1 15 3 https://www.nal.usda.gov/exhibits/speccoll/files/original/f11abf4af56863a9455a349556df7a88.pdf b4a09025dca32840bda0307cf9a00341 PDF Text Text Item ID Number °0631 Author Erickson, Mitchell D. Research Triangle Institute, Research Triangle Park, NC Ronort/ArtiClO TltlO Acquisition and Chemical Analysis of Mother's Milk for Selected Toxic Substances Journal/Book Title Year 190 Month/Day December ° Color Number of Images 168 DBSOrlptOn NOtOS Alvin L Youn 9 filcd this item under the category "Human Exposure to Phenoxy Herbicides and TCDD" Tuesday, February 20, 2001 Page 631 of 680 �J H.fc. e.V.«A. l r t * A Wl PB81-231029 Acquisition and Chemical Analysis of Mother's Milk for Selected Toxic Substances * Research Triangle Inst. Research Triangle Park, NC Prepared for Environmental Protection Agency Washington, DC Dec 80 U.S. DEPARTMENT OF COMMERCE National Technical Information Service �United States Environmental Protection Agency Office of Pesticides and Toxic Substances Washington, DC 20460 EPA-560/13-80-029 December 1980 L - 2 3 1029 Toxic Substances &EPA Acquisition and Chemical Analysis of Mother's Milk for Selected Toxic Substances REPRODUCED 3V NATIONAL TECHNICAL INFORMATION SERVICE U.S. DEPARWfNI Of COMMERCE SPRWflflELO. VA 22161 �TECHNICAL REPORT DATA (Pleat read Inttructiom on the reverie before completing) 3. R 1. REPORT NO. 560/13-80-029 4. TITLE ANOSUBTITLS 6. REPORT DATE ACQUISITION AND CHEMICAL ANALYSIS OF MOTHER'S MILK FOR SELECTED TOXIC SUBSTANCES December, 1980 S. PERFORMING ORGANIZATION CODE 31U-1521-21 + 22 '__ 7.AUTH0R(S)Mitchell D. Erickson, Benjamin S. H. Harris, III, Edo D. Pellizzari, Kenneth B. Tomer, Richard D. Waddell and Donald A. Whitaker m. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO. Research Triangle Institute It. CONTRACT/GRANT NO. P. 0. Box 12194 Research Triangle Park, NC 27709 68-01-3849 - Task 2 12. SPONSORING AGENCY NAME ANO ADDRESS 13. TYPE OF REPORT ANO PERIOD COVERED Field Studies Branch, Exposure Evaluation Division Office of Pesticides and Toxic Substances, U. S. Environmental Protection Agency, Washington, DC 20460 Task Final 1/23/78-4/18/80 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES Project Officer: Joseph Breen . ABSTRACT RACT Samples of mother1s milk were collected from Bayonne, NJ; Jersey City, NJ; Pittsburgh, PA; Baton Rouge, LA; and Charleston, WV, and analyzed for volatile (purgeables) and semivolatile (extractable) organics using glass capillary gas chromatography/mass spectrometry/computer. In the volatile fraction, 26 halogenated hydrocarbons, 17 aldehydes, 20 ketones, 11 alcohols, 2 acids, 3 ethers, 1 epoxide, 14 furans, 26 other oxygenated compounds, 4 sulfur-containing compounds, 7 nitrogen-containing compounds, 13 alkanes, 12 alkenes, 7 alkynes, 11 cyclic hydrocarbons, and 15 aromatics were found, including major peaks for hexanal, limonene, dichlorobenzene, and some esters. The levels of dichlorobenzene appeared to be significantly higher in the samples from Jersey City and Bayonne than in samples from other sites. Jersey City samples also appeared to have significantly higher levels of tetrachloroethylene. Charleston and Jersey City samples appeared to have significantly higher levels of chloroform; however, chloroform was observed in the blanks at about 20% of that in the samples. Due to the small sample size and lack of control over the solicitation of sample donors, the data cannot be used to extrapolate to the general population. Fewer semivolatile compounds of interest were found. Polychlorinated naphthalenes, polybrominated biphenyls, chlorinated phenols, and other compounds were specifically sought and not detected (limit of detection about 20-100 ng/mL milk). Polychlorinated biphenyls (PCBs) and DDE were found. 17. a. DESCRIPTORS KEY WORDS AND DOCUMENT ANALYSIS b.lDENTIFIERS/OPEN ENDED TERMS c. COSATi Field/Croup Mother's Milk Purge and Trap GC/MS Sampling Milk Chlorinated Organics f f. It. DISTRIBUTION STATEMENT 19. SECURITY CLASS <TMs Rtpon) UNCLASSIFIED RELEASE TO PUBLIC EPA f*rm 2220.1 (R««. 4-77) 20. SECURITY CLASS <TM* page) PNCVIOU* COITION i* OMOLCTC 21. NO. Of PAGES 164 22. PRICE �EPA 560/13-80-029 ACQUISITION AND CHEMICAL ANALYSIS OF MOTHER'S MILK FOR SELECTED TOXIC SUBSTANCES by Mitchell 0. Erickson, Benjamin S. H. Harris, III, Edo D. Pellizzari, Kenneth B. Tomer, Richard D. Waddell and Donald A. Whitaker Contract No. 68-01-3849 Task 2 Project Officer: Joseph Breen Field Studies Branch Exposure Evaluation Division Office of Pesticides and Toxic Substances U. S. Environmental Protection Agency Washington, DC 20460 December 1980 / -a. �DtCLAIMER This document has been reviewed and approved for publication by the 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. ii �ABSTRACT Samples of mother's milk were collected from Bayonne, NJ; Jersey City, NJ; Pittsburgh, PA; Baton Rouge, LA; and Charleston, WV, and analyzed for volatile (purgeables) and semivolatile (extractable) organics using glass capillary gas cbromatography/mass spectrometry/computer. In the volatile fraction, 26 halogenated hydrocarbons, 17 aldehydes, 20 ketones, 11 alcohols, 2 acids, 3 ethers, 1 epoxide, 14 furans, 26 other oxygenated compounds, 4 sulfur-containing compounds, 7 nitrogen-containing compounds, 13 alkanes, 12 alkenes, 7 alkynes, 11 cyclic hydrocarbons, and 15 aromatics were found, including major peaks for hexanal, limonene, dichlorobenzene, and some esters. The levels of dichlorobenzene appeared to be significantly higher in the samples from Jersey City and Bayonne than in samples from other sites. Jersey City samples also appeared to have significantly higher levels of tetrachloroethylene. Charleston and Jersey City samples appeared to have significantly higher levels of chloroform; however, chloroform was observed in the blanks at about 20% of that in the samples. Due to the small sample size and lack of control over the solicitation of sample donors, the data cannot be used to extrapolate to the general population. Fewer semivolatile compounds of interest were found. Polychlorinated naphthalenes, polybrominated biphenyls, chlorinated phenols, and other compounds were specifically sought and not detected (limit of detection about 20-100 ng/mL milk). Polychlorinated biphenyls (PCBs) and DDE were found. iii �CONTENTS Abstract Figures Tables List of Abbreviations and Symbols Acknowledgments 1. 2. 3. 4. 5. 6. 7. Introduction Summary and Conclusions Recommendations Selection of Sampling Sites Sample Collection Sample Analysis Methods Results References Appendices A. B. C. 0. E. iii vi vii ix x 1 15 16 18 32 36 45 61 Data Collection Instruments Sampling and Analysis of Volatile Organics in Milk Analysis of Semivolatile Organics Compounds in Milk Volatile Compounds Identified in Selected Purges of Mother's Milk Semivolatile Compounds Identified in Selected Extracts of Mother's Milk Preceding page blank 68 104 112 119 144 �FIGURES Number Page B-l Diagram of headspace purge and trap system D-l Total ion current chromatogram from GC/MS analysis in sample no. 1081 (Bayonne, NJ) D-2 Total ion current chromatogram from GC/MS analysis in sample no. 1040 (Bayonne, NJ) D-3 Total ion current chromatogram from GC/MS analysis in sample no. 1107 (Jersey City, NJ) D-4 Total ion current chromatogram from GC/MS analysis in sample no. 1115 (Jersey City, NJ) D-5 Total ion current chromatogram from GC/MS analysis in sample no. 2048 (Pittsburgh, PA) D-6 Total ion current chromatogram from GC/MS analysis in sample no. 2071 (Pittsburgh, PA) D-7 Total ion current chromatogram from GC/MS analysis in sample no. 3053 (Baton Rouge, LA) D-8 Total ion current chromatogram from GC/MS analysis in sample no. 3111 (Baton Rouge, LA) E-l Total ion current chromatogram from GC/MS analysis in sample 1032 (Bayonne, NJ) E-2 Total ion current chromatogram from GC/MS analysis in sample 2121 (Pittsburgh, PA) E-3 Total ion current chromatogram from GC/MS analysis in sample 3095 (Baton Rouge, LA) E-4 Total ion current chromatogram from GC/MS analysis in sample 4093 (Charleston, WV) 107 VI for volatiles 122 for volatiles 125 for volatiles 129 for volatiles 132 for volatiles 135 for volatiles 138 for volatiles 141 for volatiles 143 for semivolatiles 146 for semivolatiles 148 for semivolatiles 150 for semivolatiles 152 �TABLES Number 1 Comparison Between Human and Cow's Milk 2 Levels of Organic Compounds Found in Human Milk in the United States 3 Ranking of Pesticides and PCBs by Reported Concentrations in Human Milk 4 Levels of Organic Compounds Found in Human Milk Outside the United States 5 Summary of PCN Concentrations Found Near Manufacturing and Use Sites 6 7 8 9 10 11 12 13 14 15 16 Prevalent Halogenated Compounds in Ambient Air and Water of Rahway/Woodbridge, Boundbrook and Passaic, NJ Estimated Daily Intake of Selected Volatile Compounds and Expected Concentrations in Blood in Northern New Jersey Total Daily Intake of Target Compounds, Pesticides, PCBs, BaP and Metals and Concentrations in Blood in Northern New Jersey . . Potential Emissions from Chemical Industry in Baton Rouge, LA. . . Prevalent Halogenated Compounds Occurring in Ambient Air and Water of Baton Rouge, Geismar and Plaquemine, LA Potential Emissions from Chemical Industry in Plaquemine, Geismar, and St. Gabriel, LA Method Validation Recovery of Selected Volatile Standards from Milk Method Validation Recovery of Semivolatile Compounds Spiked into Raw Cow's Milk Operating Conditions for GC/MS Analysis of Purgeables Operating Conditions for the GC/MS Analysis of Semivolatiles . . . Summary of Qualitative Identifications of Volatile Compounds in Mother's Milk vii 5 9 10 19 22 23 24 27 29 30 38 39 42 43 46 �TABLES CONT'D. Number Page 17 Volatiles Quantitated in Mother's Milk Samples (ng/mL) 52 18 Summary Statistics for Volatile Compounds by Site 55 19 Significance of the Differences in the Geometric Means by Site . . 56 20 Spearman Correlation Coefficients for Volatile Organics Found in Mother's Milk 57 21 Quality Control Results for Volatiles in Milk 58 22 DDE and Tetrachlorobiphenyl Levels in Selected Mother's Milk Samples B-l Instrumental Operating Conditions D-l Volatile Compounds Identified in Purge of Sample No. 1081 (Bayonne, NJ) D-2 Volatile Compounds Identified in Purge of Sample No. 1040 (Bayonne, NJ) D-3 Volatile Compounds Identified in Purge of Sample No. 1107 (Jersey City, D-4 Volatile Compounds (Jersey City, D-5 Volatile Compounds D-6 NJ) Identified in Purge of Sample No. 1115 NJ) Identified in Purge of Sample No. 2048 (Pittsburgh, PA) Volatile Compounds Identified in Purge of Sample No. 2071 (Pittsburgh, PA) D-7 Volatile Compounds Identified in Purge of Sample No. 3053 (Baton Rouge, LA) D-8 Volatile Compounds Identified in Purge of Sample No. 31]1 (Baton Rouge, LA) E-l Semivolatile Compounds Identified in Extract of Sample 1032 (Bayonne, NJ) E-2 Semivolatile Compounds Identified in Extract of Sample 3095 (Baton Rouge, LA) E-3 Semivolatile Compounds Identified in Extract of Sample 3095 (Baton Rouge, LA) E-4 Semivolatile Compounds Identified in Extract of Sample 4093 (Charleston, WV) viii 60 108 120 123 126 130 133 136 139 142 145 147 149 151 �LIST OF ABBREVIATIONS AND SYMBOLS ABBREVIATIONS DDT dpra ECD GC MS NICIMS OMB PBBs PCBs PCF PCN PLF SQ l,l-Bis(p_-chlorophenyl)-2,2-trichloroethane Disintegrations per minute Electon capture detection Gas chromatography Mass spectrometry (electron impact ionization) Negative ion chemical ionization mass spectrometry Office of Management and Budget Polybrominated biphenyls Polychlorinated biphenyls Participant Consent Form Polychlorinated Naphthalene Participant Listing Form Study Questionnaire ix �ACKNOWLEDGMENTS The authors thank the Project Officer, Dr. Joseph Breen, for his guidance and advice. Sandra P. Parks, David L. Newton, and Larry C. Michael are acknowledged for their assistance in the laboratory. Nora P. Castillo and Kent W. Thomas are thanked for their assistance with mass spectral interpretation. Pamela A. Gentry, Fred A. McKinney, Stephen P. Burke, and Barbara L. Bickford are thanked for sample analysis using mass spectrometry. Personnel who assisted in the milk collection are greatly appreciated: Elizabeth Bartholomew, Bayonne Hospital, Bayonne, NJ; Jules Rivkind and Trudy Strunk, Medical Center Hospital, Jersey City, NJ; Ian Holtzman, MageeWomen's Hospital, Pittsburgh, PA; Lewis Tracbtman, Louisiana Health Department, New Orleans, LA; Maxine Parker, Baton Rouge Area Regional Nursing Consultant, Baton Rouge, LA; Clementine Martine, Public Health Nursing Supervisor of the East Baton Rouge Parish Health Unit, Baton Rouge, LA; and N. N. Sehgal, Charleston Area Medical Center (Memorial Division), Charleston, WV. Finally we would like to thank the 42 women who so kindly donated the samples. �SECTION 1 INTRODUCTION BACKGROUND It is becoming increasingly important to correlate ambient environmental pollutant levels with human body burden. Establishment of this correlation ("exposure assessment") may provide a link between pollution and health effects. This correlation is of interest for both scientific research and regulatory risk assessment. Measurement of pollutant body burden levels generally requires invasive techniques (exceptions are breath and urine sampling) which are undesirable from the subjects' viewpoint. Some invasive techniques are generally regarded as acceptable (e.g., blood samples), while others are generally considered unacceptable from living donors (e.g. adipose tissue, internal organs, etc.). Mother's milk is an attractive medium for several reasons: (1) sample collection is reasonably straightforward; (2) milk contains a high amount of fat (about 3.5 percent, as shown in see Table 1), so fat-soluble pollutants such as DDT and polychlorinated biphenyls (PCBs) are likely to be found in higher concentrations in milk than in blood or urine; (3) large (50-100 ml) volumes are easily collected for analysis, increasing analytical reliability and detection limit; and (4) the population of nursing mothers is large relative to pathology samples such as adipose tissue. In addition, an assessment of pollutant concentrations in mother's milk may be used to predict the pollutant intake by the nursing infant. The major disadvantages of mother's milk as a human-sampling medium relate to the sampling demography: only young-to-middle-aged females are nursing. Thus, any use of mother's milk in a probability-based sampling framework extrapolated to the general population would be fraught with difficulties, such as locating donors. �Table 1. COMPARISON BETWEEN HUMAN AND COW'S MILK (1) Parameter Human Milk Cow's Milk Water and solid content Same in both; 87 to 87.5 percent is water Calories Same in both; 20 calories per ounce Protein 1 to 1.5 percent; 60 percent of this is lactalbumin and 40 percent car.ein 3.5 percent; 15 percent of this is lactalbumin and 85 percent casein Carbohydrate (in form of lactose) 6.5 to 7.5 percent 4.5 to 5.0 percent Fat(s) Variable, but both have approximately 3.5 percent. (Differs qualitatively) Contains more olein, which is Contains more volatile fatty is readily adsorbed acids, which are irritating to the gastric mucosa Digestion of fat easy Digestion of fat sometimes difficult Minerals 0.15 to 0.25 percent Vitamins 0.7 to 0.75 percent. Contains more of all minerals with the exception of iron and copper Iron content is low in both milks, approximately: 1.5 mg/1 0.5 mg/1 Varies with maternal intake (continued) �Table 1 (cont'd.) Parameter Vitamin A Vitamin B Vitamin C Thiamine Riboflavin Vitamin D Vitamin E Digestion Human Milk Cow's Milk Relative large amounts in both milks Probably adequate in both milks More is found in human milk Higher content in cow's milk Higher content in cow's milk Relatively small amount in both milks Satisfactory level in breath milk Cow's milk has a higher buffer content and can therefore adsorb much more gastric acid than breast milk before it reaches the acidity necessary for digestion. The large amount of casein on cow's milk make large, tough curds in the stomach as compared with the fine, easily broken down curds of breast milk �The purpose of this study was to measure levels of environmental pollutants in human milk by gas chromatography/mass spectrometry (GC./HS) and to evaluate the utility of using this body fluid in specific pollutant studies for populations in the vicinity of chemical manufacturing plants and/or industrial user facilities. All routes of exposure, i.-e., air, water, particulate, clothing and food were of interest. Mother's milk samples were acquired and analyzed for selected industrial chemicals. The chemicals of interest included: polychlorinated naphthalenes (PCNs), tetrachloroethylene, trichloroethane, dichloropropanes, benzene, polybrominated biphenyls (PEBs), chlorinated phenols, toluene, chlorinated benzenes, and chloroform. Where possible, any other chemicals found in the extracts were identified and quantitated. The levels of selected organic compounds in mother's milk were investigated to assess the possibility of using this medium as an indicator of body burden for a wide range of organic compounds. For this feasibility study, no attempts were made to develop a statistically valid sample; sites were selected as having a high probability of pollutant detection and subjects were selected on a volunteer basis. LITERATURE REVIEW A review of the literature concerning pollutants in mother's milk was conducted. A computer search of MEDLARS II and ORBIT—III yielded 108 citations. These citations, plus personal contacts and manual searches yielded the data discussed below. By far, most of the literature on environmental pollutants in mother's milk deals with chlorinated insecticides (e.£. DDT). PCBs have also been studied. Only a few references discuss the presence of other compounds in milk. Table 2 lists the levels of pollutants found in mother's milk in the United States. Table 3 summarizes these findings. Table 4 summarizes pollutants found in mother's milk outside the United States. With the exception of one reference (27) regarding 1,2-dichloroethane exposure, all of the compounds found in mother's milk are semivolatile (extractable) halogenated compounds. �Table 2. LEVELS OF ORGANIC COMPOUNDS FOUND IN HUMAN MILK IN THE UNITED STATES Compound Sample Matrix B-BHC Milk Milk Y-BHC Milk Fat Total BUG Milk Milk Milk £,p_' -DDD Milk Milk Fat Milk o,p_'-DDE Milk £,p_* -DDE Milk Milk Milk Milk Milk Fat Milk DDE Milk Milk Milk Milk Milk Mean (ppb) Range (ppb) Number of Determinations Locations References T-10 T-28 57 40 AR, MS CO 2 3 30-270 53 PA 4 6.5 7.7 6.2 <0.1-20.2 n.d.-37.0 3.6-9.0 14t 28 7 US TX Houston, TX 5 6 6 4.7 <0.1-14 n.d.-30 T-5 14t 53 40 US PA CO 5 4 3 <0.1-2.8 14t US 5 10-1720 5.2-981 13.4-236 16.7-138 790-4350 79-386 57 14t 28 7 53 40 AS, MS US TX Houston, TX PA CO 2 5 5 6 4 3 30* 4 5 1** AZ Chicago, IL Wenatche, WA Phoenix, AZ US 7 8 8 8 8 0.5 83 10.8 1.0 227 29 84.1 92.4 1766 194 60 30 30 100 74-314 20-90 <10-140 _** 70-120 (continued) �Table 2 (cont'd.) Compound o,£' -DDT Sample Matrix Milk Milk Milk Milk Mean (ppb) Range (ppb) 92 25 10 10-840 <0.1-10.8 5-36 T-13 Number of Determinations Locations References 57 AR, MS 14 1 30* 40 US AZ CO 2 5 7 3 £,£'-DDT Milk Milk Milk Fat Milk 29 114 513 7.8-89 9-383 90-2120 7-109 14t 30* 53 40 US AZ PA CO 5 7 4 3 DDT (unspecified) Milk Milk Milk Milk Milk Milk 100 60 60 70 80-130 < 10-220 50-90 10-110 n.d.-770 4 5 1 ** 40 32 Chicago, IL Wenatche, WA Phoenix, AZ US CO DC 8 8 8 8 3 9 Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk 334 20-2760 40.4-156 SD=100 SD=130 SD=100 SD=170 SD=150 SD=80 SD=120 59-1899 15-133 185-721 n.d.-770 57 14 14 20 19 27 34 6 18 38 14 7 32 AR, MS 2 Total DDT Equiv. 130 70.5 100 170 180 220 170 150 180 447 75 323 130 _** US Long Island, NY Rochester, NY Chicago, IL Lexington, KY Nashville, TN Memphis, TN Los Angeles, CA MS, AK Nashville, TN MS, AK Washington, DC (continued) 10 10 10 10 10 10 11 11 11 11 9 �Table 2 (cont'd.) Compound Sample Matrix Mean (ppb) Range (ppb) Number of Determinations 0.4 6.2 3.3 7.5 T-50 2.9-14.6 n.d,-21 1.9-21 T-ll 57 14t 28 7 40 AR, MS US TX Houston, TX CO 2 5 5 5 3 T-30 <0.1-4.4 40-460 T-5 57 14t 53 40 AR, MS US PA CO 2 5 4 3 Locations References Dieldrin Milk Milk Milk Milk Milk Heptachlor Epoxide Milk Milk Milk Fat Milk t-Nonachlor Milk 1 T-10 57 AR, MS 2 Oxychlordane Milk 5 T-20 57 AR, MS 2 PCBs Milk Milk Milk T MO T <40-100 40-100 57 39 40 AR, MS CO CO 2 12 3 Nicotine NOTES: 4 1.7 160 n.d.-195 6 Breast CA 13 Fluid BUG = benzenehexachloride (hexachlorocyclohexane) ODD a 2,2-bis(chlorophenyl)-l,l-dichloroethane DDE = l,l-dichloro-2,2-bis(chlorophenyl)ethylene DDT = l,l,l-trichloro-2,2-bis(chlorophenyl)ethane Total DDT equiv. = sum of all DDT-related peaks calculated as if all were DDT PCBs = polychlorinated biphenyls. Quantitation generally based on comparison to an Aroclor mixture T = trace n.d. = not detected SD = standard deviation t = 5 women. Separate determinations make total of 14 samples. * = 6 women. Separate samples makes total of 30 samples. ** = unspecified pool of donors in Denver and other US areas, no range given. Missing values indicate no data in original article �Table 2 (cont'd.) NOTES (cont'd.): Mean values were taken from original citation where available; otherwise arithmetic mean was calculated, counting "ND" values as zero and "T" values as 0.5 times the lowest reported value. oe �Table 3. RANKING OF PESTICIDES AND PCBs BY REPORTED CONCENTRATIONS IN HUMAN MILKa Weighted Mean Concentration (ppb) Number of Samples DDEC 99 103 DDTC 94 100 PCBsC <10 96 Oxychlordane 5 57 Dieldrin 4 92 DDDC 4 54 Hepta'chlor epoxide 4 71 BHCC 3 106 t-Nonachlor 1 57 Compound milk only. Mean value calculated from a weighted mean of values in Table 2. Where either the mean or number of samples analyzed were unavailable, the data were excluded from calculation. C A11 isomers summed. �Table 4. LEVELS OF ORGANIC COMPOUNDS FOUND IN HUMAN MILK OUTSIDE THE UNITED STATES Compound Suple Matrix a- me HI Ik t-rnc H1U Milk Hi Ik HI Ik Hi Ik Milk Y-MC HI Ik Milk Milk Fit Milk Fat HI Ik Milk Mrnn (ppb) Range <PPI>) 0.58 0.1-1.9 Number of Positive! Location Date Reference 50 17 Norwujr 1975 14 4.69 70 200 2 BO 4 2 1.2-17.8 NO -900 80-910 10-850 1-16 ND-21 50 96 22 9 SO 100 49 64 19 7 42 91 Norway Gemanjr Vienna Kuril Austria Leiden (Neth.) Canada 1975 1971 1973 1973 1969 1975 14 IS 16 16 17 13 10.91 1.0-35.8 NO 26-114 40- 100 17 0 19 7 <l-35 50 96 22 9 29 147 Norway Cenany Vienna Rural Austria Israel Canada 1975 1971 1973 1973 1975 1967-8 14 15 16 16 18 19 0.3-3.2 SO 34 Norway 197S 14 1.7-45.5 7-33 50 19 SO 19 Norway En f land 1975 1964 14 20 21 48 63 10.1 1 I-8HC Milk 1.14 Total MIC HI Ik Milk 9.4 13 Milk 9.9 £.