three weeks. Leucine uptake was depressed throughout the study. Madge (1977) studied the effects of 2,3,7,8-TCDD and OCDD on function of the small intestine in mice. He found that absorption of D-glucose decreased following a single oral dose of each of the compounds. No effect was noted on the absorption of D-galactose, L-arginine, or L-histidine. Total fluid transfer was generally unaffected by treatment with either compound, and D-mannose, an exogenous energy source, abolished the apparent malabsorptive effects of D-glucose in treated animals. Neuropsychiatric Effects Two studies have examined the neuropsychological function of rats exposed to 2,3,7,8-TCDD. Creso et al. (1978) found that exposure induced 179 �irritability, aggressiveness, and restlessness in rats, without acquisition or loss of a conditioned avoidance reflex. In this study, the dioxin stimulated the activity of adenyl cyclase in the rat brain striatum and hypothalmus in vitro. It also enhanced the stimulatory effect of dopamine on striatal adenyl cyclase; however, this action was blocked by haloperidol. The study also showed that 2,3,7,8-TCDD acted synergistically with histamine in stimulating the hypothalmic adenyl cyclase. Elovaara et al. (1977) showed that treatment with 2,3,7,8-TCDD caused: (1) an increase in acid proteinase activity in the brains of normal Wistar rats, (2) reduction of RNA and protein contents in heterozygous Gunn rats, and (3) no changes in homozygous Gunn rats. Purkyne et al. (1974) found various psychiatric and neurological complaints in a cohort of 55 workers occupationally exposed to 2,3,7,8-TCDD. Seventeen subjects showed neurological abnormalities. The most common disorder was polyneuropathy of the lower extremities (confirmed by electromyography). Most of these patients suffered from psychiatric disorders such as severe neurasthenia syndromes with vegetative symptoms. These workers complained of weakness and pain in the lower extremities, somnolence, insomnia, excessive perspiration, headache, and various sexual disorders. DEVELOPMENTAL EFFECTS A brief review of the pertinent nomenclature is given here to characterize the several developmental effects discussed in this section. The terms embryotoxicity and fetotoxicity denote all transient or permanent toxic effects induced in an embryo or fetus, regardless of the mechanism of action. These are the most comprehensive terms. A special fetotoxic effect is teratogenicity, which is defined as an abnormality originating from impairment of an event that is typical in embryonic or fetal development. For example, fetal growth retardation is a fetotoxic but not a teratogenic effect of 2,3,7,8-TCDD (Neubert et al. 1973). The first clue to the teratogenic and fetotoxic potential of 2,3,7,8TCDD resulted from a National Cancer Institute study begun in 1964 to evaluate the carcinogenic and teratogenic potential of a number of herbicides (Collins and Williams 1971). In this study, 2,4,5-T and 2,4-D were shown to induce increased proportions of abnormal fetuses in hamsters. Courtney (1970) demonstrated the teratogenicity of 2,4,5-T containing approximately 30 ppm of 2,3,7,8-TCDD in two strains of mice. Subsequent investigations studied the fetotoxicity and teratogenicity of both 2,4,5-T and 2,3,7,8-TCDD in a number of species. 180 �Teratogenicity Courtney (1970) showed that 2,4,5-T containing 2,3,7,8-TCDD increased the incidence of cleft palate in both C57BC/6 and AKR mice. Neubert et al. (1972), using the purest available sample of 2,4,5-T, showed that at doses higher than 20 mg/kg given orally during days 6 to 15 of gestation, the frequency of cleft palate was significantly increased in NMRI mice. The maximal teratogenic effect was produced when the drug was administered on days 12 or 13 of gestation. In the same study, doses exceeding 1 ug/kg of 2,3,7,8-TCDD produced an increased rate of cleft palate; maximal teratogenicity occurred with administration on days 8 and 11 of gestation. Although Courtney and Moore (1971) found no potentiation of teratogenicity with combinations of 2,4,5-T and 2,3,7,8-TCDD, Neubert and coworkers found that 1.5 ppm of 2,3,7,8-TCDD administered with 30 to 60 mg/kg 2,4,5-T potentiated the increase in cleft palate frequency. Moore and coworkers (1973) found that the mean average incidence of cleft palate was 55.4 percent in mice exposed to 3 ug/kg 2,3,7,8-TCDD on days 10 to 13 of gestation. In 1976, the threshold teratogenic dose of 2,3,7,8-TCDD in CF-1 mice was estimated to be 0.1 (jg/kg per day (Smith, Schwetz, and Nitchke 1976). In golden hamsters, oral administration of 2,4,5-T containing dioxin on days 6 to 10 of gestation increased the incidence of absence of the eyelid (Collins and Williams 1971). Although 2,3,7,8-TCDD is fetotoxic in primates at doses as low as 50 ppt, it has not been shown to be teratogenic in this species (Schantz et al. 1979)'. Fetotoxicity and Embryotoxicity In general, 2,4,5-T and 2,3,7,8-TCDD produce fetotoxicity at doses that do not produce teratogenic effects in a wide variety of species. Fetotoxic effects of 2,4,5-T containing 2,3,7,8-TCDD were first noted in Courtney's original work (1970). Both species of mice studied showed increased incidences of cystic kidneys, while in rats, fetal gastrointestinal hemorrhages and increased ratios of liver to body weight were also noted. Highman and Schumacher (1977) later demonstrated that cystic kidneys in mice exposed to 2,4,5-T containing 2,3,7,8-TCDD were due to retardation in fetal renal development and downgrowth of the renal papilla into the pelvis. The results of this study demonstrated a retarded development of fetal renal alkaline phosphatase, and thus support the hypothesis that cystic kidneys in mice are a fetotoxic and not truly a teratogenic effect. Moore et al. (1973) proved that prenatal and postnatal kidney anomalies had a common etiology, and the incidence and degree of hydronephrosis* was a function of dose and of the length of exposure of a target organ. Other fetotoxic effects of 2,4,5-T and 2,3,7,8-TCDD include thymic atrophy, fatty infiltration of the liver, general edema, delayed head ossification, low birthweight, fetal resbrptions, and embryolethality. Many studies have examined the fetotoxic effects of 2,4,5-T and 2,3,7,8-TCDD on various species. In a study of the effects of 2,3,7,8-TCDD on the rat, no adverse effects were noted at the 0.03 ug/kg level; but fetal mortality, early and late resorptions, and fetal intestinal hemorrhage were * Dilation of renal pelvis usually associated with an obstructed ureter. 181 �observed in groups given 0.125 to 2.0 (JQ/kg, the incidence increasing as the dose increased (Sparschu, Dunn, and Rowe 1971). In the CD rat, 2,4,5-T was neither teratogenic nor fetotoxic; however, 2,3,7,8-TCDD produced kidney anomalies (Courtney and Moore 1971). In golden hamsters, 2,4,5-T containing 2,3,7,8-TCDD caused delayed head ossification in a dose-dependent fashion (Collins and Williams 1971). Neubert and Dillman (1972) determined that the threshold dose of 2,4,5-T that produced an increase in embryo!ethality was 10 to 15 mg/kg, whereas 2,3,7,8-TCDD doses of 4.5 ug/kg produced marked increases in embryo!ethality in NMRI mice. Cystic kidneys occurred unilaterally in 58.9 percent and bilaterally in 36.3 percent of mice pups exposed to 1 (JQ/kg 2,3,7,8-TCDD (Moore et al. 1973). Murray (1978) reports a three-generation study of rats exposed to 0.001, 0.01, or 0.1 M9/kg of 2,3,7,8-TCDD. Through three successive generations the reproductive capacity of rats ingesting the dioxin was clearly affected at dose levels of 0.01 and 0.1 ug/kg per day, but not at 0.001 ug/kg per day. In the most recent primate study, eight adult female rhesus monkeys were fed a diet containing 50 ppt 2,3,7,8-TCDD for 20 months (Schantz et al. 1979). After 7 months attempts were made to breed the females. In this group there were four abortions and one stillbirth. All eight control animals reproduced successfully. In the dioxin-exposed group, two animals were not able to conceive and two were able to carry their infants to term. One study examined the fetotoxic potentials in mice of other members of the dibenzo-para-dioxin class of compounds (Courtney 1976). None of the dibenzo-para-dioxins studied were as toxic as 2,3,7,8-TCDD, and some of the compounds could be considered relatively nontoxic. Although the mixture of di-CDD and tri-CDD produced a slight increase in the number of abnormal fetuses, it is doubtful that the malformations were produced by the mixture. Most of the malformations (a mild form of hydronephrosis) were in mouse pups from one litter, and no malformations were observed at a higher dose level. The 1,2,3,4-TCDD compound did not increase the incidence of malformation at any dose level. Oral administration of 5 or 20 mg/kg per day of OCDD to pregnant mice did not alter fetal development. In summary, related dibenzo-para-dioxins were relatively nontoxic and were not teratogenic at the doses studied. CARCINOGENICITY Several studies of rats and one study of Swiss mice demonstrated an increased incidence of neoplasms in animals exposed to 2,3,7,8-TCDD (Van Miller, Lalich, and Allen 1977; Kociba et al. 1978; Toth et al. 1979). Van Miller and coworkers exposed rats to diets containing the dioxin at concentrations of 1, 5, 50, or 500, ppt, or 1, 5, 50, 500, or 1000 ppb. In this study, the overall incidence of tumors in the experimental groups was 38 percent, with no neoplasms observed in the 1 ppt group. As indicated in Table 36, among the 23 animals with tumors, 5 had two primary neoplastic (cancerous) lesions. Ingestion by rats of 0.1 ug/kg per day 2,3,7,8-TCDD 182 �TABLE 36. SUMMARY OF NEOPLASTIC ALTERATIONS OBSERVED IN 3 RATS FED SUBACUTE LEVELS OF 2,3,7,8-TCDD FOR 78 WEEKS Level No. of animals, of 2,3,7,8-TCDD with neoplasms 0 1 pptc 5 ppt 0 0 5 Diagnosis No. of neoplasms 0 0 6 1 1 1 1 ear duct carcinoma lymphocytic leukemia adenocarcinoma (kidney) malignant histiocytoma (peritoneal) 1 angiosarcoma (skin) 1 Leydig cell adenoma (testes) 50 ppt 3 3 1 fibrosarcoma (muscle) 1 squamous cell tumor (skin) 1 astrocytoma (brain) 500 ppt 4 4 1 fibroma (striated muscle) 1 carcinoma (skin) 1 adenocarcinoma (kidney) 1 sclerosing semi noma (testes) 1 ppb6 4 5 1 chol angi ocarci noma (liver) 1 angiosarcoma (skin) 1 glioblastoma (brain) 2 malignant histiocytomas (peritoneal) 5 ppb 7 10 ? Source: VanMiller, Lalich, and Allen 1977. 10 animals per group. J 1 ppt = 10"12g 2,3,7,8-TCDD/g food. Metastases 9 observed. e 1 ppb = 10" g 2,3,7,8-TCDD/g food. 183 4 squamous cell tumors (lung) 4 neoplastic nodules (liver) 2 cholangiocarcinomas (liver) �for 2 years caused an increased incidence of hepatocellular carcinomas and squamous cell carcinomas of the lung, hard palate/nasal turbi nates, or tongue, and a reduced incidence of tumors of the pituitary, uterus, mammary glands, pancreas, and adrenal glands (Kociba et al. 1978). Figures 26 and 27 illustrate the morphology of some of these lesions. In a recent study with Swiss mice, Toth et al. (1979) showed that 2,4,5-trichlorophenoxyethanol and 2,3,7,8-TCDD enhanced liver tumors in male mice in a dosedependent fashion. In this study, the increase in liver tumors was statistically significant only at 2,3,7,8-TCDD doses greater than 0.112 Multiple studies have examined the effects of 2,3,7,8-TCDD administered in combination with other known carcinogens in experimental animal test systems. Two studies used the two- stage tumori genes is assay of mouse skin (Digiovanni et al. 1977; Berry et al. 1978). Berry and coworkers noted that a dose of 0.1 pg 2,3,7,8-TCDD twice weekly was not sufficient to promote skin tumors in mice treated with 7,12-demethylbenz(a) anthracene (DMBA). Digiovanni found that at doses of 2 (jg per mouse given concurrently with DMBA, the number of tumors observed increased slightly. These data suggest that 2,3,7,8-TCDD is a weak tumor initiator in the two-stage system of mouse skin tumorigenesis. In a more recent study, Digiovanni et al. (1979) found that 2,3,7,8-TCDD could strongly inhibit the initiation of skin tumors by DMBA in female CD-I mice. In a study with mice that were genetically nonresponsive to the known carcinogen, 3-methy Ichol anthrene (MCA), exposure to 2,3,7,8-TCDD markedly increased the carcinogenic index of MCA when the compounds were administered simultaneously (Kouri et al. 1978). These data imply that the dioxin could act as a potent cocarcinogen. GENOTOXICITY Only four of the dibenzo-para-dioxins have been subjected to genotoxicity testing. These are unsubstituted dibenzo-para-dioxin, the 2,7dichloro-isomer, 2,3,7,8-TCDD, and OCDD (Wassom, Huff, and Loprieno 1978). As expected, 2,3,7,8-TCDD has been the most extensively tested, but results of these studies are inconclusive. Information implicating 2,3,7,8-TCDD as a mutagen is scarce and conflicting. Mammalian studies with dibenzo-paradioxin derivatives have been infrequent. To date, 2,3,7,8-TCDD has shown negative results when tested for dominant lethal effects in rats and weakly positive results when tested for the ability to produce chromosomal abberations in bone marrow cells of rats (Khera and Ruddick 1973; Green, Moreland, and Sheu 1977). Mutagem'city Table 37 summarizes the results of studies of the mutagenic effects of dioxins. None of the Salmonella strains capable of detecting base-pair substitutions were positive when tested with 2,3,7,8-TCDD. Some investigations have obtained positive responses in Strain TA 1532, which detects frameshift mutations. Hussain et al. (1972) report the following results of mutagenicity studies with 2,3,7,8-TCDD (99 percent) on three bacterial systems: 184 �FAT DROPLETS CANCER CELLS Figure 26. Lesion classified morphologically as hepatocellular carcinoma in liver of rat given 0.1 yg of 2,3,7,8-TCDD/kg per day. Note adjacent fibrosis, inflammation, and fatty infiltration on left. H&E stain. X200. (Source: Redrawn from Kociba et al. 1978) 185 �KERATINIZED MATERIAL Figure 27. Lesion within lung of rat given 0.1 yg of 2,3,7,8-TCDD/kg per day. Classified morphologically as squamous cell carcinoma. Note accumulation of keratinized material within lesion. H&E stain. X100. (Source: Redrawn from Kociba et al. 1978) 186 �TABLE 37. MUTAGENICITY OF DIOXIN COMPOUNDS IN SALMONELLA TYPHIMURIUMC Strains detecting base-pair substitutions Dioxin isomer G46 Strains detecting frameshifts TA1530 TA1535 TA100 2,3,7,8-TCOD 0 0 - 0 0 - 0 0 - 0 0 0 - 0 0 0 co TA1531 TA1537 TA1538 0 - - McCann 1975 0 0 0 - Nebelt 1976 + 0 0 0 Hussain 1972 0 0 Seller 1973 TA1532 TA1534 Reference - - 0 0 ? + ? OCDO - - 0 0 - 7 ? 0 0 Seller 1973 Dibenzo-pdioxin 0 0 - - 0 0 0 - - Commoner 1976 . Source: Wassom, Huff, and Loprieno 1978. 0, not tested; -, negative results; +, positive results; ?, doubtful mutagen. protocols. Results obtained with different experimental �(1) 2,3,7,8-TCDD significantly increased the incidence of reverse mutations from streptomycin-dependence to streptomycinindependence in the bacteria Escherichia coli SD-4 treated with 2 ug/ml 2,3,7,8-TCDD. This was the only concentration at which mutations were clearly observed. (2) Evaluation of reverse mutation from histidine-dependence to histidine-independence in Salmonella typhimurium strains TA 1532 and TA 1530 indicated that 2,3,7,8-TCDD was positive in TA 1532 but negative in TA 1530. This finding indicates that the dioxin may act as a frameshift mutagen. ICR-170 was used as a positive control in the test with 1532, but no positive or negative controls were tested with TA 1530. (3) Slight prophage inductions in Escherichia coli K-39 were observed, although data were difficult to evaluate because the DMSO solvent used in this test caused cellular effects on its own. Seiler (1973) studied the effects of 2,3,7,8-TCDD and OCDD in several strains of Salmonella typhimurium. The 2,3,7,8-TCDD was strongly mutagenic only in strain TA 1532, whereas the OCDD was questionably mutagenic in strains TA 1532 and TA 1534. McCann (1976) obtained no positive mutagenic responses in several Salmonella strains exposed to 2,3,7,8-TCDD, including TA 1532. Commoner (1976) demonstrated that unsubstituted dibenzo-paradioxin was nonmutagenic in four strains of Salmonella typhimurium. Khera and Ruddick (1973) performed dominant lethal studies with 2,3,7,8-TCDD. Groups of male Wistar rats were dosed orally with 4, 8, or 12 ug/kg per day for 7 days before they mated. Although the incidence of pregnancies from all matings was reduced, there was no evidence of induction of dominant lethal mutations during postmeiotic phases of spermatogenesis. Cytotoxicity Highly purified samples of 2,4,5-T and 2,3,7,8-TCDD were evaluated for cytological effects in the African Blood Lily plant (Jackson 1972). The tests included treatments involving both compounds in varying proportions. In contrast to a no-effect result with a highly purified sample of 2,4,5-T, dramatic inhibition of mitosis was observed in cells exposed either to a 10"4 molar solution of 2,4,5-T containing 0.2 to 1.0 H9 2,3,7,8-TCDD per liter or to a 10~4 molar solution of 2,4,5-T containing an unknown level of 2,3,7,8-TCDD. Similar results were obtained when treatments were limited to 2,3,7,8-TCDD alone. These treatments also induced formation of dicentric bridges and chromatin fusion, with formation of multi-nuclei or a single large nucleus. Because these effects were not evident in the pure 2,4,5-T sample, Jackson concluded that the cytological effects were due to the 2,3,7,8-TCDD contaminant. Tests for cytological effects in a wild type Drosophila fly were conducted with 2,4,5-T containing less than 0.1 ppm 2,3,7,8-TCDD (Davring and Summer 1971). Twenty-four hours after eclosion the adult flies were exposed 188 �to 250 ppm 2,4,5-T in their food. Results indicated that this formulation affected early oogenesis and caused sterility. It is not stated unequivocally that the observed sterility was of genetic origin. In an animal study (Greig et al. 1973), male Portion rats were treated with single oral doses (50 to 400 MQ/kg) of 2,3,7,8-TCDD dissolved in either dimethyl sulfoxide or arachis (peanut) oil. In the rat livers, parenchymal cell structures were altered and many cells were multinucleated. No mitoses were observed, and there were occasional pyknotic nuclei. The investigators postulate that 2,3,7,8-TCDD interfered with the capacity of the liver cells to maintain their correct morphology and thus led to death or structural disorganization. Similar results have been obtained by others (Buu-Hoi et al. 1971; Kimbrough et al. 1977). Vos et al. (1974) suggest that 2,3,7,8TCDD could be a hepatocarcinogen because of its specific cytological effects on the proliferating cells of the liver. Chromosomal abberations in bone marrow cells of 2,3,7,8-TCDD-treated Osborne-Mendel rats have also been reported (Green, Morel and, and Sheu 1977). No chromosomal abberations or cytogenetic damage was found, however, in bone marrow of male Osborne-Mendel rats treated with 2,7-di-CDD or unsubstituted dibenzo-p-dioxin (Green and Morel and 1975). 2,3,7,8-TCDD may be mutagenic to humans. Chromosomal abnormalities have been reported in 4 in vitro cytogenetic studies of human lymphocytes exposed to 10"7 to 10" m«molar solutions of 2,4,5-T that contained 0.09 ppm 2,3,7,8-TCDD (U.S. EPA 1978h). Breaks, deletions, and rings were observed. Chromatid breaks increased with increasing concentrations of 2,4,5-T. It was not possible to distinguish whether this was a toxic effect or a potential genetic effect. Pathophysiology Many investigators have tested apparently logical mechanisms of action for 2,3,7,8-TCDD toxicity. For the most part, these investigations have served only to disprove proposed mechanisms of action (Beatty et al. 1978; Neal 1979). The following proposed mechanisms for toxicity induced by 2,3,7,8-TCDD have been disproved: Inhibition of protein synthesis Inhibition of DMA synthesis Inhibition of mitosis Inhibition of oxidative phosphorylation Interference with the action of thyroxine Interference with glucocorticoid metabolism 189 �Increased serum ammonia levels Depletion of reduced pyridine nucleotides Production of superoxide anion Decreased hepatic ATP content Impairment of hepatic mitochondria! respiration The most promising explanations for at least the first step in the mechanism of 2,3,7,8-TCDD toxicity result from studies of hepatic ATPase activities (Jones 1975; Madhukar et al. 1979b). Jones administered 200 ug/kg of the dioxin to male albino rats, then sacrificed groups of animals at 24 hours and at 3, 5, 6, 8, 34, and 42 days. Hematoxylin and eosin stains of liver sections showed no abnormalities in the groups sacrificed in the 24-hour to 8-day intervals; however, in the remaining two groups (34 and 42 days) the liver sections showed centrilobular zone necrosis. As early as 3 days after exposure, a significant change in the pattern of the ATPase reaction was seen in all animals studied. In an area five to six cells deep around the central vein, there was no reaction along the canalicular borders of the parenchymal cells. Similar results were obtained by Madhukar, who studied Na, K, and Mg -ATPase activities in hepatocyte surface membranes isolated from male rats given 10 or 25 mg/kg 2,3,7,8-TCDD. As early as 2 days after administration of the dioxin, all of the ATPase activities were depressed in treated animals. A dose-response relationship was observed only for depression of Mg -ATPase activity. In further studies, Madhukar demonstrated that ATPase depression was not produced by in vitro exposures to 2,3,7,8-TCDD. EPIDEMIOLOGICAL STUDIES AND CASE REPORTS The most notable human exposures to 2,3,7,8 tetrachlorodibenzo-p-dioxin have occurred through accidental releases in chemical factories, or by exposure to contaminated materials or areas. Most of the studies reported in the literature, such as those cited below, are investigations of the effects of such exposures. General Acute Toxicity The immediate results of dioxin exposure are burning sensations in eyes, nose, and throat; headache; dizziness; and nausea and vomiting (U.S. NIEHS IARC 1978). Itching, swelling, and redness of the face may occur just prior to chloracne. Chloracne, similar to acne vulgaris, is one of the most consistent and prominent features of dioxin exposure, occurring within weeks of initial exposure (May 1973; Oliver 1975; Poland et al. 1971). Mclnty (1976) showed that as little as 20 pg of 2,3,7,8-TCDD on the skin can lead to chloracne development. Chloracne may appear first on the face and then spread to the arms, neck, and trunk (U.S. NIEHS IARC 1978; May 1973). Other symptoms of exposure include arthralgias (pains in the joints without asso- 190 �dated arthritic changes), extreme fatigue, insomnia, loss of libido, irritability, and nervousness (Ensign and Uhi 1978; U.S. NIEHS IARC 1978). High levels of blood cholesterol and hyperlipoproteinaemia may also develop (Oliver 1975). Other effects, which may be delayed or immediate, are porphyria cutanea tarda, hepatic dysfunction, hyperpigmentation, and hirsutism (U.S. NIEHS IARC 1978). Disorders of the cardiovascular, urinary, respiratory, and pancreatic systems (Goldman 1973), along with disorders of fat and carbohydrate metabolism also have been found (U.S. NIEHS IARC 1978). Emotional disorders, difficulties with muscular and mental coordination, blurred vision, and loss of taste and smell also may occur (Oliver 1975). Several deaths related to 2,3,7,8-TCDD have been recorded, some due to liver damage and others to chronic exposure to the chemical. Additionally, symptoms such as chloracne can be passed by an exposed person to close associates such as family members through clothing, hands, or other close contact (Mclnty 1976). General Chronic Toxicity Poland et al. (1971) studied possible toxic effects on 73 male workers in a factory producing the 2,3,7,8-TCDD-contaminated pesticide 2,4,5-T. The workers were classified according to job location. The medical or toxicological symptoms were grouped into three categories: (1) chloracne and mucous membrane irritation, (2) hepatotoxicity, neuromuscular symptoms, psychological alterations, and other systemic symptoms, and (3) porphyria cutanea tarda (PCT). Of the 73 subjects, 66 percent experienced some degree of chloracne, 18 percent of which was classed as moderate to severe. The presence of hyperpigmentation and hirsutism correlated with the severity of the acne. Among maintenance men, who were subject to the greatest exposure, the acne was more severe than that of administrative personnel, whose exposure was minimal. Urinary porphyrin values, although within normal limits, were elevated in the maintenance men as compared with the other workers. Although 2,3,7,8-TCDD and other chemicals produced in 2,4,5-T synthesis may be hepatotoxic in humans, demonstrable chemical liver dysfunction among workers in this plant was minimal. The toxic effect of 2,3,7,8-TCDD on three young laboratory scientists was reviewed in a case study by Oliver (1975). Two of the subjects worked with the dioxin for approximately 6 to 8 weeks, and the third, for approximately 3 years before onset of symptoms. The latter scientist worked only with a diluted sample of the material, whereas the other two worked on the synthesis of dioxins. Chloracne was the first symptom experienced by two of the scientists. Two of them also suffered from delayed reactions, experiencing abdominal pain, headache, excessive fatigue, uncharacteristic episodes of anger, diminished concentration, other neurological disturbances, and hirsutism approximately 2.5 years after exposure. None of the scientists showed liver damage or porphyrinuria; all three showed elevated serum cholesterol levels, evidence of hypocholesterolemia, and hyperlipoproteinaemia. No other biochemical abnormalities were noted. Over a period 191 �of 6 months (after the onset of the delayed symptoms), the symptoms subsided. All three scientists were aware of the danger involved in the substance with which they were working; they wore protective clothing, gloves, and masks, and worked under a vented hood. The author speculated that the exposures must have been extremely low. Accidental release of 2,3,7,8-TCDD occurred in an explosion at a chemical plant in Derbyshire, England. This exposure of workers resulted in 79 cases of chloracne recorded approximately 3 weeks after the explosion (May 1973). Young men with fair complexions were affected first, but the symptoms persisted longer in sallow-skinned men ages 25 to 40. Chloracne was present, in order of prevalence, on the face, extensor aspects of arms, lateral aspects of thighs and calves, back, and sternum. Most workers recovered in 4 to 6 months. Of 14 employees who were present during the explosion, 13 showed abnormal liver function and 9 developed chloracne. Those with chloracne had handled pipes, joints, and cables with bare hands and thus may have absorbed the dioxin through the skin; this finding suggests that excretion of absorbed dioxin or its products may occur through facial pores. Jirasek et al. (1973, 1974, 1976) cite many studies done on 80 industrial workers in Czechoslovakia who showed signs of intoxication from dioxin formed as a byproduct in production of the sodium salts of 2,4,5-T and pentachlorophenol. Symptoms included 76 cases of chloracne, ranging from mild to so severe that it covered the entire body and left scars. Twelve workers had hepatic lesions with symptoms of porphyria cutanea tarda. Symptoms in 17 of the workers included polyneuropathy, psychic disorders, weakness and pain in the lower extremities, somnolence or insomnia, excessive perspiration, headache, and disorders of the mental and sexual functions. One worker suffered and died from severe atherosclerosis, hypertension, and diabetes; two workers died from bronchogenic carcinoma (lung cancer) (ages 47 and 59). Periods of latency differed; in some instances severe dermatological and internal damage developed after brief exposure, whereas in others apparently long-term and massive exposure caused only mild symptoms. Another study (Poland and Kende 1976) deals with 29 workers who were accidentally exposed to 2,3,7,8-TCDD. Of the 29, all contracted chloracne, 11 developed porphyrinuria, and several developed porphyria cutanea tarda. The workers also showed signs of mechanical fragility, hyperpigmentation, hirsutism, and photosensitivity of the skin, in which sunlight exposure caused blistering. Measures were taken at this plant to decrease 2,3,7,8TCDD production and worker exposure. Within 5 years there was no evidence of porphyria or severe acne, and severity of the other symptoms was also reduced. In all cases reviewed, an acute exposure to dioxins resulting in chloracne and other acute symptoms and followed by a period of nonexposure to the substance resulted in the disappearance or diminution of the symptoms. In early May of 1971, an accidental poisoning incident killed or intoxicated many horses and other animals that came in contact with the soil of an arena sprayed with contaminated oil. Investigators identified 2,3,7,8192 �TCDD and polychlorinated byphenyls as the causitive agents (Carter et al. 1975; Kimbrough et al. 1977). A 6-year old girl who played in the arena soil developed symptoms of headache, epistaxis (nosebleed), diarrhea, and lethargy. In August 1971, she developed hemorrhagic cystitis (inflammation of the urinary bladder). The patient's symptoms resolved in 3 to 4 days and did not recur. Proteinuria and hematuria (protein and blood in the urine) disappeared within 1 week of onset. A voiding cystogram obtained 3 months later appeared normal; however, cystoscopy demonstrated numerous punctate hemorrhagic areas, especially in the trigone region of the bladder. The patient was reexamined 5.3 years after dioxin exposure. Physical examination was performed, as well as urinalysis, a voiding cystogram, an intravenous pyelogram, renal function chemistries, an electrocardiogram, stress test, liver-function tests, uroporphyrin excretion, and thyroid-function studies. Results of all tests were essentially within normal limits (Beale et al. 1977). Three other individuals exposed to the arena developed recurrent headaches, skin lesions, and polyarthralgia (Kimbrough et al. 1977). In another sprayed arena, two 3-year-old boys developed small, pale, nonpruritic, firm papules covered by blackheads on the exposed skin surfaces. These symptoms arose 1.5 months after the spraying. They increased in severity and lasted more than a year before gradually subsiding (Carter et al. 1975). Perhaps the most publicized incident of dioxin poisoning was that in Seveso, Italy. On July 10, 1976, at a plant where trichlorophenol was manufactured, an accident created temperature conditions ideal for the formation of 2,3,7,8-TCDD (Zedda, Circla, and Sala 1976). Trichlorophenol crystals and 2,3,7,8-TCDD in the form of dust were spread over the area (Hay 1976a). In addition to 170 plant employees, approximately 5000 persons were exposed (Zedda, Circla, and Sala 1976). Shortly after the accident, cases of chloracne were reported. Over the ensuing years more than 134 confirmed cases of chloracne have occurred in children, some of whom had not been in the area during July and August 1976. These latter cases indicate that enough dioxin persisted in the environment several months after the accident to cause the chloracne (Zedda, Circla, and Sala 1976). Reports of disorders among the 170 workers exposed include 12 cases of chloracne in directly contaminated workers, 29 cases of hepatic insufficiency, 28 cases of chronic bronchitis, 17 cases of arterial hypertension, 9 cases of coronary insufficiency, 8 cases of muscular astenia (weakness), and 3 cases of reduced libido (Zedda, Circla, and Sala 1976). Reported symptoms occurring among the exposed residents include chloracne, nervousness, changes of character and mood, irritability, and loss of appetite. Legal and illegal abortions were estimated at 90, and there were 51 spontaneous abortions (U.S. EPA 1978h). Several additional followup studies of the initially identified cohort have been reported recently (Reggiani 1978, 1979a,b; Pocchiari, Silano and Zampieri 1979). In 1978, Reggiani reported that chloracne had appeared almost only in children and young people. These cases tended to be mild, 193 �and spontaneous healing occurred in most. Transient lymphocytopenia and liver function abnormalities were detected. Reports at that time indicated no overt pathology of the liver, kidney, blood, reproductive organs, central and peripheral nervous systems, or metabolism of carbohydrate, fat, or porphyrin. In 1979, Reggiani reported that the incidence of chloracne remained between 0.6 and 1.5 percent in the surveyed population and other toxic manifestations initially observed remained at subclinical levels. Pocchiari, Silano, and Zampieri (1979) reported a somewhat more detailed followup of the cohort. In the cohort with highest exposure, chloracne was identified in approximately 13 percent of the screened population. About 4 percent of the workers from the plant (Pocchiari sets the number at 200) showed signs and symptoms of polyneuropathy. Subclinical peripheral nerve damage, confirmed by nerve conduction studies, was also observed fairly frequently in nonoccupationally exposed groups, and the incidence ranged from 1.2 to 4.9 percent in the screened population. Of note, there were no documented immunologic alterations in the exposed population. Eight percent of the screened population showed hepatomegally of undetermined etiology, and some of the screened population showed elevated levels of liver transaminases. The long-term effects of exposure to 2,3,7,8-TCDD in Seveso are not clear at this time. An epidemiologic survey now in progress includes general and specialized medical examinations, laboratory tests, and data on the outcome of pregnancies. Data will be collected over a period of 5 years. Cancer registries, hospital discharge forms, notifications of infectious diseases, and birth and death certificates will be used to detect any abnormalities of the health of the community (Fara 1977). Fetotoxicity and Teratogenicity Hexachlorophene (HCP) is a derivative of 2,4,5-TCP that has been used as an antibacterial agent for the past 20 years. Although there are no reports of 2,3,7,8-TCDD contamination in HCP, this drug has been shown to cause fetal malformations, some of which are severe (U.S. NIEHS IARC 1978). A study of mothers who were nurses exposed to hexachlorophene soap during early pregnancy showed that of 65 children born, 5 had severe and 6 had slight malformations. One slight malformation was observed in 68 children of an unexposed control group. Five babies died who had been washed more than three times with 3 percent hexachlorophene in a hospital. Autopsies revealed considerable brain damage in each case. In 1972, many infant fatalities were reported in France. The cause was cited as a new talc powder called "Bebe," which contained 6 percent HCP (dioxin content, if any, is unknown) (Mclnty 1976). It is reported that the local spontaneous abortion rate has increased to twice the national level in Italy since the chemical contamination of Seveso in 1976, and that similar results have occurred in Vietnam since the spraying of Herbicide Orange (Nature 1970). Unfortunately, doctors in Vietnam are unable to document increased abortion and birth defects because of inadequate medical records (U.S. EPA 1978a). 194 �In the sprayed areas of Vietnam, doctors have cited increased incidences of babies being born with extra fingers or without fingers, hands, or feet (Lawrence Eagle Tribune 1978). Recently, a group of U.S. military veterans who were in South Vietnam at the time of the spraying have reported birth defects in their offspring similar to those reported in South Vietnam (Ensign and Uhi 1978; Lawrence Eagle Tribune 1978; Peracchio 1979). An EPA study has been done on the relationship of dioxin-containing herbicides to miscarriages; specifically the study concerns the relationship between spraying 2,4,5-T on forested areas of Oregon and miscarriages among women living in Alsea, a town near a sprayed area. Scientists from Colorado State University and the University of Miami medical school compared miscarriages in the Alsea basin with those in a control area in rural eastern Oregon. The miscarriage rate in the Alsea area was significantly higher than in the control area, where 2,4,5-T was not sprayed. Miscarriage rates peaked dramatically in June of each of the 6 years studied, occurring 2 or 3 months after the yearly spring applications. From 1972 through 1977 the spontaneous abortion indexes in June were 130 per 1000 births in Alsea and 46 per 1000 in the control area. Although these data do not prove a cause and effect relationship, they are highly suggestive (Cookson 1979). A recent study deals with the relationship of neural-tube defects in New South Wales and annual usage rates of 2,4,5-T in the whole of Australia (Field and Kerr 1979). Table 38 gives data showing the annual New South Wales combined birth rates of anencephaly (congenital absence of the cranial vault), and meningo-myelocele (defect through which part of the spinal cord communicates with the environment), together with data on the usage of 2,4,5-T in Australia in the previous year. The plot in Figure 28 indicates linear correlation. Highest rates of neural-tube defects occurred for conceptions during the summer months, and maximum spraying of 2,4,5-T in New South Wales occurs during the summer months. Again, although these data are suggestive, they do not prove a cause and effect relationship. The linear correlation disappeared in 1975 and 1976; monitoring of 2,4,5-T herbicide was established in Australia to ensure that concentrations of 2,3,7,8-TCDD remain below 0.1 ppm. Nelson et al. (1979) report a retrospective study of the relationship between use of 2,4,5-T in Arkansas and the concurrent incidence of facial clefts in children. Occurrences of facial cleft generally increased with time; however, no significant differences were found in any of the study groups. The authors conclude that the general increase in facial cleft incidence in the high- and low-exposure groups resulted from better case finding rather than from maternal exposure to 2,4,5-T. Among 182 babies delivered in Seveso in the 2 months after the accident, only 16 birth anomalies were found. This level is not significantly higher than the national level. Women in early stages of pregnancy when the accident happened were not studied in this survey (U.S. EPA 1978a). 