1 15 6 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 197 Folder The folder containing the original item. 5527 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 Stanley, John S. Kathy E. Boggess John E. Going Gregory A. Mack Janet C. Remmers Joseph J. Breen Frederick W. Kutz Joseph Carra Philip Robinson Source A related resource from which the described resource is derived Environmental Epidemiology Date A point or period of time associated with an event in the lifecycle of the resource 1986 Title A name given to the resource Chapter 14: Broad Scan Analysis of Human Adipose Tissue from the EPA FY82 NHATS Repository ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/d205ca78ad089482ff3793305792b3eb.pdf 3807b98f708558d9a6bf31a1091fa8bb PDF Text Text Item D Number 05526 Author Stanley, John S. Corporate Author D Midwest Research Institute RBDOTt/ArtiClO Tltb Analysis for Polychlorlnated Dlbenzo-p-DioxIns (PCDD) and Dibenzofurans (PCDF) in Human Adipose Tissue: Method Evaluation Study: Final Report with attached letter transmitting the report to Alvin L Young, from Janet C. Remmers, Field Studies Branch, Exposure Evaluation Division, United States Environmental Protection Agency Journal/Book Title Year 1986 Month/Day October Color D Number of Images 149 DUSCrbtOn NOtBI EPA Prime Contract No. 68-02-3938, Work Assignment No. 46, MRI Project No. 8501-A(46). See Item 5522 for draft version Tuesday, March 19, 2002 Page 5526 of 5611 �UNITED STATES ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 JAN I 3 1987 OFFICE OF PESTICIDES AND TOXIC SUBSTANCES A l v i n Young, P h . D . , Lt. Col. , USAF Senior Policy Analyst for Life Sciences Executive Office of the President Office of Science and Technology Policy Room 5005 New Executive Office Building Washington, DC 20506 Dear Dr. Young: Enclosed for your information is the final report entitled "Analysis for Polychlorinated Dibenzo-p_-dioxins (PCDD) and Dibenzofurans (PCDF) in Human Adipose Tissue: Method Evaluation Study." Sincerely, Janet C. Remmers Field Studies Branch Exposure Evaluation Division (TS-798) Enclosure �United Stale-. Environment'*! Aqency Toxic Sub.itan &EPA ANALYSIS FOR POLY-CHLORINATED DIBENZO-p-DIOXINS (PCDD) AND DIBENZOFURANS (PCDF) IN HUMAN ADIPOSE TISSUE: METHOD EVALUATION STUDY 80t- C o co w *C 0) o C o o TJ C 3 O 95% Confidence Limits for Individual Analyses 2, 3, 7, 8-TCDD I I 50 Spiked Concentration (pg/g) �ANALYSIS FOR POLYCHLORINATED DIBENZO-£-DIOXINS (PCDD) AND DIBENZOFURANS (PCDF) IN HUMAN ADIPOSE TISSUE: METHOD EVALUATION STUDY by John S. Stanley, Randy E. Ayling, Karin M. Bauer, Michael J. McGrath, Thomas M. Sack, and Kelly R. Thornburg FINAL REPORT EPA Prime Contract No. 68-02-3938 Work Assignment No. 46 MRI Project No. 8501-A(46) and EPA Prime Contract No. 68-02-4252 Work Assignment No. 24 MRI Project No. 8824-A(01) Field Studies Branch, TS-798 Office of Toxic Substances U.S. Environmental Protection Agency 401 M Street, S.W. Washington, DC 20460 Attn: Ms. Janet Remmers, Work Assignment Manager Dr. Joseph J. Breen, Project Officer �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. The use of trade names or commercial products does not constitute Agency endorsement or recommendation for use. �PREFACE This report provides a summary of the results from a method evaluation study for the determination of 2,3,7,8-substituted polychlorinated dibenzo-p_-dioxins (PCDD) and dibenzofurans (PCDF) in human adipose tissues at the parts-per- trill ion (ppt) level. This method evaluation is an integral part of a collaborative program between the U.S. Environmental Protection Agency's Office of Toxic Substances and the Veterans Administration to determine if significant differences exist in the 2,3,7,8-substituted PCDD and/or PCDF levels in human adipose tissues for Vietnam veterans compared to the general adult male population. The study design will focus on specimens within EPA's National Human Adipose Tissue Survey (NHATS) repository. The method evaluation described in this report was necessary to establish method performance (accuracy and precision) before proceeding with actual sample analysis. This method evaluation study was completed under EPA Contract Nos. 68-02-4252, Work Assignment 24 and 68-02-3938, Work Assignment 46, "Analysis for Dioxins and Furans in Human Adipose Tissue," Ms. Janet Remmers, Work Assignment Manager, and Dr. Joseph Breen, Project Officer. MIDWEST RESEARCH INSTITUTE aul C. Constant Program Manager Approved: Jack Balsinger^ Quality Assurance Coordinator JoMi E. Going, Director Chemical Sciences Department October 30, 1986 m �TABLE OF CONTENTS I. Introduction 1 II. Summary 3 III. Recommendations 4 IV. Experimental 5 A. Preparation of Homogenized Tissue B. Analytical Standards 1. Calibration Standards 2. Spiking Solutions 5 5 6 6 C. Analytical Procedure 10 D. 12 HRGC/MS Analysis E. Data Interpretation 1. Qualitative 2. Quantitation F. Quality Assurance/Quality Control (QA/QC) G. Preliminary Method Studies 1. Gravimetric Studies 2. Carbon-14 Recovery Studies V. Results 23 A. Analytical Results B. Statistical Analysis 1. Recovery of Internal Quantitation Standards. . 2. Estimation of Background Levels of PCDDs and PCDFs 3. Day-to-Day HRGC/MS Analysis Precision VI. Quality Assurance/Quality Control (QA/QC) A. Initial Calibration B. Daily Verification of Response Factors C. Blanks D. 17 17 17 20 20 21 21 23 23 57 57 59 63 63 65 65 Absolute Recoveries of the Internal Quantitation Standards 77 VII. Glossary of Terms 82 VIII. References 84 Appendix A - Analytical Protocol for Determination of PCDDs and PCDFs in Human Adipose Tissue A-l �LIST OF FIGURES Figure 1 2 3 4 5 6 7 8 9 10 11 12 Page Comparison of the HRGC/MS-SIM reconstructed ion chromatogram (RIC) from the analysis of unspiked homogenized human adipose tissue matrix and a calibration standard for PCDDs and PCDFs 30 Example of the TCDF (m/z 304) and TCDD (m/z 320) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 31 Example of the PeCDF (m/z 338) and PeCDD (m/z 354) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 32 Example of the HxCDF (m/z 374) and HxCDD (m/z 390) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 33 Example of the HpCDF (m/z 408) and HpCDD (m/z 424) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 34 Examples of the OCDF (m/z 442) and OCDD (m/z 458) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 35 Measured concentrations versus concentrations of 2,3,7,8TCDD spiked into the homogenized human adipose lipid matrix 36 Measured concentrations versus concentrations of 1,2,3,7,8-PeCDD spiked into the homogenized human adipose lipid matrix 37 Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDD spiked into the homogenized human adipose lipid matrix 38 Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDD spiked into the homogenized human adipose lipid matrix 39 Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDD spiked into the homogenized human adipose lipid matrix 40 Measured concentrations versus concentrations of 1,2,3,4,6,7,8-HpCDD spiked into the homogenized human adipose lipid matrix 41 �LIST OF FIGURES (continued) Figure 13 14 15 16 17 18 19 20 21 22 23 24 25 Page Measured concentrations versus concentrations of OCDD spiked into the homogenized human adipose lipid matrix. . 42 Measured concentrations versus concentrations of 2,3,7,8-TCDF spiked into the homogenized human adipose lipid matrix 43 Measured concentrations versus concentrations of 1,2,3,7,8-PeCDF spiked into the homogenized human adipose lipid matrix 44 Measured concentrations versus concentrations of 2,3,4,7,8-PeCDF spiked into the homogenized human adipose lipid matrix 45 Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDF spiked into the homogenized human adipose lipid matrix 46 Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix 47 Measured concentrations versus concentrations of 2,3,4,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix 48 Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDF spiked into the homogenized human adipose lipid matrix 49 Measured concentrations versus concentrations of 1,2,3,4,7,8,9-HpCDF spiked into the homogenized human adipose lipid matrix 50 Measured concentrations versus concentrations of l,2,3,4,6,7,8~HpCDF spiked into the homogenized human adipose lipid matrix 51 Measured concentrations versus concentrations of OCDF spiked into the homogenized human adipose lipid matrix. . 52 Method accuracy estimates as determined from the slopes of the least squares regression lines for the 17 target PCDD and PCDF analytes 54 Control charts showing response factors by date for 2,3,7,8-TCDF and 2,3,7,8-TCDD 66 VII �LIST OF FIGURES (continued) Figure 26 27 28 29 30 31 32 33 34 Page Control charts showing response factors by date for 1,2,3,7,8-PeCDF and 2,3,4,7,8-PeCDF 67 Control chart showing response factors by date for 1,2,3,7,8-PeCDD 68 Control charts showing response factors by date for 1,2,3,4,7,8-HxCDF and 1,2,3,6,7,8-HxCDF 69 Control charts showing response factors by date for 2,3,4,6,7,8-HxCDF and 1,2,3,7,8,9-HxCDF 70 Control charts showing response factors by date for 1,2,3,4,7,8-HxCDD and 1,2,3,6,7,8-HxCDD 71 Control chart showing response factors by date for 1,2,3,7,8,9-HxCDD 72 Control charts showing response factors by date for 1,2,3,4,6,7,8-HpCDF and 1,2,3,4,7,8,9-HpCDF 73 Control chart showing response factors by date for 1,2,3,4,6,7,8-HpCDD 74 Control charts showing response factors by date for OCDF and OCDD 75 v i ii �LIST OF TABLES Table 1 Page Analytical Standards Available for the Method Evaluation Studies 7 2 Concentration Calibration Solutions 8 3 Native PCDD and PCDF Spiking Solution 9 4 Internal Standard Spiking Solutions 11 5 HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis . . . 13 6 Ions Monitored for HRGC/MS Analysis of PCDD/PCDF 14 7 Typical Daily Sequence for PCDD/PCDF Analysis 16 8 Ion Ratios for HRGC/MS Analysis of PCDD/PCDF 18 9 Summary of the Results of the Sample Preparation Method Evaluation Using Carbon-14 PCDDs 22 Spiked Versus Measured Concentrations of 2,3,7,8-TCDF and 2,3,7,8-TCDD in Homogenized Human Adipose Lipid Samples 24 Spiked Versus Measured Concentrations of 1,2,3,7,8-PeCDF, 2,3,4,7,8-PeCDF, and 1,2,3,7,8-PeCDD in Homogenized Human Adipose Tissue Samples 25 Spiked Versus Measured Concentrations of 1,2,3,4,7,8-; 1,2,3,6,7,8-; 2,3,4,6,7,8-; and 1,2,3,7,8,9-HxCDF in Homogenized Human Adipose Lipid Matrix 26 Spiked Versus Measured Concentration of 1,2,3,4,7,8-; 1,2,3,6,7,8-; and 1,2,3,7,8,9-HxCDD in Homogenized Human Adipose Lipid Samples 27 Spiked Versus Measured Concentrations of 1,2,3,4,6,7,8HpCDF, 1,2,3,4,7,8,9-HpCDF, and 1,2,3,4,6,7,8-HpCDD in Homogenized Human Adipose Lipid Samples 28 Spiked Versus Measured Concentrations of OCDF and OCDD In Homogenized Human Adipose Lipid Samples 29 16 Regression Line Slopes with 95% Confidence Limits 55 17 Results of the Analysis of the Low and High Level Native Spike Solutions 56 10 11 12 13 14 15 IX �LIST OF TABLES (continued) Table Page 18 Background Level Estimates with 95% Confidence Limits . . . 58 19 Day-to-Day Precision of Analysis of Specific Sample Extracts for Tetra- and Pentachloro PCDF and PCDD . . . . 60 Day-to-Day Precision of Analysis of Specific Sample Extracts for Hexa- and Heptachloro PCDF and PCDD 61 Day-to-Day Precision of Analysis of Specific Sample Extracts for OCDF and OCDD 62 Relative Response Factors (Grand Means) Determined from Multipoint Concentration Calibration Standards 64 Summary of Results from the Analysis of a Laboratory Method Blank 76 Recovery of Radiolabeled PCDDs from Precleaned Activated Alumina 78 Absolute Recoveries of the Internal Quantitation Standards from the Human Adipose Lipid Matrix 79 Recovery of Carbon-14 Labeled 2,3,7,8-TCDD, 1,2,3,4,7,8HxCDD, and OCDD as a Function of Final Concentration Conditions 81 20 21 22 23 24 25 26 �I. INTRODUCTION The Environmental Protection Agency Office of Toxic Substances (EPA/OTS) and the Veterans Administration (VA) have established an interagency agreement to study the level of polychlorinated dibenzo-£-dioxins (PCDDs) and dibenzofurans (PCDFs) in human adipose tissues. The occurrence and levels of PCDDs and PCDFs with chlorine substitution in the 2,3,7,8 positions (especially 2,3,7,8-TCDD) of the parent molecules are of primary interest. As part of this interagency effort, it has been proposed to use selected adipose tissue samples that were collected for the Field Studies Branch (FSB) of EPA's Office of Toxic Substances (OTS) through the National Human Adipose Tissue Survey (NHATS) to determine exposure to PCDDs and PCDFs. The available adipose tissues include specimens obtained from young men whose age indicates that they could have served in Vietnam and could have been exposed to Agent Orange. The tissues were originally collected as part of a broadly based and statistical random sampling of the continental United States. The analysis of these tissues may provide information on the differences of exposure of the general adult male population and Vietnam veterans to the 2,3,7,8-substituted PCDDs and PCDFs. The overall objectives of the proposed EPA/VA collaborative studies are: 1. Evaluate the reliability, accuracy, precision, and sensitivity of a proposed method for the determination of 2,3,7,8substituted PCDDs and PCDFs (tetra- through octachloro homologs) in human adipose tissue at the parts-per-tri1 lion (ppt) level. 2. Determine if these compounds can be detected in adipose tissues of the American male adult population; and 3. Determine if individuals with military service in Vietnam have significantly different levels of 2,3,7,8-substituted PCDDs and PCDFs (particularly 2,3,7,8-TCDD) than other American men. As a prelude to this work assignment, MRI conducted an extensive literature review of applicable analytical methods and conducted a meeting with recognized experts in this field to identify critical aspects of analytical methodology.1'2 Based on the information gathered through the literature review and the meeting with the recognized experts, a special report was prepared for OTS proposing a framework for an analytical method for analysis of human adipose tissues.3 Several studies have been completed since the issuance of that report which reflect the advances in analytical techniques for adipose tissue analysis.4 16 The salient features of these methods have been combined into a single protocol for the routine analysis of tetra- through octachloro PCDDs and PCDFs at the low-parts-per-tril1 ion level for the EPA/VA tissue study. This report focuses on a method evaluation study that was conducted to achieve the first objective of the interagency agreement. Clarification �of method performance is necessary before proceeding with the analysis of actual samples retrieved from the NHATS repository. This report includes a summary of the method evaluation study results (Section II). Recommendations to be implemented before proceeding with the actual tissue samples from the NHATS repository are presented in Section III. A description of the actual experimental procedures is provided in Section IV. Results of sample analyses are summarized in Section V, and quality assurance/quality control (QA/QC) aspects of the study are detailed in Section VI. Pertinent references are listed in Section VII. Appendix A contains the detailed analytical protocol that will be followed for the analysis of the NHATS specimens designated in the study design to be provided by EPA/VA. �II. SUMMARY The results of the replicate analysis of spiked and unspiked homogenized human adipose tissue matrix demonstrate that the analytical method produces accurate and precise data for 17 specific 2,3,7,8-substituted PCDD and PCDF (tetra- through octachloro homologs) compounds. Accuracy of the analytical method was demonstrated to range from 90 to 120% for the 17 2,3,7,8-substituted PCDD and PCDF compounds. Data are reported for three or four replicate analyses of samples spiked at three different concentration levels. The endogenous or background levels of the PCDD and PCDF congeners in the homogenized adipose lipid matrix were estimated through regression analyses of measured (found) versus spiked concentrations for each compound. The analytical method is capable of providing quantitative data for tetra- through octachloro PCDD and PCDF congeners to concentration levels as low as 1 pg of the tetrachloro congeners per gram of adipose tissue. However, an interference was noted at m/z 304 which coeluted with 2,3,7,8-TCDF, resulting in a detection level of approximately 4 pg/g. Average absolute recoveries of the internal quantisation standards ranged from 52% for 13C12-TCDD up to 89% for 13C12-OCDD. The agreement of the measured concentrations versus the spiked concentrations for each PCDD and PCDF congener demonstrates that the internal standard quantitation procedure provides an accurate measure of concentration which is independent of the absolute recovery. Final concentration conditions were noted to have pronounced effect on the absolute recoveries of the lower chlorinated compounds, particularly 2,3,7,8-TCDD. Experiments with carbon-14 labeled 2,3,7,8-TCDD demonstrated that final concentration at temperatures of 55 to 60°C resulted in recoveries as low as 54% while the same procedure conducted at ambient conditions resulted in greater than 90% recovery. Analysis of method and reagent blanks provided information on potential artifacts in the sample preparation scheme. Additional experiments were conducted with carbon-14 labeled PCDDs to evaluate the cleanup efficiency and recovery of PCDDs from chromatographic materials, particularly acidic alumina. �III. RECOMMENDATIONS Some minor modifications have been made in the written protocol (Appendix A) that were not included in this phase of the method validation. These include: a cleanup procedure for activated acidic alumina prior to fractionation of sample extracts to remove artifacts; and final concentration of the sample extracts using nitrogen blowdown at room temperature rather than heating to 55-60°C. The spiking solutions used to prepare the spiked quality control samples should be submitted for replicate (minimum of three/per spike level) HRGC/MS analysis to assist the interpretation of positive or negative bias in the accuracy of QC sample data. The accuracy bounds should be extended to 50-130% from 50-115% as specified in the draft quality assurance program plan. The method should include additional internal quantitation standards to pair with the HpCDF 13 OCDF congeners. Also, an additional internal recovand ery standard, possibly C12-l,2,3,4,7,8-HxCDD, is required to provide better estimates of absolute method recovery. These additional compounds, if available, will be incorporated into the method before initiating sample analyses. Analysis for 2,3,7,8-TCDF may require high resolution mass spectrometry to avoid interferences occurring at m/z 304. This will require modification of the HRGC/HRMS portion of the protocol to include the specific acquisition parameters for the characteristic ions of 2,3,7,8-TCDF. �IV. EXPERIMENTAL A. Preparation of Homogenized Tissue A bulk lipid sample was prepared from the extracts of human adipose tissue samples collected through the NHATS program. The adipose tissue samples have been stored in a deep freezer at approximately -10°C since collection. The homogenized tissue extract or bulk lipid was used in this method evaluation study for preparation of replicate samples spiked with varying levels of specific PCDD and PCDF isomers. This homogenized matrix will also be used for preparing control and spiked quality control samples for the actual NHATS sample analysis phase of the program. A total of 2,465 g of adipose tissue was extracted, dried, and concentrated to yield 1,652 g (62% of original weight) of homogenized lipid. Specific procedures for preparing this matrix are described below. The adipose tissue samples were thawed at room temperature for 1 to 2 h. Portions of the samples were added to a blender cup of a Waring® blender and covered with methylene chloride. The volume of methylene chloride was approximately equal to the sample volume (100 to 200 ml). This mixture was blended at high speed for approximately 10 min, and the contents were transferred to a 500-mL Erlenmeyer flask and further blended with a Tekmar® Tissumizer, also at high speed for 10 min. A powder funnel was plugged with a wad of glass wool (silanized, methylene chloride extracted) and filled with •v 50 g of sodium sulfate (heated overnight to 600°C in a muffle furnace). The sodium sulfate was wetted with methylene chloride prior to elution of the sample extract. The dried effluents were refiltered in the same way using a fresh bed of sodium sulfate to remove particulate and residual water. The samples were transferred to 1-L round bottom flasks, and the solvent was removed by rotary evaporation. The water bath on the rotary evaporator was kept at 60°C using a thermostatted heating element. Once the solvent appeared to have been removed (constant volume in flask, no visible condensation in condenser), the heating and evaporation process was continued for at least 2 h. The flask and contents were removed and stored in a refrigerator. The extracted lipid solidified upon refrigeration and was visually checked for homogeneity. No precipitates or phase separation was observed. The lipid residue was allowed to liquify at room temperature and was transferred to a 4-L glass bottle with a Teflon®-!ined lid. The lipid residue was brought to room temperature and heated just enough to allow the lipid to achieve an oily state prior to aliquotting portions for the method evaluation studies. B. Analytical Standards Analytical standards including native PCDD and PCDF congeners, stable isotope (carbon-13) labeled standards and radiolabeled (carbon-14) standards were purchased from Cambridge Isotope Laboratories, Woburn, Massachusetts, and Pathfinder Laboratories, St. Louis, Missouri. The 2,3,7,8TCDD was received from the EPA Reference Materials Branch as a solution in �isooctane. The other native PCDD and PCDF congeners were received as 1-mg neat standards. The stable and radiolabeled isotopes were received as solutions in n-nonane or isooctane and toluene, respectively. Table 1 provides a summary of the standards used for this study. Stock solutions of the individual PCDD and PCDF congeners were prepared from the neat standards. The neat materials were weighed using a Cahn 27 electrobalance calibrated versus a 1-mg (Class M) standard. The neat compounds were transferred to glass vials and were dissolved in 2.0 to 3.0 mL of toluene (Burdick and Jackson, distilled in glass). Toluene was added to each standard using volumetric pipettes (Class A). The OCDD required dilution to 10.0 ml using a 50:50 mixture of toluene and anisole. A working solution consisting of the 17 native PCDD and PCDF congeners was prepared in toluene at a concentration of 2 (J9/mL for the TCDD, TCDF, PeCDD, and PeCDF congeners, 5 ug/mL for the HxCDD, HxCDF, HpCDD, and HpCDF congeners, and 10 ug/mL for the OCDD and OCDF. The working solution was used to prepare both the lipid matrix spiking solution and the calibration standards. The stable isotope labeled internal standards were obtained as solutions13in n-nonane or isooctane at 50 ug/mL concentration with the exception of the C12-OCDD, which was provided at 10 ug/mL. Separate working solutions containing mixtures of the carbon-13 labeled PCDDs and PCDFs were prepared in isooctane for use in the calibration standards and the sample spiking solutions. The carbon-14 radiolabeled PCDDs were used for preliminary method evaluation studies. The specific activity of the 14C-2,3,7,8-TCDD (117.56 mCi/mmole) was high enough to allow recovery studies at spike levels equivalent to 10 pg/g for a 10-g sample. 1. Calibration Standards Eight concentration calibration standards containing the 17 native and the 9 carbon-13 labeled internal standards were prepared for determining the consistency of response factors for the native PCDDs and PCDFs versus the corresponding carbon-13 congeners. Table 2 presents a summary of the calibration standards prepared for the method calibration study. The solution concentrations (pg/pL) can also be considered as equivalent to residue levels in picograms per gram of adipose. For example, a 1 pg/uL concentration standard corresponds to a tissue concentration of 1 pg of PCDD or PCDF congener per gram of adipose assuming a 10-g sample is available for analysis. 2. Spiking Solutions a. Native PCDD and PCDF A solution containing the 2,3,7,8-substituted PCDD and PCDF congeners was prepared in isooctane for spiking the homogenized lipid materials for the method evaluation study. Table 3 specifies the levels of each of the native PCDD and PCDF congeners present in this solution. �Table 1. Analytical Standards Available for the Method Evaluation Studies Source Compound Native 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2, 3,4,7, 8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF EPA QA Reference Materials Branch Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Lot/Code 20603 AWN 1203-74/EF-903C MLB-706-53/ED-950C AWN-729-21/EF-953C AWN-729-45/EF-956C 830244/ED-961C MLB-706-47/ED-960C MLB-706-73/ED-969C AWN-729-20/EF-964C MB 13106-7/EF-962-C MB 13106-47/EF-967-C MB 13106-3/EF-968-C MLB-706-21/ED-971C AWN-729-22/EF-973C MB-13-106-77/EF-975C 8465-F-982-C/EF-982C F2832/ED-980C 13 C12-Internal stand;irds Cambridge 2,3,7,8-TCDD Cambridge 2,3,7,8-TCDF 1,2,3,7,8-PeCDD Cambridge 1,2,3,7,8-PeCDF Cambridge Cambridge 1,2,3,6,7,8-HxCDD Cambridge 1,2,3,4,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD Cambridge OCDD Cambridge 37 Isotope Isotope Isotope Isotope Isotope Isotope Isotope Isotope Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories R00208/ED-900 R00236/EF-904 R00241/ED-955 R00221/EF-952 R00249/ED-966C R00234/EF-963C R00248/ED-972 R00263/ED-981 Cl-Internal standaird 37 C14-1,2,3,4,6,7,8HpCDD KOR Isotopes 580012/SSY-4-32 l4 C 12 -Radiolabeled standards 2,3,7,8-TCDD Pathfinder Laboratories Pathfinder Laboratories 1,2,3,4,7,8-HxCDD OCDD Pathfinder Laboratories S.A. = specific activity. S.A. = 117.56 mCi/mmole S.A. = 24.16 mCi/mmole S.A. = 20.50 mCi/mmole �Table 2. Concentration Calibration Solutions' Concentration in calibration solutions in p flA iL CS1 CS2 CS3 CS4 CS5 CS6 CS7 CSS Compound Native 200 200 200 200 200 500 500 500 500 500 500 500 500 500 500 1 ,000 1 ,000 25 5 5 5 5 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 25 25 100 100 100 100 100 250 250 250 250 250 250 250 250 250 250 500 500 50 50 50 50 50 125 125 125 125 125 125 125 125 125 125 250 250 25 25 25 25 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 125 125 Internal quantitation standards 13 50 50 C12-2,3,7,8-TCDD 13 C12-2,3,7,8-TCDF 50 50 13 50 50 C12-l,2,3,7,8-PeCDD 13 C12-l,2,3,7,8-PeCDF 50 50 13 C12-l,2,3,6,7,8-HxCDD 125 125 13 C12-l,2,3,4,7,8-HxCDF 125 125 13 C12-1,2,3,4,6,7,8125 125 HpCDD 13 C12-OCDD 250 250 50 50 50 50 125 125 125 50 50 50 50 125 125 125 50 50 50 50 125 125 125 50 50 50 50 125 125 125 50 50 50 50 125 125 125 50 50 50 50 125 125 125 250 250 250 250 250 250 50 50 50 50 50 50 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF 10 10 10 10 10 25 25 25 25 25 25 25 25 25 25 50 50 2. 5 2. 5 2. 5 2. 5 2. 5 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 12. 5 12. 5 1 1 1 1 1 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 5 5 Internal recovery standard 13 C12-1,2,3,4-TCDD Prepared in tridecane. 50 50 �Table 3. Native PCDD and PCDF Spiking Solution3 Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF a Prepared in isooctane. Concentration (pg/pL) 5 5 5 5 5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 25 25 �b. Internal Standards Two different internal standard spiking solutions were prepared for quantitation of native PCDD and PCDF congeners. The compositions of each of the spiking solutions are presented in Table 4. The internal quantitation standards were spiked into the lipid aliquots prior to any cleanup procedures and hence were carried throughout the method exactly as the corresponding native congeners. The internal recovery standard was added in 10 uL of a keeper solution (tridecane) during final extract concentration prior to analysis. The recovery standard was used to measure the absolute method recoveries of the internal quantitation standards. C. Analytical Procedure The homogenized human adipose lipid matrix was allowed to come to room temperature and then warmed in a water bath until the matrix changed to an oily state. Approximately 10.0 g of the oily material was transferred by pipette to preweighed glass vials, and the actual weight of the lipid was determined to the nearest 0.01 g by difference using an analytical balance. Four 10.00-g aliquots were spiked with 20 uL of the native spiking solution presented in Table 3, another four aliquots were spiked with 50 (jL of the same solution, and three additional aliquots were spiked with 100 p.L of native PCDD and PCDF solution. These spikes were equivalent to concentrations ranging from 10, 25, and 50 pg/g in the lipid matrix for the tetra- and pentachloro PCDD and PCDF congeners up to 50, 125, and 250 pg/g for the OCDD and OCDF for the low, medium, and high level spikes. In addition to the spiked samples, three aliquots of the lipid material were transferred for determining the endogenous levels of each of the PCDD and PCDF congeners in the control matrix. Each of the sample aliquots was fortified with 100 |jL of the internal quantitation standard spiking solution (Table 4). The spiked samples were each quantitatively transferred to 500-mL Erlenmeyer flasks using hexane. The residues were diluted with a total of 200 mL of hexane, and 100 g of sulfuric acid (H2S04) modified silica gel (40% w/w) was added to each solution with stirring. The mixtures were stirred for approximately 2 h, and the supernatants were decanted and filtered through filter funnels packed with anhydrous sodium sulfate (Na2S04). The H2S04 modified silica adsorbents were washed with at least two additional aliquots of hexane and dried by elution through Na2S04. The combined hexane extracts for each sample were eluted through a column consisting of the 40% H2S04 modified silica gel (4.0 g) and silica gel (1.0 g). The eluates were concentrated to approximately 15 mL and added to columns of acidic alumina (Bio-Rad, AG-4, 6.0 g). The acidic alumina columns were eluted first with 20 mL of hexane, which was collected but not analyzed, followed by elution with 30 mL of 20% methylene chloride in hexane. The PCDDs and PCDFs were eluted from the acidic alumina using the 20% methylene chloride in hexane. The PCDDs and PCDFs in the eluates were isolated from other chlorinated planar aromatics using columns (5-mL disposable pipettes containing 500 mg of 18% Carbopak C and Celite-545). The Carbopak C/Celite 10 �Table 4. Internal Standard Spiking Solutions Concentration (pg/uL) Compound Internal quantisation standard 13 5 13 5 13 5 13 5 C12-2,3,7,8-TCDD C12-2,3,7,8-TCDF C12-l,2,3,7,8-PeCDD C12-l,2,3,7,8-PeCDF 13 12.5 13 12.5 13 12.5 C12-l,2,3,6,7,8-HxCDD C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,6,7,8-HpCDD 13 C12-OCDD 25 Internal recovery standard 13 C12-1,2,3,4-TCDD 50 .Solution prepared in isooctane. Solution prepared in tridecane. 11 �columns were pre-eluted with 2 ml of toluene, 1 mL of 75:20:5 methylene chloride/methanol/benzene, 1 mL of 1:1 methylene chloride/cyclohexane, and 2 ml of hexane. The sample extracts (30 ml) were added to the columns, which were eluted with 2 mL of hexane, 1 mL of 1:1 methylene chloride/cyclohexane, and 1 mL of the 75:20:5 methylene chloride/methanol/benzene. These eluents were collected and combined but were not analyzed. The Carbopak C/Celite columns were turned upside down, and the PCDDs and PCDFs were eluted with 20 mL of toluene. The toluene was concentrated to less than 1 mL using flowing nitrogen and a heated water bath (55-60°C) and transferred to 1.0-mL conical vials using a solution of 1% toluene in methylene chloride. Tridecane (10 uL) containing 500 pg of the internal recovery standard 13C12-1,2,3,4-TCDD was added as a keeper when the solution had concentrated to approximately 200 uL. The extracts were concentrated to final volume using nitrogen and the heated water bath. D. HRGC/MS Analysis The analyses of the spiked and unspiked lipid samples were completed using a Kratos MS50TC double-focusing magnetic sector mass spectrometer. The determination for the tetra- through octachloro homologs was achieved in a single analysis using the conditions described in Table 5. Table 6 provides the characteristic ions monitored for each PCDD and PCDF homolog. As noted from Table 6, the analysis requires five different parameter descriptions that were switched automatically during the course of the analysis. Parameters monitored included two characteristic molecular ions for each PCDD and PCDF homolog and the corresponding carbon-13 labeled internal standard. In addition, a fragment ion of perfluorokerosene (PFK), m/z 380.976, was monitored throughout each analysis to ensure that proper mass calibration was maintained. The parameter descriptors also included an ion characteristic of specific homologs of chlorinated diphenyl ethers to demonstrate that responses meeting the qualitative criteria for specific PCDF congeners were not due to these potential interferences. Triplicate analyses of six of the eight calibration solutions (Table 2) were completed, and the variability in relative response factors across this range was calculated. The analyst was required to demonstrate on a daily basis that the relative response factors (RRF) were in agreement within ± 20% of the established averages for 2,3,7,8-TCDD and 2,3,7,8-TCDF and within ± 30% of the average RRF values for the other congeners. The equation used to calculate the relative response factors for each PCDD and PCDF congener are discussed later in this report (Section E - Data Interpretation, 2 - Quantitation, p. 17). The analyst was also required to determine column performance by analyzing a mixture of TCDD isomers before proceeding with sample analysis. Table 7 gives an example of the typical daily sequence for PCDD/PCDF analysis. 12 �Table 5. HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis Mass spectrometer 8,000 V 500 MA 70 eV -1,800 V 280°C Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: Overall SIM cycle time: ^ 3,000 (10% valley definition) 1s Gas chromatograph Column coating: Film thickness: Column dimensions: He linear velocity: He head pressure: DB-5 0.25 60 m •v 25 1.75 Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 s 30 mL/min 6 mL/min 270°C 300°C 1-2 pL 200°C 2 min 200°C to 330°C at 5°C/min 13 (jm x 0.25 mm ID cm/sec kg/cm2 (25 psi) �Table 6. Ions Monitored for HRGC/MS Analysis of PCDD/PCDF Descriptor Al ID TCDF 13 C12-TCDF TCDD 13 C12-TCDD HxCDPE PFK (lock mass) A2 TCDF TCDD PeCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD PFK (lock mass) HpCDPE A3 HxCDF PFK (lock mass) 13 C12-HxCDF HxCDD 13 C12-HxCDD OCDPE Mass Nominal dwell time (sec) 303.902 305.899 315.942 317.939 319.897 321.894 331.937 333.934 373.840 380.976 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 303.902 305.899 319.897 321.894 337.863 339.860 349.903 351.900 353.858 355.855 365.898 367.895 380.976 407.801 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.035 0.035 373.821 375.818 380.976 385.861 387.859 389.816 391.813 401.856 403.853 443.759 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 �Table 6 (continued) Descriptor A4 Mass ID PFK (lock mass) HxCDD 380.976 389.816 391.813 407.782 409.779 419.822 421.819 423.777 425.774 435.817 437.814 429.768 431.765 477.720 C12-HpCDF HpCDD 13 37 C12-HpCDD Cl4-HpCDD NCDPE A5 PFK (lock mass) OCDF 13 C12-OCDF OCDD 13 C12-OCDD DCDPE 15 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 380.976 441.743 443.740 453.784 455.781 457.738 459.735 469.779 471.776 511.681 HpCDF 13 Nominal dwell time (sec) 0.060 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.060 �Table 7. Typical Daily Sequence for PCDD/PCDF Analysis 1. Tune and calibrate mass scale versus perf1uorokerosene (PFK). 2. Determine column performance by injecting the TCDD isomer mixture. 3. Inject concentration calibration solution 2.5 to 12.5 pg/|A (CS-7) solution. 4. Inject blank (tridecane). 5. Inject samples 1 through "n". 6 Inject concentration calibration solution 2.5 to 12.5 pg/pL (CS-7) solution. 16 �E. Data Interpretation 1. Qualitative The HRGC/MS elution profiles of the tetra- through octachloro PCDD and PCDF homologs were established through the analysis of environmental sample extract (fly ash from a municipal waste incinerator). The characteristic ions for each homolog were plotted within the retention window established using this mixture. The criteria for identification of a response as a PCDD or PCDF were the coincidental response of the characteristic ions monitored within the established retention window, and within ± 20% of the theoretical ion ratio. Table 8 presents the range of ion ratios used for the qualitative criteria for the specific PCDD and PCDF homologs and internal standards. 2. Quantisation Quantisation of the specific PCDD and PCDF congeners was achieved using the respective internal quantisation standards. For example, TCDD was quantitated versus the 13C12-2,3,7,8-TCDD; PeCDD versus the 13C12-1,2,3,7,8PeCDD, etc. The HpCDF and OCDF responses were quantitated versus the carbon-13 labeled hepta- and octachlorodibenzo-p_-dioxin internal standards since the corresponding dibenzofuran internal standards were not available for this study. The absolute recovery of the internal quantisation standards was achieved using 13C12-1,2,3,4-TCDD. A second internal recovery standard, 37 Cl4-l,2,3,4,6,7,8-HpCDD, was evaluated but was not used for recovery measurements due to interference arising from the corresponding native HpCDD. Relative response factors (RRF) were calculated for each of the native PCDD and PCDF compounds listed in Table 2. The RRF values were calculated as shown in Equation 1. "crn x STD RRF = A R T<^ 1S X c IS X LSTD Eq. 1 where Arjn = the sum of the area responses for the two characteristic ions of the standard compound; AT<- = the sum of the area responses for the two characteristic ions of the corresponding internal quantisation standard; CIS = concentration (pg/|jL) of the internal quantisation standard; and = concentration (pg/uL) of the standard compound. The relative response factors for the internal quantisation standards (RRFjS) were calculated as shown in Eq. 2. 17 �Table 8. Compound TCDF 13 C12-TCDF TCDD 13 C12-TCDD PCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD HxCDF 13 C12-HxCDF HxCDD 13 C12-HxCDD HpCDF 13 C12-HpCDF HpCDD 13 C12-HpCDD OCDF 13 C12-OCDF OCDD 13 C12-OCDD Ion Ratios for HRGC/MS Analysis of PCDD/PCDF Ions monitored Theoretical ratio 304/306 316/318 320/322 332/334 338/340 350/352 354/356 366/368 374/376 386/388 390/392 402/404 408/410 420/422 424/426 436/438 442/444 454/456 458/460 470/472 0.76 0.76 0.76 0.76 0.61 0.61 0.61 0.61 1.22 1.22 1.22 1.22 1.02 1.02 1.02 1.02 0.87 0.87 0.87 0.87 Acceptable range is ± 20% of the theoretical value. 18 Acceptabl e range' 0.61 - 0.91 0.61 - 0.91 0.61 - 0.91 0.61 0.49 0.49 0.49 0.49 0.98 0.98 0.98 0.98 0.82 0.82 0.82 0.82 0.70 0.70 0.70 0.70 - 0.91 0.73 0.73 0.73 0.73 1.46 1.46 1.46 1.46 1.22 1.22 1.22 1.22 1.04 1.04 1.04 1.04 �TC * RRF IS IS = QG. Eq AH RS x CLjg X "2 where A JS and GIS are defined as in Equation 1 and C10 = concentration (pg/uL) of the internal recovery standard, DC 13 C12-1,2,3,4-TCDD, and A RS = the sum of the area responses for the two characteristic ions (m/z 332 and 334) corresponding to the internal recovery standard. A calibration curve was established using six concentration levels of standards; for example, the calibration curve for 2,3,7,8-TCDD was initially established with standards at concentrations of 1, 2.5, 5, 10, 50, and 100 pg/(jL. The 2.5 pg/uL standard was analyzed daily to verify response factors and instrument sensitivity. The RRF values for each of the internal quantisation standards were calculated versus the internal recovery standard, 13 C12-1,2,3,4-TCDD, using Equation 2. The concentration of a PCDD or PCDF congener in a composite sample was calculated as shown in Equation 3. _ Asamp1e x QIS Vr AIS x RRF x Wt r where P tq> ~ * CUT = wet tissue concentration of the PCDD or PCDF congener in each tissue (pg/g); A sample = sum of the area responses for the two characteristic ions of , thfi pC[JD Qr pCDF congener; A-,<- = sum of the area responses for the two characteristic ions of the respective internal quantitation standard; Q T r = amount of the internal quantitation standard added to the 1:> 13 sample (500 pg of 13C12-TCDD,13 C12-TCDF, 13C12-PeCDD, 13 and 13C12-PeCDF; 1,250 pg of C12-HxCDD, 13C12-HxCDF, and C12-HpCDD; or 2,500 pg of 1§C12-OCDD); RRF = the relative response factor for the PCDD or PCDF congener from Equation 1; and Wt = mass of the sample (grams). The absolute recovery of the internal quantitation standard was calculated using Equation 4. 19 �L Recovery (%) = •? * ^ ARS x RRF IS x QIS where x 100 Eq 4 AR<- = sum of the area responses for the two characteristic ions of the internal recovery standard, 13C12-1,2,3,4-TCDD; QR<- = amount of the internal recovery standard (13C12-1,2,3,4-TCDD) added to the final extract (500 pg); and RRF IS = response factor of the internal quantisation standard relative to the internal recovery standard. These values are calculated as defined in Equation 2. The RRF T< - values were all calculated versus 13C12-1,2,3,4-TCDD. All data were qualified to reflect that the response for a particular compound was a positive quantifiable parameter, present as a trace value only, or was not detected. Positive quantifiable values were identified for responses greater than 10 times background signal to noise. Trace (TR) values were assigned to responses which were in the range of 2.5 to 10 times background signal to noise. A value of not detected (ND) was used to reflect that a response was not detected at greater than 2.5 times signal to noise. A limit of detection (LOD) was calculated for all trace and not detected values using the peak height response of the respective internal standard and the average measured signal to noise for the characteristic ions of the PCDDs and PCDFs. F. Quality Assurance/Quality Control (QA/QC) The QA/QC procedures included analysis of multipoint calibration concentration standards to establish relative response factor (RRF) curves for each of the 17 native PCDD and PCDF congeners. Triplicate analyses of 6 concentration calculation standards (Table 2; CS2, CSS, CSS, CS6, CS7, and CSS) were determined to vary by less than ± 20% for TCDD and TCDF and ± 30% for all other PCDD and PCDF congeners. The mean RRF values were also determined to vary by less than this criteria over the entire calibration range. The mean RRF (RRF) values and instrument sensitivity were verified daily by bracketing the sample analyses with an injection of a standard that ranged from 2.5 pg/uL for 2,3,7,8-TCDD and 2,3,7,8-TCDF up to 12.5 pg/(jL for OCDD and OCDF. The criterion for continuing with the sample analysis was agreement of the measured RRF value with the mean RRF within ± 20% for 2,3,7,8TCDD and TCDF and ± 30% for all other PCDD and PCDF congeners. Other activities included the analysis of laboratory method blanks and reagent blanks and measurement of the absolute recoveries of the internal quantisation standards. Laboratory method blanks were samples that were handled exactly as an adipose sample except no lipid matrix was used. G. Preliminary Method Studies Prior to analysis of the homogenized human adipose lipid matrix by HRGC/MS, several experiments were conducted to confirm that the sample preparation scheme was feasible. 20 �1. Gravimetric Studies The first concern was the efficient removal of up to 10 g of lipid matrix from extracted adipose tissue. A series of experiments was conducted with 10-g lipid aliquots to demonstrate removal of lipid using the H2S04-Si02 slurry technique. Initially, 50 g of the acid modified silica was added to the lipid extract in 100 ml of hexane. The acid modified silica was noted to turn dark brown immediately on contact with the lipid solution. The hexane was recovered and the adsorbent was extracted with additional hexane. The extracts were combined and concentrated to 5 mL with Kuderna-Danish evaporators. The extract was eluted through a column of 4.0 g of acid modified silica and 1.0 g of silica with 45 ml of hexane. The acid modified silica was noted to be highly discolored throughout, and the extract required a second slurry treatment of the eluent with an additional 50 g of acid modified silica gel. The adsorbent from the second slurry procedure was noted to discolor significantly, indicating that lipid materials had not been efficiently removed from the first step of the procedure. The hexane supernatant from the second slurry cleanup was reduced in volume and taken to dryness in a preweighed glass vial. The final residue was measured at approximately 10 mg for duplicate samples, which translates into a removal efficiency of 99.9% based on the initial 10-g aliquot. This lipid cleanup procedure was modified such that 100 g of acid modified silica gel was used in the initial slurry cleanup, followed by elution of the resulting extract through a column containing 4.0 g of acid modified silica and 1.0 g of silica gel. The lipid removal efficiency of duplicate samples through the cleanup procedure was determined to average 99.8% (20 to 30 mg of the initial 10-g lipid remained after cleanup). The column cleanup step in this procedure did not exhibit any significant color change. Thus this step of the procedure was incorporated into the method as a check of the efficiency of lipid removal to prevent overloading of the acidic alumina fractionation column. 2. Carbon-14 Recovery Studies The carbon-14 radio!abeled PCDD standards listed in Table 1 were used to estimate overall method recoveries for the tetra- through octachloro homologs prior to proceeding with the HRGC/MS evaluation. Triplicate analyses (10-g aliquots of lipid materials) were conducted with each of the three radiolabeled standards. The first experiment addressed the recovery of the compounds from bulk lipid cleanup. Triplicate analyses using 10-g aliquots were completed for the three compounds at the following concentrations: 14 C-2,3,7,8-TCDD, 10 pg/g; 14C-l,2,3,4,7,8-HxCDD, 100 pg/g; and 14C-OCDD, 250 pg/g. The results of these analyses indicated that all compounds were recovered in the range of approximately 70 to 80%. Following this experiment, the total sample preparation procedure described earlier in this report was evaluated using triplicate analysis of 10-g lipid samples. Table 9 provides a summary of the results from this study. These data indicate that overall method recovery is limited by the initial bulk lipid removal procedure. This assumption is based on the similar recoveries of the carbon-14 labeled compounds noted for evaluation of the bulk lipid removal step as compared to the total sample preparation scheme. 21 �Table 9. Summary of the Results of the Sample Preparation Method Evaluation Using Carbon-14 PCDDs Spike levels Bulk lipid removal, recovery Analytes (pg/g) Total method a recovery ( ) % 14 C-2,3,7,8-TCDD 10 68 75 14 100 79 66 250 82 76 C-l,2,3,4,7,8-HxCDD 14 C-OCDD Average value for triplicate analyses taken through the total sample preparation scheme. Precision of measurements varied by less than ± 10% (relative standard deviation). Average value for triplicate analyses taken through bulk lipid cleanup only. Precision of measurement varied by less than ± 6% (relative standard deviation). 22 �V. RESULTS A. Analytical Results The analytical results for the quantisation of the 17 target PCDD and PCDF 2,3,7,8-substituted congeners in the spiked and unspiked homogenized human adipose lipid samples are presented in Tables 10 to 15. These data demonstrate that 13 of the 17 congeners were definitely detected in the unspiked lipid matrix. Although 2,3,7,8-TCDF is reported as not detected, responses for the characteristic ions (m/z 304 and 306) greater than 10 times signal to noise were noted to be coincident with the internal standard, 13C122,3,7,8-TCDF. The ratio of the responses (m/z 304/306) in each of the triplicate analyses of the unspiked matrix were well outside the acceptable ratio of 0.90 to 0.61 established in Table 8. Figure 1 provides a comparison of the HRGC/MS-SIM responses noted for the unpsiked human adipose lipid matrix as compared to a concentration calibration standard. Figures 2 through 6 provide examples of the individual PCDD and PCDF responses observed for the unspiked lipid samples as compared to fortified matrices. In general, the precision of the replicate measurements at each spike level is good (relative standard deviations typically less than 10%) for PCDD and PCDF congeners that were detected with responses greater than 10 times signal to noise. The precision of the measurements for the unspiked matrices for 1,2,3,7,8-PeCDF (Table 11), 1,2,3,7,8,9-HxCDF (Table 12), and OCDF (Table 15) ranges from 21.6% to 43.1% as a result of little or no response at the specified retention windows. B. Statistical Analysis The regression results for each of the 17 specific PCDD and PCDF congeners are plotted separately in Figures 7 to 23. These plots provide the results for the individual sample analyses, a line defining the results of a least squares regression analysis, and boundaries that depict the confidence limits for the range of spiked concentrations. The regression lines were obtained by the method of least squares using the sample measurements at the three spiking levels and the unspiked level. Two types of upper and lower 95% confidence limits or bounds were calculated for the least square regressions of measured (found) concentrations versus spiked levels. The first set of confidence limits (defined by the inner pair of curves closest to the regression line) is the 95% confidence bounds for the regression line. These bounds are interpreted as follows: The true regression line (as would be determined if the experiment were repeated a countless number of times at the same spiked levels) lies within these confidence limits unless the analytical results are sufficiently unusual to be among those expected to occur less than 5% of the time. The second set of confidence bounds, depicted by the outer pair of lines, constitutes the 95% confidence limits for the result of a single analysis at a particular spiking level. The interpretation is as follows: the result (reported value) of a single analysis of a sample spiked at a given level can be predicted to fall between these 95% confidence bounds unless the analytical result is among those sufficiently unusual to be expected less than 5% of the time. 23 �Table 10. Spiked Versus Measured Concentrations of 2,3,7,8-TCDF and 2,3,7,8-TCDD in Homogenized Human Adipose Lipid Samples 13 13 C12-TCDF absolute recovery ( ) % 2,3,7,8-TCDD spike level (pg/g) (pg/g) C12-TCDD absolute recovery (%) (.) 41* ( . )a 41? (.) 40 (.) 41 0.1 1.8 59 63 59 60.3 2.3 3.8 0 0 0 10.7 11.4 13.1 11.7 1.2 10.5 53 52 46 50.3 3.8 7.5 10 10 10 10 14.3 14.8 13.6 13.8 14.1 0.5 3.9 61 67 71 78 69.3 7.1 10.3 10 10 10 10 23.4 22.8 24.7 22.5 23.3 1.0 4.1 53 53 53 62 55.3 4.5 8.1 25 25 25 25 30.8 30.8 30.7 28.7 30.2 1.1 3.5 59 75 62 72 67.0 7.7 11.5 25 25 25 25 40.8 40.6 40.3 38.3 40.0 1.2 2.9 48 53 51 60 53.0 5.1 9.6 50 50 50 57.7 59.4 55.8 57.6 1.8 3.1 64 48 58 56.7 8.1 14.3 50 50 50 65.8 72.1 67.6 68.5 3.3 4.8 55 43 48 48.7 6.0 12.4 2,3,7,8-TCDF spike level 2,3,7,8-TCDF concentration (pg/g) 0 0 0 Mean STD RSD (%) ro Mean STD RSD ( ) % Mean STD RSD (%) Mean STD RSD (%) (pg/g) ND ND ND ND 2,3,7,8-TCDD concentration ND = not detected. Value in parentheses is the estimated limit of detection. A response of greater than 10 times signal-to-noise was noted for both characteristic ions (m/z 304 and 306) at the appropriate retention time for 2,3,7,8-TCDF. However, the ion ratio was considerably greater than the acceptable range of 0.61 to 0.90. �i au ic J.A. Table 11. 1,2,3.7,8-PeCDF spike level (pg/g) 0 0 0 STO RSD ( ) % 10 10 10 10 Mean STO RSD ( ) % 25 25 25 Mean STD RSD ( ) % a ND (l.l) ND ( . ) 08 ND ( . ) 08 12.2 11.5 11.9 11.4 29.2 30.1 28.5 25.5 50 51.3 55.9 55.5 54.2 2.5 4.7 C12-PeCDF absolute 2,3,4,7,8-PeCDF concentration (pg/g) (pg/g) 0 0 0 20.8 21.6 19.0 75 76 80 20.5 1.3 6.5 10 10 10 10 27.6 recovery (%) 1,2,3,7,8-PeCDD concentration (pg/g) (pg/g) 0 0 0 20.2 19.9 18.1 51 54 57 19.4 1.1 5.7 54.0 3.0 5.6 32.1 37.6 30.8 30.4 60 57 55 62 30.2 1.9 6.3 58.5 3.1 5.3 48.0 43.5 55 60 57 64 46.7 59.0 2.5 5.4 3.9 6.6 3.4 . 32.4 4.9 15.2 50 50 50 54.9 48.5 41.7 48.2 62 81 87 84 78.5 11.3 14.4 74.6 62.9 71.9 64 69.8 69.3 5.0 7.3 10 10 10 10 78.0 11.7 15.1 48.4 5.4 11.1 25 25 25 25 36.3 90 63 84 75 37.0 28.7 6.1 8.8 NO = not detected. The value in parentheses is the estimated limit of detection. 74 70 13 1,2,3,7,8-PeCDO spike level 77.0 2.6 28.3 2.0 7.0 50 50 Mean STD RSD (%) (pg/g) 13 2,3,4,7,8-PeCDF spike level 11.8 0.4 3.1 25 CJl 1,2,3,7,8-PeCDF concentration ,t versus neasureij uunueniT at IUMS ui .L ,£ , j , / ,o rex-ur f &(j , / ,o rev,Lrr „ anu x ,£. , i ,o rc^uu t j, iii Homogenized Human Adipose Tissue Sampl es ND (0.9) 0.2 21.6 Mean ro j\j 1 1«;u 25 25 25 25 50 50 50 46.1 49.3 C12-PeCDO absolute recovery (%) 72.2 72.8 58 54 56 71.6 56.0 69.7 1.6 2.3 2.0 3.6 �Table 12. in Homogenize(1 Human Adipose L ipid Matrix 1,2,3,4,7,8HxCDF spike level (P9/Q) 0 0 0 Mean STD RSD ( ) % Mean STD RSD (X) Mean STD RSD ( ) % 125 125 125 1,2,3,6,7,8HxCOF concentration 2,3,4,6,7,8HxCDF spike level 2,3.4,6.7.8HxCDF concentration 1,2,3,7,8,9HxCDF spike level 1,2,3,7,8,9HxCDF concentration 13 C12-HxCDF (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) absolute recovery () % 22.0 22.1 22.4 0 12.3 12.4 12.7 0 0 0 4.9 4.2 4.2 0 NO (0.5)a NO ( . ) 07 NO ( . ) 09 55 57 52 ND ( . ) 07 0.2 25.3 55.7 2.5 4.6 concentration 0 0 46.7 52.0 48.8 49.0 25 25 25 25 90.2 89.1 90.3 90.7 156.3 0.9 0.6 25 25 25 25 62.5 62.5 62.5 62.5 83.7 80.5 79.6 76.0 62.5 62.5 62.5 62.5 3.2 3.9 151.0 157.7 149.1 32.3 33.6 32.0 32.5 25 25 25 25 74.7 75.5 74.2 71.1 125 125 125 152.6 4.5 2.9 ND = not detected. The value in parentheses reflects the estimated limit of detection. 141.9 141.6 151.0 144.9 5.3 3.7 34.8 28.5 30.9 29.2 30.9 2.8 • 9.1 59 57 60 67 60.8 4.3 7.2 125 125 125 82.7 89.4 82.2 76.3 60 63 63 70 82.7 5.4 6.5 62.5 62.5 62.5 62.5 73.9 2.0 2.6 79.9 125 125 125 0 32.6 0.7 2.1 39.0 1.8 4.7 90.1 0.7 0.7 157.2 155.4 156.4 36.6 38.6 40.8 39.8 0 4.4 0.4 8.9 12.4 0.2 1.7 49.1 2.2 4.4 62.5 62.5 62.5 62.5 Mean STD RSD (X) 1,2,3,6,7,8HxCDF spike level 22.2 0.2 1.1 25 25 25 25 no en 1,2,3,4,7,8HxCDF 64.0 4.2 6.6 143.0 144.1 163.4 57 54 57 150.1 11.4 7.6 56.0 1.7 3.1 �Table 13. in Homogenized Human Adipose Lipid Samples 1 1,2,3,4,7,8-HxCDD concentration 1,2,3,7,8,9-HxCDD 1,2,3,7,8,9-HxCDD concentration spike level concentration (pg/g) 1,2,3,6,7,8-HxCDD spike level 1,2,3,6,7,8-HxCDO spike level (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) 0 21.6 22.7 20.3 0 0 0 157.0 162.0 154.0 0 0 0 19.1 26.0 2,3,4,7,8-HxCDD 0 0 Mean STD RSD ( ) % 25 25 25 25 Mean STD RSD ( ) % 62.5 Mean STD RSD ( ) % 82.9 96.7 90.3 81.9 141.1 150.8 146.0 145.9 4.9 3.3 184.0 165.0 198.0 193.0 62.5 62.5 62.5 62.5 220.0 239.0 220.0 220.0 288.0 299.0 266.0 284.3 16.8 5.9 63.5 46.1 40.8 57.3 65 61 65 70 51.9 10.4 20.0 65.3 3.7 5.6 3.7 25 25 25 25 125 125 125 114.0 99.9 101.2 107.4 64 66 66 77 105.6 6.5 6.1 62.5 62.5 62.5 62.5 224.8 9.5 4.2 125 125 125 58 60 58 58.7 1.2 2.0 23.2 185.0 14.5 7.9 87.9 7.0 7.9 125 125 125 Mean STD RSD ( ) * 25 25 25 25 51.6 2.6 5.1 62.5 62.5 62.5 ro --J 47.8 52.0 53.7 52.7 C12-HxCDD absolute recovery (%) 15.8 157.7 4.0 2.6 21.5 1.2 5.5 24.7 13 68.3 5.9 8.7 141.3 152.7 189.4 62 57 61 161.1 .25.2 15.6 60.0 2.6 4.4 �Table 14. Spiked Versus Measured Concentrations of 1,2,3,4,6,7,8-HpCDF, 1,2,3,4,7,8,9-HpCDF, and 1,2,3,4,6,7,8-HpCDD in Homogenized Human Adipose Lipid Samples 1 ?,3,4,6,7,8-HpCDF 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF spike level concentration (pg/g) 0 0 0 Mean RSO ( ) % 25 25 25 25 Mean STD RSD ( ) % ro 00 (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) 30.6 27.3 28.6 0 0 0 NO (1.3)* NO ( . )a 11* NO ( . ) 10 0 0 0 214.7 210.8 215.5 71 74 69 213.7 2.5 1.2 71.3 2.5 3.5 214.7 239.0 243.9 248.5 83 75 76 78 243.3 4.0 1.7 78.0 3.6 4.6 281.5 288.1 269.9 274.9 77 84 90 99 278.6 8.0 2.9 87.5 9.3 10.7 353.2 355.4 343.7 62 61 69 350.8 6.2 1.8 64 4.4 6.8 48.6 48.6 51.0 56.3 ND ( . ) 11" 0.2 13 25 25 25 25 Mean STD RSD ( ) * 86.3 85.7 80.9 83.4 62.5 62.5 62.5 62.5 84.1 2.5 2.9 125 125 125 154.5 157.3 154.4 155.4 1.6 1.0 ND - not detected. 