£'-DDD Nurt.fr of Determinations 29 ODD Hi Ik 3-14 67 O.g'-DDB Milk Milk IS. 02 9.5 1.6-43.8 50 29 30 £.,>'-Doe Milk HI Ik Milk Milk HI Ik HI Ik Milk Milk Milk Milk 65.10 0.9-113.2 6-699 ND-600 SO 168 96 29 147 SO 6 9 100 19 SO 167 95 7 90 21.7 97 30 35 19 35 73 6-770 17-68 9-40 . •-I44 40-100 18 Israel 12 SO 6 100 19 Australia 1970 Norway Israel 197S 1975 14 18 Norway Portugal Ccnany Israel Canada Leiden (Neth.) N«w Irunswlck Nova Scotia Canada Enfland 1975 1972 1971 1975 1967-8 1969 2973 3973 !97S 1964 14 22 (continued) IS 18 19 17 23 23 24 20 �Table 4 (cont'd.) Compound DDE o.p_'-DDT E.P/-DDT DOT Sample Matrix HI Ik Milk Fat Milk Fat Milk Mean (PRb) IOS 3180 1920 61 HI Ik Milk Milk Milk 1S.S2 7.1 Milk Milk Milk Milk HI Ik Milk Milk Milk Hllk Milk 17.89 Milk Hllk Fat Milk Fat Milk s 3 90 7.1 12 16 13 6 6 45 36 1060 1760 10 Range <H>b) Nunber of Positives Location 67 22 9 26 67 22 9 26 Australia Vienna Rural Austria M. Australia 1970 1971 1971 1970-1 21 16 16 25 50 21 147 100 49 Honey Israel Canada Canada 1975 1975 1967-8 1975 14 16 19 24 SO 167 95 6-10 «2-ll t-21m 20-75 50 16S 96 29 147 50 6 9 100 19 Norway Portugal Gentanjr Israel Canada Leiden (N«th.) Hex Brunswick Nova Scotia Canada England 1975 1972 1971 1975 1967-8 1969 1971 1971 1975 1964 14 22 IS 18 19 17 21 21 24 20 7-160 100-2680 1030-2530 2-25 67 22 9 26 67 21 9 26 Australia 1970 1971 1971 1970-1 21 16 16 25 12-450 1930-7950 3420-5970 15-112 1.6-120.9 <1-31 ND-4B 2.1-118.1 1-145 10-250 1-144 Nudier of Detenlnationa 12 SO 6 9 100 19 Date Vienna Rural Austria M. Australia (continued) Reference �Table 4 (cont'd.) Compound Sample Matrix Mean (PPb) Milk Milk Milk Fit Milk Fat Milk F«t Milk Milk Milk Milk Milk Milk 81.74 106 1390 3480 34 BO 320 Milk Milk Milk Fit Milk Fit Milk Fat Milk Milk Milk Milk Milk Milk Milk Z.75 40 40 90 90 6 7.0 5 Aldrin Milk 21. « Heptachlor Epoxid* Milk Milk Milk Milk Milk Range (.b |p) Totil DOT Equlv. Dieldrln 141 139 71 378 12R s I 2 6 1.57 9.1 3 1.2 1 NintfMT of Determinations Nimbrr of Positives S. 2-349.0 < 10-780 220-2580 330-18800 110-11400 30-870 15-580 10-1020 19-137 3-5868 7S-170 50 160 19 34 48 96 67 147 26 290 19 0.3-3.6 5-31 <10-BO <10-170 < 10- 250 1-29 SO 168 19 34 48 67 29 147 26 50 100 19 29 50 1 50 29 147 SO 100 18 1-60 3-11 0.1-10.7 ND-6 1-13 0.6-2.6 <l-23 0.3-3.5 ND-3 50 167 19 34 48 96 67 26 290 19 6 IS 26 40 84 19 SO 69 Location Date Reference Norway Portugal Ontario Ontario Ontario Germany Australia Canada W. Australia Guatemala England 19 75 1972 1973-4 1971-2 1969-70 1971 1970 1967-8 1970-1 1973-4 1964 14 22 26 26 26 15 21 19 25 27 24 Norway Portugal Ontario Ontario Ontario Australia. Israel Canada W. Australia Leiden (Heth.) Canada England 1975 1972 1973-4 1971-2 1969-70 1970 1975 1967-8 1970-1 1969 1975 1969 14 14 26 26 25 21 18 19 25 17 24 20 Norway 1975 14 Norway Israel Canada Leiden (Heth.) Canada 1975 1975 1967-1 1969 1975 14 IS 19 17 24 (continued) �Table 4 (cont'd.) Coepound HCB KB Suple MatrU Milk HI Ik Fit Milk Fat Ml Ik Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk Milk Mem (ppb) 9.1 100 1240 1670 25 2 Fit Fat Fat 1200 1200 1000 Fit Fit IS40 1290 Ring* (I) P* 1.T-60.S ND-2SO 260-4160 2140-5110 12-14 NO- 21 100-2500 200-1000 700-12000 90 22 ia 12 S80-3780 9SO-1S70 1S-JO 12-12 ND-68 HiMber of Determinations Number of Positives location Date Reference Honey Ontario Vienna Rural Austria M. Australia Canada 197S 1971-4 1973 1971 1970-1 197S 14 26 16 16 2S 24 19 34 41 64 22 9 6 9 100 Ontario Ontario Ontario Germany Vienna New Brunswick Nova Scotia Canada 1971-4 1971-2 1969-70 1971 1971 197] 1971 1971 I97S 26 26 26 IS 16 16 21 21 24 SO 19 22 9 26 100 SO 19 14 48 96 22 9 6 9 100 22 9 26 81 Oxychlordane Milk 1 KD-2 100 77 Canada 197S 24 truisNonachlor Milk 1 NO-2 100 77 Canada 197S 24 1.2-DichloroethiM Milk 60 00 1 t 2B NOTES: BUG = ODD DDE = DDT = Total PCB = benzenehexachloride (hexachlorocyclohexane) 2,2-bis(chlorophenyl)-l,l-dichloroethane l,l-dichloro-2,2-bis(chlorophenyl)ethylene l,l,l-trichloro-2,2-bis(chlorophenyl)ethane DDT equiv. = sum of all DDT-related peaks calculated as if all were DDT. polychlorinated biphenyls. Quantitation generally based on comparison to an Aroclor mixture. HCB - hexachlorobenzene ND = not detected. Mean values were taken from original citation where available; otherwise arithmetic mean was calculated, counting "ND" values as zero and "T" values as 0.5 times the lowest reported value. Missing values indicate no data in original article. Lowest value not reported. �The literature shows that mother's milk often contains seraivolatile chlorinated organic pollutants (pesticides). Presumably due to lack of analytical techniques and/or sensitivity, the presence of other pollutants has apparently not been investigated. 14 �SECTION 2 SUMMARY AND CONCLUSIONS The results show that sampling and analysis for organic compounds in mother's milk is feasible. The sample collection technique presented no significant problems. Analysis of the samples was generally satisfactory. The use of purge and trap with gas chromatography/mass spectrometry/computer (GC/MS/COMP) analysis for volatile organics was successful, although the intrusion of contaminants during analysis presented problems with some compounds. The wide range of volatile compounds found includes common air and water pollutants and possible metabolites. Thus, it may be possible to use mother's milk as an indicator of body burden if a correlation between exposure and mother's milk concentration is established. The extraction and GC/MS analysis for semivolatile organics was only marginally successful due to limited sensitivity (about 20-100 ppb milk). PCBs and ODE were the only halogenated semivolatiles found. The target semivolatile compounds (PCNs, PBBs, chlorinated phenols, and the higher chlorinated benzenes) were not present in quantities detectable by the survey techniques. The use of more sensitive (generally a factor of 100-1000) and selective methods [GC/electron capture detection (BCD), GC/negative ion chemical ionization mass spectrometry (NICIMS) or GC/single ion monitoring MS] may detect these compounds, but was outside the scope of this project. 15 �SECTION 3 RECOMMENDATIONS Further studies of the applicability of mother's milk as a matrix for assessing the human body burden of pollutants must directly compare human milk with the other available sample matrices. For example, comparison of the volatiles in breath, blood, urine, and mother's milk would determine which matrices are most suitable for measuring these compounds. It may also be advisable to use animal studies to determine the extent of environmental exposure-body burden correlation. In addition, the effects of transport of pollutants to a newborn infant should be studied. Infants may be uniquely affected by some pollutants due to their small body weight and different metabolism relative to adults. The measurement of semivolatile organics in mother's milk requires more sensitive techniques than those used in this study. For example, chlorinated compounds could best be detected using GC/ECD or GC/negative ion chemical ionization mass spectrometry and polynuclear aromatics by GC/photoionization detection. Improvement in analytical methodology could occur at several points: (1) As discussed above, more sensitive, analytical procedures could be used for specific compound classes. (2) For volatile organics, background levels could be reduced with an on-line purge and trap/GC system. Potential improvements in survey and sampling methodology include: (1) Addition of questions regarding length of nursing, age of infant, time since last nursing, etc. (2) Selection of participants according to a more statistically valid method (e.g. statistically random sampling). (3) Closer control over physical collection methodologies (e.g. all respondents gathered at one location). 16 �The 5-month time lag in the study awaiting OMB clearance was seriously detrimental to the project. The personnel and apparatus used for the validation studies had to be reassembled once OMB clearance was obtained. Restartf ing a project following a long dormant period requires retraining analytical personnel (or training new personnel if original personnel have been reassigned to other research projects), recalibration of instruments, and assembling the necessary laboratory apparatus and supplies, all of which consume government resources. Reducing this time lag is extremely important for execution of programs involving human testing. 17 �SECTION 4 SELECTION OF SAMPLING SITES Five urban areas were chosen as sampling sites. Each of these cities is a high-probability area for the presence of one or more of the chemicals of interest in mother's milk. Since many of the compounds of interest are probably specific to certain industrial sites, the samples from the other sites were intended to serve as controls for the site-specific compounds. Other compounds are considered ubiquitous and their levels in milk was probably not related to local industrial activity. The rationale for selecting the five sampling sites is discussed below. BRIDGEVILLE, PENNSYLVANIA PCNs are manufactured by Koppers Company, Inc., of Pittsburgh, PA, at the Koppers Chemical and Coatings plant in Bridgeville, about 10 km SW of Pittsburgh. (29) Reported production levels were 7 million Ib in 1956 and 5 million Ib in 1972, (29) indicating a potential long-term, relatively constant, exposure level in the surrounding area. Results from environmental monitoring in the area immediately (< 1 km) surrounding the plant indicated higher levels of PCNs in air and soil than those found near five PCN user sites, as f f\f\ O / *\ shown in Table 5. Furthermore, fish and apple samples from the same area were found to contain PCNs, indicating a potential link to the human food chain. In addition to PCNs, plants in the Bridgeville area have been reported to emit large quantities of phthalic anhydride particulate. (^5) At this plant site, Koppers is reported to manufacture chlorinated naphthalenes, phthalic anhydride, maleic anhydride, and alkyd resins. 18 �Table 5. SUMMARY OF PCN CONCENTRATIONS FOUND NEAR MANUFACTURING AND USE SITES Air, ng/m Water, pg/L Capacitor manufacturing A Low High Mean Upstream 1 25 450 150 0.2 120 2900 1400 a 1 NDb 7.3 3.1 2 PCN manufacturer (Koppers) Sampling Period 2 Site ND 3.9 1.2 1 9.8 31 19 2 Capacitor manufacturing B 9.8 33 17 o water samples collected for period 2. Not detected. Downstream Soil, ng/kg Low High Mean 1.4 130 940 ND ND ND ND 0.6 ND 2300 7.3 470 2.0 100 �NORTHERN NEW JERSEY - STATEN ISLAND, NEW YORK, AREA (NNJ) The Northern New Jersey (NNJ) area was selected as a sampling site on two bases: production of PBBs and general chemical industrial activity. Three facilities are of interest^ ' with respect to PBBs: White Chemical Co., E 22nd St., Bayonne, NJ; Marcor, Inc., Standard T. Chemical Co., subsidiary, 2500 Richmond Terrace, Staten Island, NY; and Hexcel Corp., Fine Organics Division, 880 Main St., Sayreville, NJ. White produced 45,000 kg of FBBs (specifically octabromobiphenyl and decabromobiphenyl) between 1970 /oo^ f39} and 1973. Hexcel is reported to have produced unspecified amounts of decabromobiphenyl [as well as to have produced or used decabromobiphenyl oxide, ethylene dichloride, and l,2-bis(2,A,6-tribromophenoxy)ethane]. Standard T is thought to have been a PBB user up to about 1974. (39) Results of environmental sampling in the area surrounding these three companies (40 '41) indicated the presence of PBBs, especially the more highly brominated homologs, in sediment, water, soil, human hair, fish, turtle, and plant matter. The findings in human hair oil (18 total samples), which ranged from undetectable to 310 ppm, are especially relevant to this study, since they indicate that the PBB manufacturing in this area and the resultant environmental contamination has resulted in human exposure. Northern New Jersey has a high concentration of chemical industries, (42) many of which use or produce halogenated hydrocarbons. The list of industries and locations are summarized below. Coastal Industries, Inc. (swimming pool chemicals), Diamond Shamrock (textile processing chemicals), Scientific Chemical Processing (chemical waste disposal) and Tenneco Chemicals (synthetic foam rubbers) are located in Carlstadt. Crompton & Knowles Corp. (dyes, colors and chemicals) are located in Fairlawn. Fisher Scientific (chemicals), Conoco Chemicals are in Saddle Brook. In Bayonne are CIBA-Geigy (dyes and intermediates) and ICI America (organics). In Jersey City are Mallinkrodt (analytical reagents) and Onya Chemical Co. (textile finish compounds, water repellants, germicides, and detergents). In Kearney are Standard Chlorine Chemical Co. (chlorobenzenes), Theobald Industries (bleaches), PPG Industries (paint) and Monsanto (industrial chemicals). In South Kearney is BASFWyandotte (dyestuffs and vinylidine chloride). In Newark are American Oil and Supply Co. (surfactants and chemicals), Celanese Plastics (plastics), 20 �DuPont (pigments), Inmoat (paint), Haas & Waldstein (paint), Otto B. May (dyes, surfactants), 3M (chemicals), Benjamin Moore (paint), Sherwin-Williams (paint) and Vulcan Materials (chlorome thanes ) In Elizabeth are Perk (chlori. nated solvents) and Speciality Chemicals Division of Allied Chemical Corp. Linden Chlorine Products (chlorine) is in Linden. In Rahway are M & T Chemicals (speciality chemicals) and Merck and Co. (industrial chemicals). In Edison are Gary Page Chemicals (PVC compounds) and Mobile Chemical (paint). In Parlin, Hercules manufactures chloroform. In Passaic are Pantasote Co. of New York (PVC resin film), Stauffer (vinyl sheet and film) and United Wool Piece Dyeing and Finishing (dyes). In Patterson are several dye manufacturers. In Wayne are American Cyanamid (chemicals) and Owens Illinois (plastics). Many of these and other firms in NNJ undoubtedly manufacture or use compounds which are of interest to this study. The levels of general organic pollutants in NNJ have been found to be high due to intense chemical manufacturing in the area. Environmental monitoring by RTI under separate contracts, " has found a wide variety of organic pollutants in this area. In addition, preliminary results from ground and surface water samples indicate measurable levels of a number of 45) volatile halogenated hydrocarbons. ' These data, summarized in Table 6, are indicative of environmental levels of organics in the NNJ area to which humans may be exposed and thus are indicative of the types of compounds (45) anticipated in mother's milk. Under a separate research project, the daily intake of some selected organics was roughly estimated. These estimates are given in Tables 7 and 8. Clearly there is ample exposure to pollutants which could potentially partition into milk. The statistics for cancer in two counties of NNJ are very high. ' The overall rate for all malignant neoplasms is significantly above the national average. This cancer incidence in New Jersey has been partially linked to the chemical and allied industries located there. Northern New Jersey is a metropolitan area with a relatively static population, a well-established chemical industry, known environmental levels of organics (including PBBs) and abnormally high cancer rates. These factors make this area especially suited to this study of organics in mother's milk. 21 �Table 6. PREVALGNT HALOGENATED COMPOUNDS IN AMBIENT AIR AND WATER OF RAIIWAY/WOODBRIDGE, BOUNDBROOK AND PASSAIC, Nj( 4 <0 Occurrence Medium Ubiquitous Air ro K> tetrachloroethylene trichloroethylene 1,1,1-trichloroethane 1,2-dichloroethane chloroform carbon tetrachloride p_,m,p_-dichlorobenzenes chlorobenzene Water dichlorobenzene trichloroethane chloroform trichloroethylene dichloroethane bromodichloroethane bromodichloromethane tetrachloroethylene d ibromoch1orome thane Mean Concentration 210,000 125,000 62,000 96,000 47,000 29,000 11,000 2,700 209 42 14 7 5 5 3.7 3.6 3.3 aConcentrations for air expressed in ng/m 3 and for Area Specific Mean Concentration a 1,1,2-trichloroethane vinyl chloride 1,2-dichJoroethylene 1,1,2,2-tetrachloroethane 9,000 1,200 1,000 750 chloronitrobenzene methyl trichlorophenoxy acetate methyl dichlorophenoxy acetate bromopropyIben zene bromobenzene tetrachloroethane dichloroethylene water in pg/L. 10.7 5 3.5 3 3 2.5 1.8 �Table 7. ESTIMATED DAILY INTAKE OF SELECTED VOLATILE COMPOUNDS AND EXPECTED CONCENTRATIONS IN BLOOD IN NORTHERN NEW JERSEY(45) Foodc (ng/day) Total (ng/day) Potential Blood Concentration^ (ppb) 3,600 4,150 2,108,000 88 1,250,000 7,000 18,660 1,276,000 53 1,1, 1- trichloroethane 620,000 42,000 5,290 667,000 28 1 , 2-dichloroethane 9000 6,0 5,000 965,000 40 chloroform 470,000 14,500 14,280 499,000 21 carbon tetrachloride 290,000 1,000 12,070 303,000 13 dichlorobenzene 110,000 2900 0,0 319,000 13 chlorobenzene 27,000 1,000 28,000 1.2 vinyl chloride 12,000 12.000 0.5 3,700 0.2 7,800 0.2 6,188,200 258.2 Aira (ng/day) Waterb (ng/day) tetrachloroethylene 2,100,000 tri chloroethy 1 ene Toxic Chemical 10 u> bromodichloromethane benzene total 3,700 7,500C 300f From Ref. 44, calculated on basis of 1 , 0 L/24 h respiration rate, 000 'prom Ref. 44, calculated on basis of 1 L/24 h intake. : From Ref. 47, calculated from FDA standard diet (Ref. 4 ) 8. d,. Expected blood concentration is total daily intake divided by blood volume ( . 0 mL) assuming 4 800 half-lives/day. 'From Ref. 49, 50. From Ref. 50. �Table 8. TOTAL DAILY INTAKE OF TARGET COMPOUNDS, PESTICIDES, PCBs, BaP AND METALS AND CONCENTRATIONS IN BLOOD IN NORTHERN NEW JERSEY(45> Air Toxic Chemicals a -BUG lindane heptachlor (ng/day) 10 60 30 heptachlor epoxide chlordane 0 0 7 Total (ng/day) 1,100 1,110 62 646 92 640 647 20 20 DHR to Food (ng/day) 586 Water (ng/day) Expected Blood Concentration (ppb) 0.14 0.08 0.01 0.08 *Q 3,500 DDT/ ODD 70 lO 0.44 2,500 2,570 0.32 73 50 0 3,500 123 0.02 •x.200 <60 388 648 0.08 440 <67 8,849 9,356 1.16 2 7,800 7,823 1.0 arsenic 21 2,800 < 1,000 31,300 34,100 cadmium 50 <1,000 32,000 33,000 4.4 a 7,500 3,200 105,000 115,700 PCBs Total Halogenated Compounds benzo(a)pyrene lead Ref. 56. 'Ref. 57. <io 100-500b �Table 8 (cont'd.) Sources: Pesticides and PCBs in air — Ref. Pesticides in water -- Ref. Pesticides and PCBs in food -- Ref. PCBs in water — Ref. BaP in air -- Ref. 10 51 44 48 51 52 (US) (NJ) (US) (US) (US) - Ref. 53 (World) •- Rough estimation (from Ref. S3 [World]) Metals in air -- Ref. 54 (NJ) Metals in water -- Ref. 55 (NJ) Metals in food -- Ref. 48 (N.E. NJ) BaP in water BaP in food �BATON ROUGE, LOUISIANA Baton Rouge was selected on the basis of extensive organic chemical production (especially volatile halogenated hydrocarbons) as summarized in Table 9. (43) In addition, RTI has collected and analyzed ambient air samples from this area and established the presence of a number of compounds of interest in ambient air. (43) A summary of the levels of halogenated compounds found in water and air is presented in Table 10. In addition to the industrial production in Baton Rouge, industries in Flaquemine (15 km SSW), St. Gabriel (20 km SSE) and Geismar (27 km SSE) may emit significant levels of chemicals which may contribute to the levels observed in mother's milk in Baton Rouge. These industries and their production are listed in Table ll.(36) KANAWHA VALLEY, WEST VIRGINIA Many manufacturers of organic chemicals are located in the Kanawha Valley, WV. DuPont, near Belle, W, has a large chemical complex for the synthesis of substances such as methylmethacrylate, methylamines, ammonia, hydrogen cyanide, herbicides, and insecticides. In South Charleston are production and consumption plants (Union Carbide, and FMC). Plastics, PVC, antifreeze, chlorine, halogenated organics, carbon disulfide, peroxides, etc., are the predominant chemicals produced here. The major industrial facility in the town of Institute is Union Carbide, which also processes a broad spectrum of compounds, e.g., viscose rayon and phthalate esters. There is also a large-scale olefin processing complex and a rubber accelerator plant. A major terminal loading facility in South Charleston bandies large quantities of a variety of organic compounds. Monsanto, FMC, Allied, and Fike have plants near Nitro for the production of antioxidants, rubber accelerators, industrial chemicals, and other materials. Several other chemical manufacturers, consumers, and transporters are located in the Kanawha Valley, some or all of which may contribute to the presence of organic materials in the ambient air or water and thus contribute to human exposure. Previous RTI sampling^3>46'65>66) in the Kanawha Valley found a broad range of halogenated, ketone, aldehyde, ester, aromatic, and aliphatic 3 compounds. Quantitative results included high values in air of 11,000 ng/m 26 �Table 9. POTENTIAL EMISSIONS FROM CHEMICAL INDUSTRY IN BATON ROUGE, LAa(43) Chemical chlorodifluoromethane (101) dichlorodifluoromethane (12) dichlorotetrafluoroethane (114) ethylene dichloride polyethylene resin trichlorofluoromethane (11) 1,1,2-trichloro-l,2,2-trifluoroethane (113) vinyl chloride ethyl chloride methyl chloride perchloroethylene tetraethyl lead 1,1,1-trichloroethane trichloroethylene PVC benzene butadiene n-butyl alcohol Total Production (nimlb/yr) NA 1100 460 Raw Material Company chloroform carbon tetrachloride perchloroe thy1ene ethylene ethylene Acer ACC ACC ACC, EC ACC ACC NA perchloroethylene ACC 480 ethylene dichloride ethylene methanol ethylene dichloride ethyl chloride 1,1-dichloroethane ethylene ACC, EC 210 75 100 312 40 32 144 440 428 EC EC EC EC EC EC EC petroleum ethane, etc. EXCC EXCC, CRCC EXCC NA (continued) �Table 9 (cont'd.) Chemical Total Production (mmlb/yr) Raw Material Company EXCC EXCC decanol diisodecylphthalate NA nonene NA phthalic anhydride, isodecanol dodccene ethylene isobutylene isodecanol p isooctyl alcohol isoprene isopropanol neopentanoic acid nonene phthalic anhydride propylene resin toluene ethylbenzene styrene vinyl toluene 100 propane/propylene EXCC 700 ethane, etc. EXCC NA petroleum nonene neptene ethylene by-product propylene isobutylene propane/propylene £-xylene ethylene petroleum benzene ethylbenzene toluene, ethylene EXCC £ to 00 NA NA 10 680 5.5 300 90 320 378 900 800 NA EXCC EXCC EXCC EXCC EXCC EXCC EXCC EXCC EXCC, FGC FGC FGC FGC Data provided by the Louisiana State Air Board. ACC = Allied Chemical Corp., EC = Ethyl Corp., EXCC = Exxon Chem. Corp., FGC = Foster-Grant Co. Inc. C Involves production of other alcohols also, C,, C0, C , C ln , C._, C.,. u NA = not available. o y iu j.o it) �Table 10. PREVALENT HALOGENATED COMPOUNDS OCCURRING IN AMBIENT AIR AND WATER OF BATON ROUGE, GEISMAR AND PLAQUEMINE, Occurrence Medium Air ro vo Water Ubiquitous chloroform 1,2-dichloroethane carbon tetrachloride 1,1,1-trichloroethane trichloroethylene tetrachloroethylene 1,1-dichloroethane trichloroethylene chloroform trichloroethane dichloroethane carbon tetrachloride dichlorobenzene ch1orod ibromomethane tetrachloroethylene Mean Concentration' 5,500 1,656 Sll 605 142 118 86 96 20 11 7.7 7.1 4.2 3.5 1.9 Area Specific 1,1,2-trichloroethane 1,2-dichloroethylene dichlorobutane 1,2-dichloropropane vinylidene chloride 1,1,2,2-tetrachloroethane bromobenzene 1,2-dichloroethylene hexachloroethane aConcentrations for air expressed in ng/m 3 and for water in pg/L. Mean Concentrationa 632 472 409 306 78 70 13 4 1.6 �Table 11. POTENTIAL EMISSIONS FROM CHEMICAL INDUSTRY IN PLAQUEMINE, GEISMAR, AND ST. GABRIEL, LA(36) City Annual Capacity (million pounds) Chemical Company Plaquemine chloroform 1 , 2-dichloropropane ethylene dichloride methyl chloride methylene chloride tetrachloroethylene vinyl chloride b 10 1325 150 190 150 450 Dow it it ii M ii ii Geismar chloroform ethylene dichloride methylene chloride tetrachloroethylene 1,1,1-trichloroethane phosgene phosgene vinyl chloride vinyl chloride 46 330 VCM 80 150 65 55 125 300 300 BASF RCC BOR MCJ phosgene NA SCC St. Gabriel Dow « Dow Chem. USA VMC ~ Vulcan Materials Co. BASF = BASF Wyandotte Corp. RCC «= Rubicon Chems., Inc. BOR « Borden, Inc. MCI « Monochem, Inc. SCC « Stauffer Chem Co., Agric. Chem. Div. 200 million pounds combined capacity in Plaquemine and Freeport, TX plants. 30 �for methylene chloride, 1500 ng/m 3 for tetrachloroethylene, and 72,000 ng/m 3 for benzene. Compounds identified in the air particulate fraction included .long-chain alkanes, polycyclic aromatic hydrocarbons (PAH) from naphthalene through anthanthrene (or an isomer), alkyl-PAH derivatives, and nitrogen-containing heterocycles. 