195 �TABLE 38. COMBINED RATE OF NEURAL-TUBE DEFECTS IN NEW SOUTH WALES AND PREVIOUS-YEAR USAGE OF 2,4,5rT IN AUSTRALIA3 Usage of 2,4,5-T Year 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 Neural -tube defects in N.S.W. , cases per 1000 births 1.72 1.77 1.93 1.83 2.13 2.37 1.88 2.15 2.19 2.27 2.03 2.30 Source: Field and Kerr 1979. 2,4,5-T acid in equivalent metriq tons. 196 in Australia in previous year, metric tons 90 105 188 213 201 282 170 256 241 287 466 482 �3.0 2.5 2.0 o o ? 1.5 vt <u vt 03 O IQ 0.5 50 100 150 200 250 300 2,4,5-T use, tonnes 350 400 450 Figure 28. Linear correlation of New South Wales rate for neural-tube defects with previous year's usage of 2,4,5-T in Australia. (Source: Field and Kerr 1979) 500 �Careinogenlcity Ton That et al. (1973) report an increase in the proportion of prjmary liver cancer among all cancer patients admitted to Hanoi hospitals during the period 1962 to 1968; this increase is relative to the period 1955 to 1961, just before the spraying of Herbicide Orange began. Theiss and Goldmann (1977) trace 4 cancer deaths out of 15 deaths occurring in 53 workers exposed to 2,3,7,8-TCDD after a manufacturing accident in a TCP plant in Ludwigshafer, Germany, in 1953. A followup study is in progress. Two studies show an increased incidence of malignant mesenchymal softtissue tumors in persons exposed to phenoxy acids or chlorophenols (Harden and Sandstrom 1978; Hardell 1979). In the 1978 study, 52 patients with soft-tissue sarcomas and 205 matched controls were investigated in a cohort study. The incidence of exposure was 19/52 among the tumor patients and 19/206 in the tumor-free controls (p <0.001). Relative risks were determined to be 5.3 for exposure to phenoxy acid and 6.6 for exposure to chlorophenols. In the 1979 study, Hardell prospectively studied patients with histocytic, malignant lymphoma. In the first phase of the study, 14 of 17 patients reported occupations consistent with the possibility of exposure to the chemicals under study, and 11 patients reported definite exposure to phenoxy acetic acids or chlorophenols. The median latent period between exposure and tumor detection in this group was 15 years. Rappe (1979) has reported an increased incidence of primary liver cancer in members of the Vietnamese population exposed to Herbicide Orange. Mutagenicity Chromosomal analyses in Seveso have shown an increase in chromosomal lesions in males and females aged 2 to 28 years. These lesions consist of chromosomal gaps, and chromatid and chromosomal breaks and rearrangements. Cytogenetic studies indicate chromosomal damage to cells in maternal peripheral blood and in placental and fetal tissues studied following therapeutic abortions (U.S. EPA 1978h). In similar analyses, Tenchini et al. (1977) found a higher number of structural aberrations in the fetal tissues than in the maternal blood samples of fibroblast cells from adult tissues, but the frequency of these aberrations did not appear to be greater than expected to occur spontaneously in cultures of comparable cell types. Tenchini et al. point out that these preliminary findings do not indicate whether the higher frequencies of chromosome aberrations in fetal tissues were due to chromosome damage caused by 2,3,7,8-TCDD exposure. In contrast, the chromosomes of peripheral blood cells from 90 workers at the chemical plant at Seveso showed no abnormalities; the same results were obtained in a sampling of the most severely exposed residents of the area (Wassom 1978). 198 �Czeizel and Kiraly (1976) compared the frequency of chromosome aberrations in the peripheral lymphocytes of 76 workers employed at a herbicideproducing factory in Budapest with those of 33 controls. Among these workers, 36 were exposed to 2,4,5-trichlorophenoxyethanol (TCPE) or Klorinol and 26 to Buvinol. The remaining 14 workers had never been engaged in the production or use of either herbicide. The 2,3,7,8-TCDD concentration in the herbicide products is reported to be either less than 0.1 mg/kg or not more than 0.05 mg/kg. The frequency of chromatid-type and unstable chromosome aberrations was higher (p <0.01) in the factory workers than in the controls, regardless of involvement in production of the herbicide. Aberrations were more frequent in workers preparing TCPE and Buvinol than in other factory workers, but the difference was significant only for the chromatid-type effect. 199 �VOLUME I REFERENCES Ackerman, D. G., et al. 1978. At-Sea Incineration of Herbicide Orange Onboard the M/T Vulcanus. EPA-600/2-78-086. Ahling, B., and A. Lindskog. 1977. Formation of Polychlorinated Dibenzop-dioxins and Dibenzofurans During Combustion of a 2,4,5-T Formulation. Chemosphere, 6(8):461-468. Air/Water Pollution Report. 227. 1978. Around the States—Kansas. June 5, p. Aitio, A., M. Parkki, and J. Marniemi. 1979. Different Effect of TCDD on Glucuronide Conjugation of Various Aglycones: Studies in Wistar and Gunn Rats. Toxicol. Appl. Pharmacol., 47:55-60. Albro, P. W. 1978. Effects of TCDD on Lipid Profiles in Tissue of the Fischer Rat. Chpn-Biol. Interact., 23(3):315-30. Allen, J., M.D. August 28. 1978. University of Wisconsin, personal communication, Allen, J. R. 1967. Light and Electron Microscopic Observations in Macaca Mulatta Monkey Fed Toxic Fat. American Journal of Vet. Research, 28:151326. Allen, J. R., J. P. Van Miller, and D. H. Norback. 1975. Tissue Distribution, Excretion and Biological Effects of [14C]-Tetrachlorodibenzo-pdioxin in Rats. Fd Cosmet. Toxicol., 13:501-505. Allen, J. R., et al. 1975. Morphological Changes in Monkeys Consuming a Diet Containing Low Levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Fd Cosmet. Toxicol., 15:401-410. Allen, J. R., et al. 1979. Toxicological Responses of Nonhuman Primates to TCDD. Toxicol. Appl. Pharmacol., 48(1):A180. American Broadcasting Co. 1978. Transcripts of broadcasts aired on July 25 and August 1. 20/20. New York. American Industrial Hygiene Association Journal. 1980. No Excess in Cancer Deaths Found in Largest Group Ever Studied for Long-Term Effects of Dioxin Exposure. January. 200 �American Wood Preservers Institute. 1977. Memorandum to the Office of Pesticide Programs, U.S. Environmental Protection Agency. Pentachlorophenol: A Wood Preservative. Research Triangle Park, North Carolina. Anaizi, N. H., and J. Cohen. 1979. The Effects of TCDD on the Renal Tubular Secretion of Phenolsulfonphthalein. J. Pharmacol. Exp. Ther., 207(3):748-755. Anderson, J. 1978. The Poisoning of America. The Washington Post, United Feature Syndicate, Inc., July 15. Aniline, 0. 1973. Preparation of Chlorodibenzo-p-dioxins for Toxicological Evaluation. In: Chlorodioxins--0rigin and Fate, E. H. Blair, ed. American Chemical Society, Washington, D.C., pp. 126-135. Arsenault, R. D. 1976. Pentachlorophenol and Contained Chlorinated Dibenzodioxins in the Environment. American Wood Preservers Association. Auer, C. 1979. EPA Office of Toxic Substances, personal communication with D. R. Watkins, EPA, lERL-Cincinnati, February 26. American Wood Preservers Institute Directory. Design, No. 17, September-October. 1977. AE Concepts in Wood Bage, G. , E. Cekanova, and K. S. Larsson. 1973. Teratogenic and Embryotoxic Effects of the Herbicides Di- and Trichlorophenoxyacetic Acids (2,4-D and 2,4,5-T). Acta Pharmacol. et Toxicol., 32:408-416. Bail in, L. J. 1977. Microwave Plasma Detoxification Process for Hazardous Wastes. EPA, MERL-Cincinnati, EPA Contract No. 68-03-2190, October. Bailin, L. J. 1978a. Smithsonian Science Information Exchange Notice of Research Project Microwave Regeneration of Activated Carbon. EPA Contract No. 68-03-2660, March 1. Bailin, L. J. 1978b. Summary Report: Detoxification of Navy Red Dye by Microwave Plasma. EPA-600/2-78-081, June. Bailin, L. J., and B. L. Hertzler. 1976. Development of Microwave Plasma Detoxification Process for Hazardous Wastes: Phase I, EPA Contract No. 68-03-2190. Bailin, L. J., and E. L. Littauer. 1978. Microwave Regeneration of Activated Carbon. EPA, lERL-Cincinnati, Contract No. 68-03-2660, March. Ball, L. M. , and R. S. Chhabra. 1977. Effects of Environmental Pollutants on the Intestinal Absorption of Nutrients. Environmental Health Perspectives, 20:231. Barbero, R. D. 1978. Response Report: Dioxins in Sludges From the Manufacture of: Trichlorophenol, Pentachlorophenol, and Hexachlorophene. CBC •Associates, Inc., Falls Church, Virginia. Unpublished. 201 �Barsotti, D. A., L. J. Abrahamson, and J. R. Allen. 1979. Hormonal Alterations in Female Rhesus Monkeys Fed A Diet Containing TCDD. Bulletin of Environmental Contamination and Toxicology, 21:463-469. Bartleson, F. D., Jr., D. D. Harrison, and J. B. Morgan. 1975. Field Studies of Wildlife Exposed to TCDD Contaminated Soils. Air Force Armament Lab. Eglin A.F. Base, Florida. Baughman, R. , and M. Meselson. 1973. An Analytical Method for Detecting TCDD (Dioxin): Levels of TCDD in Samples From Vietnam. Environmental Health Perspectives, 5:27-35, September. Baughman, R. W. 1974. Ph.D. Thesis, Harvard University, Boston. Beale, M. G. , et al. 1(8014):788, April. 1977. Long-Term Effects of Dioxin Exposure. Lancet, Beatty, P. 1977. Studies of Metabolism and Possible Mechanisms of Toxicity of TCDD, Thesis, Vanderbilt University, Nashville, Tennessee. Beatty, P. W. , and R. A. Neal. 1976. Induction of DT-diaphorase Activity of Rat Liver by 2,3,7,8-TCDD (Abstract No. 232). Toxicol. Appl. Pharmacol., 37:189. Beatty, P. W. , et al. 1978. Effect of Alteration of Rat Hepatic Mixed Function Oxidose (MFO) Activity On the Toxicity of 2,3,7,8-TCDD. Toxicol. Appl. Pharmacol., 45(2):513-20. Becker, D. 1973. The Effect of Folate Overdose and of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDBD) on Kidney and Liver Respectively of Rat and Mouse Embryos. Teratology, 8:215. Benedetto, A. V. , and J. S. Taylor. 1978. Cutis, 21:983-88, April. 1978. Porphyria Cutanea Tarda: Update Berry, D. L. , et al. 1978. Lack of Tumor-Promoting Ability of Certain Environmental Chemicals in A Two-Stage Mouse Skin Tumorigenesis Assay. Res. Comm. Chem. Path, and Pharm., 20(1): 101-108, April. Bertoni, G., et al. 1978. Gas Chromatographic Determination of 2,3,7,8TCDD in the Experimental Decontamination of Seveso Soil by Ultraviolet Radiation. Analytical Chemistry, 50(6):732-735. Bevenue, A., and H. Beckman. 1967. Pentachlorophenol: A Discussion of its Properties and its Occurrence as a Residue in Human and Animal Tissues. Residue Reviews, 19:83-129. Blair, G. H. 1979. The Safety of 2,4,5-T. Science, 206:1135-6, December. Blair, E. H., K. C. Kaner, and E. E. Kenaga. 1963. Synthesis and Insecticidal Activity of 0-methyl 0-(2,4,5-Trichlorophenyl) Phosphoramidothioates and Related Compounds. J. Agric. Food Chem., 11(3):237-240. 202 �Blum, B. 1979. U.S. EPA Deputy Administrator, Press Conference Statement: Emergency Action to Stop Spraying of the Herbicides 2,4,5-T and Silvex. Also U.S. EPA Environmental News release, March 1. Boer, F. P., et al. 1971. X-ray Diffraction Studies of Chlorinated Dibenzo-p-dioxins. In: Chlorodioxins--0rigin and Fate, E. Blair, ed. American Chem. Society, Washington, D.C., pp. 14-15. Bolton, L. 1978. Seveso Dioxin: 85(22):78, October 9. No Solution in Sight. Bonaccorsi, A., R. Fanelli, and G. Tognoni. Ambio, 7(5-6):234-239. Chem. Eng., 1978. In the Wake of Seveso. Boobis, A. R. 1979. Effects of Microsomal System Inducers in Vivo and Inhibitors in Vitro on the Covalent Binding of Benzo(a) pyrene Metabolites to DNA Catalyzed by Liver Microsomes from Genetically Responsive and Nonresponsive Mice. Biochemical Pharmacology, 28(1):111-122. Botre, C., A. Memoli, and F. Alhaique. 1978. TCDD Solubilization and Photodecomposition in Aqueous Solutions. Environmental Science and Technology, 12(3):335-336. Botre, C., A. Memoli, and F. Alhaique. 1979. On the Degradation of 2,3,7,8-tetrachlorodibenzo-para dioxin (TCDD) by Means of a New Class of Chloriodides. Environmental Science and Technology, 13(2):228-231. Bowes, G. W., et al. 1973. The Search for Chlorinated Dibenzofurans and Chlorinated Dibenzodioxins in Wildlife Populations Showing Elevated Levels of Embryonic Death. Environmental Health Perspectives, 5:191-8, September. Bradlaw, J. A. 1979. Induction of Enzyme Activity in Cell Culture: A Rapid Screen for Detection of Planar Polychlorinated Organic Compounds. J. Assoc. Offic. Anal. Chem., 62:904-916. Brugger, J. E. 1978. Use of Mobile Incinerator to Dispose of p-Dioxin. Written communication to K. Q. Camin, Regional Administrator, EPA Region VII, from U.S. EPA, OHMSB, lERL-Ci, June 16. Buser, H. R. 1975. Polychlorinated Dibenzo-p-dioxin: Separation and Identification of Isomers by Gas Chromatography--Mass Spectrometry. J. Chromatography, 114:95-108. Buser, H. R. 1976. Preparation of Qualitative Standard Mixtures of Polychlorinated Dibenzo-p-Dioxins and Dibenzofurans by Ultraviolet and yIrradiation of the Octachloro Compounds. J. Chromatography, 129:303-7. Buser, H. R. 1978. Polychlorinated Dibenzo-p-dioxins and Dibenzofurans: Formation, Occurrence and Analysis of Environmentally Hazardous Compounds. Department of Organic Chemistry, University of Umea, Sweden, and Swiss Federal Research Station, Waedenswil, Switzerland, pp. 9-21. 203 �Buser, H. R. 1979a. Formation and Identification of Tetra-and Pentachlorodibenzo-p-dioxins from Photolysis of Two Isomeric Hexachlorodibenzo-pdioxins. Chemosphere, 4:251-257. Buser, H. R. 1979b. Formation of Polychlorinated Dibenzofurans (PCDF's) and Dibenzo-p-dioxins (PCDD's) from the Pyrolysis of Chlorobenzenes. Chemosphere, 6:415-424. Buser, H. R., and H. P. Bosshardt. 1974. Determination of 2,3,7,8-Tetrachlorodibenzo 1,4 dioxin at Parts Per Billion Levels in Technical Grade 2,4,5-Trichlorophenoxyacetic Acid, in 2,4,5-T Alkyl Ester and 2,4,5-T Amine Salt Herbicide Formulations by Quadrupole Mass Fragmentography. J. Chromatography, 90:71-77. Buser, H. R., and H. P. Bosshardt. 1976. Determination of Polychlorinated Dibenzo-p-dioxins and Dibenzofurans in Commercial Pentachlorophenols by Combined GC-MS. Journal of the AOAC, 59(3):5$2-569. Buser, H. R., and H. P. Bosshardt. 1978. Polychlorinated Dibenzo-pdioxins, Dibenzofurans and Benzenes in the Fly Ash of Municipal and Industrial Incinerators. Mitt. Geb. Lebensmittelsunters Hyg., 69(2):191-199. Buser, H. R., and C. Rappe. 1978. Identification of Substitution Patterns In Polychlorinated Dibenzo-p-Dioxins (PCDD's) by Mass Spectrometry. Chemosphere, 7(2):199-211. Buser, H. R., H. P. Bosshardt, and C. Rappe. 1978. Identification of Polychlorinated Dibenzo-p-Dioxin-Isomers Formed in Fly Ash. Chemosphere, 7(2): 165-172. Buu-Hoi, N. P., et al. 1971. Properties Canceramimetiques de la Tetrachloro 2,3,7,8 Dibenzo-p-dioxins. C.R. Acad, Sci. (Paris), 272:1447-1450. Buu-Hoi, N. P., et al. 1972. Organs as Targets of "Dioxin" (2,3,7,8-TCDD) Intoxication. Naturwissenschaften, 59:174. Campbell, K. 1979. Effect of Exposure to Chemical Contaminants on Susceptibility to Infection By Pathogenic Organisms, Tox. Research Proj. Directory, 04:05. Carlstedt-Duke, J. M. 1979. B. Tissue Distribution of the Receptor for TCDD in the Rat. Cancer Research, Issue 8, 39:3172-3176. Carreri, V. 1978. Review of the Events Which Occurred in Seveso. In: Dioxin Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds. SP Medical and Scientific Books. Spectrum Publications, Inc., New York, London, pp. 1-4. Carter, C. D., et al. 1975. Tetrachlorodibenzodioxin: An Accidental Poisoning Episode in Horse Arenas. Science, 188(4189): 738-740. 204 �Cassito, L., and P. Magni. 1977. Method for Emission Control at the Proposed Seveso Incinerating Plant Using Gas Chromatography and Mass Fragmentography. Inquinamento, 19(5):33-38. Cattabeni, F. , et al. 1978. DIOXIN: Toxicological and Chemical Aspects. SP Medical and Scientific Books, New York. Cavolloni, L., and L. Zecca. 1977. La Decomposizione Del TCDD Mediante Ozono. Medicina Termale e Climatologia, 34:73-74. CEFIC. 1978. Biodegradation Status Report. Safety of Chemicals Committee, Biodegradation Task Force, Brussels, February, p. 9. Chem. and Eng. News. 52(9):6-7, March 4. 1974a. Chem. and Eng. News. 1974b. 52(10):18-19, March 11. Chem and Eng. News. 7, p. 6. 1978. Chemical Engineering. 1978. September 11, p. 107. Vietnam Foliage Hit Hard by Herbicides. War Herbicide Report Stirs Controversy. Old Landfill Site Poses Health Problems. August British Plant Shut Because of Health Risk. Chemical Regulation Reporter. 1978. Pentachlorophenol: Study Group Advises Using Controls to Reduce Dioxin Impurity Levels. Vol. 2, No. 1, April 7, pp. 6-7. Chemical Regulation Reporter. 1979a. Reports on Bioassays of Three Chemicals Available From National Cancer Institute. March 2, p. 216. Chemical Regulation Reporter. 1979b. Hooker Chemicals Installing Toxic Controls at Montague, Michigan, Site. June 22, pp. 457-458. Chemical Regulation Reporter. 1979c. Ban On All Remaining Uses of 2,4,5-T, Silvex to be Considered at Hearings. July 13, p. 529-30. Chemical Regulation Reporter. 1980a. NIOSH Finds Dioxin, Carcinogens in Plant Dust Samples Near Landfill. January 11, p. 1593. Chemical Regulation Reporter 1980b. Dioxins: EPA Finds No Detectable Levels of TCDD in Mother's Milk Samples. January 18, p. 1620. Chemical Week. 1977. Detoxifying Dioxin. November 23, p. 17. Chemical Week. 1978a. Dioxin Lingers On. April 12. Chemical Week. 1978b. Dioxin Down Under. May 10. 205 �Chemical Week. 1978c. Checking Dioxin. May 17. Chemical Week. 1978d. 2,4,5-T Under Fire. June 14. Chemical Week. 1978e. Scrubdown for Dioxin. June 21. Chemical Week. January 3. 1979a. Hooker Dumpsites May Pose Dioxin Threat. 124(1):16, Chemical Week. February 28. 1979b. Faulty Reporting of Seveso Birth Defects. 124(9), Chemical Week. 1979c. More Agent Orange Suits Filed in Chicago; Still Others Will Follow. 124(9), February 28. Chemical Week. Chemical Week. 1979d. Still More Hassles For Hooker. April 25, p. 23. 1979e. Dioxin is Found in Cleanup Worker's Blood. May 23. Chemical Week. 1979f. Discovery of Dioxin At Montague Could Impede Accord on Hooker Wastes. June 6, p. 18. Chemical Week. July 11. Chemical Week. Chemical Week. 25. 1979g. Global Marine Will Burn Chemical Wastes at Sea. 1979h. Seveso Cleanup Still Not Solved. July 18. 1979i. Chem. Sources-U.S.A. New Jersey. Suits Boost Claims for Dioxin-Spill Damage. 1975. July Directories Publishing Co., Inc., Flemington, Chulkov, Y. I., V. Parini, and B. Staroselets. 1937. The Action of Chlorine on Phenol in Alkaline Solution and a Possible Method for Preparation of Chloranil. Chemical Abstracts, 31:4967. Ciaccio, E. I. 1979. April 16, p. 3. Dioxin Contamination. Chemical and Engineering News, Cincinnati Enquirer. 1978a. Chemical Dumps Proving Hazardous. August 6. Cincinnati Enquirer. 1978b. 6. Residents Near Canal Fear Poisoning. August Cincinnati Enquirer. 1978c. Grass-Roots Protest Forming Over Spraying of Herbicide. August 20, p. J-6. Cincinnati Enquirer. 1979. Bayou Closed by Herbicide. August 23. 206 �Cleary, T. F. 1971. Production of 2,3-Methylenebis (3,4,6-Trichlorophenol). Centerchem., Inc., September 21. Cocucci, S., et al. 1979. Absorption and Translocation of Tetrachlorodibenzo-p-dioxine by Plants From Polluted Soil. Experientia, 35(4):482-484. Colledge, C. 1978. U.S. EPA Technical Services Div., personal communication, August 21. Collins, T. F. X., and C. H. Williams. 1971. Teratogenic Studies With 2,4,5-T and 2,4-D in the Hamster. Bulletin of Environmental Contamination and Toxicology, 6(6):559-567. Commoner, B. 1976. Reliability of Bacteria Mutagenesis Techniques to Distinguish Carcinogenic and Noncarcinogenic Chemicals, EPA-600/1-76-022, 1103. Commoner, B. pp. 31, 33. 1977. Seveso: The Tragedy Lingers On. Hospital Practice, Commoner, B., and R. E. Scott. 1976a. Accidental Contamination of Soil with Dioxin in Missouri: Effects and Countermeasures. Center for the Biology of Natural Systems. Washington University, St. Louis, Missouri. Commoner, B., and R. E. Scott. 1976b. U.S.A.F. Studies on the Stability and Ecological Effects of TCDD (Dioxin): An Evaluation Relative to the Accidental Dissemination of TCDD at Seveso, Italy. Center for the Biology of Natural Systems, Washington University, St. Louis, Missouri. Cookson, C. 1979. 'Emergency1 Ban on 2,4,5-T Herbicide in U.S. Nature 278:108-109, March 8. Cooper, J. J., and W. H. Dennis, Jr. 1978. Catalytic Dechlorination of Organochlorine Compounds IV: Mass Spectral Identification of DDT and Heptachlor Products. Chemosphere, 4:229-305. Cosmestic, Toiletry and Fragrance Association, Inc. 1977. CTFA Cosmetic Ingredient Dictionary, 2nd Ed., Washington, D.C. Courtney, K. D. 1970. 168:864-866, May. Teratogenic Evaluation of 2,4,5-T. Science, Courtney, K. D. 1976. Mouse Teratology Studies With Chlorodibenzo-pdioxins. Bulletin of Environmental Contamination and Toxicology, 16(6):674-681. Courtney, K. D., and J. A. Moore. 1971. Teratology Studies with 2,4,5Trichlorophenoxyacetic Acid and 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Toxicol. Appl. Pharmacol., 20:396-403. 207 �Courtney, K. D., M. T. Ebron, and A. W. Tucker. 1977. Distribution of 2,4,5-Trichlorophenoxyacetic Acid in the Mouse Fetus. Tox. Letters, 1:103-108. Courtney, K. D., et al. 1978. Metabolic Studies with TCDD (Oioxin) Treated Rats. Alch. Environ. Contain. Toxicol., 7:385-396. Cox, J. M., B. A. Wright, and W. W. Wright. 1965. Thermal Degradation of Poly (phenylene Oxides). J. Appl. Polymer Sci., 9:513-522. Craft, T. F., R. D. Kimbrough, and C. T. Brown. 1975. Radiation Treatment of High Strength Chlorinated Hydrocarbon Wastes, U.S. EPA-660/2-75-017. Creso, E., et al. 1978. Neuropsychopharmacological Effect of TCDD. Boll-Soc. Ital. Biol. Sper., 54:1512-1516. Crosby, D. G. 1969. Experimental Approaches to Pesticide Photodecomposition. Residue Reviews, 25:1-12. Crosby, D. G. 1978a. Conquering the Monster—The Photochemical Destruction of Chlorodioxins. In: Proceedings of American Chem. Soc. Symposium on Disposal and Decontamination of Pesticides, pp. 1-12. Crosby, D. G. 1978b. University of California, Davis, personal communication, August 21. Crosby, D. G., and A. S. Wong. 1973. Photochemical Generation of Chlorinated Dioxins. Chemosphere, 5(5):327-332. Crosby, D. G., and A. S. Wong. 1977. Environmental Degradation of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). Science, 195: 1337-1338, March 25. Crosby, D. G., K. W. Moilanen, and A. S. Wong. 1973. Environmental Generation and Degradation of Dibenzodioxins and Dibenzofurans. Environmental Health Perspectives, 5:259-265, September. Crosby, D. G., et al. 1971. Photodecomposition of Chlorinated Dibenzo-pDioxins. Science, 73:748-749, August 20. Crossland, J., and K. Shea. 15(5):35-38, June. 1973. The Hazard of Impurities. Environment, Crow, K. D. 1977. Effects of Dioxin Exposure. 9. Lancet, 8028: 82-83, July Crow, K. D. 1978. Chloracne - An Up to Date Assessment. 21(3):297-8. Ann. Occup. Hyg., Crummett, W. B., and R. H. Stehl. 1973. Determination of Chlorinated Dibenzo-p-dioxins and Dibenzofurans in Various Materials. Environmental Health Perspectives. 5:15-25, September. 208 �Czeizel, E. , and J. Kiraly. 1976. Chromosome Exam in Workers Producing Klounal and Buvinol in the Development of a Pesticide as a Complex Scientific Task, L. Banki, ed. Medicina, Budapest, pp. 239-256. Davring, L. , and M. Summer. 1971. Cytogenetic Effects of 2,4,5-Trichlorophenoxyacetic Acid oh Oogenesis and Early Embryo-genesis in Drosophila Melanogaster. Hereditas, 68:115-122. Deinaer, M. , et al. 1979. Isolation and Identification of Suspected Toxicants in Pentachlorophenol. Presented Before the Division of Environmental Chemistry, American Chemical Society, Honolulu, Hawaii, April. Dennis, W. H. , Jr. 1972. Methods of Chemical Degradation of Pesticides and Herbicides—A Review. NTIS AD-752 123. Dennis, W. H. , Jr., and W. J. Cooper. 1975. Catalytic Dechlorination of Organochlorine Compounds. I. DDT. Bulletin of Environmental Contamination and Toxicology, 14(6):738-744. Dennis, W. H. , Jr., and W. J. Cooper. 1976. II. Heptachlor and Chlordane. Bulletin of Environmental Contamination and Toxicology, 16(4):425-430. Dennis, W. H. , Jr., and W. J. Cooper. 1977. III. Lindane. Environmental Contamination and Toxicology. 18(1):57-59. Bulletin of Diamond Alkalai Company. 1952. Improvements in or Relating to the Preparation of Chlor-substituted Ethylene Derivatives. The Patent Office. Patent No. 673,565, London, June 23. Dickson, J. P. 1978. Written communication to D. Watkins, U.S. EPA, IERLCincinnati, from State Pollution Control Commission, Sydney, Australia. DiGiovanni, J. , et al. 1977. (TCDD) and Arochlor 1254 in the Two-Stage System of Mouse Skin Carcinogenesis. Bulletin of Environmental Contamination and Toxicology, 18(5):552-57. DiGiovanni, J. , et al. 1979. 2,3,7,8-TCDD: Potent Anticarcinogenic Activity in CD-I Mice. Biochem. and Biophys. Res. Comm., 86(3):577-84, February. Doedens, J. D. pp. 325-338. 1964. Chlorophenols. Kirk-Othmer, Second edition. Vol. 5, Dombrowski, D. , ed. 1978. Silver Springs, Maryland. Toxic Materials News, 5(33):236, August 16, Dorman, L. F. 1978. President of American Wood Preservers Inst., personal communication, August 22, and September 5. Dow Chemical Company. 1978. The Trace Chemistries of Fire--A Source of and Routes for the Entry of Chlorinated Dioxins Into the Environment. The Chlorinated Dioxin Task Force, the Michigan Division, Dow Chemical, U.S.A. 209 �Duqois, P., et al. 1967. Acne, Chlorique Collective et Accidentelle d'un Type Nouveau. Bull Soc. Franc. Derm. Syph., 75, 260-261. Duvall, D. S. , and W. A. Rubey. 1976. Laboratory Evaluation of High Temperature Destruction of Kepone and Related Pesticides. Tech. Rep. UDRI-TR-76-21. University of Dayton Research Institute. EPA-600/2-76-299. Duvall, D. S. , W. A. Rubey, and J. A. Mescher. 1978. Laboratory Characterization of the Thermal Decomposition of Hazardous Wastes. In: Proceedings of the Fourth Annual Research Symposium—Land Disposal of Hazardous Wastes. EPA 600/9-78-016. Eaton, D. L. , 1979b. Isolated Rat Hepatic 39(11):5337. Characteristics of Hepatic Transport Systems in Parenchynal Cells. Diss. Abstract. Int. B. , Eaton, D. L. 1979a. Alterations in Hepatic Transport Systems in Isolated Rat Hepatocytes after Treatment with Microsomal Enzyme Induced. Toxicol. Appl. Pharmacol., 48(1):A187. Elovaara, E., et al. 1977. Neurochemical Effects of 2,3,7,8-TCDD in Wistor and Gunn Rats. Res. Commun. Chem. Pathol. Pharmacol., 18(3):487-94. Elvidge, D. H. 1971. The Gas-Chromatographic Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in 2,4,5-Trichlorophenoxyacetic Acid ("2,4,5-T"), 2,4,5-T Esters and 2,4,5-Trichlorophenol. Analyst, 96:721-727. Ensign, T. , and M. Uhi. 1978. Dioxin: Uncovered by Accident, The Number of Cases is Rising Dramatically. In These Times, July 19-25, p. 16. Environment Reporter. 1978. EPA to Investigate Possible Link Between Herbicide Use, Miscarriages. 9(16):670-671, August 18. Environment Reporter. 1979a. EPA Cites Miscarriage Correlation in Imposing Emergency Ban on 2,4,5-T. March 9, p. 2074-S. Environment Reporter. 1979b. Arkansas - Administrative Order Requires Cleanup of Dioxin Wastes At Vertac, Inc., Site. June 22. Epstein, S. S. 1970. A Family Likeness. Environment, 12(6):16-25. Erk, S. D. , M. L. Taylor, and T. 0. Tiernan. 1979. Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Residues on Metal Surfaces by GC-MS. Chemosphere 1:7-14. Ernst and Ernst. 1977. Proceedings, pp. 181-217. Fadiman, A. 49. 1979. Wood Preservation Statistics A Poisoned Town. 1976. In: AWPA Life Magazine, September, pp. 43-46, Faith, R. E., and M. I. Luster. 1977. Modulation of Immune Function by Chemicals of Environmental Concern. Environmental Health Perspectives, 20:245. 210 �Fanelli, R., et al. 1978. Degradation of 2,3,7,8-Tetrachlorodibenzo-pdioxin in Organic Solvents by Gamma Ray' Irradiation. Experientia, 34(9):1126-27, September 9. Fanelli, R. , et al. 1979. Studies of TCDD Levels in Animals Living in the Seveso Area. Presented in Rome at the Meeting of the Expert Committee of the National Academy of Sciences, March 5-6. Fanelli, R., et al. 1980a. Presence of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Wildlife Living Near Seveso, Italy: A Preliminary Study. Institute de Ricerche Farmacologiche "Mario Negri," Milan, Italy. Prepublication copy. Fanelli, R., et al. 1980b. 2,3,7,8-TCDD Levels in Cow's Milk From the Contaminated Area of Seveso. Mario Negri Institute for Pharmacological Research, Milan, Italy. Prepublication copy. Fara, G. M. 1977. Seveso--Studies in Teratogenic and Other Chronic Effects of Chemical Pollutants Following an Accident in a Chemical Plant. Teratology, 16:365. Farbenind, I. G. 1934. Diary! Sulfides. Chemical Abstracts, 28:179. Federal Working Group on Pest Management. Pesticides. Washington, D.C. 1974. Occupational Exposure to Ferguson, T. L., et al. 1975. Determination of Incinerator Operating Conditions Necessary for Safe Disposal of Pesticides. EPA-600/2-75-041. Field, B., and C. Kerr. 1979. Herbicide Use and Incidence of Neural Tube Defects. Lancet, 1(8130):1341-1342. Firestone, D. 1973. Etiology of Chick Edema Disease. Perspectives, 5:59-66. Environmental Health Firestone, D. 1977. The Determination of Polychlorodibenzo-p-dioxin and Polychlorodibenzofurans in Commercial Gelatins by Gas-Liquid Chromatography. J. Agric. Food Chem., 25(6):1275-1280. Firestone, D. 1978. The 2,3,7,8-Tetrachlorodibenzo-para-dioxin Problem: A Review. In: Chlorinated Phenoxy Acids and their Dioxins: Mode of Action, Health Risks, and Environmental Effects, C. Ramel, ed. Ecol. Bull. (Stockholm). Firestone, D., et al. 1972. Determination of Polychloro-dibenzo-p-dioxins and Related Compounds in Commercial Chlorophenols. Journal of the AOAC, 55(l):85-92. Fishbein, L. 1978. Overview of Potential Mutagenic Problems Posed by Some Pesticides and their Trace Impurities. Environmental Health Perspectives, 27:125-131. 211 �Fowler, B. A., G. E. R. Hood, and G. W, Lucier. 1975. Tetrachlorodibenzop-dioxin Induction of Renal Microsomal Enzyme Systems. Toxicol. Appl. Pharmacol., 33:176-177. Fox, J. L. 1979. Research Solving Body's Detoxifying System. Eng. News, 24-25, June 6. Chem. and Frigerio, A. 1978. The Seveso Case: TCDD in Animals from Contaminated Areas. Methodol. Surv. Biochem., 7:161-6. Fuller, J. G. 1977. The Poison That Fell From the Sky. Readers Digest, August, pp. 192-236. Galet, A. 1952. The Preparation of 2,4,5-Trichlorophenoxyacetic Acid (2,4,5-T). J. Am. Chem. Soc., 74:3890. Galston, A. W. 1975. The Ungreening of South Vietnam. 83(6):10,12,14. Galston, A. W. 1979. 29(2):85-89, February. Herbicides: Natural History, A Mixed Blessing. Bioscience, Gasiewicz, T. A., and R. A. Neal. 1978. Tissue Distribution and Excretion of 2,3,7,8-TCDD and Effects Upon Clinical Chemical Parameters In the Guinea Pig. Fed. Proc. Fed. Am. Soc. Exp. Biol., 37(3):501. Gebfuigi, I., R. Baumann, and F. Korte. 1977. Photochemical Degradation of 2,3,7,8-TCDD Under Simulated Environmental Conditions. Naturwissenschaften, 64(9):486-7. General Accounting Office. 1979. Health Effects of Exposure to Herbicide Orange in South Vietnam Should be Resolved. Report by the Comptroller General, CED-79-22, April 6. Getzendaner, M. D., N. H. Mahle, and H. S. Higgins. 1977. Search for the Presence of 2,3,7,8-Jetrachlorodibenzo^p-dioxin in Bovine Milk. Bulletin of Environmental Contamination and Toxicology, 18:2. Gilman and Dietrich. 79:1439. 1957. Halogen Derivatives of Dibenzo-p-dioxin. Goldmann, P. J. 1972. Extremely Severe Acute Chloracne Due to Trichlorophenol Decomposition Products. Industrial Medicine, Social Medicine, Industrial Hygiene, 7(1):12-18, January. Goldmann, P. J. 1973. Schwerste Acute Chloracne, Eine Massenintoxikation Durch 2,3,6,7-Tetrachlordibenzo-dioxin. From Int. Agency for Research on Cancer. Der Hautarzt, 24:149-152. Goldstein, J. A. 1979. The Structure-Activity Relationships of Halogenated Biphenyls as Enzyme Inducers. Ann. New York Acad. Sci., 320:164-178. 212 �Goldstein, J. A., et al. 1978. Effects of Pentachlprophenol on Hepatic Drug Metabolism and Pprphyria Related to Contamination with Chlorinated Dibenzo-p-Dioxins. Environmental Protection Agency, Research Triangle Park, North Carolina. Goldston, A. W. 1978. 29(2):85-89, February. Herbicides: A Mixed Blessing. Bio-Science, Goodell, (Senator). 1970. Authorization of Appropriations for Military Procurement During Fiscal Year 1971-Amendment. Congressional Record-Senate, July 16, pp. 24661-24670. Grant, W. F. 65(2):83-119. 1979. The Genotoxic Effects of 2,4,5-T. Mutation Res., Gray, A. P., et al. 1976. Synthesis of Specific Polychlorinated Dibenzop-dioxins. J. Org. Chem., 41(14).-2435-2437. Green, S., and F. S. Moreland. 1975. Cytogenetic Evaluation of Several Dioxins in the Rat (Abstract No. 99). Toxicol. Appl. Pharmacol., 33:161. Green, S., F. S. Moreland, and C. Sheu. 1977. Cytogenetic 2,3,7,8-TCDD on Rat Bone Marrow Cells. FDA By Lines, 6:292-294. Effects of Greig, J. B., and G. Osborne. 1978. Changes in Rat Hepatic Cell Membranes During 2,3,7,8-Tetrachlorodibenzo-p-dioxin Intoxication. In: Dioxin-Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds., SP Medical and Scientific Books, New York, London. Greig, J. B. , D. M. Taylor, and J. D. Jones. 1974. Effects of TCDD on Stimulated DMA Synthesis in the Liver and Kidney of the Rat. Chem. Biol. Interactions, 8:31-39. Greig, J. B., et al. Toxicol., 11:585-595. Gribble, W. G. February. 1974. 1973. Toxic Effects of 2,3,7,8-TCDD. TCDD, a Deadly Molecule. Fd. Cosmet. Chemistry, 47(2):15-18, Guenthner, T. M., and D. W. Nebert. 1977. Cytosolic Receptor for Aryl Hydrocarbon Hydroxylase Induction by Polycyclic Aromatic Compounds. Biol. Chem., 24:8981-89. Guenthner, T. M.., et al. 1979a. Evidence in Rat and Mouse Liver for Temporal Control of Two Forms of Cytochrome P-450 Inducible by TCDD. Environ. J. Biochem., 91(2):449-456. Guenthner, T. M., et al. 1979b. Microsomal Aryl Hydrocarbon Hydroxylase in Rat Adrenal: Regulation by ACTH but not by Polycyclic Hydrocarbons. Mol. Pharmacol., 15:719-728. 213 �Gupta, B. N., et al. 1973. Pathologic Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Laboratory Animals, Environmental Health Perspectives, 5:125-140. Gustafsson, J. A., and M. Ingelman-Sundberg. 1979. Changes in Steroid Hormone Metabolism in Rat Liver Microsomes Following Administration of TCDD. Biochemical Pharmacology, 28:497-499. Gustafsson, J. , et al. 1979. Mechanisms Involved in Tissue Responsiveness to Chemical Carcinogens and in Formation of Carcinogenic Metabolites from Aromatic Hydrocarbons. Toxicology Research Projects Directory, 4:9. Hanna, R. M., and D. R. Goldberg, n.d. Dioxin Monitoring Project: Final Report, Region II c/o Rutgers University, U.S. EPA Files. Unpublished. Hardell, L. 1979. Malignant Lymphoma of Histiocytic Type and Exposure to Phenoxyacetic Acids or Chlorophenols. Lancet, 55-56, January 6. Hardell, L., and A. Sandstrom. 1978. Malignant Mesenchymal .Soft-Tissue Tumors and Exposure to Phenoxy Acid or Chlorophenols. A Case Control Study. Lakartidringen, 75(49):3535-36. Hardell, L., et al. 1974. Case-Control Study of Soft-Tissue Sarcomas and Exposure to Phenoxyacetic Acids of Chlorophenols. Br. J. Cancer, 39(6):711-717. Harless, R. L. 1976. Presentation Given at TCDD Workshop, Universita Di Milano, Institute Di Framacognosia. U.S. EPA Files. Unpublished. Harless, R. L. 1980. Direct Testimony before the Administrator, U.S. EPA, Re: FIFRA Docket Nos. 415 et al. EPA Exhibit No. 212. April 23. Harris, M. W., et al. 1973. General Biological Effects of TCDD in Lab Animals. Environmental Health Perspectives, 5:101-109, September. Harrison, D. D., C. I. Miller, and R. C. Crews. 1979. Residual Levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Near Herbicide Storage and Loading Areas at Eglin AFB, Florida. AFATL-TR-79-20. February. Harrison, W. S., A. T. Peters, and F. M. Rowe. 1943. Polyhalogeno-oanisidines and Their Derivatives. J. Am. Chem. Soc., pp. 235-237. Haskelberg, L. 1947. The Halogenation of Aryloxyacetic Acids and Their Homologs. J. Org. Chem., 12:426-433. Hatfield, M. 0., (Senator). 1977. Written Communication to D. Costle, U.S. EPA Administrator, November 3. Hawley, G. 1971. Condensed Chemical Dictionary. Reinhold Co., New York. 214 8th ed., Van-Nostrand �Hay, A. 1976a. Toxic Cloud Over Seveso. Nature, 262:636-638, August 19. Hay, A. 18. 1976b. Seveso, The Problems Deepen. Nature, 264:309-310, November Hay, A. 1977a. Seveso Solitude. Nature, 267:384-385, June 2. Hay, A. 1977b. Hay, A. 1978a. February 16. Identifying Carcinogens. Vietnam's Dioxin Problem. Hay, A. 1978b. Company's Claim Sparks Nature, 247:108, July. Nature, 269:468-470, October 6. Nature, 271:597-598, London, Fresh Controversy Over Seveso. Helling, C. S., et al. 1973. Chlorodioxins in Pesticides, Soils, and Plants. J. Environ. Quality, 2(2):171-178. Hepner, G. W. 1979. Detection of Carcinogen-Induced Stimulation of Cytochrome P-448-Associated Enzymes by 14C02 Breath Analysis Studies Using Dimethyl aminoazobenzene. Gastroenterology, 76(2):267-271. Higgenbotham, G. R. , et al. 1968. Chemical and Toxicological Evaluations of Isolated and Synthetic Chloro. Derivatives of Dibenzo-p-dioxin. Nature, 220:702-703. Highman, B. , T. B. Gaines, and J. J. Schumacher. Renal Alkaline Phosphatase Activity in Fetal Offspring of Maternal Mice Given 2,4,5-Trichlorophenoxyacetic Acid. University of Arkansas College of Medicine, Little Rock, Arkansas; and the National Center for Toxicological Research, Jefferson, Arkansas. Highman, B. , et al. 1977. Retarded Development of Fetal Renal Alkaline Phosphatase in Mice Given 2,4,5-T. J. Toxicol. Environ. Health, 2:1007-1018. Homberger, E. , et al. 1979. The Seveso Accident: Its Nature, Extent and Consequence. Givauden Research Company, Ltd. and F. Hoffman-LaRoche and Co., Ltd. Confidential. Hook, G. E. R. , J. K. Haseman, and G. W. Lucier. 