26 23 23 25 25 25 25 25 24.1 1.4 5.7 51.1 3.6 7.0 62.5 62.5 62.5 62.5 C12-HpCDD absolute recovery ( ) % spike level 1,2,3,4,6,7,8-HpCDD spike level 28.9 1.6 5.7 STO l3 1,2.3,4,7,8,9-HpCDF concentration 60 60 59 58 62.5 62.5 62.5 62.5 59.0 0.8 1.3 125 125 125 119 126 121 122.4 3.7 3.0 The value in parentheses is the estimated limit of detection. 125 125 125 1,2,3.4,6,7,8-HpCDD concentration �Table 15. Spiked Versus Measured Concentrations of OCDF and OCDD in Homogenized Human Adipose Lipid Samples 13 OCDF spike level OCDF concentration spike level OCDD concentration (pg/g) (pg/g) (pg/g) (pg/g) 0 0 0 4.9 0 0 0 804 833 781 88 94 806.1 26.1 3.2 91.0 3.0 3.3 Mean STD RSD (%) 50 50 50 50 Mean STD RSD (%) 125 125 125 125 Mean STD RSD ( ) % 250 250 250 Mean STD RSD ( ) % 2.3 2.6 OCDD 3.2 1.4 43.1 44.2 45.0 45.9 49.6 50 50 50 50 849 856 876 860 860.4 11.4 46.2 2.4 5.2 111.1 107.8 110.7 113.1 1.3 125 125 125 125 () % 91 100 87 91 91 92.3 5.5 6.0 932 934 944 90 96 102 907 110.7 2.2 2.0 227.8 231.0 228.6 C12-OCDD abolute recovery 104 929.1 15.7 1.7 250 250 98.0 6.3 6.5 1,080 1,140 1,070 1,096 34.7 3.2 250 229.1 1.7 0.7 29 69 67 74 70.0 3.6 5.2 �Umpiked Human Adipose Tiisue 13 C , i,2.: TCDF ,3.7.B-Kf>F ,?.3.4-?CDO icon .3.7.IMCDU B-fcCDr 13C| ?,3.< 8-PfCDf 8-N-CDD 'Vl' . 1,2.: tt. t.2,3 ,3.6,7,B-M.COD 9-Hi-COlJ 9'lliiCDf 7.B-HpCDF 7,i-HpCt>n ,3,«.4.7,ft-hpCDD .7.3.4,«,7,B-HpCDD ,7,B,9-HpCDF 75 OCOO 36 '3C17-OCDO 37 OCDf R1C 21,22.23 16.17 1B19 W^W**< JlJVWsMM. izee Calibrolion Slaiidorcl 1,12 RIC 15,16,17 6,7 21,22.23 9.10 roe 12M 14m 16M 2«M SOW Figure 1. Comparison of the HRGC/MS-SIM reconstructed ion chromatogram (RIC) from the analysis of unspiked homogenized human adipose tissue matrix and a calibration standard for PCDDs and PCDFs. 30 �Umpik.d Human Adipau 384 . Spiked Human Adipaic 2.3.7,8-TCDF 304 . 9M nee 958 use 1286 1256 SCAN Umpiktd Humon Adipou 328 . r !3 C|2-lCDF ^2.3,7,8-ICDO A. A. iee.e-i Spiked Humon Adipoie 32« . >3C|2-2.3.7,8-TCDFi 9flO tew -2.3.7,8-TCDD urn 1280 1230 SCAN Figure 2. Example of the TCDF (m/z 304) and TCDD (m/z 320) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. The spiked concentrations for TCDF and TCDD in these chromatograms were 25 pg/g each. 31 �Urapik.d Human Adlpow 2.3.4.7.B-P.CDF 338 . 1.2.3,7.8-P.CDF \ Spiked Human Adipnia 2.3.4,7,8-P.CDF 1.2,3,7,8-P.CDF-, 338 . nee use 12*9 izse I3*e t3se 1400 SCMN Unspiktd Human Adipose 3C|2-I.2.3.7,8-P.CDF 354 . (1.2,3,7.88-P.CDD .. A . Spik«d Human AdipoM 354 . .2.3.7.8-P.COD 1308 13M I4W SCAN Figure 3. Example of the PeCDF (m/z 338) and PeCDD (m/z 354) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. The spiked concentrations for each isomer of PeCDF and PeCDD in these chromatograms were 25 pg/g. 32 �Umpiked Human Adipose |l.2,3,<.7.8-HxCDf h,2.3.6.7.8-HxCDF 374 . 2.3.4.6,7,8-HxCDF Spiked Human Adipote l.2,3.<.7,8-HxCLIF | | l.2.3.6.7.8-HxCDF r2,3,4,6.7.6-. \H»CDF 374 . 1.2,3.7,8.9-HxCDF 1343 1238 1366 1363 1358 1558 U_B9 1468 MSB 1588 1558 1668 SMI Umpiked Human Adipole 1.2,3.6.7.B-HxCDD 1,2.3,4.7.8-llxCDD- 339 . '3C)2-l,2.3,4,7,8-hxCDF A 1,2.3.7.8,9-Hx Spiked Human Adi ,2.3,6.7.fl-HxCDD 3M . 1.2.3,4,7.8-HxCDD \ 1.2,3,7.6.9-HxCDD l3C|2-l,2.3,<.7,8-HxCDFl 1398 1356 1488 1450 1588 1558 16!« 1E58 SOW Figure 4. Example of the HxCDF (m/z 374) and HxCDD (m/z 390) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. The spiked concentrations for each isomer of HxCDF and HxCDD were 62.5 pg/g. 33 �Umpik.d Human AdlpoM ,2,3.4.6,7,8-HpCDF 468 . Splk.d Human Adipotl l,2.3.4.6.7.8-HpCDF 488 . Jl.2.3.4.7.8,9-HpCDF MM ISM 1556 1CM I65B 1768 [750 18M SCAN Umpikcd Human Adipau 1,2.3.4.6,7,8-HpCDD 424 . Spik.d Human Adipoit 1.2,3,4.6.7,8-MpCDO 424 . 1588 1558 1688 1638 17W 1758 1868 1858 SCtH Figure 5. Example of the HpCDF (m/z 408) and HpCDD (m/z 424) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. The spiked concentrations for each isomer of HpCDF and HpCDD were 62.5 pg/g. 34 �ll3c,2-OCDD Unipik.d Human Adipotc 442 . Spik.d Hunan Adipau 13 C|2-OCDD| AJ 17W 1758 ieee ISM 1658 1358 2988 SCAM Untpikad Human Adipose OCDD 458 . Spik.d Human Adipow OCOD 458 . 1968 2MB SCAN Figure 6. Examples of the OCDF (m/z 442) and OCDD (m/z 458) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. The spiked concentrations for OCDD and OCDF were 125 pg/g each. 35 �80 70 - Regression Line- t» tt a vx d o •H +> 60 J 95% Confidence Limifs for Regression Line 50 -J td 0 CO a ti o u 30 - 95% Confidence Limifs for Individual Analyses 10 0 I 20 10 D 30 i 40 Spiked Concentration (pg/g) lipid sample meas. Figure 7. Measured concentrations versus concentrations of 2,3,7,8-TCDD spiked into the homogenized human adipose lipid matrix. 50 �80 70 - Regression Line- ue (4 95% Confidence Limits for Regression Line 60 - o •H 50 - d 4) 0 GO -vl d 0 u •d d 4:0 - 30 - •95% Confidence Limits for Individual Analyses 10 i 10 r 20 D Figure 8. r 30 i 40 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,7,8-PeCDD spiked into the homogenized human adipose lipid matrix. 50 �160 150 Regression Line- 140 130 C4 a 95% Confidence Limits for Regression Line 130 110 - o •H 100 90 - d 80 - fl 0 70 - 0 0 CO CO u 60 - •a d 50 95% Confidence Limits for Individual Analyses 40 30 20 10 0 I 40 I 20 D Figure 9. T 60 T 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDD spiked into the homogenized human adipose lipid matrix. 120 �a v_/ o *H 55 u ti CO o u •d fl 330 310 300 390 380 370 360 350 340 330 330 310 300 190 180 170 160 150 140 H 130 Regression Line. 95% Confidence Limitsfor Regression Line 95% Confidence Limits for Individual Analyses I 40 30 D i 60 80 100 Spiked Concentration (pg/g) lipid sample meas. Figure 10. Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDD spiked into the homogenized human adipose lipid matrix. 130 �(4 M a 0 •H 4> ti u 0 0 d 0 u •0 C 200 190 180 170 160 150 140 130 120 110 100 90 80 H 70 60 50 40 30 20 10 0 -10 Regression Line 95% Confidence Limitsfor Regression Line -i - •95% Confidence Limits for Individual Analyses i 40 l 20 D Figure 11. i 60 i 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDD spiked into the homogenized human adipose lipid matrix. 120 �Lfr Fovmd. (Q c n> T3 (t) _.. 0) 7T- (/> ro c a. -? n> - . Q. 1 3 r^ n o o r*- o 3- n> fD 3 c* 3" -5 o a> 3 r+ 0 -"• CQ o fD 3 3 tn N < fD fD O. -i (ft 3" C c in 01 n 3 o 3 o> n a. n> O -5 v> Q> fD r+ _j. —' O ->• 3 •O V> Q. O -h Q) I-1 -5 ro 00 •o o a a a Ul •a I—i M« •-" te* us SP- Sg 3 S •Eg a> 5*i BR o jr- ft O M 9 •o 9Q m Concentration �1.15 Regression Line- 1.1 M ft 95% Confidence Limitsfor Regression Line 1.05 1 0 n •H _l 41 ti H C 0.95 g 0 0.9 U~ 0.85 -| (4 ti ~ •a 3 •95% Confidence Limits for Individual Analyses 0.8 0 0.75 0.7 I 40 80 D 120 160 200 Spiked Concentration (pg/g) lipid sample meas. Figure 13. Measured concentrations versus concentrations of OCDD spiked into the homogenized human adipose lipid matrix. T 340 �60 Regression Line- 50 bl 95% Confidence Limitsfor Regression Line a d o 40 - 30 ti 0 GO a d o u •0 d •95% Confidence Limits for Individual Analyses 10 - 0 I 10 I 20 n Figure 14. 30 40 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 2,3,7,8-TCDF spiked into the homogenized human adipose lipid matrix. 50 �60 Regression Line' 50 H 95% Confidence Limitsfor Regression Line M a 40 -^ c 0 30 fl 0 0 C 20 - 0 u •d a 95% Confidence Limits for Individual Analyses 10 - -10 D Spiked Concentration (pg/g) lipid sample meas. Figure 15. Measured concentrations versus concentrations of 1,2,3,7,8-PeCDF spiked into the homogenized human adipose lipid matrix. �90 Regression Line- 80 - tfl d o d 95% Confidence Limitsfor Regression Line 70 - 60 50 - 0 0 cn d u 0 30 •95% Confidence Limits for Individual Analyses 10 I I i 10 20 30 D Figure 16. i 40 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 2,3,4,7,8-PeCDF spiked into the homogenized human adipose lipid matrix. 50 �160 150 140 M Regression Line- 130 95% Confidence Limitsfor Regression Line ti o 110 100 90 - ti 0) o ti 0 u •d fl 80 70 60 - •95% Confidence Limits for Individual Analyses 50 40 30 30 10 i 40 T 20 D Figure 17. i 60 T 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �160 150 -^ 140 M 130 - (4 130 - a 0 •H +> (fl u Regression Line 95% Confidence Limitsfor Regression Line 110 100 90 80 - 0 70 - 0 60 - ti u •d ti 50 •95% Confidence Limits for Individual Analyses 40 30 30 10 J 0 0 I 30 I 40 D Figure 18. i 60 r 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 130 �/-s tt d o 0 0 CO ti 0 u •d a 160 150 140 130 130 110 100 90 80 70 60 50 40 30 20 10 0 -10 - 0 Regression Line- 95% Confidence Limitsfor Regression Line 95% Confidence Limits for Individual Analyses 20 40 D 60 80 100 Spiked Concentration (pg/g) lipid sample me as. Figure 19. Measured concentrations versus concentrations of 2,3,4,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �170 Regression Line ttt 95% Confidence Limits for Regression Line * ft o •H 0 0 a 0 u •d fl 95% Confidence Limits for Individual Analyses 100 D Spiked Concentration (pg/g) lipid sample meas. Figure 20. Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �130 120 - 110 01 100 - Regression Line- 90 - o •H 4* 95% Confidence Limitsfor Regression Line 80 70 60 - 4) 0 en O 50 - o 0 40 - TJ ti 30 - a 95% Confidence Limits for Indjvidual Analyses 20 10 H 0 -10 0 i 40 I 20 D i 60 i 80 100 Spiked Concentration (pg/g) lipid sample meas. Figure 21. Measured concentrations versus concentrations of 1,2,3,4,7,8,9-HpCDF spiked into the homogenized human adipose lipid matrix. 120 �Found Concentration. �240 220 200 U Regression Line > 180 - 95% Confidence Limitsfor Regression Line 160 - ti o 140 120 - d 100 -i fi 0 80 - 0 0 en ro u •d ti •95% Confidence Limits for Individual Analyses 60 40 20 - -20 r 80 I 40 0 D Figure 23. 120 160 200 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of OCDF spiked into the homogenized human adipose lipid matrix. 240 �The slopes of the calculated regression lines from the data points in each of the 14 analyses can be used as an indication of the accuracy of the analytical method for the 17 target analytes. Figure 24 is a plot of the slope of regression lines versus the 17 individual compounds. Table 16 provides a key to specific compounds associated with a number on the x-axis of this plot. The plot presents the estimated slope from each least squares regression line as well as the upper and lower 95% confidence limits for the slope. The slope of the regression line can be interpreted as a measure of accuracy with a value of 1.00 equivalent to 100% agreement of the measured concentration with the theoretical values (background plus spike level). The plot of the 95% confidence limits presents some confirmation on the precision of measurements across the four spike levels. These confidence bounds can also be used to determine whether the accuracy of the measurements (slope of regression line) is significantly different from 100% (or 1.00). If the vertical line connecting the lower and upper 95% confidence limits intersects with the horizontal line at 1, then the accuracy of the method (as determined from the regression line) is not significantly different from 100% (slope = 1.00). The results plotted in Figure 24 demonstrate that the method accuracies for 7 of the 17 analytes are not significantly different from 100%. On the other hand, if the upper and lower confidence limits are both greater than or both less than 1.00, then the accuracy of the method is significantly different from 100%. The data presented graphically in Figure 24 indicate that some positive bias (greater than 100%) is associated with the method accuracies for 9 of the 17 analytes while the measurements for a single analyte (OCDF) result in a slightly negative (less than 100%) bias. Table 16 provides a key to the compound identification in Figure 24 and tabulates the slope of the regression lines and the upper and lower 95% confidence limits for each of the 2,3,7,8-substituted PCDD and PCDF analytes. As noted from Table 16, method accuracy (as defined by the regression line slope) ranges from 90% for OCDF up to 121% for 1,2,3,7,8,9-HxCDF. The accuracies for all other measurements fall within a range of 97 to 115%. The overall method accuracies meet the initial accuracy objective of 50-115% identified in the project quality assurance program plan. However, the predicted accuracy results for individual analysis as defined by the 95% upper confidence limits indicate that this range should be adjusted to 50-130%. The bias in the accuracy of the measurements may be a result of slight differences in the concentration calibration standards and the internal quantisation standard and native PCDD and PCDF spiking solutions. As a preliminary check on these differences, solutions of the low level and of the high level native spike combined with the internal quantitation standards were analyzed. The results of these analyses are provided in Table 17. Accuracy was calculated as measured/spiked x 100. The results of these analyses suggest that bias observed in overall method accuracy is attributed to the differences in the spiking solutions versus the calibration standards. For instance, the four HxCDF isomers demonstrated a consistent positive bias to method accuracy based on the least squares regression analysis. The analysis of the spiking solutions, submitted as samples, also indicates a definite positive bias for the same four HxCDF isomers. 53 �ACCURACY ESTIMATES 1.3 - 1.2 n M £ o> « II -I- 1.1 - §• II :TT + 0.9 - 0.8 4(014(024(034(044(054(064(0741084(094110#11 4(124(134>144(154(1 64(17 D slope estimate Compound Number 4 lower 95* CL o upper 953C CL Figure 24. Method accuracy estimates as determined from the slopes of the least squares regression lines for the 17 target PCDD and PCDF analytes. Refer to Table 16 for the key to compound number. 54 �Table 16. Regression Line Slopes with 95% Confidence Limits Significantly Lower 95% different confidence Compound no. Compound Upper 95% confidence Slope from 1.00? limit limit 1.08 1.13 1.07 0.98 yes yes yes no 1.04 1.08 1.02 0.82 1.11 1.19 1.12 1.13 01 02 03 04 2,3,7,8-TCDF 2,3,7,8-TCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 05 1,2,3,7,8-PeCDD 1.04 no 0.98 1.11 06 07 08 09 10 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1.07 1.12 1.12 1.21 0.98 yes yes yes yes no 1.06 1.09 1.09 1.12 0.92 1.09 1.16 1.15 1.29 1.05 11 1,2,3,6,7,8-HxCDD 1.01 no 0.86 1.16 12 13 14 15 16 17 1,2,3,7,8,8-HxCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD OCDF OCDD 1.12 1.01 0.97 1.08 0.90 1.15 no no no yes yes yes 0.94 0.95 0.95 1.01 0.88 1.00 1.30 1.07 1.00 1.16 0.92 1.30 55 �Table 17. Compound Results of the Analysis of the Low and High Level Native Spike Solutions Low level spike Measured Spike concentration concentration (pg/uL) (pg/uL) High level spike Accuracy Spike concentration (pg/uL) Measured concentration (pg/uL) Accuracy 108 114 2,3,7,8-TCDF 2,3,7,8-TCDD 13 12 130 120 50 50 54 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 10 10 10 11 10 12 110 1,2,3,7,8-PeCDD en 10 10 100 120 50 50 50 52 49 53 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 25 25 25 25 25 25 25 27 32 35 40 27 24 125 125 125 125 125 125 134 137 152 39 108 128 140 160 108 96 156 125 123 132 148 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD 25 25 25 22 25 26 88 100 104 125 125 125 116 121 130 97 104 OCDF OCDD 50 50 44 48 88 96 250 250 247 250 99 100 57 185 104 98 106 107 110 122 148 98 106 118 93 �Similar trends are noted for other compounds in Table 17 compared to the data presented in Figure 24 and Table 16. The limited number of analyses of the spiking solutions does not provide an adequate comparison with the sample data to confirm the bias. However, it is recommended that at least triplicate measurements of the spiking solutions at each fortification level should be analyzed at the outset of the actual NHATS sample analysis program. This will be necessary to account for any biases that will be observed from the determination of PCDD and PCDF residue levels in spiked QC samples. It should be noted that additional homogenized spiked samples will be prepared prior to initiation of the NHATS sample analyses. 1. Recovery of Internal Quantitation Standards The absolute recoveries for the carbon-13 labeled internal quantitation standards were determined for each sample by comparing the responses to the internal recovery standard, 13C12"1,2,3,4-TCDD. The average recoveries of the compounds13in Tables 10 to 15 range from 52.1% for 13C12-2,3,7,8-TCDD up to 88.9% for C12-OCDD. The results for the absolute recoveries compared to the overall method accuracy for each compound indicate the importance of the internal standard quantisation technique for analysis of the PCDDs and PCDFs in human adipose. 2. Estimation of Background Levels of PCDDs and PCDFs The estimated background levels of the various PCDD and PCDF congeners were determined as the intercept obtained from the least squares linear regression analyses. Table 18 provides a comparison of the average measured values for the unspiked matrix and the background concentration estimates from linear regression analysis of the data. In general, the measured and estimated background levels are in good agreement. However, several analytes with concentrations of less than 5 pg/g, particularly OCDF, demonstrate some disagreement in the measured versus estimated concentrations. This apparently arises from the fact that the first spike level is significantly greater than the actual background concentration. In the case of OCDF, the first spike level estimated by linear regression was 50 pg/g compared to an average measured value of 3.2 pg/g. In order to provide a better estimate of the background level based on the linear regression analysis, additional spike levels between 5 and 50 pg/g would be required. Table 18 also provides the upper and lower 95% confidence limits for the background levels estimated by the linear regression analysis of the data. These estimated background levels and confidence limits can be viewed as the intersections of the regression line and its upper and lower 95% confidence bounds, respectively, with the y-axis (measured or found concentration). These values will be used as the initial data points for developing control charts of the unspiked lipid matrix which will be analyzed with each batch of samples throughout the EPA/VA study. 57 �Table 18. Background Level Estimates with 95% Confidence Limits Compound no. 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 Measured background level (pg/g)a Compound 2,3,7,8-TCDF 2,3,7,8-TCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2, 3,4,7, 8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8HpCDF 1,2,3,4,7,8,9HpCDF 1,2,3,4,6,7,8HpCDD OCDF OCDD ND ( . ) 41C 11.5 ND (0.9) 20.5 19.4 22.2 12.4 4.4 ND (0.7) 21.5 157.75 23.2 28.9 Estimated background level (pg/g) Level significantly different from zero? Lower 95% confidence limit Upper 95% confidence limit (pg/g) (pg/g) 4.5 13.3 2.5 26.4 21.5 23.5 13.7 6.5 8.2 29.3 169.5 38.7 29.6 (3.6) 11.8 ND (1.1) 22.2 19.8 22.4 11.2 4.3 ND (2.3) 24.9 159.4 26.5 25.7 yes yes no yes yes yes •j yes •J yes 2.7 10.4 -0.2 17.9 18.2 21.3 8.6 2.1 -3.6 20.5 149.3 14.3 21.9 ND (0.3) no -2.1 1.5 213.7 214.0 yes 208.9 219.1 3.2 806.1 ND (1.0) 799.7 no yes -1.7 779.4 3.8 820.0 ND (1.1) ND yes *j no yes •j yes yes .The measured background levels are the averages of the triplicate analyses of the unspiked matrix. The estimated background levels were derived from the linear regression analysis of data. ND = not detected. The value in parentheses reflects the estimated method detection limit. �3. Day-to-Day HRGC/MS Analysis Precision In addition to the analysis of the replicate spiked samples, four extracts were analyzed by HRGC/MS on two different dates. The results of the duplicate HRGC/MS analyses of these four samples for the 17 target compounds are presented in Tables 19 to 21. Concentration values from the second analysis date were included in the statistical analysis of data presented earlier in this section. 59 �Table 19. Day-to-Day Precision of Analysis of Specific Sample Extracts for Tetra- and Pentachloro PCDF and PCDD Analysis date 4/22/86 4/28/86 Spike level (pg/g) (pg/g) 0 0 RPD ( ) %b 4/22/86 4/28/86 0 0 RPD ( ) % ND (3.1) ND (4.1) 28 0 0 ND (3.3) ND (4.0) 1,2,3,7,8PeCDF 2,3,7,8TCDD 2,3,4,7,8PeCDF 1,2,3,7,8PeCDD (pg/g) (pg/g) (pg/g) (pg/g) 10 ND ( . 4 08) ND (1.12) 24 21 18 20 13 11 23 22 17 20 4 16 16 19 17 18 11 10 29 ND (0.78) ND (0.75) 10 11 4 10 ND (0.81) ND (0.75) 10 13 19 RPD (%) 4/22/86 4/28/86 ND ( . ) 30a ND (4.1) 29 RPD ( ) % 4/22/86 4/28/86 2,3,7,8TCDF 10 10 26 8 17 6 12 14 21 23 11 12 26 28 27 19 18 9 8 7 35 .ND = not detected. Value in parentheses is the estimated limit of detection. Relative percent difference. Calculated as the difference of the two values divided by the mean of the two values times 100%. 60 �Table 20. Day-to-Day Precision of Analysis of Specific Sample Extracts for Hexa- and Heptachloro PCDF and PCDD Analysi s date 4/22/86 4/28/86 Spike level 1,2,3, 4,7,8HxCDF 1,2,3, 6,7,8HxCDF 2,3,4, 6,7,8HxCDF 1,2,3, 7,8,9HxCDF 1,2,3, 6,7,8HxCDD 1,2,3, 7,8,9HxCDD 1,2,3,4, 6,7,8HpCDF 1,2,3,4, 7,8,9HpCDF 1,2,3,4, 6,7,8HpCDD (pg/g) 1,2,3, 4,7,8HxCDD (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) (pg/g) 0 0 21 22 12 12 5 0 22 22 12 12 0 0 20 23 12 13 14 8 10 44 47 36 37 6 3 RPD ( ) %b 4/22/86 4/28/86 0 0 RPD ( ) % 4/22/86 4/28/86 0 0 RPD ( ) % 4.2 4.9 ND (0.33)a ND (0.51) ND (0.83) ND (0.36) 210 216 20 22 149 170 16 19 28 31 10 13 17 10 21 23 150 178 18 26 25 27 17 36 8 21 20 176 171 18 25 27 29 71 5 3 33 7 116 3 28 32 30 35 43 49 171 181 46 63 46 49 23 25 235 241 13 15 13 6 31 6 8 3 15 4.0 4.2 5 3.9 4.3 43 ND (0.32) ND (0.74) 79 ND (0.41) ND (0.86) 9 ' 79 ND ( . 9 10) ND (1.14) 4 ND (1.06) ND (0.28) 3 207 211 2 223 216 CTl 4/22/86 4/28/86 RPD ( ) % a 25 25 . Relative percent difference. Calculated as the difference of the two values divided by the mean of the two values times 100%. �Table 21. Day-to-Day Precision of Analysis of Specific Sample Extracts for OCDF and OCDD Analysis date OCDF concentration OCDD concentration (pg/g) (pg/g) (pg/g) 0 0 4.9 3.5 811 810 Spike level 4/22/86 4/28/86 RPD ( ) %a 33 2.2 2.3 0 0 4/22/86 4/28/86 4 RPD ( ) % 1.9 2.6 0 0 4/22/86 4/28/86 RPD ( ) % 2 788 784 1 43 44 834 848 2 50 50 RPD ( ) % . 819 836 31 4/22/86 4/28/86 a 0.1 2 . values divided by the mean of the two values times 100%. 62 �VI. QUALITY ASSURANCE/QUALITY CONTROL (QA/QC) As discussed in the experimental section of this report, the QA/QC activities included the analysis of a multipoint calibration curve, daily verification of relative response factors for each analyte, analysis of a method blank and reagent blanks along with the samples, and determining the absolute recoveries of each of the internal quantitation standards for every sample. Each of these QA/QC activities is discussed below. A. Initial Calibration At the outset of sample analysis activity, six calibration concentration standards containing each of the target PCDDs and PCDFs at varying levels and constant concentrations of the internal quantitation and recovery standards were analyzed in triplicate. The relative response factors (RRF) for each native compound and internal quantitation standard were determined for each standard analysis. An average RRF and relative percent standard deviation (RSD) were determined for each concentration level. The average RRF values from each of the six concentration calibration standards were then used to calculate a grand mean RRF value for each compound in the calibration solution. Table 22 presents a summary of the grand mean RRF values for each component in the standards. As noted from Table 22, the average RRF values for native PCDDs and PCDFs generally varied by less than ± 10% (RSD) with the exception of the pentachloro congeners. These results fall well within the criteria established in the draft quality assurance program plan which required the variability of RRF values for the tetrachloro homologs to be within ± 20% (RSD) while the RRF criterion on all other compounds was set at ± 30% (RDS). The variability of the RRF values for the internal quantitation standards, on the other hand, was noted to increase with the degree of chlorination. This is a result of the measurement of all internal quantitation standards versus the single internal recovery standard, 13C12-1,2,3,4-TCDD. A second internal recovery standard, 37Cl4-l,2,3,4,6,7,8-HpCDD, was evaluated. However, problems resulting from contribution of native HpCDD to the characteristic ions of this internal standard resulted in variabilities in the RRF value up to 50%. Hence, this internal standard was not used for any calculations. It is anticipated that an additional internal recovery standard, such as 13C12-l,2,3,4,7,8-HxCDD, will reduce the variability in the RRF values of the higher chlorinated internal quantitation standards. This compound will be incorporated into the method if available. The sensitivity of the Kratos MS-50TC to the tetra- through octachloro PCDDs and PCDFs was demonstrated through the triplicate analysis of the low level standard (CS-8, Table 2) that ranged in concentration from 1 pg/pL for the tetra- and pentachloro congeners up to 5 pg/pL for the octachloro congeners. These solution concentration values of 1 pg/pL and 5 pg/uL correspond to residue levels in tissue of 1 pg/g and 5 pg/g, respectively. Table 22 provides an indication of the observed signal-to-noise ratio for each of the native PCDD and PCDF congeners. These data demonstrate that the low level standard is well above the instrument detection limit, which is defined as the amount of a particular compound necessary to give a signal 2.5 times the background signal to noise for each of the characteristic ions while meeting the qualitative criteria for ion ratios. 63 �Table 22. Relative Response Factors (Grand Means) Determined from Multipoint Concentration Calibration Standards Compound Wa 1.00 0.80 0.98 1. .06 1. ,33 0.94 0.93 0.86 0.86 1. ,31 1. ,44 1. ,61 2. 33 1. .89 ,19 1. 1. ,38 1.04 RRF controj RSD (%) limits 2,3,7,8'-TCDD 1,2,3,7 ,8-PeCDF 2,3,4,7 ,8-PeCDF 1,2,3,7 ,8-PeCDD 1,2,3,4 ,7,8-HxCDF 1,2,3,6 ,7,8-HxCDF 2,3,4,6 ,7,8-HxCDF 1,2,3,7 ,8,9-HxCDF 1,2,3,4 ,7,8-HxCDD 1,2,3,6 ,7,8-HxCDD 1,2,3,7 ,8,9-HxCDD 1,2,3,4 ,6,7,8-HpCDF 1,2,3,4 ,7,8,9-HpCDF 1,2,3,4 ,6,7,8-HpCDD OCDF OCDD 13 C12-l,2,3,4-TCDD ia C12-2,3,7,8-TCDD ia Ci2-l,2,3,7,8-PeCDD ia Ci2-l,2,3,6,7,8-HxCDD ia C12-l,2,3,4,6,7,8- HpCDD *C14-1,2,3,4,6,7,8- 3 5.7 6.2 5.2 10.1 11.3 3.1 2.4 2.7 6.9 4.9 3.0 1.0 4.0 3.6 4.7 3.3 2.5 0.80-1.20 0.64-0.96 0.68-1.28 0.74-1.38 0.93-1.73 0.66-1.22 0.65-1.21 0.60-1.12 0.60-1.12 0.92-1.70 1.01-1.87 1.13-2.09 1.63-3.03 1.32-2.46 0.83-1.55 0.97-1.79 0.73-1.35 0.70 1.28 0.41 0.33 7.6 4.7 4.5 7.7 15.8 19.6 25.8 1.58-2.38 1.38-2.08 0.95-1.77 0.49-0.91 0.90-1.66 0.29-0.53 0.23-0.43 0.12 51.8 0.24 2,3,7,8'-TCDF 28.0 00 98 73 36 Signal-to-noise ratio for low level standard 12 6.5 11 8.9 5.7 32 30 29 17 13 14 14 35 26 13 31 21 Calibration range (pg/uL) 1-100 1-100 1-100 1-100 1-100 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 5-500 5-500 50 50 50 50 50 125 125 125 125 0.17-0.31 250 = grand mean RRF. RRF control limits designate the acceptable range of values based on the criteria for ± 20% of the KRF for 2,3,7,8-TCDD and 2,3,7,8-TCDF and ± 30% of the REP~for all other PCDD and PCDF compounds. Value for signal-to-noise ratio based on observed response for the major characteristic ion for each native PCDD or PCDF congener (Data File ,8501017X02). Internal recovery standard. 64 �B. Daily Verification of Response Factors Before proceeding with analysis of samples, the analyst was required to verify the existing response factor calibration through the analysis of a calibration standard (CS-7, Table 2). Criteria for proceeding with sample analysis required that the measured RRF value for 2,3,7,8-TCDD and 2,3,7,8-TCDF were within ± 20% (and all other congeners within ± 30%) of the mean RRF established from the calibration curve. This standard was also analyzed at the end of each working day to demonstrate that the calibration had been maintained. All RRF values were tabulated to generate RRF control charts for each specific PCDD and PCDF congener. Figures 25 through 34 are plots (control charts) of the RRF values established for the 17 individual target analytes. The RRF data are plotted versus time of analysis. These plots contain 28 individual data points, 18 of which were generated for triplicate analysis of 6 concentration calibration solutions from initial calibration and 10 analyses of solution CS-7 (Table 2) injected over the 5 days for which actual samples were analyzed. The upper and lower boundaries (dashed lines) represent a relative standard deviation of approximately ± 10% with the exception of the plot for 1,2,3,7,8-PeCDD, for which the boundaries are plotted as ± 20%. It should be noted that the actual control limits as specified in the project QAPP were set at ± 20% for 2,3,7,8-TCDD and 2,3,7,8-TCDF and ± 30% for all other target analytes. Boundaries of ± 10% have been used in Figures 25 through 34 as a means to provide the reader with a better perspective in the actual distribution of the measured calibration points. The values for the acceptable ranges of each PCDD and PCDF compound based on the initial calibration are presented in Table 22. The data presented for the RRF values in Figures 25 through 34 are well within these established control limits. The average RRF values and corresponding standard deviations reported in each of these plots are calculated from the total 28 standard analyses. C. Blanks As specified in the quality assurance program plan, a laboratory method blank was prepared along with the 14 human adipose lipid samples. The method blank was taken through all procedures as if it were an actual sample, although no lipid matrix was introduced. The analysis of the method blank resulted in the data reported for each of the target analytes reported in Table 23. As noted in Table 23, 1,2,3,4,6,7,8-HpCDD and OCDD were detected at concentrations equivalent to 4.0 and 30 pg/g (equivalent to a 10-g lipid sample), respectively. The contribution of these PCDDs were not subtracted from the observed responses for the spiked and unspiked samples. These background levels accounted for less than 2% of the 1,2,3,4,6,7,8-HpCDD and less than 4% of the OCDD measured in the unspiked lipid samples. In addition to these compounds, responses that correspond to the elution of two TCDD isomers (1,3,6,8- and 1,3,7,9-) and a PeCDD (isomer not determined) were detected in the method blank. 65 �e.33?) (AREA/REF. AREA>/(AKT. /UEF.ANT. > 1.296 1T.CEU.i ST.DEU.6.331 X MOT I ? * -x, 8.9W X DATE 4/23/8S 4/13/85 V 0.897) ST.DPJ.» e.ew 7. ST.DEV.= 6.832 8.999 X Yx |l -T > » X 8.788 DATE 4/13/85 4/23/86 3/ 3/8S Figure 25. Control charts showing response factors by date for 2,3,7,8-TCDF and 2,3,7,8-TCDD. Dotted lines represent approximately ± 10% of the mean. 66 �Kfflfttift 9.971) <f*Efl/REF.ftKEA>/<AMT./REF.AMT.> (Ml 1.299 ST.OEU.» 9.083 7. ST.DEU.« 8.739 X 1 i ? I i 1.199 IK fl* II L1 1 — "tf. >f " 1.899 -JfC S « it ¥^ $ i^ '^ S * 1 0.999 1 ! V ^ Sjf ^ - TX % X 0.S00 DATE 4/13/86 4/23/86 5/ 3/86 CMF!2,3<4,7,8-FEI(TACHLORO-FyRrtN .?EFiC13-l,2,3,7,<)-PENTScHLORO-FlJKftH (flFEfl/REF.IWEft)/ ftMT./REF.AHT.) (flUl 1.499 1.963) ST.DEU.= 3.113 7. ST.OEU. 18.601 X 1.309 X — *•* V , 1 4 i « i) * N| 1 X 0.309 >— tt x — " * X ii >L) " * M M 1.998 ? S4- X M . " , * x ^ ix v • ! ' ^ __ X 0.809 X il.Tm DATE 4/13/86 4/23/85 3/ 3/86 Figure 26. Control charts showing response factors by date for 1,2,3,7,8-PeCDF and 2,3,4,7,8-PeCDF. Dotted lines represent approximately ± 10% of the mean. 67 �«K£A/FEF.«Efl>/<«HT./REF.flHT.> 3.800 - 1.375) ST.OEU." 8.233 7. ST.DEU.18.895 2.899 !x X MTE 4/13/86 4/23/86 3/ 3/85 Figure 27. Control chart showing response factors by date for 1,2,3,7,8-PeCDD. Dotted lines represent approximately ± 20% of the mean. 68 �Cffii 1,2.3,4,7,8-HEXACHLQRO-FUIMN SEFlC13-l,2..3,4,7,8-HEX«CHLORO-FURftH (AfiEA/REF.AREAVCAMT./REF.AMT.) <flUi MM 8.323) ST.DEU.« 9.816 !! ST.DEI'.' 4.899 .ew V V •xMTE 1 5/ 3/86 4/23/86 4/13^6 !l,2,3,6,7,8-HEXft 1013-1,2,3,4,778-1 <AREfl/REF.I»Efl)/<ftMT./l?EF.flMT. B.923) nee ST.OEU.= 8.048 X ST.DEU." 4.348 -¥• i.ese X i O m K * r 0.303 * x x 3.383 DATE 4/13/86 4/23/86 1 S/ 3/86 Figure 28. Control charts showing response factors by date for 1,2,3,4,7,8-HxCDF and 1,2,3,6,7,8-HxCDF. Dotted lines represent approximately ± IQ% of the mean. 69 �12,3,4,6,7,8 8.862) i.ew ST.DEU.0.838 7. ST. DP.'. 4.429 — — ,Jfr Y 8.309 X 1 i V * ' X V »• *• i ' Th T 1 " xf | x 0.880 'Y ' " A ' X x x x 8.788 DflTE 4/13/86 4/23/86 <flRE«/REF.<*B»/<ftMT./REF.«MT.> <(Wl 1.200 -| V 3/86 8.883) ST.OEV.9.882 Z ST.DEV.' X 1.180 9 327 ' ¥ x ^ >f H 8.900 x ? 'I 1* |5* ft fi ! i 0.388 ii x I ' u ' ' ' x X T X j ^ * •" X A. 7m DATE 4/13/86 4/23/86 !/ 3/86 Figure 29. Control charts showing response factors by date for 2,3,4,6,7,8-HxCDF and 1,2,3,7,8,9-HxCDF. Dotted lines represent approximately ± 10% of the mean. 70 �1.394) 1.580 ST.DEU.0.077 7. ST.DEU.= 5.919 V V 1.4P8 ft • X •^N '* Vfl' - fJL, k- * i RX 1 f; r II J1 I f ^ •SL i 1.399 V \j/ j^. /x ! i ! | | i ^ A I i sL * * 1 -uw A i ! 1.299 1 i 1 !W DATE 4/13/85 4/23/86 AREA/REF.AREA)/«W./REF.IVIT.) (AUt 1.70B 3/ 3/8S 1.433) ST.DEU.0.036 Z ST.DEU.» 6.01S 1.609 ? i I i ~ ' V X «< III 1.509 iii f in m £ * ¥ «i il 1.400 1.308 ( * i x • •!* i • x ii i /^ 1 ' l i X X I X I . f x >< X 1 . 7fl« 4/23/86 3/ 3/86 Figure 30. Control charts showing response factors by date for 1,2,3,4,7,8-HxCDD and 1,2,3,6,7,8-HxCDD. Dotted lines represent approximately ± 10% of the mean. 71 �1.657) 2.300 > 2.200 ST.DEU.= 0.156 7. ST.DEU. 9.384 2.1?? '} ( 2.000 1.909 1.800 X I 1 X 1.700 V u 1.600 1.590 * » f 1 ;xx 1 ; !x x1 ; ! * xx ¥ X 1 X 1 400 4/23/86 S/ 3/86 Figure 31. Control chart showing response factors by date for 1,2,3,7,8,9-HxCDD. Dotted lines represent approximately ± 10% of the mean. 72 �2.228) <(«BVREF.flREfl)/<ftMT./REF.«MT.) 2.600 •X * ST.DEU.0. 132 7. ST.OEU.» 8.163 2.508 X K X 2.488 in I * x "ft* & 2.309 * aI * i |i • i ' X t ' 2.288 A x i ' x X 2.100 X X X"1 2.088 X 1.960 X DrtTE 4/ 3/86 4/23/86 CMFil, 2,3,4, 7,8,3-HEPTflCHLORO-FURfiH *EFiC13-l, 2,3,4,6 .7,8-HEPTAa*J$J-DIOXIM i'flREfl/REF.AREfl)/< «1T./R£F.«1T.) (flUl 2. 180 ^ 1.T83) X ' ' 7 J _ * 2.088 5/ 3/86 ST.DEU.' 3. 169 7. ST.OEV.9.420 /jK i< T r> i 1.300 m ^11 ^ ; - i * 1.700 '' X x x — 1.608 —L r T X 1.598 | X X X" X ^ t.4B« DATE 4/13/8S 4/23/8S 5/ 3/96 Figure 32. Control charts showing response factors by date for 1,2,3,4,6,7,8-HpCDF and 1,2,3,4,7,8,9-HpCDF. Dotted lines represent approximately ± 10% of the mean. 73 �(AREA/REF.AREA)/(ANT./REF.MIT.) (AUl 1.486 1.175) ST.OEU.8.371 7. ST.DEV." 6.043 i.aee 1.2M *'! < !x M SAC X I I x, I X X . X 1.100 e.we DATE 4/13/86 4/23/86 5/ 3/36 Figure 33. Control chart showing response factors by date for 1,2,3,4,6,7,8-HpCDD. Dotted lines represent approximately ± 10% of the mean. 74 �lOCTflCHLORO-FUf WH :CI3-OCT«HLOR(M)tOXIH WT./FEF.flMT.) (flUl 1.238) 1.5W ST.DEU." 8.117 7. ST.OEU. 8.978 X 1 X *j (1 x U HI in •M 1.40B I' * ji >i M «l tl M M M >^f 1.390 ri 1.2W X X /x -iv -1- ^^ ^ < ; < i ina xX 3/8S 4/23/85 ;H IOXIN <AKEA/REF.ftfiEfl)/<AMT.-/R£F.flt1T.) 1.280 1.033) ST.DEU.8.943 7. ST.DEU.- 4.181 ^ X 1.180 V X L-9-ff 1 x ii ^ i —i 0.900 DATE X X 4/13/86 4/23/85 1 3/ 3/86 Figure 34. Control charts showing response factors by date for OCDF and OCDD. Dotted lines represent approximately ± 10% of the mean. 75 �Table 23. Summary of Results from the Analysis of a Laboratory Method Blank Concentration Compound3 (pg/g) 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDF ND ND ND ND ND ND ND ND ND ND 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD OCDF OCDD (0.50) (2.2) (0.5) (0.5) (1.2) (0.5) (0.5) (0.5) (0.5) (1.0) ND (0.9) ND (0.8) ND (0.5) ND (0.5) 4.0 ND (0.5) 30 a At least three other PCDD compounds were detected but not quantitated in the laboratory method blank. These included 1,3,6,8- and 1,3,7,9-TCDD and an unidentified PeCDD . isomer. Concentration based on assumption of 10.0 g equivalent lipid sample. The background concentration of 1,2,3,4,6,7,8-HpCDD and OCDD were not subtracted from the measured concentration for the spiked and unspiked lipid matrix. 76 �Further analysis of individual reagents used for preparation of the samples identified the activated acidic alumina as the source of the artifacts. Acidic alumina that had been cleaned by Soxhlet extraction but not activated at 190°C was analyzed, and the artifacts were not detected. This indicates that the artifacts are generated during activation of acidic alumina at elevated temperatures (190°C). Similar background problems from the same PCDD congeners have recently been reported by the Center for Disease Control.17'18 An experiment was designed to evaluate a procedure for cleaning the activated acidic alumina immediately prior to the fractionation of the sample extract. The acidic alumina (6.0 g) was packed in hexane. The packed column was eluted with 40 mL of methylene chloride/hexane (1:1) solution followed by 80 to 100 ml of hexane. The sample extract was added to the column and was eluted with 20 ml of hexane followed by 30 ml of 20% methylene chloride in hexane which was reserved for PCDD and PCDF analysis. The carbon-14 radiolabeled 2,3,7,8-TCDD, 1,2,3,4,7,8-HxCDD, and OCDD were used to evaluate recovery of PCDDs from the cleaned alumina. Recoveries of the radiolabeled PCDDs from the activated acidic alumina precleaned by the procedure described above are detailed in Table 24. These data demonstrate that the selected PCDDs are quantitatively (greater than 90%) recovered from the precleaned acidic alumina. This procedure for cleanup of activated acidic alumina was not initiated for the analysis of the lipid samples described in this report. However, it has been integrated into the analytical protocol (Appendix A) for routine application with sample preparation activities. D. Absolute Recoveries of the Internal Quantisation Standards The absolute recoveries of the carbon-13 labeled internal quantitation standards 13were determined by comparing responses with the internal recovery standard, C12-1,2,3,4-TCDD, which was added during final concentration prior to HRGC/MS analysis. A summary of the average and range of recoveries of the 8 internal quantisation standards from the 14 human adipose lipid samples is provided in Table 25. These data indicate that recoveries ranged from an average of 52.1% for 13C12-2,3,7,8-TCDD up to 88.9% for 13C12-OCDD. The average recoveries for the lower chlorinated internal standards were lower than the preliminary method studies with carbon-14 radiolabeled standards had indicated. This resulted in a closer evaluation of the final concentration step prior to mass spectrometry. The first extracts for the human adipose lipid extracts were concentrated with a nitrogen evaporation system equipped with a water bath at approximately 55°C. Final blowdown of the samples required addition of the internal recovery standard in 10 uL of tridecane as a keeper solution. However, it was noted at the elevated temperature final volumes from nitrogen evaporation were generally on the order of 2 to 5 uL. This required addition of another 10 uL of tridecane prior to HRGC/MS analysis. In an effort to assess the effect of reducing the final volume of tridecane at elevated temperatures on absolute recoveries of the internal quantisation standards, an experiment using the radiolabeled TCDD, HxCDD, and OCDD standards was conducted. 77 �Table 24. Recovery of Radio!abeled PCDDs from Precleaned Activated Alumina Spike level (pg) Compound Recovery (%) 14 100 300 300 92 96 97 14 1,000 3,000 3,000 103 101 100 2,500 7,500 7,500 102 99 97 C-2,3,7,8-TCDD C-l,2,3,4,7,8-HxCDD 14 C-OCDD 78 �Table 25. Absolute Recoveries of the Internal Quantisation Standards from the Human Adipose Lipid Matrix Average recovery ( ) % Standard deviation Relative standard deviation (%) 13 64.0 7.9 12.3 48-78 13 52.1 5.0 9.6 43-62 13 76.1 8.9 11.7 62-90 13 57.1 3.5 6.1 51-64 13 59.4 5.0 8.4 52-70 13 63.6 5.3 8.4 57-77 13 76.3 10.3 13.6 61-99 88.9 11.4 12.9 67-104 Internal quantitation standard C12-2,3,7,8-TCDF C12-2,3,7,8-TCDD C12-l,2,3,7,8-PeCDF C12-l,2,3,7,8-PeCDD C12-1,2,3,4,7,8HxCDF C12-1,2,3,6,7,8HxCDD C12-1,2,3,4,6,7,8HpCDD 13 C12-OCDD Values based on 14 analyses of human adipose lipid samples. 79 Range of recovery (%) �Four solutions of the same spike level were prepared with each radiolabeled compound in 1 mL of toluene. Two of the spiked solutions were heated at 55-60°C and the solvent was reduced under a gentle stream of prepurified nitrogen. The toluene solution was concentrated to 100 (jL, 500 pL of 1% toluene in methylene chloride was added, and the solution was concentrated to 200 uL. At this time 10 uL of the keeper tridecane was added and the solution was allowed to concentrate further. The remaining two solutions for each radiolabeled compound were taken through a similar solvent exchange and concentration procedure except the solution was allowed to concentrate at room temperature. One of the most obvious results was the observation that solutions held at elevated temperatures could be reduced to dryness even when tridecane had been added as a keeper. On the other hand, solutions for which tridecane had been added but remained at room temperature could only be concentrated to a 10-uL final volume. The recoveries of the radiolabeled standards from each of the solutions in this study are presented in Table 26. The results from this study indicate that the final concentration condition may have a pronounced effect on the absolute recoveries of the PCDDs and PCDFs, especially for the lower chlorinated congeners such as 2,3,7,8-TCDD. These conclusions are supported by an independent study in comparison of concentration techniques for 2,3,7,8-TCDD.19 However, it should be noted that the approach to target analyte quantisation based on the internal standard method (isotope dilution for 8 of the 17 target analytes) is not affected by absolute recoveries as low as 50%. The procedure for final concentration in the analytical protocol (Appendix A) for the analysis of the NHATS samples for the EPA/VA study has been modified to specify room temperature conditions. 80 �and OCDD as a Function of Final Concentration Conditions Spike level Compound (pg) Concentration conditions Observed final volume Observed recovery ( ) % 14 300 300 300 300 55-60°C 55-60°C 20°C 20°C 1-2 |jL dryness 10 uL 10 uL 78 54 98 93 14 3,000 3,000 3,000 3,000 55-60°C 55-60°C 20°C 20°C 1-2 uL 5 uL 10 M L 10 uL 94 102 105 107 7,500 7,500 7,500 7,500 55-60°C 55-60°C 20°C 20°C 1-2 uL 2-3 M L 10 uL 10 |jL 94 94 100 97 C-2,3,7,8-TCDD C-l,2,3,4,7,8-HxCDD 14 C-OCDD Each solution was concentrated under a gentle stream of flowing nitrogen. 81 �VII. GLOSSARY OF TERMS Accuracy - A measurement of the bias of a system, which for this study, is based on the agreement of the 2,3,7,8 substituted PCDD and PCDF to an accepted reference standard. Batch, sample - A sample batch consists of up to 10 human adipose tissue samples, one method blank, 2 internal quality control (QC) samples (spiked and unspiked), and an external performance audit sample (blind spike). Blank, laboratory method - This blank is prepared in the laboratory through performing all analytical procedures except addition of a sample aliquot to the extraction vessel. A minimum of one laboratory method blank will be analyzed with each batch of samples. Calibration standards (concentration calibration solutions) - Solutions containing known amounts of the native analytes (unlabeled 2,3,7,8-substituted PCDDs and PCDFs), the internal quantisation standards (carbon-13 labeled PCDDs and PCDFs), and the recovery standard, 13C12-1,2,3,4-TCDD. These calibration solutions are used to determine instrument response of the analytes relative to the internal quantisation standards and of the internal quantitation standards relative to the internal recovery standard. Lipid - The organic solvent extractable constituents of adipose tissue consisting of fatty oils, proteins, and carbohydrates. The concentrations of PCDDs and PCDFs are reported on the lipid content bases. Instrumental mass calibration - An internal instrumental systems check and tuning standard, perfluorokerosene (PFK), is introduced automatically by the instrument. The mass ion 380.976 is monitored by the analyst as an instrumental systems check and is also used to tune the instrument. Internal quantisation standards - Carbon-13 labeled PCDDs and PCDFs, which are added to every sample and are present at the same concentration in every method blank and quality control sample. These are added to the adipose tissue prior to extraction and are used to measure the concentration of each analyte. The concentration of each internal quantitation standard is measured in every sample, and percent recovery is determined using the internal recovery standard. Internal recovery standard - 13C12-1,2,3,4-TCDD and 13C12-l,2,3,7,8,9-HxCDD which is added to every sample extract just before the final concentration step and HRGC/MS-SIM analysis. Limit of detection (LOD) - A value, derived from the noise to signal response, which is equal to 2.5 times the average instrumental noise level is the limit of detection. Limit of quantitation (LOQ) - A value, derived from the noise to signal response, which is equal to 10 times the average instrumental noise level is the limit of quantitation. 82 �Mass resolution check - Standard method used to demonstrate static resolution of 10,000 minimum (10% valley definition). Not detected (ND) - A nonresponse or a response which is less than the limit of detection is reported as not detected. Precision - The results from analysis of replicate samples (spiked and unspiked) provide the measure of method precision. The precision of the method is reported as standard deviation or relative standard deviation. Performance check mixture, HRGC column - A mixture containing known amounts of selected TCDD standards; it is used to demonstrate continued acceptable performance of the capillary column, to separate ( 25 % valley on a 50-m CP ^ Sil 88 or 60-m SP-2330 HRGC column and 30 to 60% for a 60 m DB-5 HRGC column) 2,3,7,8-TCDD isomer from all other 21 TCDD isomers, and to define the TCDD retention time window. PCDD - Polychlorinated dibenzo-p_-dioxins. PCDF - Polychlorinated dibenzofurans. Relative response factor - Response of the mass spectrometer to a known amount of an analyte relative to a known amount of an internal standard (quantitation or recovery). Trace (TR) - A response which is greater than the limit of detection but less than the limit of quantitation is reported as a trace value. An estimated method detection limit is provided for trace value. 83 �VIII. REFERENCES 1. Stanley JS. 1984. Methods of analysis of polychlorinated dibenzo-£dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in biological matrices--!iterature review and recommendations. EPA-560/584-00. 2. Stanley JS, Going JE, Redford DP, Kutz KW, Young AL. 1985. A survey of analytical methods for measurement of polychlorinated dibenzo-£-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) in human adipose tissues. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 181-195. 3. Stanley JS. 1984 (March 28). Proposed analytical method for analysis of PCDDs/PCDFs in human adipose tissue: special report. Washington, DC: Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Contract 68-02-3938, Work Assignment 8. 4. Stanley JS. 1986 (April 23). Broad scan analysis of human adipose tissue: polychlorinated dibenzo-£-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). Draft final report. Washington, DC: Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Contract 68-023938, Work Assignment 8. 5. Albro et al. 1985. Methods for the quantitative determination of multiple, specific polychlorinated dibenzo-£-dioxin and dibenzofuran isomers in human adipose tissue in the parts-per-trillion range. An interlaboratory study. Anal Chem 57: 2717-2725. 6. Patterson DG et al. 1986. High-resolution gas chromatographic/highresolution mass spectrometric analysis of human adipose tissue for 2,3,7,8-tetrachlorodibenzo-£-dioxin. Anal Chem 58: 705-713. 7. Schecter A, Tiernan TO, Taylor ML, VanNess GF, Barrett JH, Wage! DJ, Gitlitz G, Bogdasarian M. 1985. Biological markers after exposure to polychlorinated dibenzo-p_-dioxins, dibenzofurans, biphenyls, and biphenylenes. Part I: Findings using fat biopsies to estimate exposure. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 215-246. 8. Schecter A, Ryan JJ. 1985. Dioxin and furan levels in human adipose tissue from exposed and control populations. 189th National ACS Meeting Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment III, Miami, Florida. Preprint Division of Environmental Chemistry, ACS 25:160-163, Paper No. 56. 9. Ryan JJ, Williams DT, Lau BPY, Sakuma T. 1985. Analysis of human fat tissue for 2,3,7,8-tetrachlorodibenzo-£-dioxin and chlorinated dibenzofuran residues. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 205-214. 84 �10. Ryan JJ, Schecter A, Lizotte R, Sun W-F, Miller L. 1985. Tissue distribution of dioxins and furans in humans from the general population. Chemosphere 14: 929-932. 11. Nygren M, Hansson M, Rappe C, Domellof L, Hardel1 L. 1985. Analysis of polychlorinated dibenzo-p_-dioxins and dibenzofurans in adipose tissue from soft-tissue sarcoma patients and controls. 189th National ACS Meeting Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment III, Miami, Florida, 1985. Preprint Division of Environmental Chemistry, ACS 25:160-163, Paper No. 55. 12. Smith LM, Stalling DL, Johnson JJ. 1984. Determination of part per trillion levels of polychlorinated dibenzofurans and dioxins in environmental samples. Anal Chem 58: 1830-1842. 13. Rappe C, Nygren M, Linstrom G, Hanson H. 1985. Dioxins and dibenzofurans in human tissues and milk of European origin. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 14. Ryan JJ. 1985. Variation of dioxins and furans in humans with age and organ by country. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 15. Graham M, Hileman FD, Wendling J, Wilson JD. 1985. Chlorocarbons in adipose tissue samples. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 16. Patterson DG, Holler JS, Smith SJ, Liddle JA, Sampson EJ, Needham LL. 1985. Human tissue data in certain U.S. populations. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 17. Patterson DG, Holler JS, Groce DF, Alexander LR, Lapeza CR, O'Conner RC, Liddle JA. 1986. Control of interferences in the analysis of human adipose tissue to 2,3,7,8-tetrachlorodibenzo-£-dioxin (TCDD). Environ Toxicol Chem 5: 355-360. 18. Holler JS, Patterson DG, Alexander LR, Groce OF, O'Connor RC, Lapeza CR. 1985. Control of artifacts and contamination in the development of a dioxin analytical program. 33rd Annual Conference on Mass Spectrometry and Allied Topics, San Diego, CA, May 26-31, 1985. 19. O'Keefe PN, Meyer C, Dillon K. 1982. Comparison of concentration techniques for 2,3,7,8-tetrachlorodibenzo-p_-dioxin. Anal Chem 54: 26232625. 85 �APPENDIX A ANALYTICAL PROTOCOL FOR DETERMINATION OF PCDDs AND PCDFs IN HUMAN ADIPOSE TISSUE A-l �TABLE OF CONTENTS Section Description Page 1 Scope and Application A-3 2 Summary of Method A-3 3 Definitions A-6 4 Interferences A-7 5 Safety A-7 6 Apparatus and Equipment A-8 7 Reagents and Standard Solutions A-ll 8 High Resolution Gas Chromatography/Mass Spectrometry Performance Criteria 9 A-13 Quality Control Procedures A-31 10 Sample Preservation and Handling A-33 11 Sample Extraction A-33 12 Cleanup Procedures A-35 13 Analytical Procedures A-38 14 Date Reduction A-43 15 Reporting and Documentation A-50 A-2 �ANALYTICAL PROTOCOL FOR DETERMINATION OF PCDDs AND PCDFs IN HUMAN ADIPOSE TISSUE 1. SCOPE AND APPLICATION 1.1 1.2 The minimum measurable concentration is estimated to range from 1 pg/g (1 part per trillion) for 2,3,7,8-TCDD and 2,3,7,8-TCDF up to 5 pg/g for OCDD and OCDF. However, these detection limits depend on the kinds and concentrations of interfering compounds in the sample matrix and the absolute method recovery. 1.3 2. This method provides procedures for the detection and quantitative measurement of polychlorinated dibenzo-p_-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) at concentrations ranging from 1 to 100 pg/g for the tetrachloro congeners up to 5 to 500 pg/g for the octachloro congeners in 10-g aliquots of human adipose tissue. The method will be used to determine PCDDs and PCDFs, particularly congeners with chlorine substitution in the 2,3,7,8 positions. Table 1 lists the specific PCDDs and PCDFs and target method detection limits. SUMMARY OF METHOD Figure 1 presents a schematic of the analytical procedures for determination of PCDDs and PCDFs in human adipose tissue. The analytical method requires extraction and isolation of lipid materials from human adipose samples. This is accomplished using sample sizes ranging up to 10 g. The tissue is spiked with known amounts of the carbon-13 labeled PCDDs and PCDFs (e.g., 500 pg of 13C12-TCDD/F to 2,500 pg of 13C12-OCDD/F) as internal quantitation standards. Extraction and homogenization are accomplished using methylene chloride and a Tekmar Tissuemizer®. The extract is filtered through anhydrous sodium sulfate to remove water. The extraction procedure is repeated (three to five times) until the tissue sample has been thoroughly homogenized. The final extract is adjusted to a known volume (100 ml) and the extractable lipid is determined using a minimum of 1% of the final volume. The methylene chloride in the remaining extract is concentrated until only an oily residue remains. The residue is diluted with hexane ( 200 mL), and ~ 100 g of sulfuric acid modified silica gel (40% w/w) is added to the solution with stirring. The mixture is stirred for approximately 2 h, and the supernatant is decanted and filtered through anhydrous sodium sulfate. The adsorbent is washed with at least two additional aliquots of hexane. The combined hexane extracts are eluted through a column consisting of a layer of sulfuric acid modified silica gel, and a layer of unmodified silica gel. The eluate is concentrated to approximately 1 ml and added to a column of acidic alumina. The PCDDs and PCDFs are eluted from the alumina using 20% methylene chloride/hexane. This eluate is concentrated to approximately 0.5 mL and is added to a 500-mg Carbopak C/Celite column. The PCDDs and PCDFs are eluted from the column using 20 mL of toluene. A-3 �Table 1. Target PCDD and PCDF Congeners and Target Method Detection Limits Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF CAS no.3 1746-01-6 51207-31-9 40321-76-4 57117-41-6 57117-31-4 39227-28-6 57653-85-7 19408-74-3 70648-29-9 57117-44-9 72918-21-9 60851-34-5 35822-46-9 67562-39-4 55673-89-7 3268-87-9 39001-02-0 Chemical Abstract Services number. pg/g = parts per trillion. 3 A-4 Target method detection limit (pg/g) 1.0 1.0 1.0 1.0 1.0 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 5.0 5.0 �Add Internal Quantitation Standards Initial Sample Preparation Isolation of Extractable U'pid Materials (13c-PCDDs/PCDFs) Homogenization in Methylene Chloride I Lipid Determination Solvent Exchange Bulk U'pid Removal Acid Modified Silica Gel Slurry Technique Provides Cleanup of Oxidizable Compounds with Rapid Sample Turnaround, Improved Cleanup Efficiency and Recovery i Removal of Chemical Interferences Acidic Silica/Silica Acidic Alumina Carbopak C/Celite Provides Seperation of PCBs and Other Potential Interferences from PCDDs and PCDFs Selective Adsorption and Isolation of PCDDs/PCDFs Add Internal Recovery Standards HR'GC/MS-SIM Analysis 1 i LRMS Identification/Quantitation of Tetra-Octa PCDDs/PCDFs HRMS Confirmation of 2,3,7,8-TCDD Figure 1. Schematic of the sample preparation and instrumental analysis procedures for determination of PCDDs and PCDFs in human adipose tissue. A-5 �The toluene is concentrated to less than 1 ml and transferred to conical vials. Tridecane (10 uL) containing 500 pg of an internal recovery standard is added as a keeper, and the extract is concentrated to final volume. The HRGC/MS analysis is completed in the selected ion monitoring mode (SIM). Analysis of the tetra- through octachloro PCDD and PCDF congeners is achieved using low resolution mass spectrometry. Separation of the tetra- through octachloro PCDD and PCDF congeners is achieved using a 60-m DB-5 column. Verification of the 2,3,7,8-TCDD is achieved using either a 50-m CP Sil 88 column or 60-m SP-2330 column and HRGC/MS-SIM analysis in the high resolution mode (R = 10,000). 3. DEFINITIONS 3.1 Concentration calibration solutions — Solutions containing known amounts of the native analytes (unlabeled 2,3,7,8-substituted PCDDs and PCDFs), the internal quantisation standards (Carbon-13 labeled PCDDs and PCDFs), and the recovery standard, 13C121,2,3,4-TCDD. These calibration solutions are used to determine instrument response of the analytes relative to the internal quantitation standards and of the internal quantitation standards relative to the internal recovery standard. 