31 �SECTION 5 SAMPLE COLLECTION At each of the five sites, arrangements were made to work through clinical facilities to recruit a suitable panel of respondents. These facilities included the Bayonne Hospital in Bayonne, NJ; the Medical Center Hospital in Jersey City, NJ; Magee-Women's Hospital in Pittsburgh, PA; Charleston Area Medical Center in Charleston, WV; and the East Baton Rouge Parish Health Clinic in Baton Rouge, LA. Advance arrangements were made through a contact person at each facility. This person was responsible for recruiting a professional member of the facility's staff to serve as the data collector. The data collector was usually & registered, licensed practical, or public health nurse associated with the facility. Respondents were paid $5 for their assistance in providing a milk sample and completing the survey questionnaire. The data collection effort is discussed in the following sections. OMB CLEARANCE Under the Federal Reports Act, clearance for the study of human subjects must be obtained from the Office of Management and Budget. This clearance was obtained on October 18, 1978. The OMB number is 158-578010. This study was approved with the understanding that: (1) the surveys were conducted as a pretest of the feasibility of information collection procedures; (2) the information collected will not be used to generalize to either local areas or the nation as a whole. These two caveats were invoked since the sample size was small and a nonprobability sampling method (subject selection) was used. 32 �TRAINING Before data collection began at a site, a training session was held to acquaint the facility contact person and data collector(s) with the survey. The session addressed the study objectives; use of the data collection instruments; administrative instructions; quality control procedures; and instructions for collecting, packing, and shipping milk samples to RTI. The training was conducted by an RTI survey specialist from the Survey Operations Center. A detailed manual and necessary field reporting forms were developed for use in these sessions. All training was conducted at the participating facility and lasted approximately 4 hours. SURVEY INSTRUMENTS Three data collection instruments (see Appendix A) were developed for use by the data collectors. The Participant Consent Form (FCF) was used to introduce the study, explain the study objectives and requirements of participation, present the confidentiality procedures, and obtain consent of participant. This form was signed by the respondent, who retained a copy for her files. The original was attached to the data collection instrument and a second copy was filed in the respondent's hospital record. The Participant Listing Form (PLF) provided a means of assigning unique numbers to participants at each performance site. The data collector completed this form as each participant was solicited; the form was returned to RTI with the completed questionnaires when work at the site was finished. The Study Questionnaire (SQ) was the primary data collection instrument. Information concerning participant demographic characteristics, residence information, health data, use of medications, and personal characteristics was obtained through this document. The SQ was administered after patients had been screened and prior to collection of the milk sample. PARTICIPANT SCREENING Potential participants (lactating women) were screened by the data collector to determine whether or not they met certain study criteria, which included: 33 �ability and desire to provide a milk sample of approximately 100 ml. permanent residence within the area of interest for at least the preceding 12 months, and no travel outside the area of interest for the seven days preceding sample collection. After potential participants were screened, 10 women who met all the criteria for participation were asked to provide a milk sample and complete the SQ. PLF, PCF, AND SQ COMPLETION PROCEDURES When an eligible person agreed to participate, her name was listed on the PLF anci she was assigned a unique participant number. The data collector then read the information contained on the PCF to the participant while she followed along using a second copy. After answering questions or handling problems, the data collector asked the participant to sign the PCF prior to administration of the SQ. The data collector then completed the SQ by asking the questions directly to the participant. Completion time averaged 15 minutes. An adhesive, computer-generated ID label was affixed to the SQ; a duplicate label was provided to be used for identifying the milk sample bottle. Each participant was a self-respondent unless she was under 18 years of age, in which case the SQ could have been administered in whole or part to the parent or guardian, but in the participant's presence. SAMPLE COLLECTION PROCEDURES After completion of the SQ, the data collector made the necessary arrangements for the participant to provide the milk sample. A collection bottle was taken from the shipping box and the adhesive ID label was affixed to the bottle. The milk was manually expressed directly into the bottle; no breast pumps or other devices were allowed. Immediately after the milk was collected, the bottle was capped and the sample frozen until all ten samples were collected and ready for shipment to RTI. A minimum of 60 ml (half-full bottle) was required for each sample. If insufficient milk was collected, the sample was discarded and an additional subject was added to the study. 34 �SHIPPING PROCEDURES Sample bottles were packed in the shipping container, cooled with dry ice, and sent directly to RTI via Federal Express. 35 �SECTION 6 SAMPLE ANALYSIS METHODS The milk samples were analyzed using gas chromatography/mass spectrometry/cornpute:r. Due to the broad range of volatilities, the samples were partitioned into two general classes of compounds: volatiles (e.g. benzene, chloroform) and semivolatiles (e.g. PCNs, PCBs, pesticides). The analytical protocols developed for the volatile and semivolatile components in mother's milk are reproduced in Appendices B and C, respectively. The experiments conducted which led to these protocols are discussed below. DEVELOPMENT OF ANALYTICAL PROTOCOL FOR VOLATILES The headspace purge technique was validated by determining the recovery of four model compounds from raw cow's milk samples. Compounds labeled with carbon-14 were chosen in order to examine both the amounts recovered on Tenax GC and the amounts remaining in purged samples. Twelve 50 mL cow's milk samples were spiked with methanol solutions of 14 the C-compounds. The analysis for each of the four model compounds was performed in triplicate. In addition, standards were prepared in triplicate by adding the appropriate amount of each compound in solution to a scintillation-counting vial containing 15 mL of Triton X/toluene/Omnifluor scintillation "cocktail." Milk samples were purged as described in Appendix B; Tenax cartridges were stored, and aliquots of the purged samples were retained for oxidation and counting. Tenax cartridges were desorbed at 270°C and 30 mL/min N. for 10 minutes into 15 ml, of Triton X cocktail in tandem scintillation vials. The vials were capped and refrigerated until scintillation counting. An aliquot (1 mL) of each purged milk sample was oxidized in the Packard Tricarb Sample Oxidizer, which converted all carbon-containing compounds to carbon dioxide and water. The 14C-carbon dioxide was collected in a trapping solution and 36 �referenced to a quench correction curve. All standards, Tenax samples and oxidized milk samples were counted on a Packard Liquid Scintillation Counter with automatic standardization. Counting data was analyzed by computer to obtain the number of disintegrations per minute (dpm) for each vial. The percent recovery was calculated for each milk sample as shown below: 1 rocov-rv - dpm in first vial + dpm in second vial * ^ ~ average dpm added to triplicate standards .~ * The second of the tandem scintillation vials contained <2 percent of the radioactivity in every case. The amounts of 14C compounds retained in the purged sample was calculated: v retained = dpm in oxidized, purged sample ,QQ« * average dpm added to triplicate standards * x * The data are tabulated in Table 12. The recoveries for the volatile chloroform and carbon tetrachloride were about 90 percent, as expected. The less-volatile chlorobenzene and bromobenzene exhibited correspondingly poorer recoveries. These compounds are generally considered only marginally purgeable from water, so these results from milk are not surprising. The methodology validation experiment indicated that the proposed method of analyzing human milk for volatile organic compounds was adequate. Sensitivity and detection limits were determined by the capabilities of the GC/MS/COMP system. DEVELOPMENT OF ANALYTICAL PROTOCOL FOR SEMIVOLATILES The extraction and cleanup method was validated using six model compounds (2,4-dichlorophenol, pentachlorobenzene, 1,2,3,4-tetrachloronaphthalene, 4,4'-dibi:omobiphenyl, 2,2',5,5'-tetrabromobiphenyl, and octachloronaphthalene) which were representative of the semivolatile (nonpurgeable) compounds of interest. The compounds were spiked into raw cow's milk at a level of about 1 pg/mL. Raw cow's milk was chosen as the closest readily available analog to mother's milk. The results are presented in Table 13. The overall mean recovery was about 70 percent and the mean of the relative standard deviations was 22 37 �Table 12. METHOD VALIDATION RECOVERY OF SELECTED VOLATILE STANDARDS FROM MILK Compound 14 a C- chloroform 14 C-carbon tetrachloride 14 C- ch lorobenzene 14 C-bromobenzene CO Percent . Recovered Percent. Retained Percent Accounted for 62 88 + 5 6 + 0.3 94 12 76 88 + 6 3+ 3 91 ±3 132 63 + 2 26 + 3 89 +_ 1 156 35 + 3 51 + 13 86 1 10 80,000-94,000 dpm added to each sample. Mean +_ standard deviation of three replicates. •* "Sum of percent recovered and percent retained. 3 00 b.p. ( C °) �Table 13. METHOD VALIDATION RECOVERY OF SEMIVOLATILE COMPOUNDS SPIKED INTO RAW COW'S MILK Compound Concentration in Milk mp ( C bp (°C] °) 1 (ng/mL) Mean Standard Recovery Deviation Relative Standard , Deviation 2,4-Dichlorophenol 45 207 1.12 59 12 20 Pentachlorobenzene 85 277 1.24 76 19 24 1.37 59 15 25 1.04 58 19 33 0.93 94C 10 11 1.08 78C 14 17 1,2,3,4-Tetrachloronaphthalene 4,4'-Dibromobiphenyl ig_ 164 357 2 , 2 ' , 5 , 5 ' -Tetrabromobipheny 1 vo Octachloronaphthalene 198 441 Seven replicates. 'standard deviation divided by mean multiplied by 100. replicates. �percent. These results indicated that refinements in the method should be considered prior to a large-scale study. Two methods were available for removing fat and other nonvolatile components of the milk extract: Florisil column chromatography and gel permeation chromatography (GPC). Evaluation of the two techniques indicated that the Florisil method was more suitable to this project. The Florisil method was faster and had greater sample capacity than the GPC. In addition, the GPC procedure required the use of a pumping system, UV detector, and expensive, fragile GPC columns. Initial tests with both methods revealed interference problems, although those with GPC were more severe. Using GPC, decabromobiphenyl and hexabromobiphenyl eluted with the fat peak. This was judged totally unsatisfactory. Using Florisil, some fat eluted in the fraction with the compounds of interest, but repetition of the procedure yielded samples sufficiently clean for analysis. DEPARTURES :FROM THE ANALYTICAL PROTOCOLS Emulsions The formation of an emulsion during the toluene-acetone extraction of semivolatiles (step 6, Appendix C) was an area of concern. Approximately 80 percent of the time an emulsion occurred. To eliminate this, three approaches were taken with reasonable success. The first was to avoid the emulsion formation by swirling rather than shaking the toluene and acetone extracts. The second approach was to break the emulsion by adding Na.SO, and waiting. Both the amounts of Na-SO, and the time required varied. In severe cases emulsions were broken by filtering through glass wool wetted with toluene. Lipid Removal Using Florisil Problems were also encountered during the Florisil cleanup. Some samples had a tendency to solidify while concentrating the ether/pentane eluate, apparently due to abnormally high fat content. This usually occurred when the sample volume reached 1-3 mL. The samples to which this happened were diluted with pentane and eluted through another Florisil column. The Florisil cleanup was repeated until the samples remained liquid at small (<1.0 mL) volumes. Three cleanups was the maximum required for any sample. 40 �GC/MS ANALYSIS PROCEDURES Samples were analyzed by gas chromatography/mass spectrometry using an LKB 2091 EI/CI GC/MS. Operating conditions for the analysis of purgeables is given in Table 14 and the operating conditions for the extractables is given in Table 15. Analysis of the purgeables involved the use of the desorption apparatus described in Appendix B. Quantitation of the unknowns was accomplished using relative molar responses (RMRs) as discussed in Appendices B and C. The RMRs were calculated from replicate determinations of known amounts of standards and analytes. Qualitative Analysis Initial identification of compounds by GC/MS involved comparisons of unknown spectra with data compiled in the Eight Peak Index of Mass Spectra^ . If the peaks present in the unknown spectra clearly matched the peaks of the standard compound in the tables and the intensities were about the same, then a positive identification was usually made. If peak intensities of unknowns varied from those of the standards, and there were isomers of the compounds that were not listed in the Eight Peak Index, then the compound was listed as an "isomer." Whea the background peaks interfered with the spectrum of an unknown to an extent that made identification uncertain, the compound identification was labeled as "tentative" (tent.). If no standard spectra similar to those of the unknowns appeared in the mass spectral references, but fragments characteristic of a certain class of compounds were identified, tentative identifications were made on the basis of the characteristic fragments and apparent molecular weights. These identifications were also labeled "tent". Usually tentative identifications involved alkyl derivatives or homologs of classes of compounds that were positively identified in the same sample. Positive identifications, as well as some tentative identifications, often required more detailed investigations of standard spectra in the Registry of Mass Spectral Data or standard spectra found in other literature such as scientific journals. The Registry of Mass Spectral Data presents data in the form of histograms rather than as a list of peaks and their intensities. This type of format allowed more subtle differences in mass.spectra to be considered when several similar standard spectra in the �Table 14. OPERATING CONDITIONS FOR GC/MS ANALYSIS OF PURGEABLES Instrument Coluxnn Flow LKB 2091 80m - SE-30 WCOT Capillary Column Desorption Temperature Desorption Time 270°C 8 min Desorption Flow Column Temperature 15 mL/m"in He 30°C for 2. min programmed to 240°C at 4cC/min Scan Range Scan Speed Scar Cycle 5 •* 490 Dalton 0 •*• 670 in 2 sec 1.7 mL/imtn He 1.7 sec Injector Temperature Accelerating Voltage Ionizing Energy 250°C Trap Current 50 yA Source Temperature 210°C 3500 V 70 eV 42 �Table 15. OPERATING CONDITIONS FOR THE GC/MS ANALYSIS OF SEMIVOLATILES Instrument GC Column LKB 2091 25m SE-52 WCOT capillary column Flow Column Temperature 1.5 mL/min with 15:1 split 80°C for 3 min then 8°C/min to 265°C Scan Range Scan Speed 5 •»• 530 Dalton 2 sec 0 -*• 670 Dalton Scan Cycle 2.4 sec Injector Temperature 240°C Accelerating Voltage Ionizing Energy Trap Current 3500 V 70 eV 50 UA Source Temperature 210"C �Eight Peak Xndex appeared to represent possible candidates for unknown identifications. A large number of sample components remained unidentified. These unidentified components were labeled "unknown." In order to quantify the degree of certainty with which a compound has been identified, a "level" heirarchy has been established. The compound identification criteria are listed below: Level I Computer Interpretation. The raw data generated from the analysis of samples are subjected to computerized deconvolution/library search. Compounds identified using this approach have the lowest level of confidence. In general Level I is reserved for only those cases where compound verification is the primary intent of the qualitative analysis. Level II Manual Interpretation. The plotted mass spectra are manually interpreted and compared to those spectra compiled in a data compendium by a skilled interpreter. In general a minimum of five masses and intensities (±5 percent) should match between the unknown and the library spectrum. This level does not utilize any further information such as retention time since the authentic compound may not be available for establishing retention times. Level III Manual Interpretation Plus Retention Time/Boiling Point of Compound. In addition to the effort described under Level II, the retention time of the compound is compared to the retention time that has been derived from previous chromatographic analysis. Also the boiling point of the identified component is compared to the boiling points of other compounds in the near vicinity of the one in question when a capillary coated with a nonpolar phase has been used. Level IV Manual Interpretation Plus Retention Time of Authentic Compounds. Under this Level, the authentic compound has been chromatographed on the same capillary column using identical operating conditions and the mass spectrum of the authentic compound is compared to that of the unknown. Level V Level TV Plus Independent Confirmation Techniques. This Level utilizes other physical methods of analysis such as GC/Fourier transform infrared spectrometry, GC/high resolution mass spectrometry, or nmr analysis. This Level constitutes the highest degree of confidence in the identification of organic compounds. Unless otherwise stated, all identifications in this report were Level II. 44 �SECTION 7 RESULTS VOLATILES All 42 of the purged samples were analyzed by thermal desorption/GC/MS. The mass spectra from selected samples were interpreted manually to determine which compounds should be quantitated. From these data, selected compounds were quantitated in all samples. All data were stored on magnetic tape for subsequent processing and are routinely archived for at least 5 years. Qualitative Identifications Eight samples were interpreted. The results are presented in Appendix D. Samples were selected according to the following criteria. At least two samples were required from each collection site (Jersey City and Bayonne, NJ, were counted as two separate sites). The total ion current chromatograms were inspected and the samples with the greatest number of peaks or those containing very intense unique peaks (not observed in other samples) were selected. For those samples selected, all of the mass spectra were printed and interpreted manually by experienced spectroscopists. Table 16 summarizes the compounds found and their frequency of occurrence. It is interesting to note that some compounds (e.£. 1,1,1-trichloroethane and hydrocarbons.) are common air pollutants, others (e.g.., dibromochloromethane) are common water pollutants, others (dimethyldisulfide, furans, aldehydes) appear to be metabolites, others (chlorofluorocarbons, siloxanes) are known background interferents, and others (iodopentane) are of unknown source. Quantisation Based upon the qualitative identifications summarized above, nine coopounds were selected for quantitation in all of the samples. The results for four compounds are summarized in Table 17. As discussed below, the 45 �Table 16. SUMMARY OF QUALITATIVE IDENTIFICATIONS OF VOLATILE COMPOUNDS IN MOTHER'S MILK Sample Number Compound 1081 1040 1107 1115 2048 2071 3053 3111 Halogenated Compounds _ _ 4 - . _ chlorodif;.uoromethane chlorotrlfluoromethane dichlorodifluorome thane chlorome thane chloroethane trichlorofluoromethane dichloroei:hylene Freon 113 methylene chloride chloroform 1,1, 1-trichloroethane carbon tetrachloride trichloroethylene chloropentane dibromochloromethane tetrachloroethylene dichloropropene chlorobenzene chlorohexzine iodopentarie 3-methyl-3.-iodobutane chloroethylbenzene dibromodichloromethane dichlorobenzene chlorodecane trichlorobenzene 4 — 4 4 4 — _ 4 . - 4 . 4 - - - - 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 - 4 4 4 4 4 4 4 4 - . 4 4 - 4 - - 4 4 4 4 4 - 4 4 4 4 4 4 4 4 4 4 4 4 . . . 4 4 4 4 4 - 4 4 - 4 4 . 4 - 4 4 4 4 4 4 - 4 4 4 4 - - - - - 4 4 4 4 4 - - - 4 - - 4 4 4 . 4 4 4 - 4 4 4 - - - - 4 4 4 4 4 . 4 . 4 . - . . . . . - - 4 4 - - - Aldehydes _ acetaldehyde methylpropanal n-butanal methylbutanal crotonaldehyde nj-pentanal £-hexanal furaldehyde n-heptanal benzaldehyde n-octanal phenyl acetaldehyde 4 4 4 4 - 4 4 - 4 - 4 - - - . 4 - - - - 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 - 4 4 4 4 - 4 - . - 4 4 4 - 46 4 4 4 - 4 4 4 4 - 4 4 4 - 4 - 4 4 4 . - 4 - - - - (continued) �Table 16 (cont'd.) Sample Number Compou-id n_-nonanal methyl furaldehyde n-decanal n_-undecanal iv-dodecanal 1081 1040 1107 1115 + + . _ . . _ _ _ _ _ + . . . . . - + + + + + . _ _ . + _ + . + + + + + . _ + + + + + . . 2071 3053 + . + + . + . _ 2048 3111 + + . _. _ _ . . . * + . . + . + + + - + + . + . . _ . . + _ . , _ . . . . . _ . . + . _ . _ . + . . _ Ketones acetone methyl ethyl ketone methyl isopropyl ketone methyl vinyl ketone ethyl vinyl ketone 2-pent£inone methyl pentanone methyl hydrofuranone 2-methyl-3-hexanone 4-heptanone 3-heptanone 2-heptanone methyl heptanone furyl methyl ketone octanone acetophenone 2-nonanone 2-decanone alkylated lactone phthalide + + . + + _ . . _ + + . + + + _ . . + H+ > . + + + + + _ + . _ . . . + . + + + - . . _ _ . + + _ + _ _ . _ . + - h + . _ _ _ _ + + + . + + _ _ _ _ - Other Oxygenated Isomers C 5 H 10 0 C6H80 C 6 H 10 0 Ci,H602 C 6 H 12 0 C 7 H 12 0 C 7 H 10 0 C6H602 C^mOj C8H150 (continued) �Table 16 (cont'd.) Sample Number Compound 1081 1040 1107 1115 2048 2071 3053 3111 Other Oxygenaced Isomers (continued) C10H120 C10HU0 . C10H160 - C1£)H180 - C 10K20° C10H220 . - _ + _ + _ + - C H 9 8°2 CUH200 . . . . . . . . . . . . C . . . . . . + 10E10°2 Alcohols methanol isopropanol 2-methyl-2-'?ropanol n.-propanol 1-butanol 1-pentanol o-furfuryl alcohol 2-ethyl-l-hexanol phenol 2,2,4-triiaethylpentyl1,3-diol a-terpineol _ + _ _ _ _ _ _ _ + _ _ + . _ . _ _ + _ _ + _ _ + _ . + _ _ + + + _ _ + _ + _ _ + + _ + _ _ + + + + + _ _ + _ + _ _ + _ _ _ , • , . • • . . . . _ _ . _ + - - . _ _ . _ _ _ + _ _ + _ _ + _ . . _ + _ _ _ + + _ _ -f + . _ + + + _ + _ _ . + + _ + + _ • Acids acetic acid decanoic acid Sulfur Compounds sulfur dioxide carbon disuLfide dimethyl disulfide carbonyl sulfide • . (continued) 48 �Table 16 (cont'd.) Sample Number Compound 1081 1040 1107 1115 2048 2071 3053 3111 Nitrogen Compounds nitrome thane C5H6N2 C5H8N2 _ + . . . . . . C.H.N-0 442 methyl acetamide benzonitrile methyl cinnoline _ _ - + _ _ - _ _ + + - . _ • * • _ + . _ + _ _ _ - . _ - + + _ _ . . . . . . . . - _ + - _ _ _ - _ _ Esters vinyl propionate ethyl acetate ethyl-n-caproate methyl caprylate ethyl caprylate isoamyl formate methyl decanoate ethyl decanoate _ _ . _ . - _ + _ _ + _ _ _ _ . _ _ . _ _ _ . f _ _ _ _ _ - - + + + + + + - _ . _ , _ - _ _ - - Ethers dimethyl ether £-dioxane dihydropyran _ + . . . _ . + + _ _ + . - + - - _ _ _ _ _ _ - _ _ + _ + _ + + _ + + _ + _ - _ - _ _ Epoxide l,8-cin«ole . _ _ _ . _ _ _ . . + _ . _ . . _ _ _ . + + + _ _ _ _ _ _ - _ + _ _ + _ _ _ _ + _ Furans furan tetrahydrofuran methyl furan methyl tetrahydrofuran ethylfuran dimethylfuran 2-vinylfuran furaldehyde 2-n-butylfuran 2-pentylfuran methylifuraldehyde furyl xaethyl ketone a-furfuryl alcohol benzofuran ,_ . + _ + + 4 _ . _ 4 _ _ . . + + + + _ _ _ + _ _ - _ _ . _ + _ . + _ _ _ _ - - (continued) �Table 16 (cont'd.) Sample Number Compound 1081 1040 1107 1115 2048 2071 3053 3111 Alkanes C3H8 C4H1Q + C5H12 CfiH14 + C7E16 + C8H18 + C9H2Q C10H22 + + CUH24 C12H26 + + r u L H _ C . . . 13 28 14H30 C15K32 + 4 - * _ + _ J + + + . _ 4+ . . . + _ . Alkenes C3H6 C H 48 C5H1Q + . . . . + + CgH18 C1QH20 + _ + + + T J C 11H22 12K24 + . . . + . . + + + + + - - - + _ - 4 + C - C - 13K26 isoprene . + CfiH12 CyHu CgH16 p . J + _ + L + + _ + + l _ + + _ f _ _ + L _ * L * . + . -f _ „ . . _ . . - . . _ Alkynes C5Hg C6H1Q . . . . . + + . + ... C H 7 12 50 (continued) �Table 16 (cont'd.) Sample Number Compound 1081 1040 1107 1115 2048 2071 3053 3111 Alkynes (continued) C8H14 C9Hlfi + + - - + + . . H- C - . + - + + . . + + + + + _ + + + + + _ - _ . . + + + + _ + + + - + _ + _ _ . + + _ + + _ _ • - + + + + + _ _ + + + + + + + + + + + + + 4 + + + _ + + _ + + - + + + + + + + + + + - + + •)• + + + •<• + + * + -«+ + + + + . + + . » . + + + - + + + + + + + + + + 10H18 C H 12 22 . + . . . . Cyclic Hydrocarbons cyclopentane methylcyclopentane cyclohexane ethylmethylcyclohexane C 10H14is"mers C-0H.,isomers (other) limonene methyldecalin a-pinene canphene camphor + + _ . + + -_ _ _ + _ _ _ Aromatics benzene toluene ethylbenzene xylene phenylacetylene styrene benzaldehyde C,-alkylbenzene C4-alkylbenzene methylstyrene dimethylstyrene C5-alkylbenzene naphthalene Cg-alkylbenzene isomers isomers isomers isomers _ . + - + + + + + + + + . + + + + + ' + + . . _ _ + - Arranged by class in approximate elution order. See Appendix D for sampleby-saaple identifications. + - present; - - not identified in sample. b Participant code number. 