1975. Induction and Suppression of Hepatic and Extrahepatic Microsomal Foreign-CompoundMetabolizing Enzyme Systems by 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Chem. Biol. Interactions, 10:199-214. Hook, G. E. R., et al. 1975. 2,3,7,8-TCDD Induced Changes in the Hydroxylation of Biphenyl by Rat Liver Microsomes. Biochemical Pharmacology, 24:335-340. Hook, J. B. , et al. Res., 12:381-8. 1978. Renal Effects of 2,3,7,8-TCDD. 215 Environ. Sci. �Huff, J. 'E. , and J. S. Wassom. 1974. Health Hazards From Chemical Impurities; Chlorinated Dibenzodioxins and Chlorinated Dibenzofurans. Intern. J. Environ. Studies, 6:13-17. Huetter, R. 1980. Written Communication to D. R. Watkins, U.S. EPA, lERL-Cincinnati. January 9. Hummel, R. A. 1977. Cleanup Techniques for the Determination of Parts Per Trillion Residue Levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). J. Agric. Food Chem. , 25(5): 1049-1053. Hunter, W. H. , and L. M. Seyfried. 1921. A Catalytic Decomposition of Certain Phenol Silver Salts. V. The Action of Iodine on the Sodium Salt of Trichlorophenol. J. Am. Chem. Soc. , 43:151-159. Hussain, S. L. , et al . 1972. Mutagenic Effects of TCDD on Bacterial Systems. Ambio. , 1:32-33. In Brief. 1977. A Review of Water and Pollution Control News. Poison Zone Will Close Factories. July. Italian Ingersoll, B. 1979. Creek Closed By Superpoi son. Chicago Sun-Times. 2. June Innes, J. R. M. 1969. Bioassay of Pesticides and Industrial Chemicals for Tumorisenicity in Mice: A Preliminary Note. J. National Cancer Institute, Isensee, A. R. 27:255-62. 1978. Bioaccumulation of 2,3,7,8-TCDD. Ecol. Bull., Isensee, A. R. , and G. E. Jones. 1971. Absorption and Translocation of Root and Foliage Applied 2,4-Dichlorophenol , 2,7-Dichlorodibenzo-p-dioxin, and 2,3,7,8-Tetrachlorodibehzo-p-dioxin. J. Agric. Food Chem., 19(6):1210-1214. Isensee, A. R. , and G. E. Jones. 1975. Distribution of 2,3,7, 8-Tetrachlorodibenzo-p-dioxin (TCDD) in Aquatic Model Eco- system. Environmental Science and Technology, 9(7): 668-672. Jackson, W. T.1 1972. Regulation of Mitosis 3 Cytological Effects of 2,4,5-Trichlorophenoxyacetic Acid and of Dioxin Contaminants in 2,4,5-T Formulations. J. Cell Sci., 10:15-25. J. Am. Chem. Soc. 1960. Phenethicillin, B2:3934. Jansson, B. , G. Sundstrom, and B. Ahling. 1978. Formation of Polychlorinated Dibenzo-p-dioxins During Combustion of Chlorophenol Formulations. Sci. Total Environ., 10(3):209-218. Jensen, S. , and L. Renberg. 1972. Contaminants in Pentachlorophenol: Chlorinated Dioxins and Predioxins. Ambio., 1(2):62~65. 216 �Jensen, S., and L. Renberg. 1973. Chlorinated Dimers Present in Several Technical Chlorophenols Used as Fungicides, Environmental Health Perspectives, 5:37-41, September. Jirasek, L. , J. Kalensky, and K. Kubec. 1973. Acne Chlorina Porphyria Cutanea Tarda and Other Manifestatians^of General Intoxication During the Manufacture of Herbicides. (II. ) Ceskoslovenska Dermatologies, 48(5):306-317. Jirasek, L., et al. 1974. Acne Chlorina, Porphyria Cutanea Tarda, and Other Manifestations of General Intoxication During the Manufacture of Herbicides. [II. ) Ceskoslovenska Dermatologies, 49(3):145-147. Jirasek, L. , et al. 1976. Chloracne, Prophyria Cutanea Tarda, and Other Poisonings Due to the Herbicides. Hautarzt, 27(7):328-333. Johnson, J. E. 1971. Safety in the Development of Herbicides. Down to Earth, 27(l):l-7. Johnson, R. L. 1973. Chlorinated Dibenzodioxins and Pentachlorophenol. Environmental Health Perspectives, 5:171-175, September. Jones, G. 1975. A Histochemical Study of the Liver Lesion Induced by 2,3,7,8-TCDD (Dioxin) in Rats. J. Path., 116:101-105. Jones, K. G., and G. D. Sweeney. 1977. Pathogenesis of Porphyria Cutanea Tarda. Gastroenterology, 73(5):A-29. Kapetanovic, I. M. 1979. Metabolic Structure-Activity Relationships for a Homologous Series of Phenacetic Analogs. J. Pharmacol. Exp. Ther., 209:20-24. Kearney, P. C. , E. A. Woolson, and C. P. Ellington, Jr. 1972. Persistence and Metabolism of Chlorodioxins in Soils. Environmental Science and Technology, 6(12):1017-1019, November. Kearney, P. C., et al. 1973a. Environmental Significance of Chlorodioxins. In: Chlorodioxins-Origin and Fate, E. H. Blair, ed. Advances in Chemistry, Series 120, American Chemical Society, Washington, D.C., pp. 105-111. Kearney, P. C. , et al. 1973b. TCDD in the Environment: Sources, Fate, and Decontamination. Environmental Health Perspectives, 5:275-277, September. Kende, A. S. , and. J. J. Wade. 1973. Synthesis of New Steric and Electronic Analogs of 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Environmental Health Perspectives, 5:49-59. September. Kende, A. S. , et al. 1974. J. Org. Chem., 39:931. Kennedy, M. V., B. J. Stojanovic, and F. L. Shuman, Jr. 1969. Chemical and Thermal Methods for Disposal of Pesticides. Residue Reviews, 29:89. 217 �Khera, K. S. , and W. P. McKinley. 1972. Pre- and Postnatal Studies on 2,4,5-Trichlorophenoxyacetic Acid, 2,4-Dichlorophenoxyacetic Acid and Their Derivatives in Rats. Toxicol. Appl. Pharmacol., 22:14-28. Khera, K. S. , and J. A. Ruddick. 1973. Polychlorodibenzo-p-dioxins Perinatal Effects and the Dominant Lethal Test in Wistar Rats. In: Chlorodioxins-Origin and Fate, E. H. Blair, ed. Advances in Chemistry, Series 120, American Chemical Society, Washington, D.C., pp. 70-84. Kimble, B. J. , and M. L. Gross. 1980. Tetrachlorodibenzo-p-dioxin Quantitation in Stack-Collected Coal Fly Ash. Science, 207:59-61, January 4. Kimbrough, R. D. 1971. Review of the Toxicity of Hexachlorophene. Arch. Environ. Health, 23:119-122. Kimbrough, R. D. 1972. Toxicity of Chlorinated Hydrocarbons and Related Compounds. Arch. Environ. Health, 25:125-131. Kimbrough, R. D. 1974. The Toxicity of Polychlorinated Polycyclic Compounds and Related Chemicals. CRC Critical Reviews in Toxicology, 2(4):445-498. Kimbrough, R. D. 1976. Hexachlorophene: Toxicity and Use as an Antibacterial Agent. Essays in Toxicology, 7:99-120. Kimbrough, R. D. , et al. 1977. Epidemiology and Pathology of a Tetrachlorodibenzodioxin Poisoning Episode. Arch. Environ. Health, 32(2):77-86. Kimmig, J. , and K. H. Schulz. 1957. by Chlorinated Aromatic Cyclic Ethers. Occupational Acne (Chloracne) Caused Dermatologica, 115:540-546. Kirsch, R. et al. 1975. Structural and Functional Studies of Ligandin, a Major Renal Organic Anion-binding Protein. J. Clin. Invest. 55:1009-1019. Kitchin, K. T., and J. S. Woods. 1978. 2,3,7,8-TCDD Induction of Aryl Hydrocarbon Hydroxylase in Female Rat Liver. Evidence for De Novo Synthesis of Cytochrome P-448. Mol. Pharmacol., 19:890-99. Kitchin, K. T., et al. 1979. TCDD Effects on Hepatic Microsomal Cytochrome P-448-Mediated. Toxicol. Appl. Pharmacol., 47(3): 537-546. Klecka. G. M. , and D. T. Gibson. 1979. Metabolism of Dibenzo[l,4]dioxin by a Pseudomonas Species. Biochem. J., 180:639-645. Kluwe, W. M. 1978. Selective Modification of the Renal and Hepatic Toxicities of Chloroform by Induction of Drug-Metabolizing Enzyme Systems in Kidney and Liver. J. Pharmacol. Exp. Ther., 207(2):449-473. Kocher, C. W., et al. 1978. A Search for 2,3,7,8-Tetrachlorodibenzo-pDioxin in Beef Fat. Bulletin of Environmental Contamination and Toxicology, 19:229-236. Kociba, R. J. 1979. Environ. Sci. , 281:7. Toxicology Studies of TCDD in Rats. 218 Dev. Toxicol. �Kociba, R. J., et al. 1976. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)— Results of a 13-Week Oral Toxicity Study in Rats. Toxicol. Appl. Pharmacol., 35:553-574. Kociba, R. J. , et al. 1978. Results of a Two Year Chronic Toxicity and Oncogenicity Study of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Rats. Toxicol. Appl. Pharmacol., 46(2):279-304. Kouri, R. E. , et al. 1978. 2,3,7,8-Tetrachlorodibenzo-p-dioxin as Cocarcinogen Causing 3-Methylcholanthrene-initiated Subcutaneous Tumors in Mice Genetically "Nonresponsive" at Ah Locus. Cancer Research, 38:2777-2783, September. Kozak, V. P., et al. 1979. Reviews of the Environmental Effects of Pollutants: XI. Chlorophenols. ORNL/EIS-128, EPA-600/1-79-012, pp. 464-5. Kupfer, D. 1975. Effects of Pesticides and Related Compounds on Steroid Metabolism and Function. CRC Critical Reviews in Toxicology, 4(1):83-124, October. Lancet. 1976. Seveso. 2(7980):297, August 7. Lancet. 1977. Long-term Effects of Dioxin Exposure. 1:8014, April 2. Langer, H. G, , T. P. Brady, and P. R. Briggs. 1973. Formation of Dibenzodioxins and Other Condensation Products From Chlorinated Phenols, and Derivatives. Environmental Health Perspectives. 5:3-9, September. Langer, H. G. , et al. 1973. Thermal Chemistry of Chlorinated Phenols. In: Chlorodioxins-Origin and Fate, E. Blair, ed. Advances in Chemistry, Series 120, American Chemical Society, Washington, D.C., pp. 26-32. Laporte, J. R. July 9. 1977. Effects of Dioxin Exposure. Lancet, II 8028:82-83, Lawless, E. W., T. L. Ferguson, and A. F. Meiners. 1975. Guidelines for the Disposal of Small Quantities of Unused Pesticides. EPA-670/2-75-057. Lawrence Eagle Tribune. 1978. October 30, pp. A-l, A-16. Tot, 3, May Be Vietnam War Casualty. Leng, M. L. 1976. Comparative Metabolism of Phenoxy Herbicides in Animals. In: Fate of Pesticides in Large Animals, G. W. Ivie and H. W. Dorough, eds. Academic Press, New York. pp. 53-76. Leng, M. L. 1977. Are Lawn Weed Killers Really Hazardous. U.S.A. Report, August 29. Dow Chemical Liberti, A. 1978. Field Photodegradation of TCDD by Ultraviolet Radiations. In: Dioxin—Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds., SP Medical and Scientific Books, New York, London. 219 �Lira, P., M.D. 1978. Velsicol Corp., personal communication, August 29. Lucier, G. W. 1979. Developmental Toxicology of the Halogenated Aromatics: Effects on Enzyme Development. Ann. New York Acad. Sci., 320:449-457. Lucier, G. W., et al. 1973. TCDD-Induced Changes in Rat Liver Microsomal Enzymes. Environmental Health Perspectives, 5:199-209, September. Lucier, G. W., et al. 1979. Laboratory Studies on the Immune Effects of Halogenated Aromatics. Ann.New York Acad. Sci., 320:473-486. Lucier, G. W., 0. S. McDaniel, and G. E. R. Hook. 1975. Nature of the Enhancement of Hepatic Uridine Diphosphate Glucuronyl-transferase Activity by 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Rats. Biochem. Pharmacol. 24:325-334. Luster, M. I. 1979a. Effects of Brief In Vitro Exposure to TCDD on Mouse Lymphocytes. J. Environ. Pathol. Toxicol., 2(4):965-977. Luster, M. I. 1979b. Immunochemistry of Dioxin Action on the Lymphocyte. Tox. Research Proj. Directory, 04:06. Luster, M. I., et al. 1978. Inability of Passive Antibodies to Reverse the Effects of Dioxin Toxicity. Chemosphere, 7(l):29-34. Madge, D. S. 1977. Effects of Trichlorophen Oxyacetic Acid and Chlorodioxins on Small Intestinal Function. Gen. Pharmacol., 8(5-6):319-29. Madhukar, B. V., et al. 1979a. Comparison of Induction Patterns of Rat Microsomal Mixed-Function Oxidases by Pesticides and Related Chemicals. Pestic. Biochem. Physiol., 11:301-308. Madhukar, B. V., et al. 1979b. Depression of ATPase Activity in Hepatocyte Surface Membranes of Rats by 2,3,7,8-TCDD Treatment. Toxicol. Appl. Pharmacol., 48(1):A152. Mam's, J. 1977. Induction of Intestinal From Transport by 2,3,7,8-TCDD, an Environmental Pollutant and Potent Inducer of Aryl Hydrocarbon Hydroxylase. Clin. Res., 25(3):468A. Mam's, J. 1979. Intestinal Organic Anion Transport, Glutathione Transferase and Aryl Hydrocarbon Hydroxylase Activity: Effect of Dioxin. Life Sci., 24:1373-1380. Manis, J., and G. Kim. 1979. Stimulation of Iron Absorption by Polychlorinated Aromatic Hydrocarbons. Am. J. Physiol., 236: E763-E768. March, J. 1968. Advanced Organic Chemistry: Reactions, Mechanisms, and Structure. McGraw-Hill Book Co., New York, pp. 519-520. 220 �Marselos, M. 1979. Comparison of Phenobarbital and Carcinogen-Induced Aldehyde Dehydrogenases in the Rat. Biochem. Biophys. ACTA, 583(1):110-110. Marselos, M. , et al. 1978. Responses of the D-Gluccionic Acid Pathway in Rat Tissues to Treatment with TCDD. Xenobiotica, 8(7):397-402. Matsumura, F., and H. J. Benezet. 1973. Studies on the Bio-accumulation and Microbial Degredation of 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Environmental Health Perspectives, 5:253-258, September. Mauk, C. E., H. W. Prengle, Jr., and J. E. Payne. 1976. Oxidation of Pesticides by Ozone and Ultraviolet Light, U.S. Army Mobility Equipment Research and Development Command, AD-A028 306/9ST. May, G. 1973. Chloracne From the Accidental Production of Tetrachlorodibenzodioxin. British J. Industrial Medicine, 30:276-283. McCann, J., et al. 1976. Detection of Carcinogens as Mutagens in the Salmonella/Microsome Test: Assay of 300 Chemicals. Proc. Nat. Acad. Sci. 73:950-954. McConnell, E. E., and J. A. Moore. 1976. The Comparative Toxicity of Chlorinated Dibenzo-p-dioxin Isomers in Mice and Guinea Pigs. Toxicol. Appl.' Pharmacol. , 37:146. McConnell, E. E., J. A. Moore, and D. W. Dalgard. 1978. Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Rhesus Monkeys (Macaca Mulatta) Following a Single Oral Dose. Toxicol. Appl. Pharmacol., 43:175-187. McConnell, E. E., et al. 1978. The Comparative Toxicity of Chlorinated Dibenzo-p-dioxins in Mice and Guinea Pigs. Toxicol. Appl. Pharmacol., 44:335-356. Mclnty, L 1976. 71(104):383-385. The Graveyard on Milan's Doorstep. New Science, Mclntyre, T. J. (Senator). 1970. Herbicide Program in Vietnam. Congressional Record, August 25, pp. 29932-30012. McNulty, W. P., M.D. 1978a. Pesticide Programs, July 27. Written communication to U.S. EPA Office of McNulty, W. P., M.D., 1978b. Oregon Regional Primate Research Center, personal communication, August 8. McNulty, W. P., et al. 1979a. PCB and TCDD Orofacial Teratogenesis in Macaca Mulatta. Toxicology Research Projects Directory, 04:101. / McNulty, W. P., et al. 1979b. Toxicity of TCDD for Rhesus Monkeys: Report. Pub. No. 878, Oregon Primate Research Center. 221 Brief �McQueen, E. G., et al. 1977. 2,4,5-T and Human Birth Defects, New Zealand Department of Health, June. Mercier, M. J. 1976. 2,3,7,8-Tetrachlorodibenzo-p-dioxin: An Overview. In: Proceedings of the Expert Meeting on the Problems Raised by TCDD Pollution, A. Berlin, A. Buratta, and M. T. Van der Venne, eds. Milan. In press. Merck Index. New Jersey. 1978. 9th ed., M. Windholz, ed. Merck & co. , Inc., Rahway, Merenda, J. J. 1979. TSCA Section 8(e) submission on Polychlorinated Dibenzo-p-dioxins (PCDD's). Letters and enclosures to E. Blair, vice president Dow Chemical Co., from Director, Assessment Division, EPA. EPA Document Control No. 8EHQ-1178-0209, February 15. Meselson, M., P. W. O'Keefe, and R. Baughman. 1978. The Evaluation of Possible Health Hazards From TCDD in the Environment. Presentation for Symposium on the Use of Herbicides in Forestry. Arlington, Virginia,. February 21-22. Miller, S. 1979. Source of Dioxins. Chemical and Engineering News. March 12, p. 4. Miller, R. A., L. A. Nprris, and C. L. Hawkes. 1973. Toxicity of 2,3,7,8TCDD in Aquatic Organisms. Environmental Health Perspectives, 5:177-187, September. Milnes, M. H. 1971. Formation of 2,3,7,8-Tetrachlorodibenzodioxin by Thermal Decomposition of Sodium 2,4,5-Trichlorophenate. Nature (London), 232:395-396, August 6. Mitchell, L. C. 1961. The Effect of Ultraviolet Light on 141 Pesticide Chemicals. J. Assoc. Offic. Agr. Chem., 44:643-712. MITRE Corporation. 1977. The Analysis of Existing Wood Preserving Techniques and Possible Alternatives. Technical Report 7520. Prepared for U.S. EPA, Contract No. 68-01-4310. Modell, J., R. P. deFilippi, and V. Krukonis. 1978. Regeneration of Activated Carbon With Supercritical Carbon Dioxide. Presented at the ACS Conference, Miami, Florida. September 14. Moore, J. A. 1979. A Pesticide. Science, 203(4382):741-742. February 23. Moore, J. A., and R. E. Faith. 1976. Immunologic Response and Factors Affecting its Assessment. Environmental Health Perspectives, 18:125-132. Moore, J. A., B. N. Gupta, and J. G. Vos. 1976. Toxicity of 2,3,7,8-Tetrachlorodibenzofuran--Preliminary Results. In: National Conference on Polychlorinated Biphenyls, Chicago, 1975. EPA-560/6-75-004. 222 �Moore, J. A., M. W. Harris, and P. W. Albro. 1978. Tissue Distribution of (14C) Tetrachlorodibenzo-p-Dioxin in Pregnant and Neonatal Rats. National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina. Moore, J. A., et al. 1973. Postnatal Effects of Maternal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). Environmental Health Perspectives, 5:81-85. Moye, A. L. 1972. Experiencing Relevancy in Organic Chemistry. Hexachlorophene: The Great Clean-All. Journal of Chemical Education, 49(11):770-771, November. Murray, F. J. , et al. 1978. Three-Generation Reproduction Study of Rats Ingesting TCDD. Toxicol. Appl. Pharmacol., 41:200-201. Nagayama, J., M. Kuratsune, and Y. Masuda. 1976. Determination of Chlorinated Dibenzofurans in Kanechlors and "Vusho Oil." Bulletin of Environmental Contamination and Toxicology, 15:9-13. Nash, R. G. , and M. L. Beall, Jr. 1977. Environmental Distribution of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Applied With Silvex to Turf in Microagroecosystem. Final Report. U.S. EPA-USDA Interagency Agreement, EPA-1AG-D6-0054, ARS 173, EPA No. 1001-704. Nash, R. G., and M. L. Beall, Jr. 1978. A Microagroecosystem to Monitor the Environmental Fate of Pesticides. Pesticide Degradation Laboratory, ARS, USDA, Beltsville, Maryland. Unpublished. National Academy of Sciences. 1974. Committee on the Effects of Herbicides in South Vietnam, Part A. Summary and Conclusions. Washington, D.C. National Academy of Sciences. Chapter VI and Bibliography. Nature. London. 1970. 1977. Drinking Water and Health. Another Herbicide on the Blacklist. Part II, Vol. 226:309-311. Neal, R. A. 1979. Mechanisms of Toxicity of the Chlorinated p-Dioxins. Toxicology Research Projects Directory, 04:11. Nebert, D. W., S. S. Thorgeirsson, and J. S. Felton. 1976. Genetic Differences in Mutagenesis, Carcinogenesis, and Drug Toxicity. In: In Vitro Metabolic Activation in Mutagenesis Testing, F. J. deSerrers, et al. , ed. Elsevier/North Holland Publishing Co., Amsterdam, 105-124. Nelson, C. J. , et al. 1979. Retrospective Study of the Relationship Between Agricultural Use of 2,4,5-T and Cleft Palate Occurrence in Arkansas. Teratology, 19:377-384. Nelson, G. (Senator). January 17. 1973. Statement on Dioxins. 223 Congressional Record, �Nelson, J. D. , et al. 1977. 2,3,7,8-Tetrachlorodibenzo-p-dioxin In Vitro Binding to Rat Liver Microsomes. Bulletin of Environmental Contaminants and Toxicology, 18(1):9-13. Neubert, D. , and I. Dillmann. 1972. Embryotoxic Effects in Mice Treated With 2,4,5-Trichlorophenoxyacetic Acid and 2,3,7,8-Tetrachlorodibenzo-pdioxin. Naunyn-Schmiedeberg's Arch. Exp. Path. Pharmacol, 272:243-264. Neubert, D. , et al. 1973. A Survey of the Embryotoxic Effects of TCDD in Mammalian Species. Environmental Health Perspectives, 5:67-79. Newton, M., and S. Snyder. 1978. Exposure of Forest Herbivores to 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in Areas Sprayed With 2,4,5-T. Bulletin of Environmental Contamination and Toxicology, 20(6):743-750. Nilsson, C. A., et al. 1974. Chromatographic Evidence for the Formation of Chlorodioxins From Chloro-2-Phenoxyphenols. J. Chromatography, 96:137-147. Niwa, A., et al. 1975. Genetic Expression of Aryl Hydrocarbon Hydroxylase Activity in the Mouse. Archives of Biochemistry and Biophysics, 166:559-564. Nolan, R. J. , et al. 1979. Elimination of Tissue Distribution of TCDD in Female Guinea Pigs Following a Single Oral Dose. Toxicol. Appl. Pharmacol., 48(1):A162. Norback, D. H. 1975. Tissue Distribution and Excretion of Octachlorodibenzo-p-Dioxin in the Rat. Toxicol. Appl. Pharmacol., 32:330-338. Norback, D. H. , and J. R. Allen. 1973. Biological Responses of the Nonhuman Primate, Chicken and Rat to Chlorinated Dibenzo-p-dioxin Ingestion. Environmental Health Perspectives, 5:233-240. Norman, R. L. 1978. Identification of the Major Cytochrome P-450 Form Transplacentally Induced in Neonatal Rabbits by 2,3,7,8-Tetrachlorodibenzop-dioxin. J. Biol. Chem., 253(23):8640-8647. Oberacker, D. A., and S. Lees. 1977. Microwave Plasma Detoxification of Hazardous and Toxic Materials. In: News of Environmental Research in Cincinnati. U.S. EPA, Municipal Environmental Research Laboratory. Occupational Safety and Health Reporter. 1979. Health Hazards: NIOSH Begins Epidemiclogical Study of All Workers Exposed to Dioxin TCDD. Vol. 9, No. 26, November 29. Offut, C. 1978. September 8. EPA Office of Toxic Substances, personal communication, Oishi, S., M. Morita, and H. Fukuda. 1978. Comparative Toxicity of Polychlorinated Biphenyls and Dibenzofurans in Rats. Toxicol. Appl. Pharmacol., 43:13-22. 224 �Olie, L., P. L. Vermeulen, and 0. Hutzinger. 1977. Chlorodibenzo-p-Dioxins and Chlorodibenzofurans Are Trace Components of Fly Ash and Flue Gas of Some Municipal Incinerators in the Netherlands: Chemosphere, 8:455-459. Oliver, R. J. 1975. Toxic Effects of 2,3,7,8-Tetrachlorodibenzo 1,4 Dioxin in Laboratory Workers. British J. of Industrial Medicine, 32:49-53. Ott, P., M.D. 1978. EPA, personal communication, August 29. Ottinger, R. S., et al. 1973. Recommended Methods of Reduction, Neutralization, Recovery, or Disposal of Hazardous Waste. Volume 3. NTIS PB-224 582. Parkki, M. G., and A. Aitio. 1978. Induction of Drug Metabolizing Enzymes in Different Rat Tissues by TCDD. Arch. Toxicol. ISS Supl. 1, 261-5. Parks, M. 1978. The Seveso Case. EPA Journal, 4(8):11-15, 35, September. Peracchio, A. 1979. Dow Chemical Faces Suits From Veterans of Vietnam Over Babies With Defects. Newsday, copyright. Peterson, J. 1978. Seveso: The Event. Ambio, 7(5-6):232-233. Physitians' Desk Reference. Company, Oradell, New Jersey. 1978. 32nd edition. Medical Economics Piper, W. N. 1979. Toxicant Deregulation of Endocrine Heme Biosynthesis. Toxicology Research Projects Directory, 04:06. Piper, W. N., J. Q. Rose, and P. J. Gehring. 1973. Excretion and Tissue Distribution of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in the Rat. Environmental Health Perspectives, 5:241-244. Plimmer, J. R. 1972. Principles of Photodecomposition of Pesticides: Degradation of Synthetic Organic Molecules in the Biosphere. National Academy of Sciences, Washington, D.C. Plimmer, J. R. 1973. Technical Pentachlorophenol: Origin and Analysis of Base-Insoluble Contaminants. Environmental Health Perspectives, 5:41-47. Plimmer, J. R. 1978a. Approaches to Decontamination or Disposal of Pesticides: Photodecomposition. In: Disposal and Decontamination of Pesticides, M. V. Kennedy, ed. ACS Symposium Series 73, American Chemical Society, Washington, D.C., pp. 12-23. Plimmer, J. R. 1978b. Photolysis of TCDD and Trifluralin on Silica and Soil. Bulletin of Environmental Contamination and Toxicology, 20:87-92. Plimmer, J. R., et al. 1973. Photochemistry of Dibenzo-p-dioxins. In: Chlorodioxins—Qrigin and Fate, E. Blair, ed. Advances in Chemistry, Series 201, American Chemical Society, Washington, D.C., pp. 44-54. 225 �Pocchiari, F. 1978. 2,3,7,8-Tetrachlorodibenzo-p-para-dioxin Decontamination. In: Chlorinated Phenoxy Acids and Their Dioxins, C. Ramel, ed. Ecol. Bull. (Stockholm), 27:67-70. Pocchiari, F. , V. Silano, and A. Zampieri. 1979. Human Health Effects from Accidental Release of TCDD at Seveso, Italy. Ann. New York Acad. Sci., 311:20. Pohland, A. E., and G. C. Yang. 1972. Preparation and Characterization of Chlorinated Dibenzo-p-dioxin. J. Agric. Food Chem., 20:1093-1099. Pohland, A. E., G. C. Yang, and N. Brown. 1973. Analytical and Confirmative Techniques for Dibenzo-p-dioxins Based upon their Cation Radicals. Environmental Health Perspectives, 5:9-13. Pokorny, R. 63:1768. 1941. Some Chlorophenoxyacetic Acids, J. Am. Chem. Soc., Poland, A. 1973. Chlorinated Dibenzo-p-Dioxins: Potent Inducer of a-Aminolevulinic Acid Synthetase and Aryl Hydrocarbon Hydroxylase 11. A Study of the Structure-Activity Relationship. Mol. Pharmacol., 9:736-747. Poland, A., and E. Glover. 1974. Comparison of 2,3,7,8-Tetrachlorodibenzo-p-dioxin, A Potent Inducer of Aryl Hydrocarbon Hydroxylase, With 3-Methyl Chloranthrene. Mol. Pharmacol., 10:349-359. Poland, A., and A. Kende. 1976. 2,3,7,8-Tetrachlorodibenzo-p-dioxin: Environmental Contaminant and Molecular Probe. Federal Proceedings, 35(12):2404-2411. Poland, A., et al. 1971. A Health Survey of Workers in a 2,4-D and 2,4,5-T Plant With Special Attention to Chloracne, Porphyria Cutanea Tarda, and Psychologic Parameters. Arch. Environ. Health, 22:316-327. Poland, A., et al. 1976. 3,4,3,,4,-Tetrachloro Azoxybenzene and Axobenzene: Potent Inducers of Arylhydrocarbon hydroxylase. Science, 194:627. Poole, C. 1979. U.S. EPA, personal communication. Purkyne, J. E., et al. 1974. Acne Chlorina Porphyria Cutanea Tarda A Jine Projevy Celkoue Intoxikace Pki Uyrobe Herbicid. Ceskoslovenska Dermatologie, 3(49):145-57. Ramsey, J. C. 1979. The In Vivo Biotransformation of TCDD in the Rat. Toxicol. Appl. Pharmacol., 48(1):A162. Rappe, C. 1978. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Introduction. In: Dioxin—Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds. SP Medical and Scientific Books, New York, London, pp. 9-11. 226 �Rappe, C. 1979. Polychlorinated Dibenzodioxins and Dibenzofuran Compounds of Current Interest. Lakartidrigen, 76(1-2):21-29. Rappe, C., et al. 1978. Formation of Polychlorinated Dibenzo-p-dioxins (PCDDs) and Dibenzofurans (PCDFs) by Burning or Heating Chlorophenates. Chemosphere, 7(3):269-281. Rawls, R. 1979. Dow Finds Support, Doubt for Dioxin Ideas. Engineering News, pp. 23-29, February 12. Chemical and Reece, M. 1978a. EPA, personal communication, September 18. Reece, M. 1978b. October 10. Written communication from U.S. EPA Pesticide Programs, Reece, M. 1978c. EPA, personal communication, November 2. Reggiani, G. 1977. Medical Problems Raised by the TCDD Contamination in Seveso, Italy. Presented at the 5th International Conference on Occupational Health in the Chemical Industry (Medichem), San Francisco, September 5-10. Reggiani, G. 1978. Medical Problems Raised by the TCDD Contamination in Seveso, Italy. Arch Toxicol. , 40(3):161-188. Reggiani, G. 1979a. Estimation of the TCDD Toxic Potential in the Light of the Seveso Accident. Arch. Toxicol. (Suppl), 291-302. Reggiani, G. 1979b. TCDD Contamination in Italy: The Risk Assessment of Low-Level Exposure, Cumulative Effect and Long-Term Consequences. Toxicol. Appl. Pharmacol., 48(1):A180. Revzin, P. 1979. Chemical Cloud Still Casts Long Shadow Over Seveso, Italy. The Wall Street Journal, July 10. Richards, B. 1979a. Pesticide Waste Dumping Probed Near Little Rock. The Washington Post. May 19. Richards, B. 1979b. Arkansas Pesticide Plant Being Investigated. Washington Post. June 17. Richards, B. July 25. 1979c. Arkansas Site May Hold Clue. The The Washington Post. Riley, B. T. 1975. Summation of Conditions and Investigations for the Complete Combustion of Organic Pesticides. U.S. Environmental Research Laboratory, EPA-600/2-75-044. Robbins, A. 1332. 1979. Dioxin Studies. 227 Science, Vol. 205, September 28, p. �Rogers, C. J., and R. Allen. 1978. Developing Technology for Detoxification of Pesticides and Other Hazardous Materials. In: Disposal and Decontamination of Pesticides, M. V. Kennedy, ed. ACS Symposium, Series 73, American Chemical Society, Washington, D.C. Roper, M., T. Stack, and R. A. Deitrich. 1976. Phenobarbital and Tetrachlorodibenzo-p-dioxin Induce Different Isoenzymes of Aldehyde Dehydrogenase. Fed. Proc. 35:282. Rose, J. Q., et al. 1976. The Fate of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Following Single and Repeated Oral Doses to the Rat. Toxicol. Appl. Pharmacol., 36:209-226. Rosen, J. D. 1971. Photodecomposition of Organic Pesticides. In: Organic Compounds in Aquatic Environments, S. J. Faust and J. V. Hunter, eds. Marcel Dekker, Inc., New York. Saint-Ruf, G. 1978. The Structure and Biochemical Effects of TCDD. In: Dioxin—Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds. SP Medical and Scientific Books, New York, London. Saint-Ruf, G., and Do-PhuocHien. 1975. Similarity of the Biochemical Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin and the 2,3,7,8-Tetrabromodibenzo-p-dioxin in the Rat. C. R. Hebd. Seances Acad. Sci. Paris Ser. D, 280(23):2709-2711. Salkinoja-Salonen, M. 1979a. Waste Purifying Procedure. 4,169,049, September 25. U.S. Patent No. Salkinoja-Salonen, M. 1979b. Microbial Dechlorination of Chloro-Organics. University of Helsinki, Finland. Presentation given at U.S. EPA offices, IERL, Cincinnati, December 11. Schantz, S. L., et al. 1979. Toxicological Effects Produced in Nonhuman Primates Chronically Exposed to Fifty Parts Per Trillion 2,3,7,8-TCDD. Toxicol. Appl. Pharmacol., 48(1):A180. Schwetz, B. A., P. J. Gehring, and R. J. Kociba. 1978. Toxicological Properties of Pentachlorophenol Relative to its Content of Chlorinated Dibenzo-p-Dioxins. Dow Chemical, Midland, Michigan. Schwetz, B. A., P. A. Keeler, and P. J. Gehring. 1973. The Effect of Purified and Commercial Grade Pentachlorophenol on Rat Embryonal and Fetal Development. Toxicol. Appl. Pharmacol., 28:146-150. Schwetz, B. A., et al. 1973. Toxicology of Chlorinated Dibenzo-p-dioxins. Environmental Health Perspectives, 5:87-99. Science. 24. 1979. Agent Orange Furor Continues to Build. 228 Vol. 205, August �Sconce, J. S., ed. 1967. Chlorine: Its Manufacture, Properties and Uses. Reinhold Publishing Corp., New York, pp. 448-456, and 851-852. Scurlock, A. C., et al. 1975. Incineration in Hazardous Waste Management. U.S. Environmental Protection Agency, NTIS PB-261 049. Seefeld, M. D., et al. 1979a. Effects of TCDD on Liver Function in Rats, Rabbits and Guinea Pigs. Toxicol. Appl. Pharmacol., 48(1):A153. Seefeld, M. D., et al. 1979b. Time Course of TCDD Effects on Liver Function in Rhesus Monkeys. Toxicol. Appl. Pharmacol., 48(1):A160. Seiler, J. P. 1973. A Survey on the Mutagenicity of Various Pesticides. Experientia, 29(5):622-623. Seiler, J. P. 1977. Inhibition of Testicular DNA Synthesis by Chemical Mutagens and Carcinogens. Preliminary Results in the Validation of a Navel Short Term Test. Mutation Res., 46:305-310. Severo, R. 1979. Agent Orange, A Legacy of Suspicion. The New York Times, May 27-29. Shadoff, L. A., et al. 1977. The Gas Chromatographic - Mass Spectrometric Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Fat From Cattle Fed Ronnel Insecticide. Ann. Chim., 67:583-592. Sharma, R. P. 1978. Reversal of Immunologic and Toxicologic Effects of a Single Exposure of TCDD in Mice. Trace Subst. Environ. Health, 12:299-306. Shea, K. P., and B. Lindler. 1975. Pandora and the Storage Tank. Environment, 17(6):12-15, September. Shiroishi, K. September 12. 1978. EPA Office of Enforcement, personal communication, Shiver, J. K. 1976. EPA-600/2-76-270. Converting Chlorohydrocarbon Wastes by Chlorolysis, Short, F. W., and E. F. Elslager. 1962. Intestinal Anthelmintics. I. The Preparation of Bis (2,4,5-Trichlorophenol)—Piperazine Salt (Trichlofenol Piperazine) and Other Phenol-Piperazine Salts. J. Med. Pharm. Chem., 5:642. Sidwell, A. E. 1976a. Chemistry of 2,4,5-T Production, and History of Trichlorophenol and 2,4,5-T Acid Manufacturers at Jacksonville, Arkansas. Written communications to the Arkansas Dept. of Pollution Control and Ecology, from Transvaal Inc., Jacksonville, Arkansas, September 17. Sidwell, A. E. 1976b. TCDD Content of Soil and Liquid Sample. Written communication to T. Bennett, Jr., Transvaal, Inc., from Transvaal, Inc., October 27. 229 �Sittig, M. 1969. Organic Chemical Process Encyclopedia, 2d ed. Development Corp,, Park Ridge, New Jersey. Noyes Sittig, M., 1974. Pollution Control In the Organic Chemical Industry. Noyes Data Corporation, Park Ridge, New Jersey, pp. 116-119. Smith, F. A., B. A. Schwetz, and K. D. Nitchke. 1976. Teratogenicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in CF-1 Mice. Toxicol. Appl. Pharmacol., 38:517-523. Solch, J. G. et al. 1978. The Analysis of Environmental Samples for 2,3,7,8-TCDD Utilizing High and Low Resolution Gas-Liquid Chromatogrpahy-Mass Spectrometry. Final Report. U.S. EPA Contract No. 68-01-1959. June. Solch, J. G. et al. 1980. Wright State University Quarterly Report to the U.S. EPA on Cooperative Agreement No. CR 806846-01. March 23. Sparschu, G. L., F. L. Dunn, and V. K. Rowe. 1970. Teratogenic Study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the Rat. Toxicol. Appl. Pharmacol., 17:317-318. Sparschu, G. L., F. L. Dunn, and V. K. Rowe. 1971. Study of the Teratogenicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in the Rat. Fd. Cosmet. Toxicol., 9:405-412. Stanford Research Institute. USA. Menlo Park, California. 1976-1979. Directory of Chemical Producers, Stark, H. E., J. K. McBride, and G. F. Orr. 1975. Soil Incorporation/ Biodegradation of Herbicide Orange. Vol. I. Microbial and Baseline Ecological Study of the U.S. Air Force Logistics Command Test Range, Hill Air Force Base, Utah. Final Report. Stehl, R. H., and L. L. Lamparski. 1977. Combustion of Several 2,4,5-T Compounds: Formation of 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Science, 197:1008-1009. Stehl, R. H., et al. 1973. The Stability of Pentachlorophenol and Chlorinated Dioxin to Sunlight, Heat and Combustion. In: Chlorodioxin--0rigin and Fate, E. H. Blair, ed. Advances in Chemistry, Series 120, American Chemical Society, Washington, D.C., pp. 119-125. Storherr, R. W., et al. 1971. Steam Distillation Technique for the Analysis of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Technical 2,4,5-T. J. Assoc. Off. Anal. Chem., 54:218-219. Strigini, P. and A. Torriani. 1977. Science. 197:1034, 1036, September 9. Seveso: "Encouraging Evidence"? Sugar, J. 1979. Research Related to the Herbicide Buvinol, Especially for Possible Carcinogenicity. TARC Sci. Pub!., 25:167-172. 230 �Sugar, J. , et al. 1979. Role of Pesticides in Hepatocarcinogenesis. Toxicol. Environ. Health, 5:1552-3, pp. 183-191. Sweeney, G. D. 1979. Iron Deficiency Science, 204(4390):332-335. Taylor, J. S. 1974. Prevents J. Liver Toxicity TCDD. Chloracne—A Continuing Problem. Cutis, 13(585):4174. Taylor, M. L. and T. 0. Tiernan. 1979. Written communication to L. Barnes, American Public Health Association, Washington, D. C. 20036. April 4. Taylor, M. 1980. Personal communication. State University, March 4. The Brehm Laboratory, Wright Techini, M. L. , et al. 1977. Approaches to Examination of Genetic Damage After a Major Hazard in Chemical Industry: Preliminary Cytogenetic Findings on TCDD-exposed Subjects After Seveso Accident. Presented at the Expert Conference on Genetic Damage Caused by Environmental Factors, Oslo, Norway, May 11-13. Tenzer, R., et al. Characteristics of the Mobile Field Use System for the Detoxification/Incineration of Residuals From Oil and Hazardous Material Spill Clean-up Operations. Prepared for U.S. Environmental Protection Agency by MBA Associates under Contract No. 68-03-2515. Theiss, A. M. , and P. Goldman. 1977. Uber das Trichlorphenol-DioxinUmfall-geschehen urn der BASF AG VOM 13 November 1953. In: Vortrag auf dem IV Medichem-Kongress, Haifa, 1976. International Agency for Research on Cancer. Thigpen, J. E., et al. 1975. Increased Susceptibility to Bacterial Infection as a Sequela of Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Infection and Immunity, 12:1319-1324. Thomas, P. T., and R. Hinsdill. 1979. The Effect of Perinatal Exposure to TCDD on the Immune Response of Young Mice. Drug. Chem. Tox. 2 Iss., 12:77-98. Tiernan, T. 0. 1975. Analytical and Chemical Support for Herbicide Disposition Activities. Final Report. U.S. Air Force Contract No. F41608-76-C-0464. Tiernan, T. 0. 1976. The Analysis of Environmental Samples for 2,3,7,8-Tetrachlorodibenzo-p-dioxin Utilizing High and Low Resolution GasLiquid Chromatography-Mas Spectrometry. Final Report. U.S. EPA Contract No. 68-01-1959. Tiernan, T. 0. and M. L. Taylor. 1978. Development and Application of Analytical Methodology for Determination of Hexa-, Hepta-, and Octachlorodibenzodioxins in Beef Samples. Final Report. USDA Contract No. 12-64-4-378. Tiernan, T. 0. and M. L. Taylor. D2832NAEX. March 31. 1980. 231 Interim Report on EPA Order No. �Tiernan, T. 0. et al. 1979. Determination of Tetrachlorodibenzo-p-dioxin (TCDD) in Stack Effluent Samples Obtained During At-Sea Incineration of Herbicide Orange. U.S. EPA Purchase Order No. DA-8-6414J. May 31. Tiernan, T. 0. et al. 1980. Analyses of Industrial Samples for Tetrachlorodibenzo-p-dioxins (TCDDs). Final Report. U.S. EPA Contract No. 68-03-2830 and Order Nos. 9T-1501-NTEX and OT-0267-NAEX. April 1. Tiernan, T. 0. and M. L. Taylor. 1978. Development and Application of Analytical Methodology for Determination of Hexa-, Hepta-, and Octachlorodibenzodioxins in Beef Samples. Final Report. USDA Contract No. 12-64-4-378. April. Ton That, T., et al. 1973. LeCancer Primaire du foie au Viet-nam Chirurgie International Agency for Research on Cancer, 99:427-436. Toth, K., et al 1977. Carcinogenic Bioassay of the Herbicide 2,4,5-Trichlorophenoxy Ethanol (TCPE) With Different 2,3,7,8-Tetrachlorodibenzo-p-dioxin (Dioxin) Content in Swiss Mice. In: International Conference on Ecological Perspectives on Carcinogens and Cancer Control, Basel and Karger, eds., Cremona, 1976. Toth, K., et al. 1979. Carcinogenicity Testing of Herbicide 2,4,5-T Ethanol Containing Dioxin and of Pure Dioxin in Swiss Mice. Nature, 278(5704):548-549. Toxic Materials News. February 28. 1979a. Toxic Materials News in Brief. 6(9):70, Toxic Materials News. 1979b. Remaining Uses of 2,4,5-T, Silvex Undergo Review Prior to Possible Cancellation. July 18, p. 230. Toxic Materials News. 239. 1979c. Toxic Materials News In Brief. July 25, p. Toxic Materials News. 1979d. Federal Judge Allows Agent Orange Suit Against Manufacturers. 6(48):381, November 28. Toxic Materials News. p. 383. 