3.2 Internal quantitation standards -- Carbon-13 labeled PCDDs and PCDFs, which are added to every sample and are present at the same concentration in every method blank and quality control sample. These are added to the adipose tissue and are used to measure the concentration of each analyte. The concentration of each internal quantitation standard is measured in every sample, and percent recovery is determined using the internal recovery standard. 3.3 Internal recovery standard — 13C12-1,2,3,4-TCDD and 13C121,2,3,7,8,9-HxCDD which is added to every sample extract just before the final concentration step and HRGC/MS-SIM analysis. 3.4 Laboratory method blank — This blank is prepared in the tory through performing all analytical procedures except of a sample aliquot to the extraction vessel. A minimum laboratory method blank will be analyzed with each batch ples. 3.5 HRGC column performance check mixture -- A mixture containing known amounts of selected TCDD standards; it is used to demonstrate continued acceptable performance of the capillary column, to separate (g 25% valley on a 50-m CP Sil 88 or 60-m SP-2330 HRGC column and 30 to 60% for a 60-m DB-5 HRGC column) 2,3,7,8TCDD isomer from all other 21 TCDD isomers, and to define the TCDD retention time window. 3.6 Relative response factor — Response of the mass spectrometer to a known amount of an analyte relative to a known amount of an internal standard (quantitation or recovery). A-6 laboraaddition of one of sam- �3.7 3.8 4. Mass resolution check — Standard method used to demonstrate static resolution of 10,000 minimum (10% valley definition). Sample batch -- A sample batch consists of up to 10 human adipose tissue samples, one method blank, 2 internal quality control (QC) samples (spiked and unspiked), and an external performance audit sample (blind spike). INTERFERENCES Chemicals which elute from the HRGC column with ± 10 scans of the internal and/or recovery standards and which produce within the retention time window ions at any of the masses used to detect or quantify PCDDs, PCDFs, or the internal quantitation and recovery standards are potential interferences. Most frequently encountered potential interferences are other sample components that are extracted along with the PCDDs and PCDFs, e.g., PCBs, chlorinated methoxybiphenyls, chlorinated hydroxydiphenyl ethers, chlorinated benzylphenyl ethers, chlorinated naphthalenes, DDE, DDT, etc. The actual incidence of interference by these chemicals depends also upon relative concentrations, mass spectrometric resolution, and chromatographic conditions. Because very low levels (pg/g) of PCDDs and PCDFs are anticipated, the elimination of interferences is essential. High purity reagents and solvents must be used and all equipment must be scrupulously cleaned. Laboratory method blanks must be analyzed to demonstrate absence of contamination that would interfere with measurement of the PCDDs and PCDFs. Column chromatographic procedures are used to remove coextracted sample components; these procedures must be performed carefully to minimize loss of PCDDs and PCDFs during attempts to increase their concentration relative to other sample components. 5. SAFETY 5.1 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. The 2,3,7,8-TCDD is a known teratogen, mutagen, and carcinogen. Ingestion of microgram quantities can result in toxic effects. The other 2,3,7,8-substituted PCDDs and PCDFs may exhibit teratogenic, mutagenic, and carcinogenic effects. From this viewpoint, exposure to these chemicals must be reduced to the lowest possible level by whatever means available. Only experienced personnel will be allowed to work with these chemicals. 5.2 All laboratory personnel will be required to wear laboratory coats or coveralls, gloves, and safety glasses. The neat standards, stock, and working solutions will be handled only in a Class A fume hood or glove box. When manipulating stock standards or working solutions, the analyst is advised to place the solution vials in a secure holder (sample block or glass beaker) to prevent accidental spills. A-7 �5.3 5.4 If handling of these compounds results in skin contact, immediately remove all contaminated clothing and wash the affected skin areas with soap and water for at least 15 min. 5.5 6. If these standards are spilled, absorb as much as possible with absorbent paper and place in a container clearly labeled as PCDD or PCDF waste. Solvent-wash all contaminated surfaces with toluene and absorbent paper followed by washing with a strong soap and water solution. Dispose of all contaminated materials in sealed steel containers labeled as contaminated with PCDD and/or PCDF residue and indicate the approximate level of contamination. As a final precaution, prepare a wipe sample of the exposed surface area and include the wipe as part of the sample analysis batch. This will be used to confirm that the work area is free of contamination. Disposal of laboratory wastes -- All laboratory wastes (solvents and absorbents) will be disposed of as hazardous wastes. The laboratory personnel should take care to dispose of the sodium sulfate, silica gel, and alumina in separate containers. Excess solvents should be disposed of in gallon polyethylene jugs containing a layer of activated charcoal. Excess solvent that is known to be contaminated with PCDDs or PCDFs should be kept at a minimum by evaporating the solvent with a stream of air. APPARATUS AND EQUIPMENT 6.1 High Resolution Gas Chromatograph/Mass Spectrometer/Data System (HRGC/HRMS/DS) 6.1.1 The GC must be equipped for temperature programming, and all required accessories must be available, such as syringes, gases, and a capillary column. The GC injection port must be designed for capillary columns. The use of splitless injection techniques is recommended. When using this method, a 1-pL injection volume is used. The injection volumes for all extracts, blanks, calibration solutions, and the performance check sample must be consistent. 6.1.2 High Resolution Gas Chromatograph-Mass Spectrometer Interface The HRGC/MS interface is directly coupled to the mass spectrometer ion source. All components of the interface should be glass or glass-lined stainless steel. The interface components should be compatible with 300°C temperatures. The HRGC/MS interface must be appropriately designed so that the separation of the PCDDs and PCDFs which is achieved in the gas chromatographic column is not appreciably degraded. Cold spots and/or active surfaces (adsorption sites) in the HRGC/MS A-8 �interface can cause peak tailing and peak broadening. It is recommended that the HRGC column be fitted directly into the MS ion source. Graphite ferrules should be avoided in the HRGC injection port since they may absorb PCDDs or PCDFs. Vespel or equivalent ferrules are recommended. 6.1.3 Mass Spectrometer The mass spectrometer must be capable of maintaining a minimum resolution of 10,000 (10% valley) for high resolution confirmation analysis. The mass spectrometer must be operated in a selected ion monitoring (SIM) mode with total cycle time (including voltage reset time) of 1 s or less. 6.1.4 Data System A dedicated hardware or data system is required to control the rapid multiple ion monitoring process and to acquire the data. Quantification data (peak areas or peak heights) and SIM traces (displays of intensities of each m/z (characteristic ion) being monitored as a function of time) must be acquired during the analyses. Quantifications may be reported based upon computergenerated peak areas or upon measured peak heights. 6.2 HRGC Columns For isomer-specific determinations of 2,3,7,8-TCDD, the following fused silica capillary columns are recommended: a 50-tn CP-Sil 88 column and a 60-m SP-2330 (SP-2331) column. However, any capillary column which separates 2,3,7,8-TCDD from all other TCDDs may be used for such analyses, provided that the minimum acceptance criteria in Section 8 are met. 6.3 Miscellaneous Equipment 6.3.1 Nitrogen evaporation apparatus with variable flow rate. 6.3.2 Balance capable of accurately weighing to ± 0.01 g. 6.3.3 Balance capable of accurately weighint to ± 0.0001 g. 6.3.4 Water bath — equipped with concentric ring cover and capable of being temperature-controlled. 6.3.5 Stainless steel spatulas or spoons. 6.3.6 Magnetic stirrers and stir bars. 6.3.7 High speed tissue homogenizer -- Tekmar Tissuemizer® equipped with an EN-8 probe or equivalent. 6.3.8 Vacuum dessicator. A-9 �6.4 Glassware 6.4.1 Erlenmeyer flask — 500 mL. 6.4.2 Kuderna-Danish apparatus — 500-mL evaporating flask, 15-mL graduated concentrator tubes with ground-glass stoppers, and three-ball macro Snyder column (Kontes K-570001-0500, K-503000-0121, and K-569001-0219 or equivalent). 6.4.3 Minivials -- 1-mL borosilicate glass with conical-shaped reservoir and screw caps lined with Teflon®-faced silicone disks. 6.4.4 Powder funnels -- glass. 6.4.5 Chromatographic columns for the silica and alumina chromatography -- 1 cm ID x 10 cm long and 1 cm ID x 30 cm long with 250-mL reservoir and equipped with TFE stopcocks. 6.4.6 Chromatographic column for the Carbopak cleanup -disposable 5-mL graduated glass pipets, 6 to 7 mm ID. 6.4.7 Glass rods. 6.4.8 Carborundum boiling chips -- Extracted for 6 hr in a Soxhlet apparatus with benzene and air dried. 6.4.9 Glass wool, silanized (Supelco) -- Extract with methylene chloride and hexane and air dry before use. 6.4.10 Glassware cleaning procedure -- All glassware used for these analyses will be cleaned via the following procedure. Wash the glassware in soap and water, rinse with copious amounts of tap water, distilled water, and distilled-in-glass acetone, in that order. Immediately prior to use, the glassware should be rinsed with distilled-in-glass quality solvents: methylene chloride, toluene, and hexane. The glassware should be allowed to dry fully. As an added precuation, all glassware will be marked with a unique code that should be noted in the extraction and cleanup procedures for each sample. This glassware tracking will allow background results from specific glassware to be documented. After use, each piece of glassware should be rinsed with the last solvent used in it, followed by a rinse with toluene, then acetone, before transferring it to the glassware washing facility. A-10 �7. REAGENTS AND STANDARD SOLUTIONS 7.1 Column Chromatography Reagents 7.1.1 Alumina, acidic (Biorad, AG-4) — Extract the alumina in a Soxhlet apparatus with methylene chloride for 18 h (minimum of two cycles per hour). Air dry and activate it by heating in a foil-covered glass container for 24 h at 190°C. 7.1.2 Silica gel -- High purity grade, type 60, 70-230 mesh; extract the silica gel in a Soxhlet apparatus with methylene chloride for 10 h (minimum of 2 cycles per hour). Air dry and activate it by heating in a foilcovered glass container for 24 h at 130°C. 7.1.3 Silica gel impregnated with 40% (by weight) sulfuric acid -- Add two parts (by weight) concentrated sulfuric acid to three parts (by weight) silica gel (extracted and activated) (e.g., 40 g of H2S04 plus 60 g of silica gel) in a glass screw-cap bottle. Tumble for 5 to 6 h, shaking occasionally until free of lumps. 7.1.4 Sulfuric acid, concentrated — ACS grade, specific gravity 1.84. 7.1.5 Graphitized carbon black (Carbopack C, Supelco), surface of approximately 12 m2/g, 80/100 mesh -- Mix thoroughly 3.6 g of Carbopack C and 16.4 g of Celite 545® in a 40-mL vial. Activate at 130°C for 6 h. Store in a desiccator. 7.1.6 Celite 545® (Fischer Scientific), reagent grade, or equivalent. 7.2 Desiccating agents — Sodium sulfate; granular, anhydrous. Before use extract with methylene chloride for 16 h (minimum of two cycles per hour), air dry and then muffle for ^ 4 h in a shallow tray at 400°C. Let it cool in a desiccator and store in oven at 130°C. 7.3 Solvents -- High purity, distilled in glass: methylene chloride, toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent grade: tridecane. High purity solvents are dispensed from Teflon® squirt bottles. 7.4 Concentration Calibration Solutions (Table 2) Eight tridecane solutions containing native calibration standards, 13 C12-labeled internal quantisation standards, and two internal recovery standards are required. The complete compound list is A-11 �CO o Ln Ln CM l—I r H r H i —I T —( C \ J C \ J C N J C \ I C \ J C \ I C \ I C s J C \ J C \ J L T ) L n 1 CM CJ> Q . to o CM CM to o ooooiDLnLnmo to c o o to -p 3 c o O OO c. o rO s- u o rH rH rH rH CM O Ln o c o i—( T—1 rH rH CM to o rH rH rH rH CM CM O O O O O O O O O O O O O O O O O ooooLninLnLno 1—lrHrHrHrHCMCMCMCMCMCMCMCMCMCMLnLn rH i—I rH rH CM oo O) CD (J c o o o o C_J o ooooLnunLnLno LnmLnLncMCMCMCMLn T-HrHrHrHrHrHrHrHrHrHCMCM C_3 c. CM CM to o +J ro s- Ln CM i—Ir-lr-I T —I t —4 C \ J C S J C S I C X J C \ J C \ J C \ J C v J C \ I C \ I L n L n CO rO C_3 CM 1—I i—1 i— 1 i—I T —1 i— 1 r-1 i—l i —I r H C M C N J CM 00000000000000000 to ooooooooooooooooo O Ln ID CM CM CM CM C3 Ll_ (~*\ CD -Q rO a u_ u_ o oo QC3OLL.L1_L1_L1_C_3OO a a a o a o o. a. a. a. 3 O Q. r_3c_3c_5n:^x:3;3i^^cooocr> o u _ a > a ) 0 ) i i i i i i i > > > caoQ-Q-Q-oooocricoooo-ioor^f^-co . C-3 C_3 I I I o * . * . * * * . * . * ! * . * ! * litation 0 C3 O t_3 (— 3 in -a c c S*» C2 ' ' c CMCMrHi—ICMi—IrHrHrHi—IrHCMrHi—IrHOO 1—1 ^\ 0 C_3 0 Q o Q a. a. u_ 0 0 o X X 0 3C ^ CO ai o X 00 a i oo f~^ CT* •s r-~ in <J2 to ^> <d- *d- 00 CO r--. 0) r-. c S- v . * . ^ » . ^ . « . « ^ « t # . A ^ i ^ ^ u_ o 1— C_3 O) u_ 0 !. - r-~ CO CO CO ro CO CO 1— CO CO CM CM CM CM CM CM 1— CO 1 a> CO CO a i- o ro T3 CM CM i—t rH rH i-H rH rH 1 1 1 1 0J 0J 0 C1 CM CN IN <N <N c_> c_> CO O CO CO C_3 O CO CO c i. rH C ro to O 000 :0 "0 CO •o (—H *^ CM CM rH i—1 I 1 <N CM c_> o CO CO T-H TH �given in Table 2. The native 2,3,7,8-TCDD is supplied as a certified standard solution from the U.S. EPA QA Reference Materials Branch. All other native compounds were supplied in crystalline form by Cambridge Isotope Laboratories (Woburn, MA). 13C12~ Labeled internal quantitation standards were supplied in solution in n-nonane by Cambridge Isotope Laboratories. Portions of the native standards were accurately weighed to the nearest 0.001 mg with a Cahn 27 electrobalance and dissolved in toluene. 7.5 Column Performance Check Mixture The column performance check mixture consists of several TCDD isomers which will be used to document the separation of 2,3,7,8TCDD from all other isomers. This solution will contain TCDDs (A) eluting closely to 2,3,7,8-TCDD, and the first- (F) and lastel uting (L) TCDDs. Analyte Approximate amount per ampule 10 10 10 10 10 10 10 10 Unlabeled 2,3,7,8-TCDD 13 C12-2,3,7,8-TCDD 1,2,3,4-TCDD (A) 1,4,7,8-TCDD (A) 1,2,3,7-TCDD (A) 1,2,3,8-TCDD (A) 1,3,6,8-TCDD (F) 1,2,8,9-TCDD (L) 7.6 ng ng ng ng ng ng ng ng Spiking Solutions Three solutions are prepared using the same stock as in Section 7.4. A native standard solution and a 13C12 internal quantitation standard solution are prepared in isooctane (Tables 3 and 4). A recovery standard solution is prepared in tridecane (Table 4). Samples are spiked with 100 uL of internal quantitation standard solution and final sample extracts are spiked with 10 (jL of internal recovery standard solution. 8. HIGH RESOLUTION GAS CHROMATOGRAPHY/MASS SPECTROMETRY PERFORMANCE CRITERIA Samples and standards are analyzed by using a Carlo Erba MFC500 gas chromatography (GC) coupled to a Kratos MS50TC double-focusing mass spectrometer (MS) to be operated in the electron impact mode. The HRGC/MS interface is simply a direct connection of the fused silica HRGC column to the ion source of the MS via a heated interface oven. Data acquisition and processing are controlled by a Finnigan-MAT Incos 2300 data system. A-13 �Table 3. Native Spiking Solution3 Concentration (pg/jjL) Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 5 5 5 5 5 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 12.5 12.5 12.5 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF 12.5 12.5 12.5 12.5 12.5 12.5 12.5 25 25 a Prepared in isooctane. A-14 �Table 4. Internal Standard Spiking Solutions Concentration (pg/uL) Compound Internal Quantitation Standards3 13 C12-2,3,7,8-TCDD 5 13 5 13 5 13 5 C12-2,3,7,8-TCDF C12-l,2,3,7,8-PeCDD C12-l,2,3,7,8-PeCDF 13 12.5 13 12.5 13 12.5 13 12.5 C12-l,2,3,6,7,8-HxCDD C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,6,7,8-HpCDD C12-l,2,3,4,6,7,8-HpCDF 13 C12-OCDD 25 Internal Recovery Standard 13 C12-1,2,3,4-TCDD 13 C12-l,2,3,7,8,9-HxCDD ^Prepared in isooctane. Prepared in tridecane. A-15 50 125 �8.1 HRGC/MS Analysis of PCDD/PCDF Single run selected ion monitoring (SIM) analysis of the tetrachloro through octachloro-dioxins and furans is carried out with the instrumental conditions and parameters outlined in Table 5. For each HRGC/MS run, five distinct groups of ions, which correspond to each chlorine level, are sequentially monitored. These ion descriptors are shown in Table 6. The masses of the two most abundant ions in the molecular ion cluster of each dioxin and furan and isotopically labeled standard are monitored. In addition, the masses corresponding to the molecular ions of the hexachloro through decachlorodiphenyl ethers (PCDEs) are monitored to aid in the confirmation of positive furan results. Interference from the presence of PCDE is noted by coincident response to the characteristic ions for PCDFs. A lock mass, m/z 381 from PFK (perfluorokerosene), is used to observe and correct any magnet/instrument drift during the analysis. 8.1.1 Tuning and Mass Calibration The mass spectrometer is tuned on a daily basis to yield optimum sensitivity and peak shape using an ion peak (m/z 381) from PFK. The resolution is visually monitored and maintained at S 3,000 (10% valley definition) to provide adequate noise rejection while maintaining good ion transmission. Mass calibration of the mass spectrometer for the HRGC/MS analysis of PCDD/PCDF is carried out on a daily basis. The magnetic field is adjusted to pass m/z 300 at full accelerating voltage. PFK is admitted to the MS and an accelerating voltage scan from 8,000 to 4,000 V is acquired by the data system. This corresponds to an effective mass range of 301 to 593 amu. Upon completion of a successful calibration step, the five ion descriptors shown in Table 6 are updated to reflect the new mass calibration. 8.1.2 Ion Descriptor Switching The ion descriptors shown in Table 6 are sequentially monitored during a PCDD/PCDF analysis to cover the retention windows of each chlorination level. The retention windows and hence the descriptor switch points are determined initially and whenever a new HRGC column is installed by injection of a mixture of PCDD and PCDF congeners. Daily adjustment of the descriptor switch times are performed when careful monitoring of the standard retention times shows this to be necessary. The descriptors are designed to ensure acquisition of all isomers of each homolog. A-16 �Table 5. HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis Mass spectrometer 8,000 V 500 uA 70 eV -1,800 V Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: Overall SIM cycle time: 280°C ^ 3,000 (10% valley definition) 1s Gas chromatograph Column coating: Film thickness: Column dimensions: He linear velocity: He head pressure: DB-5 0.25 urn 60 m x 0.25 mm ID ~ 25 cm/sec 1.75 kg/cm2 (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 s 30 mL/min 6 mL/min 270°C 300°C 1-2 uL 200°C 2 min 200°C to 330°C at 5°C/min A-17 �Table 6. Ions Monitored for HRGC/MS of PCDD/PCDF Descriptor Al Mass ID 303.902 305.899 315.942 317.939 319.897 321.894 331.937 333.934 373.840 380.976 TCDF 13 C12-TCDF TCDD 13 C12-TCDD HxCDPE PFK (lock mass) A2 TCDF TCDD PeCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD PFK (lock mass) HpCDPE A3 Nominal dwell time (sec) HxCDF PFK (lock mass) 13 C12-HxCDF HxCDD 13 C12-HxCDD OCDPE A-18 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 303.902 305.899 319.897 321.894 337.863 339.860 349.903 351.900 353.858 355.855 365.898 367.895 380.976 407.801 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.035 0.035 373.821 375.818 380.976 385.861 387.858 389.816 391.813 401.856 403.853 443.759 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 �Table 6 (continued) Descriptor A4 Mass ID 380.976 389.816 391.813 407.782 409.779 419.822 421.819 423.777 425.774 435.817 437.814 429.768 431.765 477.720 HpCDF C12-HpCDF HpCDD 13 37 C12-HpCDD Cl4-HpCDD NCDPE A5 PFK (lock mass) OCDF 13 C12-OCDF OCDD 13 C12-OCDD DCDPE A-19 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 380.976 441.743 443.740 453.783 455.780 457.738 459.735 469.779 471.776 511.681 PFK (lock mass) HxCDD 13 Nominal dwell time (sec) 0.06 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.06 �8.1.3 HRGC Column Performance (60-m DB-5) The HRGC column performance must be demonstrated at the start of each 12-h analysis period. 8.1.3.1 Inject 1 uL of the column performance check solution (Section 7.5) and acquire selected ion monitoring (SIM) data for m/z 320, 322, 332, and 334. 8.1.3.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers should be resolved with a valley of 30-60%, where Valley % = ( / ) 1 0 xy(0) x = measured height of the valley between the chromatographic peak corresponding to 2,3,7,8-TCDD and the peak of the nearest TCDD isomer; and y = the peak height of 2,3,7,8-TCDD. Figure 2 is an example of the separation of a TCDD isomer mixture and the calculation of isomer resolution. It is the responsibility of the laboratory to verify the conditions suitable for the appropriate resolution of 2,3,7,8-TCDD from all other TCDD isomers. The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol, thus defining the retention time window for total TCDD determination. Any individual selected ion current profile or the reconstructed total ion current (m/z 320 + m/z 322) consititutes an acceptable form of data presentation. 8.1.4 Initial Calibration for PCDD/PCDF Analysis Initial calibration is required before any samples are analyzed for PCDD/PCDF. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.1.7. 8.1.4.1 Tune and calibrate the instrument with PFK as outlined in Section 8.1.1. A-20 �TCDD Isomer Mixture 78.7 2, 3, 7, 8.-TCDD 2359290. 1, 2, 3, 4-/1, 2, 3, 7-/L 2, 3, 8- TCDD 319.856 <*= 0.500 320 100.0 2998270. •*- x 100% = 29% Y 322 y\ A 321.855 ± 0.500 X = 10 mm 59.6 1785850. Q2-2, 3, 7, 8,-TCDD f\> 331.851 ± 0.500 332 75.3 2256890. 334 333.850 ± 0.500 28:00 28:24 Figure 2. 28:48 29:12 29:36 30:00 30:24 30:48 Example of the separation of 2,3,7,8-TCDD from other TCDD isomers on a 60 m DB-5 column. Time �8.1.4.2 8.1.4.3 Using the HRGC and MS conditions in Table 5 and the SIM monitoring descriptors in Table 6, analyze a 1-uL aliquot of each of the six concentration calibration solutions in triplicate. 8.1.4.4 Compute the relative response factors (RRFs) for each analyte in the concentration calibration solution using the criteria for positive identification of PCDD/PCDF's given in Section 14.1 and the computational methods in Section 14.2. 8.1.4.5 Compute the means and their respective relative standard deviations (% RSD) for the RRFs from each triplicate analysis for each analyte in the standard. 8.1.4.6 8.1.5 Six of the eight concentration calibration solutions listed in Table 2 will be analyzed for the initial calibration phase. These must include solutions CS4 through CSS (Table 2). The analyst may select any of the remaining solutions for demonstrating calibration at the upper concentration range. Calculate the grand means (RRF) and their respective RSDs using the six mean RRFs for each analyte. Criteria for Acceptable Initial Calibration 8.1.5.1 The % RSD for the response factors for each triplicate analysis of a single concentration calibration standard for each analyte must be less than ± 30% except for the TCDD and TCDF, which must be less than ± 20%. 8.1.5.2 The variation of the mean RRFs for the six concentration calibrated standards (Section 8.1.5.1) must be less than 30% except for the TCDD and TCDF which must be less than 20%. 8.1.5.3 The SIM traces for all ions used for quantitation must present a signal-to-noise (S/N) ratio of ^ 2.5. This includes analytes and isotopically labeled standards. A-22 �8.1.5.4 Isotopic ratios must be within ± 20% of the theoretical values (see Table 7). NOTE: If the criteria for acceptable calibration listed above have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The grand mean RRF from the initial calibration for unlabeled PCDD/ PCDFs and for the isotopically labeled standards will be used for all calculations until routine calibration criteria (Section 8.1.7) are no longer met. At such time, new mean RRFs will be calculated from a new set of six triplicate determinations. 8.1.6 Routine Calibrations Routine calibrations must be performed at the beginning of every day before actual sample analyses are performed and as the last injection of every day. 8.1.6.1 8.1.6.2 8.1.7 Inject 1 uL of the concentration calibration solution CS 7 (see Table 2) as the initial calibration check on each analysis day. It is recommended that the analyst select a concentration calibration solution that brackets the sample concentrations observed on a single analysis date as the last injection of each analysis date. Compute the RRFs for each analyte in the concentration calibration solution using the criteria for positive identification of PCDD/Fs given in Section 14.1 and the computational methods in Section 14.2. Criteria for Acceptable Routine Calibration 8.1.7.1 The measured RRF for all analytes must be within ± 30% of the grand mean values established by triplicate analysis of the calibration concentration solutions, except for TCDD and TCDF, which must be within ± 20% of the mean values established in the initial calibration step. 8.1.7.2 Isotopic ratios must be within ± 20% of the theoretical value for each analyte and isotopically labeled standard (see Table 7). A-23 �Table 7. Ion Ratios for HRGC/LRMS Analysis of PCDD/PCDF Compound TCDF 13 C12-TCDF TCDD 13 C12-TCDD PeCDF 13 Cl2-PeCDF PeCDD 13 Cl2-PeCDD HxCDF 13 Cl2-HxCDF HxCDD 13 Cl2-HxCDD HpCDF 13 Cl2-HpCDF HpCDD 13 Cl2-HpCDD OCDF 13 C12-OCDF OCDD 13 C12-OCDD Ions monitored Theoretical ratio 304/306 316/318 320/322 332/334 338/340 350/352 354/356 366/368 374/376 386/388 390/392 402/404 408/410 420/422 424/426 436/438 442/444 454/456 458/460 470/472 0.76 0.76 0.76 0.76 0.61 0.61 0.61 0.61 1.22 1.22 1.22 1. 22 1.02 1.02 1.02 1.02 0.87 0.87 0.87 0.87 A-24 Acceptabl e range 0.61 0.61 0.61 0.61 0.49 0.49 0.49 0.49 0.98 0.98 - 0.91 0.91 0.91 0.91 0.73 0.73 0.73 0.73 1.46 1.46 0.98 - 1.46 0.98 - 1.46 0.82 - 1.22 0.82 - 1.22 0.82 - 1.22 0.82 0.70 0.70 0.70 0.70 - 1.22 1.04 1.04 1.04 1.04 �8.1.7.3 8.2 If any of the above criteria is not met, a second attempt may be made before repeating the entire initialization process. HRGC/HRMS Analysis (Isomer Specific TCDD Analysis) Isomer specific analysis for 2,3,7,8-TCDD is carried out with the instrumental conditions and parameters shown in Table 8. In addition to monitoring the masses of the most abundant molecular ions of TCDD, an ion corresponding to the loss of COC1 from the molecular ion is monitored for verification purposes. Mass spectrometer resolution is maintained at or above 10,000 (10% valley definition) in order to increase the specificity of the analysis. 8.2.1 Tuning and Mass Calibration 8.2.1.1 The mass spectrometer must be operated in the electron (impact) ionization mode. Static resolving power of at least 10,000 (10% valley) must be demonstrated before any analysis of a set of samples is performed. Static resolution checks must be performed at the beginning and at the end of each 12-h period of operation. However, it is recommended that a visual check (i.e., not documented) of the static resolution be made before and after each analysis. 8.2.1.2 The MS shall be tuned daily using PFK to yield a resolution of at least 10,000 (10% valley) and optimal response at m/z 254.986. This step is followed by calibration of an accelerating voltage scan of PFK beginning at m/z 254 (typical calibration range is 255 to 493 amu). Other voltage scans from the same data file are used to establish and document both the resolution at m/z 316.983 and the mass measurement accuracy at m/z 330.979. 8.2.1.3 Following calibration, the SIM experiment descriptor is updated to reflect the new calibration. Six masses (see Table 8) are monitored by scanning ^ m/10,000 amu (atomic mass units) over each mass. The total cycle time is kept to 1 s. The m/z 280.983 ion from PFK is used as a lock mass because it is the most abundant PFK ion within the range of m/z 255 to 334 and therefore permits the use of low partial pressures of PFK, which minimizes PFK interferences at the analytical masses. A-25 �Table 8. HRGC/HRMS Operating Conditions Mass spectrometer 8,000 V 500 MA 70 eV 2,000 V 280° C 10,000 (10% valley definition) Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: SIM Parameters Identity TCDD-COC1 TCDD TCDD 13 Ci2-TCDD 13 C12-TCDD PFK (lock mass) Mass Nominal dwell times (s) 0.15 0.15 0.15 0.15 0.15 0.10 258.930 319.897 321.894 331.937 333.934 280.983 Overall SIM cycle time = 1 s Gas chromatograph Column coating: Film Thickness: Column dimensions: CP-Sil 88 0.2 Mm 50 m x 0.22 mm ID Helium linear velocity: Helium head pressure: ~ 25 cm/s 2 1.75 kg/cm (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 s 30 mL/min 6 mL/min 270°C 240°C 2 ML 200°C 1 min 200°C to 240°C at 4°C/min A-26 �8.2.2 Mass Measurement and Resolution Check Using a PFK molecular leak, tune the instrument to meet the minimum required resolving power of 10,000 (10% valley) at m/z 254.986 (or any other mass reasonably close to m/z 259). Calibrate the voltage sweep at least across the mass range m/z 259 to m/z 334 and verify that m/z 330.979 from PFK (or any other mass close to m/z 334) is measured within ± 5 ppm (i.e., 1.7 mmu, if m/z 331 is chosen) using m/z 254.986 as a reference. Documentation of the mass resolution must then be accomplished by recording the peak profile of the PFK reference peak m/z 318.979 (or any other reference peak at a mass close to m/z 320/322). The format of the peak profile representation must allow manual determination of the resolution; i.e., the horizontal axis must be a calibrated mass scale (amu or ppm per division). The results of the peak width measurement (performed at 5% of the maximum which corresponds to the 10% valley definition) must appear on the hard copy and cannot exceed 100 ppm (or 31.9 mmu if m/z 319 is the chosen reference ion). 8.2.3 HRGC Column Performance (50-m CP Sil 88/60-m SP-2330) Prior to any HRGC/HRMS analysis of calibration solutions or samples for 2,3,7,8-TCDD, the resolution of the HRGC columns must be documented to be within allowable limits in order to provide conditions adequate for unambiguous isomer-specific analysis of 2,3,7,8-TCDD. This column performance check must be demonstrated at the start of each 12-h analysis period. 8.2.3.1 Inject 2 uL of the column performance check solution and acquire selected ion monitoring (SIM) data for m/z 258.930, 319.897, 321.894, 331.937, and 333.934 within a total cycle time of ^ 1 s (Table 8). 8.2.3.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers must be resolved with a valley of ^ 25%, where Valley % = (x/y)(100) x = measured height of the valley between the chromatographic peak corresponding to 2,3,7,8-TCDD and the peak of the nearest TCDD isomer; and y = the peak height of 2,3,7,8-TCDD. A-27 �8.2.3.3 8.2.3.4 8.2.4 If the above resolution requirement is not met, corrective action must be taken and acceptable resolution documented prior to any further analyses. Corrective action may include removal of the first meter of the HRGC column, replacement or clearing of the injector port, or complete replacement of the GC column. The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol, thus defining the retention time window for total TCDD determination. The peaks representing 2,3,7,8-TCDD and the first and the last eluting TCDD isomer should be labeled and identified as such on the chromatograms (F and L, respectively). Any individual selected ion current profile or the reconstructed total ion current (m/z 259 + m/z 320 + m/z 322) constitutes an acceptable form of data presentation. Initial Calibration for HRGC/HRMS 2,3,7,8-TCDD Analysis Initial calibration is required before any samples are analyzed for 2,3,7,8-TCDD. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.2.6. 8.2.4.1 At least six of the concentration calibration solutions listed in Table 2 must be utilized for the initial calibration. These must include solutions CS4 through CSS. The analyst may select any of the remaining solutions for demonstrating calibration at the upper concentration range. 8.2.4.2 Tune and calibrate the instrument with PFK as described in Section 8.2.1. 8.2.4.3 Inject 1 uL of the column performance check solution (Section 8.2.3) and acquire SIM mass spectra data for m/z 258.930, 319.897, 321.894, 331.937, and 333.934 using a total cycle time of ^ 1 s (see Table 8). The laboratory must not perform any further analysis until it has been demonstrated and documented that the criterion listed in Section 8.2.3.2 has been met. A-28 �8.2.4.4 8.2.4.5 Calculate the RRFs for unlabeled 2,3,7,8TCDD relative to 13C12-2,3,7,8-TCDD and the RRF for 13C12-2,3,7,8-TCDD relative to 13 C12-1,2,3,4-TCDD using the criteria for positive identification of TCDD by HRGC/ HRMS given in Section 14.1 and the computational methods in Section 14.2. 8.2.4.6 Calculate the six means (RRFs) and their respective relative standard deviations (% RSD) for the response factors from each of the triplicate analyses for both unlabeled and 13C12-2,3,7,8-TCDD. 8.2.4.7 8.2.5 Using the same GC and MS conditions (Table 8) that produced acceptable results with the column performance check solution, analyze a 1-uL aliquot of each of the six concentration calibration solutions in triplicate. Calculate the grand mean RRFs and their respective relative standard deviations ( RSD) using the six mean RRFs. % Criteria for Acceptable Initial Calibration The criteria listed below for acceptable calibration must be met before analysis of any sample is performed. 8.2.5.1 The percent relative standard deviation (RSD) for the response factors from each of the triplicate analyses of a single concentration calibration standard for both unlabeled and 13C12-2,3,7,8-TCDD must be less than 20%. 8.2.5.2 The variation of the mean RRFs from the six concentration calibration standards unlabeled and 13C12-2,3,7,8-TCDD must be less than 20% RSD. 8.2.5.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise ratio of s 2.5 for m/z 258.930, m/z 319.897, and m/z 321.894. 8.2.5.4 SIM traces for 13C12-2,3,7,8-TCDD must present a signal-to-noise ratio ^2.5 for m/z 331.937 and m/z 333.934. 8.2.5.5 Isotopic ratios for 320/322 and 332/334 must be within the allowed range (0.61 to 0.91). A-29 �NOTE: If the criteria for acceptable calibration listed above have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The grand mean RRF from the initial calibration for unlabeled 2,3,7,8-TCDD and for 13C12-2,3,7,8-TCDD will be used for all calculations until routine calibration criteria (Section 8.2.6) are no longer met. At such time, new mean RRFs will be calculated from a new set of six triplicate determinations. 8.2.6 Routine Calibrations Routine calibrations must be performed at the beginning of a 12-h period after successful mass resolution and HRGC column performance check runs and before analysis of actual samples. The response factor calibration must also be verified at the end of each analysis date. 8.2.6.1 8.2.7 Inject 1 uL of the concentration calibration solution (CS7, Table 2) which contains 2.5 pg/uL of unlabeled 2,3,7,8-TCDD, 50.0 pg/uL of 13C12-2,3,7,8-TCDD, and 50 pg/uL of 13C12-1,2,3,4-TCDD. Using the same HRGC/ MS/DS conditions as used in Table 8, determine and document acceptable calibration as provided below. Criteria for Acceptable Routine Calibration The following criteria must be met before further analysis is performed. If these criteria are not met, corrective action must be taken and the instrument must be recalibrated. 8.2.7.1 The measured RRF for unlabeled 2,3,7,8-TCDD must be within 20% of the mean values established in the initial calibration by triplicate analyses of concentration calibration solutions. 8.2.7.2 The measured RRF for 13C12-2,3,7,8-TCDD must be within 20% of the mean value established by triplicate analysis of the concentration calibration solutions during the initial calibration. A-30 �8.2.7.3 Isotopic ratios must be within the allowed range (0.61 to 0.90). 8.2.7.4 If one of the above criteria is not satisfied, a second attempt can be made before repeating the entire initialization process. NOTE: An initial calibration must be carried out whenever the routine calibration solution is replaced by a new one from a different lot. 9. QUALITY CONTROL PROCEDURES 9.1 Summary of QC Analyses 9.1.1 Initial and routine calibration and instrument performance checks. 9.1.2 Analysis of a batch of samples with accompanying QC analyses: Sample batch -- 10 NHATS adipose tissue samples plus additional QC analyses including 1 method blank, a control tissue and a spiked tissue sample. "Blind" QC (external QC) samples may be submitted by an external source (quality assurance group or independent laboratory) and included among the batch of samples. Blind samples include spiked samples, unidentified duplicates, and performance evaluation samples. 9.2 Performance Evaluation Solutions -- Included among the samples in every third batch will be a solution provided by the quality control coordinator containing known amounts of unlabeled 2,3,7,8TCDD and/or other PCDD/PCDF isomers. The accuracy of measurements for performance evaluation samples should be in the range of 70-130%. 9.3 Column Performance Check Solutions 9.3.1 At the beginning of each 12-h period during which samples are to be analyzed, an aliquot of the HRGC column performance check solution shall be analyzed to demonstrate adequate HRGC resolution for selected TCDD isomers. A-31 �9.4 Method Blanks 9.4.1 A minimum of one method blank is generated with each batch of samples. A method blank is generated by performing all steps detailed in the analytical procedure using all reagents, standards, equipment, apparatus, glassware, and solvents that would be used for a sample analysis, but omit addition of the adipose tissue. 9.4.1.1 The method blank must contain the same amounts of Carbon-13 labeled internal quantitation standards that are added to samples before bulk lipid cleanup. 9.4.1.2 An acceptable method blank exhibits no positive response for any of the characteristic ions monitored. 9.4.1.2.1 If the above criterion is not met, solvents, reagents, spiking solutions, apparatus, and glassware are checked to locate and eliminate the source of contamination before any samples are extracted and analyzed. 9.4.1.2.2 If new batches of reagents or solvents contain interfering contaminants, purify or discard them. 9.5 Control Samples -- Control samples are prepared from a bulk sample(s) of human adipose tissue or similar matrix (e.g., porcine fat). This material is prepared by blending the tissue with methylene chloride, drying the extract by eluting through anhydrous sodium sulfate, and removing the methylene chloride using rotoevaporation at elevated temperatures (80°C). The evaporation process should be extended to ensure all traces of the extraction solvent have been removed. The resulting oily matrix (lipid) is subdivided into 10-g aliquots which are analyzed with each sample batch. The results of the individual analysis will be used to give a measure of precision from batch to batch over an entire program. Sufficient tissue should be extracted to provide a homogeneous lipid matrix that can be used over the total analysis program. Enough lipid matrix is necessary to prepare the spiked samples describe in Section 9.6. 9.6 Spiked Samples — Spiked lipid samples are prepared using a portion of the homogenized lipid described in Section 9.5. Sufficient spiked lipid matrix is prepared to provide a minimum of one spiked sample per sample batch. It is recommended that a minimum A-32 �of three spiked levels of the matrix are prepared ranging from 10 to 50 times the estimated limit of detection for each compound. Each analysis of spiked sample must be accompanied by analysis of a control sample in order to make the necessary corrections for background contribution before determining the accuracy of the method (Equation 9-1). A fw\ mn<v Cone, spiked sample-cone, control sample Eq 9 n1, ,. K Accuracy ( ) = 100% x % Spike level ' " 9.7 9.8 10. Duplicate Sample Analysis -- When possible a duplicate analysis of specific samples is included in the sample batch as an additional measure of method precision. It is suggested that the total tissue sample is extracted to isolate lipids material and then subdivided for duplicate analysis. Precision is calculated as relative percent difference (RPD) where the differences in the duplicate measurements (for each analyte) is divided by the average of the two measurements and multiplied by 100%. External Samples — Samples submitted as blinds to the analyst may consist of either performance solutions of PCDD and PCDF congeners or spiked sample matrices. These performance solutions or samples should be submitted by a source external to the analytical program (QA unit of analysis laboratory or independent laboratory). Performance audit solutions are intended to evaluate instrument calibration and quantisation procedures. Spiked blind samples must be accompanied by the corresponding unspiked samples to correct concentrations for background concentration. The blind spiked samples are intended to evaluate the total analytical procedure. The analyst must keep in mind that it is necessary to compare differences in standard sources for each type of external sample. SAMPLE PRESERVATION AND HANDLING All adipose tissue samples must be maintained at less than -20°C from time of collection. The analyst should instruct the collaborator collecting the sample(s) to avoid the use of chlorinated materials. Samples are handled using stainless steel forceps, spatulas, or scissors. Aliquots of samples removed from sample bottles not used for analysis are disposed rather than returned to the sample vial. All sample bottles (glass) are cleaned as specified in Section 6.4.10. Teflon®-lined caps should be used. As with any biological sample, the analyst should avoid any undue exposure. 11. SAMPLE EXTRACTION 11.1 Extraction of Adipose Tissue 11.1.1 Accurately weigh to the nearest 0.01 g a 10-g portion of a frozen adipose tissue sample into a culture tube (2.2 x 15 cm). Note: Sample size may be smaller, depending on availability. A-33 �11.1.2 Addition of internal quantitation standards Allow the adipose tissue specimen to reach room temperature and then add the carbon-13 internal quantitation spiking solution (Section 7.6) such that it delivers 500 to 2,500 pg of each of the surrogates specified in Table 4 in a 100-uL volume. 11.1.3 11.1.4 Allow the mixture to separate and decant the methylene chloride extract from the residual solid material using a disposable pipette. The methylene chloride is eluted through a filter funnel containing a plug of clean glass wool and 5 to 10 g of anhydrous sodium sulfate. The dried extract is collected in a 100-mL volumetric flask. 11.1.5 A second 10-mL aliquot of methylene chloride is added to the sample and homogenized for 1 min. The methylene chloride is decanted, dried, and transferred to the 100-mL volumetric flask as specified in Section 11.1.3 11.1.6 The culture tube is rinsed with at least two additional aliquots (10 mL each) of methylene chloride, and the entire contents are transferred to the filter funnel containing the anhydrous sodium sulfate. The filter funnel and contents are rinsed with additional methylene chloride (20 to 40 mL). The total eluent from the filter funnel is collected in the 100-mL volumetric flask. Discard the sodium sulfate. 11.1.7 11.2 Add 10 mL of methylene chloride and homogenize the mixture for approximately 1 min with a Tekmar Tissuemizer®. The final volume of the extract for each sample is adjusted to 100 mL in the volumetric flask using methylene chloride. Lipid Determination 11.2.1 Preweigh a clean 1-dram glass vial to the nearest 0.0001 g using an analytical balance tared to zero. 11.2.2 Accurately transfer 1.0 mL of the final extract (100 mL) from Section 11.1.7 to the 1-dram vial. Reduce the volume of methylene chloride from the extract using a water bath (50-60°C) gentle stream of purified nitrogen until an oil residue remains. A-34 �11.2.3 Accurately weigh the 1-dram vial and residue to the nearest 0.0001 g and calculate the weight of lipid present in the vial based on difference. Nitrogen blow-down is continued until a constant weight is achieved. 11.2.4 Calculate the percent lipid content of the original sample to the nearest 0.1% as shown in Equation 11-1. " I D * FYT Lipid content, LC ( ) = w % x 100% v AT AL Eq. 11-1 where: W.R = weight of the lipid residue to the nearest 0.0001 g calculated from Section 11.2.3; = total volume of the extract in mL from Section 11.1.6 (100.0 ml); WAT = weight of the original adipose tissue samples to the nearest 0.01 g from Section 11.1.1; and V.. 11.2.5 11.3 = volume of the aliquot of the final extract in ml used for the quantitative measure of the lipid residue (1.0 mL). Record the lipid residue measured in Section 11.2.3 and the percent lipid content calculated from Section 11.2.4. Extract Concentration 11.3.1 Quantitatively transfer the remaining extract volume (99.0 mL) to a 500-mL Erlenmeyer flask. Rinse the volumetric flask with 20 to 30 mL of additional methylene chloride to ensure quantitative transfer. 11.3.2 Place the Erlenmeyer flask on a hot plate at 40°C to remove solvent until an oily residue remains. 12. CLEANUP PROCEDURES 12.1 Bulk Lipid Removal 12.1.1 Add a total of 200 mL of rrhexane to the spiked lipid residue in the 500-mL Erlenmeyer flask. A-35 �12.1.2 Slowly add, with stirring, 100 g of the 40% w/w sulfuric acid impregnated silica gel (Section 7.1.3). Stir with a magnetic stir-plate for 2 h. 12.1.3 Allow solids to settle and decant liquid through a powder funnel containing 20 g of anhydrous sodium sulfate and collect in a 500-mL sample bottle. 12.1.4 Rinse solids with two 50-mL portions of hexane. Stir each rinse for 15 min, decant, and dry by elution through sodium sulfate combining the hexane extracts from Section 12.1.3. 12.1.5 After the rinses have gone through the sodium sulfate, rinse the sodium sulfate with an additional 25 ml of hexane and combine with the hexane extracts from Section 12.1.4. 12.1.6 Prepare an acidic silica column as follows: Pack a 1 cm x 10 cm chromatographic column with a glass wool plug, add approximately 25 ml of hexane, add 1.0 g of silica gel (Section 7.1.2) and allow to settle, then add 4.0 g of 40% w/w sulfuric acid impregnated silica gel (Section 7.1.3) and allow to settle. Pack a second chromatographic column (1 cm x 30 cm) with a glass wool plug, add approximately 25 mL of hexane, add 6.0 g of acidic alumina (Section 7.1.1), and allow to settle and then top with a 1-cm layer of sodium sulfate (Section 7.2). Elute the excess hexane solvent through the columns until the solvent level reaches the top of the chromatographic packing. Inspect columns to ensure they are free of channels and air bubbles. Wash the alumina column with 40 ml of 50% v/v methylene chloride/hexane. Remove the methylene chloride from the adsorbent by eluting the column with an additional 100 mL of hexane. Elute the excess solvent from the column until the solvent level reaches the top of the sodium sulfate layer. 12.1.7 Quantitatively transfer the hexane extract from the Erlenmeyer flask (Sections 12.1.3 through 12.1.5) to the silica gel column reservoir. Allow the hexane extract to percolate through the column and collect in a KD concentrator. 12.1.8 Complete the elution of the extract from the silica gel column with 50 ml of hexane in the KD concentrator. Concentrate the eluate to approximately 1.0 ml, using nitrogen blow-down as necessary. A-36 �Note: If the 40% sulfuric acid/silica gel is noted to be highly discolored throughout the length of the adsorbent bed it is necessary to repeat the cleaning procedure beginning with Section 12.1.1. 12.2 Separation of Chemical Interferences 12.2.1 Transfer the concentrate (1.0 mL) to the top of the alumina column. Rinse the K-D assembly with two 1.0-mL portions of hexane and transfer the rinses to the top of the alumina column. Elute the alumina column with 18 mL of hexane until the hexane level is just below the top of the sodium sulfate. Discard the eluate. Columns must not be allowed to reach dryness (i.e., a solvent "head" must be maintained). 12.2.2 Place 30 ml of 20% (v/v) methylene chloride in hexane on top of the alumina and elute the TCDDs from the column. Collect this fraction in a 50-mL culture tube. 12.2.3 Prepare an 18% Carbopak C/Celite 545® mixture by thoroughly mixing 3.6 g of Carbopak C (80/100 mesh) and 16.4 g of Celite 545® in a 40-mL vial. Activate at 130°C for 6 h. Store in a desiccator. Cut off a clean 5-mL disposable glass pipet (6 to 7 mm ID) at the 4-mL mark. Insert a plug of glass wool and push to the 2-mL mark. Add 500 mg of the activated Carbopak/Celite mixture followed by another glass wool plug. Using two glass rods, push both glass wool plugs simultaneously towards the Carbopak/Celite mixture and gently compress the Carbopak/Celite plug to a length of 3 to 3.5 cm. Pre-elute the column with 2 ml of toluene followed by 1 ml of 75:20:5 methylene chloride/methanol/ benzene, 1 ml of 1:1 cyclohexane in methylene chloride, and 2 mL of hexane. The flow rate should be less than 0.5 mL/min. While the column is still wet with hexane, add the entire eluate (30 ml) from the alumina column (Section 12.2.2) to the top of the column. Rinse the culture tube which contained the extract twice with 1 mL of hexane and add the rinsates to the top of the column. Elute the column sequentially with two 1-mL aliquots of hexane, 1 mL of 1:1 cyclohexane in methylene chloride, and I mL of 75:20:5 methylene chloride/methanol/benzene. Turn the column upside down and elute the PCDD/PCDF fraction with 20 mL of toluene into 6-dram vial. 12.2.4 Using a stream of nitrogen, reduce the toluene volume to approximately 1 mL. Carefully transfer the concentrate into a 1-mL mini vial and reduce the volume to about 200 (jL using a stream of nitrogen. A-37 �12.2.5 Rinse the concentrator tube with three washings using 500 uL of 1% toluene in methylene chloride. Concentrate to 200-500 |jL and add 10 |jL of the tridecane solution containing the internal recovery standard and store the sample in a refrigerator until HRGC/MS analysis. 12.2.6 Immediately prior to analysis, using a gentle stream of nitrogen at room temperature, remove toluene and methylene chloride. Submit sample to HRGC/MS once a stable 10 uL volume of tridecane is attained. 13. ANALYTICAL PROCEDURES 13.1 HRGC/MS Analysis for PCDD/PCDF 13.1.1 Once routine calibration criteria are met, the instrument is ready for sample analysis. Prior to the first sample, a blank injection of tridecane should be analyzed to document system cleanliness. If any evidence of system contamination is found, corrective action must be taken and another tridecane blank analyzed. The typical daily sequence of injections is shown in Table 9 and Figure 3. Note: Syringe Technique -- Congeners of PCDD/PCDF in the syringes used for HRGC/MS analysis can be problematic unless the syringes are properly handled between samples. The following procedure has been found to be very effective for PCDD/PCDF removal from contaminated syringes and will be used throughout these analyses. • Rinse the syringe 10 times with isooctane. Fill the syringe with toluene and sonicate syringe and plunger in toluene for 5 min and repeat at least twice. • Rinse the syringe 10 times with tridecane and pull up 1 pL of clean tridecane. • Syringe is ready for use. At no time should air be introduced into the HRGC column by using an air plug in the syringe. The oxygen present in the air plug will quickly degrade a nonbonded GC phase. 13.1.2 Inject a 1-uL aliquot of the extract into the GC, operated under the conditions previously used (Section 8.1) to produce acceptable results with the performance check solution. A-38 �Table 9. Typical Daily Sequence for PCDD/PCDF Analysis 1. Tune and calibrate mass scale versus perfluorokerosene (PFK). 2. Inject column performance mixture. 3. Inject concentration calibration solution 2.5 to 12.5 pg/nL (CS-7) solution. 4. Inject blank (tridecane). 5. Inject samples 1 through "N". 6 Inject concentration calibration solution 2.5 to 12.5 pg/pL (CS-7) solution or other concentration calibration solutions CS1 to CSS to bracket observed sample concentration. A-39 �INSTRUMENTAL ANALYSIS Instrument Mass Calibration vs PFK Mass Resolution Check 1 Column Performance Evaluation Does Column Performance Meet Minimum Resolution Requirements? No Adjust Column Length or Install New Column Calibration Standard Analysis Do Relative Response Factors Meet Criteria Based on Initial Calibrations ? No Reanalyze or Prepare Fresh Calibration Standards and Calibration Curve Yes Proceed with Sample Analysis Figure 3. Daily QA procedures for proceeding with sample analysis. A-40 �13.1.3 Acquire SIM data according to the same acquisition and MS operating conditions previously used (Section 8.1) to determine the relative response factors. 13.1.3.1 13.1.3.2 13.2 Acquire SIM data for the characteristic ions designated in Table 6. Instrument performance shall be monitored by examining and recording the peak areas for the recovery standard, 13C12-1,2,3,4-TCDD. If this area should decrease to less than 50% of the calibration standard, sample analyses shall be stopped until the problem is found and corrected. HRGC/HRMS Confirmation of 2,3,7,8-TCDD The presence of 2,3,7,8-TCDD observed through the general PCDD and PCDF procedure should be confirmed using HRGC/HRMS (resolution 10,000). 13.2.1 Once the daily criteria of mass calibration, mass resolution, HRGC performance, and routine calibration are met and documented, the instrument is ready for sample analysis. Prior to the first sample, a blank injection of tridecane will be made to document system cleanliness. The typical daily schedule for HRGC/HRMS analysis of TCDD is shown in Table 10 and Figure 3. 13.2.2 Inject a 1-uL aliquot of the extract into the GC, operated under the conditions previously used (Section 8.2) to produce acceptable results with the column performance check solution. 13.2.3 Acquire SIM data according to Section 8.2.4.3. Use the same acquisition and MS operating conditions previously used to determine the relative response factors. 13.2.3.1 Acquire SIM data for the following selected characteristic ions: m/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD 321.894 Unlabeled TCDD 331.937 13 C12-2,3,7,8-TCDD, 13 C12-1,2,3,4-TCDD 333.934 13 C12-2,3,7,8-TCDD, 13 C12-1,2,3,4-TCDD A-41 �Table 10. Typical Daily Schedule for HRGC/HRMS Analysis of TCDD 1. Tune and calibrate mass scale. 2. Perform mass measurement check and mass resolution check. 3. Inject column performance check solution. 4. Inject the routine concentration calibration solution (CS7) and confirm response factor consistency. 5. Inject tridecane blank. 6. Inject samples 1 through "N". 7. Inject concentration calibration solution and confirm response factor consistency. 8. Mass resolution check. A-42 �14. DATA REDUCTION In this section, the the analysis of data HRGC/HRMS method for qualitative criteria 14.1 procedures for the data reduction are outlined for from both the HRGC/MS method for PCDD/PCDF and the 2,3,7,8-TCDD. Figure 4 presents a schematic of the for identifying PCDDs and PCDFs. Qualitative Identification 14.1.1 14.1.2 The ion current intensities for a particular PCDD/PCDF must be ^ 2.5 times the noise level (S/N ^ 2.5) for positive identification of that isomer. 14.1.3 The integrated ion current ratios of the analytical masses for a particular PCDD/PCDF must fall within the ranges shown in Table 7. 14.1.4 14.2 The ion current responses for each mass for a particular PCDD/PCDF analyte must be within ± 1 s to attain positive identification of that analyte. For example, m/z 338 and m/z 340 must have maximum peak responses that are within ± 1 s to be positively identified as a pentachlorodibenzofuran. The recovery of the internal quantisation standards should be between 50 and 115%. Quantitative Calculations 14.2.1 Relative response factors for native PCDD and PCDF analytes (RRF). RRFs are calculated from the data obtained during the analysis of concentration calibration solutions using the following formula: A •r STD IS RRF = .b'U r i;> tt U IS ' STD where A... = the ,,, ion For sum m/z Eq. 14-1 sum of the areas of the integrated abundances for the analyte in question. example, for TCDD, ASTD would be the of the integrated ion abundances for 320 and 322; A,,, = the sum of the areas of the integrated ion abundances for the labeled PCDD/F used as the internal quantisation standard for the above analyte. For example, for 13C12~ 2,3,7,8-TCDD, A JS would be the sum of the integrated ion abundance for m/z 332 and 334. Cc-rn = concentration of the analyte in pg/(jL; A-43 �HRGC/MS-SIMData Response to Characteristic Molecular Ions within the Appropriate Homolog Retention Window? Report Compounds as Not Detected (ND) Calculate Sample LOP Characteristic Ion Ratios within ±20% Theoretical? Response Due to Coextracted Interference Response Corresponds to Specific Isomer Retention Time? Quantitate Compound as Per Protocol Report as Isomer Unknown Quantitate Specific Isomer as per Protocol Figure 4. Qualitative criteria for identifying PCDDs and PCDFs. A-44 �GIS = concentration of the internal quantisation standard in pg/uL; and Table 11 provides the pairing of target analytes to internal quantitation standards for determining RRF values for PCDD and PCDF compounds. 