51 �Table 17. VOLATILES QUANTITATED IN MOTHER'S MILK SAMPLES (ng/ml.) Site Bayonne, NJ Sample Number' 1016 1032 1040 1057 1073 Chloroform1' d Tetrachloroethylene Chlorobenzene Dichlorobenzenec 0.2 0.1 1081 0.3 0.1 0.7 0.7 1.3 1.5 1.5 1.1 0.9 3.8 6.3 0.1 0.1 0.1 0.1 6.7 9.1 66 0.2 2.2 32 Jersey City, NJ 1024 1107 1115 1123 1164 13 17 1.7 20 65 43 7.4 8.1 17 4.0 0.1 0.2 0.3 0.1 0.1 2.8 68 49 2.2 0.9 Pittsburgh, PA 2014 2022 2048 2055 2063 2071 2089 2097 2105 2113 2121 2139 0.9 1.5 0.6 0.8 0.6 1.2 0.7 6.7 2.8 1.2 0.8 0.6 0.8 1.8 1.8 1.0 1.6 1.0 26 1.8 1.3 0.7 2.4 0.7 0.2 0.1 0.1 0.05 0.1 0.1 0.2 0.4 0.1 TRe 0.1 0.2 1.1 8.9 0.7 3.1 1.4 0.5 0.3 1.1 0.4 2.0 ' 0.9 Baton Rouge, LA 3012 3020 3038 3046 2.9 0.7 0.8 21 0.1 0.5 1.7 2.5 0.3 0.1 0.2 0.1 4.2 0.6 1.3 2.2 tn (continued) �Table 17 (cont'd) Charleston, WVg OJ £* Sample Number3 Chloroform Tetrachloroethylene 3053 3079 3087 3095 3103 3111 Site 0.3 0.8 0.7 1.3 0.6 1.8 0.4 0.6 0.4 1.0 0.2 0.5 4010 4028 4036 4051 4069 4085 4093 4101 4119 5.0 7.2 1.2 1.4 3.9 0.6 0.4 0.4 1.0 1.0 >19f 7.5 8.2 5.3 12 8.7 11 Participant code number. See text for caveats with respect to chloroform. All isomers summed. Not detected. Trace. Instrument saturated. p Sample 4044 lost due to instrumental malfunction. CMorobenzene 0.2 0.1 0.2 0.3 0.1 0.1 0.2 10 0.2 0.1 0.1 0.1 0.04 Dichlorobenzenec 1.8 0.2 5.2 4.2 >22 44 0.7 1.9 0.2 1.1 3.6 3.8 0.04 26 1.4 �quantitation of the other five compounds is not reported, since the levels in milk were not judged sufficiently greater than background to be reliable. Upon inspection, it is obvious that most values are low relative to only a few high "outliers." These high values suggest that there is a range of levels of these compounds which may correlate with exposure. These results were analyzed statistically to determine if any of the values correlated significantly. As can be seen in Table 18, the arithmetic mean and median values generally are quite different. The arithmetic mean is skewed toward the high end, generally due to one or two relatively high values. mean. A more realistic representation of the central data is the geometric These geometric mean values were tested for their significance (^.6., are the geometric means significantly different from site to site?). Table 19 summarizes this data, prom this table, it appears that samples from Jerse}' City have significantly higher levels of chloroform, tetrachloroethylene, and dichlorcbenzene than the other study samples. Charleston samples appear to have significantly higher levels of chloroform, and Bayonne samples appear to have significantly higher levels of dichlorobenzene. To test if any of the compound levels were related, the Spearman correlation coefficients (nonparametric correlation based on the sample, designed to lessen the weight of a single high outlier) were determined as shown in Table 20. There does not appear to be any compound-to-compound correlation among the subjects. In interpreting these data, it must be remembered that this is a very small data set. Therefore these data should not be used to extrapolate to the city or area from which the samples were collected. Quality Control Table 21 presents the quality control results for chloroform, tetrachloroethylene, chlorobenzene, and dichlorobenzene. The very high recovery of chloroform from the controls indicates either a miscalculation of the amount actually spiked or contamination of the samples used as controls. Since the procedural blanks contained about 15 times less chloroform, the former explanation is most reasonable. However, the chloroform values reported in Table 17 must be interpreted subject to the following 54 �Table 18. SUMMARY STATISTICS FOR VOLATILE COMPOUNDS BY SITE Site Bayonne, NJ Maximum Meanb Median S.D. n Jersey City, NJ Maximum Meanb Median S.D. n Pittsburgh, PA Maximum Meanb Median S.D. n Baton Rouge, LA Maximum Msanb Median S.D. n Charleston, WV Maximum Meanb Median S.D. n Overall Maximum Mean Median S.D. n Tetrachloroethylene Chloroform 1.3 0.52 0.5 0.48 6.3 2.52 1.5 6 6 2.13 Chlorobenzene 0.2 0.12 0.004 0.1 6 66 19.37 7.9 68 24.48 31.69 8.1 0.3 0.16 0.1 24.3 15.9 0.089 5 5 65 23.34 43 15.9 17 6.7 1.53 0.85 1.74 26 3.41 1.45 7.13 12 12 21 3.09 2.5 0.79 0.8 0.5 6.34 0.75 10 Dichlorobenzene 25.54 6 2.8 5 5 0.4 0.12 8.9 1.71 0.1 1 0.11 2.41 12 0.3 0.16 0.15 0.096 12 44 8 3.2 13.98 10 10 >19 3.21 10 1.20 26 4.30 10 12 7.21 7.5 1 0.1 1.4 3.55 6.02 8.25 9 9 3.30 9 9 65 5.57 1.25 10.9 43 4.10 1.25 8.15 10 0.37 0.1 1.53 68 9.15 1.95 17.3 42 42 42 42 Tlaximum, mean and median values are ng/mL. Arithmetic mean. 55 �Table 19. SIGNIFICANCE OF THF. DIFFERENCES IN THE GEOMFTRIC MEANS BY SITE Geometric Mean (ng/mF,) Ch loroform Site 0.45 Nayonne 14.7 Jersey City Tetrach loroethylcno Chlorobcnzene Di ch lorobcn zene 2.09 0.12 8.33 11.5 0.16 8.55 Pittsburgh 1.23 1.82 0.12 1.21 Baton Rouge 1.53 0.67 0.15 3.83 Charleston 5.92 1 . 65 0.42 1.98 0.01 0.01 N.S.b 0.05 rt Significance g 0.01 implies 99 percent confidence that the numbers are statistically different, while 0.05 implies 95 percent confidence. Not significant. �Table 20. SPEARMAN CORRELATION COEFFICIENTS FOR VOLATILE ORGANICS FOUND IN MOTHER'S MILK .. , , . Chloroform Chloroform Tetrachloro. . ethylene „. , , Chlorobenzene Dichloro, benzene 1.0 0.37a -0.02b -0.13b 1.0 Chlorober.zene Dichlorobenzene 0.007b 0.05b 1.0 Tetrachloroethylene 0.03b 1.0 Significant at 0.05 level (95 percent confidence). b Not significant Sample size - 42 57 �Table 21. Type of Sample Chloroform QUALITY CONTROL RESULTS FOR VOLATILES IN MILK Tetrachloroethylene Chlorobenzene Dichlorobenzene 17 0.12 0.19 159 Blanksa Mean (ng/mL) b S.D. RSD (Z) 7 1.2 1.3 108 7 0.22 0.11 49 7 0.03 0.025 84 14.02 8.20 58 8 1.12 0.41 37 8 0.62 0.34 55 Controls 0 Mean Recovery® Vn oo S.D. RSD (Z) Blanks consisted of two field water blanks and five water blanks purged with the milk samples to monitor procedural background. No difference between the two types of blanks was observed. Arithmetic mean. Controls consisted of two spiked raw cow's milk samples carried to the field and returned, two spiked raw cow's milk samples stored in the laboratory, two spiked water samples carried to the field and returned, and two spiked water samples stored in the laboratory. No major differences were observed between the four types of samples. Samples were spiked at 30-90 ng/volume purged (or about 1 ng/mL). Not included in control spiking solution. 1.0 • 100 percent recovery. Extremely high recovery probably a result of improper loading of controls. �considerations: the mean reported levels in the samples were only 4.9 times the blank levels; the recovery from controls was about 1400 percent, invalidating the recovery study; and chloroform is known to be a laboratory atmospheric contaminant. The compounds presented in Table 17 represented significant levels above the background in blanks. Several other compounds were quantitated that did not exhibit substantial concentrations. These compounds, with the ratio of the mean in the samples to the mean in the background given in parenthesis, were: 1,1,1-trichloroethane (1:1), benzene (2:1), toluene (2:4), trichloroethylene (1:2) and carbon tetrachloride ( : ) These levels 14. in the samples cannot be reliably assigned to either the milk sample or to laboratory contamination. If these compounds are present in milk, they are very low and cannot be regarded as significant, given the limitations of the technique employed. Apparently, mother's milk does not represent a bioconcentration matrix for these compounds. SEMIVOLATILES Three samples were fully interpreted, as presented in Appendix E. As can be seen from the data, few compounds of interest were observed in the mass spectra. The data were searched on the GC/MS data system for target compounds (PCNs, PBBs and PCBs) using single ion plots called up from the full data set. No evidence for any of these compounds was observed at a detection limit of about 20 ppb. DDE was quantitated in five samples as shown in Table 22. These values were in the range generally reported by previous investigators (see Tables 2-4). Since none of the target compounds were present in detectable quantities, no further identification or quantitation was attempted. 59 �Table 22. DDE AND TETRACHLOROBIPHENYL LEVELS IN SELECTED MOTHER'S MILK SAMPLES Sice Sample Number DDE ng/mL Milk Tetrachlorobiphenyl Pittsburgh 2105 45 NDb Pittsburgh 2121 73 Tc Charleston, WV 4069 107 ND Charleston, WV 4085 38 ND Charleston, WV 4093 d Mean 91 ND 71 S.D. 29 RSD (2) 42 Median 73 Samples selected as having the most intense total ion current chromatograms. Not detected. Trace. Arithmetic mean. 60 �REFERENCES 1. Ziegel, E. and C. C. Van Blarcom, Obstetric Nursing, 6th ed., Macmillan, New York, 651 ( 9 2 . 17) 2. Strassman, S. C. and F. W. Kutz, "Insecticide Residues in Human Milk from Arkansas and Mississippi, 1973-74," Pest. Moo. J., 10, 130-133 (97. 17) 3. Savage, E. P., et. al., "Organochlorine Pesticide Residues and Polychlorinated Biphenyls in Human Milk 1971-72," Pest. Mon. J., 7_, 1-3 (1973). 4. Kroger, M., "Insecticide Residues in Human Milk," J. Pediat., 80, 401405 ( 9 2 . 17) 5. Curley, A. and R. Kimbrough, "Chlorinated Hydrocarbon Insecticides in Plasma and Milk of Pregnant and Lactating Women," Arch. Environ. Health, 18, 156-164 ( 9 9 . 16) 6. Dyment, P. G., et al., "Relationship Between Levels of Chlorinated Hydrocarbon Insecticides in Human Milk and Serum," Bull. Environ. Contamin. Toxicol., 6, 449-452 (1971). 7. Halyard, S. B., W. H. Brown, J. W. Stull, F. M. Whiting, and S. R. Kemberling, "DOT and ODE Content of Human Milk in Arizona," Bull. Environ. Contamin. Toxicol., 9, 169-172 (1973). 8. Quinby, G. E., J. F. Armstrong and W. F. Durham, "DDT in Human Milk," Nature, 207. 726-728 (1965). 9. Lang, E. P., F. M. Kunze and E. S. Prickett, "Occurrence of DDT in Hunan Fat and Milk," Arch. Indust. Hyg., 3, 245-246 (1951). 10. Wilson, 0. J., et al., "DDT Concentrations in Human Milk," Am. J. Dis. Child, 125. 814-817 (1973). 11. Woodard, B. T., B. B. Ferguson and D. J. Wilson, "DDT Levels in Milk of Rural Indigent Blacks," EPA-600/1-76-032 (1976). 61 �12. Savage, E. P., et al., "A Search for Polychlorinated Biphenyls in Human Milk in Rural Colorado," Bull. Environ. Contamin. Toxicol., 9, 222-226 (1973). 13. Petrakis, N. L., L. D. Gruenka, T. C. Beelen, N. Castagnoli, Jr., and J. C. Craig, "Nicotine in Breast Fluid of Nonlactating Women," Science, 199. 303-305 (1978). 14. Bakken, A. F. and M. Seip, "Insecticides in Human Breast Milk," Acta Paediatr. Scan., 65, 525-529 (1976). 15. Knoll, W. and S. Jayaraman, "Zur Kontamination Von Humanmilch tnit chlorierten Kohlenwasserstoffer," Die Nahrung, 17, 599-615 (1973). 16. Psendorfer, Von H., "Ruchstande von Organochlorpestiziden (DDT u.a.) und polychlorierten Biphenylen (PCBs) in der Muttennllch," Wiener Klinische Wochenschrift, 87, 731-736 (1976). 17. Th. TuLustra, L. G. M., "Organochlorine Insecticide Residues in Human Milk ia the Leiden Region," Neth. Milk Diary J., 25, 24-32 (1971). 18. Polishiik, Z. W., M. Ron, M. Wasserman, S. Cucas, 0. Wasserman, and C. Lemesca, "Organochlorine Compounds in Human Blood Plasma and Milk," Pest. Mon. J., 10, 121-129 (1977). 19. Ritchy, W. R., G. Savary and K. A. McCulley, "Organochlorine Insecticide Residues in Human Milk, Evaporated Milk, and Some Milk Substitutes in Canada," Can. Publ. Health J., 63, 125-132 (197?.). 20. Egan, H., R. Goulding, J. Roburn and J. O'G. Tatton, "Organo-chlorine Pesticide Residues in Human Fat and Human Milk," Bxit. Med. J., 2, 6669 (1965). 21. Newton, K. G. and N. C. Greene, "Organocblorine Pesticide Residue Levels in Human Milk — Victoria, Australia 1970," Pest. Mon. J., 6, 4-8 (1972). 22. Graca, I., A. M. S. Silva Fernandes and H. C. Mourao, "Organochlorine Insecticide Residues in Human Milk in Portugal," Pest. Mon. J., 8, 148-156 ( 9 4 . 17) 23. Musial, C. J., 0. Hutzinger, V. Zitko and J. Crocker, "Presence of PCB, DDE, and DDT in Human Milk in the Providences of New Brunswick and Nova Scotia, Canada," Bull. Environ. Contamin. Toxicol., 12, 258-267 (1974). 62 �24. Hes. J. and D. J. Oavies, "Presence of Polychlorinated Biphenyl and Organochlorine Pesticide Residues and the Absence of Polychlorinated Terphenyls in Canadian Human Milk Samples," Bull. Environ. Contain. Toxicol., 21, 381-387 ( 9 9 . 17) 25. Stacey, C. I. and B. W. Thomas, "Organochlorine Pesticide Residues in Human Milk, Western Australia — 1970-71," Pest. Mon. J., 9, 64-66 (1975). 26. Van House Holdrinet, M., H. E. Braun, R. Frank, G. J. Stopps, M. S. Smout, and J. W. McWade, "Organochlorine Residues in Human Adipose Tissue and Milk from Ontario Residents," Can. J. Pub. Health, 68, 74-83 ( 9 7 . 17) 27. Winter, M., M. Thomas, S. Wernick, S. Levin and M. T. Farver, "Analysis of Pesticide Residues in 290 Samples of Guatemalan Mother's Milk," Bull. Environ. Contamin. Toxicol., 16, 652-657 ( 9 6 . 17) 28. "Criteria for a Recommended Standard...Occupational Exposure to Ethylene Bichloride (1,2-dichloroethane)," HEW Publ. No. (NIOSH) 76-139 (March 1976). 29. Kover, F. D., Environmental Hazard Assessment Report. Chlorinated Naphthalenes. EPA 560/8-75-001 (December 1975). 30. Erickson, M. D., R. A. Zweidinger, L. C. Michael and E. D. Pellizzari, "Environmental Monitoring Near Industrial Sites: Polychloroaaphthalenes," EPA-560/6-77-019 (1977). 31. Erickson, M. D., L. C. Michael, R. A. Zweidinger, and E. D. Pellizzari, "Development of Methods for Sampling and Analysis of Polychlorinated Naphthalenes in Ambient Air," Environ. Sci. Techno1., 12, 927-931 (1978). 32. Erickson, M. D., L. C. Michael, R. A. Zweidinger and E. D. Pellizzari, "Sampling and Analysis for Polychlorinated Naphthalenes in the Environment," JAOAC, 61, 1335-1346 (1978). 33. Erickson, M. D., L. C. Michael, R. A. Zweidinger, and E. D. Pellizzari, "Development of Methods for Sampling and Analysis of Polychlorinated Naphthalenes in Ambient Air," 1977 Annual Meeting, American Chemical Society, Chicago, IL (August 31, 1977). 63 �34. Erickscn, H. D., L. C. Michael, R. A. Zweidinger, and E. D. Pellizzari, "Sampling and Analysis for Polychlorinated Naphthalenes in the Environment," 1977 Annual Meeting AOAC, Washington, DC (October 20, 1977). 35. Unpublished data, E. Roessler, Borough of Bridgeville, PA (1976). 36. 197? Directory of Chemical Producers-USA, Chemical Information Services, Stanford Research Inst., Menlo Park, CA (1977). 37. Environmental Sciences and Engineering, "Trip Report for Sampling of Polybrominated Biphenyls (PBBs)," submitted to OTS, EPA, Washington, DC, Contract No. 68-01-3248 (April 1977)). 38. Mumma, C. E. and D. D. Wallace, "Survey of Industrial Processing Data. Task I - Pollution Potential of Polybrominated Biphenyls," EPA-560/375-004 (June 1975). 39. Unpublished data, E. J. Londres, New Jersey Dept. of Environmental Protection via G. E. Parris, OTS, EPA, Washington, DC (1977). 40. Erickson, M. D., R. A. Zweidinger, and E. D. Pellizzari, "Analysis of a Series of Samples for Polybrominated Biphenyls (PBBs)," EPA-560/6-77-020 (Augus-: 1977). 41. Environmental Science and Engineering, "Data Report for Polybrominated Biphenyl Near Manufacture (sic) in the Northeast," submitted to OTS, EPA, Washington, DC Contract No. 68-01-3248 (June 16, 1977). 42. 1974 New Jersey State Industrial Directory, New Jersey State Industrial Directory, 2 Perm Plaza,. NY, 10001 (1974). 43. Pellizzari, E. D., "The Measurement of Carcinogenic Vapors in Ambient Atmospheres," EPA-600/7-77-055 (June 1977). 44. Pellizzari, E. D., M. D. Erickson, and R. A. Zweidinger, "Formulation of a Preliminary Assessment of Halogenated Organic Compounds in Man and Environmental Media," EPA-560/13-79-006 (July 1979). 45. Pellizzari, E. D., M. D. Erickson, T. D. Hartwell, S. R. Williams, C. M. Sparacino and R. D. Waddell, "Preliminary Study on Toxic Chemicals in Environmental and Human Samples. Part I: Formulation of an Exposure and Body Burden Monitoring Program," submitted to IT. S. Environmental Protection Agency, Washington, DC, Contract No. 68-01-3849 (June 1980). 64 �46. Pellizzari, E. D., "Analysis of Organic Air Pollutants by Gas Chromatography and Mass Spectrescopy," Publication No. EPA-600/2-77-100, Contract No. 68-02-2262, (June 1977). 47. McDonnell, G., D. H. Ferguson and C. R. Pearson, "Chlorinated Hydrocarbons and the Environment," Endeavour, 34, 13-18 ( 9 5 . 17) 48. FDA Compliance Program, Evaluation, "FY 74 Total Diet Studies (7320.08)," Date accepted: January 21, 1977. 49. State of New Jersey Department of Environmental Protection, "Initial Report on the Findings of the State Air Monitoring Program for Selected Volatile Organic Substances in Air," (October 1979). 50. Zweidinger, R. A., A. Sherdon, B. S. Harris, III, H. Zelon, T. Hartwell, and E. D. Pellizzari, "Measurement of Benzene Body Burden of Potentially Environmentally Exposed Individuals," Final Report, EPA Contract No. 68-OL-3849, Task 1 (May 1980). 51. Hartwell, T., P. Piserchia,.S. White, N. Gustafson, A. Sherdon, R. Lucas, D. Lucas, D. Myers, J. Batts, R. Handy, and S. Williams, "Analysis of E?A Pesticide Monitoring Networks," Office of Toxic Substances, Washington, DC. Draft Report (1979). 52. U.S. Environmental Protection Agency, Office of Research and Development, "Health Assessment Document for Polycyclic Organic Matter," (May 1978). 53. Stanford Research Institute, "The Environmental Fate of Selected Polynuclear Aromatic Hydrocarbons," Prepared for U. S. Environmental Protection Agency (February 1976). 54. State of New Jersey Department of Environmental Protection, "Initial Report on the Findings of the State Air Monitoring Program for Selected Heavy Metals in Air," (October 1979). .55. Unpublished'data, William J. Librizzi, U.S. Environmental Protection Agency, Region II (October 1977). 56. Fribers, L., M. Piscator, G. F. Nandberg and T. Kjellstrom, "Cadmium in the Environment," CRC Press, Cleveland, OH (1974). 65 �57. National Academy of Sciences, "Lead," Washington, DC ( 9 2 . 17) 58. Mason, T. J., F. W. McKay, "U.S. Cancer Mortality by County: 1950-69," DHEW Publ. No. (NIH), 74-615, Washington, DC, U.S. Govt. Printing Office (1974). 59. Mason, T. J., F. W. McKay, J. R. Hoover, W. Blot and J. F. Fraumeni, Jr., "Atlas of Cancer Mortality for U.S. Counties: 1950-69," DHEW Publ. No. (NIH) 75-780, Washington, DC, U.S. Govt. Printing Office (1975). 60. Greenberg, Michael R., "The Spacial Distribution of Cancer Mortality and of High and Low Risk Factors in the New Jersey-New York-Philadelphia Metropolitan Regions, 1950-1969, Part I," New Jersey Dept. of Environmental Protection, Program on Environmental Cancer and Toxic Substances (January 1979). 61. Greenberg, M., F. McKay, and P. White, "A Time-Series Comparison of Cancel Mortality Rates in the New Jersey-New York-Philadelphia Metropolitan Region and the Remainder of the United States, 1950-1969," Am. Jour, of Epidemiology, 111. 166 (I960). 62. Greenberg, M. R., P. W. Preuss, and R. Anderson, "Clues for Case Control Studies of Cancer in the Northeast Urban Corridor," Soc, Sci. & Med., 14D, 37-43 (1980). 63. Greenberg, M. R., J. Caruana, B. Holcomb, G. Greenberg, R. Parker, J. Louis, and P. White, "High Cancer Mortality Rates from Childhood Leukemia and Young Adult Hodgkin's Disease and Lymphoma in the New Jersey-New York-Philadelphia Metropolitan Corridor, 1950-1969," Cancer Research, 40, 439-443 ( 9 0 . 18) 64. Cross, J. and G. B.tWiersma, "Preliminary Analysis of Cancer Rates in Organic Chemical-Producing Counties," EPA-600/1-79-022 (June 1979). 65. Pellizzari, E. D., and M. D. Erickson, "Analysis of Organic Air Pollutants in the Kanawha Valley, WV and the Shenandoah Valley, VA," Publication No. EPA-903/9-78-007, Contract No. BOA 68-02-2543 (June 1978). 66. Erickson, M. D., S. P. Parks, D. Smith and E. D. Pellizzari, "Sampling and Analysis of Organic Air Pollutants in Two Industrialized Valleys," FACSS V, Boston (October 30 - November 3, 1978. 67. McLafferty, F. W., E. Stenhagen, and S. Abrahammson, Ed., Registry of Mass Spectral Data, John Wiley and Sons, New York (1974). 66 �68. Eight Peak Index of Mass Spectra. Vol. I (Tables 1 and 2) and II (Table 3), Mass Spectrometry Data Centre, AWRE, Aldermaston, Reading, RG74PR, UK (1970). 67 �APPENDIX A DATA COLLECTION INSTRUMENTS 68 �STUDY OF ORGANIC COMPOUNDS IN HUMAN MILK EPA Contract No. 68-01-3849 RTI Project No. 31U-1S21-22 DATA COLLECTION INSTRUCTIONS Performed for Office of Toxic Substances Environmental Protection Agency Washington, DC 20460 69 �1.0 Introduction Under contract to the Office of Toxic Substances, Environmental Protection Agency (EPA), the Research Triangle Institute (RTI) is conducting a limited study designed to measure environmental pollutant levels in human Bilk and to evaluate the utility of using this body fluid it. specific pollutant studies for populations in the. vicinity of manufacturing plants and/or industrial user facilities. RTI is responsible for all phases of the study, including study design, subject recruitment, chemical analysis of milk samples, and report writing. RTI is a not-for-profit contract research organization located in North Carolina's Research Triangle Park between Raleigh, Durham, and Chapel Hill. The Institute was incorporated as a separate operating entity in 1958 by the University of North Carolina (UNO at Chapel Bill, Duke University at Durham, and North Carolina State University at Raleigh, and is still closely affiliated with the three universities. 2.0 Overview Four urban areas have been chosen as performance sites; they are Bridgeville, Pennsylvania; the area which includes Linden and Bayonne, New Jersey and western Staten Island, New York; Baton Rouge, Louisiana; and South Charleston and Nitro, West Virginia. These sites represent high-probability areas for the presence of one or more of the chemicals of interest in human milk. The selected industrial chemicals of interest include polychlorinated naphthalenes, tetrachlorethylene, trichloroethane, diehloropropane, benzene, polybrominated biphenyls, chlorinated phenols, toluene:, chlorinated benzenes, and chloroform. 70 �At each of the four sites, arrangements will be made to work through clinical facilities such as hospitals, clinics, or physician's offices, In order to recruit a panel of respondents. At each site ten participants will be recruited, for a total of 40. Potential participants (lactating females) will be screened to determine that they live in one of the areas of interest and are willing and able to provide the milk sample. A questionnaire will be administered for each participant to obtain information on demographic variables, residence histories, and potential exposure situations; for each participant, a sample of milk will be collected and analyzed for the compounds of interest by gas chromatography/mass spectrometry or high pressure liquid chromatography. A professional member of the facility's staff, such as a registered nurse, will be trained in the proper procedures to administer the questionnaire and obtain the milk sample, To try to reduce the non-participation rate due to refusals, and to reimburse the subject for the time spent on the study, volunteers will be offered a $5.DO incentive for participating. 3.0 Data Collection 3.1 General Remarks Data collection for this research effort consists of the following steps: 1. Screening of potential participants (lactating women) to determine that they live in one of the areas of interest (see below), that they have resided in that area for at least the preceding 12 months, that they have remained in that area continuously for the preceding week, and that they are willing and able to provide a milk sample. 71 �2. When an eligible person is encountered, the nature and purpose of the study will be explained and their participation solicited. 3. When an eligible person agrees to participate, the person will be required to sign a Participant Consent Form (PCF) in order to participate in the study. 4. Once the participant has signed the PCF, the person should be listed on the Participant Listing Form (PLF), a Patient Number assigned, and the data collector will proceed to administer the Study Questionnaire (SQ) and collect the tr.Uk sample. 