1979e. Toxic Materials News In Brief. November 28, U.S. Code of Federal Regulations Title 21. 1978. Subpart D, Section 250.250. Hexachlorophene, as a component of drug and cosmetic products. Effective January 19, 1978. U.S. Dept. HEW. 1975. Center for Disease Control, NIOSH. Evaluation Determination Report No. 74-117-251, December. Health Hazard U.S. Dept. HEW. 1978. Bioassay of Hexachlorophene for Possible Carcinogenicity. National Cancer Institute. Carcinogenesis Technical Report. CAS No. 70-30 NCI-CG-TR-40, Series No. 40, p. 1. 232 �U.S. Environmental Protection Agency. 1975a. Destructing Chemical Wastes in Commercial Scale Incinerators. Volume I, Technical Summary. July. NTIS PB-257 709. U.S. Environmental Protection Agency. 1975b. Hazardous Waste Disposal Damage Reports. Document No. 2. EPA/530/SW- 151.2. U.S. Environmental Protection Agency. 1977a. Destroying Chemical Wastes in Commercial Scale Incinerators, Facility Number l--The Marquat Company. NTIS PB-265 451. U.S. Environmental Protection Agency. Health, PB-269519. U.S. Environmental Protection Agency. teria 2,3,7,8-TCDD. Washington, D.C. 1977b. 1978a. NAS. Drinking Water and Ambient Water Quality Cri- U.S. Environmental Protection Agency. 1978b. Computer Data Files on Priority Pollutants in the Leather Tanning Industry, Lines 2261-8615. Research Triangle Park, North Carolina. U.S. Environmental Protection Agency. 1978c. Draft Report of the Ad Hoc Study Group on Pentachlorophenol Contaminants. Environmental Health Advisory Committee, Science Advisory Board. Research Triangle Park, North Carolina. U.S. Environmental Protection Agency. 1978d. Draft Status Report. Dioxin in Industrial Sludges. Office of Solid Waste, Research Triangle Park, North Carolina. U.S. Environmental Protection Agency. 1978e. Notice of Rebuttable Presumption Against Registration and Continued Registration of Pesticide Products Containing Pentachlorophenol. Federal Register, 43(202): 48443-48617. U.S. Environmental Protection Agency. 1978f. Pesticide Product Information on Microfiche, Chemical Reference File (Alpha). Research Triangle Park, North Carolina. U.S. Environmental Protection Agency. 1978g. Phase I Report on Hexachlorobenzene (HCB). Research Triangle Park, North Carolina. U.S. Environmental Protection Agency. 1978h. Rebuttable Presumption Against Registration and Continued Registration of Pesticide Products Containing 2,4,5-T. Federal Register, 43(78):17116-17157. U.S. Environmental Protection Agency. 1978i. Rebuttable Presumption Against Registration and Continued Registration of Pesticide Products Containing 2,4,5-Trichlorophenol and Its Salts. Federal Register, 43(149): 34026-34054. U.S. Environmental Protection Agency. 1979. The Presence of Priority Pollutants in the Synthetic Manufacture of Pharmaceuticals. Draft. 233 �U.S. International Trade Commission. 1974, 1976, 1977. Synthetic Organic Chemicals, United States Production and Sales. U.S. National Institute of Environmental Health Sciences/International Agency for Research on Cancer. 1978. Long-Term Hazards of Polychlorinated Dibenzodioxins and Polychlorinated Dibenzofurans. Joint Working Group Report, IARC, Lyon, France. U.S. Patent Office. phenolates. 1935. Patent No. 1,991,329. U.S. Patent Office. chlorophenol. 1939. AlkaliMetal Trichloro- Patent No. 2,176,417. Preparation of Penta- U.S. Patent Office. 1941. Patent No. 2,250,480. Diphenyl Methane and Methods of Producing the Same. U.S. Patent Office. Halogenated Cresols. Dihydroxy Hexachloro 1943. Patent No. 2,319,960. Process for Producing U.S. Patent Office. 1948. Patent No. 2,435,593. Bis-(3,5,6-Trichloro-2-hydroxyphenyl) Methane. Process for Making U.S. Patent Office. 1949. Patent No. 2,471,575. 2,4-Dichlorophenoxyacetic Acid. Process of Preparing U.S. Patent Office. 1950. Patent No. 2,509,245. 2,4,5-Trichlorophenol. 1950. Preparation of U.S. Patent Office. 1952. Patent No. 2,599,516. 0-2,4,5-Trichlorophenyl 0,0-Dialkyl-thiophosphates. U.S. Patent Office. ilides. 1955a. Patent No. 2,703,322. U.S. Patent Office. Hexachloride. 1955b. Patent No. 2,728,799. U.S. Patent Office. 1956a. Patent or2,4,5-Trichlorophenoxy-propionic Acid. No. U.S. Patent Office. 1956b. Patent No. 2,754,324. of the Haloaryloxy Loweralkanols. Poly Halo-salicylanRefinement of Benzene 2,749,360. Esters of crorDichloropropionates U.S. Patent Office. 1956c. Patent No. 2,756,260. Method of Making Trichlorophenol Mixtures Which Are Rich in the 2,4,5-isomer. U.S. Patent Office. 1956d. Patent No. 2,765,224. Herbicide. U.S. Patent Office. 1957a. Patent No. 2,792,434. tion of Hexachlorobenzene. Process for the Produc- U.S. Patent Office. 1957b. Patent No. 2,799,713. Method of Making Trichlorophenols From Tetrachlorobenzenes. 234 �U.S. Patent Office. 1957c. Di- and Trichlorobenzenes. Patent No. 2,799,714. Method of Hydrolyzing U.S. Patent Office. 1957d. Patent No. 2,812,365. Bis-(3,5,6-Trichloro-2-hydroxyphenyl) Methane. Process of Preparing U.S. Patent Office. 1957e. Patent No. 2,812,366. Preparation of Polychlorophenols. Improvements in the U.S. Patent Office. 1957f. Patent No. 2,812,367. Preparation of Polychlorophenols in Aqueous Medium. Improvements in the U.S. Patent Office. 1958a. Aliphatic Carboxylic Acids. Patent No. 2,830,083. Production of Aryloxy U.S. Patent Office, phenols). Patent No. 2,849,494. 2,2-Thiobis (polyhalo- 1958b. U.S. Patent Office. 1958c. Patent No. 2,852,548. Process for the Production of Salts of Aryloxyalkanol-Sulfuric Acid Semiesters. U.S. Patent Office. 1960a. Patent No. 2,922,811. Method for the Manufacture of 0-(Chlorophenyl)0,0-dialkylphosphorothiolates. U.S. Patent Office. 1960b. Patent No. 2,947,790. Process for the Manufacture of 0-(Chlorophenyl)0,0-dialkylphosphorothiolates. U.S. Patent Office. 1961a. 2,4,5-Trichlorophenol. Patent No. 2,980,681. Salt of Piperazine and U.S. Patent Office. 1961b. and Method of Manufacture. Patent No. 3,005,720. Bacteriostatic Articles U.S. Patent Office. 1962. Patent No. 3,024,163. Bacteriostats. U.S. Patent Office. 1963a. of Undesired Vegetation. Patent No. 3,074,790. Method for the Control U.S. Patent Office. 1963b. Patent No. 3,076,025. Process for Production of Substituted Phenoxyalkanoic Acid. U.S. Patent Office. 1967a. Composition and Method. Patent No. 3,297,427. Synergistic Herbicidal U.S. Patent Office. Patent No. 3,347,937. 2,4,5-Trichlorophenol. Patent No. 3,481,991. Preparation of Chlori- 1967b. U.S. Patent Office. 1969. nated Hydroxy Compounds. U.S. Patent Office. 1971. Patent No. 3,607,949. Methylenebis-(3,4,6-Trichlorophenol). U.S. Patent Office. 1972. Patent No. 3,676,508. facture of Carbon Tetrachloride. 235 Production of 2,2Process for the Manu- �U.S. Patent Office. 1974a. of Pentachlorophenol. Patent No. 3,816,268. Stabilized Distillation U.S. Patent Office. 1974b. -Patent No. 3,852,160. Distillation of Pentachlorophenol with Salicylaldehyde and Water. U.S. Patent Office. chlorophenol. 1974c. Patent No. 3,852,161. Distillation of Penta- U.S. Tariff Commission. 1968. Production and Sales. Synthetic Organic Chemicals, United States Van Miller, J. P., and J. R. Allen. 1977. Chronic Toxicity of 2,3,7,8-TCDD in Rats. Fed. Am. Soc. Exp. Biol., 35:396. Van Miller, J. P., J. J. Lalich, and J. R. Allen. 1977. Increased Incidence of Neoplasms in Rats Exposed to Low Levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin. Chemosphere, 6:537-544. Van Miller, J. P., R. J. Marlar, and J. R. Allen. 1976. Tissue Distribution and Excretion of Tritiated Tetrachlorodibenzo-p-dioxin in Non-human Primates and Rats. Fd. Cosmet. Toxicol., 14:31-34. Verrett, J. 1970. Statement of Dr. J. Verrett, Food and Drug Administration, Department of HEW. In: Hearings Before the Subcommittee on Energy, Natural Resources and the Environment of the Committee on Commerce, U.S. Senate, Second Session on Effects of 2,4,5-T on Man and the Environment. Serial 96-60, pp. 190-203. Villanueva, E. C., V. W. Burse, and R. W. Jennings. 1973. Chlorodibenzo-pdioxins Contamination of Two Commercially Available Pentachlorophenols. J. Agric. Food Chem., 21(4):739-740. Villanueva, E. C. , et al. 1974. Evidence of Chlorodibenzo-p-dioxin and Chlorodibenzofuran in Hexachlorobenzene. J. Agric. Food Chem., 22(5):916-17. Vinopal, J. H., and J. E. Casida. 1973. Metabolic Stability of 2,3,7,8Tetrachlorodibenzo-p-dioxin in Mammalian Liver Microsomal Systems and in Living Mice. Arch. Environ. Contam. Toxicol., 1:122-132. Vinopal, J. H., I. Yamamotp, and J. E. Casida. 1973. Preparation of Tritium Labeled Dibenzo-p-dioxin and 2,3,7,8-Tetrachlorodibenzo-p-dioxin. In: Chlorodipxins—Origin and Fate, E. H. Blair, ed. Advances in Chemistry, Series 120, American Chemical Society, Washington, D.C. Viviani, A., et al. 1978. Time Course of the Induction of Aryl Hydrocarbon Hydroxylase in Rat Liver Nuclei and Microsomes by Phenobarbital, 3-Methyl Cholanthrene, TCDD, Dieldrin and Other Inducers. Biochemical Pharmacology, 27:2103-2108. 236 �Vos, J. G., and L. Kater. 1978. Immune Suppression by TCDD. In: Dioxin-Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallaro, and G. Galli, eds. SP Medical and Scientific Books, New York, London. Vos, J. G., and J. A. Moore. 1974. Suppression of Cellular Immunity in Rats and Mice by Maternal Treatment With 2,3,7,8-TCDD. Int. Arch. Allergy Appl. Immunol., 47:777-779. Vos, J. G. , J. A. Moore, and J. G. Zinkl. 1973. Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on the Immune System of Laboratory Animals. Environmental Health Perspectives, 5:149-162. Vos, J. G., J. A. Moore, and J. G. Zinkle. 1974. Toxicity of 2,3,7,8-TCDD in C57B1/6 Mice. Toxicol. Appl. Pharmacol., 29:229-241. Vos, J. G. , et al. 1978. Studies on 2,3,7,8-TCDD-Induced Immune Suppression and Decreased Resistance to Infection: Endotoxin Hypersensitivity, Serum Zinc Concentrations and Effects of Thymosin Treatment. Toxicology, 9:75-86. Wade, N. 1971. Hexachlorophene: Chemical. Science, 174:805-807. FDA Temporizes on Brain Damaging Walker, A. E., and J. V. Martin. 1979. Workers. Lancet, 446-7, February 29. Lipid Profiles in Dioxin-Exposed Wall Street Journal. 1979. EPA Orders Immediate Halt to Most Uses of Herbicide 2,4,5-T and Similar Products. March 2. Walsh, J. 1977. Seveso-The Questions Persist Where Dioxin Created a Wasteland. Science, 197:1064-1067. Ward, C., and F. Matsumura. 1977. Fate of 2,4,5-T Contaminant, 2,3,7,8Tetrachlorodibenzo-p-dioxin (TCDD) in Aquatic Environments. Department of Entomology, University of Wisconsin. Technical Completion Report, Project Number OWRT-A-058-Wis. Ward, C. T. , and F. Matsumara. 1978. Fate of 2,3,7,8-TCDD in a Model Aquatic Environment. Arch. Environ. Contam. Toxicol; 7(3):349-57. Warnick, H. 1977. Memorandum to U.S. Environmental Protection Agency File. Representative 2,4,5-T Labels. Research Triangle Park, North Carolina. Warnick, H. 1978a. tion, August 21. EPA Office of Pesticide Programs, personal communica- Warnick, H. 1978b. tions, October 20. EPA Office of Pesticide Programs, personal communica- Wassom, J. S., H. E. Huff, and N. Loprieno. 1978. A Review of the Genetic Toxicology of Chlorinated Dibenzo-p-dioxins. Draft Report. National Inst. of Environ. Health Sciences, Inter-agency Agreement 40-247-70. 237 �Watkins, D. A. 1974. Implications of the Photochemical Decomposition of Pesticides. Chemistry and Industry (London), 5:185-190. Watkins, D. R. 1978b. Program for Prevention of Dioxin Exposure. U.S. EPA, lERL-Cincinnati, Ohio. Unpublished. Watkins, D. R. 1979a. History of Industrial Sample Containing Dioxin (TCDDs). U.S. EPA, lERL-Cincinnati memo. March 27. Watkins, D. R. 1979b. Personal communication. Research Laboratory, Cincinnati, Ohio. U.S. EPA, Environmental Watkins, D. R. 1980. U.S. EPA, lERL-Cincinnati, Ohio, personal communication. Weissberg, J. B., and J. G. Zinkl. 1973. Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Upon Hemostasis and Hematologic Function in the Rat. Environmental Health Perspectives, 5:119-123. Wertheim, E. Philadelphia. 1939. Textbook of Organic Chemistry. The Blakiston Co. , Weitzman, L. n.d. Michigan Technological University grant proposal entitled "The Utility of Wet Oxidation in the Treatment of Hazardous Organic Wastes." Memorandum to Dr. E. Berkau, U.S. EPA, lERL-Cincinnati. Westing, A. H. 1979. The Safety of 2,4,5-T. WGBH Educational Foundation. Transcript. Science, 206:1135, December. 1979. A Plague On Our Children. Nova Whitemore, F. C. 1975. A Study of Pesticide Disposal in a Sewage Sludge Incinerator. EPA/530/SW116c, NTIS PB-253 485. Whiteside, T. 1977. A Reporter at Large: Cloud. The New Yorker, July 25, pp. 30-55. The Pendulum and the Toxic Wilkinson, R. R., G. L. Kelso, and F. C. Hopkins. 1978. Report: Pesticide Disposal Research. EPA-600/ 2-78-183. State-of-the-Art Wilkinson, R. D., et al. 1978. State-of-the-Art Report on Pesticide Disposal Research. In: Disposal and Decontamination of Pesticides, M. V. Kennedy, ed. ASC Symposium Series 73, American Chemical Society, Washington, D.C., pp. 73-80. Wilson, J. 1978. 1(8066):722. Breast Enlargement at an Italian School. Lancet, Wipf, H. K. , et al. 1978. Field Trials of Photodegradation of TCDD on Vegetation After Spraying with Vegetable Oil. In: Dioxin—Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallero, and G. Galli, eds. SP Medical and Scientific Books, New York, London, pp. 201-216. 238 �Wong, A. S. , and D. G. Crosby. 1978. Decontamination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) by Photochemical Action. In: Dioxin—Toxicological and Chemical Aspects, F. Cattabeni, A. Cavallero, and G. Galli, eds. SP Medical and Scientific Books, New York, London. Woods, J. S. 1973. Studies of the Effects of 2,3,7,8-Tetrachlorodibenzo-pdioxin of Mammalian Hepatic S-aminolevulinic Acid Synthetase. Environmental Health Perspectives, 5:221-225. Woolson, E. A., and P. D. Ensore. 1973. Dioxin Residues in Lakeland Sand and Bald Eagle Samples. In: Chlorodioxins: Origin and Fate, E. Blair, ed. American Chemical Society, Washington, D.C., pp. 112-118. Woolson, E. A., R. F. Thomas, and P. D. Ensore. 1972. Survey of Polychlorodibenzo-p-dioxin Content in Selected Pesticides. J. Argic. Food Chem., 20(2):351-354. World Health Organization. 1977. International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Man. Some Fumigants, the Herbicides 2,4-D and 2,4,5-T, Chlorinated Dibenzodioxins and Miscellaneous Industrial Chemicals. Vol. 15, Lyon, France, August. World Health Organization. 1978. International Agency for Research on Cancer. Information Bulletin on the Survey of Chemicals Being Tested for Carcinogenicity. No. 7, Lyon, France. Wright State University. 1979a. Report on Analyses of Love Canal Samples for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). New York State Department of Health Purchase Order No. 5975. January 21. Wright State University. 1979b. Report on Analyses of Love Canal Samples for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD). For New York State Department of Health. April 20. Yang, K. H. , and R. E. Peterson. 1979. Differential Effects of Halogenated Aromatic Hydrocarbons on Pancreatic Excretory Function in Rats. University of Wisconsin, Madison. Yang, G. C. , and A. E. Pohland. 1973. Cation Radicals of Chlorinated Dibenzo-p-dioxins. In: Chlorodioxins--0rigin and Fate, E. Blair, ed. American Chemical Society, Washington, D.C., pp. 33-43. Yates, P. B. 1979. Written communication to D. Watkins, U.S. EPA, IERLCincinnati, from New South Wales State Pollution Control Commission, Australia, January 18. Yockim, R. S. , A. R. Isensee, and G. T. Jones. 1978. Distribution and Toxicity of TCDD and 2,4,5-T in an Aquatic Model Ecosystem. Chemosphere, 7(3):215-220. 239 �Young, A. L. 1974. Ecological Studies on a Herbicide-Equipment Test Area. (TA C-52A) Eglin AFB Reservation, Florida. Air Force Armament Lab, Technical Report AFATL-TR-74-12. Young, A. L. 1978. Written communication to D. Watkins, U.S. EPA, IERL, from USAF, August 23. Young, A. L., E. Arnold, and A. M. Wachinsk. 1974. Field Studies on the Soil Persistence and Movement of 2,4-D, 2,4,5-T and TCDD. Presentation to the Weed Science Society of America, Las Vegas, Nevada, Abstract No. 226, February 13. Young, A. L., C. E. Thalken, and W. E. Ward. 1975. Studies of the Ecological Impact of Repetitive Aerial Applications of Herbicides on the Ecosystem of Test Area C-52A, Eglin AFB, Florida. A.F. Armament Lab. AFATL-TR-75-142. Young, A. L., et al. 1976. Fate of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in the Environment: Summary and Decontamination Recommendations. USAFA-TR-76-18. Young, A. L., et al. 1978. The Toxicology, Environmental Fate, and Human Risk of Herbicide Orange and its Associated Dioxin. USAF, OEHL Technical Report TR-78-92. Zedda, S., A. M. Cirla, and C. Sala. 1976. Accidental Contamination by TCDD, The ICMESA Incident. Medicina Del Lavoro, 67(5):371-378. Zinkl, J. G., et al. 1973. Hematologic and Clinical Chemistry Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Laboratory Animals. Environmental Health Perspectives, 5:111-118. Zitko, V., 0. Hutzinger, and P. M. Choi. 1972. Contamination of the Bay of Fundy-Gulf of Maine Area With Polychlorinated Biphenyls, Polychlorinated Terphenyls, Polychlorinated Dibenzodioxins, and Dibenzofurans. Environmental Health Perspectives, 1:47. 240 �INDEX Accumulation in plants, 129, 130 Acute toxicity, 169-173, 190-191 Aminophenols, 70 Aquatic toxicity, 169, 173 Bioaccumulation, 119-129 Bioconcentration, see bioaccumulation Biodegradation, 98-102 Biomagnification, see bioaccumulation Biological methods of disposal: .Soil conditioning, 144 Wastewater treatment systems, 144, 145 Micropit disposal, 145, 146 Biological transport in animals, 118-129 Bithionol, 53, 54 Brominated phenols, 66, 67 Carcinogenicity, 182-184, 198 Chemical methods of disposal: Ozone treatment, 140, 141 Chloroiodide degradation, 140, 142 Wet air oxidation, 142 Chlorinolysis and chlorolysis, 142, 143 Catalytic dechlorination, 143 Chlorophenols, 14-63 Manufacture, 17-24 Production, 24-26, 60, 61-63 Wastes, 58-59, 97, 131-133 Chronic toxicity, 173-180, 191-194 Combustion residues, dioxins in, 70-74, 89, 118, 132-133 241 �Comparative lethal doses, 169-171 Contaminated industrial wastes, 80-83 Cytotoxicity, 188, 189 2,4-D, 33-38, 41, 85, 86, 108, 122, 126, 134, 139 2,4-DB, 33-38 2,4-DEP, 33, 34, 36 DMPA, 39 2,4-DP, 33 Dermatologic effects, 174, 175 Dicamba, 56, 57 Dioxins produced for research purposes, 75, 76 Disposal or destruction of dioxins Biological treatment, 58, 143-145 Catalytic dechlorination, 143 Chlorinolysis and chlorolysis, 142 Chloroiodide degradation, 140, 142 Concentration, 136, 138 Incineration, 132-134 Landfill ing, 59, 131-132 Microwave plasma, 136, 137 Molten salt combustion, 134-136 Ozonolysis, 140, 141 Photolysis, 138-139 Radiolysis, 139-140 Storage, 131 Wet air oxidation, 142 Dowlap, 57, 58 Embrotoxicity, 180-182, 194-197 Endocrine effects, 176 Enzyme effects, 154-157 Epidemiology, 190-199 Erbon, 46-48 Exposure, 77-97, 190-199 From foods, 86-88 From herbicide applications, 84-86 242 �From industrial accidents, 77-80, 191-194 From transportation accidents, 84 From waste handling, 80-83, 97 From water supplies, 87 In chemical laboratories, 96, 97, 191 In other related industries, 94-96 Occupational, 90-97 Fetotoxicity, 180-182, 194-197 Foods, dioxins in, 86-88 Gastrointestinal effects, 179 Gross and histopathologies, 159-169 Hematologic effects, 179 Hepatic effects, 175 Herbicide applications, 84-86 Herbicide Orange, 33, 36, 41, 82, 85, 96, 99, 100, 108, 111-114, 119, 120, 122, 133, 194, 198 Hexachlorobenzene, 59, 64-66 Hexachlorophene, 50-53, 77, 89, 97 Hexachlorophene exposures, 89, 90 Immunologic effects, 176-179 Incineration disposal methods, 132-137 Industrial accidents, 77-80, 90-94 Irgasan B5200, 57 Irgasan DP300, 57 Isopredioxin, 8 Lipids, effects on, 157-159 Metabolism, 149-159 Miscellaneous pesticide uses, 89 Mutagenicity, 188, 189 Neuropsychiatric effects, 179, 180 0-Nitrophenol, 67, 68 Particulate air emissions, dioxins in, 70-74 Pathophysiology, 189, 190 243 �Pentachlorophenol, 11, 12, 16, 17, 24, 25, 94-96 Pharmacokinetlcs and tissue distribution, 151"154 Photodegradation, 103-110 Physical methods of disposal, 136-140 Concentration, 136, 138 Photolysis, 138, 139 Radiolysis, 139, 140 Physical transport in air, 118 Physical transport in soil, 110-115 Physical transport in water, 115-118 Plastic, dioxins in, 74 Precursors, 3, 6, 8, 11, 12, 19 Predioxin, 8, 10-12, 57, 64 Renal effects, 175 Ronnel, 48, 49 Salicylic acid, 68-70 Sesin, 54, 55 Sesone, 36-39 Seveso, 77-80 Silvex, 43, 45, 46, 84-86, 115, 126 Smiles rearrangement, 11 Soils, persistence in, 98-102, 110-115 2,4,5-T, 40, 41-44, 51, 81, 82, 84-87, 108, 111, 115, 117, 122, 123, 126, 127, 139, 195 2,4,5-Trichlorophenol, 27-33 Uses, 27 Manufacture, 28-31 Production, 31-33 Teratogenicity, 180, 181, 194-197 Transportation accidents, 84 Triclofenol piperazine, 55, 56 Tyrene, 58 Water supplies, dioxins in, 87 244 �TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 3. RECIPIENT'S ACCESSIOr+NO. 2. 1. REPORT NO. EPA-600/2-80-156 4. TITLE A N D S U B T I T L E 5. REPORT DATE June 1980 Dioxins: Volume I. Sources, Exposure, Transport 6. PERFORMING ORGANIZATION and Control 7. A U T H O R ( S ) M. P. Esposito, H. M. Drake, J. A. Smith, and T. W. Owens CODE 8. P E R F O R M I N G O R G A N I Z A T I O N REPORT NO. 9. P E R F O R M I N G O R G A N I Z A T I O N NAME AND ADDRESS 10. PROGRAM ELEMENT NO. 1BB610 PEDCo Environmental, Inc. 11499 Chester Road Cincinnati, Ohio 45246 11. CONTRACT/GRANT NO. Contract No. 68-03-2577 12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED Industrial Environmental Research Laboratory Office of Research and Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 ' Final 14. SPONSORING AGENCY CODE EPA/600/12 15. SUPPLEMENTARY NOTES Volume I of a three-volume series on dioxins 16. ABSTRACT Concern about the potential contamination of the environment by dibenzo-p-dioxins through the use of certain chemicals and disposal of associated wastes prompted this study. This volume reviews the extensive body of dioxin literature that has recently become available. Although most published reports deal exclusively with the hiahly toxic dioxin 2,3,7,8-TCDD, some include information on other dioxins. These latter reports were sought out so that a document covering dioxins as a class of chemical compounds could be prepared. A brief description of what is known about the chemistry of dioxins is presented first. This is followed by a detailed examination of the industrial sources of dioxi ns Chemical manufacturing processes which are likely to give rise to 2,3,7,8-TCDD and other dioxin contaminants are thoroughly discussed. Other sources are also addressed including incineration processes. Incidents of human exposure to dioxins are reviewed and summarized. Reports on possible routes of degradation and transport of dioxins in air, water, and soil environments are characterized. Current methods of disposal of dioxin-containing materials are described, and possible advanced techniques for ultimate disposal are outlined. Finally, an extensive review of the known health effects of 2,3,7,8-TCDD and other dioxins is presented. This review emphasizes the results of recent toxicological studies which examine the effects produced by chronic exposures and also the various possible mechanisms of action for these toxicants. KEY WORDS AND DOCUMENT ANALYSIS 17. DESCRIPTORS Organic chemicals Pesticides; Herbicides Btddeterioration Toxicology Waste disposal 18. D I S T R I B U T I O N STATEMENT RELEASE TO PUBLIC EPA Form 2220-1 (9-73) b.lDENTIFIERS/OPEN ENDED TERMS Dioxins 2,3,7,8-TCDD Environmental biology Chemistry Health effects Hazardous wastes 19. SECURITY CLASS (ThisReport) Unclassified 20. SECURITY CLASS (Thispage) Unclassified c. COSATI Field/Group 07C 06F 11M 06T 13B 21. NO. OF PAGES 259 22. PRICE ft U.S. GOVERNMENT PRINTING OFFICE: 1980 -657-146/5719 245 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 183 Folder The folder containing the original item. 5331 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Esposito, M. P. H. M. Drake J. A. Smith T. W. Owens Description An account of the resource <strong>Corporate Author: </strong>PEDCo Environmental, Inc. Date A point or period of time associated with an event in the lifecycle of the resource 1980-06-01 Title A name given to the resource Dioxins: Volume I. Sources, Exposure, Transport, and Control ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 207 Folder The folder containing the original item. 5810 Series The series number of the original item. Series VIII Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Feinsilber, Mike Date A point or period of time associated with an event in the lifecycle of the resource December 7 1983 Title A name given to the resource VA's Agent Orange Tape Hit ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/b8d6f94b61b6d3e97a58be8159728340.pdf 4fb00cd3d900f16eacfd15c8efb485c9 PDF Text Text Item D Number °5312 Author Finch, Edward B. Not Scanned . ^- Environmental Protection Agency (EPA) Report/Article Title United States Environmental Protection Agency Before the Administrator, In re: The Dow Chemical Company, et al. Petitioners., Docket Nos. 415, et al., Order Granting in part and Denying in part Motion for Compulsory Document Production Against Dr. James Allen Journal/Book Title Year 1980 Month/Day February 1 Color D Number of Images 3 Descrlpton Notes Tuesday, March 05,2002 Page 5312 of 5363 �UNITED STATES ENVIRONMENTAL PROTECTION AGENCY u.-1" J • \ L lilt In re nDi'i.u- J i : _•. i O'~v ) The Dow Chemical Company et al. 80 FEB I A|Q: 4 ? Docket Mo. 415 et al. Petitioners. ORDER GRANTING IN PART AND DENYING IN PART MOTION FOR COMPULSORY DOCUMENT PRODUCTION AGAINST DR. JAMES ALLEN Motion filed by The Dow Chemical Company (Dow) in this proceeding for compulsory document production against Dr. James Allen is granted to the extent that it relates solely to the Schedule of Documents to be produced for: 1. 500 ppt TCDD Monkey 2. 50 ppt TCDD Monkey 3. 25 ppt TCDD Monkey 4. 5 ppt TCDD Monkey Study.. Study.; Study.i Study.' To the extent that Mr. John Van Miller participated in and may possess "documents" related to said studies, the Motion is also granted. Duly executed subpoenas, under seal, are attached hereto and directed to Dr. James R. Allen and also to Mr. John Van Miller. These subpoenas are issued pursuant to 40 CFR 164.70 which states that the Administrative Law-Judge shall be guided by the Federal Rules of Civil Procedure. In the sprit of Rule 45 thereof, I will entertain a motion to quash or modify the terms of either or both of said subpoenas if such motion is filed within the time prescribed in that rule, and such motion contains specific statements or declarations of Dr. Allen relating to: �1 L 1. Specify which studies and "documents have been supplied to Dow and the completeness thereof as relates to the Schedule of Documents to be produced attached to the subpoenas. 2. The ownership of the subject "documents." 3. The availability of each of the studies in published literature. 4. What disposition will be made of the "documents" upon the resignation of Dr. Allen? Will they be retained by Dr. Allen or remain in the possession of an investigator or co-principal employed by the University of Wisconsin? 5. Will Dr. Allen be available as a witness to these studies after his resignation? ._.'. 6. Which of the studies in question and the raw data relating thereto have not yet produced significant, reliable and accurate results and why? Give stage of completeness. The uncertainties which have arisen due to the fact that Dr. James R. Allen plans to resign his position with the University of Wisconsin, and to insure that all relevant material and probative evidence which may be available is brought to light at the commencement of the hearing while Dr. Allen is appearing as a witness, have had a large part in the decision to issue these subpoenas. I have considered the fact that some of the studies have been represented as being incomplete and may or may not be considered discoverable at this time, but without considering the issuance of the subpoenas as precedent for such discovery, I have concluded that the uncertainties which have arisen are an overriding factor. The Motion for Compulsory Document Production Against Dr. James R, Allen and Mr. John Van Miller as it relates to "Van filler — Allen Carcinogenicity Study with TCDD in Rats," is denied. — �Based upon an analysis of arguments and documents relating to the content and value of this study filed by the parties, it is concluded that while this study was completed, any reference thereto in this proceeding will not serve to assist the court in reaching a decision on the merits of this case, since nothing in the study is of such a substantive nature as to be probative in any way. Edward B. Finch Administrative Law Judge February 1, 1980 � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 182 Folder The folder containing the original item. 5312 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Finch, Edward B. Description An account of the resource <strong>Corporate Author: </strong>U. S. Environmental Protection Agency (EPA) Date A point or period of time associated with an event in the lifecycle of the resource February 1 1980 Title A name given to the resource United States Environmental Protection Agency Before the Administrator, In re: The Dow Chemical Company, et al. Petitioners., Docket Nos. 415, et al., Order Granting in part and Denying in part Motion for Compulsory Document Production Against Dr. James A ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 201 Folder The folder containing the original item. 5670 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Fingerhut, M. A. M. H. Sweeney W. E. Halperin Date A point or period of time associated with an event in the lifecycle of the resource 1986-04-01 Title A name given to the resource The Epidemiology of Populations Exposed to Dioxin, Draft ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 201 Folder The folder containing the original item. 5675 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Fingerhut, M. A. Date A point or period of time associated with an event in the lifecycle of the resource April 17 1986 Title A name given to the resource Review of Epidemiologic Data on Humans Exposed to Dioxin-Contaminated Substances ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/06251789affe8b41708c046ccfb49a5d.pdf 45dd2d6cafa2a7c6b98cff4ba5238cc4 PDF Text Text item n Number °5699 Author Fingerhut, Marilyn n Rot scanned Corporate Author Report/Article Title Report of the AOWG Science Subpanel, June 3,1986, Appendix VII: Utilization of Biological Samples to Assess Exposure to Agent Orange Journal/Book Title Year 000 ° Month/Day Color Number of Images D ° Descrlpton Notes Tuesday, March 26, 2002 Page 5699 of 5743 �REPORT OF THE AOWG SCIENCE SUBPANEL June 3, 1986 APPENDIX VII UTILIZATION OF BIOLOGICAL SAMPLES TO ASSESS EXPOSURE TO AGENT ORANGE Prepared by Marilyn Fingerhut, Ph.D. �Ultilization of Biological Samples to Assess Exposure to Agent Orange Recent advancements In the analytic sensitivity of laboratory Instruments have made it possible to analyze very low concentrations of 2,3,7,8-TCDD in samples of human fat (1). The results of several independent efforts (2-4) indicate that there is a background average level of 2,3,7,8-TCDD in human fat of approximately 7 parts per trillion (ppt) (range 0-20 ppt) . One study analyzed fat samples from volunteer Vietnam veterans (4). The results indicated that two veterans classified by the Veterans Administration as "heavily exposed" to Agent Orange had fat levels of 2,3,7,8-TCDD of 35 and 99 ppt. The remaining 10 veterans who were classified as "lightly exposed" and "possibly exposed" had levels between 3 and 13 ppt. Four veterans who had no service in Vietnam had levels between 4 and 8 ppt. The results of this study indicate that it may be possible to distinguish high exposure to Agent Orange by analysis of fat samples. The results also indicate that veterans classified as "lightly exposed" to Agent Orange have only background levels of 2,3,7,8-TCDD in their fat, the same levels as are found in the U.S. population in general. Analysis of fat is a difficult method for several reasons. A surgical or suction procedure is necessary to obtain 20 grams of fat (about the size of an egg) and the cost is about $1,000 per sample. Efforts are underway currently to analyze a large volume of serum (200 ml) to detect low levels of 2,3,7,8-TCDD. Data are also being sought which would describe the distribution of 2,3,7,8-TCDD between adipose tissue and serum in the human body. Success with the serum method would provide a method to recognize levels of exposure which were high enough to raise levels of 2,3,7,8-TCDD above background levels in the population. The recent advances in laboratory analytic techniques could be used to ascertain whether veterans in the various exposure categories of the CDC Agent Orange study have levels of 2,3,7,8-TCDD above the background levels in the population. For example, a sample of veterans currently meeting criteria for the CDC Agent Orange study category of "high likelihood of exposure" and a sample of veterans from the non-exposed category could be asked to provide fat (or possibly serum) specimens for analysis. An evaluation of the results should provide insight into the adequacy of the military records to select truly exposed and truly unexposed Individuals. Additionally, the results should indicate whether the levels of 2,3,7,8-TCDD are significantly different from the levels in the general U.S. population. Analysis of fat (or serum) from other populations could also provide valuable insights. Several studies are currently underway in which analysis of fat is being conducted on Vietnam veterans, chemical workers, and persons with residential and recreational exposures to 2,3,7,8-TCDD. Analysis of fat (or serum) could also be conducted on selected individuals in the CDC Vietnam Experience study who have known high or low levels of exposure. Samples of fat already collected from Ranch Hand participants during elective surgery could be analyzed and compared to the levels of exposure experienced by the individuals. -1- �1. Patterson DG, et al. High resolution gas chromatography/high-resolution mass spectrophotometric analysis of human adipose tissue for 2,3,7,8-TCDD. Anal. Chem 1986; 58:705-716. 2. Graham M, Hileman FD, Kirk D, et al.: Background human exposure to 2,3,7,8-TCDD. Fourth International Symposium on Chlorinated Dioxins and Related Compounds, 1984; Ottawa, Canada; October 16-18. 3. Graham M, Hileman FD, Wendlong J, Wilson JD. Chlorocarbons in adipose tissue samples. Fifth International Symposium on Chlorinated Dioxins and Related Components, 1985. Bayreuth FRG, September 16-19.2508R 4. Gross ML, Lay JO, Lyon PA, et al.: 2,3,7,8-tetrachlorodibenzo-p-dioxin levels in adipose tissue of Vietnam veterans. Environ Res 1984; 33:261-268. —2— � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 201 Folder The folder containing the original item. 5699 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Fingerhut, Marilyn Title A name given to the resource Report of the AOWG Science Subpanel, June 3, 1986, Appendix VII: Utilization of Biological Samples to Assess Exposure to Agent Orange ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 194 Folder The folder containing the original item. 5491 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Flaherty, Francis J. Source A related resource from which the described resource is derived The National Law Journal Date A point or period of time associated with an event in the lifecycle of the resource November 12 1984 Title A name given to the resource Congress Drops Bombs on Vets ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/99eadb14ab1e5b7b30bff65c6fff5abc.pdf 7f1a8f24befd50f7453732cdb7952295 PDF Text Text Item D Number osseo Author D NotScannBfl Fox R ' °9er p- Office of Air Force History, United States Air Force, Was RBpOrt/ArtlClB Title IV JOUmal/BOOk TltlB Air Base Year 1979 - The Target Air Bases Defense in the Republic of Vietnam, 1961 -197 Month/Day Color D Number of Images ° DeecrlptOlI NotOS ^'so included are two maps. One shows provinces and major cities of South Vietnam, and the other is a 1974 National Academy of Sciences Computer Printout of all defoliation missions in South Vietnam, 1965-1971. Tuesday, March 19, 2002 Page 5560 of 5611 �AIR BASE DEFENSE ^ IN THE REPUBLIC OF VIETNAM 1961 -1973 Roger P. Fox OFFICE OF AIR FORCE HISTORY UNITED STATES AIR FORCE WASHINGTON, D.C., 1979 �on maximum alert with particular attention to the defense of headquarters complexes, logistical installations, airfields, population centers, and billets." •* Sabotage Of the four threats posed by the VC/NVA to the local security of U.S. The enemy unleashed his main air bases, sabotage was the least sigattack between 0300 and 0400 local nificant. Despite unlimited opportunitime on 31 January with about 84,000 ties for sabotage afforded by the thoutroops. In addition to Saigon they as- sands of Vietnamese civilians working saulted 36 of the 44 provincial capi- on these installations, this classic tals, 5 of the 6 autonomous cities, 64 weapon of insurgency warfare was a of the 115 district capitals, and 50 curiosity rather than a commonplace. 