14.2.2 Relative response factors for the internal quantitation standards (RRF,<.)- The RRF TS values are calculated from data obtained auring the analysis of concentration calibration solutions using the following formula. C Eq. 14-2 ~ RS A XL x M RS IS where ATC. and CT<; are defined as given in Section 14.2.1 1:> i:> and CRS = concentrations of the internal recovery standard in pg/uL; and Apc; = the sum of the areas of the integrated ion 5 abundances for the labeled PCDD (13C121,2,3,4-TCDD or 13C12-l,2,3,7,8,9-HxCDD). For example, for 13C12-1,2,3,4-TCDD, A R$ would be the sum of the integrated ion abundance for m/z 332 and 334. Refer to Table 11 for pairing of the internal quantitation standards with the appropriate internal recovery standard. 14.2.3 Concentrations of sample components. Figure 5 presents a schematic for quantitation of PCDDs and PCDFs which meet the criteria specified in Section 14.1. Calculate the concentration of PCDD/Fs in sample extracts using the formula: A -.«nn~ " QT C . i r\n X Csampl e A sampl e » IS 100 Eq. 14-3 IS RRF wAT • LC • where C , = the lipid adjusted concentration of PCDD or sample congener 1n pg/g. A , = sum of the integrated ion abundances determined for the PCDD/PCDF in question; sample AIS = sum of the integrated ion abundances determined for the labeled PCDD/F used as the internal quantitation standard for the above analyte; A-45 �Table 11. Target Analyte/Internal Quantitation Standard and Internal Quantisation Standard/Internal Recovery Standard Pairs Internal standards Quantitation Target analyte 2,3,7,8-TCDD 13 13 2,3,7,8-TCDF 13 13 1,2,3,7,8-PeCDF 13 13 2,3,4,7,8-PeCDF 13 13 1,2,3,7,8-PeCDD 13 13 1,2,3,4,7,8-HxCDF 13 13 1,2,3,6,7,8-HxCDF 13 13 2,3,4,6,7,8-HxCDF 13 13 1,2,3,7,8,9-HxCDF 13 13 1,2,3,4,7,8-HxCDD 13 13 1,2,3,6,7,8-HxCDD 13 13 1,2,3,7,8,9-HxCDD 13 13 1,2,3,4,6,7,8-HpCDF 13 13 1,2,3,4,7,8,9-HpCDF 13 13 1,2,3,4,6,7,8-HpCDD 13 13 OCDF 13 13 OCDD 13 13 Recovery C12-2,3,7,8-TCDD C12-2,3,7,8-TCDF C12-l,2,3,7,8-PeCDF C12-l,2,3,7,8-PeCDF C12-l,2,3,7,8-PeCDD C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,7,8-HxCDF C12-l,2,3,6,7,8-HxCDD C12-l,2,3,6,7,8-HxCDD C12-l,2,3,6,7,8-HxCDD C12-l,2,3,4,6,7,8-HpCDF C12-l,2,3,4,6,7,8-HpCDF C12-l,2,3,4,6,7,8-HpCDD C12-OCDD C12-1,2,3,4-TCDD C12-1,2,3,4-TCDD C12-1,2,3,4-TCDD C12-1,2,3,4-TCDD C12-1,2,3,4-TCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-l,2,3,7,8,9-HxCDD C12-OCDD C12-l,2,3,7,8,9-HxCDD A-46 �QUANTITATION HRGC/MS-SIMData Response Meets Al I Qualitative Criteria ? Report as Not Detected Calculate Sample LOD Response >2.5times S/N? Response >10 times S/N? Calculate as per Protocol Report as Trace (tr) Value Quantitate as per Protocol Report as Positive Quantifiable Value Figure 5. Procedure for quantitation of PCDDs and PCDFs in human adipose tissue. A-47 �QIS = the amount (total pg) of the labeled internal quantisation standard added to the sample prior to extraction; RRF = relative response factor of the above analyte relative to its labeled internal quantitation standard determined from the initial triplicate calibration; WAT = weight (g) of original adipose tissue sample; and LC = percent extractable lipid determined from Eq. 11-1. Refer to Table 11 for pairing of target analytes with the appropraite internal quantitation standard. Quantitative data should be classified to indicate the intensity of the signal response. Suggested qualifiers include: not detected, ND (signal-to-noise ratio is less than 2.5); trace, TR (signal-to-noise ratio is greater than or equal to 2.5 but less than 10); and positive quantifiable, PQ (signal-to-noise ratio is greater than or equal to 10). 14.2.4 Recovery of internal quantitation standards. Calculate the recovery of the labeled internal quantitation standards measured in the final extract using the formula: A IS • X0 Internal Quant. Std. _ RS . ,. . fl0 Percent Recovery AR$ - Q J$ • RRF 1UU f tq ' 1^4 where AT{. = sum of the integrated ion abundances determined for the labeled PCDD/PCDF internal quantitation standard in question; A0<- = sum of the integrated ion abundances deterKb mined for m/z 332 and m/z 334 of 13C121,2,3,4-TCDD or m/z 390 and m/z 392 of 13 C13-l,2,3,7,8,9-HxCDD (recovery standards) QR<- = amount (pg) of the respective recovery standard, added to the final extract; QTC. = amount (pg) the labeled internal quantitation standard added to the sample prior to extraction; and A-48 �= relative response factor for the labeled internal quantitation standard in question relative to the internal recovery standard. This value shall be the RRF determined from the initial calibration. Refer to Table 11 for pairing of the internal quantitation standards with the appropriate target analytes. Note: The result of calculations as presented in Section 14.2 may be off by as much as 1% due to the fact that 1 ml of the final 100 ml volume from the extraction was used for lipid determination. 14.3 Estimated Method Detection Limit Estimated where (1) sponse is sponse is method detection limits must be calculated in situations no response is noted for a specific congener; (2) a renoted but ion ratios are incorrect; and (3) where a requantitated as a trace value. 14.3.1 For samples in which no unlabeled PCDD or PCDF is detected, calculate the estimated minimum detectable concentration. The background area is determined by integrating the ion abundances for the characteristic ions in the appropriate region and relating the product area to an estimated concentration that would produce that product area. Use the formula: 2.5 sample RRF "IS Eq. 14-5 where C p = estimated concentration of unlabeled PCDD or PCDF required to produce A ,; sample 'IS sum of integrated ion abundances or peak heights for the characteristic ions of the unlabeled PCDD or PCDF isomer in the same group of ^ 5 scans used to measure A JS ; and sum of integrated ion abundances for the appropriate ions characteristic of the respective internal quantitation standard. Qyc> RRF, and W.T retain the definitions previously stated in Section 14.2. Alternatively, if peak height measurements are used for quantification, measure the estimated detection limit by the peak height of the noise in the 2,3,7,8-TCDD RT window. A-49 �14.3.2 14.3.3 15. For samples for which a response at the retention time of a specific PCDD or PCDF congener is noted, but the qualitative criteria for ion ratios are outside the acceptable range (Table 7), the estimated detection level is calculated as given in Eq. 14.3 except the values are qualified as not detected, ND, and the concentration is reported in parenthesis. If a response for a specific PCDD or PCDF congener is qualified as a trace, TR, value (signal to noise is greater than or equal to 2.5 but less than 10) the analyst must also provide an estimated method detection limit. This is accomplished by using the observed signal to noise on either side of the response and calculating as given in Eq. 14-5. REPORTING AND DOCUMENTATION All data should be reported on an individual sample basis using the data report format shown in Figure 6. The analyst is required to maintain all raw data, calculations, and control charts in a format as to allow a complete external data review. Suggested data formats for tracing calculations are provided in Figure 7. A-50 �P>g. 1 ot 1 U.S. ENVIRONMENTAL PROTECTION AGENCY OFFICE OF TOXIC SUBSTANCES EXPOSURE EVALUATION DIVISION (TS-798) WASHINGTON, DC 20460 NATIONAL HUMAN ADIPOSE TISSUE SURVEY ANALYSIS REPORT FORM EPA SAMPLE NUMBER. ANALYSIS DATE LAB NUMBER MS ANALYST BATCH NUMBER REPORT DATE _ REPORTED BY _ NATIVE COMPOUNDS CONCENTRATION DATA (pg/g)l/ QUALIFIER!/ INTERNAL OUANTITATION STANDARD 2.3.7.8-TCDD 1 . . . ! • ( , 1 1 2.3,7,8-TCOF 1 . . . 1*1 1 1 1: 1 ,2,3.7.8-PeCDD 1 ...!•!, 1 13 1 ,,.!•!, 1 13 2.3.4,7,8-PeCDF 1 i . i Ul . 1 13 1.2.3.4,7.8-HxCOD 1 ,..!•( i 1 13 1 ,2.3,6.7,8-HxCDD 1 , . , ! • ! , 1 13 1,2.3,7.8,9-HxCDD 1 , • , Ul iJ l3 . 1 1 1,2,3.7,8-PeCDF 1,2.3,4,7.8-HxCDF 1 , . , Ul 1 ,2.3.6,7.8-HxCOF 1 , ,.!•!, 1 1 2.3 7,8.9-HxCOF 1 ... 1,2,3,4.6.7,8-HpCDD 1 , . , Ul Ci2-'.2.3.7,8-PeCDF Ci2-'-2'3'6'7-8-HxCDD Ci2-l,2,3.4,7,8-HxCDF C(2-1,2.3,4,6.718-HpCOD C,2-'.2.3.4,a,7,8-HpCDF 30,2-0000 13 C,2-OCDF 1 . , , !•! , 1 1 ,2,3.4.6.7,8-HpCDF Ci2-'.2.3.7.8-PeCDD 1 , . , 1*1 , 1 2.3,4.6.7,8-HxCDF 3C,2-2.3.7.8-TCDD >C12-2.3,7,8-TCDF ULi_J . 1 1.2.3.4,7,8,9-HpCOF 1 . . , Ul . 1 OCDO 1 , , , Ul OCDF 1 , , , Ul , I . 1 REMARKS Jj Concentration reported is based on total axtractable iipid (g). 2j ND - Not Detected. TR - Trace. PG - Positive Quantifiable. Figure 6. Analysis report form. A-51 SPIKED LEVEL (pg) PERCENT (%) RECOVERY �RAW DATA SUMMARY FOR DETERMINATION OF 1.2.3.7.8-PeCDD IN HUMAN ADIPOSE TISSUE Sample no. Sample weight (xx.xx q) Extractable lipid content (XX. X % ) Analysis date Amount C12-PeCDD (pg) 13 13 C12-PeCDO n/z 332 13 Cl2-PeCDD m/z 334 Ion ratio 366/368 1,2,3,7,8PeCDD m/z 354 1,2,3,7,8 m/z 356 I tn ro Value reported as concentration in extractable lipid. Figure 7. Example of raw data summary format for the determination of 1,2,3,7,8-PeCDD in human adipose tissue. Ion ratio 354/356 1,2,3,7,8PeCOD cone. (pg/g) �TECHNICAL REPORT DATA (Please read Inunctions on the reverse before completing) 3. RECIPIENT'S ACCESSlON>NO. I a. NO. EPA 560/5-86-020 S. REPORT DATE 4. 7i7LS AND SUBTITLE Analysis for Polychlorinated Dibenzo-p_-dioxins (PCDD) and Dibenzofurans (PCDF) in Human Adipose Tissue: Method Evaluation Study September 17. 1986 6. PERFORMING ORGANIZATION CODE Midwest Research Institute 7.AUTHOR<s> js Stanley, RE Ayling, KM Bauer, MJ McGrath, TM Sack, and KR Thornberg 8. PERFORMING ORGANIZATION REPORT NO. a-JeRFOHMLNG ORGANIZATION NAME ANO ADDRESS 10. PROGRAM ELEMENT NO. Thawest Research Institute 425 Volker Boulevard Kansas City, MO 64110 11. CONTRACT/GRANT NO. 68-02-3938 68-02-4252 12. SPONSORING AGENCY NAME ANO ADDRESS Field Studies Branch (TS-798) Exposure Evaluation Division Office of Toxic Substances U.S. 8824-A(01) 13. TYPE OF REPORT ANO PERIOD COVERED Final Washinqton, DC 20460 14. SPONSORING AGENCY CODE Environmental Prntprt.inn Aqpnr.v 13. SUPPLEMENTARY NOTES J Remmers, Work Assignment Manager; J Breen, Project Officer 16. ABSTRACT This report focuses on the evaluation of an HRGC/MS analytical method for determination of 2,3,7,8-substituted polychlorinated dibenzo-p_-dioxins (PCDD) and dibenzofurans (PCDF) in human adipose tissue. This method will be used for analysis of samples from EPA's National Human Adipose Tissue Survey (NHATS) as part of a collaborative effort between EPA's Office of Toxic Substances and the Veterans Administration. The method was evaluated using aliquots of a bulk lipid matrix that was extracted from human adipose tissue. The results of the replicate analysis of spiked and unspiked homogenized human adipose tissue matrix demonstrate that the analytical method produces accurate and precise data for 17 specific 2,3,7,8-substituted PCDD and PCDF (tetra- through octachloro homologs) congeners. The endogenous or background levels of the PCDD and PCDF congeners in the homogenized adipose lipid matrix were estimated through regression analyses of measured versus spiked concentrations for each compound. This unspiked matrix will be used as a control sample with each batch of samples analyzed. KEY WORDS ANO DOCUMENT ANALYSIS 17. DESCRIPTORS b.lOENTIFIERS/OPSN ENDED TERMS C. 19. SECURITY CLASS < rhu Report/ 21. NO. OP PAGES COSATt Fteid'CfOUp D olychlorinated dibenzo-jp_-dioxin (PCDD) Polychlorinated dibenzofuran (PCDF) Human adipose tissue Method evaluation 2,3,7,8-Tetrachlorodibenzo-g-dioxin (2,3,7,8-TCDD) I «3.7.8-Tetrachl orodi benzofuran (2.3.7.f 13. ^ i S T P i a u T i O N STATEMENT Release unlimited Unclassified 20. SECURITY CLASS Unclassified EPA fttm 2220-1 (3-73) 145 22. � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 197 Folder The folder containing the original item. 5526 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 Stanley, John S. et al. 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 1986-10-01 Title A name given to the resource Analysis for Polychlorinated Dibenzo-p-Dioxins (PCDD) and Dibenzofurans (PCDF) in Human Adipose Tissue: Method Evaluation Study: Final Report with attached letter transmitting the report to Alvin L. Young, from Janet C. Remmers, Field Studies Branch, Expo ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/40acfd2554f623d0f152b8286123a60b.pdf e2abc03f964ecc5b150083b7f6d4a72e PDF Text Text Item D Number 0552 Author Stanley, John S. Corporate Author Midwest Research Institute Report/Article Title Analysis for Polychlorinated Dibenzo-p-Dioxins and Dibenzofurans in Human Adipose Tissue: Method Evaluation Study: Draft Final Report with attached letter transmitting the report to Alvin L. Young, from Janet C. Remmers, Field Studies Branch, Exposure Evaluation Division, United States Environmental Protection Agency 2 Not Scanned Journal/Book Title Year 1986 Month/Day Color D 143 DOSCrtotOU NOtOS EP/flk Prime Contract No. 68-02-3938, Work Assignment No. 46, MRI Project No. 8501 -A(46). See Item 5526 for final version Tuesday, March 19,2002 Page 5522 of 5611 �A\ Sfa | UNITED STATES ENVIRONMENTAL PROTECTION AGENCY ,/ WASHINGTON, D.C. 20460 JUNE 10, 1986 OFFICE OF PHT1CIDES AND TOXIC SUBSTANCES Alvin Young, Ph.D., Lt Col., USAF Senior Policy Analyst for Life Sciences Executive Office of the President Office of Science and Technology Policy Room 5005 New Executive Office Building Washington, DC 20506 Dear Dr . Young: Enclosed for your information is a report entitled, "Analysis for Polychlorinated Dibenzo-p_-dioxins and Dibenzofurans in Human Adipose Tissue: Method Evaluation Study", May 30, 1986. This report includes the analytical protocol which will be used in the EPA/VA dioxin study of adipose tissue from veterans and nonveterans. Also included is the method evaluation data from a single laboratory study. If you have any comments or questions, please feel f r e e to call me at 382-3583. Sincerely, fanet C. Remmers Field Studies Branch Exposure Evaluation Division (TS-798) Enclosure cc: M a r t i n Halper Joseph Breen (-11 �MIDWEST RESEARCH INSTITUTE ANALYSIS FOR POLYCHLORINATED DIBENZQ-£-DIOXINS AND DIBENZOFURANS IN HUMAN ADIPOSE TISSUE: METHOD EVALUATION STUDY DRAFT FINAL REPORT EPA Prime Contract No. 68-02-3938 Work Assignment No. 46 MRI Project No. 8501-A(46) For Field Studies Branch, TS-798 Office of Toxic Substances U.S. Environmental Protection Agency 401 M Street, S.W. Washington, DC 20460 Attn: Ms. Janet Remmers, Work Assignment Manager Dr. Joseph J. Breen, Project Officer MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD, KANSAS CITY, MISSOURI 64110 • 816 753-7600 �ANALYSIS FOR POLYCHLORINATED DIBENZO-jD-DIOXINS AND DIBENZOFURANS IN HUMAN ADIPOSE TISSUE: METHOD EVALUATION STUDY by John S. Stanley, Randy E. Ayling, Karin M. Bauer, Michael J. McGrath, Thomas M. Sack, and Kelly R. Thornburg DRAFT FINAL REPORT EPA Prime Contract No. 68-02-3938 Work Assignment No. 46 MRI Project No. 8501-A(46) For Field Studies Branch, TS-798 Office of Toxic Substances U.S Environmental Protection Agency 401 M Street, S.W. Washington, DC 20460 Attn: Ms. Janet Remmers, Work Assignment Manager Dr. Joseph J. Breen, Project Officer MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD, KANSAS CITY, MISSOURI 64110 • 816 753-7600 �DISCLAIMER This document is a preliminary draft. It has not been released formally by the Office of Toxic Substances, Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. It is being circulated for comments on its technical merit and policy implications. �PREFACE This report provides a summary of the results from a method evaluation study for the determination of 2,3,7,8-substituted polychlorinated dibenzo-p_-dioxins (PCDDs) and dibenzofurans (PCDFs) in human adipose tissues. This method evaluation is an integral part of a collaborative program between the U.S. Environmental Protection Agency's Office of Toxic Substances and the Veterans Administration to determine if significant differences exist in the 2,3,7,8-substituted PCDD and/or PCDF levels in human adipose tissues for Vietnam veterans compared to the general adult male population. The study design will focus on specimens within EPA's National Human Adipose Tissue Survey (NHATS) repository. The method evaluation described in this report was necessary to establish method performance (accuracy and precision) before proceeding with actual sample analysis. This method evaluation study was completed under EPA Contract No. 68-02-3938, Work Assignment 46, "Analysis for Dioxins and Furans in Human Adipose Tissue," Ms. Janet Remmers, Work Assignment Manager, and Dr. Joseph Breen, Project Officer. MIDWEST RESEARCH INSTITUTE Paul C. Constant Program Manager Approved: Jack Balsinger QuaH\y Assurance Coordinator E. Going, Director' Chemical Sciences Department May 30, 1986 11 �TABLE OF CONTENTS Page I. Introduction II. Summary III. Recommendations 4 IV. Experimental 5 A. B. C. D. E. F. G. 5 5 10 12 17 20 20 V. 1 • Preparation of Homogenized Tissue Analytical Standards Analytical Procedure HRGC/MS Analysis Data Interpretation Quality Assurance/Quality Control (QA/QC) Preliminary Method Studies. Results 23 A.. Analytical Results B. Statistical Analysis VI. Quality Assurance/Quality Control (QA/QC) A. B. C. D. VII. 3 Initial Calibration Daily Verification of Response Factors Blanks Absolute Recoveries of the Internal Quantisation Standards References 23 23 62 62 64 64 76 81 Appendix A - Analytical Protocol for Determination of PCDDs and PCDFs in Human Adipose Tissue m A-l �LIST OF FIGURES Figure 1 2 Page Comparison of the HRGC/MS-SIM reconstructed ion chromatogram (RIC) from the analysis of unspiked homogenized human adipose tissue matrix and a calibration standard for PCDDs and PCDFs Example of the TCDF (m/z 304) and TCDD (m/z 320) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 3 4 5 6 7 8 9 10 11 12 30 31 Example of the PeCDF (m/z 338) and PeCDD (m/z 354) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 32 Example of the HxCDF (m/z 374) and HxCDD (m/z 390) HRGC/MS-SIM elusion profiles in unspiked and spiked human adipose 33 Example of the HpCDF (m/z 408) and HpCDD (m/z 424) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 34 Examples of the OCDF (m/z 442) and OCDD (m/z 458) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose 35 Measured concentrations versus concentrations of 2,3,7,8TCDD spiked into the homogenized human adipose lipid matrix . 36 Measured concentrations'versus concentrations of 1,2,3,7,8-PeCDD spiked into the homogenized human adipose lipid matrix 37 Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDD spiked into the homogenized human adipose lipid matrix 38 Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDD spiked into the homogenized human adipose lipid matrix 39 Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDD spiked into the homogenized human adipose lipid matrix 40 Measured concentrations versus concentrations of 1,2,3,4,6,7,8-HpCDD spiked into the homogenized human adipose lipid matrix. 41 �LIST OF FIGURES (continued) Figure 13 14 15 16 17 18 19 20 21 22 23 24 25 Page Measured concentrations versus concentrations of OCDD spiked into the homogenized human adipose lipid matrix. . 42 Measured concentrations versus concentrations of 2,3,7,8-TCDF spiked into the homogenized human adipose lipid matrix 43 Measured concentrations versus concentrations of 1,2,3,7,8-PeCDF spiked into the homogenized human adipose lipid matrix 44 Measured concentrations versus concentrations of 2,3,4,7,8-PeCDF spiked into the homogenized human adipose lipid matrix 45 Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. . . . . 46 Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix 47 Measured concentrations versus concentrations of 2,3,4,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix 48 Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDF spiked into the homogenized human adipose lipid matrix 49 Measured concentrations versus concentrations of 1,2,3,4,7,8,9-HpCDF spiked into the homogenized human adipose lipid matrix 50 Measured concentrations versus concentrations of 1,2,3,4,6,7,8-HpCDF spiked into the homogenized human adipose lipid matrix 51 Measured concentrations versus concentrations of OCDF spiked into the homogenized human adipose lipid matrix. . 52 Method accuracy estimates as determined from the slopes of the least squares regression lines for the 17 target PCDD and PCDF analytes 54 Control charts showing response factors by date for 2,3,7,8-TCDF and 2,3,7,8-TCDD . . 65 �LIST OF FIGURES (continued) Figure 26 27 28 29 30 31 Page Control charts showing response factors by date for 1,2,3,7,8-PeCDF and 2,3,4,7,8-PeCDF 66 Control chart showing response factors by date for 1,2,3,7,8-PeCDD 67 Control charts showing response factors by date for 1,2,3,4,7,8-HxCDF and 1,2,3,6,7,8-HxCDF 33 34 68 Control charts showing response factors by date for 2,3,4,6,7,8-HxCDF and 1,2,3,7,8,9-HxCDF 69 Control charts showing response factors by date for 1,2,3,4,7,8-HxCDD and 1,2,3,6,7,8-HxCDD 70 Control chart showing response factors by date for 1,2,3,7,8,9-HxCDD 32 , 71 Control charts showing response factors by date for 1,2,3,4,6,7,8-HpCDF and 1,2,3,4,7,8,9-HpCDF 72 Control chart showing response factors by date for 1,2,3,4,6,7,8-HpCDD 73 Control charts showing response factors by date for OCDF and OCDD 74 �LIST OF TABLES Table 1 Page Analytical Standards Available for the Method Evaluation Studies 7 2 Concentration Calibration Solutions 8 3 Native PCDD and PCDF Spiking Solution 9 4 Internal Standard Spiking Solutions 11 5 HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis . . . 13 6 Ions Monitored for HRGC/MS Analysis of PCDD/PCDF 14 7 Typical Daily Sequence for PCDD/PCDF Analysis 16 8 Ion Ratios for HRGC/MS Analysis of PCDD/PCDF 18 9 Summary of the Results of the Sample Preparation Method Evaluation Using Carbon-14 PCDDs 22 Spiked Versus Measured Concentrations of 2,3,7,8-TCDF and 2,3,7,8-TCDD in Homogenized Human Adipose Lipid Samples 24 Spiked Versus Measured Concentrations of 1,2,3,7,8-PeCDF, 2,3,4,7,8-PeCDF, and 1,2,3,7,8-PeCDD in Homogenized Human Adipose Tissue Samples 25 Spiked Versus Measured Concentrations of 1,2,3,4,7,8-; 1,2,3,6,7,8-; 2,3,4,6,7,8-; and 1,2,3,7,8,9-HxCDF in Homogenized Human Adipose Lipid Matrix 26 Spiked Versus Measured Concentration of 1,2,3,4,7,8-; 1,2,3,6,7,8-; and 1,2,3,7,8,9-HxCDD in Homogenized Human Adipose Lipid Samples 27 10 11 12 13 14 Spiked Versus Measured Concentrations of 1,2,3,4,6,7,8HpCDF, 1,2,3,4,7,8,9-HpCDF, and 1,2,3,4,6,7,8-HpCDD in Homogenized Human Adipose Lipid Samples 15 28 Spiked Versus Measured Concentrations of OCDF and OCDD In Homogenized Human Adipose Lipid Samples 29 16 Regression Line Slopes with 95% Confidence Limits 55 17 Results of the Analysis of the Low and High Level Native Spike Solutions. 56 vii �LIST OF TABLES (continued) Table Page 18 Background Level Estimates with 95% Confidence Limits . . . 58 19 Day-to-Day Precision of Analysis of Specific Sample Extracts for Tetra- and Pentachloro PCDF and PCDD . . . . 59 Day-to-Day Precision of Analysis of Specific Sample Extracts for Hexa- and Heptachloro PCDF and PCDD 60 Day-to-Day Precision of Analysis of Specific Sample Extracts for OCDF and OCDD 61 Relative Response Factors (Grand Means) Determined from Multipoint Concentration Calibration Standards 63 Summary of Results from the Analysis of a Laboratory Method Blank 75 20 21 22 23 24 Recovery of Radiolabeled PCDDs from Precleaned Activated Alumina 25 26 77 Absolute Recoveries of the Internal Quantisation Standards from the Human Adipose Lipid Matrix 78 Recovery of Carbon-14 Labeled 2,3,7,8-TCDD, 1,2,3,4,7,8HxCDD, and OCDD as a Function of Final Concentration Conditions 80 �I. INTRODUCTION The Environmental Protection Agency Office of Toxic Substances (EPA/OTS) and the Veterans Administration (VA) have established an interagency agreement to study the level of polychlorinated dibenzo-p_-dioxins (PCDDs) and dibenzofurans (PCDFs) in human adipose tissues. The occurrence and levels of PCDDs and PCDFs with chlorine substitution in the 2,3,7,8 positions (especially 2,3,7,8-TCDD) of the parent molecules are of primary interest. As part of this interagency effort, it has been proposed to use selected adipose tissue samples that were collected for the Field Studies Branch (FSB) of EPA's Office of Toxic Substances (OTS) through the National Human Adipose Tissue Survey (NHATS) to determine exposure to PCDDs and PCDFs. The available adipose tissues include specimens obtained from young men whose age indicates that they could have served in Vietnam and could have been exposed to Agent Orange. The tissues were originally collected as part of a broadly based and statistical random sampling of the continental United States. The analysis of these tissues may provide information on the differences of exposure of the general adult male population and Vietnam veterans to the 2,3,7,8-substituted PCDDs and PCDFs. The overall objectives of the proposed EPA/VA collaborative studies are: 1. Evaluate the reliability, accuracy, precision, and sensitivity of a proposed method for the determination of 2,3,7,8substituted PCDDs and PCDFs (tetra- through octachloro homologs) in human adipose tissue; 2. Determine if these compounds can be detected in adipose tissues of the American male adult population; and 3. Determine if individuals with military service in Vietnam have significantly different levels of 2,3,7,8-substituted PCDDs and PCDFs (particularly 2,3,7,8-TCDD) than other American men. As a prelude to this work assignment, MRI conducted an extensive literature review of applicable analytical methods and conducted a meeting with recognized experts in this field to identify critical aspects of analytical methodology.1'2 Based on the information gathered through the literature review and the meeting with the recognized experts, a special report was prepared for OTS proposing 3 a framework for an analytical method for analysis of human adipose tissues. Several studies have been completed since the issuance of that report which reflect the advances in analytical techniques for adipose tissue analysis.4 16 The salient features of these methods have been combined into a single protocol for the routine analysis of tetra- through octachloro PCDDs and PCDFs at the low-parts-per-trillion level for the EPA/VA tissue study. This report focuses on a method evaluation study that was conducted to achieve the first objective of the interagency agreement. Clarification �of method performance is necessary before proceeding with the analysis of actual samples retrieved from the NHATS repository. This report includes a summary of the method evaluation study results (Section II). Recommendations to be implemented before proceeding with the actual tissue samples from the NHATS repository are presented in Section III. A description of the actual experimental procedures is provided in Section IV. Results of sample analyses are summarized in Section V, and quality assurance/quality control (QA/QC) aspects of the study are detailed in Section VI. Pertinent references are listed in Section VII. Appendix A contains the detailed analytical protocol that will be followed for the analysis of the NHATS specimens designated in the study design to be provided by EPA/VA. �II. SUMMARY The results of the replicate analysis of spiked and unspiked homogenized human adipose tissue matrix demonstrate that the analytical method produces accurate and precise data for 17 specific 2,3,7,8-substituted PCDD and PCDF (tetra- through octachloro homologs) compounds. Data are reported for three or four replicate analyses of samples spiked at three different concentration levels. The endogenous or background levels of the PCDD and PCDF congeners in the homogenized adipose lipid matrix were estimated through regression analyses of measured (found) versus spiked concentrations for each compound, The analytical method is capable of providing quantitative data for tetra- through octachloro PCDD and PCDF congeners to concentration levels as low as 1 pg/g (tetrachloro congeners). However, an interference was noted at m/z 304 which coeluted with 2,3,7,8-TCDF, resulting in a detection level of approximately 4 pg/g. Average absolute recoveries of the internal quantisation standards ranged from 52% for 13Ci2-TCDD up to 89% for 13C12-OCDD. The agreement of the measured concentrations versus the spiked concentrations for each PCDD and PCDF congener demonstrates that the internal standard quantisation procedure provides an accurate measure of concentration which is independent of the absolute recovery. Final concentration conditions were noted to have pronounced effect on the absolute recoveries of the lower chlorinated compounds, particularly 2,3,7,8-TCDD. Experiments with carbon-14 labeled 2,3,7,8-TCDD demonstrated that final concentration at temperatures of 55 to 60°C resulted in recoveries as low as 54% while the same procedure conducted at ambient conditions resulted in greater than 90% recovery. Analysis of method and reagent blanks provided information on potential artifacts in the sample preparation scheme. Additional experiments were conducted with carbon-14 labeled PCDDs to evaluate the cleanup efficiency and recovery of PCDDs from chromatographic materials, particularly acidic alumina. �III. RECOMMENDATIONS Some minor modifications have been made in the written protocol (Appendix A) that were not included in this phase of the method validation. These include: a cleanup procedure for activated acidic alumina prior to fractionation of sample extracts to remove artifacts; and final concentration of the sample extracts using nitrogen blowdown at room temperature rather than heating to 55-60°C. The spiking solutions used to prepare the spiked quality control samples should be submitted for replicate (minimum of three/per spike level) HRGC/MS analysis to assist the interpretation of positive or negative bias in the accuracy of QC sample data. The accuracy bounds should be extended to 50-130% from 50-115% as specified in the draft quality assurance program plan. The method should include additional internal quantisation standards to pair with the HpCDF 13 OCDF congeners. Also, an additional internal recovand ery standard, possibly C12-l,2,3,4,7,8-HxCDD, is required to provide better estimates of absolute method recovery. These additional compounds, if available, will be incorporated into the method before initiating sample analyses. �IV. EXPERIMENTAL A. Preparation of Homogeni zed Jissue A bulk lipid sample was prepared from the extracts of human adipose tissue samples collected through the NHATS program. The adipose tissue samples have been stored in a deep freezer at approximately -10°C since collection. The homogenized tissue extract or bulk lipid was used in this method evaluation study for preparation of replicate samples spiked with varying levels of specific PCDD and PCDF isomers. This homogenized matrix will also be used for preparing control and spiked quality control samples for the actual NHATS sample analysis phase of the program. A total of 2,465 g of adipose tissue was extracted, dried, and concentrated to yield 1,652 g (62% of original weight) of homogenized lipid. Specific procedures for preparing this matrix are described below. The adipose tissue samples were thawed at room temperature for 1 to 2 h. Portions of the samples were added to a blender cup of a Waring® blender and covered with methylene chloride. The volume of methylene chloride was approximately equal to the sample volume (100 to 200 ml). This mixture was blended at high speed for approximately 10 min, and the contents were transferred to a 500-mL Erlenmeyer flask and further blended with a Tekmar® Tissumizer, also at high speed for 10 min. A powder funnel was plugged with a wad of glass wool (si 1 anized, methylene chloride extracted) and filled with ~ 50 g of sodium sulfate (heated overnight to 600°C in a muffle furnace). The sodium sulfate was wetted with methylene chloride prior to elution of the sample extract. The dried effluents were refiltered in the same way using a fresh bed of sodium sulfate to remove particulate and residual water. The samples were transferred to 1-L round bottom flasks, and the solvent was removed by rotary evaporation. The water bath on the rotary evaporator was kept at 60°C using a thermostatted heating element. Once the solvent appeared to have been removed (constant volume in flask, no visible condensation in condenser), the heating and evaporation process was continued for at least 2 h. The flask and contents were removed and stored in a refrigerator. The extracted lipid solidified upon refrigeration and was visually checked for homogeneity. No precipitates or phase separation was observed. The lipid residue was allowed to liquify at room temperature and was transferred to a 4-L glass bottle with a Teflon®-lined lid. The lipid residue was brought to room temperature and heated just enough to allow the lipid to achieve an oily state prior to aliquotting portions for the method evaluation studies. B. Analytical Standards Analytical standards including native PCDD and PCDF congeners, stable isotope (carbon-13) labeled standards and radiolabeled (carbon-14) standards were purchased from Cambridge Isotope Laboratories, Woburn, Massachusetts, and Pathfinder Laboratories, St. Louis, Missouri. The 2,3,7,8TCDD was received from the EPA Reference Materials Branch as a solution. The �other native PCDD and PCOF congeners were received as 1-mg neat standards. The stable and radio!abeled isotopes were received as solutions. Table 1 provides a summary of the standards used for this study. Stock solutions of the individual PCDD and PCDF congeners were prepared from the neat standards. The neat materials were weighed using a Cahn 27 electrobalance calibrated versus a 1-mg (Class M) standard. The neat compounds were transferred to glass vials and were dissolved in 2.0 to 3.0 ml of toluene (Burdick and Jackson, distilled in glass). Toluene was added to each standard using volumetric pipettes (Class A). The OCDD required dilution to 10.0 ml using a 50:50 mixture of toluene and anisole. A working solution consisting of the 17 native PCDD and PCDF congeners was prepared at a concentration of 2 ug/mL for the TCDD, TCDF, PeCDD, and PeCDF congeners, 5 ug/ml for the HxCDD, HxCDF, HpCDD, and HpCDF congeners, and 10 |jg/mL for the OCDD and OCDF. The working solution was used to prepare both the lipid matrix spiking solution and the calibration standards. The stable isotope labeled internal standards were obtained as^solutions at 50 (jg/mL concentration with the exception of the 13Ci2-OCDD, which was provided at 10 |jg/mL. Separate working solutions containing mixtures of the carbon-13 labeled PCDDs and PCDFs were prepared for use in the calibration standards and the sample spiking solutions. The carbon-14 radiolabeled PCDDs were used for preliminary method evaluation studies. The specific activity of the 14C-2,3,7,8-TCDD (117.56 mCi/mmole) was high enough to allow recovery studies at spike levels equivalent to 10 pg/g for a 10^-g sample. 1. Calibration Standards Eight concentration calibration standards containing the 17 native and the 9 carbon-13 labeled internal standards were prepared for determining the consistency of response factors for the native PCDDs and PCDFs versus the corresponding carbon-13 congeners. Table 2 presents a summary of the calibration standards prepared for the method calibration study. 2. Spi king Solutions a. Native PCDD and PCDF A solution containing the 2,3,7,8-substituted PCDD and PCDF congeners was prepared for spiking the homogenized lipid materials for the method evaluation study. Table 3 specifies the levels of each of the native PCDD and PCDF congeners present in this solution. �Table 1. Analytical Standards Available for the Method Evaluation Studies Compound Source Native 2,3,7,8-TCDD EPA QA Reference Materials Branch 2,3,7,8-TCDF Cambridge Isotope Laboratories 1,2,3,7,8-PeCDD Cambridge Isotope Laboratories 1,2,3,7,8-PeCDF Cambridge Isotope Laboratories 2,3,4,7,8-PeCDF Cambridge Isotope Laboratories 1,2,3,4,7,8-HxCDD Cambridge Isotope Laboratories 1,2,3,6,7,8-HxCDD Cambridge Isotope Laboratories 1,2,3,7,8,9-HxCDD Cambridge Isotope Laboratories 1,2,3,4,7,8-HxCDF Cambridge Isotope Laboratories 1,2,3,6,7,8-HxCDF Cambridge Isotope Laboratories 1,2,3,7,8,9-HxCDF Cambridge Isotope Laboratories 2,3,4,6,7,8-HxCDF Cambridge Isotope Laboratories 1,2,3,4,6,7,8-HpCDD Cambridge Isotope Laboratories 1,2,3, 4,6, 7,8-HpCDF Cambridge Isotope Laboratories 1,2,3,4,7,8,9-HpCDF Cambridge Isotope Laboratories OCDD Cambridge Isotope Laboratories OCDF Cambridge Isotope Laboratories Lot/Code 20603 AWN 1203-74/EF-903C MLB-706-53/ED-950C AWN-729-21/EF-953C AWN-729-45/EF-956C 830244/ED-961C MLB-706-47/ED-960C MLB-706-73/ED-969C AWN-729-20/EF-964C MB 13106-7/EF-962-C MB 13106-47/EF-967-C MB 13106-3/EF-968-C MLB-706-21/ED-971C AWN-729-22/EF-973C MB-13-106-77/EF-975C 8465-F-982-C/EF-982C F2832/ED-980C 13 Ci2-Internal standards 2,3,7,8-TCDD Cambridge 2,3,7,8-TCDF Cambridge 1,2,3,7,8-PeCDD Cambridge 1,2,3,7,8-PeCDF Cambridge 1,2,3,6,7,8-HxCDD Cambridge 1,2,3,4,7,8-HxCDF Cambridge 1,2,3,4,6,7,8-HpCDD Cambridge OCDD Cambridge Isotope Isotope Isotope Isotope Isotope Isotope Isotope Isotope Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories Laboratories R00208/ED-900 R00236/EF-904 R00241/ED-955 R00221/EF-952 R00249/ED-966C R00234/EF-963C R00248/ED-972 R00263/ED-981 14 Ci2"Radiolabe1ed standards 2,3,7,8-TCDD Pathfinder Laboratories 1,2,3,4,7,8-HxCDD Pathfinder Laboratories OCDD Pathfinder Laboratories S.A. = specific activity. S.A.a = 117.56 mCi/mmole S.A. = 24.16 mCi/mmole S.A. = 20.50 mCi/mmole �ininminininininmin o ,E o co c o o in o in inininininininininin m inininincMCMCMCMCMCMCM CM CM CM in in CMCMCMCMCMeDeoeoeDeoeovDeoeoeocMCM o o o o in in in in in in in CM CM CM CD in o m o ooooinLnLninininininininocD in mininincMCMCMCMCMCMCMCMCMCMinin o o o o in in in in in in m CM CM CM o in o m oCDOOOOOOOOCDOOOOOO o ooooinLnininininininminoo m S- ,-. ,f co fU CJ L) c C O «0 5-P jjfc u e o o o o CD in in in in m in m CM CM CM o o o o in in in in in un m CM CM CM o in o m -a O O O CD in in in c. in in in m CM CM CM o in o •o in rH rH «o i. CO co o CM CM rH r-4 rH rH r H r H r H r H e M C M C M C M C M C M C M C M C M C M i n i n CM r-4 rH rH CM V) T3 S- • oCDOOOOOOOOCDOOOOOCD CD O O C D O O O O O O O O O O C D O O CM cMCMCMCMinininininininmininoo •* «t CM rH rH CJ CM rH rH rH T*H r~( i~H i~l t~4 i~~f r~f r~f r^l rH Csl CSJ c co 0) CJ (J c o CM rH rH 1-4 -M •i— c Q) o m rH rH .rH o ooocDinininininininininmocD c in o CO r-p- o o in m CM 000 rHt—f r H r H r H r H r H r H r H r H C M C M •r- CJ (0 O o in in m in in in m CM CM CM rH r H r H r H i - H C M C M C M C M C M C M C M C M C M C M i n i n ^_) ~J CJ ^— O >r_ JQ CD in r-4 rH inininininininininin in inininincMCMCMCMCMCMCMCMCMCMinin eo eo *.p s_ 0 CM C M C M C M C M C D C Q C D C O C D e D C D C D l O t O C M C M rH rH co C •£ O CD O o in in m in in m m CM CM TH in m in in CM CM CM ininininininininLnm m inininincMCMCMCMCMCMCMCMCMCMinin •r- CJ in c CM o m O OO o IX. C CD in rH r H r H r H r H C M C M C M C M C M C M C M C M C M C M i n i n eo —' ^x CD in in m rH rH rH oo r-4 CO CJ QJ («- .Q to a u. u. o oa QQQU_U_LL.U_OOCJ Q Q Q Q Q Q C3 O.O.O. QU_LJ_CJCJOC_)CJOOnia:ni r- •o SSS^miini^zoooocn a a) CO rH rH r-4 V) a u_ aQ +3 c. o +3 to 4«> •r- a U_ oQ 0 O CJ H- U X X 00 > -P m Is**. ooi^r^r^^vor^^-^r^vD^j-^'* f*^« CO CO ^ CO CO CO CO CO CO ^ CO CO CO S- Z CM C M r H r H C M r H r H i - H r H r H r H C M r H r H r H O O 0) * I I I en M ev CO r—i ^» i- co oo r^ r*~ eo ^~ •3r**^ r*^ co co co co co CO CO o <u o o ^— » 1 L_] N IN CV I tNO CJ CJ CJ CJ 0 CJ CO C O C M C M C O C M C M C M C M C M C M C O C M C M C M O a . * UJ «v CO CO CM CM CM CM CM o CO rH CO rH CO rH <0 ro -M CJ CJ . o. i. -a c 1 CJ CJ 3: IT 00 U. 0) <J) 1 1 Q 0- 0. 00 00 rO 1 1 - ' h- 00 oo r*- r^- CD I I C. CM I M CJ 31 O M *A Vi rH rH o o o i— i co CM «* rH i. st\ -P t— LH 1—4 1 CJ CO 00 �Table 3. Native PCDD and PCDF Spiking Solution Compound Concentration (pg/nl_) 2,3,7,8-TCDD 5 2,3,7,8-TCDF 5 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 5 5 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 5 12.5 1,2,3,6,7,8-HxCDD 12.5 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 12.5 12.5 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 12.5 12.5 2,3,4,6,7,8-HxCDF 12.5 1,2,3,4,6,7,8-HpCDD .. 12.5 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 12.5 12.5 OCDD OCDF 25 25 �b. Internal Standards Two different internal standard spiking solutions were prepared for quantisation of native PCDD and PCDF congeners. The compositions of each of the spiking solutions are presented in Table 4. The internal quantisation standards were spiked into the lipid aliquots prior to any cleanup procedures and hence were carried throughout the method exactly as the corresponding native congeners. The internal recovery standard was added in 10 uL of a keeper solution (tridecane) during final extract concentration prior to analysis. The recovery standard was used to measure the absolute method recoveries of the internal quantisation standards. C. Analytical Procedure The homogenized human adipose lipid matrix was allowed to come to room temperature and then warmed in a water bath until the matrix changed to an oily state. Approximately 10.0 g of the oily material was transferred by pipette to preweighed glass vials, and the actual weight of the lipid was determined to the nearest 0.01 g by difference using an analytical balance. Four 10.00-g aliquots were spiked with 20 uL of the native spiking solution presented in Table 3, another four aliquots were spiked with 50 uL of the same solution, and three additional aliquots were spiked with 100 uL of native PCDD and PCDF solution. These spikes were equivalent to concentrations ranging from 10, 25, and 50 pg/g in the lipid matrix for the tetra- and pentachloro PCDD and PCDF congeners up to 50, 125, and 250 pg/g for the OCDD and OCDF for the low, medium, and high level spikes. In addition to the spiked samples, three aliquots of the lipid material were transferred for determining the endogenous levels of each of the PCDD and PCDF congeners in the control matrix. Each of the sample aliquots was fortified with 100 pL of the internal quantisation standard spiking solution (Table 4). The spiked samples were each quantitatively transferred to 500-mL Erlenmeyer flasks using hexane. The residues were diluted with a total of 200 ml of hexane, and 100 g of sulfuric acid (H2S04) modified silica gel (40% w/w) was added to each solution with stirring. The mixtures were stirred for approximately 2 h, and the supernatants were decanted and filtered through filter funnels packed with anhydrous sodium sulfate (Na2S04). The H2S04 modified silica adsorbents were washed with at least two additional aliquots of hexane and dried by elution through Na2S04. The combined hexane extracts for each sample were eluted through columns consisting of the 40% H2S04 modified silica gel (4.0 g) and silica gel (1.0 g). The eluates were concentrated to approximately 15 ml and added to columns of acidic alumina (Bio-Rad, AG-4, 6.0 g). The acidic alumina columns were eluted first with 20 ml of hexane, which was collected but not analyzed, followed by elution with 30 ml of 20% methylene chloride in hexane. The PCDDs and PCDFs were eluted from the acidic alumina using the 20% methylene chloride in hexane. The PCDDs and PCDFs in the eluates were isolated from other chlorinated planar aromatics using columns (5-mL disposable pipettes containing 500 mg of 18% Carbopak C and Celite-545). The Carbopak C/Celite 10 �Table 4. Internal Standard Spiking Solutions Concentration (pg/(jL) Compound Internal quantitation standard8 13 5 13 5 13 5 13 5 C12-2,3,7,8-TCDD C12-2,3,7,8-TCDF C12-l,2,3,7,8-PeCDD C12-l,2,3,7,8-PeCDF 13 12.5 13 12.5 13 12.5 C12-l,2,3,6,7,8-HxCDD C12-l,2,3,4,7,8-HxCDF C12-1,2,3,4,6,7,8-HpCDD 13 C12-OCDD 25 Internal recovery standard 13 C12-1,2,3,4-TCDD 5 uSolution prepared in isooctane. Solution prepared in tridecane. 11 �columns were pre-eluted with 2 ml of toluene, I mi of 75:20:5 methylene chloride/methanol/benzene, 1 mL of 1:1 methylene chloride/cyclohexane, and 2 ml of hexane. The sample extracts (30 ml) were added to the columns, which were eluted with 2 ml of hexane, 1 ml of 1:1 methylene chloride/cyclohexane, and 1 ml of the 75:20:5 methylene chloride/methanol/benzene. These eluents were collected and combined but were not analyzed. The Carbopak C/Celite columns were turned upside down, and the PCDDs and PCDFs were eluted with 20 ml of toluene. The toluene was concentrated to less than 1 ml using flowing nitrogen and a heated water bath (55-60°C) and transferred to 1.0-mL conical vials using a solution of 1% toluene in methylene chloride. Tridecane (10 nD containing 500 pg of the internal recovery standard 13C12-1,2,3,4-TCDD was added as a keeper when the solution had concentrated to approximately 200 pL. The extracts were concentrated to final volume using nitrogen and the heated water bath. D. HRGC/MS Analysis The analyses of the spiked and unspiked lipid samples were completed ,using a Kratos MS50TC double-focusing magnetic sector mass spectrometer. The determination for the tetra- through octachloro homologs was achieved in a single analysis using the conditions described in Table 5. Table 6 provides the characteristic ions monitored for each PCDD and PCDF homolog. As noted from Table 6, the analysis requires five different parameter descriptions that were switched automatically during the course of the analysis. Parameters monitored included two characteristic molecular ions for each PCDD and PCDF homolog and the corresponding carbon-13 labeled internal standard. In addition, a fragment ion of perfluorokerosene (PFK) was monitored throughout each analysis to ensure that proper mass calibration was maintained. The parameter descriptors also included an ion characteristic of specific homologs of chlorinated diphenyl ethers to demonstrate that responses meeting the qualitative criteria for specific PCDF congeners were not due to these potential interferences. Triplicate analyses of six of the eight calibration solutions (Table 2) were completed, and the variability in relative response factors across this range was calculated. The analyst was required to demonstrate on a daily basis that the relative response factors (RRF) were in agreement within ± 20% of the established averages for 2,3,7,8-TCDD and 2,3,7,8-TCDF and within ± 30% of the average RRF values for the other congeners. The analyst was also required to determine column performance by analyzing a mixture of TCDD isomers before proceeding with sample analysis. Table 7 gives an example of the typical daily sequence for PCDD/PCDF analysis. 12 �Table 5. HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis Mass spectrometer Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: Overall SIM cycle time: 8,000 V 500 MA 70 eV -1,800 V 280°C £ 3,000 (10% valley definition) 1s Gas chromatograph Column coating: Film thickness: Column dimensions: He linear velocity: He head pressure: DB-5 0.25 Mm 60 m x 0.25 mm ID * 25 cm/sec 1.75 kg/cm2 (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 s 30 mL/min 6 mL/min 270°C 300°C 1-2 ML 200°C 2 min 200°C to 330°C at 5°C/min 13 �Table 6. Ions Monitored for HRGC/MS Analysis of PCDD/PCDF Descriptor Al ID Mass TCDF 303.902 305.899 315.942 317.939 319.896 321.894 331.937 333.934 373.840 380.976 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 303.902 305.899 319.896 321.894 337.863 337.860 349.903 351.900 353.858 355.855 365.898 367.895 380.976 407.801 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.035 0.035 HxCDF 373.821 375.819 PFK (lock mass) 13 C12-HxCDF 380.976 385.861 387.859 389.816 391.813 401.856 0.080 0.080 0.080 0.080 0.080 13 C12-TCDF TCDD 13 C12-TCDD HxDPE PFK (lock mass) A2 TCDF TCDD PeCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD PFK (lock mass) HpCDPE A3 Nominal dwell time (sec) HxCDD 13 C12-HxCDD 403.853 443.759 OCDPE 14 0.080 0.080 0-. 080 0.080 0.080 �Table 6 (continued) Descriptor A4 ID Mass 380.976 389.816 391.813 407.782 409.779 419.822 421.819 423.777 425.774 435.817 437.814 429.768 431.765 477.720 13 C12-HpCDF HpCDD 13 C12-HpCDD 37 Cl4-HpCDD NCDPE PFK (lock mass) OCDF 13 C12-OCDF OCDD 13 C12-OCDD DCDPE 15 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 380.976 441.743 443.740 453.784 455.781 457.738 459.735 469.779 471.776 511.681 PFK (lock mass) HxCDD HpCDF A5 Nominal dwell time (sec) 0.060 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.060 �Table 7. Typical Daily Sequence for PCDD/PCDF Analysis 1. Tune and calibrate mass scale versus perfluorokerosene (PFK). 2. Determine column performance by injecting the TCDD isomer mixture. 3. Inject concentration calibration solution 2.5 to 12.5 pg/uL (CS-7) solution. 4. Inject blank (tridecane). 5. Inject samples 1 through "n". 6 Inject concentration calibration solution 2.5 to 12.5 pg/uL (CS-7) solution. 16 �E. Data Interpretation 1. Qualitative The HRGC/MS elution profiles of the tetra- through octachloro PCDD and PCDF homologs were established through the analysis of environmental sample extract (fly ash from a municipal waste incinerator). The characteristic ions for each homolog were plotted within the retention window established using this mixture. The criteria for identification of a response as a PCDD or PCDF were the coincidental response of the characteristic ions monitored within the established retention window, and within ± 20% of the theoretical ion ratio. Table 8 presents the range of ion ratios used for the qualitative criteria for the specific PCDD and PCDF homologs and internal standards. 2. Quantitation Quantisation of the specific PCDD and PCDF congeners was achieved using the respective internal quantisation standards. For example, TCDD was quantitated versus the 13C12-2,3,7,8-TCDD; PeCDD versus the 13C12-1,2,3,7,8PeCDD, etc. The HpCDF and OCDF responses were quantitated versus the carbon-13 labeled hepta- and octachloro dibenzo-p_-dioxin internal standards since the corresponding dibenzofuran internal standards were not available for this study. The absolute recovery of the internal quantisation standards was achieved using 13C12-1,2,3,4-TCDD. A second internal recovery standard, 37 Cl4-l,2,3,4,6,7,8-HpCDD, was evaluated but was not used for recovery measurements due to interference arising from the corresponding native HpCDD. Relative response factors (RRF) were calculated for each compound listed in Table. 2. The RRF values were calculated as shown in Equation 1. "crn x Cjr STD IS A H Xc IS x ^STD where A Eq. 1 = the sum of the area responses for the two characteristic ions of the standard compound; = the sum of the area responses for the two characteristic ions of the corresponding internal quantisation standard; = concentration of the internal quantisation standard (pg/|jL); and = concentratlon °f tne standard compound (pg/uL). A calibration curve was established using six concentration levels of standards; for example, the calibration curve for 2,3,7,8-TCDD was initially established with standards at concentrations of 1, 2.5, 5, 10, 50, and 100 pg/uL The 2.5 pg/uL standard was analyzed daily to verify response factors and instrument sensitivity. The RRF values for each of the internal quantisation standards were calculated versus the internal recovery standard, 13 C12-1,2,3,4-TCDD, using Equation 1. 17 �Table 8. Ion Ratios for HRGC/MS Analysis of PCDD/PCDF Compound TCDF 13 C12-TCDF TCDD 13 C12-TCDD PCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD HxCDF 13 C12-HxCDF HxCDD 13 C12-HxCDD HpCDF 13 C12-HpCDF HpCDD 13 C12-HpCDD OCDF 13 C12-OCDF OCDD 13 C12-OCDD Ions monitored Theoretical ratio 304/306 316/318 320/322 332/334 338/340 350/352 0.76 0.76 0.76 0.76 0.61 0.61 0.61 0.61 1.22 1.22 1.22 1.22 1.02 1.02 1.02 1.02 0.87 0.87 0.87 0.87 354/356 366/368 374/376 386/388 390/392 402/404 408/410 420/422 424/426 436/438 442/444 454/456 458/460 470/472 18 Acceptabl e range 0.61 0.61 0.61 0.61 0.49 0.49 0.49 0.49 0.96 0.96 0.96 0.96 0.82 0.82 0.82 0.82 0.70 0.70 0.70 0.70 - 0.91 0.91 0.91 0.91 0.73 0.73 0.73 0.73 1.46 1.46 1.46 1.46 1.22. 1.22 1.22 1.22 1.04 1.04 1.04 1.04 �The concentration of a PCDD or PCDF congener in a composite sample was calculated as shown in Equation 2. r - Asample x QIS L WT " A x RRF x Wt where Fn 9 L tq< CWT = wet tissue concentration of the PCDD or PCDF congener in each tissue (pg/g); A sample = sum of the area responses for the two characteristic ions of i pcD[) Qy pCDp congerier; A,o = sum of the area responses for the two characteristic ions of the respective internal quant i tat ion standard; QT<: = amount of the internal quantisation standard added to the i:> sample (500 pg of 13C12-TCDD or 2,500 pg of 13C12-OCDD); RRF = the relative response factor for the PCDD or PCDF congener from Equation 1; and Wt = mass of the sample (grams). The absolute recovery of the internal quantisation standard was calculated using Equation 3. Recovery ( ) = ,-/* % RS IS where K ; 4 , x 100 Eq. 3 IS ARS = sum of the area responses for the two characteristic ions of . the internal recovery standard, 13C12-1,2,3,4-TCDD; QR<. = amount of the internal recovery standard added to the final extract (500 pg); and RRF jg = response factor of the internal quantisation standard relative to the internal recovery standard. All data were qualified to reflect whether the compound was a positive quantifiable parameter, present as a trace value only, or was not detected. Positive quantifiable values were identified for responses greater than 10 times background signal to noise. Trace (TR) values were assigned to responses which were in the range of 2.5 to 10 times background signal to noise. A value of not detected (ND) was used to reflect that a response was not detected at greater than 2.5 times signal to noise. A limit of detection (LOD) was calculated for all trace and not detected values using the peak 19 �height response of the respective internal standard and the average measured signal to noise for the characteristic ions of the PCDDs and PCDFs. F. Quality Assurance/Quality Control (QA/QC) The QA/QC procedures included analysis of multipoint calibration concentration standards to establish relative response factor (RRF) 'curves for each of the 17 native PCDD and PCDF congeners. The mean RRF (RRF) values and instrument sensitivity were verified daily by bracketing the sample analyses with an injection of a standard that ranged from 2.5 pg/uL for 2,3,7,8TCDD and 2,3,7,8-TCDF up to 12.5 pg/uL. for OCDD and OCDF. The criterion for continuing with the sample analysis was agreement of the measured RRF value with the mean RRF within ± 20% for 2,3,7,8-TCDD and TCDF and ± 30% for all other PCDD and PCDF congeners. Other activities included the analysis of laboratory method blanks and reagent blanks and measurement of the absolute recoveries of the internal quantisation standards. G. Preliminary Method Studies Prior to analysis of the homogenized human adipose lipid matrix by HRGC/MS, several experiments were conducted to confirm that the sample preparation scheme was feasible. 