5. Once the SQ has been administered and the milk sample collected, the participant will be offered a $5.00 incentive for participating. * 6. Milk samples and completed data collection instruments will be returned to RTI. 3.2 Survey Instruments As indicated in the preceding section, there are 3 data collection instruments for this research effort, the PCF, the PLF, and the SQ; subsequent sections contain instructions for the use of each instrument as well as item-by-tiem explanations for their completion, and general descriptions are provided below. The survey instruments have been designed hopefully to provr.de an efficient means of collecting and recording the requisite data for the study. It is imperative that all s-urvey instruments be completed accurately. The success and reliability of the study and its results are dependent upon the quality of data collected, which will be fully dependent 72 �on the accuracy of your execution of your assignment. As you complete a fora, conduct a thorough edit to verify that required data have been entered and entered correctly. Copies of the data collection instruments appear in Attachment 1. 3.2.1 Participant Consent Fora (PCF) • Purpose; The purposes of the PC? are to introduce the study; explain its objectives, sponsorship (the relationship and roles of 811 and EPA), and requirements of and risks, burdens, and benefits to participants; and stress that participation is completely voluntary and that all data collected will be kept confidential. . General Description; The PCF is a single page fora printed on special paper which makes three copies from a single impression. The survey title appears at the top, along with the name of RII; spaces for necessary identifying information appear at the bottom. . Administration; The PCF will be signed by the participant and contains an agreement to provide the necessary information and milk sample. Participants may freely withdraw from the study at any time; however, in order to encourage participation RII offers an incentive of five dollars to each participant to be paid after each data set (PCI, SQ, and milk sample) is obtained. Again, confidentiality of data is stressed, including steps 73 �taken to disassociate the name of the participant from the data once collected; for example', the PCF is the only data collection instrument which bears the name of the participant and allows its association to study identification numbers, but will be maintained in hard copy only and stored in a restricted area. To further emphasize this disassociation, the incentive will be paid in cash rather than by check or money order, although the participant will sign the PCF indicating that the incentive was received. A signed PCF must be obtained for each participant before proceeding with Study Questionnaire (SQ) administration and collection of the milk sample. • Disposition; The top (white) copy will be attached to the appropriate SQ until it is received at RTI and verified; the yellow copy will be provided to the participant; the pink copy will be retained by the data collector. 3.2.2 Participant Listing Fora r(PLF) . Purpose; The purpose of the FLF is to provide a means of assigning unique numbers to participants at each performance site. . General Description; The PLF is a single page form printed on pink paper; space for Comments is provided on the reverse side. The survey title appears at the top, along with the names and addresses of RTI and EPA/OTS and a confidentiality statement. 74 �. Administration; Aa each participant is enlisted up Co the required number ( 0 , that participant should 1) be listed on the PLT. . Disposition; When data collection at a site or facility is completed, the FLF (or a copy) should be sent to RTI. 3.2.3 Study Questionnaire (SQ) : The purpose of the SQ is to obtain information on participants, Including demographic characteristics such as age, sex, race, and occupation; residence information; health information such as current health status and prescription medications; and personal characteristics such as hobbies. . General Description; The SQ is divided into six sections, dealing respectively with demographic characteristics, occupation, health and personal habits, residence and household information, information on the interviewer and respondent, and information regarding the milk sample, including an indication as to whether or not the milk sample was obtained, the date and time of acquisition of the sample, and the date the sample was shipped to RTI. Participants will be identified by a unique study number used to correlate and cross-identify the questionnaires and samples by way of pre-printed self-adhesive labels. The SQ is 5 pages long, with space provided for comments, 75 �. Administration; An SQ is to be completed for each participant for whoa a signed PCF is obtained. • Disposition! The SQ1* are to be sent to RTI as instructed. 3.3 Screening As indicated in section 3.1, potential participants (lactating women) should be screened to determine that they meet certain study criteria for participation: 1. That they are willing and able to provide a milk sample of sufficient quantity (approximately 100 ml.), 2. That they live in one of the areas of interest (see below), 3. That they have resided in that area for at least the preceding 12 months, and A. That they have remained in that area continuously for the preceding 7 days. As indicated in section 2.0, four areas have been chosen as performance situs, with a specific Site Number assigned to each which will remain constant ilor each site and is to be entered where appropriate on data collection instruments as follows: Site Site Number Northern New Jersey/Staten Island, New York Bridgeville, Pennsylvania Baton Rouge, Louisiana Nitro/South Charleston, West Virginia 1 2 3 4 With the exception of Bridgeville, Pennsylvania, participants residing in some areas at each site are of considerably more interest to the study than those living in others, as discussed in the following sections. 76 �3.3.1 Northern Sew Jersey/Staten Island. New York Within the Northern New Jersey/Staten Island area, potential participants residing in some communities are of more interest than those residing in others, more or less in the order listed below: 1. Bayonne, NJ 2. Northern Staten Island (Port Richmond), NY 3 ' Linde0' » 4. Carlstadc, NJ 5. Saddle Brook, NJ 6. Jersey City, NJ 7. Kearney, NJ 8. Newark, NJ 3.3.2 9. Elizabeth, NJ 10. Sayreville, NJ ,, _ . ... 11. Rahway, NJ 12. Edison, NJ ^ Parlin> NJ ^ Patterson, NJ ^ Wayne> HJ Baton Rouge, Louisiana Potential participants residing in Baton Rouge are of primary interest to this study; other communities in the Baton Rouge area of interest are Placquemine, St. Gabriel, and Geismar. 3.3.3 Nitro/South Charleston, West Virginia Potential participants residing in Nitro and South Charleston are of primary interest to this study; other communities of interest in the area are Belle and Institute. 3.4 Participant Listing Fora When an eligible person is encountered who agrees to participate, that person should be listed on a PLF in order to be assigned a unique Participant Number. The PLF is completed by entering the appropriate Site Number (see section 3.3 above); then, each time that an eligible participant is encountered who agrees to participate, up to the number required, enter the Participant's Name (Last, Firstt Middle) on the PLF and assign a Porticipois Number in the left-hand column, beginning with 0001 at each site unless other77 �vise instructed. Assign Participant Numbers consecutively for all study participants. Where appropriate, enter the participant's Medical Record Kuniber in the right-hand column. When making numerical entries, right-adjust and enter leading zeros. 3.5 Participant Consent Fora Potential participants must understand exactly -what is involved in participation in the study and what benefits may be realized by participation; this understanding and agreement must be documented by a signed PCF. In the event that the potential participant is under the age of 18 years, the person's parent or other legal guardian must sign the PCF in order for the designated eligible to participate. More specifically, the potential participant and/or that person's parent, guardian or other spokesman, must understand that full participation in the study consists of providing answers to a questionnaire related to environmental exposure, part of which relates to the individual's household in general and part of which is related to the individual participant (be prepared to show the person the SQ), and providing a ntf.lk sample of approximately 100 ml. (be prepared to show the person one of the collection bottles.) The individual must further understand that she will only enjoy certain limited benefit.1! in return for her time and inconvenience, primarily a $5.00 incentive to be disbursed after administration of the questionnaire and collection of the milk sample. The individual must understand that participation in the study iu completely voluntary and that she may withdraw at any tine, but that payment of the incentive is dependent on full pcrtiGipavion. The individual must also understand that all data collected in the study will be held strictly confidential, and that names will not be disclosed. 78 �If the participant or that perons's parent, guardian or other spokesman agrees to participate, read through the PC? with then and make entries where appropriate. At the bottom, record the Date (month, day, aid year) that the PCS' is signed and print the Participant's full Same (First, Middle or Maiden, Last - do not abbreviate); record the appropriate Site Sianber (see section 3.3 above) and Participant Number (from the PLF); have the participant (or other appropriate person) sign the PCF; enter jour signature as witness; and record the participant's home Address (Street Number and None, City, State, and Zip Code) in the spaces provided. After data collection (administration of SQ and collection of milk sample) is completed, the participant (or that person's parent or guardian) should be given $5.00. The recipient must sign in the space provided at the bottom of the PCF to indicate receipt of the incentive. Should the signatures on the PCF for Participant and Recipient be other than the participant's, please explain in the Comments section of the SQ. Finally, as indicated in section 3.2.1, the top (white) copy of the PCF is to be attached to the appropriate SQ; the yellow copy is to be provided to the participant or her guardian; and the pink copy is to be retained by the data collector. 3.6 Study Questionnaire Before proceeding with administration of the SQ, read the justification and confidentiality statement in the box on the cover. Enter the appropriate Site (see section 3.3 above) and Participant (from the PL?) Numbers. Stapled inside the SQ you will find a set of pre-printed, selfachesive labels which are necessary to identify corresponding SQs and samples. Each label contains a unique Study Number, which should be the same on all 79 �labels in a set, and an indication of what the label is for. You should also have some labels that have Cj.ly a Study Number and a few that are completely blank; these are for your use in the event that a label is damaged or missing. If you use a label that has a Study Number only, you will have to write on the label what it is intended for, such as MILK; if you use a blank label, you must write on the label the Study Number and what i'C is intended for. Check to be sure that all the labels in a given SQ contain the same Study Number; if not, do not use the SQ and return it to RTI. If the Study Number is the. same on all labels, remove the one for the QUESTIONNAIRE and place it on the cover of the SQ over the spaces provided for the Study Number. Space for Comments is provided on page 5. If the.participant is under 18 years of age, the SQ may have to be administered in whole or part to the parent or guardian, and must be administered in that person's presence.- If the participant suffers from a speech or hearing deficit, or is otherwise incapacitated, the SQ may have to be administered to the spouse or some other spokesman. Item 1 - Race; Indicate the participant's race by placing an X in the appropriate box. This question may be answered by observation; however, if there is any doubt whatsoever, ask. Item 2 - Age; Determine and enter the participant's age in years as of the last birthday. Item 3 — Birthdate; Determine and enter the participant's exact birthdste (month, day and year). Again, remember to rightadjust and enter leading zeros. A note on dates: accept and record partial datest if that is all that the respondent can provide; in that case, indicate missing elements of the date 80 �with • dash ( ) — for example, April 1977 would be recorded as |01AJ - H j - |7|7j . Item 4 - Weight; Determine and enter the participant's approximate weight In pounds (to the nearest pound—no fractions!) or kilograms, in which case observe the decimal. Item 5 - Height; Determine and enter the participant's approximate height in inahea or centimeters. Item 6 - Current Employment; Determine if the participant is currently employed in any capacity and place an X in the appropriate box. If the answer is Yes, continue to Item 7; if the answer is No, skip to Item 10. Item 7 - Length of Present Employment: Determine and record the length of time that the participant has been employed by her•present employer; enter the units in the spaces provided and then place an X in the appropriate box to indicate whether the units represent days, months, or years. Item 8 - Occupation Away From Home; Determine if the participant's occupation usually takes her away from home and place an X in the appropriate box. If Yes, continue to Item 9; if No, skip to Item 11. This question, and Item 9 below, are aimed at eliciting information regarding the location of the participant's various exposure to the environment. Item 9 - Location of Present Employment; If the participant is currently employed, determine the nature (not the name) and location (street address, city, state, and Zip Code, if known) 81 �of Che employer. By naturet we mean Che type of business, such as service station, school, hospital, grocery store, doctor's office, hotel, restaurant, etc. Item 10 - Employment Statusr If the participant is not presently employed, determine which of the provided categories best describes the participant's status and place an X in the appropriate box. If the response is choice 1 or 2, skip to Item 15; if the response is choice 3-5, continue to Item 11. Itec 11 - Usual Occupation; Determine and record the participant's usual (or most common) occupation (when employed); be succinct e.g., high school coach, waitress, hotel desk clerk, taxi driver. Item 12 - Present Occupation; Determine if the participant is presently employed in her usual occupation (indicated in Item 11) and place an X in the appropriate box. Items 12 and 13 may be skipped for unemployed, retired and disabled persons. Item 13; If the response to Item 12 was positive, determine how long the participant has been employed in her usual occupation (recorded in Item 11) and record; enter the units in the spaces provided and then place an X in the appropriate box to indicate whether the units represent days, months or years. Item 14; Determine if the participant presently works at or in any of the listed occupations or establishments and place an X in each appropriate box. Item 15 - Present JSmoking Status.; Ascertain if the participant currently smokes cigarettes, and place an X in the appropriate box. If YES, continue to Item 16; if NO, skip to Item 18. 82 �*• I_ten_16 -_ Age at First Smoke; If the participant is a smoker (a positive response to Item 15), ascertain the age (in years) at which the participant started smoking and record in the spaces provided. Item 17 - Smoking Frequency: Ascertain how many cigarettes the participant smokes per day, an the average, and place an X in the appropriate box. If the participant uses tobacco in some form other than cigarettes, such as snuff, record in the space provided. Item 18 -Time Outdoors; Ascertain the average number of hours that the participant spends out of doors each day and record in the spaces provided — another indication of environmental exposure. Item 19 "Time Away From Hornet Determine how many hours of the day on the average the participant normally spends more than 2 miles away from home, and record in the spaces provided. This determination should be done separately for weekdays and weekends. Item 20 - General Health Status; Using the four qualifiers provided, ascertain the participant's general current health status and place an X in the appropriate box. Item 21 - Prescription Medications; Inquire as to whether the participant is currently taking any prescription mediaavionfs) on a regular daily basis and place an X in the appropriate box; if YES, determine and record the drug name - e.g., penicillin, oral contraceptives, Vallum, phenobarbital, etc. 83 �Item 22 - Non-prescription Medications; Inquire as to whether the participant has taken any non-prescription medications in the past 24 hours, and place an. X in the appropriate box; If YES, determine and record the drug name -e.g., aspirin, vitamins, Dristan, Bufferin, Alka-Seltzer, etc. Item 23 - Gasoline; Inquire as to whether the participant pumps her own gasoline, for example at self-service pumps, and place and X in the appropriate box. Item 24 - Egg Consumption; Determine and record the approximate number of -eggs that the participant has eaten in the past 48 hours. Again, in recording numerical entries, remember to right-adjust and enter leading zeros. Item 25 - Hobbies; Determine if the participant pursues any of the listed avocations and place an X in each appropriate box. Item 26; Determine if the participant pursues any activity that includes regular use of solvent glue or model airplane cement, and place an X in the appropriate box. Item 27 - Length of Residence in Area; Determine hew many years the participant has lived in the area, of interest, and record in the spaces provided. Round to the nearest year, except that if the response is less than one year record as [ < ! I I and terminate the interview; the individual is ineligible to participate further in the study. This situation should be detected during the screening process. 84 �Item 28 - Length of Residence at Current Address; Determine how long Che participant has lived at her current address; record the units in the spaces provided and place an X in the appropriate box to indicated whether the units represent days, months, or years. Use the most appropriate units and round to the nearest appropriate unit. For example, more than 28 days should be expressed in months and more than 11 months should be expressed in years. If the participant has resided at her current address for less than 12 months, but has lived in the area of interest for at least 12 months, record any previous addresses during the preceding 12 months (city and state is sufficient) in the Comments section. Item 29 - Cooling Appliances; Determine whether any of the indicated appliances or others, in which case specify, are used to cool the participant's home and place an X in the appropriate box(es) for all that apply. Item 30 - Home Garden; Determine if the participant's household consumes food grown in a home garden and indicate the response by placing an X in the appropriate box. If a positive response is obtained, determine the location of the garden and record. Location could be participant's backyard, or another community, in which case specify city and state; be as specific as possible. Item 31 __- Commercial Food Source; Determine where the participant's household usually obtains fruit and/or vegetables and record. 85 �Again, be as specific as possible. For example, if Che city or town has more than one store by the same name, the store name alone would not be an adeuqate answer; as a matter of course, record the name and location of the store, market, or vendor. Items 32-34 - Water Sources; In Item 32, try to determine the primary source of drinking water for the participant's household and place an X in the appropriate box. In Item 33, determine if the same primary drinking water source indicated in Item 32 is used for drink mixes such as coffee and tea; if it differs, indicate how. In Item 24, try co determine the primary source of water for cooking in the participant's household and place an X in the appropriate box. For example, some households in some areas of the country use bottled water for drinking and drink mixes but tap water (from whatever source) in cooking. Itea 35 - Other Household Tobacco Use; Inquire as to whether other members of the participant's household smoke, and place an X in the appropriate box; if YES, dpr.ernine if the other members smoke cigarettes, cigars, a pipe, etc. and place an X in each appropriate box. Item 36 - Occupation of Other Household Members: Determine if any other members of the participant's household work at any of the listed occupations or businesses, and place an X in each appropriate box. 86 �Item 37 - Hobbies of Other Household Members; Determine if any other members of the participant's household pursue any of the listed avocations, and place an X in each appropriate box. Respondent/Interviewer Information Item 38 - Respondent; Indicate, by placing an X in the appropriate box, whether the person who served as the primary respondent was the participant or some other person, in which case specify in the space provided. Iten 39 - Interviewer Number; Enter your assigned 3-digit Interviewer identification Number. Item 40 - Date of Interview; Enter the date (month, day and year) that the interview was conducted and the questionnaire completed. Item 41 - Interviewer Name; The name of the person administering the questionnaire should be printed in the space provided. Sample Information Item 42; Indicate, by placing an X in the appropriate box, whether or not a milk sample was collected; if not, explain in the • Comments section below. Item 43 - Date, and Tine of Milk Saaple Collection; If a milk sample is collected, record the date (month, day and year) and approximate time (using a 24-hour clock) of such collection. The time should correspond to the time that collection was completed; on a 24-hour clock, add 12 to the p.m. hours - e.g., 1:00 p.m. would be 13:00, 5:30 would be 17:30, etc. 87 �Iten 44 - Date Shipped to RTI; Record the date (month, day and year) that the respective milk sample was shipped to RTI, or turned over to an RTI representative. 3.7 Collection of the Milk Sample 3.7.1 General Remarks As indicated in section 1.0 above, the milk samples are being collected for chemical analysis by RTI as part of an EPA study to measure pollutant levels in human milk and evaluate the utility of using this body fluid in specific pollutant studies. The chemical compounds for which the samples will be analyzed are present in extremely low levels, so the utmost care and cleanliness must be used to prevent either contamination or loss. The instructions below are designed to preserve the integrity of the sample and should be followed precisely. 3.7.2 Sample Collection Instructions 1. The bottles provided have been thoroughly cleaned and should be kept tightly closed, except during sampling; do not wash or otherwise clean them. 2. Remove the MILK SAMPLE label from the sheet of labels in the appropriate SQ and place on one of the collection bottles. 3. The milk should be manually expressed directly into the the bottle; do not use breast pumps or other devices as the plastics in such devices would contaminate the sample. Hands should be cleaned and thoroughly rinsed to remove any residual soap; do not use rubber gloves. 88 �4. Collect as ouch milk as possible. Unless the mother has recently nursed her infant, at least half a bottle should be easily obtainable. Less than half a bottle is unuseable and does not constitute a sample. The ability of the participant to provide an adequate sample should be determined during the screening process. 5. Immediately cap the bottle and double check to see that the study numbers on the bottle and questionnaire match. 6. The milk sample should be immediately frozen following collection and remain so until shipping. 7. Note any deviations from this procedure in the Comments . section of the appropriate SQ. 3.7.3 Shipping Instructions 1. Pack the container as it was received. 