242 hamlets. Responding to USMACV Records reveal but one notable case of alerting orders, the Seventh Air Force sabotage at an American base during Commander directed all bases to adopt the entire war. On 8 February 1967 Security Condition Red (Option 1), at Bien Hoa, Soviet-made explosive a readiness posture in which all base devices, secretly planted, destroyed defense forces were mobilized and about 2,600 napalm bombs valued at ST deployed to repel an impending at- $342,000. During 1968, a year of tack. Hence at both Bien Hoa and intense enemy activity, not a single Tan Son Nhut, the VC/NVA forces instance of actual or attempted sabofound themselves opposed at once by tage was reported at any Seventh Air 88 U.S. defenders.6* It was generally Force base. Why the VC/NVA all but ignored this simple and potentially agreed that this fact alone accounted for the successful defense of the two highly rewarding tactic cannot be explained by available evidence. bases. Security check The VC/NVA showed they could do serious damage to air bases, notably by standoff attacks and sapper raids. Much of their capability derived from a high degree of military expertise that reflected sound doctrine, meticulous planning and preparation, deeply instilled discipline, and an aptitude for fusing available manpower and weapons with proper tactics to produce a mission-effective force. Such ingenuity and skill helped surmount many of the inherent advantages of the defense and to retain a broad initiative to strike at grounddeployed U.S. air power at times and places of their choice. Accordingly, it is perplexing that the VC/NVA never sought to redress more vigorously the air power imbalance by fully exploiting their notable counterair base capabilities to the extent permitted by the vulnerability of Allied defense measures. IV. THE TARGET AIR BASES The majority of bases do not have a positive approach or active planning program for the protection of their operational assets. . . . There are no criteria established for the construction of air bases in a combat environment. New construction and redesigning is [sic] based on peacetime criteria. Seventh Air Force Base Defense Study Group, 1967. Major targets for Viet Cong/ North Vietnamese Army attacks embraced the 10 primary bases that supported USAF operations in Southeast Asia. Da Nang, Phu Cat, Tuy Hoa, Nha Trang, Cam Ranh Bay, and Phan Rang are located in the narrow coastal zone bordering the South China Sea. (See page 56.) Pleiku is situated in the Central Highlands less than 70 kilometers from Cambodia. Tan Son Nhut and Bien Hoa are in the environs of Saigon. Binh Thuy, the southernmost base, lies on the outskirts of Can Tho in the middle of the Mekong Delta. Florida, the country extends more than 1,300 kilometers from north to south, while its width from east to west varies from 50 to 200 kilometers. Saigon, usually considered an east coast city, lies less than 60 kilometers from the Cambodian frontier to the west. The Republic of Vietnam is a classic example of exposed territory. So lengthy are its boundaries in relation to its size, that points for infiltration by land and sea are almost unlimited—a circumstance fully exploited by the VC/NVA. The Ho Chi Minh Trail, stretching the whole The Geographic Impact length of the western boundary with branches extending into most interior Geography had a vital bearing areas, was their main route for inon all facets of the war. Its impact on filtration of men and materiel throughlocal ground defense of these bases came chiefly from the conformation, out the war. Secondary but much more topography, climate, and vegetation of limited infiltration occurred along the 1,300-kilometer sea frontier. Hence, the Republic of Vietnam. due in part to the physical conformaA geopolitical principle holds that tion of RVN, logistic support for VC/ a compact country is much easier to NVA operations against USAF bases defend than a large sprawling one. was available along well-established Clearly, the Republic of Vietnam fits lines of communication reaching from the latter category. Slightly larger than North Vietnam to within tactical M �striking distance of the target installations. Topography also favored the insurgency forces. Nearly 60 percent of RVN consists of relatively high mountains and plateaus rising to 2,500 meters. These mountains, the Annamite Chain, extend southeastward from China forming the border between RVN and Laos and, further south, between RVN and Cambodia. They terminate at a point in the Mekong Delta about 80 kilometers north of Saigon. Numerous spurs extending to the east insure broken and rugged terrain in close proximity to all USAF bases but Binh Thuy. Low- NORTH VIETNAM \ INFILTRATION ROUTES 1968 DEMARCATION LINE 102 1 ' Primary USAF Operating Bases NORTH VIETNAM DMZ TUYHOA AB NHA TRANG AB CAM RANH BAY AB . SIHANOUKVILLE LEGEND PERSONNEL ROUTES. LOGISTICAL ROUTES. BASE AREAS SEA INFILTRATION AREAS BY PRIORITY NOTE: CORPS TACTICAL ZONES WERE REDESIGNATED MILITARY REGIONS IN 1970. 56 57 �lands with little or no relief comprise the remaining 40 percent of the country and are located chiefly in the Mekong Delta where the land is seldom more than 4 to 5 meters above sea level and is intersected by numerous waterways. Consequently, almost the whole countryside offered cover and concealment to the VC/NVA while presenting obstacles to observation, penetration, and movement by RVN and Allied ground forces. Each of the 10 USAF primary bases was accessible by land and/or water to insurgency forces. Except in the mountains and plateaus of the Annamite Chain—for example the Pleiku AB area—high temperatures prevail throughout the year, the average annual range varying only from 77°F in the north to 81 °F in the south. There high temperatures accompanied by high humidity create a climate that saps human energy and enormously increases maintenance requirements for all equipment. As in other countries with similar climates, the afternoon siesta is an institution observed, except for U.S. forces, by friend and foe alike. It appeared that by tacit agreement mutual hostilities were suspended during the early afternoon hours. Except for about six standoff attacks during the Tet and May offensives of 1968, air bases were rarely threatened during siesta.* Annual average rainfall is heavy in all regions of RVN and torrential in many. It is heaviest in the Da Nang-Hue area with 128 inches. At Saigon it amounts to 80 inches. For most of Southeast Asia the rainy season occurs in the summer (JuneNovember), when an average of 10 typhoons off the South China Sea bring yet more rain. In the Da Nang * Siesta, it appeared, was the preferred time for launching a coup d'etat. 58 area the wettest period lasts from December through January. This heavy rainy season crippled Allied and VC/NVA operations alike and marked the yearly low point in attack on U.S. air bases.* Abundant rainfall joins the yearround high temperatures to give much of RVN a 12-month growing season that results in luxuriant vegetation. More than 80 percent of the country has a natural cover of rain forests, monsoon forests, and savanna lands, which provide extensive concealment for insurgents. Around and within the U.S. air bases, plant life flourished in overwhelming and unwanted profusion. Several varieties of grasses and weeds created a critical problem for base defense. Especially widespread is tranh grass which reaches a height of 1 to 2 meters, easily tall enough to hide a man or even to imperil a helicopter landing. Yen-bach, another common weed and a serious countrywide pest, grows from 1.25 to 1.60 meters. Lau, cane of frequent occurrence grows in clumps 2 to 3 meters tall. Also widespread are the bamboos, the most common of which, mai pha, occurs throughout Southeast Asia to form dense, almost impenetrable brakes that ascend 12 to 16 meters in height. Obviously, the height and density of such vegetation afforded ideal concealment for ambush and infiltration, f 1 * John F. Fuller, historian of the Air Weather Service, gives a good account of the impact of weather on military operations in his monograph, Weather and War (Hist Ofc, MAC, December 1974). t The botanical designations for these plans are: tranh grass (Imperta cylindrical, yen-bach (Eupatorium odoratum), lau (Saccharum spontaneum), and mai pha (Bambusa arundinacea). Vietnamese fishing village engulfed by dense tropical vegetation Effective vegetation control was made vastly more urgent and onerous by the year-long growing season and the exceptional growth rate. The latter was a truly incredible phenomenon and one on which information is surprisingly incomplete.2 Security Police at Tan Son Nhut recorded that vegetation grew 1V4 to 2V2 inches per day during the rainy season,3 an observation consistent with the findings of plant life specialists. A botanical study of one giant bamboo (Dendrocalmus giganteus) established that growth could occur as rapidly as 46 centimeters within 24 hours.4 Vegetation was probably least troublesome at Tuy Hoa where the entire eastern perimeter fronted directly on the South China Sea and where ground cover around the remainder of the circumference was the lighter variety common to savannas. The most extreme vegetation problem existed at Binh Thuy, the smallest USAF operating base. Situated in the center of the waterlogged Mekong Delta near Can Tho, it had an elevation of only .75 to 1.5 meters above mean sea level. The base was ringed by exceptionally dense tropical vegetation 3-4 meters high. This growth engulfed the perimeter fences constructed on the outer face of the levee that enclosed the installation. Likewise concealed were navigable canals, used occasionally by the VC/NVA to float munitions and weapons to the base perimeter. In the interior of Binh Thuy the same vegetation flourished. At other bases vegetation growth fell somewhere between the extremes represented by Tuy Hoa and Binh Thuy. At all bases, however, it was a permanent security threat that varied only hi the urgency of its impact.* So on the whole, the geography of RVN greatly favored the VC/NVA either directly by facilitating their military operations or indirectly by restricting activities of Allied forces. In the case of air base defense, the tactical unbalance was perpetuated and * The combination of dank vegetation and abundant rainfall created a breeding ground for mosquitoes and other disease-bearing insects. 59 �accentuated by other factors, notably decisions on location and layout of USAF operating bases. Location and Layout of Air Bases Among the most critical decisions affecting air base defense was the determination to make maximum use of existing airfields, however inadequate, in order to speed the introduction of USAF combat elements. The six bases in question were Da Nang, Pleiku, Nha Trang, Bien Hoa, Tan Son Nhut, and Binh Thuy. All dated from the French regime and all were located in or near population centers. Tan Son Nhut with its southern and eastern perimeters abutting metropolitan Saigon and with numerous villages and hamlets situated to the north and west was literally engulfed in a sea of humanity. Da Nang AB joined and shared the name of the second largest city in RVN. At Nha Trang the perimeter fence bordered upon civilian dwellings and often served as a clothesline. Nor were conditions radically improved at Phu Cat, Tuy Hoa, Cam Ranh Bay, and Phan Rang—bases expressly built for the USAF. All four were close to settlements of varying size. In the spring of 1969 a study compared an old and a new base in this regard. It found that clearing a 1-mile security zone around Bien Hoa would displace 13,998 people, 2,478 homes, and 555 shops. A like strip circling Tuy Hoa would expel 16,180 persons and Dong Tac, a refugee village newly erected by the Agency for International Development (AID)." Relocation of all people inhabiting air base approaches was probably the ideal technical solution to the defense problem. But politically it was out of the question, even though many of those concerned were squatters without legal title to the land they occupied. There was the unacceptable risk that those relocated would be alienated from the Government of Vietnam and converted to the VC/ NVA cause. Such an outcome would have simply aggravated an already unsatisfactory situation. As it was, problems of this nature faced the Air Force at Phu Cat, Tuy Hoa, and Phan Rang where construction had forced small landowners from their property. Many did not desire to sell in the first place, or feared that family graves might be disturbed or the land gods displeased. Some owners were underpaid or not paid at all. After waiting 2 years, former residents of Phu Cat petitioned the GVN to compensate them for their property. Such grievances created a receptive audience for VC/ NVA propaganda and bred a distinct antipathy toward U.S. forces.6 Tan Son Nhut Air Base Vietnamese so displaced posed fresh security problems. Former residents frequently desired to return to the base to worship at pagodas .left standing, to care for graves, to harvest tree or garden crops, or to tend to other affairs. Security personnel had to accompany the returnees and to search for boobytraps after their departure. At one USAF base under VNAF control, the faithful regularly came on the base without clearance or escort to visit a pagoda located near unguarded VNAF napalm stocks and ordnance-loaded aircraft.7 This episode will illustrate the exasperating and hazardous idiosyncrasies encountered in security operations at the six old airfields where VNAF had primary responsibility for base defense and exercised control over base access. Concentrations of civilian dwellings adjacent to the 10 USAF operating bases afforded the enemy an abso- lute tactical advantage since they provided cover and concealment to the threshold of the target base. These same conditions seriously restricted defense forces by prohibiting or limiting use of boobytraps, tripflares, sensors, freifire zones, and exclusion areas around, base perimeters. Also totally or critically curtailed was the delivery of artillery, aircraft, or helicopter counterfire. Thus, like the Allied conduct of the overall war, base defense operations were profoundly influenced by the necessity to enlist the widespread active support of the population. The USAF and VNAF buildup soon saturated the six older air bases to a point that invited enemy attack. Near the peak of the war, 76 percent of the total aircraft and 60 percent of all USAF aircraft operated from these more vulnerable airfields, whose target value was further heightened Aircraft Assigned To Primary RVN Bases 3 January 1969 RAAF VNAF USAF *Bien Hoa *Binh Thuy Cam Ranh Bay *Da Nang *Nha Trang Phan Rang Phu Cat *Pleiku *Tan Son Nhut Tuy Hoa Total 220 52 117 59 158 47 110 141 90 48 77 105 97 301 1,138 75 43 8 USA USN USMC Total 515 220 59 89 22 2 69 24 69 48 416 * Older bases. SOURCE: USAF Management Summary Southeast Asia, 3 Jan 69, p 39. �by large stores of munitions and aviation fuel. At many of them, conditions were further aggravated by the presence of major military headquarters and/or key political facilities. The ARVN II Corps was at Pleiku and the USMACV I Field Force Vietnam (FFV) at Nha Trang. Da Nang hosted the ARVN I Corps and the III Marine Amphibious Force. , But in this respect Tan Son Nhut was unique. It not only supported an aerial combat mission but housed the headquarters of the Vietnamese Air Force, Seventh Air Force, and United States Military Assistance Command, Vietnam.* The base was also Saigon International Airport and in 1965 became the VNAF induction center. For much of the time, it served as the residence for the RVN premier or vice president. Location at the seat of government gave Tan Son Nhut a farreaching political and psychological importance as a military target. Population saturation was noted as early as August 1965 in an Air Staff report which stated that the base was designed for 3,000 people but had 25.000.7 An April 1968 estimate placed the permanent population at * USARV Headquarters was housed at Tan Son Nhut until it moved to Long Binh in July 1967. At that time, USMACV Headquarters relocated most of its activities from various points in Saigon to the newly built "Pentagon East," situated on Tan Son Nhut near the Saigon International Air Terminal. 25,000 but that the influx of daily workers and military members living off base raised this number to 55,000 during duty hours.8 Overcrowding .seriously degraded security at the older bases. As congestion mounted, new combat-support facilities for the expanding aerial mission had to be sited solely on the basis of unoccupied real estate without regard to security factors. Dispersal to protect parked aircraft was impossible due to lack of space to enlarge or decentralize the ramps. At Tan Son Nhut, Da Nang, and Pleiku aviation fuel tanks and bladders were sited within 10-30 meters of the base perimeter. On every older base except Da Nang, munitions were stored in equally exposed locations.8 The USAF tenant status greatly complicated these troubles. As host, the VNAF insisted on exercising approval authority over all new construction. Thus a command change like that at Tan Son Nhut in early 1966 often necessitated renegotiation of many planning actions previously approved by the former commander. Agreements were also subject to cancellation for routine reasons. As one USAF base civil engineer plaintively observed, "Boy it's discouraging to get a project all set to go and then have the host say 'Sorry about that, you'll have to put it some place else.'" Usually no alternative site was offered or, if one were proposed, it was invariably in the rice paddies and required extensive fill before use.10 The task of unsnarling •<•*•%. v View of the crowded flight line at Tan Son Nhut Vulnerable fuel storage bladders adjacent to the Pleiku Air Base perimeter these tangles fell to the base engineer, one of the much-abused heroes of USAF deployment to RVN. Because there were no USAF criteria for constructing air bases in a combat area, peacetime standards governed the design of Tuy Hoa, Cam Ranh Bay, Phan Rang, and Phu Cat.11 Some of the more glaring drawbacks of this approach showed up in the siting and configuration of these bases. Perhaps from a location standpoint, Phan Rang was the most vulnerable because it received its water and aviation fuel from offbase sources through pipelines exposed to enemy interdiction.* In contrast, a peninsular site made Cam Ranh Bay the most defensible base in the Republic of Vietnam. Critics, however, leveled their sharpest barbs at the internal layout of the four new installations. Security police officials, themselves partly to blame for the lack of proper planning guidance, pointed out that although the bases had "ample real estate to permit * Of the older bases, Pleiku and Binh Thuy also relied on vulnerable offbase sources for water. [Final Report, 7th AF Base Def Study Gp, 17 Aug 67.] the locations of critical resources consistent with optimum security/ defense criteria . . . this was not done." As a consequence, they asserted, vital resources and facilities were without exception sited at vulnerable locations or were so positioned that excessive manpower were required for their protection.18 Munitions were stored in the northwest and aviation fuel in the southeast corner of Phan Rang, both within easy small-arms range from the base perimeter. At Cam Ranh Bay combat essential facilities were so scattered that additional multiple guardposts were created. The security police claimed that a little forethought in planning could have incorporated dispersal into the general scheme while grouping resources in a tighter-knit layout that would have reduced manpower, increased security, and simplified defense operations.19 The siting of noncritical facilities also impaired base defense. For example, at Tuy Hoa a raised railroad bed along the south and west perimeters afforded excellent cover and concealment to enemy forces approaching from the rice paddies in these areas. And base defense forces launching a 63 �counterattack were placed at a disadvantage, since the flat terrain from the track inward provided no cover against an enemy operating from the shelter of the embankment.14 As these and other incongruities reveal, new bases were located and laid out with scant concern for security. Active Defense Facilities, 1961-1972 After siting and layout, the most critical physical element in base defense operations was the status of security facilities—fences, barriers, lighting, sensors, minefields, towers, bunkers, and roads. But, from 1961 to 1965 USMACV viewed base defense as a primary responsibility of the overextended and hard-pressed RVNAF. Therefore the USAF did little more than post a few interior guards around parked aircraft and/ or base billets, and file periodic reports on the unsatisfactory status of security safeguards. As early as November 1961, the Farm Gate Commander at Bien Hoa informed CINCPACAF of security problems posed by uncontrolled vegetation and the need to lay "adequate concertina wire and mines throughout the perimeter."" During 1962 a USMACV survey rated Da Nang's perimeter fence as inadequate.16 In anticipation of VC/NVA reprisals for U.S. air raids on the DRV, USMACV and 2d Air Division in late 1964 jointly inspected the physical defenses at Tan Son Nhut. This inspection revealed that the base perimeter fence—none too sturdy when new—was in an advance state of deterioration. There were improvised gates and numerous holes which permitted uncontrolled access by civilians and military dependents. Threequarters of its length was overgrown by foliage so dense that a companysize unit could have infiltrated undetected. Minefields laid in 1957 along some sections were not chartered or maintained, and livestock grazed in allegedly mined areas. No perimeter lighting system existed, and from 40 to 50 percent of the 18-kilometer perimeter was neither under surveillance nor covered by fire, due to the distance between observation posts and bunkers.17 As Tan Son Nhut was the most prestigious air base in RVN, its defenses were likely the best to be found. USAF assumption of responsibility for base defense facilities dated from December 1965 when COMUSMACV directed 2d Air Division and all other Service components to initiate measures for the local defense of their RVN bases.18 Southern perimeter of Tan Son Nhut Air Base. In May 1968 the VC/NVA attacked the base through this area, abetted by the overgrowth on the fences and the close proximity of private dwellings Progress was halting and meager. After 18 months, a detailed survey by a Seventh Air Force Base Defense Study Group in the summer of 1967 reported widespread defects in physical security safeguards.18 Of the 10 primary air bases, Da Nang alone boasted both permanent perimeter fencing and lighting systems installed by the USMC in early 1966. This double cyclone-type fence was the only one of its kind at RVN air bases.* At Tan Son Nhut a new but * The French considered the best obstacle a vertical fence, 2 meters high, imbedded 40 centimeters into the ground to prevent tunneling, made of barbed wire with a maximum mesh of 20 cm, and equipped with a conventional doubleapron fence at its base. [V. J. Croizat, trans, A Translation from the French: Lessons of the War in Indochina (RM5271-PR, The RAND Corp, May 1967), 11 138-39.] Mine field on the perimeter of Phu Cat Air Base less durable perimeter barrier complex had been installed at the direction of COMUSMACV, after the 4 December 1966 sapper raid. It consisted of from one to three lines of triple-tier concertina wire, minefields, and permanent lighting.20 Both Da Nang and Tan Sori Nhut possessed good observation' towers And fighting bunkers. Elsewhere the ' picture was bleak. Perimeter sighting was unsatisfactory at six bases; fencing was inadequate at 2; minefields were not utilized at 4; and bunkers were inadequate or unsafe at 5. By February 1969 Phu Cat and Tuy Hoa were still "aggressively pursuing" fencing programs. Phu Cat had constructed a single line of triple-tier concertina wire along 16 kilometers of its main line of resistance (MLR),* but its perimeter fence remained in the programming stage. Tuy Hoa's perimeter was 68 percent fenced, but the beach area was still unenclosed. Plans for a perimeter fence at Cam Ranh Bay were abandoned due to scope, configuration, and soil conditions, and an approved fencing project was confined to the MLR alone.21 Perimeter * A line at the forward edge of a battle position, designated for the purpose of coordinating the fire of all units and supporting weapons. It defines the forward limits of a series of mutually supporting defensive areas. �Base control tower and "big light" used in defense of Phu Cat Air Base $1,090, and a requirement of 100 for a single base was not unreasonable. But the initial outlay was only the beginning. Not designed for continuous 8- to 12-hour daily operation, these units required daily maintenance service, a task which at a large base employed two airmen full time. The NF-2s were also vulnerable to smallarms fire, and the loss of a single unit darkened that segment of the perimeter it serviced.22 lighting continued to lag at five bases. Thirty-two percent of Tuy Hoa's perimeter was unlighted. As with fencing, the lights programmed for Cam Ranh Bay were limited to the MLR. Procurement delayed Phan Rang's permanent lighting system, and the one planned for Bien Hoa in July 1969 was never installed. A basic obstacle to adequate security lighting was a chronic shortage of electricity from sources both on and off base. In most cases, therefore, installation of a permanent perimeter lighting system included an organic power source. Field expedients were widely used as substitutes. These makeshifts ranged from mobile Fresnel units to jury-rigged flares that had been condemned for aerial use. However, the most common interim answer was the NF-2 Light-All unit. One generator fed up to 10 floodlights spaced along 100 meters of perimeter. Each NF-2 unit cost 66 Hand-held slapflares* and 81-mm mortar illumination rounds supplemented lighting at all bases and constituted the primary source at some. Air-dropped flares routinely enhanced these ground efforts. In April 1969, Seventh Air Force reported to PACAF a monthly cost of $81,000 for slapflares and $100,000 for mortar shells.t M At best, none of these interim solutions, even coupled with sophisticated night observation devices, provided more than a bare minimum level of lighting. It was asserted that "the cost of aircraft destroyed by sappers at one base [Tuy Hoa] in July 1968 would have been sufficient to ade* A slapflare looked like a paper towel cylinder with a cap on the bottom. The steps for igniting were to remove the cap, hold the flare in the left hand, and slap the bottom with the right hand. t Cost data on the air-dropped flares was not available. Sandbag bunker at Cam Ranh Bay Air Base quately fence and light all our bases in RVN.* 24 Construction of fighting bunkers was equally troublesome. Experience and experimentation led to the use of a wide assortment of materials and designs. Initially bunkers of sandbags were nearly universal. But deterioration due to weather and hard usage normally necessitated replacement of the bags every 90 days and created a monumental work load. Waterproofing was not feasible and all timbers were vulnerable to rot and termites. Accordingly, the trend was to replace sandbags with more durable materials. By 1968 each base had for the most part produced a bunker best adapted to local conditions. The French had found the ideal to be a facility of permanent construction and low silhouette. At Cam Ranh Bay, however, the shifting sands rendered this type undesirable. And at Binh Thuy, because of the high water table of the delta, bunkers had to be built above ground. Accordingly, building materials adapted to varying conditions and terrain, but most bunkers * This sapper raid on Tuy Hoa on 29 July 1968 resulted in 2 C-30s destroyed, and 5 C-130s, 1 C-47 and 1 F-100 damaged. Four USAF personnel were wounded. (7AF/IGS WEINTSUM, No. 68-13, 27 Jul-2 Aug 68, p 23) were designed to withstand a direct hit by a B-40 rocket. Most but not all bunkers at the bases had some type of overhead protection. All enjoyed a standoff weaponry screen, usually cyclone or other heavy fencing. Placed 3-4 meters forward of the bunker, the screen predetonated rocket propelled grenades.25 In the spring of 1969, bunker construction was least advanced at Phu Cat and Cam Ranh Bay. At the former limited fire necessitated shifting bunkers from perimeter sites to the MLR, where in conjunction with the planned fencing and lighting, they would contribute to a sound defense complex. At Cam Ranh Bay bunker construction was deferred pending action on programmed MLR fencing and lighting. After 4 years of massive USAF involvement, physical safeguards in 1969 were still judged inadequate by the Director of Security Police, Seventh Air Force. This was attributed to profound USAF disinterest as reflected by the lack of an active planning program and the absence of any criteria for air base construction in a combat area. General apathy and indifference were only intermittently dispelled by a near-disaster such as the 1968 Tet Offensive, or by a destructive sapper raid like that on Tuy Hoa (57 �in July 1968. Contributing to the problem was the continuous turnover of commanders at all echelons. New commanders not exposed to enemy attack usually stressed more spectacular but less vital construction. Highly visible recreation facilities received top priority while defense works at obscure or remote locations were ignored. For example, at the time of the Tuy Hoa sapper raid the perimeter was only partially fenced and totally unlighted. Yet, a year before, the base had been equipped with air-conditioned recreation facilities that included a base exchange, open messes for officers and noncommissioned officers, a library, and a recreation center. The latter offered a poolroom, reading room, and complete snackbar. Under these conditions which prevailed at all bases, security police undertook the construction of security safeguards as a selfhelp project with a corresponding degrading of their primary security mission capability." By 1970 construction projects in support of base defense had been overtaken by events. Shortly after assuming office in January 1969, President Richard M. Nixon decided to Vietnamize the war and to begin the phased withdrawal of U.S. forces from RVN. His decision was swiftly reflected in such actions as the Nha Trang Project which, begun in 1969, aimed at early USAF relinquishment of that air base to VNAF.27 Consistent with this policy, the Secretary of Defense refused Military Construction Program (MCP) funds for the perimeter fence at Phu Cat. Because concertina wire was an expendable item, he recommended that construction be accomplished with Operation , and Maintenance (O&M) funds.28 This policy was soon extended by USMACV to other security fence projects. Seventh Air Force instructed base commanders to draw fencing material 68 through base supply and install it by self-help.20 At the same time security lighting requests were also deleted from the MCP with the recommendation that they be resubmitted in the O&M Program, "selecting the most critical area for accomplishment within the $25,000.00 limitation." 80 Clearly, for all practical purposes, USAF construction of physical safeguards at RVN air bases was at an end. Passive Defense Facilities, 1961-1972 Passive defense facilities directly complemented the physical security safeguards of active defense operations. Their purpose was to reduce the probability of and to minimize the damage from enemy action without taking the initiative. In RVN such facilities consisted chiefly of shelters, revetments, and hardened structures installed to protect USAF personnel and resources not engaged in a base defense mission. From 1961 through 1965 the only USAF passive defense construction to speak of was the erection of aircraft revetments. The stimulus for this program came initially from the necessity to reduce explosive safety hazards arising from wingtip-to-wingtip parking a bomb-laden aircraft. On 16 May 1965 at Bien Hoa, an accidental explosion aboard a B-57 triggered a series of blasts that killed 28 and injured 77 people. The aircraft toll reached 10 B-57s, 2 A-2Hs, 1 A-1E, and 1 F-8U destroyed, plus 30 A-lHs and 1 H-43 damaged. Also demolished were 12 pieces of aerospace ground equipment (AGE), 10 vehicles, and the JP-4 fuel dump. This one incident was more destructive than any single VC/NVA attack on any air base during the entire war.81 It resulted in a USAF directed emergency program for revetment construction. F-100 Super Sabres parked in aircraft revetments at Tan Son Nhut Air Base For revetment construction the Air Force chose a prefabricated facility, developed by the Air Force Logistics Command (AFLC) and produced by the American Rolling Mill Company (ARMCO). It consisted of earth-filled corrugated steel bins 12 feet high and 5.5 feet wide. Built up on three sides of an aircraft hardstand, the bins afforded considerable protection against such dangers as near-miss blasts, secondary explosions, fragmentation effects, surface ordnance, and secondary damage and proliferation. Three 28-man Prime Beef* teams were deployed to RVN to do the work, the first one arriving in August 1965. Aided by troop and * Prime Beef (Base Engineer Emergency Forces) are worldwide base civil engineer forces. They are organized to provide trained military elements, used in direct combat support or emergency recovery from natural disaster. local-hire labor, they erected 12,040 linear feet of revetments at these bases by the end of the year.82 Tan Son Nhut Bien Hoa Da Nang 4,700 3,800 3,540 During 1966 through 1969, USAF interest in passive defense facilities continued to center chiefly on aircraft revetments which totalled 506 at all bases by 30 June 1967.83 However, the Seventh Air Force Base Defense Study Group reported on 17 August the improper siting of many revetments. Explosives-laden aircraft stood face to face, their forward-firing weapons pointed toward maintenance facilities or other planes. The study group asserted that this arrangement severely curtailed protection against blast or fragment damage, and could not prevent an explosive chain reaction from aircraft to aircraft. Of the �then owned about 1,000 revetments and 373 shelters for a total 1,373 protective structures. This number compared favorably with the 1,164 USAF aircraft permanently assigned at that time to RVN air bases.88 Damaged revetments at Bien Hoa Air Base following an attack In June 1969 10 primary bases, Bien Hoa alone had positioned its revetments so that the bay opening of one faced the rear wall of another.84 The corrective action action recommended by the study group was rejected by Gen. William W. Momyer, Seventh Air Force Commander, because "we are too far committed to change now. Cost in time and manpower is prohibitive."8B Static aircraft protection embarked on a new phase in 1968 as the Air Force launched a crash shelter construction program. The switch from revetments to shelters stemmed from the VC/NVA spring offensive when standoff attacks had destroyed 25 (valued at $94 million) and damaged 251 USAF aircraft. These strikes bared the weaknesses of revetments, mainly the absence of overhead cover. The adopted shelter design called for a double corrugated' steel arch with a poured-in-place concrete cover 18 inches thick. An added freestanding backwall extended protection equal to the cover's and included an opening to let out jet exhaust. A small 70 number of the shelters were also fitted with a front closure device. Production of materials began in CONUS in mid-1968, and the first concrete cover was poured in RVN in October 1968. Civilian contractors such as RaymondMorrison-Knudson and Brown-RootJames (RMK-BRJ) erected a few of these shelters. But USAF civil engineer Red Horse* squadrons augmented by troop labor built the majority.86 In contrast to revetments, siting of shelters received careful consideration. Wherever possible they were placed nose to tail with the front ends oriented away from the most likely direction of a ground attack.87 The capping of the last shelter at Tuy Hoa on 13 lanuary 1970 completed the program. Seventh Air Force * Red Horse (Rapid Engineer Deployment, Heavy Operational Repair Squadrons, Engineering) are controlled by Headquarters USAF. They give the Air Force a highly mobile, self-sufficient, rapidly deployable civil engineer capability required in a potential theater of operations. and no one had come up with a way to exit quickly from the unprotected upper floors. Quarters of key personnel were equally unsafe, and working areas were unsheltered.40 Popular response to these exposed conditions were echoed in these earnest lines: The protection afforded aircraft by hardened shelters confirmed the soundness of the program. Responding to a PACAF query, Seventh Air Force on 3 June 1969 cited two cases in which aircraft parked in shelters escaped destruction by direct rocket hits. On another occasion shelters saved several aircraft from damage or destruction when a nearby munitions storage area exploded. In spring 1970 a USN EC-121 crashed and burned at Da Nang, but adjacent hardened shelters saved three USAF F-4Ds from destruction and two others from major damage. The estimated dollar savings attributed to shelters in these incidents more than paid for the $15.7 million program in RVN.89 I arrited at i Da, Nang and my heart felt a pang ' As I viewed my new home for the year For the sheetmetal top, I was told would not stop The rockets intended for here. Men, like aircraft, were for much of the war without safe shelter. Inspection by the 1967 Seventh Air Force Base Defense Study Group found personnel bunkers unroofed and in disrepair. They were often too dispersed to give real protection. Revetment construction to safeguard the lower floors of barracks was slow, Brick revetments constructed about billets at Pleiku Air Base to protect against shell fragments. Such revetments were useless against direct hits When the sirens go off, or the rocket tubes cough "Get under your bed!" reads Directive But try (and I strive), I can't stop the drive To seek shelter a bit more protective." The steps to a final solution of the barrack-revetment problem were drawn-out and wasteful. Initially revetments consisted of earth-filled sandbags, stacked to a height and thickness necessary for protection and stability. These bags as a rule deteriorated within 90 days and were replaced with new earth-filled ones. As local conditions stabilized and further replacement was required, plywood shells packed with earth took the place of the sandbags. These wood revetments also �rotted, and the substitute became brick or concrete materials that lasted for the useful life of the facility protected. By 1968 precast concrete slabs were adopted as the least expensive revetments for both personnel and equipment. A forklift operator and a welder were the only skilled labor required to erect them.42 Concrete slab revetments promised impressive savings. At Da Hang, for example, more than 40,000 linear feet of sandbag revetments shielded barracks and operational facilities. An estimate showed that replacement of sandbag revetments by concrete slabs around the barracks alone would save $521,340 in 1 year.43 The delay in protecting essential facilities and services matched that in sheltering personnel. Again, in the absence of combat construction criteria, most bases made no plans for such protection. For example, in 1967 all bases were constructing centralized electric powerplants, but only Cam Ranh Bay had a protection plan for this facility. Even at that base, less than 25 percent of all mobile and alternate generators—those used chiefly for ground controlled approach (GCA) and other navigational aids—were protected. Disregarding the principle of dispersion, alternate generators were frequently located next to primary power sources.44 USAF munitions storage areas— priority enemy targets—were adequate at all bases except Pleiku and Binh Thuy. However, those of the VNAF were substandard at each base, save Bien Hoa. Large unprotected quantities of munitions cluttered every VNAF parking ramp, a serious hazard to USAF personnel and resources. Barring the bases of Tan Son Nhut, Phan Rang, and Cam Ranh Bay, munitions at aerial ports awaiting shipment had little or no protection. 