1. Gravimetric Studies The first concern was the efficient removal of up to 10 g of lipid matrix from extracted adipose tissue. A series of experiments was conducted with 10-g lipid aliquots to demonstrate removal of lipid using the H2S04-Si02 slurry technique. Initially, 50 g of the acid modified silica was added to the lipid extract in 100 ml of hexane. The acid modified silica was noted to turn dark brown immediately on contact with the lipid solution. The hexane was recovered and the adsorbent was extracted with additional hexane. The extracts were combined and concentrated to 5 ml with Kuderna-Danish evaporators. The extract was eluted through a column of 4.0 g of acid modified silica and 1.0 g of silica with 45 mL of hexane. The acid modified silica was noted to be highly discolored throughout, and the extract required a second slurry treatment of the eluent with an additional 50 g of acid modified silica gel. The adsorbent from the second slurry procedure was noted to discolor significantly, indicating that lipid materials had not been efficiently removed from the first step of the procedure. The hexane supernatant from the second slurry cleanup was reduced in volume and taken to dryness in a preweighed glass vial. The final residue was measured at approximately 10 mg for duplicate samples, which translates into a removal efficiency of 99.9% based on the initial 10-g aliquot. This lipid cleanup procedure was modified such that 100 g of acid modified silica gel was used in the initial slurry cleanup, followed by elution of the resulting extract through a column containing 4.0 g of acid modified silica and 1.0 g of silica gel. The lipid removal efficiency of duplicate samples through the cleanup procedure was determined to average 99.8% 20 �(20 to 30 mg of the initial 10-g lipid remained after cleanup). The column cleanup step in this procedure did not exhibit any significant color change. Thus this step of the procedure was incorporated into the method as a check of the efficiency of lipid removal to prevent overloading of the acidic alumina fractionation column. 2. Carbon-14 Recovery Studies The carbon-14 radiolabeled PCDD standards listed in Table 1 were used to estimate overall method recoveries for the tetra- through octachloro homologs prior to proceeding with the HRGC/MS evaluation. Triplicate analyses (10-g aliquots of lipid materials) were conducted with each of the three radiolabeled standards. The first experiment addressed the recovery of the compounds from bulk lipid cleanup. Triplicate analyses using 10-g aliquots were completed for the three compounds at the following concentrations: 14 C-2,3,7,8-TCDD, 10 pg/g; 14C-l,2,3,4,7,8-HxCDD, 100 pg/g; and 14C-OCDD, 250 pg/g. The results of these analyses indicated that all compounds were recovered in the range of approximately 70 to 80%. Following this ^experiment, the total sample preparation procedure described earlier in this re*port was evaluated using triplicate analysis of 10-g lipid samples. Table 9 provides a summary of the results from this study. These data indicate that overall method recovery is limited by the initial bulk lipid removal procedure. This assumption is based on the similar recoveries of the carbon-14 labeled compounds noted for evaluation of the bulk lipid removal step as compared to the total sample preparation scheme. 21 �Table 9. Summary of the Results of the Sample Preparation Method Evaluation Using Carbon-14 PCDDs Spike levels Bulk lipid removal, recovery Analytes (pg/g) Total method recovery ( ) % 14 C-2,3,7,8-TCDD 10 68 75 14 100 79 66 250 82 76 C-l,2,3,4,7,8-HxCDD 14 C-OCDD * Average value for triplicate analyses taken through the total sample preparation scheme. Precision of measurements varied by less than ± 10% .(relative standard deviation). Average value for triplicate analyses taken through bulk lipid cleanup only. Precision of measurement varied by less than ± 6% (relative standard deviation). 22 �V. RESULTS A. Analytical Results The analytical results for the quantisation of the 17 target PCDD and PCDF 2,3,7,8-substituted congeners in the spiked and unspiked homogenized human adipose lipid samples are presented in Tables 10 to 15. These data demonstrate that 13 of the 17 congeners were definitely detected in the unspiked lipid matrix. Although 2,3,7,8-TCDF is reported as not detected, responses for the characteristic ions (m/z 304 and 306) greater than 10 times signal to noise were noted to be coincident with the internal standard, 13C12~ 2,3,7,8-TCDF. The ratio of the responses (m/z 304/306) in each of the triplicate analyses of the unspiked matrix were well outside the acceptable ratio of 0.90 to 0.61 established in Table 8. Figure 1 provides a comparison of the HRGC/MS-SIM responses noted for the unpsiked human adipose lipid matrix as compared to a concentration calibration standard. Figures 2 through 6 provide examples of the individual PCDD and PCDF responses observed for the unspiked lipid samples as compared to fortified matrices. In general, the precision of the replicate measurements at each spike level is good (relative standard deviations typically less than 10%) for PCDD and PCDF congeners that were detected with responses greater than 10 times signal to noise. The precision for estimated detection limits for 1,2,3,7,8-PeCDF (Table 11), 1,2,3,7,8,9-HxCDF (Table 12), and OCDF (Table 15) ranges from 21.6% to 43.1% as a result of little or no response at the specified retention window. B. Statistical Analysis The regression results for each of the 17 specific PCDD and PCDF congeners are plotted separately in Figures 7 to 23. These plots provide the results for the individual sample analyses, a line defining the results of a least squares regression analysis, and boundaries that depict the confidence limits for the range of spiked concentrations. The regression lines were obtained by the method of least squares using the sample measurements at the three spiking levels and the unspiked level. Two types of upper and lower 95% confidence limits or bounds were calculated for the least square regressions of measured (found) concentrations versus spiked levels. The first set of confidence limits (defined by the inner pair of curves closest to the regression line) is the 95% confidence bounds for the regression line. These bounds are interpreted as follows: The true regression line (as would be determined if the experiment were repeated a countless number of times at the same spiked levels) lies within these confidence limits unless the analytical results are sufficiently unusual to be among those expected to occur less than 5% of the time. The second set of confidence bounds, depicted by the outer pair of lines, constitutes the 95% confidence limits for the result of a single analysis at a particular spiking level. The interpretation is as follows: the result (reported value) of a single analysis of a sample spiked at a given level can be predicted to fall between these 95% confidence bounds unless the analytical result is among those sufficiently unusual to be expected less than 5% of the time. 23 �Table 10. Spiked Versus Measured Concentrations of 2,3,7,8-TCDF and 2,3,7,8-TCDD in Homogenized Human Adipose Lipid Samples 2,3,7,8-TCDF spike level 2,3,7,8-TCDF concentration (pg/g) (pg/g) 13 C12-TCDF absolute recovery ( ) % 13 2,3,7,8-TCDD spike level 2,3,7,8-TCDD concentration (pg/g) (pg/g) C12-TCDD absolute recovery (%) (.) 41a a (.) 41a (.) 40 (.) 41 0.1 1.8 59 63 59 60.3 2.3 3.8 0 0 0 10.7 11.4 13.1 11.7 1.2 10.5 53 52 46 50.3 3.8 7.5 10 10 10 10 14.3 14.8 13.6 13.8 14.1 0.5 3.9 61 67 71 78 69.3 7.1 10.3 10 10 10 10 23.4 22.8 24.7 22.5 23.3 1.0 4.1 53 53 53 62 55.3 4.5 8.1 25 25 25 25 30.8 30.8 30.7 28.7 30.2 1.1 3.5 59 75 62 72 67.0 7.7 11.5 25 25 25 25 40.8 40.6 40.3 38.3 40.0 1.2 2.9 48 53 51 60 53.0 5.1 9.6 50 50 50 57.7 59.4 55.8 57.6 1.8 3.1 64 48 58 56.7 8.1 14.3 50 50 50 65.8 72.1 67.6 68.5 3.3 4.8 55 43 48 48.7 6.0 12.4 0 0 0 Mean STD RSD (%) ro Mean STD RSD (%) ND ND ND ND -pi Mean STD RSD ( ) % Mean STD RSD ( ) % ND = not detected. Value in parentheses is the estimated limit of detection. A response of greater than 10 times signal-to-noise was noted for both characteristic ions (m/z 304 and 306) at the appropriate retention time for 2,3,7,8-TCDF. However, the ion ratio was considerably greater than the acceptable range of 0.61 to 0.90. �Table 11. Spiked Versus Measured Concentrations of 1,2,3,7,8-PeCDF, 2,3,4,7,8-PeCDF, and 1,2,3,7,8-PeCDD In Homogenized Human Adipose Tissue Samples 1.2,3,7,8-PeCDF spike level (pg/g) 0 0 0 10 10 10 10 Mean STD RSO ( ) % 2,3,4,7,8-PeCDF concentration (pg/g) (pg/g) 0 0 0 20.8 21.6 19.0 75 76 80 20.5 1.3 6.5 77.0 2.6 3.4 12.2 11.5 11.9 11.4 10 10 10 10 Mean STO RSO (X) 29.2 30.1 28.5 25.5 Mean STD RSD ( ) % ND = not detected. 51.3 55.9 55.5 54.2 2.5 4.7 . 90 63 84 75 50 50 50 54.9 48.5 41.7 48.2 62 81 87 84 64 74 70 69.8 6.1 8.8 69.3 5.0 7.3 The value in parentheses is the estimated limit of detection. 1,2,3,7,8-PeCDD concentration »3C,2-PeCDO absolute recovery ( ) % (pg/g) (pg/g) 0 0 0 20.2 19.9 18.1 51 54 57 19.4 1.1 5.7 54.0 3.0 5.6 32.1 37.6 30.8 30.4 60 57 55 62 30.2 1.9 6.3 58.5 3.1 5.3 48.0 46.1 49.3 43.5 55 60 57 64 46.7 2.5 5.4 59.0 3.9 6.6 69.7 72.2 72.8 58 54 56 71.6 1.6 2.3 56.0 2.0 3.6 10 10 10 10 78.5 11.3 14.4 74.6 62.9 71.9 1.2,3,7,8-PeCDD spike level 78.0 11.7 15.1 48.4 5.4 11.1 25 25 25 25 28.3 2.0 7.0 50 50 50 27.6 37.0 28.7 36.3 C12-PeCOF absolute recovery ( ) % 32.4 4.9 15.2 11.8 0.4 3.1 25 25 25 25 CJ1 NO ( . ) 11" NO ( . ) 08 NO (0.8) 13 2,3,4,7,8-PeCDF spike level ND ( . ) 09 0.2 21.6 Mean STD RSO ( ) % ro 1,2,3,7,8-PeCDF concentration (P9/g) 25 25 25 25 50 50 50 �Table 12. Spiked Versus Measured Concentrations of 1,2,3,4,7,8-; 1,2,3,6,7,8-; 2,3,4,6,7,8-; and 1,2,3,7,8,9-HxCDF in Homogenized Human Adipose Lipld Matrix 1,2,3,4,7,8HxCOF spike level (pg/g) 0 0 0 Mean STO 25 25 25 25 Mean STO RSD ( ) % 22.0 22.1 22.4 (pg/g) 0 0 0 46.7 52.0 48.8 49.0 Mean STD RSD (X) 125 125 125 90.2 89.1 90.3 90.7 25 25 25 25 156.3 0.9 0.6 (pg/g) 12.3 12.4 12.7 2.3,4.6,7,8HxCDF spike level (pg/g) 0 0 0 36.6 38.6 40.8 39.8 62.5 62.5 62.5 62.5 83.7 80.5 79.6 76.0 25 25 25 25 151.0 157.7 149.1 (pg/g) 4.9 4.2 4.2 1,2.3,7,8,9HxCOF spike level (pg/g) 0 0 0 32.3 33.6 32.0 32.5 62.5 62.5 62.5 62.5 74.7 75.5 74.2 71.1 25 25 25 25 152.6 4.5 2.9 HO = not detected. The value in parentheses reflects the estimated limit of detection. 141.9 141.6 151.0 144.9 5.3 3.7 (pg/g) a 13 Ct2-HxCDF absolute recovery () « NO ( . ) 05 NO ( . ) 07 NO (0.9) 55 57 52 55.7 2.5 4.6 34.8 28.5 30.9 29.2 30.9 2.8 • 9.1 59 57 60 67 60.8 4.3 7.2 125 125 125 82.7 89.4 82.2 76.3 60 63 63 70 82.7 5.4 6.5 62.5 62.5 62.5 62.5 73.9 2.0 2.6 125 125 125 1,2,3,7.8,9HxCOF concentration NO ( . ) 07 0.2 25.3 32.6 0.7 2.1 79.9 3.2 3.9 125 125 125 2,3,4,6,7,8HxCDF concentration 4.4 0.4 8.9 39.0 1.8 4.7 90.1 0.7 0.7 157.2 155.4 156.4 1,2,3,6,7,8HxCDF concentration 12.4 0.2 1.7 49.1 2.2 4.4 62.5 62.5 62.5 62.5 Mean STD RSD (X) (pg/g) 1,2,3,6,7.8HxCDF spike level 22.2 0.2 1.1 RSD (X) ro 1,2,3.4,7,8HxCDF concentration 64.0 4.2 6.6 143.0 144.1 163.4 57 54 57 150.1 11.4 7.6 56.0 1.7 3.1 �in Homogenized Human Adipose Lipid Samples 1,2,3,4,7,8-HxCDD spike level (pg/9) 0 0 0 25 25 25 25 (pg/g) 0 0 0 4. 78 52.0 53.7 52.7 62.5 62.5 62.5 62.5 Mean STO RSD (X) 8. 29 96.7 90.3 81.9 25 25 25 25 141.1 150.8 146.0 145.9 4.9 3.3 (pg/g) 157.0 162.0 154.0 1,2,3,7,8,9-HxCDD spike level (pg/g) 0 0 0 184.0 165.0 198.0 193.0 62.5 62.5 62.5 62.5 220.0 239.0 220.0 220.0 280 8. 299.0 266.0 284.3 16.8 5.9 (pg/g) 13 Ct2-HxCDD absolute recovery (X) 19.1 26.0 24.7 58 60 58 58.7 1.2 2.0 63.5 46.1 40.8 57.3 65 61 65 70 51.9 10.4 20.0 25 25 25 25 65.3 3.7 5.6 125 125 125 114.0 99.9 101.2 107.4 64 66 66 77 105.6 6.5 6.1 62.5 62.5 62.5 62.5 224.8 9.5 4.2 125 125 125 1,2,3,7,8,9-HxCDD concentration 23.2 3.7 15.8 185.0 1' 5 4. 7.9 8. 79 7.0 7.9 125 125 125 1.2,3,6,7,8-HxCDD concentration 157.7 4.0 2.6 51.6 2.6 5.1 Mean STD RSD (X) Mean STD RSD (X) 21.6 22.7 20.3 1,2,3,6,7,8-HxCDD spike level 21.5 1.2 5.5 Mean STD RSD (X) ro 1.2,3,4.7,8-HxCDD concentration (pg/9) 68.3 5.9 8.7 141.3 152.7 189.4 62 57 61 161.1 .25.2 15.6 60.0 2.6 4.4 �Table 14. Spiked Versus Measured Concentrations of 1,2,3,4,6,7,8-HpCDF, 1,2,3,4,7.8,9-HpCOF, and 1,2,3,4,6,7,8-HpCOO in Homogenized Hunan Adipose Lipid Samples 1 ?,3,4,6,7,8-HpCDF spike level (99 P/) 0 0 0 Mean STD RSD ( ) % (pg/g) 30.6 27.3 28.6 Mean STD RSO (X) (pg/g) 0 0 0 48.6 48.6 51.0 56.3 Mean STD RSD ( ) % 86.3 85.7 80.9 83.4 25 25 25 25 154.5 157.3 154.4 155.4 1.6 1.0 (pg/g) NO ( . ) 13* ND ( . )a 11* HD ( . ) 10 1,2,3,4,6,7,8-HpCDD spike level 26 23 23 25 62.5 62.5 62.5 62.5 60 60 59 58 119 126 121 122.4 3.7 3.0 ND = not detected. The value in parentheses is the estimated limit of detection. 13 C,z-HpCDD absolute recovery ( ) % (pa/g) 0 0 0 214.7 210.8 215.5 71 74 69 213.7 2.5 1.2 71.3 2.5 3.5 214.7 239.0 243.9 248.5 83 75 76 78 243.3 40 . 1.7 78.0 3.6 4.6 281.5 288.1 269.9 274.9 77 84 90 99 278.6 8.0 2.9 87.5 9.3 10.7 353.2 355.4 343.7 62 61 69 350.8 6.2 1.8 64 4.4 6.8 25 25 25 25 62.5 62.5 62.5 62.5 59.0 0.8 1.3 125 125 125 1,2,3,4.6,7,8-HpCDD concentration (pg/g) 24.1 1.4 5.7 84.1 2.5 2.9 125 125 125 1,2,3,4,7,8,9-HpCDF concentration ND ( . ) lla 0.2 13 51.1 3.6 7.0 62.5 62.5 62.5 62.5 a 1,2,3,4,7,8,9-HpCDF spike level 28.9 1.6 5.7 25 25 25 25 ro CO 1,2,3,4,6,7,8-HpCDF concentration 125 125 125 �Table 15. Spiked Versus Measured Concentrations of OCDF and OCDD in Homogenized Human Adipose Lipid Samples OCDF spike level (pg/g) 0 0 0 Mean STD RSD ( ) % 50 50 50 50 Mean STD RSD ( ) % 125 125 125 125 Mean STD RSD ( ) % 250 250 250 Mean STD RSD ( ) % OCDF concentration (pg/g) 4.9 2.3 2.6 (pg/g) (pg/g) 0 0 0 804 833 781 91 88 94 806.1 26.1 3.2 91.0 3.0 3.3 3.2 1.4 43.1 44.2 45.0 45.9 49.6 860.4 11.4 1.3 125 125 125 125 932 934 944 907 110.7 2.2 2.0 227.8 231.0 228.6 OCDD concentration 849 856 876 860 50 50 50 50 46.2 2.4 5.2 111.1 107.8 110.7 113.1 13 C12-OCDD abolute recovery () % OCDD spike level 929.1 15.7 1.7 250 250 250 100 87 91 91 92.3 5.5 6.0 90 96 102 104 98.0 6.3 6.5 1,080 1,140 1,070 1,096 34.7 3.2 229.1 1.7 0.7 29 69 67 74 70.0 3.6 5.2 �Umplked Hutmn Adipo» Tillue 1. 2. 3. . . . . . . t . 2.3.7.8-ICOf '3C|J-2.3.7.»-FCnf "CI2-I.J.3.4-TCDO 7.3.7.J-FCOD "CI2-2.3.7.MCDD 1,2.], 7,1.F.CDF "ClJ-1.2 3.7,1 -P.CDF 2.3,4. 7.». FrCDF l,2,3.7,t-NCOD '3C|}>l.2.3,7.a-r«CDD 1 . '3Cn-l.2.3.4,7.«-H.Crif Df 1 . 1.2.3.6.7 RIC 4. 2.J.4.i.7.e-M.CDr S. l,2.3.4,7.e-ll<CDD «. 1,2,3. 6,7. »-H»COD 7. '3Cl2-l.2.3.6.7.t-H.CDD I. l.!.3.7.B.»-HrCDl> '. 1,2.3,7,1,9-HiCDF 20. 1, 2,3.4. *.7.»-HpCDF 21. l,2,3,4.«.7.«-HpCDI> 'K. I3CI2-1. 2.3.4. 4.7,1-HpCOD 23. 37ci4.|.2,3,4(«,7,«.HpCDO ,24 " 25. OCOD 26. "CI2-OCOO 27. OCDF i 21.:2,23 1 14.17 1 2 7 4,5 \{ li 9, 10 I i r-.fl 880 IBM IS 12 18 19 1488 1288 20 1 2888 1688 SCAN Calibration Standard 11,12 1,2 P.IC 15,14,17 6,7 19 21,22,23 9.10 1888 1288 1488 1888 1988 2888 SCAN Figure 1. Comparison of the HRGC/MS-SIM reconstructed ion chromatogram(RIC) from the analysis of unspiked homogenized human adipose tissue matrix and a calibration standard for PCDDs and PCDFs. 30 �Uraplk.d Human Adipou 384. Spiked Human AdipoM |2,3.7.8-TCDF 384 350 1888 last 1188 1158 1200 1290 SCAN Umpilnd Human Adipax r '3C|2-lCDF /-2.3,7,8-ICDD AJL lee.eSpikwi Human Ad!po» 320 . I3C 12-2.3,7.8-ICDF-i - A A me y-2.3,7,8-TCDO A UN lisa i2w 1230 sew Figure 2. Example of the TCDF (m/z 304) and TCDD (m/z 320) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. 31 �Unpllwd Hunan AdlpoM 2.3,4,7.8-PeCDF 1.2,3.7,8-f.CDF \ Spilwd Hunan Adlpou 2.3.4.7,8-P.CDF 1.2.3,7,8-PeCDF 1 I'"'1 ' ' '"I ' 1859 nee USe 1288 12S8 1308 1358 1488 SCAN 1458 SCAN Uiapikld Hunan Adlpoi. 354 . .2,3.7,8-P«COO Spiked Human AdipoM l3C|2-l,2.3.7.8-PeCOF 334. I.2.3.7.8-P.COD USB 1288 USB 1388 1310 1468 Figure 3. Example of the PeCDF (m/z 338) and PeCDD (m/z 354) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. 32 �Unipikod Human Adipose |l,2.3,.4.7,8-HxCDF ll.2.3.6,7,8-HxCDF 374 . .8-H»CDF 1 ' i I '' Spik.d Human Adipou 1.2.3,4,7,8-HxCDF | 11.2.3.6.7.8-HxCDF [2.3,4,4.7.8-. \H»CDF 374 . 1.2.3,7,8,9-HxCDF 1343 1258 1398 1983 1358 H93 1480 J498 USB I \. ISM 1.556 1550 1609 SCAN Uraplked Human Adipose 1.2,3,6,7,8-HxCDD 1,2,3.4,7.8-HxCDD- 339 . 13C]2-l,2,3.4.7,8-HxCDF A A 1.2.3,7,8,9-H.CDD Spik.d Human Adipou 1,2.3.6,7.8-HxCDD 399 . 1.2.3,4.7,8-HxCDD 1.2.3.7,8.9-HxCDD l3 C|2-l.2,3,4,7,8-HxCDF|l 1 1358 1488 1438 1589 1559 1689 1658 SCAN Figure 4. Example of the HxCDF (m/z 374) and HxCDD (m/z 390) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. 33 �Urapilwd Human AdlpaM 11,2,3.4.6,7.8-HpCDF 498. Spllud Human Adipon 1.2,3,4.6.7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1458 15B8 1556 1669 1658 1766 1758 1809 SCAN Umplk.d Human Adlpou l,2.3.4.6,7.B-HpCDD 424 . Spilwd Human Adipou 1.2,3,4.6,7,8-HpCDD 424. •I 1586 1559 1686 ' ' "> '! '•" 1658 I 1706 • ' ' I 1730 "I"' 1806 I1 1838 SCAN Figure 5. Example of the HpCDF (m/z 408) and HpCDD (m/z 424) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. 34 �Urapilw) Hunan Adtpou 442. Spiked Human Mlpow OCDF 442 . '3C|2-OCDD| A) me (798 teee test ts«e 1950 2eee SCAN isee ISM zeea sow Umplk.d Hunan Adlpon lOCDD 438 . Spilwd Hunan Adipoj. lOCDD 4S8 . 1766 ITM teee IBM Figure 6. Examples of the OCDF (m/z 442) and OCDD (m/z 458) HRGC/MS-SIM elution profiles in unspiked and spiked human adipose. 35 �80 70 - Regression Line95% Confidence Limits for Regression Line 60 - d 0 50 - d CO en <D 0 d 0 u 30 - •a • 95 % Confide nee Li mi rs for Individual Anafyses d 10 0 0 10 I 30 30 I 40 Spiked Concentration (pg/g) lipid sample meas. Figure 7. Measured concentrations versus concentrations of 2,3,7,8-TCDD spiked into the homogenized human adipose lipid matrix. 50 �80 Regression Line 70 95% Confidence Limits for Regression Line 60 - d 0 50 d 0 0 CO & 0 o S3 d 3 o 30 '95% Confidence for individual Analyses 10 - "T ~ 0 10 20 D 30 i 40 Spiked Concentration (pg/'g) lipid sample meas. Figure 8. Measured concentrations versus concentrations of 1,2,3,7,8-PeCDD spiked into the homogenized human adipose lipid matrix. 50 �160 150 Regression Line' 140 130 95% Confidence Limits for Regression Line W a Vx' 120 110 100 90 80 - co CO G i3 0 o •a 70 60 50 •95% Confidence Limits for Individual Analyses 40 30 30 -t 10 - 0 20 40 n 60 80 100 Spiked Concentration (pg/g) lipid sample meas. Figure 9. Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDD spiked into the homogenized human adipose lipid matrix. 120 �Regression Line \ 95% Confidence Limits for Regression Line v.,/ d 0 0 co d o o •a d ^ 95% Confidence Limits for Individual Analyses 0 60 0 n 80 100 Spiked Concentration (pg/g) lipid sample meas. Figure 10. Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDD spiked into the homogenized human adipose lipid matrix. 120 �200 ~r a d o <sJ I, 4> 0 d 0 u •d d 3 o 190 180 170 160 150 140 H 130 130 110 H 100 90 H 80 70 60 50 40 30 ~ 20 4 10 0 -10 0 Regression Line 95% Confidence Li mil's • for Regression Line 95% Confidence Limits for Individual Analyses '"I" 20 40 D Figure 11. "T 60 80 100 Spiked Concentration (pg/'g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDD spiked into the homogenized human adipose lipid matrix. 120 �370 Regression Line /•""N W <0! a 95 % Confidence Limits for Regression Line v-> d 0 •H +» flj d a) o d o o -d d 95% Confidence Limits for Individual Analyses 200 0 100 20 n Spiked Concentration (pg/g) lipid sample meas. Figure 12. Measured concentrations versus concentrations of 1,2,3,4,6,7,8-HpCDD spiked into the homogenized human adipose lipid matrix. 120 �1.19 Regression Line- 1.1 95% Confidence Limitsfor Regression Line 1.05 - s~ 1- d 0.95 - 0 a rf & :; o o ro o O •0 •95% Confidence Limits for individual Analyses 0.75 0.7 - 0 r- j r_ 40 80 n j j 120 __ r j 160 300 Spiked Concentration (pg/g) lipid sample meas. Figure 13. Measured concentrations versus concentrations of OCDD spiked into the homogenized human adipose lipid matrix. 240 �60 Wi \ fcfi Regression Line- 50 - 95% Confidence^ Limits for Regression Line a 0 •H •P 30 0 co ti o o 20 - •0 •95% Confide nee Li mits for Individual Analyses d 10 0 0 i 10 I 20 30 Spiked Concentration (pg/g) lipid sample meas. Figure 14. Measured concentrations versus concentrations of 2,3,7,8-TCDF spiked into the homogenized human adipose lipid matrix. 50 �60 Regression Line- 50 U 95% Confidence Limitsfor Regression Line 40 - d o 30 - o d o o •d • 95 % Confidence Limits for Individual Analyses 10 - -10 0 I 10 T 20 30 i 40 Spiked Concentration (pg/g) D lipid sample meas. Figure 15. Measured concentrations versus concentrations of 1,2,3,7,8-PeCDF spiked into the homogenized human adipose lipid matrix. 50 �90 Regression Line- 80 - W) ft d 0 95%JTonfidence Limitsfor Regression Line 70 - 60 - 50 <D 0 d 0 o 40 - •a 30 95% Confidence Limits for Individual Analyses 20 10 I 0 10 I 20 i 30 i 40 Spiked Concentration (pg/g) D lipid sample meas. Figure 16. Measured concentrations versus concentrations of 2,3,4,7,8-PeCDF spiked into the homogenized human adipose lipid matrix. 50 �160 ~r 150 r~\ \ 140 130 - A 110 - •p 100 - t CTl 95% Confidence Limitsfor Regression Line 130 - d o •pa Regression Line- 90 - d 0 0 d 0 80 70 - u 60 - d 50 - 0 •95% Confidence Limits for Individual Analyses 40 30 20J 10 - "T 0 Figure 17. 20 40 60 80 100 Spiked Concentration (pgXg) n lipid sample meas. Measured concentrations versus concentrations of 1,2,3,4,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �__ 160 150 140 Regression Line- 130 0. d 0 rf 130 95% Confidence Limitsfor Regression Line 110 100 90 80 - 0 ti 0 u Tl d 3 0 70 60 50 95 % Confidence Limits for Individual Analyses 40 30 20 ^ 10 0 0 i 40 i 20 o Figure 18. i 60 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 130 �160 150 W 0. vx {3 0 >r4 +J at 140 130 120 Regression Line95% Confidence Limitsfor Regression Line 110 100 90 80 70 0 CO d 60 o 50 0 95% Confidence Limits for Individual Analyses 30 0 20 10 0 10 0 20 40 60 80 100 Spiked Concentration (pg/g) n lipid sample meas. Figure 19. Measured concentrations versus concentrations of 2,3,4,6,7,8-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �bi' P, s*> d o IO d fi) 0 d 0 u •d 170 160 150 140 130 120 110 100 90 Regression Line- 95% Confidence LimiHfor Regression Line 80 70 60 50 40 30 20 10 0 -10 -20 •95% Confidence Limits for Individual Analyses 0 20 40 n Figure 20. 60 80 100 Spiked Concentration (pg/g) lipid sample ineas. Measured concentrations versus concentrations of 1,2,3,7,8,9-HxCDF spiked into the homogenized human adipose lipid matrix. 120 �\ 95 for Regression Line ti 0 0 0 01 o d 0 95% Confidence Limits for Individual Analyses u •a d -10 80 0 D 100 Spiked Concentration (pg/g) lipid sample meas. Figure 21. Measured concentrations versus concentrations of 1,2,3,4,7,8,9-HpCDF spiked into the homogenized human adipose lipid matrix. 120 �170 160 150 140 W P. d 0 Regression Line- 130 95% Confidence Limitsfor Regression Line 130 110 100 - d 0) 0 d 0 u •d d 3 0 90 80 70 - 60 - _ _ _ for Individual Analyses 50 40 - fcl 30 20 H 10 0 20 40 D Figure 22. 60 1 80 100 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of 1,2,3,4,6,7,8-HpCDF spiked into the homogenized human adipose lipid matrix. 120 �240 220 200 Regression Line 180 H 95% Confidence Limits for Regression Line 160 d o 140 120 H d 0 100 - 0 Ol IN3 d 0 u -a d 95% Confidence Limits fpr Individual Analyses 80 60 -20 0 —, 80 n Figure 23. 120 160 200 Spiked Concentration (pg/g) lipid sample meas. Measured concentrations versus concentrations of OCDF spiked into the homogenized human adipose lipid matrix. 240 �The slopes of the calculated regression lines from the data points in each of the 14 analyses can be used as an indication of the accuracy of the analytical method for the 17 target analytes. Figure 24 is a plot of the slope of regression lines versus the 17 individual compounds. Table 16 provides a key to specific compounds associated with a number on the x-axis of this plot. The plot presents the estimated slope from each least squares regression line as well as the upper and lower 95% confidence limits for the slope. The slope of the regression line can be interpreted as a measure of accuracy with a value of 1.00 equivalent to 100% agreement of the measured concentration with the theoretical values (background plus spike level). The plot of the 95% confidence limits presents some confirmation on the precision of measurements across the four spike levels. These confidence bounds can also be used to determine whether the accuracy of the measurements (slope of regression line) is significantly different from 100% (or 1.00). If the vertical line connecting the lower and upper 95% confidence limits intersects with the horizontal line at 1, then the accuracy of the method (as determined from the regression line) is not significantly different from 100% (slope = 1.00). The results plotted in Figure 24 demonstrate that the method accuracies for 7 of the 17 analytes are not significantly different from 100%. On the other hand, if the upper and lower confidence limits are both greater than or both less than 1.00, then the accuracy of the method is significantly different from 100%. The data presented graphically in Figure 24 indicate that some positive bias (greater than 100%) is associated with the method accuracies for 9 of the 17 analytes while the measurements for a single analyte (OCDF) result in a slightly negative (less than 100%) bias. Table 16 provides a key to the compound identification in Figure 24 and tabulates the slope of the regression lines and the upper and lower 95% confidence limits for each of the 2,3,7,8-substituted PCDD and PCDF analytes. As noted from Table 16, method accuracy (as defined by the regression line slope) ranges from 90% for OCDF up to 121% for 1,2,3,7,8,9-HxCDF. The accuracies for all other measurements fall within a range of 97 to 115% with the exception of the 1,2,3,7,8,9-HxCDF (compound no. 9). The overall method accuracies meet the initial accuracy objective of 50-115% identified in the project quality assurance program plan. However, the predicted accuracy results for individual analysis as defined by the 95% upper confidence limits indicate that this range should be adjusted to 50-130%. The bias in the accuracy of the measurements may be a result of slight differences in the concentration calibration standards and the internal quantisation standard and native PCDD and PCDF spiking solutions. As a preliminary check on these differences, solutions of the low level and of the high level native spike combined with the internal quantisation standards were analyzed. The results of these analyses are provided in Table 17. Accuracy was calculated as measured/spiked x 100. The results of these analyses suggest that bias observed in overall method accuracy is attributed to the differences in the spiking solutions versus the calibration standards. For instance, the four HxCDF isomers demonstrated a consistent positive bias to method accuracy based on the least squares regression analysis. The analysis of the spiking solutions, submitted as samples, also indicates a definite positive bias for the same four HxCDF isomers. 53 �ACCURACY ESTIMATES 1.4 1.3 <D o 1.2 H V) Oi £ II 1.1 t) + n o 0) J! i 09 . OJ T—r _j j j r—r j.. -T r #01 #Q2#03#04#05#06#Q7#08#09#10#11#12#13#14#15#16# J 7 a slope estSmote Compound Number -tlower 95% CL o upp«r 35% GL Figure 24. Method accuracy estimates as determined from the slopes of the least squares regression lines for the 17 target PCDD and PCDF analytes. Refer to Table 16 for the key to compound number. 54 �Table 16. Regression Line Slopes with 95% Confidence Limits Compound no. Compound 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 2,3,7,8-TCDF 2,3,7,8-TCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,8-HxCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD OCDF OCDD Significantly Lower 95% Upper 95% confidence confidence different Slope from 1.00? limit limit 1.08 1.13 1.07 0.98 1.04 1.07 1.12 1.12 1.21 0.98 1.01 1.12 1.01 0.97 1.08 0.90 1.15 yes yes yes no no yes yes yes yes no no no no no yes yes yes 55 1.04 1.08 1.02 0.82 0.98 1.06 1.09 1.09 1.12 0.92 0.86 0.94 0.95 0.95 1.01 0.88 1.00 1.11 1.19 1.12 1.13 1,11 1.09 1.16 1.15 1.29 1.05 1.16 1.30 1.07 1.00 1.16 0.92 1.30 �Table 17. Results of the Analysis of the Low and High Level Native Spike Solutions Low level spike Compound Spi ke concentration (pg/uL) Measured concentration (pg/uL) Accuracy () % High level spike Spike Measured concentration concentration (pg/ML) (pg/ML) 50 Accuracy () % 2,3,7,8-TCDF 2,3,7,8-TCDD 13 12 130 120 50 54 57 108 114 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD en 10 10 10 10 10 11 10 12 110 100 120 50 50 50 52 49 53 104 98 106 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 25 25 25 25 25 25 25 27 32 35 40 27 24 39 108 128 140 160 108 96 156 125 125 125 125 125 125 125 134 137 152 185 123 132 148 107 110 122 148 98 106 118 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD 25 25 25 22 25 26 88 100 104 125 125 125 116 121 130 93 97 104 OCDF OCDD 50 50 44 48 88 96 250 250 247 250 99 100 �Similar trends are noted for other compounds in Table 17 compared to the data presented in Figure 24 and Table 16. The limited number of analyses of the spiking solutions does not provide an adequate comparison with the sample data to confirm the bias. However, it is recommended that at least triplicate measurements of the spiking solutions at each fortification level should be analyzed at the outset of the actual NHATS sample analysis program. This will be necessary to account for any biases that will be observed from the determination of PCDD and PCDF residue levels in spiked QC samples. It should be noted that additional homogenized spiked samples will be prepared prior to initiation of the NHATS sample analyses. 1. Recovery of Internal Quantitation Standards The absolute recoveries for the carbon-13 labeled internal quantitation standards were determined for each sample by comparing the responses to the internal quantisation standard, 13C12-1,2,3,4-TCDD. The 13 average recoveries of the compounds in Tables 10 to 15 range from 52.1% for C12-2,3,7,8TCDD up to 88.9% for 13C12-OCDD. The results for the absolute recoveries compared to the overall method accuracy for each compound indicate the importance of the internal standard quantisation technique for analysis of the PCDDs and PCDFs in human adipose. 2. Estimation of Background Levels of PCDDs and PCDFs The estimated background levels of the various PCDD and PCDF congeners were determined as the intercept obtained from the least squares linear regression analyses. Table 18 summarizes the estimated background levels along with their upper and lower 95% confidence limits. These background levels and confidence limits can be viewed as the intersections of the regression line and its upper and lower 95% confidence bounds, respectively, with the y-axis (measured or found concentration). These values will be used as the initial data points for developing control charts of the unspiked lipid matrix which will be analyzed with each batch of samples throughout the EPA/VA study. 3. Day-to-Day HRGC/MS Analysis Precision In addition to the analysis of the replicate spiked samples, four extracts were analyzed by HRGC/MS on two different dates. The results of the duplicate HRGC/MS analyses of these four samples for the 17 target compounds are presented in Tables 19 to 21. Concentration values from the second analysis date were included in the statistical analysis of data presented earlier in this section. 57 �Table 18. Background Level Estimates with 95% Confidence Limits Compound no. 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 Compound 2,3,7,8-TCDF 2,3,7,8-TCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8HpCDF 1,2,3,4,7,8,9HpCDF 1,2,3,4,6,7,8HpCDD OCDF OCDD Lower 95% Upper 95% Estimated Level background significantly confidence confidence limit level limit different from zero? (pg/g) (pg/g) (pg/g) ND (3.6)a 11.8 ND (1.1) 22.2 19.8 11.2 4.3 ND (2.3) 24.9 159.4 26.5 25.7 yes yes no yes yes yes yes yes no yes yes yes yes 2.7 10.4 -0.2 17.9 18.2 21.3 8.6 2.1 -3.6 20.5 149.3 14.3 21.9 4.5 13.3 2.5 26.4 21.5 23.5 13.7 6.5 8.2 29.3 169.5 38.7 29.6 ND (0.3) no -2.1 1.5 214.0 yes 208.9 219.1 ND (1.0) 799.7 no yes -1.7 779.4 3.8 820.0 22.4 ND = not detected, The value in parentheses reflects the estimated method detection limit. 58 �Table 19. Day-to-Day Precision of Analysis of Specific Sample Extracts for Tetra- and Pentachloro PCDF and PCDD Spike 2,3,7,8Analysis level TCDF date (pg/g) (pg/g) 4/22/86 4/28/86 0 0 RPD ( ) %b 4/22/86 4/28/86 25 0 0 RPD ( ) % 4/22/86 4/28/86 RPD ( ) % ND (3.1) ND ( . ) 41 28 0 0 ND (3.3) ND ( . ) 40 1,2,3,7,8PeCDD 1,2,3,7,8PeCDF (pg/g) (pg/g) 10 11 ND ( . 4 08) ND (1.12) 24 21 18 20 14 12 23 22 17 20 5 14 16 19 17 18 (pg/g) 4 29 ND (0.78) ND (0.75) 10 11 4 13 10 13 ND (0.81) ND (0.75) (pg/g) 19 RPD ( ) % 4/22/86 4/28/86 ND (3.0)a ND ( . ) 41 2,3,4,7,8PeCDF 2,3,7,8TCDD 10 10 31 8 15 5 12 14 21 23 11 12 26 28 27 19 14 12 13 7 36 a .ND = not detected. Value in parentheses is the estimated limit of detection. Relative percent difference. Calculated as the difference of the two values divided by the mean of the two values times 100%. 59 �Table 20. Day-to-Day Precision of Analysis of Specific Sample Extracts for Hexa- and Heptachloro PCOF and PCDO 4/22/86 42/6 /88 0 0 RPO ( ) Xb 0 0 RPO (%) 4/22/86 4/28/86 0 0 RPD (%) 42/6 /28 4/28/86 RPO (X) 25 25 20 22 16 19 28 31 10 13 20 8 NO ( . 2 03) ND ( . 4 07) 21 23 150 178 18 26 25 27 79 9 17 38 9 21 20 176 171 18 25 27 29 102 5 3 32 8 116 3 28 32 30 35 43 49 171 181 46 63 46 49 23 25 235 241 13 15 15 6 31 5 12 3 12 12 1 22 22 12 12 5 20 23 12 13 9 10 44 47 36 37 6 1 (pg/g) 1,2,3,4, 7,8,9HpCDF 149 170 21 22 (pg/g) 1.2,3, 6.7,8HxCDD 1.2,3,4, 6,7.8HpCDF 10 (pg/g) (pg/g) 4.2 4.9 15 4.0 4.2 5 3.9 4.3 1.2,3, 7,8,9HxCDF 1,2,3, 4,7,8HxCDO 1,2,3, 7,8,9HxCDD 1,2,3, 6,7,8HxCOF 1 Analysis date 4/22/86 4/28/86 2,3,4, 6,7,8HxCOF 1,2,3, 4,7,8HxCDF 6 Spike level (pg/g) NO ( . 3 a 03) ND ( . 1 05) 43 ND ( . 1 04) ND ( . 6 08) (pg/g) (pg/g) (pg/g) (pg/g) rND = not detected. The value in parentheses is the estimated limit of detection. Relative percent difference. Calculated as the difference of the two values divided by the mean of the two values times 100%. (pg/g) ND ( . 3 08) NO ( . 6 03) 79 ND ( . 9 10) ND ( . 4 11) 4 NO ( . 6 10) ND ( . 8 02) 1,2,3,4, 6,7,8HpCDO (pg/g) 210 216 3 207 211 2 223 216 �Table 21. Day-to-Day Precision of Analysis of Specific Sample Extracts for OCDF and OCDD Analysis date 4/22/86 4/28/86 OCDF concentration OCDD concentration (pg/g) (pg/g) (pg/g) 0 0 4.9 3.5 811 810 Spike level RPD ( ) %a 4/22/86 4/28/86 0 0 2.2 2.3 RPD ( ) % 4/22/86 4/28/86 0 0 1.9 2.6 2 788 784 31 1 43 44 834 848 3 50 50 RPD ( ) % a 819 836 4 RPD (%) 4/22/86 4/28/86 0 33 2 Relative percent difference. Calculated as the difference of the two values divided by the mean of the two values times 100%. 61 �VI. QUALITY ASSURANCE/QUALITY CONTROL (QA/QC) As discussed in the experimental section of this report, the QA/QC activities included the analysis of a multipoint calibration curve, daily verification of relative response factors for each analyte, analysis of a method blank and reagent blanks along with the samples, and determining the absolute recoveries of each of the internal quantisation standards for every sample. Each of these QA/QC activities is discussed below. A. Initial Calibration At the outset of sample analysis activity, six calibration concentration standards containing each of the target PCDDs and PCDFs at varying levels and constant concentrations of the internal quantisation and recovery standards were analyzed in triplicate. The relative response factors (RRF) for each native compound and internal quantitation standard were determined for each standard analysis. An average RRF and relative percent standard deviation (RSD) were determined for each concentration level. The average RRF values from each of the six concentration calibration standards were then used to calculate a grand mean RRF value for each compound in the calibration solution. Table 22 presents a summary of the grand mean RRF values for each component in the standards. As noted from Table 22, the average RRF values for native PCDDs and PCDFs generally varied by less than ± 10% (RSD) with the exception of the pentachloro congeners. These results fall well within the criteria established in the draft quality assurance program plan which required the variability of RRF values for the tetrachloro homologs to be within ± 20% (RSD) while the RRF criterion on all other compounds was set at ± 30% (RDS). The variability of the RRF values for the internal quantisation standards, on the other hand, was noted to increase with the degree of chlorination. This is a result of the measurement of all internal quantisation standards versus the single internal 37 recovery standard, 13C12-1,2,3,4-TCDD. A second internal recovery standard, Cl4-l,2,3,4,6,7,8-HpCDD, was evaluated. However, problems resulting from contribution of native HpCDD to the characteristic ions of this internal standard resulted in variabilities in the RRF value up to 50%. Hence, this internal standard was not used for any calculations. It is anticipated that an additional internal recovery standard, such as 13C12-l,2,3,4,7,8-HxCDD, will improve the variability in the RRF values of the higher chlorinated internal quantisation standards. This compound will be incorporated into the method if available. The sensitivity of the Kratos MS-50TC to the tetra- through octachloro PCDDs and PCDFs was demonstrated through the triplicate analysis of the low level standard (CS-8, Table 2) that ranged in concentration from 1 pg/uL for the tetra- and pentachloro congeners up to 5 pg/uL for the octachloro congeners. Table 22 provides an indication of the observed signal-tonoise ratio for each of the native PCDD and PCDF congeners. These data demonstrate that the low level standard is well above the instrument detection limit, which is defined as the amount of a particular compound necessary to give a signal 2.5 times the background signal to noise for each of the characteristic ions while meeting the qualitative criteria for ion ratios. 62 �Table 22. Relative Response Factors (Grand Means) Determined from Multipoint Concentration Calibration Standards Compound RRFa 1.00 0.80 0.98 1.06 1.33 0.94 0.93 0.86 0.86 1.31 1.44 1.61 2.33 1.89 1.19 1.38 1.04 5.7 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,6,7,8-HpCDD 1,2,3,4,7,8,9-HpCDD 1,2,3,4,6, 7, 8-HpCDD OCDF OCDD 13 C12-l,2,3,4-TCDDb'c 13 C -2,3,7,8-TCDF 13 12 1.00 1.98 C -2,3,7,8-TCDD 1.73 13 12 1.36 C12-l,2,3,7,8-PeCDF 13 C12-l,2,3,7,8-PeCDD 0.70 13 C12-l,2,3,4,7,8-HxCDF 1.28 13 C -l,2,3,6,7,8-HxCDD 0.41 13 12 C12-1,2,3,4,6,7,80.33 HpCDD 37 C14-1,2,3,4,6,7,80.12 HpCDD 13 C12-OCDD 0.24 6.2 5.2 10.1 11.3 3.1 2.4 2.7 6.9 4.9 3.0 1.0 4.0 3.6 4.7 3.3 2.5 _ Calibration range (pg/MD 12 6.5 11 8.9 5.7 32 30 29 17 13 14 14 35 26 13 31 21 _ RSD ( ) % 2,3,7,8-TCDF 2,3,7,8-TCDD Signal-to- noise ratio for low. level standard 1-100 1-100 1-100 1-100 1-100 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 2.5-250 5-500 5-500 - 50 50 50 50 50 125 125 125 51.8 - 125 28.0 ™ 250 7.6 4.7 4.5 7.7 15.8 19.6 25.8 . - JIJRRF = grand mean RRF. characteristic ion for each native PCDD or PCDF congener (Data File 8501D17X02). Internal recovery standard. 63 �B. Daily Verification of Response Factors Before proceeding with analysis of samples, the analyst was required to verify the existing response factor calibration through the analysis of a calibration standard (CS-7, Table 2). Criteria for proceeding with sample analysis required that the measured RRF value for 2,3,7,8-TCDD and 2,3,7,8-TCDF were within ± 20% (and all other congeners within ± 30%) of the mean RRF established from the calibration curve. This standard was also analyzed at the end of each working day to demonstrate that the calibration had been maintained. All RRF values were tabulated to generate RRF control charts for each specific PCDD and PCDF congener. Figures 25 through 34 are plots (control charts) of the RRF values established for the 17 individual target analytes. The RRF data are plotted versus time of analysis. These plots contain 28 individual data points, 18 of which were generated for triplicate analysis of 6 concentration calibration solutions from initial calibration and 10 analyses of solution CS-7 (Table 2) injected over the 5 days for which actual samples were analyzed. The upper and lower boundaries (dashed lines) represent a relative standard deviation of approximately ± 10% with the exception of the plot for 1,2,3,7,8-PeCDD, for which the boundaries are plotted as ± 20%. It should be noted that the actual control limits as specified in the project QAPP were set at ± 20% for 2,3,7,8-TCDD and 2,3,7,8-TCDF and ± 30% for all other target analytes. The average RRF values and corresponding standard deviations reported in each of these plots are calculated from the total 28 standard analyses. C. Blanks As specified in the quality assurance program plan, a laboratory method blank was prepared along with the 14 human adipose lipid samples. The method blank was taken through all procedures as if it were an actual sample, although no lipid matrix was introduced. The analysis of the method blank resulted in the data reported for each of the target analytes reported in Table 23. As noted in Table 23, 1,2,3,4,6,7,8-HpCDD and OCDD were detected at concentrations equivalent to 4.0 and 30 pg/g (equivalent to a 10-g lipid sample), respectively. In addition to these compounds, responses that correspond to the elution of two TCDD isomers (1,3,6,8- and 1,3,7,9-) and a PeCDD (isomer not determined) were detected in the method blank. Further analysis of individual reagents used for preparation of the samples identified the activated acidic alumina as the source of the artifacts. Acidic alumina that had been cleaned by Soxhlet extraction but not activated at 190°C was analyzed, and the artifacts were not detected. This indicates that the artifacts are generated during activation of acidic alumina at elevated temperatures (190°C). Similar background problems from the same PCDD 18 congeners have recently been reported by the Center for Disease Control.17' 64 �(flRBVREF.flKeW/CflMT./REF.flHT.) 1.208 9.337) ¥ i ST.DEV.9.865 Z ST.OEU.« S.5S1 X 1.189 Jl • . y T if * i • DATE " n ii ii y 1 x 3.390 t i X" i TI •If if 'j v _ i «! X _ S/ 4/23/85 4/13/SS 9.897) ST.OPJ.8.035 2 ST.OEU.5.932 X 1 9.909 . 1 v • ? , t | <! J 1 tl 1 Jl ^X 11 11 n^ 1 1 .** \ J ^ V 1 1J II H 1 X X 9.789 DftTE 5/3/86 Figure 25. Control charts showing response factors by date for 2,3,7,8-TCDF and 2,3,7,8-TCDD. Dotted lines represent approximately ± 10% of the mean. 65 �F<C13-l/2(3.7/ MKEft/KEF.ARBWAANT./REF.flHT.) 1.268 1.188 _____ 9.971) T [ x f r _ j^_ L. Tr 1 ( ST.OEU.' 9.835 7. ? ST.OEU.8.739 _ _ \(. -T n • n /£ n v( ii w * * | 1 '*lC 9.966 _ w 1 * \L! 7W ^ y X* -f Tl x r 1 w A *i ~~Tx X 8.866 x 9.709 DATE 4/13/86 4/23/86 v'fiKEft/REF.flfi£ft)/<fW.--REF.ft«T.) (AUt 1.469 1.353) ST.DEU.» 9.113 r. ST.OEU. 18.691 X 1.366 1 1 1 1 1.290 i i " a n S V MM 1.668 i^ x 8.369 *r ^ " 1111 >1 -uA x ' w i I X /\ ^f x X 8.789 DATE 3/ 3/86 4/13/86 4/23^5 5/ 3/86 Figure 26. Control charts showing response factors by date for 1,2,3,7,8-PeCDF and 2,3,4,7,8-PeCDF. Dotted lines represent approximately ± 10% of the mean. 66 �.AHT.) (flUi 1.375) 3.000 ST.OEU.0.233 Z ST.DEU.' 18.8135 ¥ i i 2.889 i i i x X DATE 4/13/86 4/23/86 5/ 3/85 Figure 27. Control chart showing response factors by date for 1,2,3,7,8-PeCDD. Dotted lines represent approximately ± 20% of the mean. 67 �<f*£ft/R£F.f*BV/««T./REF.ftl1T.) (flVi 9.329) 1.188 ST.DEU.» 9.046 2 ST.OEU.» 4.899 it ii 1.999 ft II ii II 4» y» 11 ' (ft jj AL A 0.309 v v 1 A^ y T v v A1 i t ' '' ! ^v A 1 X v ^ -^ 9.390 4/ 13/86 MTE 4/23/86 (flREft/REF.fWEft>/<ftWT./ReF.fl«T.) «Wi 1.168 -| S/ 3/85 8.923) ST.OEU.0.S48 ?. ST.OEU.' 4.348 • 1.808 ' x 1 ¥ * k J ! v 1. , K v X 1. f 0.388 If y ! j! X X x x 9.399 DATE 4/13/86 4/23/86 V 3/86 Figure 28. Control charts showing response factors by date for lj2,3,4,7,8-HxCDF and 1,2,3,6,7,8-HxCDF. Dotted lines represent approximately ± 10% of the mean. 68 �((WEfl/R£F.ftR£rt)/(AnT./REF.ArtT.) (flUi 0.862) l.VW ST.DEU." 0.038 7. ST. OP.'.4.429 i i i 0.300 ? 1 % $ Jf i* ii !* ' 9.899 x i 0.760 VI 3/86 DOTE 4/23/86 5/3/86 « 3 M U <ftR£fl/REF.AREA>/«OT./REF.ftMT.) <(Wl 1.299 8.883) ST.OE«.« 9.982 Z ST.OEV.- x 9 327 - 1.199 ¥ 1.999 ¥ * 0.308 ! H ! ' ' r w ' JJ « e.309 X X 1 i X ¥ ' ! x x 1 x X a. ran 4/23/86 3/ 3/86 Figure 29. Control charts showing response factors by date for 2,3,4,6,7,8-HxCDF and 1,2,3,7,8,9-HxCDF. Dotted lines represent approximately ± 10% of the mean. 69 �PiCl3-l,2;376,7,^flBSiCHLOR(M)IOXIH <ARBVR£F.flRe»/<flNT./REF.««r.) ((Wt 1.589 1.394) ST.OEU.« 9.977 Z ST.DEV.3.310 1.4TO x T 1.399 1 i ' X w A 1.299 X 1.199 DATE 4/13/8S V 3/BS 4/23/85 ./REF.AHT.) <W| 1.433) 1.796 ST.DEU.» 9.936 Z ST.OEV.' 6.015 1.680 -Ti i HI ni in HI in in 1.569 ,. x 1.409 x « i X i 1.399 ,_*.- x 1.299 DATE 4/13/86 4/23/86 S/ 3/86 Figure 30. Control charts showing response factors by date for 1,2,3,4,7,8-HxCDD and 1,2,3,6,7,8-HxCDD. Dotted lines represent approximately ± 10% of the mean. 70 �1.637) 2.399 s h 2.299 e ST.DEU.0.135 Z ? ST.OEU.9.384 2.199 < ) 2.999 1.899 X 1 1 1.799 1 1 V& sx ^ i 1.S99 X i v "T i X 1.48B WTE V13/8S 4/23/8S Figure 31. Control chart showing response factors by date for 1,2,3,7,8,9-HxCDD. Dotted lines represent approximately ± 10% of the mean. 71 �mmm <ftREft/REF.AR£ft>/< AHT./R£F.fl«T.) (flUl 2.660 2.228) « * J i 2.500 ST.DEV." 0.182 2 ST.OEU.8.163 . k * xi 2.460 H a 2.309 *x • * » i * 2.260 ! x x ;; A i •! x X 2.160 " W A x 1.309 X x OrtTE 4/13/85 4/23/86 S/ 3/86 S:-fa§i^I;3ji^aiBWKirsra««.(fKEA/f£F.AREA>/( WT./REF.filfT.) (flU« 2.160 1.783) X i ,1 $ 2.669 ii ST.OEU." 0.169 Z ST.OE1^.9.426 i 1.360 1! 1.3W !i !j ! X' 1.760 x 1.660 j ii M *J ! • X x 1.500 f x fb «•«••» X ;: X I.46A DATE V13^S 4/23/SS Figure 32. Control charts showing response factors by date for 1,2,3,4,6,7,8-HpCDF and 1,2,3,4,7,8,9-HpCDF. Dotted lines represent approximately ± 10% of the mean. 72 �«*Eft/R£F.ftRB»/<AMT./REF.<*T.) (AUi 1.460 1.175) ST.OEU.8.071 2 ST.OEU.' S.043 ¥ 1.369 _ *• . . _ . » .. — — j *» _ M * W M _ t a «...-,~._VyJ__-.« iI • •H* K* M x Y PS A • ' '!• A « if I * X 1.106 -M • , ' ' X i x ii ' lv 'i y t • * « « . • '1 i! ' X| i x .'!-•««,«,. • . X e.see DATE 4/13/86 4/23/8S 5/ 3/8S Figure 33. Control chart showing response factors by date for 1,2,3,4,6,7,8-HpCDD. Dotted lines represent approximately + 10% of the mean. 73 �0IOXIN <«R»/R£F.f*EA>/<«1T./P.EF.fl«T.> (AUi 1.238) 1.500 X ' 5fi it V i i i i i 1.480 w III ST.DEV.9.117 * ST.DEW.» 8.378 Jt I III t «. - l«Zvv M • « M M - i 'I { X ~ i : X x 4 ^.j.i_ _ ___________ X 4/23/86 4/ 3/86 DATE w AJy X 1.189 5/3/86 ffiS»K»?«M 1.033) (AREft/KEF.rtRBWAAHT.v'REF.AMT.) <AUi 1.283 -i ST.OEU.3.843 Z ST.OEU.. 4.181 1.183 i i X X X m m J\ 1.309 , i * ^ 1 x ' X Xx —1-»— • ^ X X X i i i X 8.300 4/23/86 5/ 3/86 Figure 34. Control charts showing response factors by date for OCDF and OCDD. Dotted lines represent approximately ± 10% of the mean. 74 �Table 23. Summary of Results from the Analysis of a Laboratory Method Blank Concentration3 (pg/g) Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDD 1,2,3,4,7,8-HxCOF 1,2,3,6,7,8-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD NO ND ND ND ND ND ND ND ND ND ND ND ND ND (0.50) (2.2) (0.5) (0.5) (1.2) (0.5) (0.5) (0.5) (0.5) (1.0) (0.9) (0.8) (0.5) (0.5) 4.0 ND (0.5) 30 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF 1,2,3,4,6,7,8-HpCDD OCDF OCDD 75 �An experiment was designed to evaluate a procedure for cleaning the activated acidic alumina immediately prior to the fractionation of the sample extract. The acidic alumina (6.0 g) was packed in hexane. The packed column was eluted with 40 ml of methylene chloride/hexane (1:1) solution followed by 80 to 100 ml of hexane. The sample extract was added to the column and was eluted with 20 ml of hexane followed by 30 ml of 20% methylene chloride in hexane which was reserved for PCDD and PCDF analysis. The carbon-14 radiolabeled 2,3,7,8-TCDD, 1,2,3,6,7,8-HxCDD, and OCDD were used to evaluate recovery of PCDDs from the cleaned alumina. Recoveries of the radiolabeled PCDDs from the activated acidic alumina precleaned by the procedure described above are detailed in Table 24. These data demonstrate that the selected PCDDs are quantitatively (greater than 90%) recovered from the precleaned acidic alumina. This procedure for cleanup of activated acidic alumina has been integrated into the analytical protocol (Appendix A) for routine application with sample preparation activities. D. Absolute Recoveries of the Internal Quantisation Standards The absolute recoveries of the carbon-13 labeled internal quantitation standards 13were determined by comparing responses with the internal recovery standard, C12-1,2,3,4-TCDD, which was added during final concentration prior to HRGC/MS analysis. A summary of the average and range of recoveries of the 8 internal quantisation standards from the 14 human adipose lipid samples is provided in Table 25. These data indicate that recoveries ranged from an average of 52.1% for 13C12-2,3,7,8-TCDD up to 88.9% for 13C12-OCDD. The average recoveries for the lower chlorinated internal standards were lower than the preliminary method studies with carbon-14 radiolabeled standards had indicated. This resulted in a closer evaluation of the final concentration step prior to mass spectrometry. The first extracts for the human adipose lipid extracts were concentrated with a nitrogen evaporation system equipped with a water bath at approximately 55°C. Final blowdown of the samples required addition of the internal recovery standard in 10 uL of tridecane as a keeper solution. However, it was noted at the elevated temperature final volumes from nitrogen evaporation were generally on the order of 2 to 5 uL. This required addition of another 10 uL of tridecane prior to HRGC/MS analysis. In an effort to assess the effect of reducing the final volume of tridecane at elevated temperatures on absolute recoveries of the internal quantisation standards, an experiment using the radiolabeled TCDD, HxCDD, and OCDD standards was conducted. Four solutions of the same spike level were prepared with each radiolabeled compound in 1 ml of toluene. Two of the spiked solutions were heated at 55-60°C and the solvent was reduced under a gentle stream of prepurified nitrogen. The toluene solution was concentrated to 100 uL, 500 uL of 1% toluene in methylene chloride was added, and the solution was concentrated to 200 uL. At this time 10 pL of the keeper tridecane was added and the solution was allowed to concentrate further. The remaining two solutions for each radiolabeled compound were taken through a similar solvent exchange 76 �Table 24. Recovery of Radiolabeled PCDDs from Precleaned Activated Alumina Spike level (pg) Compound Recovery 14 100 300 300 92 96 97 14 1,000 3,000 3,000 103 101 100 14 2,500 7,500 7,500 102 99 97 C-2,3,7,8-TCDD C-l,2,3,4,7,8-HxCDD C-OCDD 77 �Table 25. Absolute Recoveries of the Internal Quantitation Standards from the Human Adipose Lipid Matrix Internal quantisation standard Average recovery ( ) % Standard deviation Relative standard deviation (%) 13 64.4 7.9 12.3 46-78 13 52.1 5.0 9.6 43-62 13 76.1 8.9 11.7 62-90 13 57.1 3.5 6.1 54-64 13 59.4 5.0 8.4 54-70 13 63.6 5.3 8.4 57-77 76.3 10.3 13.6 61-99 88.9 11.4 12.9 67-104 C12-2,3,7,8-TCDF C12-2,3,7,8-TCDD C12-l,2,3,7,8-PeCDF C12-l,2,3,7,8-PeCDD C12-1,2,3,4,7,8HxCDF C12-1,2,3,6,7,8HxCDD 13 c12-i,2,3,4,6,7,8- Range of recovery (%) HpCDD 13 C12-OCDD Values based on 14 analyses of human adipose lipid samples. 78 �and concentration procedure except the solution was allowed to concentrate at room temperature. One of the most obvious results was the observation that solutions held at elevated temperatures could be reduced to dryness even when tridecane had been added as a keeper. On the other hand, solutions for which tridecane had been added but remained at room temperature could only be concentrated to a 10-uL final volume. The recoveries of the radiolabeled standards from each of the solutions in this study are presented in Table 26. The results from this study indicate that the final concentration condition may have a pronounced effect on the absolute recoveries of the PCDDs and PCDFs, especially for the lower chlorinated congeners such as 2,3,7,8-TCDD. However, it should be noted that the approach to target analyte quantisation based on the internal standard method (isotope dilution for 8 of the 17 target analytes) is not affected by absolute recoveries as low as 50%. The procedure for final concentration in the analytical protocol (Appendix A) for the analysis of the NHATS samples for the EPA/VA study has been modified to specify room temperature conditions. 79 �Table 26. Recovery of Carbon-14 Labeled 2,3,7,8-TCDD, 1,2,3,4,7,8-HxCDD, and OCDD as a Function of Final Concentration Conditions Spike level Compound (pg) Concentration conditions Observed final volume Observed recovery ( ) % 14 300 300 300 300 55-60°C 55-60°C 20°C 20°C 1-2 uL dry ness 10 |jL 10 ML 78 54 98 93 14 3,000 3,000 3,000 3,000 55-60°C 55-60°C 20°C 20°C 1-2 ML 94 102 105 7,500 7,500 7,500 7,500 55-60°C 55-60°C 20°C 20°C 1-2 2-3 10 10 C-2,3,7,8-TCDD C-l,2,3,4,7,8-HxCDD 14 C-OCDD 5 ML 10 ML 10 ML ML ML ML ML 107 94 94 100 97 Each solution was concentrated under a gentle stream of flowing nitrogen. 80 �VII. REFERENCES 1. Stanley JS. 1984. Methods of analysis of polychlorinated dibenzo-pdioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in biological matrices—literature review and recommendations. EPA-560/584-00. 2. Stanley JS, Going JE, Redford DP, Kutz KW, Young AL. 1985. A survey of analytical methods for measurement of polychlorinated dibenzo-p_-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) in human adipose tissues. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 181-195. 3. Stanley JS. 1984 (March 28). Proposed analytical method for analysis of PCDDs/PCDFs in human adipose tissue: special report. Washington, DC: Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Contract 68-02-3938, Work Assignment 8. 4. Stanley JS. 1986 (April 23). Broad scan analysis of human adipose tissue: polychlorinated dibenzo-jD-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). Draft final report. Washington, DC: Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency. Contract 68-023938, Work Assignment 8. 5. Albro et al. 1985. Methods for the quantitative determination of multiple, specific polychlorinated dibenzo-p_-dioxin and dibenzofuran isomers in human adipose tissue in the parts-per-trillion range. An interlaboratory study. Anal Chem 57: 2717-2725. 6. Patterson DG et al. 1986. High-resolution gas chromatographic/highresolution mass spectrometric analysis of human adipose tissue for 2,3,7,8-tetrachlorodibenzo-p_-dioxin. Anal Chem 58: 705-713. 7. Schecter A, Tiernan TO, Taylor ML, VanNess GF, Barrett JH, Wagel DJ, Gitlitz G, Bogdasarian M. 1985. Biological markers after exposure to polychlorinated dibenzo-p_-dioxins, dibenzofurans, biphenyls, and biphenylenes. Part I: Findings using fat biopsies to estimate exposure. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 215-246. 8. Schecter A, Ryan JJ. 1985. Dioxin and furan levels in human adipose tissue from exposed and control populations. 189th National ACS Meeting Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment III, Miami, Florida. Preprint Division of Environmental Chemistry, ACS 25:160-163, Paper No. 56. 9. Ryan JJ, Williams DT, Lau BPY, Sakuma T. 1985. Analysis of human fat tissue for 2,3,7,8-tetrachlorodibenzo-jD-dioxin and chlorinated dibenzofuran residues. In: Chlorinated dioxins and dibenzofurans in the total environment II. Keith LH, Rappe C, Choudhary G, eds. Butterworth Publishers, pp. 205-214. 81 �10. Ryan JJ, Schecter A, Lizotte R, Sun W-F, Miller L. 1985. Tissue distribution of dioxins and furans in humans from the general population. Chemosphere 14: 929-932. 11. Nygren M, Hansson M, Rappe C, Domellof L, Hardell L. 1985. Analysis of polychlorinated dibenzo-p_-dioxins and dibenzofurans in adipose tissue from soft-tissue sarcoma patients and controls. 189th National ACS Meeting Symposium on Chlorinated Dioxins and Dibenzofurans in the Total Environment III, Miami, Florida, 1985. Preprint Division of Environmental Chemistry, ACS 25:160-163, Paper No. 55. 12. Smith LM, Stalling DL, Johnson JJ. 1984. Determination of part per trillion levels of polychlorinated dibenzofurans and dioxins in environmental samples. Anal Chem 58: 1830-1842. 13. Rappe C, Nygren M, Linstrom G, Hanson H. 1985. Dioxins and dibenzofurans in human tissues and milk of European origin. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 14. Ryan JJ. 1985. Variation of dioxins and furans in humans with age and organ by country. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 15. Graham M, Hileman FD, Wendling J, Wilson JD. 1985. Chlorocarbons in adipose tissue samples. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 16. Patterson DG, Holler JS, Smith SJ, Liddle JA, Sampson EJ, Needham LL 1985. Human tissue data in certain U.S. populations. 5th International Symposium on Chlorinated Dioxins and Related Compounds, Bayreuth, FRG, September 16-19, 1985. 17. Patterson DG, Holler JS, Groce DF, Alexander LR, Lapeza CR, 0'Conner RC, Liddle JA. 1986. Control of interferences in the analysis of human adipose tissue to 2,3,7,8-tetrachlorodibenzo-p_-dioxin (TCDD). Environ Toxicol Chem 5: 355-360. 18. Holler JS, Patterson DG, Alexander LR, Groce OF, O'Connor RC, Lapeza CR. 1985. Control of artifacts and contamination in the development of a dioxin analytical program. 33rd Annual Conference on Mass Spectrometry and Allied Topics, San Diego, CA, May 26-31, 1985. 