2. Fill the can with dry ice. 3. Make sure that there is adequate padding to prevent breakage, that all excess space is filled with packing material. 4. Fill out enclosed Federal Express forms, attach to the outside of the box, and seal the box. 5. Call Federal Express and have them pick up the package. 6. When Federal Express picks up the package, call Dr. Mitch Erickson at RTI (see below) to notify him that Federal Express has picked up the package; if Dr. Erickson is ouc, leave an appropriate message with his secretary. 89 �7. Mail Che corresponding questionnaires to RTI in one of the envelopes provided. 8. When the questionnaires are in the. mail, call Ben Harris at RTZ (see below) to notify him that the questionnaires are in the mail; if Mr. Harris is out, leave an appropriate message with his secretary. 4.0 Confidentiality All survey research conducted by RTI is based on highest ethical standards, including those related to confidentiality. These standards are applied from the earliest steps of deciding whether or not RTI should participate in a proposed survey to the final steps of analyzing and reporting the information obtained. Strict precautions must be observed at all times to protect the rights of those whom we interview or about whom we collect data. Such 9 precautions are built into the study design, so that promises of confidentiality and anonymity will be upheld during all phases of data handling and analysis. Ho amount of effort to insure confidentiality will be successful, however, unless those responsible for data collection in the field maintain equally rigid standards, treating with utmost confidence all information offered or observed during data collection. Successful and meaningful survey research is dependent on the establishment of trust between individuals engaged in data collection and sources of information, and maintaining this sense of responsibility to the public throughout all survey activities. Each data collector will be required to sign in duplicate a contractual agreement which Includes provisions on confidential treatment of data. This agreement is designed to protect you as well as RTI and participating institutions and individuals. A copy of this agreement appears in Attachment 2. 90 �The Importance of cotal confidentiality cannot be over-emphasized. Any breach of confidence could result in litigation. 5.0 Contacts with Project Staff During the data collection period it will be necessary for data collectors to maintain regular contact with RTI project staff by telephone. While you are collecting data, problems or confusing issues may arise that are not addressed in these instructions. You are encouraged to telephone RTI whenever you experience a problem or encounter a situation which you feel you cannot adequately handle. All supplies required for data collection will be furnished by RTI. Should you require additional supplies during the conduct of data collection, inform your RTI contact so that proper arrangements can be made. Need for additional supplies should be anticipated so that your work will not be delayed while you await receipt of needed items. All study-related items that are in your possession at the conclusion of data collection are to be returned to RTI or disposed, of according to instructions from your RTI contact. Calls to RTI should be made between the hours of 8:30 a.m. and 5:00 p.m. (Eastern Time), Monday through Friday, to RTI'3 toll-free number, 800-334-8571. Request to speak to the appropriate project staff member listed below: Or. Mitch Erickson Extension 6505 (regarding milk sample collection) MX. Ben Harris Extension 6055 (regarding participant selection and questionnaire administration) 91 �If che problem is particularly acute, and you have trouble getting through on .the toll-free line, call collect 919-541-6505 (Dr. Erickson) or 919-541-6055 (Mr. Harris). After 6:00 p.m. Eastern Time you may call Mr. Harris collect at work (919-541-6055) or person-to-person at home (919-942-6988). 92 �Attachment 1 Data Collection Instruments 93 �OMVNo. IM-ITtOI im Snnmoir 1H RESEARCH TRIANGLE INSTITUTE STUDY OF ORGANIC COMPOUNDS IN HUMAN MILK PARTICIPANT CONSENT FORM I jndtrttand thit Rntireh Triangla Innitutt ii angagad in * nudy of vinous organic compounds u thtv aopaar in hunan milk. I undantand th*t tha wrvty ii baing eonduettd in onitr to maBura th» Itvili of various organic compciundi in human milk, and it limittd to tha purpota natad. I hirthar undamtnd thit thi lurvtv '« baing conductad undar tht autpica* Of thi Unitld Stitu EnvironmantaJ Protaetion Agtnev in emmmion with [Kaae of Local Agency] • I do htnfav fncly eonunt to piriieiott* in thit nudy of ergviie compound* in human milk ind undtnond thai my airbcipttion will contiit of eroviding miwtn to a sutiuonnain ralawd to tnvironmtntai txpotun and providing a milk tampia of aopraximtttlv 100 ml. I undtrnand that an agtnt of Rmareh Tnanolt Initituta will acmm ttar tha auaitionnaira and collact tfw milk tampia, aftar which I will nMtint an incannva of hva dollara for my participation. I unaarnand thai my na.1"* will not bt voluntarily ditdoiad. or raltrrad to in any way whan compiling and avaluating tni nrauln of th* nxdy. I undarnand tfiat oarticipation in tfiii ttudy may rawlt in no dirtet banafia to rrn. om«r tnan tnou datcribad harain. and that I am fraa to wi^draw from thii nudy at any tima. It hat bnn axpiainad to m« that than ara no lignificant nik: to ma from participation in thii ttudy. I funtiar undarnand that whila panicioating in thi itudv I will b* frat to atk any quattiont eoncaming ma nudy: if I hava any funnar Ouanioru about trw projaet, I know that I am fra« to contact or Mr Banjamin S. K Harrit, III, Survay Ooarationi Canttr, Haaarch Triangla Innituta, Raiaarch Triangli Park. Norm Carolina Z770S, vlaphona numbtr 91S-S*1-605S. One . , " • , i . ' " Ifinml u D SIGNATURES: ISotti Htimetr mt Html Karl Oaal 94 TTTI t£p Can; �STUDY OF ORGANIC COMPOUNDS IN HUMAN MILK Conduced by: *nmtra\ ~rmrq* Inuai* Off to e/ To«« SUHWMM IfwRuinwitM PmnctKNi tyitt 2MW _ RnHn« Trian«i« fw*. Norti CmliiM 2T708 PARTICIPANT LISTING FORM I on 4i* c NOT1CI!: All i ta K«d in wict eenfldinei. mil « MM only by gmoni rmit Mnnfladon «f HI indn^uw or m •DDil«uiMni mil in md tor ow purpoMi IORM (or *• mrtv. m) will net M tfiKlOMtl Of (VWUMl TO OCflVT PWtt Of UMd "Of OAV OCflOf D 95 �COMMENTS 96 �OMINO. 1U-S7W Aoorewi txeirn StmrnMt 19 STUDY OF ORGANIC COMPOUNDS IN HUMAN MILK Iponmtf by. Conam.ua by: <MiM of Ton tmiromwnul Prannion A^aiQ NMAinfton. O.C. «M«0 ft«Mrah Triwigto Imtaiu fttMKft Triwifli Nrt. Nam Cwaiw 27709 QUESTIONNAIRE THE RESEARCH TRIANGLE INSTITUTE OF RESEARCH TMIANCLE PARK. NORTH CAROLINA. IS UNDERTAKING A RESEARCH STUDY FOR THE U.S. ENVIRONMENTAL PROTECTION AGENCY OF LEVELS OF VARIOUS ORGANIC COMPOUNDS IN HUMAN MILK. THE INFORMATION RECORDED IN THIS QUESTIONNAIRE WILL BE HELD IN STRICT CONFIDENCE AND WILL BE USED SOLELY FOR RESEARCH INTO THE EFFECTS OF ENVIRONMENTAL FACTORS ON PUBLIC HEALTH. ALL RESULTS WILL BE SUMMARIZED FOR GROUPS OF PEOPLE; NO INFORMATION ABOUT INDIVIDUAL PERSONS WILL BE RELEASED WITHOUT THE CONSENT OF THE INDIVIDUAL. THIS QUESTIONNAIRE IS AUTHORIZED BY LAW (P.L. 94-469). WHILE YOU ARE NOT REQUIRED TO RESPOND. YOUR COOPERATION IS NEEDED TO MAKE THE RESULTS OF THIS SURVEY COMPREHENSIVE. ACCURATE. AND TIMELY. Stu*.r nuflitar: • D Stanumbtn 97 Prttetpim number: �Fint, I would Ilk* to «k torn* gutmi qu*ftiom obout you. 1. Mm: QHWWI. [[]j f"*™^ I ™ fl|»'«*-""*( p. E "'•«••>«*•''•'« B I I Hlventa *ri*ifi I m-m-cn I l«(« 2. Whn «•» your •(• in y«n K IBM hirtm»y7 1 Whn • your htightr | j |ircn« | | DID- n j | I ureuld like to uk lomt quottiem about your oeeupnien. B. An vou priMRily *mDlov*ri IK Wf cmehy? | ' | Vot ICamtnni \ t ] No <0e ta 0. 101 7. How long hw*you bion unploy** Bv your prownt unployir? [__JU|»K» LlJt>*v' LiJMDmn uJ1>Mr> (. Doo) your eecupttion iou*liy t>k* you M*V from horn*? jjj Voi ICotmnutl \ * |No/6«io 0. Ill I. flhf. a Vw nnurt *nd loe*>.?n Ittrxt iddr»«l of *<• compmy for whWi you wort? 1C. If net prmntlv •mpley*a. whteh of «u foUONin) b*n d*Briboi your n*B»7 HOUHWlt* ICc u a »» 11. Whn atwm youruwd occuorton? tSe*aty> . 12. Art von prmnoy •mplt^Bl in thto <xcuORien7 [ < [Y» a No 13. If y«i to ibon ouMOon, haw long hm you bun vnoloyid In Ml i PT~| 1 I I I" "" lCh«e*«oteri . Do you work n of in «ny of *• following oeeupni I ' I 'wrong [_SJ Dry dwung ITIrwL_J°"" [T • I Serve*ration/pr*j»/«nj>««r»p*» PnralMii pteni hirnlBira r«f hiMikii or rap** 98 EV- �tteit I would Ilk* to «k urn* quarternragudlno,your hctltti wd pononol tabta. 11 Oo you motor Ifc H Voi ICum^MJ I {TJNo»o«af« HQW old won you Mritvn you flnn icmtd 17. On On oiorin. haw m«ny cltirvRM do you ««oti oor ON' QJ LM ttun M port (M «praMd QJ About 1H p«la I2M4 c*nml [jj About » OMk 15-14 lifBuart Qj About J p«ki O8-49 clpnmi) I pock I1H4 iljatnoJ NCTh QMor(*in2poela(SQarfflara*itnttHl /f *w ttneiptn tarn Mawco /« »<n» MMr ftm feoXr MM 11 Win it «M n«raai numew o< houn am you omd out ol doon forii doy? 19. Haun Mowfflmyhaun of *• diy. M *• »mi|lL do you namully OMI* unv from Kamnf Mmnaf •**»!•» «x n»i«unt Mount ~" [ | ~"l ». H.*-"* ( f| Wan do you eoniov «M eumm lonii ol your koil«7 ffiktv* ancj QfaortomQ Hood Qj »>* 21. Ait you aurrwrty «Ui«j ony »i«u»iloa modkabodUl on * rojulv <My br^tf J2. M««you t^w*nvnon«>«BrioitoniMdlador«kl«M»on4«lH)un7 33. Oo you puma yaufoMipir X. 8L Voi How mony *)gi h*x you «on * M pon < Oo you punuo «iy of tn MHo«Hn| hobela? tttac* o« otn ojp^rJ |»umHuioro)lnWiini XL Q] Yoi | « | PUnant Q] Q] leoto mod^i Oo you purojo ony onMiy «in Indudoi ropilor ino of tolvtm gluo or « [JJYoi Q]"*o 99 QJ Voi Q Q No �L*rty, I would Ilk* to mk lomt quortc™ (bout your rodrianc* md houMhoJd. 27. HIM mony y«n hw* you «•* <" «* "»^ I I ~| Vnn a. Now lor* hm yen IM n your cwnm oddraorf 9. DC TOM ceo! your norm with Mr of Ow hXlc«rln» •ollvotf ffikM* •» Ultr •Pft'J \ \ | Unta [_jj ten [* I » | Window *Jr oondltKmrli) | » | C*imt oihoyn hnUI |_jj Do AM know [T] EvoBomw* MOtorta) jjjj Other f* Q Cirwtatlni Mi) Don your heu«*old ore* ony of la OMI food In a hem, pr4«n? Qj V« 31. Do mn know Q WK« dew v«uf heu»HcW etoin fmti tntt milt, MoratHM? Ootdtrl 12. Whit • 9li pnmw mm 0f ytur nmr tar dnnkinf? [»JT«p.i [TJ Tip • municipil wppi, n [Tj »c If no. how *o« h dlttorr «3iMBnV/ Whit a Bit prirMry mm* of your «•»» for cooking? [ * I Top • munlekHl wpery }& | * | Top - DnVoo wdl Com xvon. MM m your imMMd motor If ytr. duct til Mff MPV/ M. Qj Voi | * | Oo not [JJ No Qj Oo not know JjJ QoHitai Qj OpMl [JJ p|p,| «] OOw t&atityl Oo» onyom •• In your houMhaW work • ony of th« followini oceuoKurK/buvnMir*? lOuet off Kwr oppryj Q feimlni [ * | Chomal ptam [_^1 Dry elo»n| | * | Nmlowii plwn 17. [_*J Oa Ml know li that th« ovnt prinury mm ft xnor nx dnnk mi»« ueri • eoffoi. iov KookAid. me' Q]VII 34. | « | T«s> • prMB Mil [_»J fonriei n»don/oii»oi/>noino npoir ( * | •vmnur* wflnWilnH or n»lr Do* invent *IM if your ttaafheU pumn ony of «MtoMowmthobbW /Due* *» «Ht*&Y.> r~i r—I r—i IJJ.rHirnhun nftnMiki| I » I la RESPONDENT/INTERVIEWER INFORMATION M. Movendonc [j] [_»J »« 41. I0*rl lYmrl en - CD - nn I 100 �SAMPLI INFORMATION 43. W» i n*k wnpi* mUmuit M V« I I No «. ,,^-» en-nn-en —• Won*/ 44. On *«Md » I (Dnt -m-m COMMENTS 101 Houn : Minuw I : �Attachment 2 Research Triangle Institute Data Collection Agreement 102 �for Research Triangle Institute DATA. COLLECTION Project Mo. AGREEMENT X, . agree to provide, as in employee of Poverforee Company, Inc., field data collection services for Research Triangle Institute in connection with the project named above. a. X agree to provide services within the guideline* aod •pacifications for project data collection activities provided by Research Triangle Institute; b. X an aware that the research being conducted by the Institute is being perfomed under contractual arrangeaent with t c. X agree to treat as confidential all information secured during interviews or obtained la any project-related way during the period X aa providing services to the Institute; d. I shall at all tines recognize and protect the confidentiality of all information secured while providing ay services throughout the conduct of this research project; e. I aa aware that the survey instruments completed fora the-basis from which all the analysis will be dravn. and therefore agree that all work for which X subait invoices will be of high quality •and la accordance with project specifications; and f. X fully agree to conduct myself at all tlaet in a manner that will obtain the respect and confidence of all individuals from whom data will be collected and I will not betray this confidence by divulging information obtained to anyone other than authorized representatives of Research Triangle Institute. Dated at (City/Town) (State) thia . day of 19_ Employee For Research Triangle Institute Disposition: Original to KTX; yellow copy retained by Employee. 103 �APPENDIX B SAMPLING AND ANALYSIS OF VOLATILE ORGANICS IN MILK 104 �SAMPLING AND ANALYSIS OF VOLATILE ORGANICS IN MILK 1.0 Principle of the Method Volatile compounds are recovered from an aqueous or solid sample by warming the sample and purging helium over it. The vapors are then trapped on a Tenax cartridge which can be introduced by thermal desorption directly into the GC/HS for analysis. This protocol is the result of extensive development efforts.(1'9) 2.0 Range and Sensitivity For a typical organic compound approximately 30 ng is required to obtain mass spectral identification using high resolution gas capillary GC/MS analysis. Based on a 50 g milk sample, a detection limit of about 0.6 JJg/kg would be possible. The dynamic range (limit of detection to saturation on the mass spectrometer) for a purged sample is M.O ; however, smaller samples may be purged and the upper end of the range increased commensurately. 3.0 Interferences Two possible types of interferences must be considered: (1) material present in the sample which physically prevents the effective purge of the •ample, and (2) material which interferes with the analysis of the purged sample. In the former case, several techniques have been developed to . handle such problems (e.g., foaming) by diluting and stirring the sample. The second case is minimized by the use of GC/MS for the analysis, since unique combinations of m/£ and retention time can be selected for most compounds. This permits the evaluation of compounds even though chromatographic resolution is not obtained. 4.0 Precision and Accuracy The purge and trap technique has been evaluated for a variety of matrices using model compounds which are expected to be typical of volatile halogenated compounds. 105 �The recovery of the purge step was validated using cow's milk samples spiked w-'th 14C-chlorofonn, 14C-carbon tetrachloride, 14C-chlorobenzene and 14C-bromobenzene. The average recoveries were 88, 88, 63, and 35 percent, respectively. The recoveries correlate roughly with volatility (inversely with boiling point), so anticipated recovery for other compounds may be interpolated from these data. 5.0 Apparatus 5.1 Purge Apparatus The purge apparatus is shown in Figure 1. 5.2 Sampling Cartridges The sampling tubes are prepared by packing a 10"en long x 1.5-cm i.d. glass tube containing 6 cm of 35/60 mesh Tenax GC with glass wool in the ends 3} to provide support. (2 ' Virgin Tenax is extracted in a Soxhlet extractor for a minimum of 24 h with redistilled methanol and pentane prior to preparation (•> 3) of cartridge samples. ' After purification of the Tenax GC sorbent and drying in a vacuum oven at 100°C for 2-3 h all of the sorbent material is meshed to provide a 35/60 mesh-size range. Sample cartridges are then prepared and conditioned at 270°C with helium flow at 30 ml/min for 30 minutes. The fit conditioned cartridges are transferred to Kiroax (2.5 cm x 150 cm) culture tubes, immediately sealed using Teflon-lined caps, and cooled. This procedure is performed in order to avoid recontamination of the sorbent bed. (2' 3) 5.3 GC/MS/COMP The volatile halogenated hydrocarbons purged from water are analyzed on either an 1KB 2091 GC/MS with an LKB 2031 data system or a Varian HAT CH-7 GC/MS with « Varian 620/i data system. The sample, concentrated on a Tenax GC (2 4) cartridge, is thermally desorbed using an inlet manifold system. ' The operating conditions for the thermal desorption unit and the analysis Tenax GC cartridges lire given in Table 1. 6.0 Materials 6.1 Sampling Clean, 120 ml, wide-mouth glass bottles with Teflon-lined caps are used for the collection of milk samples. 106 �TENAX CARTRIDGE THERMOMETER -20tcl50°c HELIUM 'PURGE THERMOMETER ADAPTER -with 0-ring HELIUM INLET TUBE ¥ 10/18 LIQUID LEVEL 100 ml ROUND BOTTOM FLASK MAGNETIC STIRRING BAR Figure B-l. Diagram of headspace purge and trap .yatea. 107 �Table H-l. INSTRUMENTAL OPERATING CONDITIONS UC* 2091 Deaorptlon chamber temperature 265 IS *L/»ln 10 HlVmia Deaorption tlM 8.0 Bin 8.0 mla Capillary trap temperature during deeorptlon -196»C -196"C Temperature *f capillary trap darlag injection onto rolMui -196*C to 2SO*C - tbea held at 190*C Tiae of He flow through capillary trap 12 3/4 •• ! 12 3/4 ail. He flow through column laveep time] 00 270 Deaorptloa chamber Re flew o VarUn HAT CH-7 f .5 »tn 4 •• ! Carrier flow 2.0 BL/.ln 1.0 ml/mlo Capillary column 100 • SK-30 SCOT 20 • 8C-30 VCOT Column temperature 30*C for 2 »1«, then « / ! to 2*0* ••• 20 •» 240* at 4»/»ln Bcaa range 5-490 dalton 20 * 500 dalton Scan rate 2 aec full acale 1 aec/decade Scan cycle tl«* 2.4 aec 4.3 aec Scan • 4 oe parabolic eiponeatial Trap current 4A Fll«Bent current SO|iA 30i 0|A Accelerating Tolctage 3.5 kV IkV �6.2 Purge Teuax cartridges - 16-mm o.d. x 10.5 cm glass tubes filled with 6 cm of Tenax with 1-cm glass-wool plugs in each end. Charcoal cartridges - 16-mm o.d. x 6 cm filled with 4 on of charcoal and glans-wool plugs in each end. Glass culture tubes with Teflon-lined screw caps. 7.0 Procedure 7•1 Collection of Field Samples Milk (60-120 ml) is expressed directly into the wide-mouth bottle, capped tightly, and frozen for shipment and storage. To preserve the integrity with respect to volatiles, handling and transfer must be minimized. 7.2 Purging of Volatiles The apparatus is assembled as depicted in Figure 1, including the Tenax GC cartridges (1.5-cm diameter x 6.0-cm length). A carbon cartridge 1.5-cm diameter x 4.0-cm length is connected to the effluent end of the Tenax cartridge to prevent contamination of the cartridge by laboratory vapors. The milk sample is cooled to ~A°C, shaken vigorously and 100 ml diluted with 350 ml distilled water. The pH of the solution is adjusted to 4.0 with sulfuric acid. A glass-wool plug is inserted into the center neck of the flask just above the level of the solution and, with the flask in a heating mantle, the solution is heated to 70°C while it is stirred with a magnetic stirrer. The sample is purged at 15 ml helium/min and 70°C for 90 minutes. The loaded cartridge is removed and stored in a culture tube containing 1-2 g CaSO, desiccant for 2-12 h. The desiccant is removed from the culture tube and the dry, loaded cartridge stored at -20°C. 7.3 Analysis of Sample Purged on Cartridge The instrumental conditions for the analysis of volatile compounds of (2-9) the sorbent Tenax GC sampling cartridge are shown in Table 1. The thermal desorption chamber and six-port valve are maintained at 270°C and 200°C, respectively. The helium purge gas through the desorption chamber is adjusted to 15-20 mL/min. The nickel capillary trap at the inlet manifold is cooled with liquid nitrogen. In a typical thermal desorption cycle a sampling cartridge is placed in the preheated desorption chamber and helium gas is channeled through the cartridge to purge the vapors into the liquid 109 �nitrogen cooled nickel capillary trap. After desorption the six-port valve is rotated and the temperature on the capillary loop is rapidly raised; the carrier gas then introduces the vapors onto the high resolution GC column. The glass capillary column is temperature programmed from 20°C to 240°C at 4°/min and held at the upper limit for a minimum of 10 minutes. After all of the components have eluted from the capillary column, the analytical column is cooled to ambient temperature and the next sample is processed. 7.4 Quantitation All data are acquired in the full scan mode. Quantitation of the halogenateci compounds of interest is accomplished by utilizing selected ion plots (SIPs), which are plots of the intensity of specific ions (obtained from full t.can data) versus time. Using SIPs of ions characteristic of a given compound in conjunction with retention times permits quantitation of components of overlapping peaks. Two external standards, perfluorobenzene and perfluorotoluene, were added to each Tenax GC cartridge in known quantities just prior to analysis. In order to eliminate the need to construct complete calibration curves for each compound quantitated, the method of relative molar response (RMR) is used. In this method the relationship of the RMR of the unknown to the RMR of the standard is determined as follows: std where A std unk g GMW = = = = = - Auak/m°leSunk "A.td'"lM.td peak response of a selected ion, standard unknown number of grams present, and gram molecular weight. Thus, in the sample analyzed: . tAtok)(OWunk)(l.td) 110 �The value of an RMR is determined from at least three independent analyses of standards of accurately known concentration prepared using a gas permeation system. ' The precision of this method has been determined to be generally ±10 percent when replicate sampling cartridges are examined. 8.0 References 1. Michael, L. C., M. D. Erickson, S. P. Parks, and E. D. Pellizzari, Anal. Chem., 52, 1836-1841 ( 9 0 . 18) 2. Pellizzari, E. D., "Development of Analytical Techniques for Measuring Ambient Atmospheric Carcinogenic Vapors," Publication No. EPA-600/276-076, Contract No. 68-02-1228, 185 (November 1975). 3. Pellizzari, E. D., "Development of Analytical Techniques for Measuring Ambient Atmospheric Carcinogenic Vapors," EPA 600/2-75-075, 187, (November 1975). 4. Pellizzari, E. D., J. E. Bunch, R. E. Berkley and J. McRae, Anal. Chem., 48, 803 (1976). 5. Pellizzari, E. D., J. E. Bunch, B. H. Carpenter and E. Savicki, Environ. Sci. Tech., 9, 552 (1975). 6. Pellizzari, E. D., B. H. Carpenter, J. E. Bunch, and E. Sawicki, Environ. Sci. Tech., 9, 556 (1975). 7. Pellizzari, E. D., Quarterly Report No. 1, EPA Contract No. 68-02-2262, February, 1976. 8. Pellizzari, E. D., J. E. Bunch, R. E. Berkley and J. McRae, Anal. Lett., 9, 45 ( 9 6 . 17) 9. Pellizzari, E. D., Analysis of Organic Air Pollutants by Gas Chromatography and Mass Spectroscopy. EPA-600/2-79-057, 243 pp., March, 1979. Protocol Prepared, June, 1980 111 �APPENDIX C ANALYSIS OF SEMIVOLATILE ORGANIC COMPOUNDS IN MILK 112 �ANALYSIS OF SEMIVOLATILE ORGANIC COMPOUNDS IN MILK 1.0 Principle of the Method Milk samples are collected from nursing mothers and frozen until ready for analysis. An aliquot of the thawed sample is then extracted, cleaned up by Flori.sil column chromatography and analyzed by GC/MS/COMF. The extraction procedure used here is preferable to that used by the AOAC , since both polar and nonpolar compounds are extracted from the milk. Tie AOAC method is designed for pesticide residues and would not efficiently extract polar and/or acidic compounds. Opea column chromatography is a necessary prerequisite to GC/MS/COMF analysis. Although some loss of sample may occur during the extraction and cleanup, these procedures remove proteins and fats from the sample which would otherwise create overwhelming interferences for GC/MS/COMP analysis. Since the compounds of interest in these fractions cover such a broad range of volatilities, the GC/MS/COMF analysis can be rather complex. The higher PBBs of interest in the extracted fraction must be chromatographed on a very short column (45 cm x 0.2-cm i.d., 2 percent OV-101 on Gas-Chrom Q) at high temperatures to elute them as sharp peaks which may be identified and quantitated. These chromatographic conditions are not applicable to more volatile compounds since they are not resolved from the solvent. Thus, the extracted fraction is analyzed a second time using a nonpolar SCOT capillary column (either OV-101 or SE-30 liquid phase) to separate and identify semivolatile constituents (e.g. chlorobenzenes, PCNs, pesticides, etc.). The chromatographic conditions are typically 60°C initially, programmed to 2AO°C (or the column limit) at 6°/min. The: mass spectral data are stored on magnetic tape. The mass spectra of interest: will be printed out by the instrument operator for qualitative analysiii. Quantitation from this data may be achieved by integrating the area of selected ions and comparing them to the area of the external standard. 