72 Storage was either on or immediately adjacent to aircraft parking areas.48 Security of petroleum storage tanks—also priority enemy targets— needed upgrading. Other than at Tan Son Nhut, the protection of these storage tanks was after the fact. It relied on earthen dikes to contain escaping fuel and head off a holocaust. When rockets struck Da Nang on 27 April 1971 and Cam Ranh Bay on 25 May, the dikes let firemen limit the blaze to tanks taking direct hits.46 On Tan Son Nhut the tanks belonged to commercial petroleum companies who encased them in costly masonry shells. The wisdom of this move was doubtful, due to the high silhouette of the tanks and the deep penetration of rocket propelled grenades. Fuel storage in rubber bladders became widespread in South Vietnam. Often set adjacent to aircraft hydrant fueling systems, the bladders posed a grave fire hazard. No shielding from blast or fragmentation existed for most aircraft maintenance and civil engineering control centers, supply control systems using UNI VAC 1050 computers, and base command posts and communications centers. Fire and crash vehicles crucial to damage control were normally parked in rows at one central open area on each base. Few bases had any plans to disperse this critical recovery equipment. None provided a hardened parking area. Dikes constructed to protect petroleum supplies at Tuy Hoa Air Base storage points at Cam Ranh Bay were situated in the fighter aircraft area, a choice target for enemy attack.47 •The stimulus given passive defense by the 1968 Tet Offensive carried over into 1969. But this momentum focused almost exclusively on protection of aircraft with only limited attention to personnel and facilities. As the year wore on, the program began a gradual phaseout, owing to the decision to begin withdrawal of American forces and the cutback in funds for RVN operations. With the completion of the last hardened aircraft shelter on 13 January 1970, significant USAF passive defense construction in RVN came to an end. Thereafter, general policy was to perform minimum maintenance on the minimum number of existing facilities needed to protect the diminishing USAF forces. Vegetation Control Water sources, purification equipment, and storage points were unprotected at all bases. Pleiku, Phan Rang, and Binh Thuy depended on water from vulnerable offbase sources. Several bases put in fire hydrant systems, but only Bien Hoa had dispersed emergency water storage. Two water No element of the Vietnamese environment was more detrimental to base defense than the invincible ground cover described earlier. This rampant vegetation hid the enemy, shut off friendly observation and fields of fire, neutralized fencing and other defense barriers, slowed security forces, and nullified detection by sentry dog teams. The need to control this jungle was evident and urgent— how to do it was the sticking point. Clearing approaches to the base was the first order of business. This meant defoliating a zone around the outside circumference of the installation, an area outside the Air Force's accepted defense responsibility. Hence it became the task of the Allied ground commander whose TAOR was confined to the base. Actually internal and external security overlapped in this zone, creating a joint and at times unequal interest in common defensive measures. This diffusion of military responsibility and the necessity for political clearance vastly diminished the prospects of winning approval for any defoliation program. Another critical area calling for the most complete defoliation was the air base perimeter. Here physical factors crippled or canceled out progress. From the outset the six old bases took security steps, and the four new bases followed. These safeguards embodied fencing, tactical wire, minefields, and tripflares set in divers numbers and mixes along the perimeter. The skill of the VC/NVA sapper in clearing manmade obstacles and in disarming 73 �explosives devices dictated that this complex be kept free of concealing vegetation. Ignoring the French experience, the USAF discovered anew the problems associated with defoliation of the perimeter barrier system.48 Rarely if ever charted, the minefields of the perimeter barrier prohibited use of manual labor to cut and remove the vegetation. The mines, fencing, and wiring prevented mowing or scraping by mechanized equipment. Burning was unsatisfactory on several counts. Vegetation was highly fire resistant, particularly during the rainy season when growth was most rapid. It ignited slowly, even if sprayed with a flammable such as contaminated jet fuel. Because fire hardly ever consumed the vegetation, the residue went on obscuring the barrier system and offering cover to penetrators. Burning also detonated or destroyed mines and flares within the complex. Next in importance was defoliation of the base interior. Here too, the ideal was to clear the ground cover that concealed penetrators and reduced surveillance by defense forces. For example, the defense vegetation ne- gated sentry dog detection—the base's most reliable alarm. And the exertion in plowing through this thicket sapped dog and handler. Because the interior was without the perimeter's hazards or obstructions, it seemed that the clearing methods mentioned earlier could be given full play. In practice this was not the case. Safety factors forbade burning in or near fuel and munitions storage areas. The immense labor entailed in clearing a sizable area in a reasonable time curtailed manual cutting. Cutting by hand nonetheless left the root system intact, and so was well-suited to Cam Ranh Bay's very unstable soil. Elsewhere, however, an undisturbed root system meant rapid regrowth of vegetation. Even though scraping served well in the base interior, the conventional USAF civil engineer squadron usually lacked the needed mechanized equipment. In light of these facts, the answer to vegetation control in the interior as on the perimeter appeared to be herbicides. By the time the Air Force turned to herbicides for base vegetation control, they were in full-scale military use in support of other ground operations. The dispensing of defoliants centered on foliage along thoroughfares to deny the enemy ambush cover. Spraying also focused over VC/NVA camps and assembly areas, as well as over crops intended for feeding the foe. The acreage treated with agents from the 1,000-gallon tanks of USAF UC-123 (Ranch Hand) aircraft rose from 17,119 in 1962 to 608,106 in 1966.« None of these herbicides was believed to endanger humans or animals. All had been widely used in the United States for more than 20 years on foods and other crops, rangeland, and forests. None persisted in the soil and periodic respraying was required to kill regrowth. All were liquids. Those dispensed in RVN were designated Orange, White, and Blue. Appendix 5 gives general data on their composition, application, effect, and safety precautions. The use of these herbicides was a GVN program supported by the United States. The U.S. Ambassador and COMUSMACV acted jointly on GVN requests for herbicide operations on the basis of policy formed by State and Defense Departments and approved by the President.60 Senior U.S. Army advisors at ARVN corps and' division level were delegated authority to approve requests in which dispersal of the herbicides was limited to hand or ground-based power-spray methods. A herbicidal defoliation request from a USAF air base was prepared and documented by the base civil engineer, using a set checklist. (See page 77.) It was then processed through U.S. military channels to the senior U.S. Army headquarters in the corps tactical zone. If approved there, it was sent on to the ARVN commanding general of the same CTZ for military approval and political clearance. It was at this point that delay most frequently occurred, due to opposition from the district and/or province chief. These officials were influenced by such things as superstition, concern for local crop damage, and possible propaganda value to the VC/NVA. Final action on requests for ground-delivered herbicides was taken at this level. If aerial delivery was desired, the request could only be approved at USMACV/JCS level. A C-123 sprays defoliation chemicals over South Vietnamese jungles �Technical factors also entered into the dispensing of herbicides. Dry weather was essential, because rain quickly washed chemicals from the target vegetation to nearby crops and other desirable growth. Ideally, spraying was done between dawn and 1000, at ambient temperatures under 30° C (86° F), and in calm or very low wind conditions to minimize drift. Storage and mixing points had to be kept to a minimum, isolated from cultivated areas. Empty herbicide drums required close control to avoid accidental contamination."1 Approval and execution of herbicidal defoliation projects were time- consuming and uncertain. In February 1968 Phan Rang requested defoliation of a 200-meter strip both inside and outside the perimeter, around the entire circumference of the base. The approving authority reduced the scope of the project to one-half the perimeter. In addition, problems in obtaining herbicide and other obstacles delayed 'completion of the project for 1 year." ' Excessive vegetation at Tan Son Nhut and Bien Hoa hindered the base defenders throughout the 1968 Tet attacks.53 At Bien Hoa the approval process for aerial defoliation was termed "hopelessly complicated," one Checklist for Defoliation Requests 1. Overlays or annotated photographs depicting the exact area. 2. Target list: a. Area—province and district. b. UTM coordinates. c. Length and width. d. Number of hectares. e. Type of vegetation. 3. Justification: a. Objectives and military worth. b. Summary of incidents. 4. Psychological warfare annex (prepared by sector): a. Leaflets. b. Loudspeaker texts. 5. Civil affairs annex (prepared by sector): a. No crops within 1 kilometer. b. Contingency plan to provide food or money to families whose crops are accidentally damaged by the defoliation operation. 6. Certification by province chief: a. Province chief approval. b. Indemnification will be made by the Republic of Vietnam for accidental damage to crops. SOURCE: Lib of Cong Rprt, 8 Aug 69, to the House Subcommittee on Science and Astronautics, 91st Cong, 1st sess, A Technological Assessment of the Vietnam Defoliant Matter: A Case History, p 19. 77 �that might take two or more months. Plant growth meanwhile continued unabated. Even when authorized, a project was apt to be fettered with restrictions. Thus aerial delivery of Orange was denied at Bien Hoa, and only parts of its perimeter were approved for chemical defoliation. Accordingly, because Blue and White were not suited to local conditions, Orange had to be 'dispensed from a tank truck by a power spray that did not reach beyond the second fences. Local terrain made it impossible to go outside the third and fourth fence and spray inward.84 As noted earlier, Binh Thuy faced the most extreme defoliation problem. Here the one herbicide approved for use was Blue, which killed only those portions of plants with which it came in contact. With the root systems left intact, regrowth was rapid. In 1 month, 2,420 gallons of Blue valued at $22,000 were sprayed over limited areas of the interior and a narrow zone around the perimeter of the 550acre installation without making any significant inroads against the teeming vegetation.85 Herbicides for air base defense seldom if ever improved the horizontal view at installations by the desired 40 to 60 percent.56 Defoliation needs of the 10 primary bases were specific, permanent, and known in advance. Still no ongoing long-term program to satisfy them was ever set up. Instead the job was done piecemeal, with each base handling defoliation requests. Despite administrative and technical controls, chemical agents remained the single sure way to control vegetation in places where other means could not—notably in the critical perimeter complexes. As the war drew t6 a close, however, curbs on the use of herbicides grew more and more rigid. The last herbicide mission by fixed-wing aircraft was flown on 7 January 1971. 78 On 1 May, a presidential directive ended all U.S. herbicide operations.57 In the ensuing months, mines killed eight and injured seven Army personnel who were trying to clear vegetation by hand from wire entanglements and fields of fire.58 With the Ambassador's full backing, COMUSMACV urged Washington to alter at once the ban on chemical herbicides because immediate defoliation was "essential to security of bases." 5e On 18 August the President permitted the resumption of chemical defoliation until 1 December 1971. He authorized the use of Blue and White but not Orange. Approved herbicide operations were restricted to the perimeters of firebases and installations, with delivery limited to solely helicopter or ground-based spraying equipment, under the same regulations applied in the United States.60 As the expiration date for this authority neared, COMUSMACV asked for an extension. On 26 November 1971 the President authorized continued use of herbicides and set no termination date. At the same time, he stipulated that U.S. defoliation assistance to the Government of Vietnam be confined to "base and installation perimeter operations and limited operations for important lines of communications." This policy prevailed until the last U.S. forces departed RVN in 1973." No defoliant method tried for air base defense purposes in South Vietnam proved to be at once efficient, economical, and politically acceptable. The practical value of herbicides was much impaired by technical, administrative, and political constraints. For chiefly technical reasons, the same could be said for techniques such as burning and scraping. For the United States—as it had for France—vegetation remained a major unresolved problem. V. USAF GROUND DEFENSE FORCES The enormous mass of non-combatant personnel who look after the very few heroic pilots, who alone in ordinary circumstances do all the fighting, is an inherent difficulty in the organization of the air force. Here is the chance for this great mass to add a fighting quality to the necessary services they perform. Every airfield should be a stronghold of fighting air-groundmen, and not the abode of uniformed civilians in the prime of life protected by detachments of soldiers. Sir Winston Churchill, 1941. By late 196S it became certain that U.S. ground combat forces would take part in offensive operations, and that the Air Force would be expected to protect its own installations. The USAF reaction to this unwelcome task was alien to the U.S. armed forces.1 It was to ship the basic means of air base defense to South Vietnam—man by man and item by item. Then in the combat zone the Air Force assembled, organized, and trained these troops. More than 8 months passed before this process began to turn out forces that showed elementary skill in executing their unit mission.2 Security police squadrons were formed in this manner at the 10 major bases in RVN. These units became the focal point of USAF ground defense during the entire war. Tactical versus Nontactical Organization The governing USAF directives* were silent on how to organize and employ security police in a hot war. Hence USAF ground defense forces in RVN were structured to cope with CONUS contingencies in a cold war. A security police squadron in RVN * Air Force Manual (AFM) 207-1, Doctrine, and Requirements for Security of Air Force Weapons Systems, 10 June 1964 (superseded by AFM 207-1, 10 Jun 68, and in turn by AFM 207-1, 10 Apr 70); AFM 205-3, Air Police Security Operations, 15 February 1963 (replaced by AFM 207-2, Handbook for Security Forces, 15 Jul 66, which was supplanted by AFM 207-2, 15 June 69). 79 �South China Sea Kontum , x \ Binh Din CambPb\dia pong Thorn Kratie \ Qu^ng^ Due ^. ._.>-A /Longr-*C_~- South Vietnam International boundary ..__ Province boundary ® National capital O Province capital Da Lat Autonomous municipality Railroad Road Base 500876 5-72 ./ �SOUTH VIETNAM DEFOLIATION MISSIONS ALL MISSIONS Mission Trocb < Woe = Defofint I Greei = Uikiowi fruf* � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 197 Folder The folder containing the original item. 5560 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Fox, Roger P. Description An account of the resource <strong>Corporate Author: </strong>Office of Air Force History, United States Air Force, Washington, D. C. Source A related resource from which the described resource is derived Air Base Defense in the Republic of Vietnam, 1961-1973 Date A point or period of time associated with an event in the lifecycle of the resource 1979 Title A name given to the resource IV. The Target Air Bases ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/c1367427a79c5b7abd832c1ddd5ef9b1.pdf 4b5dc567a7d86e5a5412fcf7652c1cd9 PDF Text Text Item D Number 05224 D NotSMnnBfl Author Fullerton, Raymond W. Corporate Author United States Department of Agriculture (USDA), Office Report/Article Title Memorandum with attachment: From USDA, Office of the General Counsel with enclosed final report on the 2,4,5 -T scientific workshop, March 8-9, 1974, Washington, D.C. Journal/Book Title Year 1974 Month/Day Color Number of Images D 47 Dascripton Notes Friday, March 01, 2002 Page 5224 of 5263 �DEPARTMENT OF AGRICULTURE OFFICE OF THE GENERAL COUNSEL W A S H I N G T O N , D.C. 20250 TO: Administrative Law Judge Frederick Denniston Workshop Participants Parties to 2, 4, 5-Trichlorophenoxyacetic Acid Hearing FROM: USDA, Office of the General Counsel "\c(W ' Enclosed you will find the final report onjAe 2, 4, 5-T scientific workshog &@ld in Washington, D.. C. on March ,8 and 9, 1974. The report is broken down into the following headings. I. Toxicology 1. Teratology 2. Human Toxicology 3. Other Toxicology 4. Carcinogenicity/Mutagenicity II. Chemistry 1. Environmental Impact 2. Analytical Methods 3. Residues 4. Sources of Dioxin 5. Statistics I.II. Rule of Reason Raymond W. Fullerton Margaret Bresnahan Carlson Alfred R. Nolting �1. Toxicology 1. Teratology The workshop was opened by Dr. Gehring stating the objectives of the conference. These objectives were to answer a series of questions which had been posed to Che participants prior to the meeting, to test the validity of the data, and to discuss and hopefully elucidate the meaningfulness of the data for assessing the risk of using 2,4,5-T for currently registered purposes. A list of individuals attending the conference is attached. In order to initiate the conference discussion, Dr. Gehring presented the essence of his testimony on the pharmacokinetics of 2,4,5-T and TCDD. Immediately followingj Dr. Schwetz presented the essence of his testimony on the teratology of 2,4,5-T and TCDD. Pertinent points alluded to in a short dicussion following these presentations were more thoroughly discussed when the participants addressed themselves to the questions posed prior to the workshop. The questions and the subsequent significant discussion were as follows: 1. Has an adequate no-effect level for teratogenicity been determined in experiments for 2.4.5-T and TCDD? The consensus of the group was that a no-effect level cannot be established statistically. No-effect is an absolute term and it cannot be rigorously demonstrated experimentally. Dr. Gaylor pointed out that an experimental no-effect level may be established but that the large confidence limits for even an experimental no-effect preclude utilization of the terminology in the absolute sense it implies. Thus it is necessary to use the judgment of the experimentalists and other qualified people to assess the hazard of any material. There was some discussion about the pertinence of the pharmacokinetic studies in the projection of a dose-response curve. As indicated in his proposed testimony, Gehring asserted that it is scientifically unsound to estimate the incidence of an untoward effect of a trace dose of an agent from studies in which does superseding excretory and/or degradation thresholds have been administered, Drs. Gaylor and Holson from the NCTR took some issue with the assertion, indicating that the mechanism for such effects, whatever they may be, may be the same at smaller doses. Dr. Gehring agreed that the mechanism, that is the molecular interaction �between 2,4,5-T and various receptors, may not change qualitatively with ihe dose. However, the a priori assumption of dose-response methodology assumes that the kinetics for the clearance of a chemical from tissues of the body do not change. Otherwise, one is in essence comparing two different populations. It would be invalid to compare the dose-response for animals excreting or clearing the material from the body with a half-life of 48 hours with the dose-response curve for animals excreting the material from the body in 24 hours. Drs. Young and Holson indicated that they are initiating pharmacokinetic studies.in mice and relating the results to teratogenicity. 2. What are the quantative and qualitative teratogenic characteristics of 2,4.5-T and TCDD? This subject was covered very adequately by Dr. Schwetz and thus there was little additional information presented during the discussion. Dr. Moore stated that 2,4,5-T and TCDD have very little teratogenic potential in rats. In mice, 2,4,5-T and TCDD produce terata—cleft palates and abnormal kidneys. The question was raised as to whether the mouse may be a false lead in assessing the teratogenic hazard of either 2,4,5-T or TCDD by Dr. Gehring. The mouse is very susceptible to stress of various types including airplane rides and it has been demonstrated that such stress may cause terata (cleft palates). Dr. Holson from the NCTR pointed out that although this was the case perhaps humans are also susceptible to such stress and we 'must, therefore, use the mouse to assess the hazard of 2,4,5-T and TCDD. Dr. Gehring asked Dr. Holson if they had measured water consumption and urinary output in the teratogenic studies of 2,4,5-T in mice. He indicated that this was not done, however, it was being considered. Urinary output appears important because 2,4,5-T does cause diuresis. 3. To What extent are results of extreme dosage tests relative to the evaluation of teratological potential at anticipated exposures? This question was alluded to in Question 1 above. The consensus , was that when regimens supersede thresholds for excretion and/or degradation, the data have very limited value for assessing what effects may be incurred with regimens which do not supersede the thresholds. �Ur. Moore indicated that it was important to ascertain whether retarded kidney development may continue with continued postnatal exposure of mice to 2,4,5-T or TCDD.. This is important because kidney development is not complete at birth. Dr. Gaylor raised the point of whether all defects in teratology studies should be combined and evaluated in toto or should specific defects bo evaluated. Dr. Schwetz stated both should be done. Dr. Holson agreed saying rodents are polytocious species and embryos in the same uterus may be in different stages of development. Therefore, the same agent may produce multiple effects in the same litter; the specific effect seen in each individual will depend on its stage of development when exposure to the agent occurs. Ur. Golberg stated that metabolic data are essential for assessing the teratogenic potential of different species. In man, imipramlne is rapidly demethylated to desmethylimipramine. Rabbits are unable to demethylate imipramine as readily and in rabbits the compound is a tcratogen. Teratogenicity in man given recommended regimens would not be expected. Dr. Poland indicated that thus far experiments have demonstrated that TCDD is not degraded to a polar compound and is not very reactive. Therefore, one is hard pressed to conclude that TCDD reacts irreversibly with genetic material to induce teratogenesis. 4. What is the statistical reliability of teratology tests? Since projections of dose-response curves to guesstimate what may occur at lower doses is stochastic, such procedures are useful only to guesstimate the extreme of the potential risk. Dr. Holson asked if effects discerned at doses below those superseding thresholds be used to predict responses to lower doses. Dr. Gehring agreed that this is valid. However, it must be pointed out such projections are stochastic. The tailing of a normal distribution curve for dose-response to either lower or higher doses is ficticious but useful representation of data. 5. What is the teratogenic impact of other dioxins? This question was, for the most part, skipped over because very little information is available. That which is available was presented in Dr. Schwetz's summary. �6. "What are the factors to be considered in extrapolating i'rum the teratogenicity animal testing to humans? The consensus was that the factors are many and many are unknown. Basically, it boils down to a matter of judgment. Dr. Golberg indicated that dilantin may be a human teratogen. He suggested epidemiological studies should be conducted to ascertain whether 2,4,5-T is a teratogen in man. Ur. Holson asked if an epidemiological study had been conducted. Gehring said no and suggested that such a study may be impossible because lie doubts whether many women have been exposed in a manner which would allow characterization of the degree of exposure even if it had occurred. Dr. Schwetz added that for the most part such studies are only feasible for prescribed drugs. 7. What is the significance of the thalidomide instance to tlic current .teratologies.! prognosis? This was only briefly discussed. Dr. Gehring indicated in a reference by B. B. Brody it was stated "that there was a-very good correlation between the blood levels of thalidomide and its teratogenic effect in various species". Dr. Holson from the NCTR, pointed out, however, that the determination of thalidomide is a very difficult one and thus any such correlation may be meaningless. Thalidomide is too labile to allow gathering of data that could meaningfully be interpreted scientifically. 8. What is the significance of the chick embryo tests? This question was discussed for a very short time, because the group quickly reached the concensus that chick embryo is a very poor test system for teratogens as well as toxic effects of chemicals. Dr. Golberg indicated that in work supported by the FDA the chick embryo test system was clearly shown to be inadequate. The chick embryo is in a captive environment with no possibility of eliminating the chemical from its environment. In addition, the chick's metabolic capabilities to degrade and detoxify chemicals are minimal at best. 9. By definition, what is a teratogen? �Although tliis question wasn't in the set of questions supplied to I ho. participants, it was alluded to in Dr. Schwetz's presentation. In Uf.neral, there was a consensus agreement with the definitions proposed l>y Dr. Schwetz. . However, again judgment must be used to differentiate between the fine lines in these definitions. For example, delay in ossification may not constitute a teratogenic response if ossification iollowing birtli is sufficient to quickly catch up. If ossification was so lacking that it would result in physical deformities or abnormal mobilization, this would, of course, have to be termed teratogenic. Dr. Uolson pointed out that it is important to consider not only physical deformities, terata, but also functional deformities. For example, the effects of a chemical on the central nervous system function. It was concurred that such assessment is indeed in order. Also pointed out by Dr. Sciiwetz was that teratologists are now beginning to involve themselves in such evaluations. An additional point which was not alluded to above is that an experimental no-effect level for TCDD has not been established in the mouse. Evidence of embryo and fetotoxicity has been shown at 1 ug/kg when given from the day 6 through 15 of pregnancy. Dr. Moore indicated that essentially equivalent results were obtained in a study in which 0.1 g/kg/day was given to mice. Another point which was not presented above was a short discussion of the approach of Jusko for evaluation of teratogenic effects. The consensus was that Jusko's approach is appropriate only for irreversible teratogens. That is to say for materials which react irreversibly with biological material such as protein and DNA. finally, Dr. Dougherty discussed briefly his data collected from teratology studies of 2,4,5-T in monkeys. His studies confirm the previously observed negative results reported by Wilson. Dr. Holson pointed out that in Wilson's studies a higher incidence of abortions occurred in monkeys given the higher dose levels. Dr. Golberg responded that in Wilson's studies the high incidence of spontaneous abortion in monkeys precludes interpreting this as being related to treatment. The normal incidence of spontaneous abortion in monkeys is 15-20%. Dr. Dougherty added that in his studies the incidence of spontaneous abortion in monkeys was 20%. In monkeys receiving the highest dose of 2,4,5-T (10 rag/kg) the incidence was lower. �2, Human Toxicology , 6 Dr. Kramer reported on the medical surveillance of the Dow 2,4,5-T worker population with exposures dating back to 1940. There was no statistically significant increase in morbidity of disease processes monitored or mortality when compared to standard male population in the United States. The Chairman, V. K. Rowe called for information regarding medical surveillance of any Vietnamese population and none was presented. Dr. Morgan discussed some of the symptoms relayed by applicators such as headache, dizziness or not feeling well; and Dr. Kramer related that this was not a pattern heard from 2,4,5-T workmen. The report of finding prophyria in 2,4,5-T exposed workmen by Dr. Jacob Bleiberg was discussed at some length. This finding has not been duplicated by other investigators in the field. It was the consensus ol. the group that investigation of this parameter would not be productive of." success in developing a monitoring technique for 2,4,5-T exposure. Dr. Kramer reported on the study of 61 workers exposed to dioxins iu a chlorinated phenol process. Forty-nine workers developed some degree oi" chloracne and medical surveillance of this group is continuing. The present epidemiological survey revealed no increase in mortality or change iu the morbidity rate except for the skin disease itself. Dr. Kramer will have a detailed report of this study at a later date. He emphasized that no case of chloracne has been seen in any of the 2,4,5-T workmen. The question of immunological significance of 2,4,5-T exposure was raised and no one had any data to answer this question. It was suggested that following the human exposed population for infectious disease incidence or absenteeism rate could provide meaningful data in this area. Dr. Kilian reported on the cytogenetic studies done on 2,4,5-T workers. A group of 49 workmen were evaluated approximately two years ago and recently a follow-up reevaluation of 40 employees was done. Neither group revealed cytogenetic evidence of an effect from 2,4,5-T exposure. He pointed out that groups of humans had been identified who had had exposures to uranium dust, radium and benzene and studies had shown a correlation between and an increased incidence of cancer. The negative cytogenetic data and the normal epidemiological findings are mutually supportive of the conclusion that 2,4,5-T exposure to this group of workmen had no effect on their health. If 2,4,5-T had rautagenic significance, then one should see a change in the disease patterns of this group, and also see some significant chromosomal abnormalities in their serial chromosomal analysis. �Dr. Golberg raised the question of what realistic human exposures exist with the use of 2,4,5-T in our society. Data has been developed but was not available at this meeting showing that a several thousandfold safety factor exists if one were to directly extrapolate animal data to man. Considerable interest was expressed by the group in population monitoring to determine the distribution and concentration of 2,4,5-T and dioxin in humans. Dr. Kilian pointed out that it had been a relatively simple matter to enlist the aid of lactating mothers to cooperate with this type of study. Dr. Jack Moore related that he was familiar with animal work which indicated that TCDD was readily excreted in milk. It was the consensus of the group that fat biopsies of a large population group would not be practical since a considerable amount of tissue would be required for a part per trillion assay. However, a smaller human study group involving surgical biopsies or autopsy material would be possible. The workshop recommended that: 1. Dr. Bleiberg be contacted to see if he has any additional information on porphyria since writing this paper; «> 2. Be looked at closely in order absenteeism and infectious disease patterns to evaluate evidence of possible effect on immune systems; 3. A larger study on distribution and concentration of 2,4,5-T and TCDD utilizing human milk as the sample tissue be considered; 4. A study utilizing adipose tissue from postmortem and surgical specimens be undertaken to determine if they contain 2,4,5-T and TCDD. 3. Other Toxicology This workshop discussed acute and repeated dose toxicity, and absorption, excretion and tissue distribution of 2,4,5-T and TCDD. This was drawn from information presented in the Dow pre-hearing memorandum No. 2 and in two drafts of testimony (P. Gehring and J. Morris) in which studies conducted by The Dow Chemical Company as well as literature reports were discussed, including, for example, those of relevance from the 1971 American Chemical Society "Chlorodioxin" Symposium and work reported at the NIEHS Meeting in April, 1973, �8 Specifically referenced information was as follows: Dow ['re-hearing Memorandum No. 2, corrected copy February 8f 1974 for 2,4,5-T; Single dose toxicology: pages 9-10, pages 15-16. Repeated dose toxicology: pages 16-19, pages 108-111. Metabolism: pages 20-22, pages 32-37. Metabolism from P. J. Gehring draft testimony: pages 3-20. For 2,3,7?8-tetrachlorodibenzoparadioxin: * Single dose toxicology, draft testimony J. M. Norris: pages 3-11. Repeated dose toxicology: page 22 Metabolism, P. J. Gehring draft testimony: beginning page 20 and from Dow pre-hearing No. 2, pages 111-113. The acute and repeated dose toxicity information of 2,4,5-T was substantially that which is widely available. Much of the information on TCDD, however, is of recent date. In fact, Dr. George Fries, USDA; Dr. -Joliu Moore, NIEHS; and Dr. Alan Poland, University of Rochester presented data from current, ongoing investigations. Dr. Fries reported on a rat feeding study Involving TCDD at dietary concentrations of 7 or 20 ppb (parts per billion) given over a period of over 42 days. He will present tills paper at the National Meeting of the American Chemical Society beginning March 31, 1974 in Los Angeles, California. OC particular importance to the workshop was the presentation by Dr. R. J. Kociba (Dow) of the results currently available from a 90 day study in which rats were given repeated oral dose daily by gavage of 1, 0.1, 0.01, or 0.001 micrograms TCDD per kilogram per day. Light and electron microscopic examination of the tissues is in the final stages. The most important findings were from the pathological examination in which there appears to be definite liver changes and minimal changes in the thymus seen in those animals maintained on the 0.1 micrograms/kilogram/ day dose. Very minimal to minimal cloudy swelling in liver tissues (male rats only) was seen at the two lower levels, 0.01 and 0.001 microgratns/kilogram/day by light microscopy. Preliminary examination by electron microscropy indicate normal appearance of the liver cells, but with dispersion and a possible increase of the smooth endoplasmic reticulum �seen in both male and female rats. These hepatic alterations are similar to those reported with many other compounds and indicate a physiological adaptation on behalf of the liver to metabolizing foreign compounds. The results of the metabolism studies for 2,4,5-T and TCDD (P. Gehring draft testimony) were reviewed by James Rose. It was emphasized that a "steady state" was indicated as having been achieved in the C-TCDD work in rats. Therefore, it is suggested that a steady state would have also been achieved during the 90-day study period reported by Dr. Kociba. Steady state in this instance is believed to mean that the body burden had been established at a maximum level and that the additional input of TCDD into the animal was matched by the rate of excretion. There was a considerable amount of discussion about all the aspects of the single dose, repeated dose, and metabolism of both 2,4,5-T and TCDD. Insofar as possible, this discussion was directed toward evaluating the adequacy of the "other toxicology" irrespective of the other workshops on metabolism, teratogeniclty, carcinogenicity, or mutagenicity. The general concensus of the scientists in this workshop was that adequate data on 2,4,5-T was available on which conclusions for the safety evaluation of levels of exposure to residues which might be ingested due to their occurrence under practical conditions of use of the herbicide could be based. In the government regulatory sense, it was pointed out that negligible residues (less than 0.1 ppm) were indicated for any food crop use. Actually, results of "market basket" studies reported from the U.S. Department of Agriculture would indicate nil residues of 2,4,5-T occurring in the human food supply. Even so, the 90-day dietary feeding studies done in rats and dogs show a "no illeffect" level of 10 mg/kg/day. Should the total diet of humans contain as much as 0.1 ppm of 2,4,5-T (a highly unlikely assumption), a safety factor of 5,000-fold exists for human consumption over that which caused no ill-effect in the total diet of rats and dogs. One or two of the participants indicate that the results of longterm feeding and multi-generation studies of 2,4,5-T in rats would be desirable. The workshop did not have sufficient qualitative and quantitiative data on the amount of TCDD that are occurring in the human food supply. This must be further defined by the analytical and residue chemists. Finalization of these analytical studies and those of the repeated dose toxicity studies on TCDD (90-daysj are necessary before it will be possible to judge adequate margins for TCDD. These may well prove to be �10 sufficient. However, due to the very intricate toxicological and biological manifestations of this extremely toxic material, the workshop recommended that serious consideration be given to conducting longer term studies, i.e. 2 year dietary feeding studies and multi-generation studies in rats. It was reported that a 2 year study on TCDD may be in progress at the Illinois Institute of Technology. However, information relative to thia study was not forthcoming in this workshop. It was recognized that much of the preliminary toxicological and pharmacological data essential for the proper planning of such studies has become only recently available. However, it is now believed that these essential data are in the hands of the toxicologists who should now be in a position to plan the protocols and proceed to organize the accomplishment of such long-term studies. 