82 �APPENDIX A ANALYTICAL PROTOCOL FOR DETERMINATION OF PCDDs AND PCDFs IN HUMAN ADIPOSE TISSUE A-l �TABLE OF CONTENTS Section Description Page 1 Scope and Application A-3 2 Summary of Method A-3 3 Definitions 4 Interferences A-7 5 Safety A-7 6 Apparatus and Equipment A-8 7 Reagents and Standard Solutions A-ll 8 High Resolution Gas Chromatography/Mass Spectrometry Performance Criteria. A-13 Quality Control Procedures A-30 10 Sample Preservation and Handling A-32 11 Sample Extraction A-33 12 Cleanup Procedures A-35 13 Analytical Procedures A-37 14 Date Reduction A-42 15 Reporting and Documentation A-47 9 . A-2 A-6 �ANALYTICAL PROTOCOL FOR DETERMINATION OF PCDDs AND PCDFs IN HUMAN ADIPOSE TISSUE 1. SCOPE AND APPLICATION 1.1 1.2 The minimum measurable concentration is estimated to range from 1 pg/g (1 part per trillion) for 2,3,7,8-TCDD and 2,3,7,8-TCDF up to 5 pg/g for OCDD and OCDF. However, these detection limits depend on the kinds and concentrations of interfering compounds in the sample matrix and the absolute method recovery. 1.3 2. This method provides procedures for the detection and quantitative measurement of polychlorinated dibenzo-p_-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) at concentrations ranging from 1 to 100 pg/g for the tetrachloro congeners up to 5 to 500 pg/g for the octachloro congeners in 10-g aliquots of human adipose tissue. The method will be used to determine PCDDs and PCDFs, particularly congeners with chlorine substitution in the 2,3,7,8 positions. Table 1 lists the specific PCDDs and PCDFs and target method detection limits. SUMMARY OF METHOD Figure 1 presents a schematic of the analtyical procedures for determination of PCDDs and PCDFs in human adipose tissue. The analytical method requires extraction and isolation of lipid materials from human adipose samples. This is accomplished using sample sizes ranging up to 10 g. Extraction and homogenization are accomplished using methylene chloride and a Tekmar Tissuemizer®. The extract is filtered through anhydrous sodium sulfate to remove water. The extraction procedure is repeated (three to five times) until the tissue sample has been thoroughly homogenized. The final extract is adjusted to a known volume (100 ml) and the extractable lipid is determined using a minimum of 1% of the final volume. The methylene chloride in the remaining extract is concentrated until only an oily residue remains. The residue is spiked with known amounts of the carbon-13 labeled PCDDs and PCDFs (e.g., 500 pg of 13C12-TCDD/F to 2,500 pg of 13C12-OCDD/F) as internal quantisation standards. The residue is diluted with hexane ( 200 ml), and 100 g of ^ sulfuric acid modified silica gel (40% w/w) is added to the solution with stirring. The mixture is stirred for approximately 2 h, and the supernatant is decanted and filtered through anhydrous sodium sulfate. The adsorbent is washed with at least two additional aliquots of hexane. The combined hexane extracts are eluted through a column consisting of a layer of sulfuric acid modified silica gel, and a layer of unmodified silica gel. The eluate is concentrated to approximately 1 ml and added to a column of acidic alumina. The PCDDs and PCDFs are eluted from the alumina using 20% methylene chloride/hexane. This eluate is concentrated to approximately 0.5 ml and is added to a 500-mg Carbopak C/Celite column. The PCDDs and PCDFs are eluted from the column using 20 ml of toluene. A-3 �Table 1. Target PCDD and PCDF Congeners and Target Method Detection Limits Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3, 7,8- PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4, 7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1, 2,3,7,8, 9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF CAS no. Target method detection limit (pg/g)D 1746-01-6 51207-31-9 40321-76-4 57117-41-6 57117-31-4 39227-28-6 57653-85-7 19408-74-3 70648-29-9 57117-44-9 72918-21-9 60851-34-5 35822-46-9 67562-39-4 55673-89-7 3268-87-9 39001-02-0 1.0 1.0 1.0 1.0 1.0 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 5.0 5.0 Chemical Abstract Services number. = parts per trillion. A-4 �Initial Sample Preparation Isolation of Extractable U'pid Materials Homogenization in Methylene Chloride I Llpid Determination Solvent Exchange Add Internal Quantitation Standards (13C-PCDDs/PCDFs) Bulk Lipid Removal Acid Modified Silica Gel Slurry Technique Provides Cleanup of Oxidizable Compounds with Rapid Sample Turnaround, Improved Cleanup Efficiency and Recovery Removal of Chemical InterferencesAcidic Silica/Silica Acidic Alumina Provides Seperation of PCBs and Other Potential Interferences from PCDDs and PCDFs. I Carbopak C/Cel 7 te Selective Adsorption and Isolation of PCDDs/PCDFs Add Internal Recovery Standards HRGC/MS-S1M Analysis I 1 LRMS I de nti fi cat i on/Quant? tati on of Terra-6'cta PCDDs/PCDFs HRMS Confirmation of 2,3,7,8-TCDD Figure 1. Schematic of the sample preparation and instrumental analysis procedures for determination of PCDDs and PCDFs in human adipose tissue. A-5 �The toluene is concentrated to less than 1 mL and transferred to conical vials. Tridecane (10 jjL) containing 500 pg of an internal recovery standard is added as a keeper, and the extract is concentrated to a final volume. The HRGC/MS analysis is completed in the selected ion monitoring mode (SIM). Analysis of the tetra- through octachloro PCDD and PCDF congeners is achieved using low resolution mass spectrometry. Separation of the tetra- through octachloro PGDD and PCDF congeners is achieved using a 60-m DB-5 column. Verification of the 2,3,7,8-TCDD is achieved using a 50-m CP Sil 88 column and HRGC/MS-SIM analysis in the high resolution mode (R = 10,000). 3. DEFINITIONS 3.1 Concentration calibration solutions ~ Solutions containing known amounts of the native analytes (unlabeled 2,3,7,8-substituted PCDDs and PCDFs), the internal quantisation standards13(Carbon-13 labeled PCDDs and PCDFs), and the recovery standard, C121,2,3,4-TCDD. These calibration solutions are used to determine instrument response of the analytes relative to the internal quantisation standards and of the internal quantisation standards relative to the internal recovery standard. 3.2 Internal quantitation standards — Carbon-13 labeled PCDDs and PCDFs, which are added to every sample and are present at the same concentration in every method blank and quality control sample. These are added to the lipid residue extracted from the adipose tissue and are used to measure the concentration of each analyte. The concentration of each internal quantitation standard is measured in every sample, and percent recovery is determined using the internal recovery standard. 3.3 Internal recovery standard -- 13C12-1,2,3,4-TCDD which is added to every sample extract just before the final concentration step and HRGC/MS-SIM analysis. 3.4 Laboratory method blank — This blank is prepared in the laboratory through performing all analytical procedures except addition of a sample aliquot to the extraction vessel. A minimum of one laboratory method blank will be analyzed with each batch of samples. 3.5 HRGC column performance check mixture ~ A mixture containing known amounts of selected TCDD standards; it is used to demonstrate continued acceptable performance of the capillary column, i.e., separation ( 25% valley on a 50-m CP Sil 88 or 60-m SP-2330 ^ HRGC column) of 2,3,7,8-TCDD isomer from all other 21 TCDD isomers, and to define the TCDD retention time window. 3.6 Relative response factor -- Response of the mass spectrometer to a known amount of an analyte relative to a known amount of an internal standard (quantitation or recovery). A-6 �3.7 3.8 4. Mass resolution check — Standard method used to demonstrate static resolution of 10,000 minimum (10% valley definition). Sample batch -- A sample batch consists of up to 10 human adipose tissue samples, one method blank, 2 internal quality control (QC) samples (spiked and unspiked), and an external performance audit sample (blind spike). INTERFERENCES Chemicals which elute from the HRGC column with ± 10 scans of the internal and/or recovery standards and which produce within the retention time window ions at any of the masses used to detect or quantify PCDDs, PCDFs, or the internal quantisation and recovery standards are potential interferences. Most frequently encountered potential interferences are other sample components that are extracted along with the PCDDs and PCDFs, e.g., PCBs, chlorinated methoxybiphenyls, chlorinated hydroxydiphenyl ethers, chlorinated benzylphenyl ethers, chlorinated naphthalenes, DDE, DDT, etc. The actual incidence of interference by these chemicals depends also upon relative concentrations, mass spectrometric resolution, and chromatographic conditions. Because very low levels (pg/g) of PCDDs and PCDFs are anticipated, the elimination of interferences is essential. High purity reagents and solvents must be used and all equipment must be scrupulously cleaned. Laboratory method blanks must be analyzed to demonstrate absence of contamination that would interfere with measurement of the PCDDs and PCDFs. Column chromatographic procedures are used to remove coextracted sample components; these procedures must be performed carefully to minimize loss of PCDDs and PCDFs during attempts to increase their concentration relative to other sample components. 5. SAFETY 5.1 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. The 2,3,7,8-TCDD is a known teratogen, mutagen, and carcinogen. Ingestion of microgram quantities can result in toxic effects. The other 2,3,7,8-substituted PCDDs and PCDFs may exhibit teratogenic, mutagenic, and carcinogenic effects. From this viewpoint, exposure to these chemicals must be reduced to the lowest possible level by whatever means available. Only experienced personnel will be allowed to work with these chemicals. 5.2 All laboratory personnel will be required to wear laboratory coats or coveralls, gloves, and safety glasses. The neat standards, stock, and working solutions will be handled only in a Class A fume hood or glove box. When manipulating stock standards or working solutions, the analyst is advised to place the solution vials in a secure holder (sample block or glass beaker) to prevent accidental spills. A-7 �5.3 5.4 If handling of these compounds result's in skin contact, immediately remove all contaminated clothing and wash the affected skin areas with soap and water for at least 15 min. 5.5 6. If these standards are spilled, absorb as much as possible with absorbent paper and place in a container clearly labeled as PCDD or PCDF waste. Solvent-wash all contaminated surfaces with toluene and absorbent paper followed by washing with a strong soap and water solution. Dispose of all contaminated materials in sealed steel containers labeled as contaminated with PCDD and/or PCDF residue and indicate the approximate level of contamination. As a final precaution, prepare a wipe sample of the exposed surface area and include the wipe as part of the sample analysis batch. This will be used to confirm that the work area is free of contamination. Disposal of laboratory wastes — All laboratory wastes (solvents and absorbents) will be disposed of as hazardous wastes. The laboratory personnel should take care to dispose of the sodium sulfate, silica gel, and alumina in separate metal containers. Excess solvents should be disposed of in gallon polyethylene jugs containing a layer of activated charcoal. Excess solvent that is known to be contaminated with PCDDs or PCDFs should be kept at a minimum by evaporating the solvent with a stream of air. APPARATUS AND EQUIPMENT 6.1 High Resolution Gas Chromatograph/Mass Spectrometer/Data System (HRGC/HRMS/DS) 6.1.1 The GC must be equipped for temperature programming, and all required accessories must be available, such as syringes, gases, and a capillary column. The GC injection port must be designed for capillary columns. The use of splitless injection techniques is recommended. When using this method, a l-pL injection volume is used. The injection volumes for all extracts, blanks, calibration solutions, and the performance check sample must be consistent. 6.1.2 High Resolution Gas Chromatograph-Mass Spectrometer Interface The HRGC/MS interface is directly coupled to the mass spectrometer ion source. All components of the interface should be glass or glass-lined stainless steel. The interface components should be compatible with 300°C temperatures. The HRGC/MS interface must be appropriately designed so that the separation of the PCDDs and PCDFs which is achieved in the gas chromatographic column is not appreciably degraded. Cold spots and/or active surfaces (adsorption sites) in the HRGC/MS A-8 �interface can cause peak tailing and peak broadening. It is recommended that the HRGC column be fitted directly into the MS ion source. Graphite ferrules should be avoided in the HRGC injection port since they may absorb PCDDs or PCDFs. Vespel or equivalent ferrules are recommended. 6.1.3 Mass Spectrometer The mass spectrometer must be capable of maintaining a minimum resolution of 10,000 (10% valley) for high resolution confirmation analysis. The mass spectrometer must be operated in a selected ion monitoring (SIM) mode with total cycle time (including voltage reset time) of 1 s or less. 6.1.4 Data System A dedicated hardware or data system is required to control the rapid multiple ion monitoring process and to acquire the data. Quantification data (peak areas or peak heights) and SIM traces (displays of intensities of each m/z (characteristic ion) being monitored as a function of time) must be acquired during the analyses. Quantifications may be reported based upon computergenerated peak areas or upon measured peak heights. 6.2 HRGC Columns For isomer-specific determinations of 2,3,7,8-TCDD, the following fused silica capillary columns are recommended: a 50-m CP-Sil 88 column and a 60-m SP-2330 (SP-2331) column. However, any capillary column which separates 2,3,7,8-TCDD from all other TCDDs may be used for such analyses, provided that the minimum acceptance criteria in Section 8 are met. 6.3 Miscellaneous Equipment 6.3.1 Nitrogen evaporation apparatus with variable flow rate. 6.3.2 Balance capable of accurately weighing to ± 0.01 g. 6.3.3 Water bath — equipped with concentric ring cover and capable of being temperature-controlled. 6.3.4 Stainless steel spatulas or spoons. 6.3.5 Magnetic stirrers and stir bars. 6.3.6 High speed tissue homogenizer -- Tekmar Tissuemizer® equipped with an EN-8 probe or equivalent. 6.3.7 Vacuum dessicator. A-9 �6.4 Glassware 6.4.1 Erlenmeyer flask — 500 ml. 6.4.2 Kuderna-Danish apparatus -- 500-mL evaporating flask, 15-mL graduated concentrator tubes with ground-glass stoppers, and three-ball macro Snyder column (Kontes K-570001-0500, K-503000-0121, and K-569001-0219 or equivalent). 6.4.3 Minivials -- 1-mL borosilicate glass with conical-shaped reservoir and screw caps lined with Teflon®-faced silicone disks. 6.4.4 Powder funnels -- glass. 6.4.5 Chromatographic columns for the silica and alumina chromatography -- 1 cm ID x 10 cm long and 1 cm ID x 30 cm long with 250-mL reservoir and equipped with TFE stopcocks. 6.4.6 Chromatographic column for the Carbopak cleanup — disposable 5-mL graduated glass pi pets, 6 to 7 mm ID. 6.4.7 Glass rods. 6.4.8 Carborundum boiling chips -- Extracted for 6 hr in a Soxhlet apparatus with benzene and air dried. 6.4.9 Glass wool, silanized (Supelco) — Extract with methylene chloride and hexane and air dry before use. 6.4.10 Glassware cleaning procedure -- All glassware used for these analyses will be cleaned via the following procedure. Wash the glassware in soap and water, rinse with copious amounts of tap water, distilled water, and distilled-in-glass acetone, in that order. Immediately prior to use, the glassware should be rinsed with distilled-in-glass quality solvents: methylene chloride, toluene, and hexane. The glassware should be allowed to dry fully. As an added precuation, all glassware will be marked with a unique code that should be noted in the extraction and cleanup procedures for each sample. This glassware tracking will allow background results from specific glassware to be documented. After use, each piece of glassware should be rinsed with the last solvent used in it, followed by a rinse with toluene, then acetone, before transferring it to the glassware washing facility. A-10 �7. REAGENTS AND STANDARD SOLUTIONS 7.1 Column Chromatography Reagents 7.1.1 Alumina, acidic (Biorad, AG-4) -- Extract the alumina in a Soxhlet apparatus with methylene chloride for 18 h (minimum of two cycles per hour). Air dry and activate it by heating in a foil-covered glass container for 24 h at 190°C. 7.1.2 Silica gel -- High purity grade, type 60, 70-230 mesh; extract the silica gel in a Soxhlet apparatus with methylene chloride for 10 h (minimum of 2 cycles per hour). Air dry and activate it by heating in a foilcovered glass container for 24 h at 130°C. 7.1.3 Silica gel impregnated with 40% (by weight) sulfuric acid -- Add two parts (by weight) concentrated sulfuric acid to three parts (by weight) silica gel (extracted and activated) (e.g., 40 g of H2S04 plus 60 g of silica gel) in a glass screw-cap bottle. Tumble for 5 to 6 h, shaking occasionally until free of lumps. 7.1.4 Sulfuric acid, concentrated -- ACS grade, specific gravity 1.84. 7.1.5 Graphitized carbon black (Carbopack C, Supelco), surface of approximately 12 mVg, 80/100 mesh -- Mix thoroughly 3.6 g of Carbopack C and 16.4 g of Celite 545® in a 40-mL vial. Activate at 130°C for 6 h. Store in a desiccator. 7.1.6 Celite 545® (Fischer Scientific), reagent grade, or equivalent. 7.2 Desiccating agents -- Sodium sulfate; granular, anhydrous. Before use extract with methylene chloride for 16 h (minimum of two cycles per hour), air dry and then muffle for ^ 4 h in a shallow tray at 400°C. Let it cool in a desiccator and store in oven at 130°C. 7.3 Solvents -- High purity, distilled in glass: methylene chloride, toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent grade: tridecane. High purity solvents are dispensed from Teflon® squirt bottles. 7.4 Concentration Calibration Solutions (Table 2) Eight tridecane solutions containing native calibration standards, 13 C12-labeled internal quantisation standards, and two internal recovery standards are required. The complete compound list is A-11 �Table 2. Concentration Calibration Solutions Compound Native Concentration in cal ibration solutions i n pg/uL CS1 CS2 CS3 CS4 CSS CS6 CS7 CSS 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3, 7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4, 7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF Internal Quantisation Standards 13 C12-2,3,7,8-TCDD 13 C12-2,3,7,8-TCDF 13 C12-l,2,3,7,8-PeCDD 13 C12-l,2,3,7,8-PeCDF 13 C12-l,2,3,6,7,8-HxCDD 13 C12-l,2,3,4,7,8-HxCDF 13 C12-l,2,3,4,6,7,8-HpCDD 13 C12-OCDD Internal Recovery Standard 13 C12-1,2,3,4-TCDD 200 200 200 200 200 500 500 500 500 500 500 500 500 500 500 1,000 1,000 100 100 100 100 100 250 250 250 250 250 250 250 250 50 50 50 50 50 125 125 125 125 25 25 25 25 25 62. 5 250 500 500 250 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 62. 5 125 250 125 250 125 125 125 125 125 125 10 10 10 10 10 25 25 25 25 25 25 25 25 25 25 50 50 50 50 50 50 125 125 125 250 50 50 50 50 50 50 50 50 50 125 125 125 250 125 125 125 250 50 50 50 125 125 125 250 50 50 50 50 125 125 125 250 50 50 50 50 50 A-12 5 5 5 5 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 12. 5 25 25 50 50 2. 5 2. 5 2. 5 2. 5 2. 5 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 6. 25 12. 5 12. 5 1 1 1 1 1 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 5 5 50 50 50 50 125 125 250 50 50 50 50 125 125 125 250 50 50 50 50 50 125 125 125 125 250 �given in Table 1. The native 2,3,7,8-TCDD is supplied as a certified standard solution from the U.S. EPA QA Reference Materials Branch. All other native compounds were supplied in crystalline form by Cambridge Isotope Laboratories (Woburn, MA). 13C12~ Labeled internal quantisation standards were supplied in solution in ji-nonane by Cambridge Isotope Laboratories. Portions of the native standards were accurately weighed to the nearest 0.001 mg with a Cahn 27 electrobalance and dissolved in toluene. 7.5 Column Performance Check Mixture The column performance check mixture consists of several TCDD isomers which will be used to document the separation of 2,3,7,8TCDD from all other isomers. This solution will contain TCDDs (A) eluting closely to 2,3,7,8-TCDD, and the first- (F) and lasteluting (L) TCDDs. Analyte Approximate amount per ampule Unlabeled 2,3,7,8-TCDD 13 C12-2,3,7,8-TCDD 1,2,3,4-TCDD (A) 1,4,7,8-TCDD (A) 1,2,3,7-TCDD (A) 1,2,3,8-TCDD (A) 1,3,6,8-TCDD (F) 1,2,8,9-TCDD (L) 7.6 10 10 10 10 10 10 10 10 ng ng ng ng ng ng ng ng Spiking Solutions Three solutions are prepared using the 13 same stock as in Section 7.4. A native standard solution and a C12 internal quantitation standard solution are prepared in isooctane (Tables 3 and 4). A recovery standard solution is prepared in tridecane (Table 4). Samples are spiked with 100 (jL of internal quantisation standard solution and final sample extracts are spiked with 10 |jL of internal recovery standard solution. 8. HIGH RESOLUTION GAS CHROMATOGRAPHY/MASS SPECTROMETRY PERFORMANCE CRITERIA Samples and standards are analyzed by using a Carlo Erba MFC500 gas chromatography (GC) coupled to a Kratos MS50TC double-focusing mass spectrometer (MS) to be operated in the electron impact mode. The HRGC/MS interface is simply a direct connection of the fused silica HRGC column to the ion source of the MS via a heated interface oven. Data acquisition and processing are controlled by a Finnigan-MAT Incos 2300 data system. A-13 �Table 3. Native Spiking Solution Concentration (pg/uO Compound 2,3,7,8-TCDD 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 1,2,3,7,8-PeCDF 2,3,4,7,8-PeCDF 1,2,3,4,7,8-HxCDD 1,2,3,6,7,8-HxCDD 1,2,3,7,8,9-HxCDD 1,2,3,4,7,8-HxCDF 1,2,3,6,7,8-HxCDF 1,2,3,7,8,9-HxCDF 2,3,4,6,7,8-HxCDF 1,2,3,4,6,7,8-HpCDD 1,2,3,4,6,7,8-HpCDF 1,2,3,4,7,8,9-HpCDF OCDD OCDF 5 5 5 5 5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 12.5 25 25 A-14 �Table 4. Internal Standard Spiking Solutions Concentration (pg/^L) Compound Internal Quanti tati on Standards C12-2,3,7,8-TCDD 13 5 13 5 ^Cia-l^.SJ.S-PeCDD 5 13 5 Ci2-2,3,7,8-TCDF C12-l,2,3,7,8-PeCDF 13 12.5 13 12.5 13 12.5 C12-l,2,3,6,7,8-HxCDD C12-l,2,3,4,7,8-HxCDF C12-l,2,3,4,6,7,8-HpCDD 13 C12-OCDD 25 Internal Recovery Standard 13 C12-1,2,3,4-TCDD A-15 5 �8.1 HRGC/MS Analysis of PCDD/PCDF Single run selected ion monitoring (SIM) analysis of the tetrachloro through octachloro-dioxins and furans is carried out with the instrumental conditions and parameters outlined in Table 5. For each HRGC/MS run, five distinct groups of ions, which correspond to each chlorine level, are sequentially monitored. These ion descriptors are shown in Table 6. The masses of the two most abundant ions in the molecular ion cluster of each dioxin and furan and isotopically labeled standard are monitored. In addition, the masses corresponding to the molecular ions of the hexachloro through decachlorodiphenyl ethers (PCDEs) are monitored to aid in the confirmation of positive furan results. A lock mass, m/z 381 from PFK (perfluorokerosene), is used to observe and correct any magnet/instrument drift during the analysis. 8.1.1 Tuning and Mass Calibration The mass spectrometer is tuned on a daily basis to yield optimum sensitivity and peak shape using an ion peak (m/z 381) from PFK. The resolution is visually monitored and maintained at £ 3,000 (10% valley definition) to provide adequate noise rejection while maintaining good ion transmission. Mass calibration of the mass spectrometer for the HRGC/MS • analysis of PCDD/PCDF is carried out on a daily basis. The magnetic field is adjusted to pass m/z 300 at full accelerating voltage. PFK is admitted to the MS and an accelerating voltage scan from 8,000 to 4,000 V is acquired by the data system. This corresponds to an effective mass range of 301 to 593 amu. Upon completion of a successful calibration step, the five ion descriptors shown in Table 6 are updated to reflect the new mass calibration. 8.1.2 Ion Descriptor Switching The ion descriptors shown in Table 6 are sequentially monitored during a PCDD/PCDF analysis to cover the retention windows of each chlorination level. The retention windows and hence the descriptor switch points are determined initially and whenever a new HRGC column is installed by injection of a mixture of PCDD and PCDF congeners. Daily adjustment of the descriptor switch times are performed when careful monitoring of the standard retention times shows this to be necessary. The descriptors are designed to overlap to ensure acquisition of all isomers of each homolog. A-16 �Table 5. HRGC/LRMS Operating Conditions for PCDD/PCDF Analysis Mass spectrometer Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: Overall SIM cycle time: 8,000 V 500 |jA 70 eV -1,800 V 280°C £ 3,000 (10% valley definition) 1s Gas chromatograph Column coating: Film thickness: Column dimensions: He linear velocity: He head pressure: DB-5 0.25 pm 60 m x 0.25 mm ID * 25 cm/sec 1.75 kg/cm2 (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 s 30 mL/min 6 mL/min 270°C 300°C 1-2 (JL 200°C 2 min 200°C to 330°C at 5°C/min A-17 �Table 6. Ions Monitored for HRGC/MS of PCDD/PCDF Descriptor Al ID Mass TCDF 303.902 305.899 315.942 317.939 319.896 321.894 331.937 333.934 373.840 380.976 TCDD 13 C12-TCDD HxDPE PFK (lock mass) TCDF TCDD PeCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD PFK (lock mass) HpCDPE A3 HxCDF PFK (lock mass) 13 C12-HxCDF HxCDD 13 C12-HxCDD OCDPE A-18 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.035 0.035 373.821 375.818 380.976 385.861 387.858 389.816 391.813 401.856 403.853 443.759 C12-TCDF 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 303.902 305.899 319.896 321.894 337.863 337.860 349.903 351.900 353.858 355.855 365.898 367.895 380.976 407.801 13 A2 Nominal dwell time (sec) 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 0.080 �Table 6 (continued) Descriptor A4 ID Mass PFK (lock mass) HxCDD HpCDF 13 C12-HpCDF HpCDD 13 C12-HpCDD 37 Cl4-HpCDD NCDPE A5 PFK (lock mass) OCDF 13 C12-OCDF OCDD 13 C12-OCDD DCDPE A-19 Nominal dwell time (sec) 380.976 389.816 391.813 407.782 409.779 419.822 421.819 423.777 425.774 435.817 437.814 429.768 431.765 477.720 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.040 380.976 441. 743 443.740 453.783 455.780 457.738 459.735 469.779 471.776 511.681 0.06 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.06 �8.1.3 HRGC Column Performance (60-m DB-5) 8.1.3.1 Inject 1 uL of the column performance check solution (Section 7.7) and acquire selected ion monitoring (SIM) data for m/z 320, 322, 332, and 334. 8.1.3.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers should be resolved with a valley of 30-60%, where Valley % = (x/y)(100) x = measured height of the valley between the chromatographic peak corresponding to 2,3,7,8-TCDD and the peak of the nearest TCDD isomer; and y '= the peak height of 2,3,7,8-TCDD. Figure 2 is an example of the separation of a TCDD isomer mixture and the calculation of isomer resolution. It is the responsibility of the laboratory to verify the conditions suitable for the appropriate resolution of 2,3,7,8-TCDD from all other TCDD isomers. The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol, thus defining the retention time window for total TCDD determination. Any individual selected ion current profile or the reconstructed total ion current (m/z 320 + m/z 322) consititutes and acceptable form of data presentation. 8.1.4 Initial Calibration for PCDD/PCDF Analysis Initial calibration is required before any samples are analyzed for PCDD/PCDF. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.1.6. 8.1.4.1 Tune and calibrate the instrument with PFK as outlined in Section 8.1.1. 8.1.4.2 Six concentration calibration solutions listed in Table 2 will be analyzed for the initial calibration phase. A-20 �TCDD Isomer Mixture 78.7 2. 3. 7, 8,-TCDD 1, 2, 3, 4-/1, 2359290. 2, 3, 7-/L 2. 3, 8- TCDI 319.856 ±0.500 A 320 100.0, 2998270. A x 100% = 29% 322 3> 321.855 ± 0.500 A 59.6 1785850. Q2-2. 3. 7, 8.-TCDD INS 331.851 ± 0.500 332 75.3 2256890. 333.850 ± 0.500 334 28:00 28:24 28:48 29:12 29:36 30:00 30:24 30:48 Figure 2. Example of the separation of 2,3,7,8-TCDD from other TCDD isomers on a 60 m DB-5 column. Time �8.1.4.3 8.1.4.4 Compute the relative response factors (RRFs) for each analyte in the concentration calibration solution using the criteria for positive identification of PCDD/PCDF's given in Section 14.1 and the computational methods in Section 14.2. 8.1.4.5 Compute the means and their respective relative standard deviations (% RSD) for the RRFs from each triplicate analysis for each analyte in the standard. 8.1.4.6 8.1.5 Using the HRGC and MS conditions in Table 5 and the SIM monitoring descriptors in Table 6, analyze a 1-jjL aliquot of each of the six concentration calibration solutions in triplicate. Calculate the grand means (RRF) and their respective RSDs using the six mean RRFs for each analyte. Criteria for Acceptable Initial Calibration 8.1.5.1 The % RSD for the response factors for each triplicate analysis for each analyte must be less than 30% except for the TCDD and TCDF, which must be less than 20%. 8.1.5.2 The variation of the mean RRFs obtained from the triplicate analysis must be less than 30% except for the TCDD and TCDF which must be less than 20%. 8.1.5.3 The SIM traces for all ions used for quantitation must present a signal-to-noise (S/N) ratio of S 2.5. This includes analytes and isotopically labeled standards. 8.1.5.4 Isotopic ratios must be within ±20% of the theoretical values (see Table 7). NOTE: If the criteria for acceptable calibration listed abo've have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The mean RRF from triplicate determinations for unlabeled PCDD/ PCDFs and for the isotopically labeled standards will be used for all calculations until routine calibration criteria (Section 8.1.7) are no longer met. At such time, new mean RRFs will be calculated from a new set of six triplicate determinations. A-22 �Table 7. Ion Ratios for HRGC/LRMS Analysis of PCDD/PCDF Compound TCDF 13 C12-TCDF TCDD 13 C12-TCDD PeCDF 13 C12-PeCDF PeCDD 13 C12-PeCDD HxCDF 13 C12-HxCDF HxCDD 13 C12-HxCDD HpCDF 13 C12-HpCDF HpCDD 13 C12-HpCDD OCDF 13 C12-OCDF OCDD 13 C12-OCDD Ions monitored Theoretical ratio 304/306 316/318 320/322 332/334 338/340 350/352 354/356 366/368 374/376 386/388 390/392 402/404 408/410 420/422 424/426 436/438 442/444 454/456 458/460 470/472 0.76 0.76 0.76 0.76 0.61 0.61 0.61 0.61 1.22 1.22 1.22 1.22 1.02 1.02 1.02 1.02 0.87 0.87 0.87 0.87 A-23 Acceptabl e range 0.61 0.61 0.61 0.61 0.49 0.49 0.49 0.49 0.96 0.96 0.96 0.96 0.82 0.82 0.82 0.82 0.70 0.70 0.70 0.70 - 0.91 0.91 0.91 0.91 0.73 0.73 0.73 0.73 1.46 1.46 1.46 1.46 1.22 1.22 1.22 1.22 1.04 1.04 1.04 1.04 �8.1.6 Routine Calibrations Routine calibrations must be performed at the beginning of every day before actual sample analyses are performed and as the last injection of every day. 8.1.6.1 8.1.6.2 8.1.7 Inject 1 uL of the concentration calibration solution CS 7 (see Table 2). Compute the RRFs for each analyte in the concentration calibration solution using the criteria for positive identification of PCDD/Fs given in Section 14.1 and the computational methods in Section 14.2. Criteria for Acceptable Routine Calibration 8.1.7.1 8.1.7.2 Isotopic ratios must be within 20% of the theoretical value for each analyte and isotopically labeled standard (see Table 8). 8.1.7.3 8.2 The measured RRF for all analytes must be within 30% of the mean values established by triplicate analysis of the calibration concentration solutions, except for TCDD and TCDF, which must be within 20% of the mean values established in the initial calibration step. If the above criteria are not met, a second attempt may be made before repeating the entire initialization process. HRGC/HRMS Analysis (Isomer Specific TCDD Analysis) Isomer specific analysis for 2,3,7,8-TCDD is carried out with the instrumental conditions and parameters shown in Table 8. In addition to monitoring the masses of the most abundant molecular ions of TCDD, an ion corresponding to the loss of COC1 from the molecular ion is monitored for verification purposes. Mass spectrometer resolution is maintained at or above 10,000 (10% valley definition) in order to increase the specificity of the analysis. 8.2.1 Tuning and Mass Calibration 8.2.1.1 The mass spectrometer must be operated in the electron (impact) ionization mode. Static resolving power of at least 10,000 (10% valley) must be demonstrated before any analysis of a set of samples is performed. Static resolution checks must be performed at the beginning and at the end A-24 �Table 8. HRGC/HRMS Operating Conditions Mass spectrometer Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: 8,000 V 500 (jA 70 eV 2,000 V 280°C 10,000 (10% valley definition) Sim Parameters Identity Mass TCDD-COC1 TCDD TCDD 13 C12-TCDD 13 C12-TCDD 0.15 0.15 0.15 0.15 0.15 0.10 258.930 319.897 321.894 331.937 333.934 PFK (lock mass) Nominal dwel1 times (s) 280.983 Overall SIM cycle time = 1 s Gas chromatqgraph Column coating: Film Thickness: Column dimensions: CP-Sil 88 0.2 pm 50 m x 0.22 mm ID Helium linear velocity: Helium head pressure: •v 25 cm/s 2 1.75 kg/cm (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Split!ess, 45 s 30 mL/min 6 mL/min 270°C 240°C 2 ML 200°C 1 min 200°C to 240°C at 4°C/min A-25 �of each 12-h period of operation. However, it is recommended that a visual check.(i.e., not documented) of the static resolution be made before and after each analysis. 8.2.1.2 8.2.1.3 8.2.2 The MS shall be tuned daily using PFK to yield a resolution of at least 10,000 (10% valley) and optimal response at m/z 254.986. This step is followed by calibration of an accelerating voltage scan of PFK beginning at m/z 254 (typical calibration range is 255 to 493 amu). Other voltage scans from the same data file are used to establish and document both the resolution at m/z 316.983 and the mass measurement accuracy at m/z 330.979. Following calibration, the SIM experiment descriptor is updated to reflect the new calibration. Six masses (see Table 8) are monitored by scanning ^ m/10,000 amu (atomic mass units) over each mass. The total cycle time is kept to 1 s. The m/z 280.983 ion from PFK is used as a lock mass because it is the most abundant PFK ion within the range of m/z 255 to 334 and therefore permits the use of low partial pressures of PFK, which minimizes PFK interferences at the analytical masses. Mass Measurement and Resolution Check Using a PFK molecular leak, tune the instrument to meet the minimum required resolving power of 10,000 (10% valley) at m/z 254.986 (or any other mass reasonably close to m/z 259). Calibrate the voltage sweep at least across the mass range m/z 259 to m/z 334 and verify that m/z 330.979 from PFK (or any other mass close to m/z 334) is measured within ± 5 ppm (i.e., 1.7 mmu, if m/z 331 is chosen) using m/z 254.986 as a reference. Documentation of the mass resolution must then be accomplished by recording the peak profile of the PFK reference peak m/z 318.979 (or any other reference peak at a mass close to m/z 320/322). The format of the peak profile representation must allow manual determination of the resolution; i.e., the horizontal axis must be a calibrated mass scale (amu or ppm per division). The results of the peak width measurement (performed at 5% of the maximum which corresponds to the 10% valley definition) must appear on the hard copy and cannot exceed 100 ppm (or 31.9 mmu if m/z 319 is the chosen reference ion). A-26 �8.2.3 HRGC Column Performance (50-m CP Sil 88/60-m SP-2330) Prior to any HRGC/HRMS analysis of calibration solutions on samples for 2,3,7,8-TCDD, the resolution of the HRGC columns must be documented to be within allowable limits in order to provide conditions adequate for unambiguous isomer-specific analysis of 2,3,7,8-TCDD. 8.2.3.1 Inject 2 pL of the column performance check solution and acquire selected ion monitoring (SIM) data for m/z 258.930, 319.897, 321.894, 331.937, and 333.934 within a total cycle time of < 1 s (Table 8). 8.2.3.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers must be resolved with a valley of ^ 25%, where Valley % = (x/y)(100) x = measured height of the valley between the chromatographic peak corresponding to 2,3,7,8-TCDD and the peak of the nearest TCDD isomer; and y = the peak height of 2,3,7,8-TCDD. 8.2.3.3 If the above resolution requirement is not met, corrective action must be taken and acceptable resolution documented prior to any further analyses. Corrective action may include removal of the first meter of the HRGC column, replacement or clearing of the injector port, or complete replacement of the GC column. 8.2.3.4 The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol, thus defining the retention time window for total TCDD determination. The peaks representing 2,3,7,8-TCDD and the first and the last eluting TCDD isomer should be labeled and identified as such on the chromatograms (F and L, respectively). Any individual selected ion current profile or the reconstructed total ion current (m/z 259 + m/z 320 + m/z 322) constitutes an acceptable form of data presentation. A-27 �8.2.4 Initial Calibration for HRGC/HRMS 2,3,7,8-TCDD Analysis Initial calibration is required before any samples are analyzed for 2,3,7,8-TCDD. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.2.6. 8.2.4.1 8.2.4.2 Tune and calibrate the instrument with PFK as described in Section 8.2.1.- 8.2.4.3 Inject 1 uL of the column performance check solution (Section 8.2.3) and acquire SIM mass spectra data for m/z 258.930, 319.897, 321.894, 331.937, and 333.934 using a total cycle time of S 1 s (see Table 8). The laboratory must not perform any further analysis until it has been demonstrated and documented that the criterion listed in Section 8.2.3.2 has been met. 8.2.4.4 Using the same GC and MS conditions (Table 8) that produced acceptable results with the column performance check solution, analyze a l-(jl_ aliquot of each of the six concentration calibration solutions in triplicate. 8.2.4.5 Calculate the RRFs for unlabeled 2,3,7,8TCDD relative to 13C12-2,3,7,8-TCDD and the RRF for 13C12-2,3,7,8-TCDD relative to 13 C12-1,2,3,4-TCDD using the criteria for positive identification of TCDD by HRGC/ HRMS given in Section 14.1 and the computational methods in Section 14.2. 8.2.4.6 Calculate the six means (RRFs) and their respective relative standard deviations ( RSD) for the response factors from each % of the triplicate analyses for both unlabeled and 13C12-2,3,7,8-TCDD. 8.2.4.7 8.2.5 At least six of the concentration calibration solutions listed in Table 2 must be utilized for the initial calibration. Calculate the grand mean RRFs and their respective relative standard deviations ( RSD) using the six mean RRFs. % Criteria for Acceptable Initial Calibration The criteria listed below for acceptable calibration must be met before analysis of any sample is performed. A-28 �8.2.5.1 The percent relative standard deviation (RSD) for the response factors from each of the triplicate analyses for both unlabeled and 13C12-2,3,7,8-TCDD must be less than 20%. 8.2.5.2 The variation of the six mean RRFs for unlabeled and 13C12-2,3,7,8-TCDD obtained from the triplicate analyses must be less than 20% RSD. 8.2.5.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise ratio of § 2.5 for m/z 258.930, m/z 319.897, and m/z 321.894. 8.2.5.4 SIM traces for 13C12-2,3,7,8-TCDD must present a signal-to-noise ratio 5 2.5 for m/z 331.937 and m/z 333.934. 8.2.5.5 Isotopic ratios for 320/322 and 332/334 must be within the allowed range (0.67 to 0.90). NOTE: If the criteria for acceptable calibration listed above have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The mean RRF from six triplicate determinations for unlabeled 2,3,7,8-TCDD and for 13C12-2,3,7,8-TCDD will be used for all calculations until routine calibration criteria (Section 8.2.6) are no longer met. At such time, new mean RRFs will be calculated from a new set of four triplicate determinations. 8.2.6 Routine Calibrations Routine calibrations must be performed at the beginning of a 12-h period after successful mass resolution and HRGC column performance check runs. 8.2.6.1 Inject 1 yL of the concentration calibration solution (CS5, Table 2) which contains 10 pg/|Jl of unlabeled 2,3,7,8-TCDD, 50.0 pg/HL of 13C12-2,3,7,8-TCDD, and 50 pg/|jL of 13C12-1,2,3,4-TCDD. Using the same GC/ MS/DS conditions as used in Table 8, determine and document acceptable calibration as provided below. A-29 �8.2.7 Criteria for Acceptable Routine Calibration The following criteria must be met before further analysis is performed. If these criteria are not met, corrective action must be taken and the instrument must be recalibrated. 8.2.7.1 The measured RRF for unlabeled 2,3,7,8-TCDD must be within 20% of the mean values established in the initial calibration by triplicate analyses of concentration calibration solutions. 8.2.7.2 The measured RRF for 13C12-2,3,7,8-TCDD must be within 20% of the mean value established by triplicate analysis of the concentration calibration solutions during the initial calibration. 8.2.7.3 Isotopic ratios must be within the allowed range (0.61 to 0.90). 8.2.7.4 If one of the above criteria is not satisfied, a second attempt can be made before repeating the entire initialization process. NOTE: An initial calibration must be carried out whenever the routine calibration solution is replaced by a new one from a different lot. 9. QUALITY CONTROL PROCEDURES 9.1 Summary of QC Analyses 9.1.1 Initial and routine calibration and instrument performance checks. 9.1.2 Analysis of a batch of samples with accompanying QC analyses: Sample batch -- 10 NHATS adipose tissue samples plus additional QC analyses including 1 method blank, a control tissue and a spiked tissue sample. "Blind" QC (external QC) samples may be submitted by an external source (quality assurance group or independent laboratory) and included among the batch of samples. Blind samples include spiked samples, unidentified duplicates, and performance evaluation samples. A-30 �9.2 Performance Evaluation Samples -- Included among the samples in every third batch will be a solution provided by the quality control coordinator containing known amounts of unlabeled 2,3,7,8TCDD and/or other PCDD/PCDF isomers. The accuracy of measurements for performance evaluation samples should be in the range of 70-130%. 9.3 Performance Check Solutions 9.3.1 9.4 At the beginning of each 12-h period during which samples are to be analyzed, an aliquot of the HRGC column performance check solution shall be analyzed to demonstrate adequate HRGC resolution for selected TCDD isomers. Method Blanks 9.4.1 A minimum of one method blank is generated with each batch of samples. A method blank is generated by performing all steps detailed in the analytical procedure using all reagents, standards, equipment, apparatus, glassware, and solvents that would be used for a sample analysis, but omit addition of the adipose tissue. 9.4.1.1 The method blank must contain the same amounts of Carbon-13 labeled internal quantitation standards that are added to samples before bulk lipid cleanup. 9.4.1.2 An acceptable method blank exhibits no positive response for any of the characteristic ions monitored. 9.4.1.2.1 9.4.1.2.2 9.5 If the above criterion is not met, solvents, reagents, spiking solutions, apparatus, and glassware are checked to locate and eliminate the source of contamination before any samples are extracted and analyzed. If new batches of reagents or solvents contain interfering contaminants, purify or discard them. Control Samples -- Control samples are prepared from a bulk sample(s) of human adipose tissue or similar matrix (e.g., porcine fat). This material is prepared by blending the tissue with methylene chloride, drying the extract by eluting through anhydrous sodium sulfate, and removing the methylene chloride using rotoevaporation at elevated temperatures (80°C). The evaporation process should be extended to ensure all traces of the extraction A-31 �solvent have been removed. The resulting oily matrix (lipid) is subdivided into 10-g aliquots which are analyzed with each sample batch. The results of the individual analysis will be used to give a measure of precision from batch to batch over an entire program. Sufficient tissue should be extracted to provide a homogeneous lipid matrix that can be used over the total analysis program. Enough lipid matrix is necessary to prepare the spiked samples describe in Section 9.6. 9.6 Spiked Samples — Spiked lipid samples are prepared using a portion of the homogenized lipid described in Section 9.5. Sufficient spiked lipid matrix is prepared to provide a minimum of one spiked sample per sample batch. It is recommended that a minimum of three spiked levels of the matrix are prepared ranging from 10 to 50 times the estimated limit of detection for each compound. Each analysis of spiked sample must be accompanied by analysis of a control sample in order to make the necessary corrections for background contribution before determining the accuracy of the method (Equation 9-1). Accuracy (%) = 100% x Cone, spiked samp^-conc. control sample Eq ^ 9.7 Duplicate Sample Analysis -- When possible a duplicate analysis of specific samples is included in the sample batch as an additional measure of method precision. It is suggested that the total tissue sample is extracted to isolate lipids material and then subdivided for duplicate analysis. Precision is calculated as relative percent difference (RPD) where the differences in the duplicate measurements (for each analyte) is divided by the average of the two measurements and multiplied by 100%. 9.8 External Samples ~ Samples submitted as blinds to the analyst may consist of either performance solutions of PCDD and PCDF congeners or spiked sample matrices. These performance solutions or samples should be submitted by a source external to the analytical program (QA unit of analysis laboratory or independent laboratory). Performance audit solutions are intended to evaluate instrument calibration and quantisation procedures. Spiked blind samples must be accompanied by the corresponding unspiked samples to correct concentrations for background concentration. The blind spiked samples are intended to evaluate the total analytical procedure. The analyst must keep in mind that it is necessary to compare differences in standard sources for each type of external sample. 10. SAMPLE PRESERVATION AND HANDLING All adipose tissue samples must be maintained at less than -20°C from time of collection. The analyst should instruct the collaborator collecting the sample(s) to avoid the use of chlorinated materials. Samples are handled using stainless steel forceps, spatulas, or scissors. A-32 �Aliquots of samples removed from sample bottles not used for analysis are disposed rather than returned to the sample vial. All sample bottles (glass) are cleaned as specified in Section 6.4.10. Teflon®-!ined caps should be used. As with any biological sample, the analyst should avoid any undue exposure. 11. SAMPLE EXTRACTION 11.1 Extraction of Adipose Tissue 11.1.1 Accurately weigh to the nearest 0.01 g a 10-g portion of a frozen adipose tissue sample into a culture tube (2.2 x 15 cm). Note: Sample size may be smaller, depending on availability. 11.1.2 11.1.3 Allow the mixture to separate and decant the methylene chloride extract from the residual solid material using a disposable pipette. The methylene chloride is eluted through a filter funnel containing a plug of clean glass wool and 5 to 10 g of anhydrous sodium sulfate. The dried extract is collected in a 100-mL volumetric flask. 11.1.4 A second 10-mL aliquot of methylene chloride is added to the sample and homogenized for 1 min. The methylene chloride is decanted, dried, and transferred to the 100-mL volumetric flask as specified in Section 11.1.3 11.1.5 The culture tube is rinsed with at least two additional aliquots (10 mL each) of methylene chloride, and the entire contents are transferred to the filter funnel containing the anhydrous sodium sulfate. The filter funnel and contents are rinsed with additional methylene chloride (20 to 40 mL). The total eluent from the filter funnel is collected in the 100-mL volumetric flask. Discard the sodium sulfate. 11.1.6 11.2 Allow the adipose tissue specimen to reach room temperature. Add 10 ml of methylene chloride and homogenize the mixture for approximately 1 min with a Tekmar Tissuemizer®. The final volume of the extract for each sample is adjusted to 100 mL in the volumetric flask using methylene chloride. Lipid Determination 11.2.1 Preweigh a clean 1-dram glass vial to the nearest 0.0001 g using an analytical balance tared to zero. A-33 �11.2.2 Accurately transfer 1.0 ml of the final extract (100 ml) from Section 11.1.6 to the 1-dram vial. Reduce the volume of methylene chloride from the extract using a water bath (50-60°C) gentle stream of purified nitrogen until an oil residue remains. 11.2.3 Accurately weigh the 1-dram vial and residue to the nearest 0.0001 g and calculate the weight of lipid present in the vial based on difference. Nitrogen blow-down is continued until a constant weight is achieved. 11.2.4 Calculate the percent lipid content of the original sample to the nearest 0.1% as shown in Equation 11-1. I D ^ PYT Lipid content, LC (%) = r^ X V r x 100% ^ W AT Eq. 11-1 AL where: W.R = weight of the lipid residue to the nearest 0.0001 g calculated from Section 11.2.3 (100.0 ml); VtAI = total volume of the extract in mL from FYT Section 11.1.6; W.,. = weight of the original adipose tissue samples to the nearest 0.01 g from Section 11.1.1; and V.. = volume of the aliquot of the final extract in ml used for the quantitative measure of the lipid residue (1.0 mL). 11.2.5 11.3 Record the lipid residue measured in Section 11.2.3 and the percent lipid content calculated from Section 11.2.4. Extract Concentration 11.3.1 11.3.2 11.4 Quantitatively transfer the remaining extract volume (99.0 mL) to a 500-mL Erlenmeyer flask. Rinse the volumetric flask with 20 to 30 mL of additional methylene chloride to ensure quantitative transfer. Place the Erlenmeyer flask on a hot plate at 40°C to remove solvent until an oily residue remains. Addition of Internal Quantisation Standards To each lipid residue add the carbon-13 internal quantisation spiking solution (Section 7.8) such that it delivers 500 to 2,500 pg of each of the surrogates specified in Table 4 in a 100-(jL volume. A-34 �12. CLEANUP PROCEDURES 12.1 Bulk Lipid Removal 12.1.1 Add a total of 200 ml of n-hexane to the spiked lipid residue in the 500-mL Erlenmeyer flask. 12.1.2 Slowly add, with stirring, 100 g of the 40% w/w sulfuric acid impregnated silica gel (Section 7.1.3). Stir with a magnetic stir-plate for 2 h. 12.1.3 Allow solids to settle and decant liquid through a powder funnel containing 20 g of anhydrous sodium sulfate and collect in a 500-mL sample bottle. 12.1.4 Rinse solids with two 50-mL portions of hexane. Stir each rinse for 15 min, decant, and dry by elution through sodium sulfate combining the hexane extracts from Section 12.1.3. 12.1.5 After the rinses have gone through the sodium sulfate, rinse the sodium sulfate with an additional 25 mL of hexane and combine with the hexane extracts from Section 12.1.4. 12.1.6 Prepare an acidic silica column as follows: Pack a 1 cm x 10 cm chromatographic column with a glass wool plug, add approximately 25 mL of hexane, 1.0 g of silica gel (Section 7.1.2), and 4.0 g of 40% w/w sulfuric acid impregnated silica gel (Section 7.1.3). Pack a second chromatographic column (1 cm x 30 cm) with a glass wool plug, approximately 25 mL of hexane, 6.0 g of acidic alumina (Section 7.1.1), and top with a 1-cm layer of sodium sulfate (Section 7.4). Elute the combined hexane solutions (Section 12.1.5) through the columns until the solvent level reaches the top of the chromatographic packing. Inspect columns to ensure they are free of channels and air bubbles. Wash the alumina column with 40 mL of 50% v/v methylene chloride/hexane. Remove methylene chloride from the adsorbent by eluting the column with an additional 100 mL of hexane. 12.1.7 Quantitatively transfer the hexane extract from the Erlenmeyer flask (Sections 12.1.3 through 12.1.5) to the silica gel column reservoir. Allow the hexane extract to percolate through the column and collect in a KD concentrator. 12.1.8 Complete the elution of the extract from the silica gel column with 50 mL of hexane in the KD concentrator. Concentrate the eluate to approximately mL, using nitrogen blow-down as necessary. A-35 �Note: If the 40% sulfuric acid/silica gel is noted to be highly discolored throughout the length of the adsorbent bed it is necessary to repeat the cleaning procedure beginning with Section 12.1.1. 12.2 Separation of Chemical Interferences 12.2.1 Transfer the concentrate (1.0 ml) to the top of the alumina column. Rinse the K-D assembly with two 1.0-mL portions of hexane and transfer the rinses to the top of the alumina column. Elute the alumina column with 18 ml of hexane until the hexane level is just below the top of the sodium sulfate. Discard the eluate. Columns must not be allowed to reach dryness (i.e., a solvent "head" must be maintained). 12.2.2 Place 30 ml of 20% (v/v) methylene chloride in hexane on top of the alumina and elute the TCDDs from the column. Collect this fraction in a 50-mL culture tube. 12.2.3 Prepare an 18% Carbopak C/Celite 545® mixture by thoroughly mixing 3.6 g of Carbopak C (80/100 mesh) and 16.4 g of Celite 545® in a 40-mL vial. Activate at 130°C for 6 h. Store in a desiccator. Cut off a clean 5-mL disposable glass pipet (6 to 7 mm ID) at the 4-mL mark. Insert a plug of glass wool (Section 7.3) and push to the 2-mL mark. Add 500 mg of the activated Carbopak/Celite mixture followed by another glass wool plug. Using two glass rods, push both glass wool plugs simultaneously towards the Carbopak/Celite mixture and gently compress the Carbopak/Celite plug to a length of 2 to 2.5 cm. Pre-elute the column with 2 ml of toluene followed by 1 ml of 75:20:5 methylene chloride/methanol/ benzene, 1 ml of 1:1 cyclohexane in methylene chloride, and 2 ml of hexane. The flow rate should be less than 0.5 mL/min. While the column is still wet with hexane, add the entire eluate (30 ml) from the alumina column (Section 12.2.2) to the top of the column. Rinse the culture tube which contained the extract twice with 1 ml of hexane and add the rinsates to the top of the column. Elute the column sequentially with two 1-mL aliquots of hexane, 1 ml of 1:1 cyclohexane in methylene chloride, and I ml of 75:20:5 methylene chloride/methanol/benzene. Turn the column upside down and elute the PCDD/PCDF fraction with 20 ml of toluene into 6-dram vial. 12.2A Warm the vial to approximately 60°C and reduce the toluene volume to approximately 1 ml using a stream of nitrogen. Carefully transfer the concentrate into a 1-mL minivial and reduce the volume to about 200 nL using a stream of nitrogen. A-36 �12.2.5 Rinse the concentrator tube with three washings using 500 ML of 1% toluene in CH2C12 (Section 12.2.5) concentrated to 200-500 ML and add 10 ML of the tridecane solution containing the internal recovery standard and store the sample in a refrigerator until HRGC/MS analysis. 12.2.6 Immediately prior to analysis, using a gentle stream of nitrogen at room temperature, remove toluene and methylene chloride. Submit sample to HRGC/MS once a stable 10 ML volume of tridecane is attained. 13. ANALYTICAL 13.1 PROCEDURES HRGC/MS Analysis for PCDD/PCDF 13.1.1 Once routine calibration criteria are met, the instrument is ready for sample analysis. Prior to the first sample, a blank injection of tridecane should be analyzed to document system cleanliness. If any evidence of system contamination is found, corrective action must be taken and another tridecane blank analyzed. The typical daily sequence of injections is shown in Table 9 and Figure 3. Note: Syringe Technique — Congeners of PCDD/PCDF in the syringes used for HRGC/MS analysis can be problematic unless the syringes are properly handled between samples. The following procedure has been found to be very effective for PCDD/PCDF removal from contaminated syringes and will be used throughout these analyses. • Rinse the syringe 10 times with isooctane. • Fill the syringe with toluene and sonicate syringe and plunger in toluene for 5 min and repeat at least twice. • Rinse the syringe 10 times with tridecane and pull up 1 ML of clean tridecane. • Syringe is ready for use. At no time should air be introduced into the HRGC column by using an air plug in the syringe. The oxygen present in the air plug will quickly degrade a nonbonded GC phase. 13.1.2 Inject a 1-ML aliquot of the extract into the GC, operated under the conditions previously used (Section 8.1) to produce acceptable results with the performance check solution. A-37 �Table 9. Typical Daily Sequence for PCDD/PCDF Analysis 1. Tune and calibrate mass scale versus perfluorokerosene (PFK). 