113 �The sensitivity of the determination may be significantly improved for quantitative purposes by using the technique of selected ion monitoring (SIM), also known as multiple ion detection (HID). This technique monitors up to 9 ions at a sensitivity 10-100 greater than the normal operating mode. This technique is used for quantitation of compounds in samples where the increased sensitivity is necessary for detection or accurate determination. 2.0 Range and Sensitivity The detection limit of the GC/MS/COMP system has been determined to be about 5*50 ng/pL for pesticides such as v-BHC, £,£(-DDE, atrazine, trifluralin and heptaculor using a 40 m SE-30 capillary column. When SIM was used, the detection limit was about one order of magnitude less (i.e., 0.5-5 ng/pL). The detection limit for tetrabromobiphenyl is about 1 ng/[.(L in the SIM mode using A5 x 0.2-cm i.d. column packed with 2 percent OV-101 coated on GasChrom Q. For an instrumental detection limit of 1 ng/pL, the overall sensitivity of the method should be about 6 ng/mL (6 ppb) milk assuming a 50 ml milk sample extracted and extract concentrated to 0.3 ml. This detection limit may be improved by using SIM and may be worsened by background interferences. 3.0 Precision and Accuracy When electron capture gas chromatography (GC/ECD) was used, the mean recoveries from cow's milk for seven replicates ranged from 57 to 93 percent for six model compounds. Thus, the results obtained may be as little as half the actual amount in the sample. The relative standard deviations (RSD) for the above replicates ranged from 11 to 33 percent, with the average RSD at 21.7 percent. Thus the precision of the method is about + 20 percent. It is anticipated that accuracy and precision will improve with experience with the method. 4.0 Apparatus 4.1 Gas Chromatograph 3 A Fisher-Victoreen 4400 gas Chromatograph with an H electron capture detector, a 10 AFS electrometer, and a 1.0 mV recorder is used. 4.2 Gas Chromatography Column For most compounds, separation is achieved using « 40 m SCOT glass capillary column coated with 1 percent SE-30 and 0.32 percent Tullanox. For 114 �the compounds of very low volatility (e.g. the higher PBBs) which will not chromatograph'on the capillary column, a 45- x 0.2-cm i.d. glass column packed with 2 percent OV-101 on Gas-Chrom Q is used. 4.3 Liquid Chromatography Column A 24-mm i.d. glass column with a Teflon stopcock is used. 4.4 Gas Chromatography/Mass Spectrometer An LKB 2091 gas chromatograph/mass spectrometer with 2 PDF 11/4 computer is used. The system is equipped with a glass jet separator and is used with either glass capillary or packed glass column. 5.0 Materials Kuderna-Danish evaporators: 5 ml receivers 250 ml KD flasks Snyder columns 500 mL flat-bottom boiling flasks 250 ml separately funnels Clean glass wool Whatman 1 P/S .filter paper Florisil Sodium sulfate (anhydrous) Acetone "Distilled in Glass", redistilled Pentane "Distilled in Glass", redistilled Toluene "Distilled in Glass", redistilled Ethyl ether "Distilled in Glass" 6.0 Procedure 6.1 Extraction (1) Mix 50 mL (or volume available up to 50 ml) of a milk sample with clean glass wool and 150 ml of acetone to precipitate the proteins. (2) Decant and filter the acetone/water layer. (3) Repeat steps 1 and 2 with two 50 ml acetone fractions. (4) Concentrate to about 20 mL using a Kuderna-Danish evaporator. (5) Extract the precipitate with 40 mL of toluene; decant and filter the toluene layer. 115 �(6) (7) (8) (9) (10) Combine the toluene extract and the acetone extract with shaking. Let the layers separate and draw off toluene (top) layer. Repeat Steps 5-7 with 40 ml toluene and then with 10-20 mL toluene. Discard the lower water layer. Dry the organic layer with anhydrous sodium sulfate and concentrate to desired volume using a flat-bottom boiling flask and Snyder column. Quantatively transfer to a vial and concentrate to 5-10 mL under a gentle stream of nitrogen. 6.2 FlqrisrLl Column Chromatography (1) Prepare Florisil by beating to 130°C for at least 5 hours. (2) Prepare a 24-mm i.d. column so that the Florisil is 10 cm high after settling. (3) Place about 1 cm of anhydrous sodium sulfate on top of the Florisil. (4) Rinse column with 40-50 mL pentane, never allowing the solvent to go below the Na.SO, layer, as channeling may result. (5) Add up to 10 mL of sample to column. (6) Elute with 200 mL of 6 percent ethyl ether/pentane solution at <5 mL/min. (7) Collect and concentrate in a Kuderna-Danish evaporator. (8) Evaporate under nitrogen stream to 1 1.5 mL. Quantitatively transfer to a vial, store in a freezer. (9) If sample solidifies after concentration, repeat the Florisil cleanup (Steps 1-8). 6.3 Standards Standards are spiked into the sample following the extraction and workup (d..-pyrene was used at 200 ng/mL). 6.4 Analysis 6.4.1 GC/MS/COMP Analysis for Semivolatiles Inject 0.2 yL onto a 40 n SE-30 SCOT capillary at 60°C initially, program at 6°/min to 240°C, then hold until no more peaks are observed. Collect mass spectral data at 2 sec/scan from m/z 20-500. Compounds amenable to this analysis include organic compounds with volatility lower than that for purgeable compounds. Only the very low volatile compounds (e.g. higher PBBs) will not elute from the capillary. 116 �6.4.2 GC/MS/COMP Analysis for Low Volatile Compounds 6.4.2.1 Normal Procedure Inject 1.0 pL onto a 45 x 0.2-cm i.d. glass column packed with 2 percent OV-101 on GasChrom Q at 220°C initially, program to 300° at 12°/min and hold until all peaks have eluted. A helium flow rate of 20 mL/min is used. The mass spectrometer is scanned from m/z 20-1000 at 2 sec/scan. 6.4.2.2 Alternate Procedure Using the same chromatographic conditions analyze the sample by SIM. Preselect up to 8 ions characteristic of the compound(s) of interest and one ion characteristic of the standard. Retention times provide qualitative identifications. Peak areas may be used for quantification as discussed below. This alternate procedure has 10-100 times better sensitivity than the full scan mode and provides faster quantitative results. The main disadvantage is that only preselected compounds may be identified. In addition, if specific halogenated compounds are found to be present with little interference in most samples, they may be analyzed by GC/ECD. This procedure improves the sensitivity and reduces the analysis time (since GC/MS/COMP requires an offline data output). If GC/ECD is used, approximately 10 percent of the analyses are verified by GC/MS/COMP. 6.4.3 Qualitative Data Interpretation Spectra are interpreted by visual comparison with standard spectral (2 3) reference collectionsv ' where possible. Where standard spectra are not available, tentative identifications are made based upon interpretation of the mass spectrum. Where possible, the GC retention time is also used to assist in the identification procedure. All identifications and interpretations are checked independently by other experienced chemists or spectroscopists to assure that the interpretations are correct. 6.4.4 Quantitative Analysis In order to eliminate the need to construct complete calibration curves for each compound to be quantified, the method of relative molar response (RMR) is used. Successful use of this method requires information on the exact amount of standard added and the relationship of RMR (unknown) to the RMR (standards). In general, the RMR for a compound is determined for a 117 �characteristic ion (parent or fragment) in its mass spectrum. The integrated ion current may also be used, but is generally less precise. The value of RMR is determined from at least three independent analyses. The method of calculation is as follows: A ./moles unknown/standard " Astd./moles std. . .. A = peak area, determined by integration or triangulation of the total ion current or for a selected mass of each compound . "/U ». A = peak area, as above g = number of grams present GMW = grain molecular weight Thus, in the sample analyzed: 7.0 References 1. Horowitz, W., ed., AOAC Methods of Analysis, 12th ed., Association of Official Analytical Chemists, Washington, DC. (1975). 2. McLafferty, F. W. , £. Stenhagen, and S. Abrahammson, ed., "Registry of Mass Spectral Data," John Wiley and Sons, New York (1974). 3. Eight Peak Index of Mass Spectra. Vol. I (Tables 1 and 2) and II (Table 3), Mass Spectrometry Data Centre, AWRE, Aldennaston, Reading, RG74PR, UK (1970). Protocol Prepared, June, 1980 118 �APPENDIX D VOLATILE COMPOUNDS IDENTIFIED IN SELECTED PURGES OF MOTHER'S MILK 119 �Table D-l. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 1081 (Bayonne, KJ) Chromltoiraphlc Faak Mo. Uiitiaa fnp. C'C) CUr o»» tographic PftAk. MOa QMBOUBd riuttoB Taap. CO CnayBUBd 41 150 £-»ct«. 42A 152 421 132 tatracblBroatbylaM CjE16 laoaar (tut.) CgE16 lanMT (MBt.) •llauaa C.E, , iaoaar (MBt.) 8 16 ehlorobman* 1-ehlorohaicaBa (cant.) •thylb.««. STlaaa laoMr r.s carbon dloxlda cblorocriflneroa* chu* propylaa* C^Uomr (•6 C 43 44 154 4 5 6 f,7 C4Bj iaOMt 45 159 V3 acauldabjrda 46 161 74 71 •13 aeatma 47 74 48 49 168 SO 171 3»hapf BTMriiw J II r imrnift SI 52 171 1-haptBBBB* 173 acTTU* .13 93 87 t ri cUoraf luenv* cbw* c.-paBtaoa iaeprapaaol •atbylua cblorida fnoa 113 carbon dlaulfida B-bucaaal cyclopaatam 163 166 89 91 92 94 14 U 1 1 58 (1 2 3 8 9 76 77 10 79 11 .to 12 13 14 IS 16 17 4*10 *"•" S34 531 S3C 173 173 174 C,B,0. iaOMT S3D 174 xylaaa laeatr Mtbyl athyl batBaa 54 173 C CjEjj laoatr 55 56 178 10H22 *«— * £-BOBaBa 179 C 57 181 584 183 184 laopropjlbaacaBa 581 S9 604 188 189 C 601 189 10E16 1"*'r CjEj.O laoMt (taut.) 18 19 95 20 21 96 97 22 99 234 102 102 104 btufluorobcntao* (1st. atd.) D-bauB* ehlarefoi* C?EK IkoMt C£BU ICOMT parfluetotaluana <iot. atd.) MebylcyelepaBcana 1.1.1-criehleroatbaaa 231 156 C E 9 16 lg—r CjEjp iaoMr £-haptaul <t nt > ' - 10E22 lt<mtr S-vatbTl-l-lBdobutaaa C 10E22 t"m'r 11B24 tK»" C 61A 191 baualdahrda 25 26 27 105 CjE^ i»am»r 611 191 B^propjl baaaaaa 108 112 twBMB* 62 193 194 C.-a.lk]tl baasaaa 28A :i3 ttbyl Tiajrl kctoat 195 C9B18 lao«T 281 a4 2-paataDaaa 29 30 31A )13 M6 C.B..O <tact.) B.p.ocu.1 65 66 67 196 197 C 11E24 l>0—r oecaaoBa laoacr 199 C :.i9 CricblBroacbylana 68 69A 200 201 691 70 202 203 204 24 311 32 33 34 35 36 37 38 39 40 119 63 M crelebaua* .L22 C E 7 12 ot C6E8° U-" B_-baptaaa .126 C E L29 134 B 16 *>olnl C E Ue 714 r 7 J4 ~ 1-ehletepaBtaBa 135 72 73A tBlUBBB 145 147 C£E^jO laoaar (Mat.) £-baxaoal C E t t 8 16 «« 205 206 210 731 73C 74 unknown 138 143 711 210 210 211 - CoBtlBIMd - 120 CjEj0 taaavr (tint.) 11E24 lM~r 2-paBt7lf«raB C 11E24 ttol"r g-octanal alleiau C 10E22 t'01"r dieblorobaataBa C 11E2* t'°-r 10E14 1K>-r CjElt ifomtr (cast.) C ant. hydroearboa aat. hydrocarbon �Table D-l (cont'd.) Chreuto- SluKion Tctip. (raphlc P**k No. CO 75 76 77* 77B 78 ' 79A 798 SO 81 82 S3 84 to 212 21.1 21.1 21ft 2111 Co-pound !.. !. ». • t hydrocarbon •. imMt. hydrocarbon •onochlorod*caM (t«nt.) C 2! 1) 22,'L 22:! 224 22'i 22'» 230 9»18° tc*tophnoM •*t. hydrocarbon •at. hydrocarbon 2-000*000* dlMthylatyrm* B-geaaaal n.-undicau Chrouto- Uutloo irapaic T««p. Puk No. CO 86 87 88 89 90 91 92 93 94 93 96 97 121 240 240 240 240 240 240 240 240 240 240 240 240 Compound mat. hydrocarbon • l U l O M naphtha l*n* C H ( 10 20°Uo-r " ' ") £-dod*can* uakaovn unaat. hydrocarbon •lloiana CUH22 laoMr alloun* ankaowa •iloxaa* �M 1 Sf........V.........fc ^Jw^M***-,^^ i i . *........r.....i " i i Figure D-l. Total ion current chromatogram from GC/MS analysis for volatiles in sample no. 1081 (Bayonne, NJ). �Table D-2. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 1040 (Bayonne, NJ) Cbrouco- t.utioo •r«p. iraphlc Fuk Me. 1 2 38 3 . 60 4 5A 67 74 59 SB 74 5C 75 SO 73 77 6A 61 78 6C 78 60 6£ 79 79 7 81 8 82 84 9 10 1 1 85 87 12 90 13 14 92 94 13 16 96 97 17 98 18 19 101 104 20A 201 107 106 21 109 22 110 23 11 1 24A 113 24B 25A Covpouad jrtphle P«ak No. Mrbaa dteaida caleratrlfluoroajatbaaa dlMtbyl ataar C.H.0 iWMr laopaatMa trlehlarofluoreajathaa* •eatoaa CjH10 laaaar n-paataaa laepraaa laoprapaaal 23B 26 27A 115 117 28B 121 28A 123 124 127 28B 29 30 31 32 33 130 132 136 138 1*0 COllMO* 35A 141 1-paataael 351 142 imkaown 36 145 37 146 C;H16 laoMr ^phtfMUuil 38 149 ISO 39A 39B 40A 40B 41 42A 42B 42C 131 152 153 153 154 154 154 43 155 44A 44B 137 157 45 161 46A 162 162 46B 47 163 48 165 49 167 30 6«12 "" parfluoroteluaM (lat. >td.) 1,1.1-trlchlemthaaa 3-Mthylbutaaal (tut.) 2-Mthylbutaaal bauaat c«rboa tatraealaxidi cyclobuun* 169 51 52A Uo 173 174 52B 173 53A 113 113 115 Compound 34 Vl2 Uo~r vtnylldlaa eblerida avtbylana cfalorid* ftaoa 113 earboa diaulflda 2-vataylprepanal eyelopaatana unknown Mthyl athyl kacoaa C.tL, laoaar buaflueTobaaiaa* (int. atd.) £-haxaaa calerefoni C Uutloa Taap. fC) 175 176 331 34 177 35 36 179 57 181 182 58A C H 7 1* •tbyl vlayl katoa* 2-p«nt«noo« Tlnyl proplaaaM (taat.) tziebloreatbylaaa 581 S9A 59B 181 183 184 185 60 191 63 66 67A aluam " - Caatlauad - 123 •at. hydrocarbon oaaat. hydrocarbon C>H16 iaeawr CgB14 laoMi ailouaa uaaat. hydrocarbon aat. hydrocarbon oBMt. hydrocarbon aakaewa ebletohtcan* athylb«uaaa *ylaat iaoawr 2-bap canon* •tyt*o* 2-£-bucylfuran (tant.) a^haptaaal sylaaa taoaar C9B18 laoaur CjBjQ laoatar •at. hydrocarbon 1 r Vis "~ 3-*aehyl-l-ladebutaa« C,HI8 l.o~r taeptopylbaaxto* aat. hydrocarbon hydrocarbon 190 63 64 7 12 " W oakaoim C7814 laoa*r CjH14 laoMr dlMtnyl diauldda 1-ealenpaataa* 189 61 62 C H 8^1 & laHMaT unknown C|B1& taoMt CjH18 UoMt tr«n»-*-oet«n« tatraehleTo«thyl«a« 190 192 194 196 196 uaaat. hydrocarbon baasaldabyda £-propylb«nt«n« (tant.) triaathylbrataaa iaoaur laoaayl format* (taat.) aakaevB �Table D-2 (cont'd.) ChroMceiraphlc ptak No. UuCioo T«*p. CO traphle raak to. Covpouad 671 197 aat. hydrocarbon •4 68A 198 681 69 70 " 199 200 CgEjg laoawr C,-«lkyl bkniaoa 85 86 aat. hydrocarbon 2-pantyl furan 67 71 72 201 203 203 204 H •9 Cj-alkjrl bcuttM C 90A 90» H 1C 20 7i 207 •llnuna dlchlorobanaana Cj-alkyl bmaaM (cent.) 76 209 C H 77 76 211 212 •aothaaa (taot.) 79 80 213 216 81 82 216 217 E3 219 73 74 206 91 92 93 94 95 96 97 98 99 100 S 14 A IAC cbylccby HVCOXWM ivoMtr llaoDaoa C1,E,2 iaowtr uaaat. hydrocarbon *at. hydrocarbon •unknown 124 butlra C-pc^ <%' 220 222 223 22} 226 228 230 231 234 239 unknown aeatopbaaona aat. hydrocarbon C 10>22 i " " " 1 diBathylatyran* 240 n-nonanal alloxan* alloxana tatraMthylbauao* (tant.) •llouna •lloxaaa napbtbaluta 240 C 240 240 240 240 240 240 12E26 i"a*t unknown allouna 2-nndaeaaoDa C 13E28 alloxana �I1 I'. Il Figure D-2. I»T-» T • > • i ••• i iTTf • i »T»y«^'i^Tr> ••• rn Total Ion current chromatogram from GC/MS analysis for volatiles In sample no. 1040 (Bayonne, HJ). �Table »-3. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 1107 (Jersey City, KJ) ChroBate- Elutiea iraphlc lap. Paak No. CC) Chr ovate- Uutlea graphic tap. Peak Ho. * ^ 113 114 Yloyl proplonat* o-puntanal 116 117 C 221 22C 118 22D 118 23 24 120 1 64 SMM 201 2 65 67 carbon dloxld* fraoa 22 dlchlorodlfluoroMtbua £-propant but«Di laoBtr ja-butana 21 22* 3* 31 4 67 69 54 56 70 SC 72 36 6A 7 73 74 8* 71 75 76 81 8C 78 78 9A 79 91 10* 79 81 10B 83 IOC 85 06 86 26B 125 127 26C 127 C B 27 28 128 129 29 131 8 16 1"°**r dlMtbyl diaulfld* dlhydropyran cblorop«BCaoa 30* 134 tOllMO* 30B 137 31 aeataldahydt butwa l*oMr chlorocthan* tctraMtbjlaUau trichlerof luoroai than* 1-pantao* accent* laopropanol «~p«itan. Mttyrlu* cbloild* Tram 113 7«14 *— « tt 1 chl oro» chjlaoa E-dloian. •tbyl furan (cunt.) B-hnptaot 2.2,4- criM tbyl-l-pcBtaoa lietwuiul CtEloO l.o.r 4-Mcb7l-2~p«Bt«noBa 139 141 25 26* 32* 321 123 124 lor 11* in 86 carbon dKulfldt (me*) •atbyl »tnyl Uton* (tract) Bttbyl propwol nitrowtlkut (teat.) 88 cyclopcs taa« 89 12* 90 35 36 121 91 13* 92 2-Mtbyl putiaa vlajrl aeatata a-butaoal 3-«*cbyl pantant 131 14* 93 94 C B 381 156 38C 39* 156 IOC 101 141 97 14C 98 100 15 164 101 161 102 16C 102 17* 104 171 17C IB* 105 106 181 19* 191 19C 19D 20* 108 109 110 110 11 1 12 1 112 Conpoiart 33* 331 34 143 146 147 148 149 37 38* 6 12 *«»" parfluorobuima (int. ltd.) v-basu* eblorofeiB dibydrofuraa Utrabydrofuran pcrfluororoluaM (1st. ttd.) MCbylcyclopaatana £-»ttbyl acacnld* 1,1.1-trlchlerMtbaB* 3. 3-diB»tbylo»ta» (Mot.) buuaa carbOB tctracblorld* 1-buunol cyclobaian. Cjlj^pO ivowr •tbjl vlnjl kctoM (teat.) 151 154 156 391 158 159 40* 160 401 161 40C 41* 161 411 163 164 42* 421 43 44* 441 44C AS* 451 46 2-p*BUBOB« - Cootlauad - 126 162 165 166 167 168 168 169 170 172 Vie uo-r C6E120 iaoMr £-huanal Vj6 i.*r ji-octana C E 8 16 t>OBtr tatrachloro«tb>l«n« Vl6 i*°**t •lloxaa* inkBCVB CjHjj Ifomn cblorobaaiao* j-bcxaaal (teat.) ehlorabuBJU C B 7 12° U""r •tbyl bm*»« CjEJ8 iMMr 4-hcptaBOBi ryl.n, IMMT BbuylacatylaBt J-hupcanon* 2-hcptnaoB* C^jO (tut.) •tyraw B-hapiaBal xylana I»OB«T •at. byrtroearkoo C.1L| laomr ..-«««« �Table D-3 (cont'd.) Chroaata- (lutlon Ta». iraphic I*C) 47 48 49* 49B 49C 49D SO* SOB SI S2A 52B 32C S3 54 55* S5B 5« 57 58* S8B 59 60* 608 61* 61B 62 63* 63B 63C 63D 64* 64B 64C 64D 64E 65 66* 668 66C 66D 47* 67B 47C 68* «8B 173 174 175 17S 176 176 176 177 178 181 182 182 184 186 187 187 187 188 190 190 192 193 193 194 194 193 196 196 197 198 200 200 201 201 201 202 203 203 204 204 205 206 206 207 207 (raphlc PMk Mo. CoapOUOd •at. hydrocarbon C10iJO laoMr aat. hydrocarbon •thyl aathyl cyclobaaaaa 69* 69B 70 71* 71B 72* 7 IB 73* 73B 73C 74 75* 7SB 76* 76B 77 78* 78B 79* 79B 80 81 82* 82B 83 84* 84B 84C BQk&OWB C7H1()0 laaan laoprepyl baaaana C H 10 221*OB<r CjH140 laoaar (tact.) tzana-2-baptanal a-plaraa busaldohydo nrpropylbauua xylana laoawr aat. hydrocarbon C 10»22 Uo*tr bauonitrlla (tract) •at. hydrocarbon phenol triaatbylbauaaa pantyl turaa £-octaaal baniofuran trfcatthylbaaxaaa laoMra C 10B20t ° t *•lluaaa C,,B_ _fl ^/^±0 o*4ccuM dichl orobtnxnt llutlon Tan. (*C) 208 209 210210 211 211 212 212 212 213 213 214 215 215 216 217 218 219 220 220 221 222 223 224 224 225 226 226 CoBnooad C- l , b . « 4.klu. C ll»22lM-r 11H221 ~ " r C l0 ll»22 < -* phthalld* (tut.) •at. hydrocarbon dacalla (taat.) aat. hydrocarbon C C 11B24Uo—r C^-alkylbauu* laoaar 2-flonanona C 11H22Uo— * C4-alkyl banaana iaowr Mt. hydrocarbon £-ooaaaal C 11822 Uo-« 10»120 »— « n^vadacana allosana C C 11H22i ° r '" r 10B18 «*" C^-alkylbanxaaa laoaar ClJHj6 laosar CUH:4 iaoMr 2-atthyldacalln (tant.) C C 12»26 C.-alfcylba»ana lioacr C^-alkylbonaana iaoaar 396 •Av 86* 86B 86C 86D 86E 86F 86C 87 88 89 90 91 92 93 94 • 5 C 11H22 U<"*r onkaown trl»tthyl baaiaoo iaontar unknovn C^-alkylbantana •at. hydrocarbon C 11H23 * * *" aat. hydrocarbon llaoa*n« CjjHjj iaoa«r •a thyl ityrana Mt. hydrocarbon C H 11 22 * " *• dlathylbutana iaoatr •at. hydrocarbon acatophanona • Coatlnatd - 127 228 228 229 229 229 230 230 MO 230 230 230 230 230 MO 230 230 C 12H24 * " ^ 11H20UOMI(««e') C H 12 24 llo~r C H 10 12° " " " C B 10 18°1>OMI unknovn C C 11H16 Uc~r •llosana •at. hydrocarbon •at. hydrocarbon •at. hydrocarbon aat. hydrocarbon aaphthalana unaat. hydrocarbon jg-dodacana aat. hydrocarbon �Table D-3 (cont'd.) Chrouto§«phic Pt»k No. Chroiatogrmphle Puk Ho. ElUtlOB fc^d Co' Elution Tap. CO CCMBDOUBd unkaovn 105 230 2-CrldKIAOM •it. hydroorbon 106 107 230 230 •»t. hydrocirboo 230 unknovB 108 230 vllouni 230 230 • llOXUM 230 230 phth»l»te •it. hydrocarbon 109 110 unknovn 11 1 230 dil»obutyr«t« itoacr diphuyl tthur 112 230 C 96 97 230 230 •lloun* 98 99 230 100 101 102 230 103 101. 230 230 2-undocuon* Mt. hydrocarbon 128 Ittctoni ivoBtr (tint.) 14R22° 1>OMr �m rt-fr* Figure D-3. (-rrvrr> *-vf--rr*T<-»r-.^f < • • » > ! < • F T I - T T I . £""•* rr^fwr »•• py,,^.....^ ^,....,^,™ Total Ion current chroroatogram from GC/MS analysis for volatiles in sample no. 1107. (Jersey City, NJ). �Table D-4. VOLATILE COMPOUNDS IDENTIFIED IN PITRGE OF SAMPLE NO. 1115 (Jersey City, NJ) ChroM tographic Ptak No. Elutiet CbroMto- Clutlon iraphic . Ta«p. Paak Ha. CO Co^d ro' CiapouBd 1A 62 carbon dioxide 24 120 1 1 63 xanon (traca) 25 122 2 SA' 65 (7 26 27 122 124 31 68 carbonyl aulflda (taat.) chleroBathane unknown 28 126 chlorepantant ORknovn 4A 41 76 trlchlorofluoroMtbkBa acataot 29A 291 128 129 1-paatanol SA 77 131 31 134 4-*etbyl-2-pKntanom s-buinal 6A 78 80 laopaataaa laopropaaal 30 SI •athylaat ehlorlda 32A 61 81 FraoB 113 321 136 137 furaldnhydt (taut.) (tract) 6C 60 62 carbon diaulilda (traca) unknown 33 138 Bj-octant 140 tatrachlorotthyjiant 7 unknown cyclopancaM 34C 140 141 dlchloroproptnt (tract) BA 83 86 34A 341 81 87 avthyl laoprep;? katont BC 89 B-butaaal ISA 331 142 76 82 142 9 90 1-baxana (tant.) 36 143 IDA 92 101 HA 92 94 htxanuorobauBae (int. atd.) Bj-baunt 37A 371 146 147 148 94 chloroform (tract) •ethyl furan 3BA 11 1 381 149 12 13 96 uaaat. hydrocarbon 39 151 98 parfluorotBluana (1st. atd.) 40 14A 99 erotonildahydt (taat.) 41A 411 151 152 141 100 14C 15 16A 100 102 104 161 103 16C 103 17 106 107 18A 181 1 . 1 . 1-triehloroa thant 3-Mtbylbutaul 42A 421 2-BttbylbvtauU (cast.) bansent 42C 43A 431 carbon tatraehlortda (traca) 1-butanol (tant.) unknown 44A 441 1S2 133 153 155 155 156 157 C E 7 J* Uo"*r dlMthyldlaulfidt dlhydropyran toluaoa C.B., iaavar B 10 unknown C H II 5 8 2 CjE^j laimr ailouut 2-haxan«l ehlorobaniant C,E,4 iaoatr 5-Mtbyl-3-hydrofuraD-2-ont (tant.) e-furfuryl alcohol athylba&tant C0Blg iioacr C^MjO (cant.) sylant laomtr phtBylMKCyltnt 5~Mthyl-3-btunoBt 2-b«pttnnM C B 7 12° we athyl vinyl katont 2-ptntanont 158 158 SH20 <trmct) atyrao* b-htpttntl 19 107 JOB vinyl propieaatt 4*0 159 ryltn* Iroacr 20A 201 20C 109 110 11C g-pantanal 45 CjRlg iaomar aat. hydrocarbon ••thylbexaat (tant.) (traca) 46 47 159 160 162 21A 111 11: l-basana 48 163 lodopantant trlehloroathylaaa athylfuraa (taat.) 49 SO 166 unknown 170 trani'2-baptanal 2.5-diMthylfuran arbaptaaa CjBj laoaar unknown S1A 511 171 172 baataldthydi SIC 172 173 17* 211 21C 22A 112 221 22C 23A 231 1* 1 15 1 116 117 11* SID S2A CjEglj (tant.) (tract) -Continute- 130 2-furyl attbyl kttont (tant.) B-nonant S-Mtbyl»2-fnrfura,l onknowB v-propy Ibaniant nylana laovtr �Table D-4 (cont'd.) Chromato- Ilutloa fraphic Tap. Ftak No. CO 52B 175 32C 52D 175 52E 53A 175 176 176 53B 176 53C 176 53D 177 53E 177 177 53F 54 178 55 56 180 57A 181 57B 58 182 182 182 182 59 60 184 184 61 62A 187 188 188 57C 370 62B 62C 63A 63B 63C 180 188 190 190 191 64 192 65A 192 65B 193 66A 66B 194 195 67 196 68 196 197 69 70A 198 70B 198 Chroaa to- Elation graphic lap. Ftak No. CO COBOOUaal 71A banaonUrUa occaoont 711 71C C 10H22 Cj-alkylbtaxtna l-chloro-3-*thylbaiat (tat. ) dibroBodlchloroMtbana (tat.) phenol •at. hydrocarbon 5-«thyl-3-haptaaou (tat.) unknown 6-**thyl-2-btptanoat patyl furan o^oc canal btnzofuran (tract) Cj-alkylbtnaaa 7U> 72 73 199 200 200 200 201 204 74 212 75 76A US 214 76B 77 215 78 US U6 79A 220 79B 80 220 221 81 223 82 Conound 2-nonanona diaithylttyrat (tract) C4-alkylb«nitat (tract) C H 10 16° 1*°-r a-nonanal undacana unaat. hydrocarbon lu 18 n—pantylbtnaana •Uoiant •at. hydrocarbon 2-dtcanont napfathalaat C12H22 Iwmtr n^dacanal n^dodacant •at. hydrocarbon unknown mathyl etanoltat (tut.) (tract) lactont iao*«r (tant.) oxyimatad hydrocarbon phanyl baxaaa Cfyf iao-r •ilosant g-dacana dichlorobtncat 83A 225 226 83B 84 227 228 C H 9 16 ' C^-alkylbaxaa 85 231 86 233 87 237 88 238 89 239 90 240 240 =10=16° 91A 91B 240 allozan* 92 240 93 94 240 93 240 anaat. hydrocarbon •at. hydrocarbon 2.2,4- trlaathylpenta-1, 3-dlol dl-laobutyrata (BKC) aat. hydrocarbon 96 240 C 97 98 240 240 14H30 lB81wr unaat. hydrocarbon aat. hydrocarbon 99 100 240 240 C H 13 32 *— « •at. hydrocarbon C H 10 20 U r llaonant — l.B-clMolt CjgBjj (tract) anaat. hydrocarbon •at. bydrocarboA acatophanoaa »-butylbaiua (tant.) C^BgOj (ttnc.) CUB22 lacnar unknown C 10B18 lgo->r aat. hydrocarbon 131 240 unknown nndacana <M "-) �Figure D-4. Total ion current chromatogram from GC/MS analysis for volatiles in sample no. 1115 (Jersey City, NJ). �Table D-5. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 2048 (Pittsburgh, PA) Chroma to- Uutlw graphic tap. •uk No. 1A 1 1 2 3' AA 41 SA SI 6A 61 7A 71 8 9A 91 10 11A 11 1 12A 121 13 14A 141 ISA 1S1 16 17 18A 181 iac 19 20A 201 21 22 23 24A 2*1 25 26 27 28 29 30 31 32 carboa dlodda 58 51 64 66 70 70 71 72 73 74 77 77 79 S3 83 84 87 87 89 89 91 96 96 96 98 102 104 106 107 109 109 112 112 113 119 126 126 127 130 131 133 134 136 138 140 143 Chroma tographic Faak *». Compound 33A 331 34 35 36A 361 37 38 39A 398 39C 390 40 41 42A 421 42C 43A 431 Uo r S«12 trlehlorof luornma thi oa aeatooa X-putaaa laopropuol fraon 113 •athylaaa chlorlda carboa dlaulfida CjHj,, l.o«r ' Vu *-r* CjHjpO iaomar (tut.) macbyl athyl tauoa • C ta 6"l2 "" ha»fluorebaa*ua (int. ltd.) a-baxana chloroform pmfluorocoluaaa (Int. atd.) •atbyHarcloputan. 1.1,1-WlcblaTMChaa* 1-butaaol (tue.) baasaoa eyclebaxaaa C 145 146 147 149 153 154 156 159 161 161 162 162 164 165 166 167 167 168 168 169 170 173 175 177 177 181 182 183 184 184 44 U 6«12 "" CjHjoO 1— r C 6«10 *—" C-pound ?o' 43 46 47 48A 481 49 50A SOI 51A 311 ehlerocrlfluoramathaaa C4H, iaomar C^Hyj laomtr aeataldahyda C UutlOB SIC 52A CjHjj iaomar •nknown 321 S3 54 55 56A 561 57 58 S9 60A 601 61 62A 186 186 187 189 190 190 192 192 19* 19* 195 195 197 198 11 1 C 7»l*° lie"r •lloiaaa Vl2° 1<c-r ohlorobaotaoa (tract) cblotoheiana (traea) athylbanm* •at. bydrocarboo •ylaaa laomtr onkaom C,H20 laa«r 3-baptaoaoa 2-baptanooa atyraaa CjHjj Iaomar (taat.) uc. hydrocarbon pbaptanal xylaaa IIOMT g-aoaaaa C10H,0 Lour CI()HJJ laomar (tant.) laopropylbanitoa C.0ljj laomtr Culj4 laomar C 10*16 1"m" CgHjjO laomar unaat. hydxecarboa busaldabyda !£ • AO* CTl CblOiTOV Cb vl V*M tatrachloroachylaoa CjH16 Iaomar D ' MPf^PT^ JAM S! ** 6«12° *«»k uaaat. bydrocarboa cblerottaacan* •aaat. hydrocarboa (cut.) tolaaaa 1-pmataaol C V, Uo-' Vl2° *— ' pbuanal C,lli taomar t-octaaa 421 133 C H 10 16 *•"" C.-aikyl baaaarc laomtr •at. hydrocarboa oaaat. bydrocarboa C ll"24 U"" U Vl6° t —r 10H22 "*" C ll«24 *— « 2-f«ntylfuraa C11H24 laomar (tut.) C3-a4kylban*«na laomtr 6 C lAfl * " " ' alloxaaa •at. bydrocarboa dleblorobaaiaM �Table D-5 (continued) Chronato- Elutlm iraphic Peak No. «3A 200 63B 6i 200 202 65 66A 66B 206 206 203 67 208 6S 69 209 211 70 212 71 72 213 214 73 74A 74B 215 216 217 75A 218 75E 76 77 220 222 219 ChroHto- Slutlon grapuie FcaV No. CoBpoimd (*o' unaat. hydrocarbon tat. hydrocarbon (teat.) uoaat. hydrocarbon 2-*chyl-l-baxanol 82 B5A 235 ilaonan* C10B180 Un*r •at. hydrocarbon (teat.) •at. hydrocarbon C4-alkylb«nt«n» acatophaooaa •at. hydrocarbon •at. hydrocarbon •at. hydrocarbon C^-alkylbanxnvt KB 236 11B20 1M-r 10B1«° ***» •HezaBii CIOHlgO UOMT 85C 236 C 86 238 240 83 84 87 88A BOA 80S 227 81 229 240 89B 240 C C ic'l4° i*0ftr unaat. hydrocarbon •at. hydrocarbon oapbthalau C1()E220 i.o«r (..«.) 95 96 240 n~t*rpln»ol (tut.) unaat. hydrocarbon .B-dodacMt* ailtntani unaat. hydrocarbon •lloxana 2-undacanonn ailoxan* 240 C 97 240 96 99 240 100 240 101 240 90 240 92 93 94 134 240 91 •at. hydrocarbon 79 240 240 89A C0Ei5° iaoaar dlMthylatyrcna aat. hydrocarbon p-nocaaa.1 223 226 226 233 8BB B-uodtcao. •iloxane C^-al ky Ibanxana C^-alkylbaiuna unknown 78 231 232 ^unaat. hydrocarbon 240 240 240 240 13B28 i"mrT •iloxaot dacanotc acid (tant.) C K » 30 Uol"r unaat. hydrocarbon alloxana �JL i Figure D-5. i I I I I 7 I Total ion current chromatogram from GC/MS analysis for volatlles in sample no, 2048 (Pittsburgh, PA). �Table D-6. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 2071 (Pittsburgh, PA) ChroMto- Eluciao traphlc Tap. Peek Da. CC) 1 59 u 60 (1 34 31 4A 63 64 41 65 3 66 68 6 71 7 8A 73 81 73 9A 75 91 9C 10 75 11* 78 11 1 11C 12 79 Chreaate- KlutUn (raphle Tea*. Feak *o. CO CaqxMBd 33 344 348 35 36 37 38 39 carbon diexld* propylen* (tract) dichlorodlfluoreaBtbaae (trace) dla*thjrldlfluor<MllaM leobutane C.E- laoaer b-bucaae (trace) acetaldefajde chlerocchaB* (trace) a* Chanel aeecwe triehloretluoTnaa f h»rie leopropaael a-peatane 40 41 42 43 44 45A 116 18 1 19 1 122 124 126 127 129 133 138 139 141 144 cE 451 CjHjj laoatr •ethylent chlcrUe 2-«etbyl-2-propai>ol Freon 113 46 145 146 147 47* 149 471 149 48 152 81 82 C B 6 » carboa dliulfldt 49 c H 501 14 15A 83 85 86 153 158 159 . a-propanol (cent.) cyclopentane 51A 160 511 160 C H 32A C H 521 162 163 164 13* 131 76 77 80 151 87 16 87 17 88 18 19 89 90 20 91 21 93 22 23* 231 24 95 96 2iA 100 251 26 100 101 27A 271 102 28 29 30A Ml 31 32 j e *"°»»r 504 « e° 6 12 * •" 6 14 U-" vinyl aeetace • £-butaaal kethyl ethyl katoae 53 54 33 56 Ue r 166 167 baxafluenbeuaiie (lac. acd.) 37 169 170 94 b-bCXUM 58A 173 94 •Cbyl acacate chlerofera 581 174 176 177 179 103 106 107 108 108 11 1 13 1 Via ~ 59 (0 IMrfluorocelueee (Int. atd.) ••thrlcyelope&cane t^ijt laoaar 1 ,1 , } -trleblerea thane CjB^ijO laoavr (tent.) Baaiena carbon tecraehlerld* (trace) fbutaaol (cent.) CTclobasaae •ethyl propjrl keton* e-peacaul (1 (2A (21 63* (31 (4 (5 (t (7A (71 MA 136 C-p-d erlchleroethyleui ^-heptane C l r C7"l4 t—I e"i6 '°~ diaechyl dlaulflde taikamn C.BU laoaaT (tent.) toluene dlbrevoehloreMtbaae (trace) a-bexanal C(Elt lecwer £.oetaae tecrachleroetbrUne CgEu Inner (tent.) ttnkwmn uneac. hydrocarbon •lloxane CfBia laoMr echylbenxeo* C.K.. leoewr xylen* iaoavr ph«oyl*eetylen» 2-heptanona etyren* SjrlaB* laevrr B-heptanal C*TI«~' iaapropjrlbeuent 10 22 **°**t C 10E20 tB™*r 180 180 baualdenyde £-prerj.ylben»en« 182 182 184 183 185 C 186 186 187 10B16 *—r 3 . C E 10 22 i90^T heucoBltrUe Mtbylheptanane laoacr a-Mthrlir^raiM trie, thy Ibenien* leoMr �Table D-6 (continued) •ChroMto- Uutloa triphlc Yap. P«ak No. 688 188 698 70A L88 190 708 70C 190 70D 71 191 192 72A 192 728 72C 193 72D 190 193 194 73 74 194 7SA 196 197 196 758 76 78 79A 798 80 205 205 207 81 208 82 210 . C H Uo 211 211 85 86 212 214 87 215 88 216 89 222 90 223 225 227 91 92 93 94A 230 948 232 95 96 235 97 10H16 lM-r <"«•> 10H16 1'°-r Ct-«lkylb*ni«n* UmotMtM unknown Mt. hydrocarbon acatophcnon* 203 Coapomid 84 C 201 Xlutlon 13 C 199 77 iraphic Puk Mo. Mt. hydrocarbon •thyl n^caproac* putylfuraa (e*nt.) baniofuran (dot.) C4-alkylMU«M trlMthylbnum* ia«*r phenol (txae<) •ilouo* C1QH22 laoMr dlchlorob*u«M unknown C1(JHI4 laomr Mt. hydrocarbon 187 691 Compound 240 240 98A 231 239 988 240 99 240 100 240 101 10 16 ~ •at. hydrocarbon unknown 2-nonanoM 240 102 r 240 240 103 104 137 240 dlJMthylacyrm* Mt. hydrocarbon canncn* (ctnt.) •iloxan* •at. hydrocarbon ••thyl eaprylac* alloxan* caaphor C.-fl, .0 (tract) (tent.) 1U 10 •lloxana trlchlorob*niw< (tract) •thyl caprylat* naphthaline £-dod«can* wait, hydrocarbon (ttnt.) •Uouat 2-und*caoon« Mt. hydrocarbon •at. hydrocarbon ••thyl dccanoac* illoxan* C, /.H,n (ttnc.) 14 30 •thyl dacanoat* uoMt. hydrocarbon �H 00 101 J|_A_A_jA_JL_JV It 1 k 1 k i Figure D-6. Total Ion current chromatogram from CC/MS analysis for volatiles in sample no. 2071 (Pittsburgh, PA). �Table D-7. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 3053 (Baton Rouge, LA) ChroMte- Elucloo iriphle T«p. f«*k No. ("0 u 11 37 2. 3 Cbrom«Cotripble Ho. 62 enrben dioxlda chlototrlfluoroMthaaa cbloroMthaM 63 e 38 4 68 SA 71 JC 72 6 74 76 81 77 9 10 79 80 11 12 86 144 C 8«16 £-oct«n« 38 145 C 8 39 85 141 37 73 7 SA 139 36 4«10 * * * " dlMthyldlfluoroillM* 136 33 70 31 X£L 34 147 •nto furaa 40 149 41 150 42A 153 |,-propanol ••thylaM chlorlda frwn 113 urbon dlaulfld* (traca) CjHjO 421 153 43A 153 431 156 44 158 6 10° fuMldibydc Vu v« C H 6 10° 8 14 C H 6 10° nknowi C H 14A 43C U 459 161 C 46 164 C 47 6 14 °iMxafluorobauaaa (Int. ltd.) 2-aathylfutaa 161 45E r 160 165 Vl6 166 481 88 159 48A 87 13B 139 451 13A 43A ybutu*l •tchyl *thyl katona (tut.) C 8 166 UA 90 90 90 92 151 93 48C 94 49* 169 491 170 171 97 30 172 18 19 20 98 8»18 167 16 17A ((•at.) a-furfuryl alcohol 141 14C 21 22A 97 100 parfluorotoliMo* (Int. ltd.) C4HjO laoMT . 1.1,1-trlehloroathan* 5U 106 JOS C 173 32 104 175 53 6»12 •tbyi 176 34 kateo* tMMT 173 311 7«a4( 178 221 108 33 180 23 109 36 181 24 110 ppaatvul 37* 182 25A 113 £ B 371 182 251 1.13 trlenlorocthylcn* 38 184 ••tbylfurcldchyd* l*e»er 2SC M4 C 39* 26A 391 187 C 261 ,U6 117 39C 188 C 27 120 60 189 28 122 m 29 30 31 32 33 .123 126 133 134 InOHAlT 0 6>8 •eitle «cid 2-rlnylf (tut.) (tint.) MCb7lfur«ld«07da IWMT 10820 10B22 61 191 621 191 C 63* 192 C 631 193 64 dlbydropfMB (tut.) tola 190 62* 194 coatlowtd - 139 ll»22 U«24 (int.) �Table D-7 (continued) Chrouto- Elutiou Ta«p. fraphtc Fiak No. CC) 65A 194 651 66 67A 194 671 68 69 70 71 72A 196 197 198 199 201 202 204 204 204 72E 73A 207 73E 74A 207 208 74B 209 75 210 211 76 77A 77E 77C 77D 78A 78£ 79 80 81 82 83 84 212 212 213 213 214 215 217 Chroaato- Klutlon (raphic T«q>Faak Mo. (C *) CoBpound 871 C.-alkyl bwutM l«e*ar unknown ailoxatw £-d*can* 88A 888 89 10E10°2("Bt-> 91 237 C H 0 r 9 16 U " C^-alkylbcn»Dt (trat.) 92A 238 921 B 6°21 ~ r '' li»on*n« 93A 239 240 unut. bydroearbon C H 931 240 aat. hydrocarbon 94A 948 240 C H 12 22i"MT £-d*eaoal un»at. hydrocarbon c E i r 95 96 n 24 — aat. hydrocarbon acitophanoni C -alkylbancua C B 1 97 98A 981 r 11 24 ' " °11E24i ° ' 'unMt. hydrocarbon aat. hydrocarbon C5Hg02 i«-r 99 C 100 H U < C 240 240 C 240 240 240 240 105 106 C naphthalant (tract) 240 240 103 104 r 12E26 Uo"er 12E26 1W~r C E 1C 20°l«"*r C C 102 C/E0 * I»OB*T (t*nt.) .., D C C 240 240 101 •at. hydrocarbon 240 240 240 240 12E24 ilol*r n.-dodaeaD« 13E28 ilHmI •at. hydrocarbon 13E26 itn*T 11E20°i"»er C E 13 28 Uo~r C 13H28itml C 13E28 lBO"r C B 1 1 °Uo-« 0 6 C K 13 24 tM~r C £-undacanal n-crldceaoi C E Uoi r IO I«° - 107 240 108 240 240 •iloxa&c unaat. hydrocarbon unaat. bydroearbon 240 240 n.-tttrad»cant 109 110 C o r 12E26 U " 228 229 C •at. hydrocarbon 90 226 227 87* 12E24 t§0"tr •lloxan* dlchlerobanxana unaat. hydrocarbon 221 222 224 85 86 234 C 235 236 11 24 ~ 10E16°UC~r n-Donanal B-uadtcan* unaat. hydrocarbon aat. hydrocarbon 218 231 234 COBDOIud 111 112 C 12H261 r •lloxaiw— 113 140 240 240 B.-dod»caaal unaat. hydrocarbon n.-p*otad*cana �*» \iil uJj*K v Figure D-7. I fc'T k' r k fc ' '1 i1 uuu 1"" fcill Total Ion current chroraatogram from GC/MS analysis for volatlles In sample no. 3053 (Baton Rouge, LA). �Table D-8. VOLATILE COMPOUNDS IDENTIFIED IN PURGE OF SAMPLE NO. 3111 (Baton Rouge, LA) Chrocato- Elucioo iraphle Ta»p. Faak No. CC) 1 2 3*. 31 4 5* 51 6* 61 6C 7* 71 8 9 10* 101 11 12* 121 13 14 15 16* 161 17* 171 16 19 20 21 22 23 24 25 26 27 26 29 30 3U 311 32 33* 59 61 65 65 71 7J 74 76 76 77 80 81 82 84 87 88 89 91 92 94 95 96 101 101 104 104 106 108 109 110 11 1 112 114 117 120 123 126 128 135 142 144 146 148 Chrovato- tlution Ta-p. CO COBpOUBd tapond raak *>. carbon dioxlda dlchloredlfluoroMthana •ulfur aioilda C^H, laoMT 331 34 35 36* 361 37 38* 381 39* 391 C}KIO i(o>ar trichaoroflueroaatbana acatona liopropanol £-panta&* CjEg iacwr •athylua chloriM rraen 113 carbon dlaulfida 2,-buuul cyclopaotana C6H14 laoa»r CjEJflO t«».r to 41 42 43* 431 44 43 C H 3 10° 1—" CjE12 Uowr haxafluorebauau (int. atd.) 46 B-btSUI 49 chloroform parfluoretoluMMi (int. atd.) •aihjrl.eyclopan.tana 1.1,1-trichloroathana CjE100 taomar (tut.) 30 47 48 51 52 53 54* 541 55* 551 56 57 58 59 60 61* 611 62 63 64 65 66 67 »6 C E 6 12° i>0—r ' ban* ana carbon tatrachlorida 1 r '."12 «~ CCE120 iaeawt (tut.) C£EI20 laoatr (tut.) £-paBt»al trichloroathylan. ft-bapcana C{EU iMwr C E 7 14 U-" dlaatbyl diaulflda teluaaa 5-ba«anal C E 8 16 *—T D-octana tatrachloroa thylnna 142 148 150 152 155 155 161 163 164 168 168 169 170 173 177 177 178 179 181 183 186 189 18S 191 192 193 194 193 196 197 198 202 204 206 208 212 213 214 217 221 233 240 240 240 uoaat. hydrocarbon C.B.. laoa*r (taut.) •iloxana CyEjg IMMT C$B20 iaoawr (tut.) •thylbaasaaa xylan* iaeaar C E 9 20 lM™r atyrana C E 9 20 U<mtr xylao* laomr C9E2Q iaoa*r CjBjp Umur Ht. hydrocarbon C,-alkyl bauana (t«nt.) C 10E22 i§c*tr 1CH22 UoMr •at. hydrocarbon •llouna baualdahydt ueknovn C c iAt 1>OI*r Cj-alkyl bauan* C 11E24 1>on" 11E24 U—r C E 11 24 1*OI*r C,-alkyl baniana ailosaaa C C 11E2* 1>cmr dicblorobauana Cj-alkyl banian* linoiiaaa aat.. hydrocarbon Ht. hydrocarbon •eatepbanoDa Ht. hydrocarbon •at. hydrocarbon aat. hydrocarbon B-imdacaoa •lleuoa c-dodacant anaat. hydrocarbon •tlesaaa �Figure D-8. Total Ion current chroroatogram from GC/MS analysis for volatiles In sample no. 3111 (Baton Rouge, LA). �APPENDIX E SEMIVOLATILE COMPOUNDS IDENTIFIED IN SELECTED EXTRACTS OF MOTHER'S MILK 144 �Table E-l. SEMIVOLATILE COMPOUNDS IDENTIFIED IN EXTRACT OF SAMPLE 1032 (Bayonne, NJ) Chrouto- Uucioa •raphlc Taap. Puk Ho. CC) U IS 2 3" 4 S 6 7 8 9 10 lit 11B 12 13 14 13 16 17 18 19 20 21 22 23 24 Chrouta- Uutlon graphic TBBD. Peak Mo. CC) Compound telum* 01*aa laoMr Ciaifniiiwl 23 26 27 28 29 30 •lloun* •lloun* •lloun* •lloun* •lloua* •lloua* •lloua* dlaatbylblpbuyl (teat.) unknown • t *nd un**t. •. unknown •lloun* •llexaa* ) d10-pyr*n* (ltd . hydrocarbon* 31 •lloun* DOE •lua loa 32 33 34A 341 unknown •lloun* 31 unknown unknown 36 37 38 39 40 41 42 43 44 41 46 47 48 •at. and unaac. •lloua* •at. hydrocarbon •lloun* •lloua* •at. hydrocarbon •at. and onaat. hydrocarbon* •ilffun-t •llosana •lloua* •at. hydrocarbon pathalac* (cant.) •lloua* •at. and uoaat. hydrocarbon* 145 unknown •llnua* hydrocarbon* •lloua* •at. aad uuat. hydrocarbon* Mt. and imut. hydrocarbon* •lloua* •lloun* •lloun* •lloun* •lloun* •lloun* lyeop*ra*a* ehol*at*ryl acatat* •lloun* �8 B S "I" wu - I, n ri K. Figure E-l. Total ion current chrotnatogram from GC/MS analysis for Semlvolatiles in sample 1032 (Bayonne, NJ). �Table E-2. Chroaato- Cluclon graphic Tamp. raak No. CO 1 2 3 4 s 6 7 t 9 10 11 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 26 27 SEMIVOLATILE COMPOUNDS IDENTIFIED IN EXTRACT OF SAMPLE 2121 (Pittsburgh, PA) Chrouto- Uutlon graphic tap. raak Mo. CO Coopouad taluaaa alloxaaa alloxana allouaa allouaa 2 ,6-dl-tir c-buty 1-4-M chylphanol •athyl dodacaaoata athyl butyrata (tant.) alloxana Mt. hydrocarbon •lloxana alloxaoa MC. hydrocarbon •iloxana alloxana alloxana MC. and unaat. hydrocarbona oat. hydrocarbon unknown alloxana aat. and uaaat. hydrocarbona nnkaovn unknown alloxaoa alloxana d10-pyrana (Int. acd.) alloxana 147 28 29A 29B 30 31 32 33 34 33 36 37 38 39 40 41* 41B 42 43 44 4S 46 47 48 49 30 SI Cua^inimil uaaat. hydrocarbon taaat. hydrocarbon »E MC. and uaaat. hydroearbona alloxana paacacbloroblpbanyl aat. and unaat. hydrocarbona alloxana •at. and unaat. hydrocarbona haxachlorobiphanyl •lloxana •at. hydrocarbon alloxana aat. and unaat. hydrocarbon* •at* and uaaat. hydrocarbooa haptaohloroblphanyl alloxana Mt. aad unaac. hydrocarbona alloxaaa •lloxana •lloxana alloxana •lloxana lycoparaana alloxaaa ebolaacaryl acatata �8 Figure E-2. Total ion current chromatogram from GC/MS analysis for Bemivolatiles in sample 2121 (Pittsburgh, PA)- �Table E-3. SEMIVOLATILE COMPOUNDS IDENTIFIED IN EXTRACT OF SAMPLE 3095 (Baton Rouge, LA) Chroma CO- ilucioa (raphlc tap. F«ak Mo. (*C) 1 t 3 4 " 3 6 7 J 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27A 271 28 29 30 31 Chrouce- ilucioa iraphle Top. P**k Mo. (*C) Cnapoumi 32 33* 333 34 35 36* 361 37A 37> 38 39 40 41 42 43 44 45 46A 46B 47 48 49 50A 5C1 51 52 53A 533 54 55 56 57 Mtbylcn* chlorid* tolwn* •llcxan* •ac. hydrocarbon •ac. hydrocarbon (tanc.) illoxua aac. hydrocarbon (cane.) . •lloxan* •ac. hydrocarbon (cane.) •iloxaa* •ac. hydrocarbon •at. hydrocarbon unknown unknown MC. hydrocarbon •lloxan* MC. hydrocarbon •Iloxaa* •ilexint MC. hydrocarbon aac. hydrocarbon•lloxan* •lloxan* •lloxan* MC. hydrocarbon •llosan* •ac. hydrocarbon unaac. hydrocarbon unknown unknown •iloxaa* •lloxan* 149 ConpiiuDd «10-W«« •ac. hydrocarbon un*ac. hydrocarbon alloxan* DDE unknown unaac. hydrocarbon •llexan* unknown MC. hydrocarbon (cane.) •lloxan* uaaac. hydrocarbon (cane.) alloxan* •ac. hydrocarbon (cane.) •lloxaea •ac. hydrocarbon MC. hydrocarbon •ac. hydrocarbon •lloxan* •lloxan* •ac. hydrocarbon •Iloxaa* (ttnt.) •Iloxaa* •ac. hydrocarbon •ac. hydrocarbon lycop*ra«n* •lloxan* chol**csryl acccac* •lloxaA* •at. hydrocarbon ODJuown •iloxan* �Figure E-3. Total Ion current chromatogram from GC/MS analyaJs for semlvolatilea in sample 3095 (Baton Rouge, LA). �Table E:-4. SEMIVOLATILE COMPOUNDS IDENTIFIED IN EXTRACT OF SAMPLE 4093 (Charleston, WV) Chroaato- Clutlon (raphle Ta»p. Paak Mo. CO 1 2 3 4A41 5 6 7 8 9A 9B 10 11 12 13 14 15 16 17 IB 19 20 21 22 23A 23B 24 26 27 28 29 Co^-d Chroaato- Uutloa graphic Tamp. Faak Ho. CO tolaaoa •iloxana •llooa* •lloua* •at. hydrocarbon •lloxaoa •lloxana 30 31 32 33 3* 35 36 37A 371 38 39 40 41 42A 421 43 44 43 46 47 48 •9 50 SI 52 butyric arhydrlda (tut.) •at. hydrocarbon CjHju laoaar onknovn •Uoxana •at. hydrocarbon •at. hydrocarbon •lloxana •lloxana Mt. hydrocarbon Mt. hydrocarbon Mt. hydrocarbon •at. hydrocarbon unknown •lloua* Mt. hydrocarbon •at. and unaat. hydrocarbon •lloxana Mt. and mwat. hydrocarbon* •lloxana S3 54 56 57 58 59 MkBonn Mt. and unaat. hydrocarbon* •at. and iBaat. hydrocarbon* 151 Coapoood •lloua* •iloxana d1Q-pyrana (Int. ltd.) Mt. and unaat. hydrocarbon* •iloun* Mt. and unaat. hydrocarbon* Mt. and uuac. hydrocarbon* Mt. and uuat. hydrocarbon* DDE Mt. and unaac. hydrocarbon* •lloxana •lloxana •at. and uaMt. hydrocarbon* •lloxan* Mthyl d*hydtoabl*tata (tant.) •lloxaoa Mt. hydrocarbon •lloxan* ut. and unaac. hydrocarbon* •lloua* •hthalata •Uoxana •nkaowa •lloxana •Uoxana •Uoxan* lyeoparMna •lloxana cbolaataryl acatact Mt. and unaat. hydrocarbona •llouna tt-tacopb*rol (rltaxdn) �59 s tn NJ s 10 u. • _ * g - & o I' u o Figure E-4. Total ion current chromatogram from GC/MS analysis .for semlvolatiles in sample A093 (Charleston, WV). �UNITED STATES DEPARTMENT OF COMMERCE National Taohnloal Information Sarvloa 5285 Port Royal Road Springfield, Virginia 221 61 OFFICE OF THE DIRECTOR £>' *^s . . Doar ProfosTsional: I wish I could tell you how much time and money the NTIS Abstract. Newsletters could save you. But, I can't. Because I don't know "now much time and money you unknowingly spend on costly research that actually duplicates the work of others. Yet this research information is readily available to you when you know how to access it. That's what NTIS Abstract Newsletters are all about. NTIS Abstract Newsletters Summarize 70,000 New Reports Annually You see, NTIS .information specialists index, analyze, and categorize 70,000 new research documents a year. Each report is judged for its potential value For thousands of current subscribers. Then it is summarized in a brief abstract and published in one or more Newsletters. You receive only those abstracts oE interest to your professional field. NTIS exists to help you find—and provide you with— research results and information. You'll know what you need to know—when you need tc know it—with NTIS Abstract Newsletters. �The Newsletters arc a speedy, comprehensive, economical way for you to keep current on the latest research and to cut your risk of duplicating work already available in a report from NTIS. Your Newsletters will suggest possible directions and starting points for your own research. Knowing research applications of others will help you get more out of every research dollar you invest. Worldwide Sources The NTIS collection also contains selected information oa scientific and technical developments around the world. NTIS is now actively expanding its sources of foreign technology to add to more than 300,000 foreign reports already on hand—including translated materials from the U.S.S.R., other Eastern European countries, Japan, and major Western countries. Low Subscription Rate By now, you may well be apprehensive over the cost of this remarkable Newsletter service. If so, you will bo pleasantly surprised to learn you can subscribe to the Enorgy Abstract Newsletter, for example, for about $1.83 a weckl NTIS is a unique self-supporting Government agency offering economical information products and services on a costrocovery basis. You Get Reports on Research Not Published By Any Other Source NTIS Abstract Newsletters let you know about all Government-sponsored research in your field. Each week ycu'll learn about research not reported or available from any other source. And when you spot a report you need, NTIS can provide it for you in a matter of days, usually for about $14 in paper copy or only $4.50 in microfiche. Your Newsletters contain a handy report order form for your convenience. Annual Subject Index Included Your subscription includes an Annual Subject Index to give permanent reference value to your Newsletter. ! II!) (iOVUKKMKNTl'UINTINtlOl'I'ICK. IWJ - 'IflL-OO') �Of course, your nubscriptiori is tax deductible as a business expense in most instances. More than that, when you consider the wasted time and costly mistakes the Newsletters; can help you avoid, you will begin to appreciate what a practical investment NTIS Abstract Newsletters really arc. For example: • MANUFACTURING TECHNOLOGY: one of the newest Newsletters that meets industry's most pressing need right now...brings news of the latent research on robotics and CAD/CAM. $125* • FOREIGN TECHNOLOGY: abstracts of latest technological and scientific research in all fields from Japan, United Kingdom, France, U.S.S.R., and countries all over the world. SPECIAL CHARTER RATE: $75* • ENVIRONMENTAL POLLUTION S CONTROL: one of the most sought after Newsletters covering air, noise, solid wastes, water, pesticides, radiation, health and safety and environmental impact statements. $95* • ENERGY: brings you a wealth of information on energy use; supply and demand; power and heat generation; energy policies, regulations and studies; engines and fuels, and much more. $95* • COMPUTERS, CONTROL, AND INFORMATION THEORY: important information, and lots of it, concerning this, one of the fastest growing and most competitive industries worldwide. $95* • MATERIAL SCIENCES: applicable to nearly every industry with abstracts touching on a huge variety of processes and products. $80* • MEDICINE & BIOLOGY: covers an enormous range of information from anatomy to zoology. Every facet of medicine and biology is touched upon. $75* And any one of the abstracts in these Newsletters may be the added element of success for your ongoing research, contract proposals, cost reduction efforts, and profits. NTIS regularly publishes information not available elsewhere, and frequently brings you news of research before it appears in professional journals and magazines. You simply should not risk acting without benefit of complete, current information on what's happening in Government-sponsored research in your field. *0ne-year subscription rate (Over, please) �You need NTIS Abstract Newsletters and here's how to order them. Complete the order form at the bottom of this letter and mail it back today to the address below. You may use any of the options listed in order to pay for your subscription. You are guaranteed satisfaction. If your Newsletters do not satisfy you for any reason, you may cancel and receive a credit or refund for the balance of your subscriptions. I invite you to subscribe today. Sincerely, Joseph F. Caponio Acting Director NTIS 5285 Port Royal Road Springfield, VA 22161 or Call (703) 487-4630 Please start my subscription to thefollowingAbstrartNeius/ettens; Special Offer Only: Foreign Technology -Save$15or$30CharterRatc: 2yrs.$150 D (Reg. Rate $180) Charter Rate: 1 yr. $75 D (Reg. Rate $90) Additional subscriptions $25 each annually. Other Titles: Title D plus. Title D plus. additional subscriptions to this Newsletter $25 each D Charge to my Deposit Account No D Here is my check to NTIS for $ Cost Name Occupation. Company — Address — City/State/ZIP. additional subscriptions to this Newsletter $25 each D BUI me Charge to D American Express D MasterCard D VISA Card No Expires TOTAL Signature PR-733-3 4/83 (required to validate order) � 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. 033 Folder The folder containing the original item. 0631 Series The series number of the original item. Series III 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. Benjamin S. H. Harris III Edo D. Pellizzari Kenneth B. Tomer Richard D. Waddell Donald A. Whitaker Description An account of the resource <strong>Corporate Author: </strong>Research Triangle Institute, Research Triangle Park, NC Date A point or period of time associated with an event in the lifecycle of the resource 1980-12-01 Title A name given to the resource Acquisition and Chemical Analysis of Mother's Milk for Selected Toxic Substances ao_seriesIII 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