4. Carcinogencity/Mutagenicity Dr. Legator gave a brief discussion on the relevancy of current mutagenic test systems. He pointed out that the relevancy of these tests were similar to other animal tests. Dr. Legator classified the current tests based on relevancy to man and ease of performing the test. Ease of Performing* Test Relevancy* Iri vitro bacterial test 10 1 I lost-media ted 3 3 Specific locus 2 10 In vivo cytogenetics 2 3 Dominant lethal 2 4 Human cytogenetics 1 3-4 Body fluid analysis (blood, urine)/ bacterial system * 1 = relevant or easy 10 = not relevant or difficult Preferred Test �11 Dr. Legator emphasized the necessity of using test systems employing metabolic activation and mentioned that the body fluid analysis technique could be used in the human. Dr. Kalian agreed with Dr. Legator and referred to the old proposed FDA protocols on mutagenesis. Dr. Kilian briefly discussed the collaborative work with various laboratories to evaluate some of the current mutagenic test systems. Also some of these tests are being used to evaluate GRA.S list compounds. Each test has its specific advantages and disadvantages and the investigator must select the most appropriate test for the specific purpose. Dr. John Moore reported that NIEHS had conducted a dominant lethal test in rats with TCDD and it was negative. His group of workers does not consider TCDD to be a mutagen. Dr. Robiuson mentioned the dominant lethal test conducted with TCDD by Dr. Khera which was also negative. Also the host mediated and dominant .LeLital tests conducted with 2,4,5-T by Buselmaier which were also negative. Dr. Alan Poland also reported sending TCDD samples to Dr. Bruce Ames Cor testing, using his tester strains of S. typhimurium; these test results were also negative. Dr. Kilian reported that human cytogenetic and epidemiological Htudics had not revealed adverse effects in humans working in the production of 2,4,5-T. Dr. Kramer further defined the human cytogenetics studies as one study being conducted while the individuals were actively involved in the manufacture of 2,4,5-T and the second study was a followup on the original group two years later when they were not involved in Llie production of 2,4,5-T. A brief discussion followed on the human exposure dose of TCDD. Dr. Geliring briefly summarized the comparative pharmacokiuetics data on 2,4,5-T in man, rat and dog. Dr. Kilian pointed out that most carcinogens require metabolic activation and if TCDD is not metabolized there would be less potential for carcinogenesis. Dr. Gehring stated there would be no reason to suspect TCDD as a-mutagen based on the rat data as the material is removable and there is no permanent association. Dr. Legator mentioned ±n vivo cytogenetics studies in man: dominant lethal, host mediated, and body fluid analysis if population is available. Dr. Robinson reiterated the tests and results that have been reported in the literature. Drs. Robinson and Emerson stated that there was a correlation between mutagenicity and carcinogenicity. Many scientists feel that carcinogenicity is the result of several mutational events within a �12 cell. Dr. Kramer mentioned "Down's Syndrome" and the abnormal karyotypes associated with it; also the Philadelphia chromosome. Dr. Legator pointed out that about three-fourths of the carcinogens require metabolic activation. Dr. Kociba briefly discussed and showed slides of the multinucleated and enlarged hepatocytes of the rats which were treated with 1 g TCDD/kg for 13 weeks. These changes were similar to those of Buu Hoi and others. Dr. Kramer asked iffeta protein determinations were made and Dr. Kociba said no. Dr. Gehring briefly discussed what Dr. Golberg had said in reference to the hepatocytes - "that multinucleated cells are observed normally in aging rats". Dr. Emerson stated that the lesion was different from those induced by AAF and that multinucleated cells could be found in aging rats and in vitamin deficiencies of primates, Ur. Kramer suggested that the lesion may be reversible. Dr. Moore read several sentences from Gupta's paper - "Besides these degenerative lesions, large multinucleated giant hepatocytes were also seen in liver of TCDD treated rats. The presence of these cells, increased numbers of mitotic figures and pleomorphism of cord cells suggest that a long term study should be done to assess the possibility of the development of hyperplastic nodules and/or neoplasm". yi^ Dr. Gehring mentioned that in the Bionetics study the mice were ft treated with an estimated ,7^g/kg/week for 7 - 2 8 days of age and then .2*/g/kg of TCDD/week as a contaminant of 2,4,5-T for 17 months. Dr. Moore referred to an article in press (Tox. App. Pharm.) that reported the U>5g of TCDD in C57BL mice as 114 ug/kg. The authors also reported similar hepatic lesions, as described by Gupta, et ad, in a subacute study, and stated there was a need for long term studies to evaluate these changes. Dr. Gehring pointed out that the.Bionetics study has done that. Dr. Robinson mentioned the work of IIT on TCDD oral rat and mice .studies. Dr. Emerson elaborated on the IIT studies by saying that the objectives of this program were to determine the chronic toxicity and carcinogenicity of chlorinated dibenzodioxins (including TCDD and hexachlorodioxin) and related compounds by skin application to mice and by oral administration to mice and rats. Dr. Emerson mentioned the 3 mouse carcinogenic studies in Europe on 2,4,5-T that were reported by the International Agency for Research on Cancer, 1973. Dr. Kramer asked if we needed inhalation studies on TCDD. Dr. Gehring said there was no evidence that TCDD was metabolized and that it was not volatile. �13 Ur. Legator stated that data now available is negative on the question of mutagenicity and carcinogenicity of 2,4,5-T and TCDD but additional tests can be added. There is not enough information available at this time. In summary, it was generally agreed that data presently available do not suggest that 2,4,5-T is a mutagen or a carcinogen. Additional studies might possibly lend more confidence. The Bionetics study in mice was long-term. The TCDD contained in the 2,4,5-T amounted to approximately 0.7 micrograms/week for 1 month and 0.2 tnicrograms/week for 17 months without an increase in incidence of tumors. Long-term studies on TCDD in rats now in progress at IIT Laboratories should help clarify the hepatic lesions seen at high dose levels in subacute studies in mice and rats. II. Chemistry 1. Environmental Impact Several participants presented data from laboratory and field studies with TCDD, alone or in conjunction with 2,4,5-T and 2,4-D. The information presented herein was developed from notes taken during the workshop supplemented by published and unpublished reports of the individual studies as listed under references. Proposed answers to the assigned questions are outlined briefly at the end of this report on Workshop B. Discussions during the workshop included attempts to define terms used to describe the relationship between concentrations of TCDD reported to be associated with different components of the ecosystems studied. Although agreement was not reached among all participants, the following definitions are hereby proposed for further consideration: (a) liioconcentration - the concentration of a chemical in or on an organism compared to its environment, due at least in part to physical adsorption on the organism. (b) bioaccumulation - the accumulation of a chemical in an organism from its environment. (c) biomagnification - the increase in concentration of a chemical in successive organisms in ascending the trophic food chain. The terms used in the following reports are the terms used by those making the presentations and-do not necessarily conform to the above proposed definitions. �14 Studies by Isensee and Matsumura were done quite differently but the data obtained were similar and generally supportive of each other. The general conclusion from Isensee's work with C-TCDD was that the distribution ratio for the radioactivity in water/soil was about 1/10,000 and the ratio for organisms/water was,about 10,000/1. The bioaccuraulation ratio calculated for TCDD (based on C count) was about 10 times less than for DDT in Isensee's experiments and about 10 to 100 times less than for DDT in Matsumura's studies. According to Matsumura, no evidence was obtained to indicate biomagnification of TCDD in the food chain. Isensee's data were reported in part on page 34 of the EPA January 18 prehearing brief. , His studies were conducted in a glass aquarium containing 4 liters of water and various amounts of Matapeake silt loam or Lakeland sandy loam in three distinct experiments. The soil was pretreated with ^C-TCDD at nine levels ranging from 7.45 parts per million to 0.0001 ppm. The amount of TCDD per tank was 149 g in 20 g of soil in the first experiment, 63 g in various amounts of soil in the second experiment, and ranged from 10 to 0.01 g in 100 g of soil in the third experiment. The organisms were introduced into the tank in sequence as follows: Algae, duckweed, snails and daphnids for 28-29 days, then Gambusia (mosquito fish) for 3 days, then catfish for 6 days. . The following table taken from Isensee1s manuscript represents the distribution of apparent TCDD in the various componetns, all based on •'•^C-countlng. Almost all the recovered radioactivity was associated with the soil, regardless of level added, indicating that solid would be the main reservoir for TCDD in the environment. The amount recovered in the water ranged from 0.05 to 3.61% of the amount added, with no apparent relation1 to the amount added. The TCDD levels reported were all based on 14C-counting. The nature of the radioactivity was examined by thin-layer-chromatography (tic). About 86 to 94% of the recovered activity was found in a single mobile spot for each extract, with up to 6% at the origin and up to 10% as a streak between the origin and the mobile spot. The major spot for tissue and water extracts had a somewhat lower mobility compared to the standard TCDD (Rf 0.71), attributed by Isensee to the presence of soluble organic material. (However, it is possible that the 1% radioactivity found in the organisms and water compared to the soil represented soluble impurities or photodegradation products of TCDD rather than TCDD itself.) The levels reported for soil and water are shown on the following page of text, giving an average distributon ratio of 1/11,350 for water/soil. �15 Table III. Recovery of 14 C in Ecosystem Components. 14 Expt. no. Soil cone, ppm Soil H Algae I 7.45 84.90 3.61 1.90 na II 3.17 97.79 1.51 0.67 II 0.53 9.9 50 0.30 II 0.29 88.45 II 0.15 III 2° Percent of C-TCDD originally added Daphnids Duckweed Snails a Gambusia Catfish Total 0.44 0.16 0.06 na 91.07 na 0.23 00 .2 0.04 0.20 104 0.6 0.12 na 0.04 ndb 0.02 0.04 95.61 0.11 0.06 na 0.02 nd 0.01 0.04 88.69 87.57 0.05 0.04 na. 0.02 nd 0.01 0.02 87.70 0.10 85.44 0.31 0.26 0.03 0.21 0.01 0.11 0.47 86.83 III 0.01 86.73 0.32 0.28 0.04 0.15 0.01 0.07 0.53 88.13 III 001 .0 87.59 1.32 0.55 0.04 0.18 0.01 0.06 0.47 90.22 III 0.0001 98.56 0.79 0.28 0.26 0.68 0.02 0.15 0.43 101.17 a not analyzed. b not detectable. �16 Experiment ppm in soil ppt in water water/soil I 7.45 1330 il 3.17 239 1/13,260 II 0.53 48 1/11,000 II 0.29 18 1/16,000 II 0.15 7 1/21,400 III 0.10 7.13 1/14,000 III 0,01 0.66 1/15,000 III 0.001 0.26 1/3,850 III 0.0001 0.05 1/2,000 1/5,600 average 1/11,350 In experiment I at 7.45 ppm TCDD in soil, the apparent 1330 parts Pcr trillion (ppt) TCDD in the water exceeded the solubility of TCDD in pure water (0.2 ppb or 200 ppt). This discrepancy may be due to increased solubility of TCDD in water containing dissolved organic matter from components in the ecosystem, or to adsorption of TCDD on colloidal particles in the sample of water which was counted, or because part or all of the dissolved ^C-activity was not TCDD. A concentration of 3.17 ppm in soil gave 239 ppt TCDD equivalent in water (close to the solubility of TCDD in water). Experiment III was conducted using higher specific activity -^C-TCDD than in Experiment I and II, and the lowest levels studied approached levels which might be encountered in soil treated with 2,4,5-T containing measurable levels of TCDD. Apparent TCDD levels in the organisms reached as high as 2 ppm in daphnids in Experiment I at 7.45 ppm in soil vs. 1330 ppt in water. The organisms survived these very high concentrations, possibly because the I4C-activity was not TCDD or was TCDD adsorbed on the surface rather than absorbed into the organisms. This view is supported by the fact that the organism/water ratio of C-activity was lower for catfish than Cor the smaller Cambusia (mosquito fish). This is the opposite to DDT in fish in natural systems where larger fish have higher residues; however, the exposure may not have been long enough in this study to �17 draw firm conclusions. At lower concentrations the relative amount in various species changed, indicating that there is a difference between bioconcentration and bioaccumulation. Raising the concentration in water two-fold resulted in a decrease in the apparent bioaccumulation ratio by half. (In all cases, the bioaccumulation ratios were calculated from the C-activity in tissue on a dry weight basis compared to the C-activity in water, emphasizing differences for tiny aquatic organisms which consist of up to 90% water.) (See Table II from Isensee, which follows.) �Table II. Bioaccunulation of Expt. no. Soil cone. f\T\ff\ ppm I II II II II III III III III 7.45 3.17 0.53 0.29 0.15 0.10 0.01 0.001 0 . 0001 H20 Cone. 14 C-TCDD by Several Aquatic Organisms as Affected by Soil and Water Concentration Algae Duckweed Snails nn t" ppc 1330 239 48 18 7 7.13 0.66 0.26 0.05 * Daphnids Gambusia Catfish ppD 6,690 + 960b 2,500 + 120 390 + 20 230 + 20 130 + 50 79.3 + 12.5 5.0 + 1.0 1.4 + 0.2 0.1 + 0.0 nac "na na na na 30.7 + 1.3 3.3 + 0.5 0.3 + 0.0 0.2 + 0.1 1,820 + 170 2,780 + 400 1,970 + 690 290 + 30 330 + 80 125 + 23 9.7 + 1.4 1.4 + 0.2 1.2 -t- 0.6 10,400 + 480 7,450 + 30 70 + 70 + 70'+ 163 + 10 17.7 + 5.9 4.7 + 2.2 2.4 + 1.1 na 1,380 + 220 2,200 + 680 720 + 130 . 540 + 250 110 + 90 120 + 5 420 + 190 90 + 20 80 + 50 439 + 76 103 + 49 41.8 + 4.5 18.4 + 5.3 5.9 + 2.7 1.2 + 0.3 1.2 + 0.6 0.1 + 0.0 Bioaccujnulation Ratio I II II II II III III III III 1330 239 48 18 7 7.13 0.66 0.26 0.05 5,000 10,500 8,100 12,800 18,600 11,100 7,600 5,400 2,000 a na na na na na 4,300 5,000 1,200 4,000 1,400 11,600 41,000 16,100 47,100 17,500 14,700 5,400 24,000 7,800 31,200 na na na 22,900 26,800 18,100 48,000 1,000 9,200 11,300 23,300 . 12,900 61,600 63,300 22,700 24,000 TCDD of 2.8 uCi/mg specific activity used in experiments I and II; 460 uCi/mg specific activity used in experiment III. Standard error of the mean for 3 replications (experiment I) and 2 replications (experiments II and III). cna - Not analyzed. Concentration of TCDD in tissue (dry wt.) divided by concentration of TCDD in water. * Solubility of TCDD in water to 200 ppt na 3,000 2,300 6,700 11,400 14,400 27,900 4,600 2,000 �19 The January 1974 prehearing brief submitted by EPA contained data derived from Isensee's study. Values cited were 0.08 to 0.44 ppm TCDD in various aquatic organisms exposed for 28-29 days or 3 days to water in contact with soil containing 0.1 ppm (100,000 ppt) TCDD. They calculated that treatment of rice with 2,4,5-T containing 0.1 ppm TCDD would result in 12 ppt TCDD in the top 1/4 inch of soil and 0.01 ppt in the water in contact with it. They extrapolated this to result in 140 ppt in fish within 3 days exposure to rice flood water. (This was based on the 1/14,000 concentration factor for water/soil at the 0.1 ppm level in soil rather than for the lower 1/2000 factor found at the more reasonable level of 100 ppt in soil. Matsumura measured the uptake of radioactivity by a variety of organisms in a 200 ml mini-ecosystem to which he added the same CTCDD used by Isensee in Experiments I and II above. In one series of experiments the TCDD was added directly to water as a solvent solution along with the primary food organism such as algae and yeast. In a second series the solvent solution was evaporated as a thin film on the inner surface of a glass container in which the food organisms were grown prior to transferring them to the aquarium. .In a third series the •^c-TCDD solution was added to sand, the solvent evaporated, and the sand added to the aquarium. All studies were conducted for only 4 to 7 days under static conditions with single and mixed populations of organisms to compare the bioaccumulation ratios for TCDD, DDT, Y-BHC and mexacarbate (the active ingredient in ZECTRAN(R) insecticide). In the first study, concentration factors for TCDD in organisms compared to water were 49 for daphnia in the presence of algae, 218 for ostracods in the presence of algae, and 121 for brine shrimp in the presence of yeast. However, the theoretical water concentrations of 32.4 and 16.2 ppb TCDD equivalent far exceeded the solubility of 0.2 ppb for TCDD in water so absorption of the TCDD on the food organisms must liave occurred. In the second experiment with algae containg 162 ppb TCDD, the concentration factors were 2198 for Daphnia compared to water containing 0.4 ppb TCDD equivalent, and 107 for Ostracod in water containing 2.6 ppb TCDD equivalent. In the third series of experiments using 1.62 ppm (ug/g) C-TCDD on sand, he found 157 ppb TCDD equivalent in brine shrimp vs. 0.1 ppb in water, and 4,150 ppb in mosquito larvae vs. 0.45 ppb in water. Under the same conditions only 2 ppb was found in fish (silverside) and none was detected in water. In a two-step study with mosquito larvae followed by fish, the level in fish was 708 ppb TCDD equivalent compared to 3700 ppb in the mosquito larvae and 1.3 ppb in the water. This gave a concentration factor of 54 as compared to 306 for DDT (not 540 as cited �20 In the January 1974 prehearing brief submitted by EPA). Based on these experiments, TCDD has a bioaccumulation factor about 1/10 to 1/100 that oi" DDT for the organisms studied or about 1/10 of that found in Isensee's studies. Matsumura stated during the workshop that we have no proof that TCUD is biomagnified, i.e. that its concentration increases as it goes up the food-chain. However, he did find bioconcentration of the ^Cactlvity in or on organisms compared to water under the conditions of the studies. He also found that the bioconcentration factor was 10 Limes less when he used lake sediment rather than sand in his miniecosystem. He also found 1-2% degradation of the TCDD in the presence of. lake sediment and a variety of organisms. He plans to do more work on microbial degradation using higher specific activity TCDD and lower concentrations in the soil reservoir of his system. Baughman and Meselson of Harvard reported finding 18 to 810 ppt TCDD in crustaceans and fish caught in rivers and near the coast of Vietnam not far from Saigon. The samples were collected in August and September 1970 and were kept frozen under liquid nitrogen until analyzed 2-1/2 years later using Baughman1s repeat scan mass spectrometry technique. Dow has requested samples of the fish and/or shrimp for confirmatory analysis using combined gas chromatography-mass spectrometry (GC/MS), but these requests have not yet been honored. Dow is interested in performing confirmatory analysis because it is possible that the TCDD reported to have been found in these samples may represent inadequate separation from high levels of interfering PCB's.or DDE, or to the presence of tetrachlorodioxins other than the toxic 2i3,7,8-isomer referred to as TCDD. Such "dioxins" could originate from pentachlorophenol used in that region for treatment of aquatic areas. Analyses of Asiatic pentaclilorophenol revealed high levels of "dioxin" compounds including TCDD whereas no TCDD has been detected in Dow pentachlorophenol. Use ot" Herbicide Orange (Agent Orange) for defoliation in Vietnam was at 3 gal/A (approx. 13 Ib 2,4,5-T acid equivalent per acre plus 13 Ib 2,4-D ae/A, both as butyl esters). Captain Young stated that herbicides were applied as a spray released at 150 ft elevation at 130 (Knots Indicated Air Speed) with average particle size 250 microns and 98% of all particles greater than 50 microns in diameter. Thus most of the material was intercepted by foliage of the target forest area. Since TCDD is considerably more soluble in Herbicide Orange than in water, and esters of 2,4-D and 2,4,5-T are readily taken up by the waxy surface of leaves, most of the TCDD in the herbicide remained on the foliage where it was subject to photodegradation without ever reaching �21 tlic water. Young added that some areas may have received four or five applications over the years and a few spots may have been grossly contaminated when defoliant loads were dumped by pilots to escape enemy attack. Leng has calculated that direct application of Herbicide Orange to a pond one foot deep would result in initial levels of 5 ppm 2,4,5-T plus 5 ppm 2,4-D as butyl esters. Such levels would be lethal to fish. If the 2,4,5-T contained 1 ppm TCDD (the specification level for Dow 2,4,5-T in the 1960's) the water would contain 5 ppt TCDD at the time of application. However, the dissolved TCDD could undergo photodegradation in the presence of dissolved organic hydrogen donors and could also be largely absorbed on the pond sediment resulting in much less than the calculated 5 ppt TCDD in the water. The chances seem slim that contaminated sediment from treated aquatic sites could end up in any one location to provide levels of TCDD sufficiently high to cause residues up to 810 ppt in fish or shrimp caught up to 30 kilometers from shore, as Implied in reports on the work by Baughman and Meselson. The general concensus of opinion among participants in the workshop was that it was unlikely that the residues-found in Vietnamese fish and shrimp collected in 1970 were due to TCDD in the 2,4,5-T used for defoliation in that area during the 1960's. Further information should be obtained as to how the analyses were conducted, i.e. whole fish including heads, fins and viscera', and whether most of the alleged residue is associated with 'Scales and skin or with fat of the fish, or with heads and tails of the shrimp as has been rumored recently. The samples should be made available for analysis in other laboratories, using slightly different methods, to confirm the nature and level of the residues claimed to have been found by Baughman. Crummett reported on analyses for TCDD in samples collected by Dow in a rangeland area in Texas and in a rice growing area in Arkansas. No TCDD was found in catfish caught in a pound draining an area of about a million acres of rangeland. According to Bovey (USDA, Texas), the area had been treated with about a million pounds of 2,4,5-T since 1949. The GC/MS methods had a sensitivity of 1 to 2 ppt TCDD and a detection limit of 6 ppt in these fish. Similarly, no TCDD was found in catfish and bass collected in a 200 acre pond adjacent to a 6000 acre rice field where 2,4,5-T had been used for many years and where the water had been recycled over the field each year. The lower limit of detection for TCDD was 8 ppt in these fish due to background interference from high levels of DDE and PCB's. No TCDD was detected in sediment from the pond (detection limit 1 ppt) nor in water from the pond (detection limit 250 parts per quadrillion). �22 Samples of human milk from women in the% rice growing area in Arkansas were also analyzed. No TCDD was detected with a sensitivity of i to 2 ppt based on recovery studies on cow's milk with much interference due to high levels of DDE and PCB's.. Page 36 of the EPA prehearing brief reported finding 6 to 41 ppt TCDU in £at and 1 to 5 ppt in liver of calves, goats and sheep fed immediately after application of 2,4,5-T to rangeland. According to information obtained from EPA, the animals grazed for 38 days prior to slaughter in an area treated with 2,4,5-T at 0.5 Ib/A. The 2,4,5-T contained 0.05 ppm TCDD. Leng calculated that measurable residues of TCDD are not likely to occur in fat and still less in liver of these animals. As shown below, the maximum theoretical residue of TCDD would be 117 ppt in fat if all the grass eaten contained the maximum calculated residue of 4 ppt TCDD for the entire 38 days, and all the ingested TCDD reamined in the fat on the animals. in reality, most grass would contain less than the maximum residue, the TCDD content of the grass would decrease with time after application, much of the TCDD ingested would be excreted during the 38 days, and only part of the retained TCDD would be in the fat. This view is supported by data from independent analyses by Dow and EPA of fat and liver from cattle fed 50 to 900 ppt TCDD with 100 to 1800 ppm 2,4,5-T continuously in the total diet for 28 days. According to the EPA data (table following p. 36 of the January 1974 EPA prehearing brief) the levels of TCDD found in fat were about 2.1 times the level in the diet and were lower in liver. Therefore, ingestion of less than 4 ppt TCDD in the grass (12 ppt on a dry weight basis) would result in less than 25 ppt in the fat of the animals. Dow values for TCDD in fat were considerably less than those found by EPA at levels of 50 or 150 ppt TCDD in the diet ' and were much higher than EPA values at 450 and 900 ppt TCDD in the diet, indicating that EPA had more background interference and poorer recoveries than Dow. EPA also reported finding up to 397 ppt TCDD in shrews trapped in rights-of-way treated with 2,4,5-T. Additional information obtained recently from EPA indicated that residues found in four samples of shrews ranged from 54 ppt to 397 ppt (average 202 ppt) from areas treated with 2,4,5-T at 10, 16 or 8 Ib/A. No information was provided as to how the material was applied, nor the dates of treatment and sampling, nor the nature of samples analyzed. Further inquiries will be made to obtain full details of how the animals were exposed and how the analyses were conducted. Dow will pursue obtaining monitoring samples from EPA for confirmatory analyses by the combined GC/MS procedure. �23 C.-jptain Young reported on studies conducted in a U.S. Air Force site (Tost Area C-52A, Eglin Air Force Base Reservation, Florida). Massive amounts of herbicide were applied undiluted by air during 196270 to an area of approximately one square mile. In 1962-64, Herbicide I'urple (Agent Purple) was used. It contained n-butyl ester of 2,4-D and mixed butyl and isobutyl esters of 2,4,5-1, and is estimated to have contained as much as 40 ppm TCDD. It was applied along the flight path on a 92 acre area at a total rate of 1894 pounds 2,4-D plus 2,4,5-T per acre. Another flight path in the 92 acre area was treated in 1964-66 with Herbicide Orange (Agent Orange) at a total rate of 1168 pounds 2,4D plus 2,4,5-T per acre. Another 240 acre area received lower rates of Herbicide Orange and Herbicide White (picloram plus 2,4-D) in 1966-70. The test site was very sandy (92% sand, 4% silt, 4% clay). SpringTod ponds originated on the test grid and drained across the flight path into the adjacent plant and animal community. In 1970-71 samples of soil were analyzed for TCDD and none was detected by the methods available at that time (sensitivity 1 ppb rather than 1 ppm as given in a USDA summary report). Recent analyses of samples taken in June and October 1973 indicate levels of 10, 11, 30, and 710 ppt TCDD in the top 6 inches of soil from various locations in the site. Residues found at lower depths were probably due to contamination from the upper level during the sampling procedure. The highest level (710 ppt) was in a sample from the area that received 947 Ib 2,4,5-T/A during 1962-64. Young estimated that initial residues may have been as high as 1 ppm TCDD in soil on one of the oldest flight paths treated at these high rates with Herbicide Purple in 1962-64. The 30 ppt level was at the intersection of flight paths receiving-Herbicide Orange in 1964-66 and 1966-68. Analysis of sediment from a bayhead near the test area revealed levels of 13 ppt near the 1962-64 flight path and 11 ppt in a pond adjacent to the intersection of the 1966-68 flight paths. The soil around the ponds also contained low levels of TCDD (10 and 11 ppt) but none was detected in aquatic organisms collected from ponds, bayheads, or streams draining the test area (limit of detection 10 ppt). Livers of beach mice trapped in 1973 were reported to contain 300 to 540 ppt TCDD after an estimated 30 generations of exposure time in this area. Cotton rats trapped near ponds on the 1966-68 test area were reported to contain 210 ppt TCDD in the liver. Analyses of livers from mice and rats trapped about a mile from Test Area C-.52A were reported as 20 ppt. Photographs of the test areas in 1969 clearly showed the effects of the massive herbicide treatment but photgraphs in 1970-71 and in 1973 showed relatively complete recovery of the vegetation cover within a few �24 yuarH. Samples of seed from panicum grass in the treated area are available for TCDD analysis to confirm the belief that this chemical is not taken up by plants from soil residues. Young also reported on mass degradation studies where Herbicide Orange was incorporated below the soil surface at rates of 1000, 2000 or 4000 Ib/A. The intitial TCDD level was about 148 ppb in sites receiving 4000 Ib/A. The half life found for TCDD was only 88 days in the presence of massive amounts of 2,4-D and 2,4,5-T under the alkaline desert conditions of this study in Utah. This is considerably faster than the one-year half life found by Kearney et al. when only TCDD was added to soil at levels of 1, 10 and 100 ppm. It is likely that the TCDD was more evenly dispersed in the soil when added as a ppm solution in .Herbicide Orange (butyl esters of 2,4-D and 2,4,5-T) and was cometabolized with these massive amounts of 2,4-D and 2,4,5-T in the soil. The soil was initially at a pH of about 8 but rapidly became acid when the esters were hydrolyzed to 2,4-D and 2,4,5-T by soil microorganisms. The degradation of TCDD is believed to occur via bacterial action. Norris commented briefly on his work with TCDD in three species of fish (guppies, coho or silver salmon, and trout) and three aquatic invertebrates (a snail, a worm and mosquito larvae). The levels studied ranged from 0.056 to 10,000 ppt TCDD in water for 24 to 96 hours and observations were made for up to 80 days. The TCDD level in water with yo.ung salmon declined significantly with time. A 50 ppt solution decreased to 50% in 24 hr. and to 20% in 96 hr. The initial rapid loss is probably due to adsorption since.a similar test without fish declined to 60% in 4 lir. Some volatilization may also have occurred. The toxic response to TCDD in fish is delayed as it is in other animals. Initial response to the chemical did not occur for 5 to 10 days after the beginning of the exposure period and mortality often extended over the next 2 months. The levels of exposure were expressed in nanograms per gram of total body weight (ng/g) of the organism based on the amount of material in the container relative to fish biomass at Lhe beginning of the experiment; this is not equivalent to total body burden in tlie fish. Norris ejt al. concluded that TCDD in water or food is toxic to fish and duration of exposure is less important than level of exposure. Irreversible effects were produced in young salmon exposed to TCDD in water al levels greater than 23 ng/g of fish and death resulted in 10-80 days. The critical exposure period may be somewhat less than 24 hours in static water toxicity tests in which TCDD concentrations may change �25 markedly witli time. Small fish are more sensitive than large fish on an equivalent exposure level basis indicating adsorption on the surface may be the major route of uptake from water. Levels of 2.3 ppm TCDD in food markedly reduced growth of young rainbow trout in tests where 10 fish were exposed to 6.3 ug per tank per week for 6 weeks. However, no effect was noted in fish when the food contained 2.3 ppb (2300 ppt) TCDD. Pupation of mosquito larvae was not affected at 0.2 ppb TCDD in water (its solubility) but this level reduced the reproductive success of the species of snail and worm studied. As noted previously, all these studies were conducted at TCDD levels lor in excess of what might be encountered in the environment from the use of 2,4,5-T containing 0.1 ppm TCDD. Norris estimated that levels o£ 0.0001 to 0.001 ppt TCDD might occur in streams shortly after aerial application of 2,4,5-T at 2 to 4 Ib/A in Western forests. Based on the above information, the following answers to questions presented to the workshop were suggested: 1. The reported finding of up to 800 ppt TCDD in Vietnamese Qsh and shrimp has little or no significance to current U.S. manufacture and use. Data reported by Young indicate that the residues found in 1973 are not derived 1'rom use of 2,4,5-T for defoliation in Vietnam during the 1960's. 2. The. results of laboratory studies on "bioaccumulation" ol: TCDD indicate that TCDD is preferentially associated with soil in the natural environment, but that the very small quantities in water in contact with the soil may become bioconcentrated in/on aquatic organisms. However, the studies also indicated that the levels in/on the organisms would not exceed the levels in the soil source. Current U.S. manufacture and use is not likely to result in detectable residues of TCDD at the ppt level in water, fish, soil, crops, meat or milk. Care must be taken in interpreting analyses for TCDD in the presence of much larger amounts of DDE and PCB's in the samples. • 3. There is little significant hazard to the non-human environment resulting from current U.S. 2,4,5-T manufacture and use. This conclusion is based on the lack of pathological effect noted in animals exposed to high �26 levels- in the environment at Eglin Air Force Base as well as in the diet in exaggerated feeding studies along with 2,4,5-T in livestock. . Calculations indicate that levels of TCDD which might occur in the environment L:rom use of 2,4,5-T are far below those which might cause an untoward effect in animals, birds, fish, or other living organisms. 