2. Inject concentration calibration solution 2.5 to 12.5 pg/(jL (CS-7) solution. 3. Inject blank (tridecane). 4. Inject samples 1 through "N". 5 Inject concentration calibration solution 2.5 to 12.5 pg/uL (CS-7) solution. A-38 �INSTRUMENTAL ANALYSIS Instrument Mass Calibration vs PFK I Column Performance Evaluation Does Column Performance Meet Minimum Resolution Requirements? No Adjust Column Length or Install New Column No Reanalyze or Prepare Fresh Calibration Standards and Calibration Curve Yes Calibration Standard Analysis Do Relative Response Factors Meet Criteria Based on Initial Calibrations ? Yes Proceed with Sample Analysis Figure 3. Daily QA procedures for proceeding with sample analysis. A-39 �13.1.3 Acquire SIM data according to the same acquisition and MS operating conditions previously used (Section 8.1) to determine the relative response factors. 13.1.3.1 13.1.3.2 13.2 Acquire SIM data for the characteristic ions designated in Table 6. Instrument performance shall be monitored by examining and recording 13 peak areas the for the recovery standard, C12-1,2,3,4-TCDD. If this area should decrease to less than 50% of the previous run, sample analyses shall be stopped until the problem is found and corrected. HRGC/HRMS Confirmation of 2,3,7,8-TCDD 13.2.1 Once the daily criteria of mass calibration, mass resolution, HRGC performance, and routine calibration are met and documented, the instrument is ready for sample analysis. Prior to the first sample, a blank injection of tridecane will be made to document system cleanliness. The typical daily schedule for HRGC/HRMS analysis of TCDD is shown in Table 10 and Figure 3. 13.2.2 Inject a 1-uL aliquot of the extract into the GC, operated under the conditions previously used (Section 8.1) to produce acceptable results with the performance check solution. 13.2.3 Acquire SIM data according to Section 12.3.1. Use the same acquisition and MS operating conditions previously used (Section 8.3.4) to determine the relative response factors. 13.2.3.1 Acquire SIM data for the following selected characteristic ions: m/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD 321.894 Unlabeled TCDD 331.937 13 C12-2,3,7,8-TCDD, 13 C12-1,2,3,4-TCDD 333.934 13 C12-2,3,7,8-TCDD, 13 C12-1S2,3,4-TCDD A-40 �Table 10. Typical Daily Schedule for HRGC/HRMS Analysis of TCDD 1. Tune and calibrate mass scale. 2. Perform mass measurement check and mass resolution check. 3. Inject column performance check solution. 4. Inject the routine concentration calibration solution. 5. Inject tridecane blank. 6. Inject samples I through "N". 7. Inject column performance check solution. 8. Mass resolution check. A-41 �14. DATA REDUCTION In this section, the the analysis of data HRGC/HRMS method for qualitative criteria 14.1 procedures for the data reduction are outlined for from both the HRGC/MS method for PCDD/PCDF and the 2,3,7,8-TCDD. Figure 4 presents a schematic of the for identifying PCDDs and PCDFs. Qualitative Identification 14.1.1 14.1.2 The ion current intensities for a particular PCDD/PCDF must be S 2.5 times the noise level (S/N S 2.5) for positive identification of that isomer. 14.1.3 The integrated ion current ratios of the analytical masses for a particular PCDD/PCDF must fall within the ranges shown in Table 7. 14.1.4 14.2 The ion current responses for each mass for a particular PCDD/PCDF analyte must be within ± 1 s to attain positive identification of that analyte. For example, m/z 338 and m/z 340 must have maximum peak responses that are within ± 1 s to be positively identified as a pentachlorodibenzofuran. The recovery of the internal quantisation standards should be between 50 and 115%. Quantitative Calculations 14.2.1 Relative response factors (RRF). RRFs are calculated from the data obtained during the analysis of concentration calibration solutions using the following formula: C A x IS RRF = -^ ' j£ A C Eq. 14-1 IS ' x where A = the sum of the integrated ion for the analyte in question. for TCDD, A would be the sum grated ion abundances for m/z abundances For example, of the inte320 and 322; A,s = the sum of the integrated ion abundances for the labeled PCDD/F used as the internal quantisation standard for the above analyte. For example, for 13C12-TCDD, A,s would be the sum of the integrated ion abundance for m/z 332 and 334. C = concentration of the analyte in A-42 �HRGC/MS-SIM Data Response to Characteristic Molecular Ions within the Appropriate Homolog Retention Window? Report Compounds as Not Detected (ND) Calculate Sample LOD Characteristic Ion Ratios within "±20% Theoretical? Response Due to Coextracted Interference Response Correspondsito Specific Isomerj Retention Time? 'Quantitate Compound as Per Protocol Report as Isomer Unknown Quantitate Specific Isomer as per Protocol Figure 4. Qualitative criteria for identifying PCDDs and PCDFs. A-43 �CjS = concentration of the internal quantisation standard in pg/uL; and RRF = relative response factor. NOTE: The above formula is also used to compute the RRFs for the various internal standards relative to the recovery standard, 13C12-1,2,3,4-TCDD. 14.2.2 Concentrations of sample components. Figure 5 presents a schematic for quantisation of PCDDs and PCDFs which meet the criteria specified in Section 14.1. Calculate the concentration of PCDD/Fs in sample extracts using the formula: y TC A,. - Qis X C x= A IS' RRF •' W AT • LC Eq. 14-2 where Cx = the lipid adjusted concentration of PCDD or PCDF congener in pg/g; Ax = sum of the integrated ion abundances determined for the PCDD/PCDF in question; AIS = sum of the integrated ion abundances determined for the labeled PCDD/F used as the internal quantisation standard for the above analyte; QjS = the amount (total pg) of the labeled internal quantisation standard added to the sample prior to extraction; RRF = relative response factor of the above analyte relative to its labeled internal quantisation standard determined from the initial triplicate calibration; WAT = weight of original adipose tissue sample; Al and LC = percent extractable lipid determined from Eq. 11-1. Quantitative data should be classified to indicate the intensity of the signal response. Suggested qualifiers include: not detected, NO (signal-to-noise ratio is less than 2.5); trace, TR (signal-to-noise ratio is greater than or equal to 2.5 but less than 10); and positive quantifiable, PQ (signal-to-noise ratio is greater than or equal to 10. A-44 �QUANTITATION HRGC/MS-SIMData Response Meets All Qualitative Criteria? Report as Not Detected Calculate Sample LOD Response >2.5t[mes S/N? Response > 10 times S/N? Calculate as per Protocol Report as Trace (tr) Value Quantitate as per Protocol Report as Positive Quantifiable Value Figure 5. Procedure for quantitation of PCDDs and PCDFs in human adipose tissue. A-45 �14.2.3 Recovery of internal quantisation standards. Calculate the recovery of the labeled internal quantisation standards measured in the final extract-using the formula: QRS Internal Quant. Std. _ Eq. 14.3 Percent Recovery ~ ARS RRF 100 where AT<; = sum of the integrated ion abundances deteri:> mined for the labeled PCDD/PCDF internal quantisation standard in question; A Q RS sum of the integrated ion abundances determined for m/z 332 and m/z 334 of 13C121,2,3,4-TCDD (recovery standard); recovery standard, RS ~ amount (pg) of the added to the final 13 C12-1,2,3,4-TCDD extract; QIS = amount (pg) the labeled internal quantitation standard added to the sample prior to extraction; and RRF = relative response factor for the labeled internal quantitation standard in question relative to the internal recovery standard. This value shall be the RRF determined from the initial calibration. 14.3 Estimated Method Detection Limit Estimated where (1) sponse is sponse is method detection limits must be calculated in situations no response is noted for a specific congener; (2) a renoted but ion ratios are incorrect; and (3) where a requantitated as a trace value. 14.3.1 For samples in which no unlabeled PCDD or PCDF is detected, calculate the estimated minimum detectable concentration. The background area is determined by integrating the ion abundances for the characteristic ions in the appropriate region and relating the product area to an estimated concentration that would produce that product area. Use the formula: (2.5) Eq. 14-4 < A IS> A-46 (RRF) • (W) �where C^ = estimated concentration of unlabeled PCDD or PCDF required to produce AX; A = sum of integrated ion abundances or peak heights for the characteristic ions of the unlabeled PCDD or PCDF isomer in the same group of ^ 5 scans used to measure A,s; and AjS = sum of integrated ion abundances for the appropriate ions characteristic of the respective internal quantitation standard. Qjc, RRF, and W retain the definitions previously stated in Section 14.2. Alternatively, if peak height measurements are used for quantification, measure the estimated detection limit by the peak height of the noise in the 2,3,7,8-TCDD RT window. 14.3.2 For samples for which a response at the retention time of a specific PCDD or PCDF congener is noted, but the quantitative criteria for ion ratios is outside the acceptable range (Table 7), the estimated detection level is calculated as given in Eq. 14.3 except the values are qualified as not detected, ND, and the concentration is reported in parenthesis. 14.3.3 If a response for a specific PCDD or PCDF congener is qualified as a trace, TR, value (signal to noise is greater than or equal to 2.5 but less than 10) the analyst must also provide an estimated method detection limit. This is accomplished by using the observed signal to noise on either side of the response and calculating as given in Eq. 14-4. 15. REPORTING AND DOCUMENTATION All data should be reported on an individual sample basis using the data report format shown in Figure 6. The analyst is required to maintain all raw data, calculations, and control charts in a format as to allow a complete external data review. Suggested data formats for tracing calculationsare provided in Figure 7. A-47 �Pig* 1 oil U.S. ENVIRONMENTAL PflOTECTION AGENCY OFFICE OF TOXIC SUBSTANCES EXPOSURE EVALUATION DIVISION (TS-7M) WASHINGTON, DC 20460 NATIONAL HUMAN ADIPOSE TISSUE SURVEY ANALYSIS REPORT FORM EPA SAMPLE NUMBER. ANALYSIS DATE LAB NUMBER MS ANALYST BATCH NUMBER REPORT DATE _ REPORTED BY _ NATIVE COMPOUNDS CONCENTRATION (pg/gLl/ DATA INTERNAL QUANTITATION STANDARD QUALIFIER!/ 2.3,7.8-TCDD 1 ,,.!•!, 1 1 2.3,7,8-TCOF 1 . . ... |*l , 1 13 1,2.3,7,8-PeCDO 1 , , . 1*1 , r 1 3C12-2.3,7.8-TCDO Ci8-2,3,7.8-TCDF 3C12-1.2,3.7,8-P«COD 1,2.3,7,8-PeODF 1 . . . |*)l , 1 13 2.3,4,7.8-PeCDF 1 , . , Ul jlT l 1 <,2.3,4.7,8-HxCDO 1 ... 1 1,2.3,8,7.8-HxCDD 1 ...!•( i 1 1,2.3,7.8.9-HxCDO 1 , i , 1*1 . 1 '3C12-1,2,3,4,8.7,a-HpCDF 1,2.3,4.7.8-HxCDF 1 ...!•), 1 13 1,2.3,8,7,8-HxCDF 1 . . . Ul , 1 'SCu-OODF 1,2.3,7.8.9-HxCDF 1 . . . I.I . 1 2.3.4,8,7,8-HxCDF 1 , i , |«4 , 1 t.2,3,4,8,7,8-HpCDD 1 ... 1,2,3,4,8,7,8-HpCDF 1 , , , 1*1 , 1 1*1 , 1 C12-1.2,3.7,8-PeCDF 3Ct2-1.2.3.6.'.a-H»CDD 3C12-t.2,3,4.7.8-HxCDF '3012-1.2.3.4,8,7,8^0000 Cia-OCDO 1*1 , 1 1.2.3.4,7.8,9-HpCDF 1 , . . Ul OCOO I , . . !•( . 1 . 1 OCOF 1 . . . 1*1 . 1 REMARKS Jj Concantratlon reported la based on total extractable lipld (a). I NO - Not Detected, TR - Traoe. PQ - Positive Quantifiable. / Figure 6. Analysis report form. A-48 SPIKED LEVEL (P9) PERCENT (%) RECOVERY �RAW DATA SUMMARY FOR DETERMINATION OF 1.2.3.7,8-PeCDD IN HUMAN ADIPOSE TISSUE Sample no. 3 Sample weight (xx.xx g) Extractable lipid content (xx.x %) Analysis date Amount C12-PeCDD (pg) 13 13 C12-PeCDD m/z 332 3 > C12-PeCOD m/z 334 Ion ratio 366/368 1,2,3,7,8PeCDD m/z 354 Value reported as concentration in extractable lipid. Figure 7. Example of raw data summary format for the determination of 1,2,3,7,8-PeCDD in human adipose tissue. 1,2,3,7,8 m/2 356 Ion ratio 354/356 1,2,3,7,8PeCDD cone. (pg/g) � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 197 Folder The folder containing the original item. 5522 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 Stanley, John S. et al. 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 1986 Title A name given to the resource Analysis for Polychlorinated Dibenzo-p-Dioxins and Dibenzofurans in Human Adipose Tissue: Method Evaluation Study: Draft Final Report with attached letter transmitting the report to Alvin L. Young, from Janet C. Remmers, Field Studies Branch, Exposure Eva ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/0bb255590ea9525829127e3b5b01688b.pdf 7ed354cf95d157f4f41304e505faca83 PDF Text Text Item D Number °5521 Author Stanley, John S. Corporate Author Midwest Research Institute Report/Article Title Protocol for the Analysis of 2,3,7,8-Tetrachlorodibenzop-Dioxin by High-Resolution Gas Chromatography/HighResolution Mass Spectrometry with attached letter transmitting the report to Alvin L. Young, from Ronald K. Mitchum, Director, Quality Assurance Division, United States Environmental Protection Agency D Not Scanned Journal/Book Title Year 1986 Month/Day January Color D Number of Images 157 Descripton Notes EPA 600/4-86-004 Tuesday, March 19, 2002 Page 5521 of 5611 �i UNITED STATES ENVIRONMENTAL PROTECTION AGENCY OFFICE OF RESEARCH AND DEVELOPMENT ENVIRONMENTAL MONITORING SYSTEMS LABORATORY-LAS VEGAS P.O. BOX 15027, LAS VEGAS, NEVADA 89114-5027-702/798-2100 (FTS 545-2100) MAR 6 1986 Dr. Alvin L, Young President, Scientific Advisory Office Office of Science and Technology Policy, R. 5005 New Executive Office Building Washington, DC 20506 Dear Dr. Young: We have recently completed the preparation and evaluation of a highresolution protocol for TCDD determination. Because of your interest in this field (vide your participation in the September 1985 Bayreuth Symposium), I am sending you a copy of our report "Protocol for the Analysis of 2,3,7,8Tetrachlorodibenzo-p-Dioxin by High-Resolution Gas Chromatography/HighResolution Mass Spectrometry" for your use. This report provides the results of the single-laboratory evaluation of a high-resolution gas chromatography/ high-resolution mass spectrometry method for the determination of 2,3,7,8tetrachlorodibenzo-p-dioxin and total tetrachlorodibenzo-p-dioxins at concentrations ranging from 10 to 200 pg/g (ppt) in soils and 100 to 2,000 pg/L (ppq) in water. Based on the data generated during this study and based on discussions at our laboratory, we revised certain parts of the protocol to lower the quantitation level for 2,3,7,8-tetrachlorodibenzo-p-dioxin to 2 ppt in soil and 20 ppq in water samples. The revised protocol is included in this report as Appendix B. Please contact me or Dr. Werner Beckert (who was the Project Officer for this task) if you have any questions or comments. Sincerely, Director Quality Assurance Division Enclosure �United States Environmental Protection Agency Environmental Monitoring Systems Laboratory P.O. Box 15027 Las Vegas NV 89114-5027 EPA 600/4-86-004 January 1986 Research and Development Protocol for the Analysis of 2,3,7,8Tetrachlorodibenzo-pby High-Resolution Gas Chromatography/ High-Resolution Mass Spectrometry �PROTOCOL FOR THE ANALYSIS OF 2,3,7,8-TETRACHLORODIBENZO-£-DIOXIN BY HIGH-RESOLUTION GAS CHROMATOGRAPHY/HIGH-RESOLUTION MASS SPECTROMETRY by John S. Stanley and Thomas M. Sack Midwest Research Institute Kansas City, Missouri 64110 Contract Number SAS 1576X Project Officer Werner F. Beckert Quality Assurance Division Environmental Monitoring Systems Laboratory Las Vegas, Nevada 89114 ENVIRONMENTAL MONITORING SYSTEMS LABORATORY Office of Research and Development U.S. Environmental Protection Agency Las Vegas, Nevada 89114 �NOTICE The information in this document has been funded wholly or in part by the United States Environmental Protection Agency under Contract Number SAS 1576X to the Midwest Research Institute, Kansas City, Missouri. It has been subject to the Agency's peer and administrative review, and it has been approved for publication as an Environmental Protection Agency document. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. �PREFACE This report describes the activities completed as part of a singlelaboratory evaluation of a high-resolution gas chromatography/highresolution mass spectrometry method for the determination of tetrachlorodibenzo-p_-dioxins in water, soil, and sediment samples. The work described in this report was completed at the Midwest Research Institute under contract to Viar and Company (Special Analytical Services SAS 1576X) for the U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory, Quality Assurance Division, Las Vegas, Nevada. The revision of the protocol to allow for lower quantitation limits for tetrachlorodibenzo£-dioxins was carried out at the Environmental Monitoring Systems LaboratoryLas Vegas. This report was prepared with assistance from M. McGrath. The authors acknowledge the technical project monitor, W. F. Beckert, as well as R. K. Mitchura and S. Billets of the Environmental Monitoring Systems LaboratoryLas Vegas and, especially, Y. Tondeur of the Environmental Research Center, University of Nevada, Las Vegas for guidance provided during this study. 111 �ABSTRACT This report provides the results of the single-laboratory evaluation of a high-resolution gas chromatography/high-resolution mass spectrometry method for the determination of 2,3,7,8-tetrachlorodibenzo-£-dioxin and total tetrachlorodibenzo-£-dioxins at concentrations ranging from 10 to 200 pg/g (ppt) in soils and 100 to 2,000 pg/L (ppq) in water. The report summarizes the data for the precision and accuracy of triplicate measurements of five solid and five aqueous samples. The results indicate that the method is capable of generating accurate and precise data within the concentration limits specified above and within absolute recoveries of 40 to 120 percent with 50 percent precision. An attempt to reach a quantitation limit for TCDD of 2 ppt (or less) for soil and 20 ppq (or less) for aqueous samples was not successful. Based on the data generated during this study and based on discussions at the Environmental Monitoring Systems LaboratoryLas Vegas, the Environmental Monitoring Systems Laboratory-Las Vegas revised certain parts of the protocol to lower the quantitation limit for tetrachlorodibenzo-p_-dioxins to 2 ppt in soil and 20 ppq in water samples. IV �CONTENTS Preface Abstract Figures Tables iii iv vi vii 1. 2. 3. 4. 1 3 5 7 7 7 9 9 11 13 13 14 14 22 Introduction Conclusions Recommendations Experimental Procedures Sample description Sample preparation Reagents HRGC/HRMS instrumentation Mass measurement accuracy Chromatographic resolution Injection technique 5. Results and Discussion Approach to cleanup column evaluation Final method evaluation References Appendices A. Validated Analytical Protocol B. Proposed Analytical Protocol 42 43 �FIGURES Number Page 1 Column cleanup procedures specified in the protocol 15 2 Column cleanup procedures proposed by the EMSL-LV 16 3 Background levels of 1,3,6,8- and 1,3,7,9-TCDD observed over the single-laboratory evaluation study 41 VI �TABLES Table Page 1 Solid Samples Used for HRGC/HRMS Method Evaluation 8 2 Aqueous Samples Used for HRGC/HRMS Method Evaluation 8 3 TCDD Isomers Used for HRGC/HRMS Method Evaluation 10 4 Composition of Concentration Calibration Solutions (pg/pL) . . 10 5 HRGC/HRMS Operating Conditions 12 6 Recovery ( ) of Several TCDD Isomers from Cleanup Option A . . % 18 7 Recovery ( ) of Several TCDD Isomers from Cleanup Option B . . % 19 8 Recovery ( ) of Several TCDD Isomers from Cleanup Option C . . % 20 9 Recovery ( ) of Several TCDD Isomers from Cleanup Option D . . % 21 10 Initial Calibration Summary 23 11 HRGC and Mass Resolution Check Summary 24 12 TCDD Data Report Form 26 13 Accuracy and Precision of the HRGC/HRMS Analysis for 2,3,7,8-TCDD from Laboratory Aqueous Matrix Spikes . . . . . 33 Precision of the HRGC/HRMS Analysis for 2,3,7,8-TCDD of Soil and Fly Ash Samples 34 Accuracy of the HRGC/HRMS Method for the Determination of TCDD Isomers Spiked into Aqueous Matrices 35 Accuracy of the HRGC/HRMS Method for the Determination of TCDD Isomers Spiked into Soil Matrices 36 Fortified Field Blank Results 37 14 15 16 17 vii �SECTION 1 INTRODUCTION The U.S. Environmental Protection Agency's (EPA) strategy for dealing with dioxin requires the development and validation of an analytical method capable of achieving detection of the tetrachlorodibenzo-p_-dioxins (TCDD), specifically 2,3,7,8-TCDD, at the parts-per-trillion (ppt) level in soil and sediment and parts-per-quadrillion (ppq) level in water.1 This validated method will be used by qualified contract laboratories to extend the analytical capabilities for such analyses to all EPA regional and program offices. This report deals specifically with the single-laboratory evaluation of a high-resolution gas chromatography/higb-resolution mass spectrometry (HRGC/HRMS) analysis method for TCDDs in soil, sediment, and water. The method (Appendix A) is intended to provide quantitative determination of TCDD at levels of 10 to 200 pg/g (soil and sediment) and 100 to 2,000 pg/L (water) at a mass resolution of 10,000. This single-laboratory evaluation has been completed as part of the validation process recommended by EPA.2 The proposed method was prepared after several candidate methods were reviewed and their best features were selected. After peer review, the proposed method was refined for completeness, technical accuracy, clarity, and regulatory applicability. The single-laboratory evaluation of the proposed analytical method has been accomplished through three tasks. The first task involved preliminary performance testing of the method using TCDDcontaminated soils and TCDD-spiked aqueous samples. The results of this study indicated that the proposed method required modification to achieve the target method detection limits and the accuracy and precision criteria. The second task focused on ruggedness testing of the chroraatographic cleanup procedures. The results of this study were used to modify the proposed method. This report is focused on the results of the triplicate analysis of five solid and five aqueous samples completed under the third task of the evaluation, using the modified method. Section 2 of this report summarizes the conclusions based on the single-laboratory evaluation of this method using TCDD-contaminated soils and TCDD-spiked aqueous samples. Section 3 presents recommendations that should be considered for inclusion in the method before proceeding with collaborative testing. Section 4 presents some specific experimental conditions, and Section 5 summarizes the analytical data for the triplicate analysis of four soil, one fly ash, and five aqueous samples completed in the third task of the single-laboratory evaluation. Triplicate analyses of a l-pg/pL calibration solution did not give satisfactory results. In order to achieve a quantitation limit of 2 ppt for soil (using a 10-g sample) and 1 �20 ppq for water (using a 2.0-L sample), the protocol evaluated in this study was modified. The rationale for the modifications and the revised protocol are included as Appendix B. �SECTION 2 CONCLUSIONS The single-laboratory evaluation of the analytical method for the determination of 2,3,7,8-TCDD in soil and aqueous samples demonstrates that the method as described is capable of achieving the target detection limits of 10 pg/g (ppt) for soils and 100 pg/L (ppq) for water. The relative response factors (RRF) determined for native 2,3,7,8-TCDD versus the internal standard 13C12-2,3,7,8-TCDD, and the RRP of the internal standard versus the recovery standard 13Ci2-l,2,3,4-TCDD over the five-point concentration calibration curve demonstrate that the HRGC/HRMS method maintains a linear response for 2,3,7,8-TCDD from 10 to 200 ppt for soils and 100 to 2,000 ppq for water. The results of the analysis of spiked aqueous samples demonstrate that internal standard (isotope dilution) quantitation provides an accurate measurement of 2,3,7,8-TCDD. The accuracy of the 2,3,7,8-TCDD measurement for triplicate analysis of four water samples spiked at various concentrations was quite good. The accuracy of measurement for 2,3,7,8-TCDD averaged 104 percent for three aqueous matrices prepared as laboratory matrix spikes. The absolute recovery of the internal standard 13Ci2"2,3,7,8-TCDD did not significantly affect the accuracy of the 2,3,7,8-TCDD determination. The precision of the analyses for 2,3,7,8-TCDD ranged from 3.6 to 16 percent for replicate analyses of the five aqueous samples. The precision of the triplicate analyses of the soil samples was somewhat higher than determined for aqueous samples. The precision of triplicate analyses of the four soil samples ranged from 19 to 50 percent. The difference in precision from that of the aqueous samples may be attributable to the potential for TCDD adsorption on the soil samples. The results from the analyses of soil and aqueous samples spiked with additional TCDD isomers demonstrate that the internal standard quantitation gives good estimates of total TCDD values. The accuracy of the analyses of fortified distilled water and influent and effluent wastewaters averaged 101 ± 14 percent for five TCDD isomers. The accuracy of the measurements of these isomers for the four fortified soil samples averaged 87 ± 24 percent. The results of the analyses demonstrate that the requirements for absolute recovery of the internal standard (40 to 120 percent) and precision of replicate analyses (RPD < 50 percent) can be achieved for relatively clean samples. �The sample matrix can severely impact the performance of the analytical method. This is evidenced by the consistent low recovery of the internal standard from the fifth aqueous sample, an industrial wastewater, and from a fly ash sample. The low recovery from the industrial wastewater is possibly due to the effect of coextractants on the elution sequence from alumina. The low recoveries observed for the fly ash sample, on the other hand, may be attributed to adsorption by the sample matrix. One of the most critical variables in the analytical method is the completeness of removal of the benzene from the extract before proceeding with the acidic alumina column fractionation. The cleanup column ruggedness testing experiments demonstrated that the recoveries of 2,3,7,8-TCDD and the other TCDDs are affected by the presence of benzene in the alumina column fractionation step. The analyst must be aware of the potential problem of interferences arising from background contamination. For example, the 1,3,6,8- and 1,3,7,9-TCDD isomers were present in the fortified field blanks in this work. From other referenced activities it becomes clear that these isomers may present problems in other laboratories as well. The fortified field blanks are important tools in assessing the background contamination problems over time. Although the 1.0-pg/|JL standard did not yiej.d satisfactory results in this study, due to unacceptable ion ratios, the response factors are within the established curve. The data for the triplicate analyses of the LO-pg/pL standard demonstrate that the characteristic ions for TCDD were greater than 20:1 for the m/z 322 S/N and approximately 10:1 for m/z 259 S/N. Thus, it should be possible to extend the detection limit to 1 pg/pL if an allowance for abundance ratios based on ion statistical errors is incorporated. Based on the column performance and bleed characteristics, the column of choice for the analysis for TCDD at ppt (for soils and sediments) and ppq (for water) levels appears to be the 50-m CP-Sil 88 with a 0.2-(Jm film thickness. To preserve the performance characteristics of the HRGC columns, an injection technique that excludes any air is highly recommended. �SECTION 3 RECOMMENDATIONS 1. Mass measurement accuracy should properly be determined relative to the lock mass (if any), rather than m/z 254.9856, because it is that relationship which will determine how accurately the masses of the TCDD ions will be measured. 2. It is recommended that the chromatographic resolution check be performed on the summed ion chromatograms of m/z 259 + m/z 320 + m/z 322. This yields a chromatogram which is less noisy and more representative of the true column performance. 3. The 5 percent peak width criterion for mass resolution should be the selected mass/1,000 mmu rather than 31.9 mmu because the protocol allows peaks other than m/z 319 to be used for resolution measurement (e.g., 31.7 mmu if m/z 317 is used). 4. It is recommended that the mass measurement accuracy be recorded and reported along with the resolution check summary table. 5. The addition of the recovery standard 13Ci2-l,2,3,4-TCDD should be achieved by using a spike volume of 25 to 50 (JL rather than 5 (JL to minimize errors resulting from volume measurement. 6. The recommended temperature program settings in the method should be converted to those presented in the experimental section of this report. These conditions were established for analysis with tridecane as the solvent. 7. Lower limits of detection can be achieved by allowing the analyst to concentrate the final extract to as low as 10 |Jl. It may be necessary to use the smaller final volume with other HRMS instruments to achieve the same levels of detection. 8. The method should recommend several techniques to break up emulsions resulting from extraction of aqueous samples. In this evaluation the emulsion phase was put through a column packed with glass wool, which was then rinsed with additional methylene chloride. Other options might include stirring or centrifugation of the emulsion phase. 9. The method should specify the procedure to deal with aqueous samples containing high levels of suspended solids. In this study it was �necessary to centrifuge the soil extract sample before proceeding with the extraction. 10. It is highly recommended that the method be modified such that the benzene extract is completely exchanged to hexane prior to cleanup on the silica column since this is apparently one of the most critical factors leading to successful sample analysis. 11. It may be worthwhile to evaluate a cleanup procedure in which the charcoal column precedes the alumina column as a means to improve method recovery. �SECTION 4 EXPERIMENTAL PROCEDURES SAMPLE DESCRIPTION Five solid samples were provided by the Environmental Monitoring Systems Laboratory-Las Vegas (EMSL-LV) to the Midwest Research Institute (MRI) for analysis for 2,3,7,8-TCDD and total TCDD using the analytical method in Appendix A. A description of the five solid samples and the estimated 2,3,7,8-TCDD concentrations from previous analyses by an independent laboratory are provided in Table 1. Each sample was analyzed in triplicate as specified in the protocol. One of the triplicate samples for each soil sample was spiked with the seven TCDD isomers (1,3,6,8-; 1,3,7,9-; 1,2,3,7-; 1,2,3,8-; 1,2,3,4-; 1,2,7,8-; and 1,2,8,9-TCDD) at approximately 10 times the estimated level of 2,3,7,8-TCDD specified in Table 1. Five aqueous samples were generated for the evaluation of the analytical method at the ppq detection level. Table 2 presents a description of each water type and lists the fortification levels of 2,3,7,8-TCDD and seven additional TCDD isomers (1,3,6,8-; 1,3,7,9; 1,2,3,7-; 1,2,3,8-; 1,2,3,4-; 1,2,7,8-; and 1,2,8,9-TCDD) in each sample. The influent and effluent wastewater samples were collected from a sewage treatment facility in metropolitan Kansas City, Missouri. The industrial wastewater was obtained from a holding pond within a hazardous waste area that was known to be highly contaminated with PCBs and possibly other chlorinated aromatic compounds (chlorobenzenes). This aqueous sample was very acidic (pH < 1) and was dark in color. The soil extract was prepared from 30 g of a soil sample, Hyde Park 002 (H2), and 1 gallon of distilled water. The mixture was stirred constantly (at least 24 hrs) until just prior to subsampling of 1.0-L aliquots. SAMPLE PREPARATION All samples listed in Tables 1 and 2 were extracted and analyzed in triplicate according to the protocol provided in Appendix A. As indicated in Tables 1 and 2, one aliquot of each sample matrix was fortified with additional TCDD isomers, which represent the compounds that elute first (1,3,6,8-TCDD), last (1,2,8,9-TCDD), and within the approximate retention window of 2,3,7,8-TCDD (1,2,3,7-; 1,2,3,8-; and 1,2,3,4-TCDD) from the HRGC columns used for sample analysis. �TABLE 1. SOLID SAMPLES USED FOR HRGC/HRMS METHOD EVALUATION Approximate sample size3 Estimated 2,3,7,8-TCDD concentration (ppt) Spike level (ppt) of TCDD isomers EPA sample no. Matrix B25-Piazza Road (B5) Soil 10 g 50 100 Hyde Park 001 (HI) Soil 10 g 70 140 B52-Shenandoah (Bl) Soil 1 g 360 720 Hyde Park 003 (H3) Soil 1 g 1,700 1,700 RRAI-5,7,8 (FA) Fly ash 10 g NRd e .Approximate sample size of each replicate sample. Estimated level of endogenous 2,3,7,8-TCDD reported to MRI by W. Beckert in letters dated April 19, 1985 and August 30, 1985. .Approximate fortification level of each of seven additional TCDD isomers. No estimate of 2,3,7,8-TCDD concentration was reported. Additional TCDD isomers were not spiked into this matrix. TABLE 2. AQUEOUS SAMPLES USED FOR HRGC/HR.MS METHOD EVALUATION Sample type Approximate sample size3 Fortification level of 2,3,7,8-TCDD (ppq) Fortification level of TCDD isomers (ppq) Distilled water (DW) 1.0 L 250 500 POTW influent (IWW) 1.0 L 500 1,000 POTW effluent (EWW) 1.0 L 1,000 2,000 Industrial wastewater (IND) 1.0 L 500 Hyde Park 002; soil extract (H2W) 1.0 L c 1,000 c .Approximate sample size of each replicate sample. Approximate fortification level of each of seven additional TCDD isomers. This aqueous sample was not fortified with TCDD isomers. �All samples were fortified with 500 pg 13Cl2-2,3,7,8-TCDD in 1.5 ml acetone. The solid samples were extracted continuously for 24 hr in a Soxhlet apparatus with benzene and the 1.0-L aqueous samples were batchextracted using 2.0-L separatory funnels and three 60-mL portions of methylene chloride. The extractions of the influent wastewater (IWW) and effluent wastewater (EWW) and the soil extract (H2W) were complicated by the formation of emulsions. In each case, the emulsion was removed by passing the methylene chloride and emulsion layer through a column of glass wool prerinsed with methylene chloride. The extract and resulting aqueous layers were collected in a sample bottle and the glass wool plug was rinsed with an additional 10 ml methylene chloride. Following the complete extraction of the aqueous sample, the contents of the bottle were transferred to a clean 250-raL separatory funnel and the methylene chloride was removed from the aqueous phase that was transferred with the emulsion. All extracts were concentrated with Kuderna-Danish evaporators and nitrogen evaporation to approximately 1.0 ml. Each extract was taken through the entire cleanup procedure including the acidic silica, acidic alumina, and Carbopak C as specified in the protocol (Appendix A). The HRGC/HRMS analysis of each extract was completed as specified below. REAGENTS All solvents for extraction and cleanup were obtained as "Burdick and Jackson distilled-in-glass" quality. The tridecane (99 percent purity) was obtained from Aldrich (TS, 740-1). The chromatographic materials, acidic alumina (100-200 mesh AG-4, Biorad Laboratories 132-1340), silica (70-230 mesh Kieselgel 60, EM Reagent, American Scientific Products C5475-2), sodium sulfate, potassium carbonate, Celite 545® (Fisher Scientific Company), and the silanized glass wool and Carbopak C (80-100 mesh Supelco 1-1025) were prepared for use as specified in Section 7 of the protocol (Appendix A). Table 3 provides the sources of standards used to prepare the calibration solutions, sample fortification solutions, recovery standard spiking solution, internal standard spiking solutions, field fortification solutions, and TCDD isomer fortification solutions. Table 4 is a summary of the concentration calibration standards prepared for the HRGC/HRMS method evaluation. These standards were prepared as specified in the protocol (Appendix A). The standard HRCC6 was included in the final evaluation of the HRGC/HRMS method as a means to demonstrate the lower limit of detection under optimum instrumental conditions. HRGC/HRMS INSTRUMENTATION Sample extracts and calibration standards were analyzed using a Carlo Erba Mega Series gas chromatograph (GC) which was coupled to a Kratos MS50 TC double-focusing mass spectrometer (MS). The GC/MS interface was simply a direct connection of the GC column to the ion source via a heated interface oven. A Finnigan 2300 Incos data system was used for data acquisition and processing. �TABLE 3. TCDD ISOMERS USED FOR HRGC/HRMS METHOD EVALUATION Stock concentration Isomer 2,3,7,8-TCDD C12-2,3,7,8-TCDD 50 ± 5 1,2,3,4-TCDD 2.7 mg/mL l3 Ci2-l,2,3,4-TCDD 50 ± 5 Mg/mL 1,3,6,8-/1,3,7,9-TCDD 0.82 mg/mL 1,2,3,7-/1,2,3,8-TCDD 0.5 mg/mL 1,2,7,8-TCDD 0.39 mg/mL 1,2,8,9-TCDD 1.46 mg/mL Column performance standard Standard code EPA QA Reference Materials Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories Cambridge Isotope Laboratories 7.87 ± 0.79 13 Source 10 20603 R00201 (Lot AWN-1203-65) ED-915C (Lot 6578) R00212 (Lot AWN-1203-93) ED-913C (Lot F2086) ED-905C (Lot 7371) ED-915C (Lot 7184) ED-916C (Lot MLB-682-26) ED-908 (Lot No. R00215) Mixture of TCDD isomers including 2,3,7,8-; 1,2,3,4-; 1,2,3,7-/1,2,3,8-; 1,2,7,8-; and 1,4,7,8-TCDD. TABLE 4. COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS (pg/ML) Recovery standard 13 C12-1,2,3,4-TCDD HRCC1 HRCC2 HRCC3 HRCC4 HRCC5 HRCC68 Analyte 2,3,7,8-TCDD Internal standard 13 C12-2,3,7,8-TCDD 2.5 5.0 10.0 20.0 40.0 1.0 2.5 5.0 10.0 20.0 40.0 1.0 10.0 10.0 10.0 10.0 10.0 10.0 This solution is not specified in the analytical method in Appendix A. 10 �The HRGC/HRMS operating conditions used in the final phase of this work are summarized in Table 5. The GC operating conditions recommended in the protocol were not used for these analyses for three reasons. First, the TCDDs have rather long retention times, and the solvent (tridecane) boils at 235°C. Thus no benefit could be realized with a low initial temperature. Second, past experience at MR I has indicated that 200°C is an acceptable starting temperature for these types of analyses when tridecane is used as a solvent. Finally, since the CP-Sil 88 and SP-2330 phases are both very polar and thinly coated, it has been recommended that they not be subjected to rapid heating or cryogenic cooling to prevent thermal shock to the column.3 The MS was tuned daily to yield a resolution of at least 10,000 (10 percent valley) and optimal response at m/z 254.986. This step was followed by calibration of an accelerating voltage scan beginning at m/z 254 (typical calibration range was 255 to 605 amu). Other voltage scans from the same data file were then used to establish and document both the resolution at m/z 316.983 and the mass measurement accuracy at ra/z 330.979. MASS MEASUREMENT ACCURACY For this work, mass measurement accuracy was measured relative to PFK m/z 254.986, as required by the protocol, by applying the mass correction, Am, to the entire spectrum, which yields an error of 0 ppm at m/z 254.986. In this way, it was possible to meet routinely the 5 ppm accuracy criterion at m/z 330.979. However, if a lock mass other than 254.986 is used, the mass measurement accuracy should be measured relative to that lock mass, since it is that peak which is used to maintain magnet alignment and will ultimately control the mass measurements during the selected ion monitoring (SIM) experiments. Mass Resolution Mass resolution at m/z 316.983 was documented by an output of the Incos PROF program. However, the computer-generated value for resolution was found to be significantly higher than the value measured manually. Thus, the manually determined resolution, which was nearly identical to the value measured by using the peak matching unit, is reported. Closer inspection of the PROF source code revealed that resolution is computed via a statistical method, not as m/Am at 5 percent height. Incos users should therefore be aware of this discrepancy, because the computer-generated value can be as much as 20 percent over the proper value. Following calibration, the SIM experiment descriptor was updated to reflect the new calibration. Six masses (see Table 5) were monitored by scanning *v m/10,000 amu over each mass. The total cycle time was kept to 1 sec. The m/z 280.983 ion from PFK was used as a lock mass because it is the most abundant PFK ion within the range of m/z 255 to 334 and therefore permits the use of low partial pressures of PFK, which minimizes PFK interferences at the analytical masses. 11 �TABLE 5. HRGC/HRMS OPERATING CONDITIONS Mass spectrometer Accelerating voltage: Trap current: Electron energy: Electron multiplier voltage: Source temperature: Resolution: 8,000 V 500 MA 70 eV 2,000 V 280°C 10,000 (10% valley definition) Ions monitored Nominal dwell times (sec) 258.930 319.897 321.894 331.937 333.934 280.9825 (lock mass) 0.15 0.15 0.15 0.15 0.15 0.10 Overall SIM cycle time = 1 sec Gas chromatograph Column coating: Film thickness: Column dimensions: CP-Sil 88 0.2 (Jin 50 m x 0.22 mm ID Helium linear velocity: Helium head pressure: ~ 25 cm/sec 1.75 kg/cm2 (25 psi) Injection type: Split flow: Purge flow: Injector temperature: Interface temperature: Injection size: Initial temperature: Initial time: Temperature program: Splitless, 45 sec 30 ml/min 6 ml/rain 270°C 240°C 2 Ml 200°C 1 min 200°C to 240°C at 4°C/min 12 �CHROMATOGRAPHIC RESOLUTION « Chromatographic resolution values were measured for the SIM plot of m/z 320. However, it may be advantageous to measure cbromatographic resolution from a plot of the sum of m'z 259, 320, and 322. The sum trace has better signal-to-noise ratio (S/N) and peak definition than the SIM plots, which permits a more accurate measurement of resolution. Selection of the HRGC Column Three different HRGC columns were evaluated in the course of this project: SP-2330 (60 m x 0.24 mm); DBS (60 m x 0.22 mm); and CP-Sil 88 (50 m x 0.22 mm). By evaluating the mass spectra of the bleed from each column at 240 to 250°C, it became apparent that the column background may be the limiting factor in achieving the desired detection limit for this method. The DBS column provided the least amount of background at 250°C, and the SP-2330 had the worst. This coincides with the fact that quantitation at the detection limit (i.e., 2.5 pg/pL) with the SP-2330 column was difficult at best. The CP-Sil 88 column appeared to offer less bleed than the SP-2330 column and indeed does permit more accurate quantitation due to reduced background contribution. The Chromatographic performance afforded by these columns is a further issue, since the column best suited for low detection limits, DB-5, cannot meet the 25 percent valley Chromatographic resolution criteria in all cases. Both the SP-2330 and CP-Sil 88 columns can easily resolve the 2,3,7,8-TCDD. However, based on the bleed considerations discussed above, the 50-m CP-Sil 88 column is recommended for the best combination of low bleed and good isomer separation. It may also be advisable that other HRGC columns (including SP-2340, Silar IOC, and SP-2331) that have been used for 2,3,7,8-TCDD analysis at the 1-ppb soil level be evaluated for background contribution and their application for HRMS analysis at ppt and ppq concentrations. INJECTION TECHNIQUE The HRGC column performance can degrade very quickly if proper injection techniques are not used. Specifically, the SP-2330 and CP-Sil 88 phases are very sensitive to QZ and will decompose rapidly at 200°C if any trace of 02 is present. Therefore, the common practice of using 1 (JL of air to flush the syringe and effect reproducible injections is to be avoided, since even that small amount of air per injection can cause column performance to degrade in less than one week of continued use. The following injection technique is recommended. First rinse the syringe copiously with isooctane (or other volatile solvent, such as toluene). Dry the syringe by drawing air through it. Pull up and expel several volumes of tridecane until all bubbles are gone, and leave 1 pL of tridecane in the barrel. Finally, pull up 2 pL of the sample solution and inject. This technique has worked very well and yields injection reproducibility comparable to that of the air purge method, without introducing air onto the analytical GC column. 13 �SECTION 5 RESULTS AND DISCUSSION The primary purpose of any method validation process is to assure that the method under consideration is adequate to meet testing and monitoring requirements.1 The single-laboratory evaluation of the analytical protocol presented in this report has been preceded by several evaluation and improvement steps. These have included the preparation of a written protocol, technical review of the protocol for completeness, technical accuracy, and clarity; preliminary testing to evaluate performance of the analytical method; and revision and refinement of the written protocol based on the results of the preliminary testing. Prior to the assessment of the refined protocol presented in Appendix A, the proposed analytical method had been evaluated for performance through the analysis of several duplicate samples. The results of the preliminary evaluation indicated that problems existed in the design and approach to the extract cleanup steps, which greatly affected the method detection limit, accuracy, and precision. This section presents a summary of the studies that have led to the refinement of the analytical protocol as provided in Appendix A and also summarizes the single-laboratory evaluation of this protocol. APPROACH TO CLEANUP COLUMN EVALUATION The initial method evaluation completed under the first task resulted in very low recoveries of the internal standard, 13Ci2-2,3,7,8-TCDD, and the accuracy and precision of duplicate sample analyses were poor. After reviewing the data, it was apparent that the problems were the result of poor chromatographic separation in the cleanup columns. The initial protocol involved reducing sample extract volumes to 1.0 ml in benzene, elution through the acidic silica column with hexane, and collection of the total eluent which was then added to the acidic alumina column. The alumina column was further eluted with hexane/20-percent methylene chloride. The eluate was concentrated and cleaned further using a Carbopak C/Celite column, and the TCDDs were eluted with 2 mL toluene. Column cleanup techniques were revised and further evaluated following the procedures depicted in Figures 1 and 2. The column evaluations were completed with triplicate measurements at three spike levels (0.10, 1.0, and 10 ng) equivalent to 10, 100, and 1,000 ppt of TCDD in solids with several TCDD isomers (2,3,7,8-; 1,3,6,8-; 1,3,7,9-; 1,2,3,4-; 1,4,7,8-; 1,2,3,7-; 1,2,3,8-; and 1,2,8,9-TCDD). 14 �OPTION A OPTION B t mL Benzene Extract 1 mL Benzene Extract 1 H2SO4 - Si02 H2SO4 - SiO2 4.0g 4.0g Si O2 l.Og SiO2 l.Og Acidic A! 203 Acidic AI2O3 6.0g 6.09 30ml 20% CH2Cl2/Hexane Concentrate to 30ml 20% CH2Cl2Aiexane Carbopak C/Celite I 6ml Toluene Carfaopak C/Celite HRGC/HRMS 6mL Toluene HRGC/HRMS Figure 1. Column cleanup procedures specified in the protocol 15 �OPTION C OPTION D 1 ml Bcnztrw Extract 1 mL S«nz«n* Extract H2SO4 - Si02 4.0g H2S04 " SIO2 4.0g SiO2 l.Og SiO2 l.Og < > Concentrate to 0.5 ml Concentrate to 0.5mL Acidic AI2O3 6.0g Acidic AI2O3 6.0g I 30ml. 20% CH2d2/H«xan« Conetnfratt to 30mL 20% CH2Cl2/Htxone Carbopak C/Ctlitt I ami Tolu«ne Carbopok C/C«iite HRGC/HRMS 6mL Tolu«n« HRGC/HRMS Figure 2. Column cleanup procedures proposed by the EMSL-LV. 16 �The TCDD isomers were added to 1-mL portions of benzene and were taken through the four sample cleanup sequences depicted in Figures 1 and 2. One of the replicates for each procedure was also spiked with 100 ng of Aroclor 1260. The results of the sample analyses are provided in Tables 6 through 9. As noted in Tables 6 and 7, recoveries of the TCDD isomers were low and quite variable for the early eluting isomers 1,3,6,8- and 1,3,7,9-TCDD as compared to 1,2,8,9-TCDD. Recovery of 1,2,8,9-TCDD was still low and variable (approximately 60 percent recovery with an RSD of ^ 20 percent). These results were generated using the procedures specified in the original protocol (see Figure 1). The results of the analyses following the cleanup options A and B demonstrate that accurate quantitation of all TCDD isomers is not possible using only the 13C12-2,3,7,8-TCDD surrogate standard. The low recoveries measured for options A and B are obviously a result of the presence of benzene in the eluent from the acid-modified silica column that is taken directly through the acidic alumina column. In contrast, options C and D (Tables 8 and 9) demonstrate quantitative recovery of the TCDD isomers. Some background contamination has been noted from the acidic alumina for the 1,3,6,8- and 1,3,7,9-TCDD isomers. This material had previously been prepared by Soxhlet extraction with methylene chloride and activation at 190°C prior to use. As noted in Tables 8 and 9, the average recovery of the other spiked TCDD isomers was greater than 84 percent. When the recoveries of the different isomers and the 13C12-2,3,7,8-TCDD are compared, the average relative percent difference ranges from 1 percent for 2,3,7,8-TCDD (Table 3) to 24 percent for 1,2,3,4-TCDD (Table 4). These results demonstrate that either of these cleanup procedures (options C and D) will provide good recovery and reliable quantitation of 2,3,7,8-TCDD and very good estimates of the concentrations of the other TCDD isomers present in the samples. No interferences were observed in the samples spiked with 100 ng Aroclor 1260. The lack of PCB interferences was especially noted in the extracts of samples spiked at 0.10 ng/TCDD isomer. In addition to the evaluations of the cleanup procedures presented above, the acid-modified silica gel/acidic alumina columns and the Carbopak C/Celite column were evaluated separately. Evaluation of the silica/alumina at the 0.10-ng spike level as shown in Figure 2 resulted in an average recovery of 120 percent for 1,2,3,4-, 1,2,3,7-, 1,2,3,8-, and 1,4,7,8-TCDD; 114 percent for 2,3,7,8-TCDD; 118 percent for 13C12-2,3,7,8-TCDD; and 118 percent for 1,2,8,9-TCDD. The results for the recovery of 1,3,6,8- and 1,3,7,9-TCDD indicated that some contamination originated from the acidic alumina. Replicate analyses of the Carbopak C/Celite column at the 0.10-ng spike level resulted in average recoveries of 97 percent for 1,3,6,8-TCDD; 88 percent for 1,3,7,9-TCDD; 81 percent for 1,2,3,4-, 1,2,3,7-, 1,2,3,8-, and 1,4,7,8-TCDD; 75 percent for 2,3,7,8-TCDD; 96 percent for 13C12-2,3,7,8-TCDD; and 90 percent for 1,2,8,9-TCDD. Elution of the Carbopak C/Celite column with additional toluene beyond 6 ml did not improve recoveries even for the samples spiked at 10 ng/TCDD isomer. 17 �TABLE 6. KECOVKKY ( ) OF SEVERAL TCIN) ISOHKRS FROM CI.EANIII' OPTION A % Spike level Lt lM U W U M d | HjfOt - UOj 4.0, SI02 10 . , , 1 AcMkAljOj 130.110% CHjCljAfoMn* 1.3,6,8 1,3,7,9 Recovery (%) ol TCDO iaomer >,2, 3,4/ 1,2,3, 11 1,2,3,8 2,3,7,8 1,4,7,8 I3 Cl2-2,3.7,8 1,2,8,9 21 14 19 27 38 40 33 39 48 64 28 17 23 31 40 19 41 31 44 38 76 6.2 12 29 20 31 36 61 10 ng 6.3 12 31 20 29 34 59 Nean 7.4 14 32 23 31 36 56 36 24 34 31 20 26 1 ng 1 ug ing" 10 ng 10 ng a 4.6 12 3.1 12 9.6 21 9.3 00 CMCMMM » NOfiL J CMJIIIfrf C/C«Mt* | . I| « « l .M 1 RSI) 51 HRGC/HRMS ""Sample was also spiki'd will) 100 iig of Arorlor 1260. �TABLE 7. RECOVERY (X) OF SEVERAL TGDD 1SOWERS FROM CLEANUP OPflON B Spike level I ml t*«l«w E*fcM 1,3,6,8 1,3,7,9 rCDI) ieooei Recovei-y_ (X) of 1 I.2.3.4/ 1,2, 3, 11 1,4,7,8 1,2,3,8 2,3,7,8 l3 C|s-2,3,7,8 1,2,8,9 I nit 7.6 15 35 14 37 38 48 1 ng 2.4 21 31 17 34 40 52 1 ng" 6.4 25 34 22 40 41 57 14 21 50 44 70 11 36 25 47 48 59 17 44 31 59 52 78 16 32 22 45 44 61 40 31 28 21 12 19 J H2SO4 - SIQ2 4.0, SIOJ 1.0* 1 10 ng 0.9 10 ng 1.8 7.6 AcMU AljOj 4.0, J30.1 JO* 10 nga 11 C«tep* C/C.HI* U_t lalun Heart HRGC/HRHS X RSI) 5.0 79 Sample was also spiked w i l d 100 ng ol Arorlur 1260. �TAWJi 8. KECOVEKY (1) Of StVEKAI. TCOO ISONEKS FROM CLEANUP OPTION C Spike level 1 l«4.l»U»«l»t»lMCl 1 1, 3.6,8" 0 . 10 ng 300 0 .10 ng l,3,7,9a Recoveir*JXlj?UCIW ieomer 1 1,2, 3.7/ 1,2.3,8 2,3,7,8 1X/!« 390 107b 13 CI2-2. 3,7,8 1,2,8.9 c 107 89 81 310 420 nu" c 102 89 84 0 .10 ngd 340 HjSO* - SIOj 440 97b c 88 92 91 1• 0 1.0 116 94 95 89 MOj , 1 CMKMM. M • 5 1 . J AcMcAHQ) 4.0, | 30.430% ro O ng 158 165 ng 140 155 162 98 96 80 75 80 .0 ngd 168 183 136 120 96 90 94 .0 ng 171 155 118 116 110 88 107 10 .0 ng 157 140 130 130 106 122 10 .0 ngj 130 130 102 99 100 85 110 Nean 210 240 115 112 101 90 95 I RSD GMC«M» to MBfil 130 39 54 14 12 14 J C^top* C/C.IM. I tml Mu« •• 9.1 HRGC/HRNS a Tlie 1,3,6,8- and 1,3.7,9-TCDD iatmere were also noted in reagent blanks fio» the acidic alumina column. Ho such illterlerenccs were noted froM the acidified allied gel or the Carbopak C/Celite . colimi. Resolution of 1,2,3,4-, l,2,3,7-/l,2,3,8-, and 1,4,7,8-TCDD was not achieved. This value represents recovery of the four isomers. ^Recovery reported with 1,2.3,4-/1,4,7,8-^:110. 'Sample was also spiked with 100 ng of Arcxlor 1260. 15 �TABLE 9. RECOVERY ( ) OF SEVERAL TCDD ISOHEKS FROM CI.EANIII' OPTION 0 1 1 Spike level 1 , 1 6,8a l,3,7,9a Recovery ( ) of TCUD isomer 1 1234 .../ I.2.3.7/ 1,4,7,8 1,2,3,8 2,3,7,8 197 113b c 12lb c 0. 10 ngd 260 200 360 nob 1.0 ng 90 93 1.0 ng ISO 1.0 ngd > 3 C|2-2,3,7,8 1,2,8,9 c 72 71 84 84 90 97 103 85 185 106 95 121 109 53 108 135 158 85 79 47 86 90 60 10.0 ng 81 50 114 105 10.0 ngJ 126 138 89 104 108 95 101 107 91 78 82 92 84 1 . ng 00 107 72 Mean 158 168 97 X RSD 46 51 21 101 14 85 23 84 13 0.10 ng 0.10 ng r\> •" "• | HRGC/HRHS| 147 290 'I'lie 1,3,6,8- ami 1,3,7,9-Tt'DD isuncrs were also noted in rcagenl blanks frov the acidic alumina colimn. No such inter!denies were noted lioiu the acidified silica gel or the Carltopak C/Celitc . cohum. Resolution of 1,2,3,4-, 1,2,:},7-/l ,2,3,8-, and 1,4,7,8-TCDD was uol achieved. This value represents recovery of the lour isoners. ^Recovery reported w i t h 1,2,3,4-/l,4,7,8-TCOO. Sample was also spiked w i t h 100 ng of Aroclur 1260. HKGC/IIRMS analysis was interrupted p r i o r to the u l u l l on of this isower. 70 71 e 63 118 112 85 29 �Three additional experiments were completed to evaluate the efficiency of reverse elution of the carbon column. The Carbopak C/Celite was placed in a 5-mL disposable pipette packed at both ends with glass wool plugs. The column was eluted in one direction for the hexane, cyclohexane/methylene chloride, and the methylene chloride/methanol/benzene mixture. The column was then turned over and eluted with 6 ml toluene. Triplicate analyses at the 0.10 ng/TCDD isomer spike level demonstrated average recoveries of 98 percent for 1,3,6,8-TCDD; 91 percent for 1,3,7,9-TCDD; 104 percent for 1,2,3,4-, 1,2,3,7-, 1,2,3,8-, and 1,4,7,8-TCDD; 116 percent for 2,3,7,8TCDD; 102 percent for 13C12-2,3,7,8-TCDD; and 93 percent for 1,2,8,9-TCDD. FINAL METHOD EVALUATION Based on the results of the column evaluation study, the analytical method was revised to specify the cleanup procedure presented as Option D in Figure 2. The final protocol, as presented in Appendix A, was then evaluated as described below. The data presented in Tables 10 through 17 are summaries of the initial column calibration, HRGC and HEMS resolution checks, and the results of the sample analysis. Calibration Table 10 summarizes the RKF data for the concentration calibration standards from the initial calibration and the routine monitoring of the RKF values over the time required to complete the sample analyses. The RRF(I) as specified in the protocol is a measure of the response of 2,3,7,8-TCDD versus the internal standard, 13Ci2-2,3,7,8-TCDD. The value for RRF(I) varied ±9.4 percent over the five concentration levels of 2,3,7,8-TCDD ranging from 2.5 pg/nL to 40 pg/(JL. The REF(II) is used to calculate the absolute recovery of the internal standard as compared to the recovery standard 13 C12-1,2,3,4-TCDD. The average RRF(II) was determined to vary by ± 19.3 percent over the calibration curve. The variability of the RRF(I) and RRF(II) were determined to be less than ± 10 percent and ± 18 percent, respectively, over all data points required to complete the sample analysis. In addition to the analysis of calibration standards specified in the protocol, solution HRCC6 was analyzed in triplicate to determine the lower limit of sensitivity (1 pg/pL). Although the calculated RRF(I) and RRF(II) values and the S/N are within the specified criteria, the ion ratio for the native compound and recovery standard indicate that these measurements fall outside the acceptable calibration window. 22 �TARI.E 10. —. Calibration standard Date 09/12/85 09/12/85 09/12/85 IIRCCI HRCCI IIRCCI Tine mil 320/322 09:05 09: 44 12: 31 INITIAL CALIBRATION SWMRT — — »/z 332/334i (IS) 0. 