2. Analytical Methods The workshop briefly discussed the following analytical methods; (1) analysis of 2,4,5-trichlorophenol, (2) analysis of 2,4,5-T acids and esters, (3) determination of 2,4,5-T acid in plant tissues and products, (4) determination of 2,4,5-T acid in animal tissues, (5) determination oi: 2,4,5-trichlorophenal in animal tissues, (6) TCDD in 2,4,5-trichlorophenol, 2,4,5-T. acid, and 2,4,5-T esters, (7) TCDD in environmental samples and (H) other dioxins in 2,4,5-T acid and esters. Participants in the workshop saw no problems with methods 1 and 2. A question was raised concerning methods 3, 4 and 5 as to whether these methods determine total 2,4,5-T acid and total 2,4, 5-trichlorophenol or if bound residues of these materials remained unextracted by the method. No problem was found with method 6. Considerable problems remained in the interpretation of the meaning of low part per trillion results in method 7, however. It was also generally agreed that method 8 was not completely developed due to lack of analytical standard of certain dioxinu. The: workshop thus agreed that the two questions proposed by those wiio set up the work shop were the correct ones to which it should address itself. These were: (1) what is the ability of current methods used to determine bound residues of 2,4,5-T and 2,4,5-trichlorophenol? Those iu attendance were in agreement that these methods determine total residues including "bound" residues in animal and plant tissues, (2) what are the criteria to be used in arriving at a determination of the valid level of detection for TCDD? This question was considered by a group of Analytical Scientists, December 13, 1973 in a meeting at the Environmental Protection Agency. The results of that meeting were summarized by Carrol Collier in a letter to the participants on January 25, 1974. lie summarized the conclusions in seven points. The first five of these points were in agreement with the notes and recollection of the workshop participants who also participated in the December 13 meeting. However, points six and seven were not and Dow Chemical was instructed to respond to points six and seven in a letter to Collier. �27 Present methods for determination of TCDD at low levels in environmental samples include: (1) gas chromatography/low resolution mass spectroscopy, (2) high resolution mass spectroscopy, and, (3) gas chromatography/high rcHolution mass spectroscopy. These methods are the most specific and sensitive methods known. But in spite of this, the exact meaning of small signals produced on the mass spectrometer is not clear. The reasons J:or this are: (1) control samples are not available, (2) ions having the same mass have been shown to arise from other materials present in the environment and, (3) interferences are easily picked up due to contamination. These reasons make interpretation of results at low parts per trillion levels very uncertain. One way to increase the certainty of an analytical procedure is to liavu an alterntitive equally specific and sensitive technique. The participants in the analytical workshop had no such technique, although radio immunological assay was suggested as a possible solution into the problem. A second suggestion proposed to add more credibility to the analytical results was to have an exchange of samples between participating laboratories. In particular, the group suggested the Dow Chemical and Environmental Protection Agency exchange samples from the environment where the TCDD level was expected to be in the range of 0-20 parts per trillion. A suggestioia was made by Phil Kearney that TCDD levels in environmental samples be reported in groups or levels of data; for example, 0-10 parts per trillion, 10-50 parts per trillion, etc. This, it was thought, would be all that would be necessary to make judgements as to the meaning ol levels in the environment. In general, however, the group felt that it was in no position to resolve this question at the level of 10 parts per trillion TCDD during Llie workshop. It suggested that we encourage the Environmental ProLection Agency to ask the American Chemical Society to select a peer group to review the methods and determine the true level of detection for these methods. Finally the workshop acknowledged that more work would need to be done to determine dioxins other than TCDD in 2,4,5-T acids and esters. Although estimates of these materials can be made, standards are not available and the precise structure of the material being measured is still in doubt. Efforts are being made by all participants to obtain standards and have them examined by Dow Chemical in relation to its products. �'\2yjfe Residue of 2 , 4 , 5 - T - ppm Residue of TCDD - PPT CT7< 3 o o to H- tr H- O O tr< M n to o o o o to o o un o o o o o O ^ o M ^ O ^ H3 O O 0 > NJ • j ••: T) M HO O ft HO i ! 1 X~ - 9/ "• ; 03 / / / / J^- U1 > I I ^_ H3 M M 0} pS I I P> O co a / ro en rt rt fl> H < 0) 21 CO - / i £> Jf ' ' ' r T ;- / / / o en ____ CO , 23 CD ft) X en cn § S to o Ul o Ul o o o o o �28 3. Residues In this session, we were primarily concerned with two questions: • ^ 1. What is the significance of low-level residue findings in fat? 2. What residues of 2,4,5-T and TCDD are likely to occur in human food as a result of registered uses of current 2,4,5-T manufacture? A summary of the uses permitted by the label was given by the chairman. ' Data were then summarized by the chairman on residues of 2,4,5-T which have previously been reported, starting with sugarcane. On growing sugarcane, two applications of 2,4,5-T at the rate of one pound/A gave an average of 10 ppm of 2,4,5-T immediately after die second application. This decreased to 0.05 ppm by harvest time, 24 weeks later. When sugarcane containing a residue was processed, the residue was distributed as follows: Stalks contained more than tops, the concentration in bagasse was greater than that in juice, syrup and molasses contained 5 and 12 times the concentration of the juice they were uuule from, and raw sugar contained less than the juice. Next a summary of data of 2,4,5-T residues on grass was discussed (Getzendaner, M. E., "Fate of Herbicides in Forage Crops", Joint Session Agronomy, Animal Science and Dairy Science, Southern Agricultural Workers, Atlanta, Georgia 2/6/73, slide 3). Specific residues averaged 100 ppm per pound.per acre at time of application, and decreased with a halflife of 1-2 weeks. In the Texas experiment which comprised one of the experiments cited, grass was also analyzed for TCDD. This treatment was made in 1969 before the specification for TCDD in 2,4,5-T was lowered from 1 ppm to 0.1 ppm maximum TCDD. It is estimated that there was approximately 0.5 ppm of TCDD, in the 2,4,5-T used in the formulation. Preliminary data were given showing that one- day after application of 12 pounds of 2,4,5-T per acre about 600 ppt of TCDD was on the grass. This decreased to about 200 ppt of TCDD one week after application, compared to 700 ppm of 2,4,5-T. Sixteen weeks after application there were residues of 10 ppm of 2,4,5-T and about 15 ppt of TCDD. It was emphasized that these are preliminary figures, and that the TCDD in the 2,4,5-T applied was much greater than the present maximum allowed. More samples need to be analyzed to get more precise data for this, but these data show the TCDD as well as 2,4,5-T disappears at a very rapid rate from grass after application. �29 A discussion of a milk residue study followed. (Bjerke, E. L., e£ ol "Residue Study of Phenoxy Herbicides in Milk and Cream", J. Agr. Food Cliem. , 20, 963-967 (1972)). Three cows were given diets which contained 2,4,5-T at successive levels of 10, 30, 100, 300 and 1000 ppm, based on total feed weight, for two week periods. The 1000 ppm level was given Cor 3 weeks, then feed without 2,4,5-T for several weeks. The 2,4,5-T used in this study was found to contain about 0.5 ppm of TCDD, or about five times the maximum level permitted in 2,4,5-T manufactured today. Average residues found in milk were 0.1 ppm of 2,4,5-T and 0.1 ppm trichlorophcnol at the 300 p'pm 2,4,5-T feeding level, and 0.4 ppm of 2,4,5-T and 0.2 ppm trichlorophenol from the 1000 ppm 2,4,5-T feeding level. These decreased in 3 days after withdrawal of 2,4,5-T from the feed to levels below the level of sensitivity of the method, 0.05 ppm. Preliminary results on analysis of milk from the 1000 ppm 2,4,5-T L'eediug level, show about 50 ppt of TCDD. It was emphasized that these are preliminary data. These same animals had received increments of 2,4,5-T containing TCDD in the diet before the start of the 1000 ppm 2,4,5-T feeding adding up to 22% of the amount consumed during the 21day feeding of 1000 ppm which would have made a contribution to this. Also, it must be remembered that the 2,4,5-T used contained about five times the concentration of TCDD as current production. Further, very limited numbers of samples have been analyzed. Seven days after withdrawal of the chemicals from the feed, a level of 40 ppt of TCDD was recorded, while about 15 ppt of TCDD was found in n .sample 60 days after withdrawal. Discussion of these data, method of analysis and possibility of contamination in the laboratory followed. Dr. Kearney pointed out the critical nature of the data. Lynn stated the need to analyze samples at lower feeding levels which would more nearly reflect the levels of TCDD on gratis sprayed in a pasture at rates actually used and with milk animals kept off for 6 weeks, as stated on the label, which would allow dissipation oE the residue. Dr. Bovey.stated that on pastures for dairy animals, 2,4-D was usually used instead of 2,4,5-T. Discussion of these data and consideration of the probability of TCDD being a residue in meat and/or milk from actual use patterns, led Lo the recommendation that we draw together information from the field people who know how 2,4,5-T is used and put that together with the data we have on residues to come up with a complete picture as to what the �30 potential is for TCDD residue in human food. Further, it was recommended that we try to find areas where 2,4,5-T is used in conjunction with dairy herds and get milk .from the market there. Also recommended is that we try to get milk samples from EPA from a study they have conducted (Dr. Bovey) grazing cattle on rangeland sprayed with 2,4,5-T. Dr. Bovey described a study of movement of 2,4,5-T in water which he had conducted on a 3 acre plot given multiple treatments with 2,4,5T. 2,4,5-T was detected at only very low levels, the highest being 26 ppb. lie concluded that the possibility of contamination of ground water was unrealistic. Even wash-off from a treated area would be very slight. Discussion next centered on "residue data in tissues of animals and sheep given 2,4,5-T" in the diet (Jensen, D. J., "Investigation for Bound Residues on Tissues from Cattle Fed presented at the 165th National Meeting of American Chemical April, 1973. beef Et al 2,4,5-T" Society, Animals were fed for 28 days with a constant level of 2,4,5-T in the diet. This was the same chemical which was used for the milk study, containing about 0.5 ppm of TCDD, roughly 5 times the amount permitted by the present specification on 2,4,5-^T. 2,4,5-T data were reviewed briefly. At the maximum feeding level 2,4,5-T residues in muscle and fat were around 2 ppm, and about 8 ppm in liver. Levels were roughly proportional to the amount in the diet. Dr. Jensen discussed the trichlorophenol data results. After 7-day withdrawals of 2,4,5-T from the diet the phenol did not disappear. A new totit has been started feeding sheep the more realistic level of 300 ppm 2,4,5-T for 4 weeks followed by withdrawal for periods up to 56 days. The tissues are in hand and an analysis for 2,4,5-T and Lr.Lchlorophcnol, as well as TCDD is planned. On analysis of liver from the cattle experiment, TCDD levels (single animal anaylses) were 13, 61, 150 and 360 ppt from feeding of 100, .300, 900 and 1800 ppm of 2,4,5-T in the diet. Half of the TCDD disappeared from the liver in 7 days. Fat from cattle on the 1800 ppm feeding level contained around 2000 ppt of TCDD. There is a big dropoff ol" TCDD level in fat in the first 7 days to about half of the level aL 0-day withdrawal. In the sheep experiment, composite samples of fat and liver from 3 animals after various periods of withdrawal of 2,4,5-T have been analyzed. Again a rapid drop-off of 7 days after withdrawal of 300 ppm 2,4,5-T containing 0.5 ppm of TCDD was seen - from about 170 ppt to 40 ppt. Little decrease has been observed from 7 days to 28 days withdrawal. In �31 livers at 0-days, a level of around 200 ppt was found, decreasing to about 70 ppt with 7 days withdrawal and to about 40 ppt with 28 days withdrawal. Samples from a group of sheep slaughtered 56 days after withdrawal ol the chemical from the feed are yet to be analyzed, and some values on Individual animals as well as other tissues have yet to be completed. Dr. Crummett reported that in fat heated to 160° C 3-15 hours, containing 1000 pptn of trichlorophenol, no TCDD was found with a limit of sensitivity of 50 ppb. This concept was discussed at length with the final general agreement Lliat formation of TCDD as a result of cooking fat containing 2,4,5-T or trichlorophenol does not pose a potential problem. With the experiment wliich has been done, it has been shown that there is a very low potential for TCDD to be formed in this way, especially in view of the low level of. trichlorophenol in fat of cattle consuming 2,4,5-T. Residue data on rice was reported. The rice had been given two applications of 2,4,5-T of 1.5 lb-/A. Rice grain at harvest time had no detectable residue of 2,4,5-T with a method sensitive to 0.025 ppm, while the straw contained about 12 ppm of 2,4,5-T. These samples will be analyzed Eor TCDD. Rice samples from an area in which 2,4,5-T is used are being procured for analysis for TCDD, to determine if this crop can be a source of TCDD in human food. Another residue study reported was on wheat treated with 1 Ib. 2,4,5-T per acre, in which no residue was found in grain 56 days after application. Dr. Dutton reported on the fate of radioactive TCDD which had been added to soybean oil during the processing of the oil. About 50% of the radioactivity followed the oil through the processing. It can be removed down to the order of 3/10% by adding norite carbon black to the bleach step in the process. It is also removed by increasing the temperature oL the deodorization process to 260°C. A discussion followed on the question of whether TCDD might be found in food in the market. It was proposed we collect beef fat, as well as milk, from areas where 2,4,5-T is used, and analyze them for TCDD. �32 Dr. Youny described some seed he has collected from areas in which TCDU is 25-30 ppt in the soil—no TCDD was found in the seed. He still has sued from plants growing in soil containing 710 ppt of TCDD, which an- being analyzed now. This will give a good fix on the translocation of TCDD from soil to the seeds. He indicated that he has some sorghum samples collected from areas where 2,4,5-T had been placed at a 6" depth in the aoil, at the rate of 1000 Ib/A. Further discussion on a market surveillance followed with ideas expressed as to how to proceed. It was concluded that a protocol should bf developed at Dow after giving some thought to what we can expect to accomplish. cs_ oi Dioxin The workshop first considered to what extent TCDD is formed from the thermal stress of 2,4,5-T under environmental conditions. As has been previously reported (EPA, Dow-Langer), the apparent maximum amount of conversion of 2,4,5-trichlorophenol (or salts) to form 2,3,7,8-dioxin (TCDD) is between 0.1% and 0.3%, certainly less than 1% when heated under laboratory conditions. The work of Buu-Hoi is not sufficiently described to be repeated and present indications are that 1% represents a maximum amount of eoavers Lon. The apparent clioxins content of a material called "Toxic Fat" has boon attributed to gross contamination by /'bad" pentachlorophenol and Lelrachlorophenol. Work by USDA and others has shown that pentachlorophenol Ls also a source of "dioxions". "The use pattern determines whether any ol these contaminants will be as bad as 2,4,5-T". Tlu: possibility of 2,3,7,8-dioxin formation from combustion of materials coated with various 2,4,5-trichlorophenoxy-containing compounds has been investigated. Recent work at Dow indicates that less than 0.0001% oi. any 2,4,5-T species is converted to 2,3,7,8-dioxin on combustion (i.e. less than 1 ppt 2,3,7,8-dioxin formed from each ppm 2,4,5T burned). Work at 1'TJA and Dow which has subjected fat containing 1000 ppm of various 2,4,5-trichlorophenolics to "deep-fat frying" conditions found that after 14 hours, no 2,3,7,8-dioxin was detected, with a detection limit of 0.05 ppm. �33 We i. It en considered to what extent other compounds bearing the 2,4,5-lrichlorophenol moiety contribute to dioxins in the environment. USUA Imu investigated the photolysis of di- and trichlorophenols, both with and without a "photoactivator", riboflavin. They have identified both chlorinated phenoxyphenols and dihydroxybiphenyls, but have not detected "dioxins". This appears to suggest that the photolysis to form "dtoxins" is slower than the photolytic decomposition of dioxins, usipec.Lally in alcohol or water. Similar experiments which subjected 2,4-D and 2,4,5-T to metabolic conditions in soils (incubation) showed no detectable "dioxins". Examination of 40 fish (107 determinations) from 2,4,5-T use areas allowed no 2,3,7,8-dioxin detectable in 38 of these and "slight" positive response-;-; from 2 samples which could not be repeated on resampling. Examination of current Dow ronnel production showed no 2,3,7,8dioxin with a detection limit of 0.01 ppm. Examination of Dow pentachlorophenol showed no detectable 2,3,7,8dioxin (0.05 ppm limit of detection). All current production 2,4,5-T materials (2,4,5-TCP, 2,4,5-T esters, Silvex esters) have less than 0.1 ppm. 2,3,7,8-Dioxin is detected (0.02 to 0.099 ppm) most often.in 2,4,5-T esters. Different chemical conditions exist at several different steps for the different products and some processing conditions can lead to 2,3,7,8-dioxin but these condtions can be controlled. Dow employs >;ood tight process quality control to keep'2,3,7,8-dioxin content in any products to less than 0.1 ppm. The workshop then turned its attention to the question of contribution by other chlorophenbls to "dioxins" in the environment. There appears to be uo significant problem from pentachlorophenol, except for some uncertainty about the toxicity of hexachlorodibenzo-p-dioxins. Dow is currently investigating the identity of various "hexachloro~dioxins". Although there is no detectable 2,3,7,8-dioxin in Dow pentachlorophenol, it has been detected in several Asian pentachlorophenol samples. Similarly, heating pentachlorophenol with hydrocarbon oils and metal appears not to produce 2,3,7,8-dioxin, although some work is still in prbgress. (Crummett, L.-mger) There is some possibility that anaerobic reductive dechlorination of. liexa- or octachlorodibenzo-p-dioxin may give rise to "tetrachlorodipxins". This should be investigated. The following experiments were suggested. �34 1. Combustion of wood or grass which has been treated with 2,4,5T should be done. Norris reported 100 ppm 2,4,5-T on "twigs" after spraying at 2 Ibs/acre. One month later, this had declined to about 3 ppm. One should, therefore, burn.wood which contains these residual amounts of 2,4,5-T. Kearney suggested that one should also examine whether any 2,3,7,8-dioxin so formed is primarily in the vapors or in the ashes. 2. Examination of "heating" products of pentachlorophenol should be done to determine extent of dechlorination. (Langer has some work in progress.) 3. Examination of various hexa-dioxins to determine identity. Dow has work in progress. 5. .S tat istics The purpose of this workshop was to evaluate the statistical questions raised in the 2,4,5-T Advisory Committee dissenting opinion report and later expanded in Science, 174, 1971, pp. 1358-1359. Specifically, two main criticisms were discussed: 1. The authors of the major 2,4,5-T studies did not "milk" the data by attempting to extrapolate the dose-response curves to "very low dose" levels in an effort to learn about expected teratogenic frequencies at these low levels. 2. Multiple t-tests and chi square tests were used in place of their nonparametric equivalents or one way analyses of variance. Regarding the first criticism it was stated that to carry out this extrapolation required the assumption that the dose-response function is the same for lower doses as it is in the experimental region. This is not a reasonable assumption unless we know the mechanism by which the teratogenic responses occur. The probit, logit and one hit model all fit equally as well for most experimental data but give dose estimates orders of magnitudes apart when extrapolated to risks as low as 10 Lower additional dose levels could perhaps have been used in some of the studies but we then get into the mega-mouse argument. Even if 100,000 animals show no difference from control this does not demonstrate a "safe" dose, it only shows 99% certainty that the true risk is less than 4.6/100,000. �35 The question was raised as to which is worse, extrapolating dose response functions assuming linearity, or applying somewhat arbitrary factors to the highest no-observed-effect levels in animals? No real answer was given (see recommendations). Some concern was expressed about the size of the type II error when estimating no observed effect levels. It was felt that perhaps type II error should be considered when planning the experiments if no observed effect levels are important. Regarding the second criticism, that the most appropriate statistical tests were not used to evaluate the data, it was pointed out that the criticism was somewhat self-contradictory. The author recommended that more "sophisticated" statistical methods such as multivariate analysis should have been used, but he also pointed out that the data is nonnormal and generally discrete (-frequency of teratogenic occurrences). With the present state of statistical methodology multivariate analysis of discrete data is not practical. Multivariate analyses are generally less robust against non-normality than their equivalent univariate methods. Part of this second criticism is technically correct, however. Chi-square tests and t-tests were used when their nonparametric counterparts, Fisher's exact probability test and the Mann-Whitney U test (or Wilcoxon's test), would have been more appropriate. Multiple t-tests were used when a one way analysis of variance should have been done. However, when the data were reanalyzed using the other methods, the results were no different. In fact, the more appropriate tests will tend to show fewer statistical differences than the tests that were used. The experimental design of the studies was discussed. It was felt that log or geometric spacing of the doses was the best choice of scale. It was suggested that sample sizes inversely proportional to expected response would enhance the power of the statistical testing for the small doses where it is most needed. From an intuitive point of view we would be learning more about the lower doses than the higher.doses, which aeems reasonable. To summarize the workshop's feelings about: the criticism, it was felt that the first criticism about extrapolation to lower dose levels was questionable, with our present knowledge of teratogenic mechanisms. The second criticism was felt to be technically justified but different methods would not affect the conclusions. �36 The workshop recommends that we obtain better estimates of baseline levels of anomalies both by pooling data when appropriate and through inter-laboratory data sharing; consider sample sizes inversely proportional to expected responses; and routinely perform dose response analyses, using for instance probit or logit models, in an effort to build up enough background information to consider establishing conservative "safe" levels using procedures such as Mantel-Bryan. 111 Rule of: Reason At the Rule oC Reason seminar on Friday March 8, 1974, the participants engaged in a general discussion of risks versus benefits. The following points were made: 1. Risks and benefits may be divided into the following categories: (a) Voluntary vs. involuntary. For example, smoking vs. environmental impact of DDT. (b) Controlled vs. Non-controlled. (c) Public vs. private. (d) informed vs. uninformed. (e) Vital vs. non-vital. Primarily, one must ask when is individual risk justified for public benefit. Example: the public risk of smallpox is now so low that the-risk of individual inoculation is not justified. Applying this theory to the case, if rice cannot be grown without herbicides, as the Rice Institute contends, then the public benefit as well as the private benefit in using herbicides is great and the individual risk is low. Generally in speaking of risks, it is the involuntary risks which must be evaluated by decision makers since the individual cannot make that decision on his own. Voluntary risks are usually definable and assuming that the hazard can be understood by the user are not often the source of major controversy in technology assessment. 2. Alternatives must be evaluated in terms of benefits vs risks. From that evaluation, society can make value judgements. One method by which to do this is to consider the possible worst outcome of all alternatives and then to select the one alternative whose worst outcome is better than the �37 worst outcome of any other alternative. In making this evaluation, the public must be made aware of the nonexistence of absolute safety. The alternative of absolute safety in many Instances would be worse than the risks•of a certain alternative. Example: in the minds of many persons, the alternative of absolute safety would be worse than the risk of using chemical substances to produce food even with their implied risks. The public wants to know what the worst outcome could be and then it will make its judgements. Example: the worst that could'happen to a truck going through town filled with gas is that it will crash and burn. If a circumferential highway is available, the truck should go around the town. If the truck is carrying vital provisions, and no route is available except through the center of town, the public must make its decision based on the worst outcome vs the benefit. 'J. It was proposed that the upper limit of risk that should be accepted in any situation should be no greater than the risk of natural disease. But, 40 percent of the population is killed by heart attacks from too much fat in their diet. Yet a 40 percent figure as an upper limit of risk is too high. Query: what standard should we use as the tool to measure the upper limit of risk. 4. Risks and benefits were defined. A benefit confers an improvement in status. A risk confers a derogation in status in an area essential to life. Nonvital risks and benefits can be valued in the market place. It is easy to make judgements with skilled advice in vital risk areas. For instance, a doctor can decide when to give penicillin and when the patient should accept the risk of the side benefits of penicillin. The difficult question is the acceptance of vital risks for nonvital benefits. However, the public on an individual basis makes such judgements every day. For example, the nonvital benefits of driving are so great to the individual that he is willing to take vital risks. This is partly due to the fact that the risks can be easily visualized and the feeling on the part of the individual driver that such risks are controllable. r.n an area such as pesticides, the risks are not so easily visualized and the individual fears them more because he cannot control them. Morever, food is a nonvital benefit for the most part. It is only wften an individual is starving that he would take a vital risk to eat; for example, eating food from a swollen can. �38 5. The DDT ban was partly based on a judgement that the risks of DDT were not as well known as risks of other substances and not as controllable by the Individual. 6. Cn some instances, were functional alternatives available there would be no question but that the alternative would be used. Example: .if there were an alternative to nitrate, it would be used without hesitation since the risks of nitrate are well known. The same applies to cyclamates. The only question which remains is cost benefit. 7. After weighing the risks and benefits, the decision maker is ultimately left with the prospect of making a value judgement. Society and its values are diverse. The judgement depends on where you stand: in rural society, weeds are bad, pulling them takes time, 2, 4, 5-T gets rid of them, and all of this increases beef production. The fact that it may decrease wildlijfe habitats is peripheral to this segment of society. A value judgement then become a question of trade offs among special interest groups- Consequently, it becomes much more difficult for the regulator to decide. 8. It is the obligation of a socially responsible agency to interpret the judgements of society as to what risk is acceptable for what benefit and then to respond to that interpretation. For instance, presently, the public will accept more air pollution when there is a gasoline shortage. 9. Coming up with the criteria to make judgements based on a rule of. reason is difficult. Several approaches have been advanced: A. Quality of life review—send proposed decision to interested agencies who will thrash out the impact given the interests they represent. B. The market place—to the extent it is safe, leave the decision to the market place. This results in a personal translation of risk: how does this affect me. C. Environmental dose commitment—the prediction of the probability of radiation getting into the environment and then the use of the Pier report, to translate that into �39 the probability of causing cancer. In this way the regulator sets the magnitude of risk acceptable: the • long term risk of cancer versus the immediate benefit of more nuclear power for man. D. The probability approach—we can have cheaper rice with the use of herbicide which carries with it one chance in a million of a birth defect. Compared to the risks of birth defects from other substances, this fades into the background of importance. Although the probability of botulism from eating home canned food is 70 times greater than in eating processed food, much of the public is willing to assume that risk because of perceived benefits. They see the probability of botulism from eating home-canned food as very low, even though it is not, because the perceived benefits are high. 10. Unknown risks enter into decisions. First, quantify the Tacts you do know and then give that, along with the uncertainties involved, to the decision maker. Then the decision maker uses his judgement. 11. It is well known that the public makes conscious choices among varying hazards and nonvital benefits. If we could quantify the differences in risk the public is willing to take, we could make socially acceptable decisions based on this quantification. For example, the hazards of smoking during pregancy are better documented and more immediate than the hazards of smoking generally. If statistics on how many women give up smoking during pregnancy and then return to it after birth were available, we might be able to make one judgement on how individuals quantify differences in risk. If such information were available on a variety of issues, it would be possible to quantify acceptable risk and therefore make socially acceptable decisions for the public. 12. We could also quantify the benefit: how much death is a certain benefit worth? The examples are not widely applicable since those decisions which we know will result in death are not widespread. The public is willing to support the building of Golden Gate Bridge though they know at least 5 lives are likely to be sacrificed. However, the perceived benefit to millions of people is great and the immediacy of the risk is �40 more distant. Each individual feels that the lives sacrificed are not likely to be theirs or their family. The public is willing to take much greater risk when the risk is not perceived as a personal one. Risks & Benefits of 2. 4, 5-T A. Risks of 2, 4. 5-T I Risks to Human Health 1. Toxicity factor (300 to 500 mg/kg acute oral 10 mg/kg chronic 90 day). 2. Chronic toxicity 3. residues in food 4. 5. teratogenicity 6. population at risk 7. metabolites 8. ' anxiety 9. II extra-sensitivity economic cost Risks to the Environment 1. toxicity to fauna (acute and chronic) 2. phytotoxicity 3. habitat modification 4. increased erosion & runoff �41 5. mobility 6. aesthetics 7. alternative products 8. fire hazard (oak) B. Benefits of 2. 4, 5-T T Range Weed Control 1. 2. aesthetics 3. wildlife habitats 4. elemination of harmful plants 5. secondary tsetse fly control 6. reduced evapotranspiration 7. water erosion 8. economic well-being of ranchers 9. TT increased food supply reduced manpower requirements Rice Weed Control 1. increased food supply 2. reduced manpower requirement 3. economic advantage to growers 4. increased bird populations �42 T I I Utility & Rights-of-Way 1. lower cost vegetation management 2. cheaper, more dependable power and communication 3. lower personnel hazard 4. reduced erosion 5. habitat diversity a. faunal diversity 6. reduced fire hazard i lv Forestry Uses 1. increased commercial timber growth 2. lower production cost 3. [altered habitat] 4 increased personnel safety > fire protection 5. V Roadside Uses including Rail 1. less traffic hazard to man & deer 2. aesthetics 3. cheaper maintenance 4. reduced fire hazard 5. water erosion reduced 6. personnel safety �43 V I! Miscellaneous Uses 1. general fire control 2. general flood control 3. general industrial vegetation control Risks & Benefits of TCDD T Risks to Man 1. acute toxicity 0.6 ug/kg a. dermal-chloracne 2. chronic toxicity (1 x 10-10 gm) a. enzyme inhibition? b. intracellular (endoplasmic vitriculura) c. liver 3. teratology a. b. 4. IT (LD-50) potential - not proven rat [no effect level 3 x 10-8 gm] oral dose of 30 mg/kg 1.25 x 10-7 pos. in other labs: no effect at 1 x 10-6 (various species & strains) residues in food (fish) Environmental Risks 1. toxicity to fauna (inter & intra - specific variation, including teratogenicity) i 2. bioccumulation 3. persistence (ingestion and retention within an organism) �44 4. .pllytoreproductive effects •5. mobility 6. conversion through fire 7. uncertainty due to limited scope of testing Order of Benefits 1. Economic Food Timber Industrial factors Alternatives ;i. b. c. mechanics other chemical combinations nothing Order of Risks 1. health a. b. 2. occupational (mfg., trade, application) teratogenicity women of child bearing age. Hazard to wildlife Noted Reference Material: i Committee on Public Engineering Policy 1972 Perspectives on Benefit-Risk Decision Making. (Washington: National Academy of Engineering) viii + 157 pp. �45 Birkhoff, George D., A Mathematical Approach to Ethics, Volume 4, Newman, The world of Mathematics, Simon & Schuster Darby, William J., 1973 Acceptable Risk and Practical Safety: Philosophy in the Decision-Making Process. J. Am. Med. Assoc., 224; 1165-1168. Environmental Studies Division 1973 The Quality of Life Concept: A Potential New Tool for Decision-Makers. (Washington: Environmental Protection Agency) xv + 397pp. Environmental Studies Division 1972 Quality of Life Indicators. (Washington: Environmental Protection Agency) ii .+ 83 pp. Panel on Chemicals and Health 1973 Democratic Representation (Washington: National Science Foundation) xi + 211 pp. � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 178 Folder The folder containing the original item. 5224 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Fullerton, Raymond W. Margaret Bresnahan Carlson Alfred R. Nolting Description An account of the resource <strong>Corporate Author: </strong>United States Department of Agriculture (USDA), Office of the General Counsel Date A point or period of time associated with an event in the lifecycle of the resource 1974 Title A name given to the resource Memorandum with attachment: From USDA, Office of the General Counsel with enclosed final report on the 2,4,5 -T scientific workshop, March 8-9, 1974, Washington, D.C. ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 198 Folder The folder containing the original item. 5574 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Gordon, Paula D. Date A point or period of time associated with an event in the lifecycle of the resource July 27 1981 Title A name given to the resource Agent Orange Assessment ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 186 Folder The folder containing the original item. 5383 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Gottlieb, Martin Source A related resource from which the described resource is derived New York Date A point or period of time associated with an event in the lifecycle of the resource October 25 1982 Title A name given to the resource The Vietnam War That's Still Being Fought ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 202 Folder The folder containing the original item. 5707 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Gough, M. Source A related resource from which the described resource is derived Dioxin, Agent Orange - The Facts Date A point or period of time associated with an event in the lifecycle of the resource 1986 Title A name given to the resource Appendix: Calculation of the Amount of Dioxin Exposure of a Person Standing under a Ranch Hand Spray Mission ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 197 Folder The folder containing the original item. 5546 Series The series number of the original item. Series VIII Subseries I Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Gough, Michael Title A name given to the resource The Political Assessment: A Congressional View ao_seriesVIII