77 0. 80 0. 73 0. 82 0. 84 0. 84 •/! 332/334 (RS) 0. 80 0. 73 0. 87 S/N 259 S/N 322 S/N 334(18) RRF(I) RRF(II) 41:1 48:1 24:1 > 65:1 > 65: 1 > 65: 1 > 65:1 > 65:1 > 65:1 0.783 0.794 0.750 0.776 2.9X 2.18 2.27 2.28 2.24 2.4X 0.829 0.853 0.799* 0.861 0.848 2. OX 1.76 1.93^ 2.03* 2.27 1.99 13. OX Hean t RSD HRCC2 HRCC2 IIRCC2 HRCC2 13:00 13: 27 13: 53 15: 39 0. 73 0. 80 0. 92 0. 78 0. 77 0. 73 0. 66 0. 77 0. 83 0. 81 0. 69 0. 73 HRCC3 HRCC3 HRCC3 09/13/85 09/13/85 09/13/85 10: 31 10: 57 11: 23 0. 78 0. 78 0. 73 0. 79 0. 80 0. 83 0. 79 0. 78 0. 78 97:1 111:1 110:1 > 65: 1 > 65: 1 > 65: 1 > 65:1 0.974 > 65:1 0.965 > 65:1 0.972 Mean 0.970 % RSD 0.5X 1.53 1.57 1.58 1.56 1.9X HRCC4 HRCC4 HRCC4 I\J 09/12/85 09/12/85 09/12/85 09/12/85 09/13/85 09/13/85 09/13/85 13: 02 13: 29 13: 56 0. 77 0. 73 0. 77 0. 80 0. 76 0. 78 0. 76 0. 78 0. 76 96:1 > 144:1 > 144:1 > 65: 1 > 65:1 > 65: I > 65:1 0.945 > 65:1 0.935 > 65:1 0.967 Hean 0.949 t RSD I.7X 1.49 1.52 1.49 1.50 1-1X IIRCC5 HRCC5 IIRCC5 09/13/85 09/13/85 09/13/85 14: 22 14: 49 15: 15 0. 78 0. 75 0. 76 0. 80 0.82 0. 78 0. 78 0. 83 0. 79 > 144:1 > 144:1 > 144.1 > 65: 1 > 65: 1 > 65: 1 > 65:1 > 65:1 > 65:1 Hean X RSD 0.964 1-01 0.989 0.987 2.21 1.52 1.42 1.43 1.46 3.9X Overall Hean (RRF) X RSD 0.906 9.4X 1.75 19. 3X 36:1 58:1 78:1 76:1 IIRCC6 HRCC6 HRCC6 09/16/85 09/16/85 09/16/85 10:43 11: 19 13:44 1. 32 1. 18 0.86 0. 71 0. 83 0. 80 0. 84 1. 14 1. 01 09/16/85 09/16/85 12: 42 14: 40 0. 87 0. 83 0. 8] 0. 79 0. 75 0. 85 36: 48: HRCC2 HRCC2 IIRCC2 HRCC2 HRCC2 IIRCC2 IIRCC2 IIRCC2 IIRCC2 09/20/85 09/23/85 09/23/85 09/24/85 09/25/85 09/26/85 09/27/85 09/30/85 10/03/85 10: 46 08: 47 10: 46 10: 47 08: 39 08: 56 09:33 09: 19 08: 53 0. 86 0.82 0. 88 0. 89 0. 76 0. 77 0. 78 0. 83 0. 68 0. 73 0. 80 0. 80 0. 69 0. 78 o. 80 0. 84 0. BO 0. K 0. 80 0. 70 0. 83 0. 69 0. 89 0. 72 0. 81 0. 85 0. 81 > 75: > 75: 42: 26: 58: 73: 49: 73: 29: fNnl inrlndrd in nean RRK computation. M«l L f i t l i i n t,. u *hl a l ;_ 1 1 IT Fn«- *-••••• t rt« 9 ra 1 i l i t - a f Inn 65: 1 65: 1 65: 1 65:1 > 65:1 > 65:1 > 65:1 > 65:1 Mean I RSD 21: 25: 30: > 65: > 65: > 65: 0.917 0.878 0.935 1.44 1.09 1.58 63: > 63: > 65: > 65: 0.876 0.850 1.86 2.11 > > > > > > > > > > > > > > > > > > 0.835 0.832 0.941 1.01 0.949 0.941 1.04 0.854 0.955 1.98b 2.65b 1.41 1.68 1.96 1.57 1.68 1.67 1.53 10: 9.6: 18: IIRCCI IIRCCI > > > > 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: 63: �TABLE 11. HRGC AND MASS RESOLUTION CHECK SUMMARY Date Inst. ID 9/12/8S 9/12/85 9/12/85 9/12/85 9/13/85 9/13/85 9/13/85 9/13/85 9/16/85 9/16/85 9/16/85 9/16/85 9/17/85 9/17/85 9/18/85 9/18/85 9/19/85 9/19/85 9/20/85 9/20/85 9/20/85 9/20/85 9/23/85 9/23/85 9/23/85 9/23/85 9/24/85 9/24/85 9/24/85 9/24/85 MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO MSSO NS50 MSSO MSSO HSSO MSSO MSSO MSSO MSSO MSSO MSSO Sol. ID Time PC 07:51 08:32 16:05 16:47 08:26 08:45 PC 15:48 PC PC - PC PC PC - 16:23 09:25 10:45 15:16 15:57 10:26 10:14 07:58 PC 08:18 - 11:14 PC 12:56 - 0:0 80 PC PC 08:16 15:44 PC 16:14 07:55 08:15 16:01 16:45 09:51 10:15 PC 16:01 - • PC PC *l 16:33 TCDD isomer resolution (% valley) File name MID254I12X1 8367I12XQ1 83671 12XQ9 MID2S4I12X2 MID254I13X1 8367I13XQ1 83671 13XQ12 MID254I13X2 MID254I16X1 8367I16XQ1 83671 16XQ8 MID254I16X2 8367I17XQ1 MID254I17X1 MID254I18X1 8367I18XQ1 MID2S4I19X1 8367I19XQ1 MID2S4I20X1 8367I20XQ1 8367I20XQ5 MID254I20X3 MID2S4I23X1 8367I23XQ1 8367I23XQ6 Manual check1 MID254I24X1 8367I24XQ1 8367I24XQ3 MID254I24X2 5.9 2.9 6.9 11.4 - 11.9 23.0 13.3 20 3.5 * 6.7 4.1 - 8.8 12.5 12.2 13.1 Mass resolution at 10% valley 10,774 5 ppm 10,450 10,230 0 ppm 10,384 10,294 24 4 ppm 10,388 10,824 11,019 2 ppm 1 ppm 11,679 4 ppra 12,068 3 ppm 10,777 10,096 1 ppm 12,500 10,374 10,567 *A manual resolution check was performed due to data system failure. (continued) Mass measurement error 3 ppra �TABLE 11. (continued) Date Inst. ID 9/25/85 9/25/85 9/25/85 9/25/85 9/26/85 9/26/85 9/26/85 9/26/85 9/27/85 9/27/85 9/27/85 9/27/85 9/30/85 9/30/85 9/30/85 9/30/85 10/3/85 10/3/85 10/3/85 10/3/85 MS50 MS50 HS50 HS50 HS50 MS50 MS50 MS50 MS50 MS50 MS50 MS50 MS50 MS50 MS50 M550 MS50 MS50 MS50 HS50 a Sol. ID PC PC - - PC PC PC PC - PC PC - PC PC " Time 07:50 08:05 16:13 16:45 08:07 08:21 15:49 16:21 08:19 09:02 16:01 16:29 08:15 08:33 15:10 15:41 07:59 08:20 15:56 16:29 TCDD isomer resolution ( vallev) U File name MID254I25X1 8367I25XQ1 8367I25XQ3 MID254I25X2 MID254I26X1 8367I26XQ1 8367I26XQ3 HID254I26X2 HID254I27X1 8367I27XQ1 8367I27XQ3 MID254I27X2 MID254I30X1 8367I30XQ1 8367I30XQ3 MID254I30X2 MID254J03X1 8367J03XQ1 8367J03XQ3 MID254J03X2 . 11.1 6.5 8.3 13.2 11.9 11.8 < 25 < 25 17 12 ' Mass resolution at 10% valley 11,165 0 ppm 11,419 10,989 3 ppm 10,499 11,564 1 ppm 10,639 11,149 5 ppm 11,321 10.567 - - 10,442 A manual resolution check was performed due to data system failure. 25 Mass measurement error 0 ppra - ' �TABLE 12. TCDD DATA REPORT FORM HRGC/HRMS Analysis Sample No. 8367-83-1576X-DWD Aliquot Air-dry wt. (g) or Vol. (L) 1.0 L TCDD Isomer TCDD (ppt Retention time or ppq) 1J CJ2-2,3,7,8 Meas. DL TCDD 2,3 9 7,8- 21:35 1,3,6,8- 16:43 1,3 I 7,9- 17:54 21:34 228 Instr. ID MS50 Date Time 09/20/85 11:27 117 86.5 - Relative Ion Abundance Ratios 320/322 332/334(15) 332/334(RS) % Rec. m/z 259 0.78 0.69 0.80 0.63 0.68 40 S/N m/z 322 m/z 334(IS) Comments 73:1 64:1 41:1 > 63:1 > 49:1 Ratio unacceptable. Rerun. 50:1 30: 1 8367-82-1576X-DW 1.0 L 2,3 J 7,8- 21:35 21:33 196 MS50 09/20/85 13:02 0.58 0.69 0.72 14 21:1 62:1 > 20:1 Ratio-unacceptable; low recovery. 8367-89-1576X-EWWD 1.0 L 2,3 * 7,8- 21:40 1,3 9 6,8- 16:48 1,3 97,9- 17:58 21:37 -2,277 HS50 09/20/B5 14:41 0.79 1.02 0.84 0.90 0.72 0.71 96 > 146:1 10:1 28:1 > 63:1 > 5:1 > 11:1 > 63:1 Sample spiked at twice requested level. 6:1 63: 1 0.77 0.89 0.71 146:1 > 63:1 7:1 31:1 0.72 1,860 0.74 0.85 0.83 0.81 27:1 22:1 30:1 54:1 34:1 22:1 33:1 > 63:1 1,840 340 ,3 1,330 0.71 0.78 0.80 45:1 78:1 22:1 > 63:1 c 8367-88-1576X-EWW 8367-90-1576X-EWWN 1.0 L 1.0 L t 134 282 137 24:10 2,3 » 7,8- 21:29 c 2:7 35 21:27 2,3 J 7,81,3 ) 86, 1,3 9 7,91,2 » 3,7/ 1,2 9 3,81,2 » 3,41,2 9 7,81, J 8,92 21:56 21:59 17:02 18:14 22:18 22:31 2:0 43 30:01 1.0 L 2,3 9 7,8- 21:55 "21:55 8367-92-1576X-IWWD 1.0 L 2,3 9z,8- 22:01 1,3 » 6,8- 17:06 1,3 9 7,9- 18:16 1,2 1 7,8- 24:33 21:59 2,3 9 7,8- 22:00 1,2 97,8- 24:31 21:59 2,3 9 7,81,3 96,81,3 9 7,91,2 f 3,11 1,2 t3,81,2 r3,41,2 9 7,81^2 98,9- 2:0 20 8-367-84-1576X-DWN 1.0 L 1/0 L 22:01 17:06 18:18 22:21 22:35 24:32 30:62 MS50 09/20/85 MS50 1,010 09/23/85 11:17 0.72 0.82 61 91 > 63:1 > 63:1 60:1 18: 1 1,290 MS50 09/23/85 12:51 0.81 0.90 0.81 23 37:1 63: 1 30:1 508 MS50 09/23/85 13:29 0.74 0.85 0.75 0.80 0.71 75 49:1 25:1 27:1 >-63:l 8:1 63:1 29: 1 29:1 16: 1 23:1 191 208 55.2 1,520 0.89 MS50 09/23/85 14:00 586 234 0.87 0.78 0.86 20 49:1 21:1 63:1 -31:1 0.71 0.74 82 23:1 73:1 53:1 42:1 . 63:1 50:1 -45^1 61:1 70:1 23:1 55:1 63:1 20:1 0.71 512 395 403 MS50 09/23/85 0.75 0/80 0.74 0.72 616 840 328 0.72 0.80 0.79 •Aqueous sample data reported as ppq and soil sample data presented as ppt. Criteria for positive identification require that the ion ratios fall between 0.67 and O.~90. Isomer could not be identified. (continued) 26 15:11 - 502 766 8367-85-1576X-IKD 8367-87-1576X-INDN 1,090 75.9 Rerun. 1:1 43 > 63:1 Low recovery. Low recovery. Rerun. �TABLE 12. (continued) HRGC/HRMS Analysis Aliquot Air-dry wt. (g) or Vol. ( ) L Sample No. 8367-9 1-1576X-IWW 1.0 I TCDD (Ppt Retention time or ppq) TCDD li:iC12-2,3,7,8 Meas . DL TCDD Isomer 2,3 ,7 ,8- 22 :00 l,3-r6 ,8- 17 :03 1,3 ,7 ,9- 18 :14 1,2 ,7 ,8- 24 :30 21:57 534 246 222 54 .7 - Instr. ID MS50 Date Time 09/23/85 15:01 - 1.0 L 2,3 ,7 ,8- 21 :50 21:48 1,430 - 8367-93-1576X-IWWN 1.0 L 2,3 ,7 ,8- 21 :39 1,3 ,6 ,8- 16 :47 1,3 ,7 ,9- 17 :57 1,2 ,3 ,11 21 :58 1,2 ,3 ,81,2 ,3 ,4- 21 :11 1,2 ,7 ,8- 24 :08 1,2 ,8 ,9- 29:35 21:38 530 582 690 940 2,3 ,7 ,8- 21 :38 1,3 ,6 ,8- 16 :45 1,3 ,7 ,9- 17 :55 1,2 ,3 ,7/ 21 :58 1,2 ,3 ,81,2 ,3 ,4- 22 :10 0 1,2 ,7 ,8- 24: 9 1,2 ,8 ,9- 29 :36 21:37 8367-70-1576X-H1N 10.01_g 29 49:1 63:1 > 33:1 09/24/85 11:19 0.76 0.82 0.80 0.72 0.72 0.73 71 23:1 42:1 43:1 49:1 49:1 40:1 41:1 49:1 > 63:1 50:1 72:1 22:1 64:1 63:1 20:1 -12:1 15:1 24:1 46:1 15:1 26:1 33:1 63:1 38:1 73:1 20:1 58:1 > 63:1 23:1 12:1 941 .: 42:1 25:1 > 63:1 MS50 - 21:46 12 .9 ND 15 .2 61 .8 - MS50 9.2 - 4 2,3 ,7 ,8- 21 : 4 1,3 ,7 ,9- IS :03 8367-67-1576X-B5N 10.00 g 21 :47 —16 :55 18 :05 22 :07 09/24/85 12:55 0.78 0.77 56 0.78 09/24/85 13:58 0.67 0.87 08 .0 07 .2 85 18:1 6.3:1 31:1 12:1 > 63:1 09/24/85 14:29 0.86 0.59 0 .67 " 0.73 0.71 48 16:1 15:1 20:1 63:1 > 63:1 0.77 9:1 11:1 14:1 4: 81 0.90 0.75 0. 3 8 35:1 73:1 26:1 58:1 63:1 22:1 54 .1 147 63 .9 - (continued) Low recovery. > 63:1 0.87 0.85 09/24/85 13:27 .Aqueous sample data reported as ppq and soil sample data presented as ppt. Criteria for positive identification require that the ion ratios fall between 0.67 and 0.90. Isomer could not be identified. 27 0.85 0.63 0.87 0.81 0.69 0.81 0.85 MS50 9.85 g 22 :19 24 :16 29 :45 - 15 .1 4.2 - 8367-66-1576X-B5D 2,3 ,7 ,81,3 ,6 ,81,3 ,7 ,91,2 ,3 , / 7 1,2 ,3 ,81,2 ,3 ,41,2 ,7 ,81,2 ,8 ,9- 0.79 0.82 0.77 - 30 .3 29 .0 51 .1 125 Comments > 63:1 7:1 ' > 62:1 48:1 32:1 16:1 0.79 21:43 21:45 72:1 47:1 36:1 0.76 MS50 2,3 ,7 ,8- 21 :47 1,3 ,7 ,9- 18 :05 77 08 .0 18 .2 8.5 - 10.00 g 0.81 09/23/85 15:31 - 118 252 100 8367-65-1576X-B5 0.83 m/z 334(IS) MS50 - 1,180 1,790 691 0.87 0.71 0.78 0.70 S/N m/z 322 MS50 8367-86- 1576X-INDD ' Relative Ion Abundance Ratios 320/322 332/334(15) 332/334(RS) % Rec. m/z 259 0.75 73 : Ratio unacceptable. �TABIE 12. (continued) HRGC/HRrtS Analysis Sample No. Aliquot Air-dry wt . (g) or Vol. (L) TCDD (ppt Retention time or ppq) TCDD ^0^-2,3,7,8 Meas . DL TCDD Isomer Instr. ID Jj Date Time Relative Ion Abundance Ratios 320/322 332/334(13) 332/334(RS) s/y % Rec. m/2 259 m/z 322 m/z 334(IS) 8367-68-1576X-H1 9.67 g 2,3*4,71,3,6,8- 21:42 16:48 21:40 34.3 4.5 - MS50 09/24/85 15:02 0.82 0.67 0.67 0.75 73 29:1 11:1 63:1 14:1 > 63:1 ~~~ 8367-69-1576x-HlD 10.00 g 2,3,7,81,3,6,81,3,7,9- 21:33 21:31 36.6 5.2 10.0 MS50 09/24785 15:32 0.70 0.69 0.88 0.74 0.81 46 39:1 11:1 15:1 31:1 > 54:1 2,3,7,81,3,6,81,3,7,9c 21:37 16:44 17:54 19:04 21:35 2,3,7,8- 21:35 1,3,6,8- 16:42 1,3,7,9- 17:53 19:02 c 21:33 2,3,7,8- 21:41 1,3,6,8- 16:48 1,3,7,9- 17:59 c 19:09 1,2,3, 6/ 22:01 1,2,3,81,2,3,4- 22:15 1,2,7,8- 24:12 1,2,8,9- 29:41 21:41 2,3,7,8- 21:44 1,3,6,8- 16:49 1,3,7,9- 18:00 c 19:10 1,2,7,8- 24:58 c .26:03 21:42 2,3,7,8- 21:36 1,3,6,8- 16:42 1,3,7,9- 17:53 19:02 c 21:34 Comments 8367-71-1576x-Bl 8367-72-1576x-BlD 1.02 g 1.05 8 • 8367-73-1576x-BlN 8367-74-1576x-H3 8367-77-1576x-FA 1.03 8 1.15 g ?r94 8 16:42 17:52 937 160 312 50.6 - 09/25/85 10:01 0.83 0.85 0.84 0,69 0.83 0.83 95 97:1 34:1 44:1 7.3:1 63:1 14:1 24:1 4:1 > 63:1 MS50 09/25/85 10:30 0.78 0.87 0.82 0.65 0.85 0.80 75 73:1 36:1 40:1 63:1 22:1 28:1 3:1 > 63:1 0.77 0.77 0.85 0-83 - MS50 1,280 333 635 52 518 .Aqueous sample data reported as ppq and soil sample data presented as ppt. 11:27 6.5:1 0.81 80 97:1 47:1 79:1 0.81 0.72 MS50 09/25/85 13:00 - MS50 09/25/85 n 67 *r,A n on (continued) 13:30 6.7:1 57:1 0.78 0.82 0_83 Isomer could not be identified. 28 09/25/85 695 1,170 463 1,720 1,880 1,750 1,250 10:1 MS50 785 201 308 ND 28.9 2,~020 164 237 70.6 31.7 27.3 5:1 51:1 87:1 23:1 0.81 0.86 0.81 0.74 0.92 0.68 0781 0.80 0.83 0.81 0.80 0.82 0.81 79 > 145:1 22:1 24:1 -9:1 2.5:1 3:1 0.82 4 67?1 109:1 94:1 60:1 Ratio unacceptable. > 63:1 23:1 35:1 4:1 27:1 29:1 63:1 29:1 > 63:1 > 6:1 > 10:1 > 3:1 6:1 6:1 > 63:1 28:1 -38:1 34:1 22:1 6:1 Low recovery. �TABLE 12. (continued) ==' Aliquot Sample No. TCDD " Retention time Isomer TCDD 1JCl2-2,3,7,8 Air-dry wt. (g) or Vol. ( ) i c -c c 1237 .../ 1238 ,,,1,2,3,4c 1,2,7,8c 8367-77-1576X-FA (concluded) c c c c c • 8367-94-1576X-H2W 1.0 I 1,220 274 109 675 2:8 20 22:35 2:5 40 2:9 44 25^:23 25:55 27:18 2:7 80 29:41 460 ,4 879 720 3,460 155 3,430 441 2,920 NDNC - c 1:7 64 17:58 19:08 2:6 00 2:9 04 2:3 41 2:5 45 26:01 2:5 72 -28:13 2,3,7,81368 ,,,1,3,7,9c 21:11 16:25 1:4 73 ^23:39 2:0 11 2378 2:0 ,,,- 11 1368 1:2 ,,,- 62 1,3,7,9- 17:32 2:9 10 c — c -C 8367-83-1576X-DWD 8367-82-1576X-DW 1.0 I 1.0 I Instr. ID Date Time 58 2:1 14 - NC NC NC NC NC NC NC MSSO MSSO 0 / 5 8 15:18 92/5 0.84 0.75 0.77 0.80 4: 41 16:1 6:1 32:1 • — -._.- —J.,-. — .,.. ,. - .. ... ,. . .. . ^ „ ,— 161 4: 31:1 23:1 80:1 7:1 8: 61 1: 41 68:1 7:1 6: 31 14:1 15:1 56:1 5:1 63:1 9:1 4: 61 4:1 NC NC 265 - MSSO ND 167 125 50.7 MSSO "Isoraer could not be identified. (continued) 0/68 92/5 0:6 95 08 .6 83 . 08 .2 ND 0.79 o.-gi 0.72 0.81 09 .6 > 141 5: Ratio unacceptable. 18:1 12:1 2: 01 6:1 3:1 ~3:1 ^: 71 63:1 9:1 3: 01 07 .3 07 .0 42 0.71 0.71 0/68 1:6 92/5 02 Comments ' 15:1 5:1 2.5:1 11:1 0.81 08 .4 0,79 09 .8 06 .8 0.75 08 .0 07 .2 08 .2 0-81 NC - ' --. b Relative Ion Abundance Ratios S/N 320/322 3 2 3 4 1 ) 332/334(RS) % Rec. m/z 259 m/z 322 m/z 3 4 I ) 3/3(3 3(S 0.75 0.78 0.79 0.68 0.57 .Aqueous sample data reported as ppq and soil sample data presented as ppt. 29 0 / 5 8 13:30 92/5 - " 300 — 140 106 -• HRGC/HRMS Analysis 0.81 0.88 08 .8 0.80 2:0 00 20.43 21:25 21:54 2 3 7 8 21:42 ,,,1,3,6,81,3,7,9c c c 1278 ,,,- — " TCDD (ppt or ppq) Meas. DL 07 .0 07 .4 16 > 65:1 ^_ 55:1 4: 01 63:1 2: 01 11:1 3:1 35:1 6: 31 81 : 25 : 1 4:1 3 2 3 4 f ) Ratio unacceptable; no amount 3/3fS computations performed. 18:1 18:1 14:1 3.2:1 31:1 24:1 21:1 7:1 4: 71 - Ratio unacceptable. 1: 01 631 .: 381 .: 21:1 12:1 11:1 2: 21 Low recovery. Rerun. J ~- �TABLE 12. (continued) HRGC/HRMS Analysis Sample No. 8367-88-1576X-EWW 8367-95-1576X-H2WD Aliquot Air-dry wt. (g) or Vol. (L) 1.0 L 1.0 L TCDD Isomer Retention time lJ TCDD Cl2-2,3,7,8 2,3,7,8- 21:18 1,3,6,8- 16:28 1,3,7,9- 17:38 1,2,7,8- 23:44 21:15 2,3,7,8- 21:12 21:13 1.0 L 2,3,7,8- 21:14 1,030 - K Instr. ID MS50 Date Time 09/26/85 12:32 119 221 119 NC - MS50 09/26/85 13:05 NC NC NC NC NC NC NC NC NC NC NC NC NC NC 1,3,6,8- 16:25 1,3,7,9- 17:35 c 18:43 c 19:38 c 20:20 c 22:11 1,2,7,8- 23:39" c 24:23 c 25:27 c 26:47 c 27:37 c 29:34 c 30:12 c 31:07 8367-96- 1576X-H2WN TCDD (ppt or ppq) Meas. DL 21:13 NC Relative Ion Abundance Ratios S/N 320/322 332/334(13) 332/334(RS) % Rec. m/z 259 B/Z 322 m/z 334(IS) 0.76 0.71 06 .8 0.76 09 .3 0.79 0.75 66 0.67 NC MS50 09/26/85 14:21 08 .7 > 146:1 > 63:1 32:1 3 2 3 4 1 ) Ratio unacceptable; no amount 3/3(3 computations performed. 76:1 42:1 4: 61 15:1 13.3 > 63:1 > 10:1 > 15:1 31:1 > 63:1 6:1 0.77 08 .2 08 .1 0.87 07 .9 08 .4 07 .0 07 .6 0.76 07 .9 0.73 0.71 08 .1 05 .8 - 73:1 12:1 17:1 Comments > 26:1 > 14:1 > 17:1 > 5:1 > 4:1 > 2:1 3.5:1 39:1 63:1 3:1 332/334(13) Ratio unacceptable; no amount 9:1 5:1 5:1 54:1 79:1 11:1 36:1 8:1 25:1 7:1 5:1 20:1 15:1 2.5:1 4:1 94 .7 08 .7 NC > 145:1 > 63:1 computations performed. 1,3,6,8- 16:26 1,3,7,9- 17:35 c 18:44 c 19:41 c 20:22 c 21:13 1,2,7,8- 23:40 c 2:4 42 c 2:0 53 c 26:51 c 27:39 c 29:36 c 30:14 c 31:12 NC NC NC NC NC NC NC NC NC NC NC NC NC NC 07 .9 07 .8 07 .8 08 .9 06 .3 0159 07 .9 07 .3 07 .5 07 .6 07 .0 0.77 07 .4 03 .3 ?Aqueous sample data reported as ppq and soil sample data presented as ppt. Criteria for positive identification require that the ion ratios fall between 0.67 and 0 9 . .0 (continued) 30 40:1 22:1 34:1 11:1 7:1 4:1 5:1 58:1 8: 91 11:1 39:1 9:1 23:1 3:1 > > > > > > > > > > > > > > 23:1 15:1 21:1 6:1 5:1 6:1 3:1 41:1 6: 91 10:1 28:1 6:1 20:1 25:1 - Ratio unacceptable. Ratio unacceptable. - Ratio unacceptable. �TABLE 12. (continued) HP.GC/HRMS Analysis Sample Mo. Aliquot Air-dry wt. (g) or Vol. (L) Retentioa time TCDD 13Ct2-2,3,7,8 TCDD Isoraer TCDD (ppt or ppq) Meas. DL Instr. ID Date Time Relative Ion Abundance Ratios S/N 320/322 332/334(13) 332/334(RS) * Rec. m/z 259 m/z 322 m/z 334(IS) Comments 8367-75-1576X-H3D 1.16 g 2,3,7,8- 21:20 1,3,6,8- 16:31 1,3,7,9- 17:41 c 18:49 21:19 2,260 116 163 - MS50 09/27/85 15:31 0.79 0.90 0.67 0.88 0.80 99 > 145:1 14:1 20:1 6:1 8367-76- 15 76-X-H3N 1.14 g 2,3,7,81,3,6,81,3,7,9c 1,2,3,7/ 1,2,3,81,2,3,41,2,7,81,2,8,9- 21:39 16:45 17:56 19:07 21:58 21:39 1,800 383 367 ND 825 - MS50 09/30/85 09:57 0.80 0.80 0.81 1.00 0.81 0.79 0.80 86 > 143:1 > 63:1 44:1 > 15:1 41:1 > 17:1 5:1 > 2.5:1 79:1 > 29:1 2,3,7,81,3,6,81,3,7,9c c c 1,2,3,7/ 1,2,3,81,2,3,4c c c c c c c 21:33 16:41 17:51 19:01 19:57 20:41 21:52 2,3,7,81,3,6,81,3,7,9c c c 1,2, 3, 11 1,2,3,81,2,3,4- 21:39 16:45 17:56 19:06 20:01 20:47 21:58 8367-78- 1576X-FAD 8367-99-15 76X-FAH 10.04 g 9.93 g 21:30 1,020 926 747 610 557 146 286 - MS50 09/30/85 21:38 1,160 1,390 1,160 881 888 194 423 - 0.73 0.77 0.80 0.87 0.89 0.71 0.67 0.88 6.7 MS50 09/30/85 13:00 0.80 0.78 0.79 0.85 0.74 0.76 0.80 0.36 3,620 .Aqueous sample data reported as ppq and soil sample data presented as ppt. cCriteria for positive identification require that the ion ratios fall between 0.67 and 0.90. Isomer could not be identified. (continued) 0.84 0.88 5 45:1 55:1 42:1 32:1 23:1 8.7:1 17:1 32:1 37:1 25:1 22:1 14:1 6:1 10:1 25:1 40:1 28:1 18:1 17:1 5:1 10:1 63:1 10:1 > 63:1 Ratio unacceptable. 63:1 11:1 8:1 56:1 63:1 10:1 48:1 47:1 56:1 94:1 80:1 60:1 35:1 13:1 29:1 131:1 32:1 > 63:1 > 24:1 63:1 23:1 75:1 16:1 13:1 92:1 97:1 17:1 80:1 83:1 0.71 0.79 0.91 0.74 0.78 0.79 0.83 0.76 0.79 2,260 ND 329 356 3,520 3,680 558 3,120 3,270 22:05 22:33 24:01 24:45 25:51 27:14 28:03 30:01 31 10:29 63:1 5:1 7:1 2.5:1 60:1 145:1 46:1 0.83 0.85 0.77 855 2.330 952 22:11 24:10 29:36 22:11 49.5 > > > > 9:1 -Recovery low. Ratio unacceptable. 6:1 Low recovery. �TABLE 12. (concluded) HRGC/HRMS Analysis Sample No. Aliquot Air-dry wt. (g) or Vol. ( ) L TCDD Isoner c 8367-99- 1576X-FAN (concluded) 1,2,7,8- c c c c c c 8367-100-1576X-DWD 1.0 L TCDD (ppt Reteation time or ppq) i-i C12-2,3,7,8 fleas. DL TCDD 22:38 24:08 24:52 25:58 27:20 28:10 30:09 30:47 562 516 4,310 4,530 SD 390 ,8 4,080 1,170 2,3,7,8- 21:04 1,3,6,8- 16:19 1,3,7,9- 17:26 21:01 500 ml. 2,3,7,8- 21:02 21:03 8367-101-1576X-EWWD 1.0 mL 2,3,7,81,3,6,81,3,7,9- 2:8 05 16:14 17:23 2:5 05 2,3,7,81,3,6,81,3,7,9- 20:57 16:17 17:23 20:56 2,3,7,81,3,6,81379 ,,,- 21:00 16:18 17:27 18:32 2:6 40 25:09 27:16 2:9 94 2:9 05 20:56 16:15 17:22 18:30 2:3 40 25:07 24:17 2:5 94 2:6 05 8367-104-1576X-H2W 500 mL 430 mL c c c c c 8367-105-1576X-H2W 430 mL 2,3,7,81,3,6,81,3,7,9- c c c c c Date Time MS50 09/30/85 13:00 632 - MS50 604 1,050 Relative Ion Abundance Ratios 320/322 332/334(13) 332/334(RS) % Rec. 5 0.81 0.82 0.75 0.73 0.95 0.77 0.88 0.71 0.86 S/N m/z 259 m/z 322 m/z 3 4 I ) 3(S 32:1 20:1 127:1 146:1 20:1 24:1 122:1 26:1 10:1 10:1 54:1 63:1 7:1 46:1 44:1 14:1 Comments Ratio unacceptable. 0.82 0.74 68.5 21:1 73:1 51:1 18:1 63:1 47:1 > 63:1 0.77 0.80 - MS50 10/03/85 12:58 0.71 0.87 0.85 59.5 16:1 31:1 > 63:1 - MS50 1/38 00/5 13:28 0.69 0.81 0.82 0.90 80 72:1 17:1 28:1 63:1 > 63:1 18:1 42:1 13:1 17:1 > 63:1 > 63:1 16:1 30:1 19:1 32:1 28:1 18:1 > 63:1 18:1 31:1 55:1 21:1 32:1 26:1 27:1 > 145:1 16:1 17:1 30:1 20:1 33:1 23:1 13:1 > 63:1 19:1 21:1 34:1 14:1 31:1 14:1 10:1 > 63:1 157 384 628 - ND 45 MS50 10/03/85 14:01 87 27,100 SD 71 164 391 SD 427 531 575 - 313 - MS50 1/38 00/5 14:33 0.75 0.78 57 6:1 8:1 0.74 0.84 0.74 78 1.05 06 .8 0.85 0.65 0.70 0.81 08 .6 - 28,100 224 302 453 564 730 518 347 0.71 0.52 0.69 MS50 .Aqueous sample data reported as ppq and soil sample data presented as ppt. Criteria for positive identification require that the ion ratios fall between 0.67 and Isomer could not be identified. * 32 10/03/85 11:32 637 489 8367-102-1576X-IND 8367- 103- 1576X-IND 246 - Instr ID 10/03/85 15:17 0.80 0.85 0.88 0.73 0.80 0.81 0.80 0.79 0.90. > 145:1 9:1 0.84 0.71 96 13:1 26:1 Ratio unacceptable. Ratio unacceptable. �TABLE 13. ACCURACY AND PRECISION OF THE HRCT/HRHS ANALYSIS FOR 2,3,7,8-TCDD FROM LABORATORY AQUEOUS HATRIX^SIMKS_ Sanple aiatrix ,7,8-TCDO *,3, Spike level (ppo.) Distilled water (DW) 250 250 250 Effluent vastevater (EWW) 1,000 1,000 1,000 Influent vastevater ( W I) 2,3,7,8-TCDD Detected (ppq) 93.6 106 103 Average rec. 101 RPR 9.3 82 42 69 Average rec. 64 RPR 63 1,090, 1,030 1,010 1,050 Average cone. 1,050 7.6 RPR 109, 103 101 105 Average ree. 105 RPR 7.6 61 , 66 91 80 Average rec. 75 40 RPR 107 102 106 Average rec. 105 RPR 4.8 77 75 71 Average rec. 74 RPR 8.1 258 304 286 Average rec. 283 RPO 16 .. 23 20 29 Average rec. 24 RPR 38 534 508 530 500 500 Average cone. RPR 500 500 500 Industrial vaatevater (IND) Soil extract (H2W) , - 524 5.0 1,290 1,520 1,430 Average cone. 1,410 16 RPR „ CO to "Clt-2 ,3,7,8-TCDD Abaolute recovery (X) 234 265 246 Average cone . 248 RPR1 12.5 500 Industrial vastevater (IND) 2,3,7,8-TCOD Recovery (X) Average cone. RPDD 604 628 616 3.9 27,100 28,100 Average cone. 27,600 RPD 3.6 . - 60 57 Average rec. 58 RPD 5.2 78 96 Average rec. 87 RPD 25 Relative percent range (calculated {torn the difference of the high anil low values divided by the average of all values and by 100 percent). Relative percent difference. b amltlplied �TABLE 14. PRECISION OF THE IIRGC/IIRHS ANALYSIS FOR 2,3,7,8-TCDD Of SOIL AND FLY ASH SAMPLES .. Sample Matrix B25-Piazza Road (B5) ^_ — . . ._ ._ _*. Endogenous 2,3,7,8-TCIW level (pi.l)a _ . ._ . _n 73 85 48 69 54 73 46 56 58 47 95 75 80 83 24 79 99 86 88 23 4 7 5 5.3 57 70 RPR 3.1 . 7 Average rec. RPR 360 1,700 Average cone. 1,000 50 RPR 2,020 2,260 1,800 Average cone. 2,010 . Average rec. 34 34.3 36.6 30.3 19 937 785 1,280 RPR Fly ash - RRAI __ 18.2 15.1 12.9 15.4 50 RPR llyrle Park 0 . (113) 03 _. 13 Ciz-2 ,3,7,8-TCDD % Absolute recovery ( ) Average cone . B52-Shenandoah (Bl) . 2,3,7,8-TCDD Detected (ppl) Average rone. RPRD Hyde Park 001 (III) .._. 23 1,720 1,020 1,160 Average cone. 1,300 f>4 RPR Average rec. RPR Average rec. RPR Average rec. RPR Estimated level of endogenous 2,3,7,8-TCDD reported to MR I by Dr. W. Beckert in letters dated 19, 1985, and August .10, 1985. Relative percent range (calculated from the difference of the high and low values, divided by the average ol .ill values, and m u l t i p l i e d liy 100 percent. h April �TABU! 15. ACCURACY OF THE HUCC/HKHS METHOD FOR THE PETERMJKATIOH Of TEMMSOM5RS SPIKED IKTO AQUEOUSt IttTRtCES . TCDO •ntlyte Effluent vaaU,'Katcr Spike Measured Recoviery (P*> (PR) __(X) 1,3,6,8 1,840 502 27 1,3,7,9 840 766 Spike (PC) 01 mil led water Heaaured Recovery Influent vaatewater Spike Heaaured Recovery () X taL_ (PS) Industrial vaatevater Spike Measured Recovery (P8> (PK) (X) (pgl (X) 460 512 III 920 582 63 920 HD* 0 91 210 395 190 420 690 164 420 NO 0 1,2,3,7/1,2,3,8 1,680 1.860 110 420 403 96 840 940 112 840 HI) 0 1,2,3,4 2,440 1,840 75 610 616 101 1,220 1,180 97 1,220 NO 0 1,2,7.8 3,080 3,430 111 770 840 110 1,540 1,790 116 1,540 586 38 1,2,8.9 1,200 1,330 HI 300 328 110 600 691 115 600 ND 2,3,7.8 a ' C,t-2,3,7,8 U> 01 1,000 500 b 0 1,010 101 250 234 94 500 530 106 500 904 181 455 91 500 410 82 500 355 71 500 100 20 *Hot detected. Heaaured value corrected for endogenous 2 ,3,7,8-TCDD content (averaged 616 ft/D. �AR836B(V), 1 TABU! 16. ACCURACY OF THE HHCC/HRMS METHOD FOR THE DETERMIHATION Of TCDD ISOHERS SPIKED IHTO SOIL MATRICES TCOO •nalyte Hyde P*rk 001 (HI) Spike Heiiured Recovery (i P) B25-Pl»*n Hold JB5J Spike He**ured Recovery (I P) (I P) r) t B52-Shen*ndo«t•l?li Spike Measured Recovery (I P) () X Hyde Pirk 003 (H3) SpTSe Heaiured Recovery (I P) (I P) () * 0 660 333 50 1,560 383 24 15.2 36 300 635 210 710 367 51 1,3,6,8 130 29.0 22 92 ND ( . ) 92* 1,3,7,9 60 51.1 86 42 1,2,3,7/1,2,3,8 120 125 106 84 61.8 74 600 518 87 1,430 825 58 1,2,3,4 170 118 69 120 54.1 44 880 695 79 2,070 855 41 1,2,7,8 220 252 117 ISO 95 1,110 1,170 106 2,620 2,330 89 1,2,8,9 84 100 119 60 63.9 107 430 463 108 1,020 952 93 12.9 - - 1,280 - 1,800 - 48 500 400 500 430 86 2,3,7,8 s « C,,-2, 3,7,8 500 30,3 280 56 500 147 240 *HD - not detected. The vclue in p»renthe«ei reflect! the eitiMted detection limit. O> 80 �TABLE 17. FORTIFIED FIELD BLANK RESULTS Sample So. ' Aliquot Air-dry at. (g) or Vol. (L) Retention time 13 Native C Date Time Relative Ion Abundance Ratios 332/334(13) 320/322 332/334(RS) Instr. ID S/N % Rec. m/z 259 m/z 322 m/z 334(IS) 8367-62- 1576X-FFWB 1.0 L 23:38 23:35 MS50 09/11/85 14:08 . 0.78 - 0.71 65 > 145:1 > 63:1 > 63:1 8367-64- 15 76X-FFSB 10 g 23:38 23:38 «S50 09/11/85 14:42 0.79 - 0.80 68 > 145:1 > 63:1 > 63:1 8367-63- 15 76X-FFSA 10 g 23:40 23:38 MS50 09/12/85 14:47 0.77 0.78 - 71 145:1 > 63:1 > 63:1 8367-61-1576X-FFWA 1.0 L 23:40 23:39 MS50 09/12/85 15:14 0.80 0.76 - 79 145:1 > 63:1 > 63:1 8367-81-1576X-FFB 10.01 g 22:17 22:15 MS50 09/20/85 13:33 0.88 - 0.74 29 144:1 > 63:1 > 63:1 8367-80- 1576X-FFA 10.01 g 21:37 21:35 MS50 09/20/85 14:09 0.86 0.73 - 83 145:1 > 63:1 > 63:1 8367-97-1576X-FFA 1.0 L 21:37 21:37 MS50 09/20/S5 09:10 0.82 0.83 - 50 23:1 > 63:1 > 63:1 8367-98- 15 76X-FFB 1.0 I 21:43 21:43 MS50 09/20/85 09:24 0.80 - 0.82 48 145:1 > 63:1 > 63:1 Comments 37 Low recovery �HRGC and Mass Resolution Table 11 presents a summary of all chromatographic and mass resolution checks completed during the final method evaluation. As per the protocol requirements the required mass resolution was demonstrated as the first and last quality control requirements for each day. The column performance check mixture was also analyzed before the first sample analysis and after the final sample analysis each day as a QC measure to assure that specificity for 2,3,7,8-TCDD was maintained. The mass measurement accuracy at m/z 330.979 is also included in this table, as it was verified on a daily basis prior to any sample analyses. Sample Analysis The results from the analysis of the aqueous and soil samples are provided in Table 12. The data in Table 12 are presented in the format specified as Form B-l in the protocol reporting requirement. The data are recorded in the chronological order that they were obtained by HRGC/HRMS. As indicated in Table 12, several samples required reanalysis due to low recovery of the internal standard, unacceptable ion ratios for 320/322, or the result of interferences at the internal standard. Two of the distilled water samples demonstrated responses for the characteristic ions at m/z 259, 320, and 322 for 2,3,7,8-TCDD. However, the ion ratio for the native 2,3,7,8-TCDD in one replicate and the ion ratio for the internal standard in another required that both samples be reanalyzed. Although both samples met all the qualitative criteria, recoveries were noted to be low (< 20 percent) for one of the samples and complete reanalysis of the replicate was required. Significant problems were encountered with the aqueous soil extract, H2W, and the fly ash sample. The problems with the soil extract resulted from an interference at m/z 332 that coeluted with the internal standard, 13 C12-2,3,7,8-TCDD. No accurate quantitative measurements could be achieved for TCDD responses observed for this sample. The original sample contained a large amount of suspended particulate in each of the three replicates. Problems with the extraction were noted with the first portion of methylene chloride. A large amount of particulate matter was noted at the interface of the aqueous and organic phases. Hence, the interference at m/z 332 and TCDD responses observed in these replicates were probably due to direct extraction of the suspended soil particulate rather than the actual watersoluble TCDD. The remaining aqueous sample for H2W was centrifuged for 10 min at approximately 2,000 rptn, and the aqueous phase was decanted from the settled particulate. The resulting aqueous sample was divided into duplicate 430-mL samples and each was analyzed according to the protocol. The HRGC/HRMS analysis was successful for both replicates with absolute recoveries of 78 percent and 96 percent of the internal standard. The triplicate analyses of the fly ash sample resulted in absolute recoveries less than 10 percent for the internal standard in each aliquot 38 �analyzed. These low recoveries may be associated with the total fixed carbon content of the fly ash material. Previous work in this laboratory with fly ash from coal-fired power plants has demonstrated low recoveries of analytes from materials with high carbon content.4 The only other sample for which successful analysis was not achieved as specified in the protocol on first analysis was the industrial wastewater (IND). The triplicate analysis of the sample resulted in absolute internal standard recoveries of 23, 20, and 29 percent. The criteria for successful analysis for TCDD as discussed in the protocol require an absolute recovery of 40 to 120 percent. In addition to the observed low recoveries, the level of 2,3,7,8-TCDD detected in the sample averaged 1,410 ppq as compared to the 500-ppq spike level. Two 500-mL aliquots of the unspiked industrial wastewater sample were reanalyzed to determine the background level of 2,3,7,8TCDD. The results of the duplicate analysis yielded an average 2,3,7,8-TCDD concentration of approximately 620 ppq and the absolute recoveries were noted to be 60 percent and 57 percent. The increase in absolute recovery of the internal standard in the unspiked sample by approximately a factor of two is possibly due to the preparation of samples one half the size of that used for the original analysis. This suggests that the sample matrix has a considerable impact on the effectiveness of the cleanup procedure. Table 13 provides a summary of the accuracy and precision of the analyses of the five aqueous sample types for 2,3,7,8-TCDD. Only the data points from Table 12 that demonstrate compliance with all QC criteria (ion ratios, absolute recovery of the internal standard, etc.) are included in Table 13. These data demonstrate that the isotope dilution method of quantitation provides accurate and precise quantitation of 2,3,7,8-TCDD in the aqueous samples. It should be noted that even when the absolute recovery of the 13Ci2" 2,3,7,8-TCDD internal standard varies by as much as 66 percent (RPR) for the triplicate distilled water samples, the accuracy of the measurement of the spiked 2,3,7,8-TCDD averaged 101 percent with less than 10 percent variability. Table 13 summarizes data for both the spiked and unspiked aliquots of industrial wastewater. The high recovery noted for the 2,3,7,8-TCDD value in the spiked samples is a result of the presence of this compound at approximately 620 ppq in the original matrix. Table 14 presents a similar summary for the five solid samples analyzed. The precision of the measurements is not quite as good as noted for the aqueous samples and may reflect the difference in adsorption of the endogenous 2,3,7,8-TCDD and the spiked internal standard to the matrices. Tables 15 and 16 provide data dealing with the accuracy of the HRGC/ HRMS methods for the determination of total TCDD isomers in aqueous and solid samples. In general, the data support the use of the internal standard method of quantitation for all but the earliest eluting isomers, 1,3,6,8- and 1,3,7,9-TCDD. The accuracy for the additional isomers is very good and more consistent than is observed for the solid samples. This may be partially due to the differences in adsorption to the soil particles. 39 �Fortified Field Blanks As part of the overall quality assurance/quality control (QA/QC) program identified in the HRGC/HRMS protocol, the analyst is required to analyze fortified field blanks to demonstrate (a) that the extraction and cleanup procedure will provide recovery of the 2,3,7,8-TCDD within the criteria of greater than 40 percent specified in the protocol and (b) that the reagents are free from contamination with TCDD isomers. Table 17 provides the results of the fortified field blanks run before proceeding with sample analysis and also those of an additional set of blanks prepared along with the actual samples. The analyses of the fortified field blanks at the outset of the study demonstrated that the recoveries of 2,3,7,8-TCDD and 1,2,3,4-TCDD ranged from 65 to 79 percent. No detectable levels of other TCDD isomers were found in this preliminary study. The field fortification blanks analyzed with the actual samples resulted in recoveries of 29 percent and 83 percent. More importantly, these analyses demonstrated some interferences arising from 1,3,6,8- and 1,3,7,9-TCDD. Previous studies involving evaluation of the cleanup procedure indicated that these isomers are associated with the acidic alumina cleanup. Figure 4 is a plot of the ratio of response of 1,3,6,8- and 1,3,7,9TCDD and the response of the recovery standard 13C12~1,2,3,4-TCDD versus the time elapsed since the acidic alumina was cleaned and activated at 190°C. The results of the analyses of the fortified field blanks and the samples not spiked with the 1,3,6,8- and 1,3,7,9-TCDD isomers are presented in Figure 4. As noted from this plot, these TCDD isomers were not initially detected in the acidic alumina immediately following cleanup by Soxhlet extraction. The first set of fortified field blanks was taken through the acidic alumina column 7 days later. Although response was observed at m/z 320 and 322 at the retention time for these isomers, the ion ratios did not indicate presence of the compounds. Since the detectable levels were well below 10 pg/g of alumina, the sample analyses were initiated. The data for the fortified field blanks and samples taken through alumina from 14 to 30 days from activation indicate that the contamination of the 1,3,6,8- and 1,3,7,9-TCDD isomers apparently occurs over time using this particular oven. The background contamination of 1,3,6,8- and 1,3,7,9-TCDD isomers has also been recently addressed by the Center for Disease Control.5 Note Added in Proof A second magnetic sector instrument (built in 1976) from a different manufacturer was tested and found to be incapable of achieving sufficient sensitivity at 10,000 resolving power to be used in experiments for this study. 40 �1.2 Fortified Field Blanks 1.1 I I Aqueous and Environmental Samples o o t! 1.0 i CM 0.9 2 0.8 u CO 0.7 g 0.6 u * 0.5 rC ^ CO < 0.4 0^ *o" n 0.3 o 0.2 o cc 0.1 10 20 Time (Days) Elapsed from Cleanup and Activation of Acidic Alumina 30 Figure 3. Background levels of 1,3,6,8- and 1,3,7,9-TCDD observed over the single-laboratory evaluation study. 41 �REFERENCES 1. U.S. Environmental Protection Agency, "Dioxin Strategy," prepared by the Office of Water Regulations and Standards and the Office of Solid Waste and Emergency Response in conjunction with the Dioxin Strategy Task Force, Washington, B.C., November 28, 1983. 2. L. R. Williams, Validation of Testing/Measurement Methods. EPA 600/X-83-060, 1983. 3. GC Bulletin 793C, Supelco Inc., Beliefonte, Pennsylvania, 1983. A. C. L. Haile, J. S. Stanley, T. Walker, G. R. Cobb, and B. A. Boomer, "Comprehensive Assessment of the Specific Compounds Present in Combustion Processes. Volume 3. National Survey of Organic Emissions from Coal-Fired Utility Boiler Plants," EPA-560/5-83-006, September 1983. 5. J. S. Heller, D. G. Patterson, L. R. Alexander, D. F. Groce, R. P. O'Connor, and C. R. Lapeza, "Control of Artifacts and Contamination in the Development of a Dioxin Analytical Program," presented at the 33rd Annual Conference on Mass Spectrometry and Allied Topics, May 26-31, 1985, San Diego, California. �APPENDICES 43 �APPENDIX A VALIDATED ANALYTICAL PROTOCOL for the Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total TCDDs in Soil/Sediment and Water by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry September 10, 1985 This analytical protocol has been written in the format used in the Superfund program, as "Exhibit D" of a Statement of Work which in turn is part of an Invitation-for-Bid package under the Superfund Contract Laboratory Program. The other exhibits of the Statement of Work, although cited in Exhibit D, do not pertain to this method evaluation study. �EXHIBIT D Analytical Methods 2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total TCDDs in Soil/Sediment and Water by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry �EXHIBIT D Section Subject Page 1 Scope and Application D-l 2 Summary of Method D-l 3 Definitions D-2 A Interferences D-3 5 Safety D-3 6 Apparatus and Equipment D-3 7 Reagents and Standard Solutions D-6 8 System Performance Criteria D-8 9 Quality Control Procedures D-13 10 Sample Preservation and Handling D-13 11 Sample Extraction D-14 12 Analytical Procedures D-l7 13 Calculations D-18 �1. SCOPE AND APPLICATION 1.1 This method provides procedures for the detection and quantitative measurement of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD; CAS Registry Number 1746-01-6; Storet number 3475) at concentrations of 10 pg/g (10 parts per trillion) to 200 pg/g (200 parts per trillion) in 10-g portions of soil and sediment and at 100 pg/L (100 parts per quadrillion) to 2000 pg/L (2 parts per trillion) in 1-L samples of water. The use of 1-g or 100-ml portions permits measurements of concentrations up to 2,000 pg/g (2 parts per billion) or 20 ng/L, respectively. This method also allows the estimation of quantities of total TCDD present in the sample. Samples containing concentrations of 2,3,7,8-TCDD greater than 2 ppb or 20 ng/L must be analyzed by a protocol designed for such concentration levels, with an appropriate instrument calibration range. 1.2 The minimum measurable concentration is estimated to be 10 pg/g (10 parts per trillion) for soil and sediment samples and 100 pg/L for water samples, but this depends on kinds and concentrations of interfering compounds in the sample matrix. 1.3 This method is designed for use by analysts who are experienced in the use of high-resolution gas chromatography/high-resolution mass spectrometry. CAUTION: TCDDs are extremely hazardous. It is the laboratory's responsibility to ensure that safe handling procedures are employed. 2. SUMMARY OF METHOD Five hundred pg of 13C,2-2,3,7,8-TCDD (internal standard) are added to a 10-g portion of a soil/sediment sample (weighed to 3 significant figures) or a 1-L aqueous sample and the sample is extracted with 200 to 250 mL benzene using a Soxhlet apparatus with a minimum of 3 cycles per hour or a continuous liquid-liquid extractor for 24 hours. A separatory funnel and 3 x 60 mL methylene chloride may also be used for aqueous samples. After appropriate concentration and cleanup, 50 uL of tridecane are added to the extract. Before HRGC-HRMS analysis, 500 pg of a recovery standard ( C^~ 1,2,3,4-TCDD) are added to the extract which is then concentrated to a final volume of 50 uL. A 2-uL aliquot of the concentrated extract is injected into a gas chromatograph with a capillary column interfaced to a high-resolution mass spectrometer capable of rapid multiple ion monitoring at resolutions of at least 10,000 (10 percent valley). Identification of 2,3,7,8-TCDD is based on the detection of the ions m/z 319.897 and 321.894 at the same GC retention time and within -1 to +3 seconds GC retention time of the internal standard masses of m/z 331.937 and 333.934. Confirmation of 2,3,7,8-TCDD (and of other TCDD isomers) is based on the ion m/z 258.930 which results from loss of COCL by the parent ion. D-l �3. DEFINITIONS 3.1 Concentration calibration solutions — solutions containing known amounts of the analyte (unlabeled 2,3,7,8-TCDD), the internal standard 13 C12-2,3,7,8-TCDD and the recovery standard 13C,2~1,2,3,4-TCDD; they are used to determine instrument response or the analyte relative to the internal standard and of the internal standard relative to the recovery standard. 3.2 Field blank — a portion of soil/sediment or water uncontaminated with 2,3,7,8-TCDD and/or other TCDDs. 3.3 Rinsate — a portion of solvent used to rinse sampling equipment; the rinsate is analyzed to demonstrate that samples have not been contaminated during sampling. 3.4 Internal standard — ^Cj2~2»3,7,8-TCDD, which is added to every sample (except the blanks described in Sections 4.2.1 and 4.2.3 of Exhibit E) and is present at the same concentration in every laboratory method blank, quality control sample, and concentration calibration solution. It is added to the soil/sediment or aqueous sample before extraction and is used to measure the concentration of each analyte. Its concentration is measured in every sample, and percent recovery is determined using an internal standard method. 3.5 Recovery standard -C,£~1,2,3,4-TCDD which is added to every sample (except for the blanks discussed in Sections 4.2.1.A.2 and 4.2.3.6, Exhibit E) extract just before HRGC-HRMS analysis. 3.6 Laboratory method blank — this blank is prepared in the laboratory through performing all analytical procedures except addition of a sample aliquot to the extraction vessel. 3*7 GC column performance check mixture — a mixture containing known amounts of selected standards; it is used to demonstrate continued acceptable performance of the capillary column, i.e., separation ( 25% valley) of 2,3,7,8-TCDD isoraer from all other 21 TCDD isomers £ and to define the retention time window. 3.8 Performance evaluation sample -- a soil, sediment or aqueous sample containing a known amount of unlabeled 2,3,7,8-TCDD and/or other TCDDs. It is distributed by EPA to potential contractor laboratories who must analyze it and obtain acceptable results before being awarded a contract for sample analyses (see IFB Pre-Award Bid Confirmations). It may also be included as an unspecified ("blind") QC sample in any sample batch submitted to a laboratory for analysis. 3.9 Relative response factor — response of the mass spectrometer to a known amount of an analyte relative to a known amount of an internal standard. 3.10 Mass resolution check — standard method used to demonstrate static resolution of 10,000 minimum (102 valley definition). D-2 �4. INTERFERENCES Chemicals which elute from the GC column within jflO scans of the internal and/or recovery standard (ra/z 331.937 and 333.934) and which produce ions at any of the masses used to detect or quantify TCDD are potential interferences. Most frequently encountered potential interferences are other sample components that are extracted along with TCDD, e.g. PCBs, methoxybiphenyls, chlorinated hydroxydiphenylethers, benzylphenylethers, chlorinated naphthalenes, DDE, DDT, etc. The actual incidence of interference by these chemicals depends also upon relative concentrations, mass spectrometric resolution, and chromatographic conditions. Because very low levels of TCDD must be measured, the elimination of interferences is essential. High-purity reagents and solvents must be used and all equipment must be scrupulously cleaned. Laboratory reagent blanks (Exhibit E, Quality Control, Section 4) must be analyzed to demonstrate absence of contamination that would interfere with TCDD measurement. Column chromatographic procedures are used to remove some coextracted sample components; these procedures must be performed carefully to minimize loss of TCDD during attempts to increase its concentration relative to other sample components. 5. SAFETY • 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 file of current 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. Additional references to laboratory safety are identified d~3) (page D-38). 2,3,7,8-TCDD has been identified as a suspected human or mammalian carcinogen. The laboratory is responsible for ensuring that safe handling procedures are followed. 6. APPARATUS AND EQUIPMENT 6.1 High-Resolution Gas Chromatograph/High-Resolution Mass Spectrometer/Data System (HRGC/HRMS/DS) 6.1.1 The GC must be equipped for temperature programming, and all required accessories must be available, such as syringes, gases, and a capillary column. The GC injection port must be designed for capillary columns. The use of splitless injection techniques is recommended. On-column injection techiques can be used but this may severely reduce column lifetime for nonchemically bonded columns. When using the method in this protocol, a 2-uL injection volume is used consistently. With some GC injection ports, however, 1-uL injections may produce improved precision and chromatographic separation. A 1-uL D-3 �injection volume may be used if adequate sensitivity and precision can be achieved. NOTE: If 1 uL is used at all as injection volume, the injection volumes for all extracts, blanks, calibration solutions and the performance check sample must be 1 uL. 6.1.2 Gas Chromatograph-Mass Spectrometer Interface The GC-MS interface may include enrichment devices, such as a glass jet separator or a silicone membrane separator, or the gas chromatograph can be directly coupled to the mass spectrometer source. The interface may include a diverter valve for shunting the column effluent and isolating the mass spectrometer source. All components of the interface should be glass or glass-lined stainless steel. The interface components should be compatible with 300°C temperatures. The GC/MS interface must be appropriately designed so that the separation of 2,3,7,8-TCDD from the other TCDD isomers which is achieved in the gas chromatographic column is not appreciably degraded. Cold spots and/or active surfaces (adsorption sites) in the GC/MS interface can cause peak tailing and peak broadening. It is recommended that the GC column be fitted directly into the MS source. Graphite ferrules should be avoided in the GC injection area since they may adsorb TCDD. Vespel* or equivalent ferrules are recommended. 6.1.3 Mass Spectrometer The static resolution of the instrument must be maintained at a minimum 10,000 (10 percent valley). The mass spectrometer must be operated in a selected ion monitoring (SIM) mode with total cycle time (including voltage reset time) of one second or less (Section 8.3.4.1). At a minimum, the following ions which occur at these masses must be monitored: m/z 258.930, 319.897, 321.894, 331.937 and 333.934. 6.1.4 Data System A dedicated hardware or data system is employed to control the rapid multiple ion monitoring process and to acquire the data. Quantification data (peak areas or peak heights) and SIM traces (displays of intensities of each m/z being monitored as a function of time) must be acquired during the analyses. Quantifications may be reported based upon computer-generated peak areas or upon measured peak heights (chart recording). MOTE: Detector zero setting must allow peak-to-peak measurement of the noise on the base line. 6.2 GC Columns D-4 �For isomer-specific determinations of 2,3,7,8-TCDD, the following two fused silica capillary columns are recommended: a 60-m SP-2330 column and a 50-m CP-Sil 88 column. However, any capillary column which separates 2,3,7,8-TCDD from all other TCDDs may be used for such analyses, but this separation must be demonstrated and documented. Minimum acceptance criteria must be determined per Section 8.1. At the beginning of each 12-hour period (after mass resolution has been demonstrated) during which sample extracts or concentration calibration solutions will be analyzed, column operating conditions must be attained for the required separation on the column to be used for samples. Operating conditions known to produce acceptable results with the recommended columns are shown in Table 2 at the end of this Exhibit. 6.3 Miscellaneous Equipment 6.3.1 6.3.2 Balance capable of accurately weighing to 0.01 6.3.3 Centrifuge capable of operating at 2,000 rpm. 6.3.A Water bath — equipped with concentric ring cover and capable of being temperature-controlled within +2°C. 6.3.5 Stainless steel spatulas or spoons. 6.3.6 Stainless steel (or glass) pan large enough to hold contents of 1-pint sample containers. 6.3.7 Glove box. 6.3.8 6.4 Nitrogen evaporation apparatus with variable flow rate. Drying oven. g. Glassware 6.4.1 Soxhlet apparatus — all-glass, Kontes 6730-02 or equivalent; 90 mm x 35 mm glass thimble; 500-mL flask; condenser of appropriate size. 6.4.2 Kuderna-Danish apparatus — 500-mL evaporating flask, 10-mL graduated concentrator tubes with ground-glass stoppers, and 3-ball macro Snyder column (Kontes K-570001-0500, K-5030000121 and K-569001-0219 or equivalent). 6.4.3 Mini-vials — 1-mL borosilicate glass with conical-shaped reservoir and screw caps lined with Teflon-faced silicone disks. 6.4.4 Funnels — glass; appropriate size to accommodate filter paper used to filter jar extract (volume of approximately 170 mL) 6.4.5 Separatory funnel — 2000 mL with Teflon stopcock. D-5 �6.4.6 Continuous liquid-liquid extractors equipped with Teflon or glass connecting joints and stopcocks requiring no lubrication (Hershberg-Wolf Extractor - Ace Glass Company Vineland, NJ, P/N 6841-10 or equivalent). 6.4.7 Chromatographic columns for the silica and alumina chromatography — 1 cm ID x 10 cm long and 1 cm ID x 30 cm long. 6.4.8 Chromatography column for the Carbopak cleanup — disposable 5-mL graduated glass pipets, 7 mm ID. 6.4.9 Desiccator. 6.4.10 Glass rods. NOTE: Reuse of glassware should be minimized to avoid the risk of cross contamination. All glassware that is reused must be scrupulously cleaned as soon as possible after use, applying the following procedure. Rinse glassware with the last solvent used in it then with high-purity acetone and hexane. Wash with hot water containing detergent. Rinse with copious amounts of tap water and several portions of distilled water. Drain dry and heat in a muffle furnace at 400°C for 15 to 30 minutes. Volumetric glassware should not be heated in a muffle furnace, and some thermally stable materials (such as PCBs) may not be removed by heating in a muffle furnace. In these cases, rinsing with high-purity acetone and hexane may be substituted for muffle furnace heating. After the glassware is dry and cool, rinse with hexane, and store inverted or capped with solvent-rinsed aluminum foil in a clean environment. 7. REAGENTS AND STANDARD SOLUTIONS 7.1 Column Chromatography Reagents 7.1.1 Alumina, acidic — Extract the alumina in a Soxhlet with methylene chloride for 6 hours (minimum of 3 cycles per hour) and activate it by heating in a foil-covered glass container for 24 hours at 190"C. 7.1.2 Silica gel — high-purity grade, type 60, 70-230 mesh; extract the silica gel in a Soxhlet with methylene chloride for 6 hours (minimum of 3 cycles per hour) and activate it by heating in a foil-covered glass container for 24 hours at 130°C. 7.1.3 Silica gel impregnated with 40 percent (by weight) g-ulfuric acid — add two parts (by weight) concentrated sulfuric acid to three parts (by weight) silica gel (extracted and activated), mix with a glass rod until free of lumps, and store in a screw-capped glass bottle. D-6 �7.1.4 Sulfuric acid, concentrated — ACS grade, specific gravity 1.84. 7.1.5 Graphitized carbon black (Carbopack C or equivalent), surface of approximately 12 m^/g, 80/100 mesh — mix thoroughly 3.6 grams Carbopak C and 16.4 grams Celite 545* in a 40-mL vial. Activate at 130° C for six hours. Store in a desiccator. 7.1.6 Celite 545®, reagent grade, or equivalent. 7.2 Membrane filters or filter paper with pore size of <25 urn; rinse with hexane before use. 7.3 Glass wool, silanized — and air-dry before use. 7.4 Desiccating Agents extract with methylene chloride and hexane 7.4.1 Sodium sulfate — granular, anhydrous; before use, extract it with methylene chloride for 6 hours (minimum of 3 cycles per hour) and dry it for >4 hours in a shallow tray placed in an oven operated at 120°C. Let it cool in a desiccator. 7.4.2 Potassium carbonate—anhydrous, granular; use as such. 7.5 Solvents — high purity, distilled in glass: methylene chloride, toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent grade: tridecane. 7.6 Concentration calibration solutions (Table 1) — five tridecane solutions containing unlabeled 2,3,7,8-TCDD and C,«-l,2,3,4-TCDD (recovery standard) at varying concentrations, and Cj2~2>3,7,8-TCDD (internal standard, CASRN 80494-19-5) at a constant concentration must be used to calibrate the instrument. These concentration calibration solutions must be obtained from the Quality Assurance Division, US EPA Environmental Monitoring Systems Laboratory (EMSL-LV), Las Vegas, Nevada. However, additional secondary standards may be obtained from commercial sources, and solutions may be prepared in the contractor laboratory. Traceability of standards must be verified against EPAsupplied standard solutions. Such procedures will be documented by laboratory SOPs as required in IFB Pre-award Bid Confirmations, part 2 f ( ) It is the responsibility of the laboratory to ascertain that ..4. the calibration solutions received are indeed at the appropriate concentrations before they are injected into the instrument. Serious overloading of the instrument may occur if the concentration calibration solutions intended for a low-resolution MS are injected into the high-resolution MS. 7.6.1 The five concentration calibration solutions contain unlabeled 2,3,7,8-TCDD and labeled 13C,2~1,2,3,4-TCDD at nominal concentrations of 2.5, 5.0, 10.0, 20.0 and 40.0 pg/uL, respectively, and labeled C,2~2,3,7,8-TCDD at a constant nominal concentration of 10.0 pg/uL. D-7 �7.6.2 7.7 Store the concentration calibration solutions in 1-mL minivials at 4°C. Column performance check mixture — this solventless mixture must be obtained fvom the Quality Assurance Division, Environmental Monitoring Systems Laboratory, Las Vegas, Nevada, and dissolved by the Contractor in 1 mL tridecane. This solution will then contain the following components (including TCDDs (A) eluting closely to 2,3,7,8-TCDD, and the first- (F) and last-eluting (L) TCDDs when using the columns recommended in Section 6.2) at a concentration of 10 pg/uL of each of these isomers: Analyte Approximate Amount Per Ampule Unlabeled 2,3,7,8-TCDD 10 ng 13 10 ng 1,2,3,4-TCDD (A) 10 ng 1,4,7,8-TCDD (A) 10 ng 1,2,3,7-TCDD (A) 10 ng 1,2,3,8-TCDD (A) 10 ng 1,2,7,8-TCDD 10 ng 1,3,6,8-TCDD (F) 10 ng 1,2,8,9-TCDD (L) 10 ng C12-2,3,7,8-TCDD 7.8 Sample fortification solution — an isooctane solution containing the internal standard at a nominal concentration of 5 pg/uL. 7.9 Recovery standard spiking solution — an isooctane solution containing the recovery standard at a nominal concentration of 100 pg/uL. Five uL of this solution will be spiked into the extract before HRGC/HRMS analysis. 7.10 Internal standard spiking solution — an isooctane solution containing the internal standard at a nominal concentration of 100 pg/uL. Five uL of this solution will be added to a fortified field blank extract (Section 4.2.1.A.2, Exhibit E). 8. SYSTEM PERFORMANCE CRITERIA System performance criteria are presented in two sections. One section deals with GC column performance criteria while the other section consists of initial calibration criteria. The laboratory may use either of the recommended columns described in Section 6.2. It must be documented that D-8 �all applicable system performance criteria specified in Sections 8.1, 8.2 and 8.3 have been met before analysis of any sample is performed. Table 2 provides recommended conditions that can be used to satisfy the required criteria. Table 3 provides a typical 12-hour analysis sequence. 8.1 GC Column Performance 8.1.1 Inject 2 uL (Section 6.1.1) of the column performance check solution (Section 7.7) and acquire selected ion monitoring (SIM) data for m/z 258.930, 319.897, 321.894, 331.937 and 333.934 within a total cycle time of <1 second (Section 8.3.4.1). 8.1.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers must be resolved with a valley of ^25 percent, where Valley Percent - (x/y)(100) x * measured as in Figures 1 and 2 y « the peak height of 2,3,7,8-TCDD. It is the responsibility of the laboratory to verify the conditions suitable for the appropriate resolution of 2,3,7,8-TCDD from all other TCDD isomers. The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol thus defining the retention time window for total TCDD determination. The peaks representing 2,3,7,8-TCDD, the first and the last eluting TCDD isomers must be labeled and identified as such on the chromatograms. 8.2 Mass Spectrometer Performance 8.2.1 The mass spectrometer must be operated in the electron (impact) ionization mode. Static mass resolution of at least 10,000 (10 percent valley) must be demonstrated before any analysis of a set of samples is performed (Section 8.2.2). Static resolution checks must be performed at the beginning and at the end of each 12-hour period of operation. However, it is recommended that a visual check (i.e., not documented) of the static resolution be made using the peak matching unit before and after each analysis. 8.2.2 Chromatography time for TCDD may exceed the long-term mass stability of the mass spectrometer and thus mass drift correction is mandatory. A reference compound (high boiling PFK is recommended) is introduced into the mass spectrometer. An acceptable lock mass ion at any mass between m/z 250 and m/z 334 (m/z 318.979 from PFK is recommended) must be used to monitor and correct mass drifts. D-9 �NOTE: Excessive PFK may cause background noise problems and contamination of the source resulting in an increase in "downtime" for source cleaning. Using a PFK molecular leak, tune the instrument to meet the minimum required mass resolution of 10,000 (102 valley) at m/z 254.986 (or any other mass reasonably close to ra/z 259). Calibrate the voltage sweep at least across the mass range m/z 259 to m/z 334 and verify that m/z 330.979 from PFK (or any other mass close to m/z 334) is measured within +5 ppm (i.e., 1.7 mmu) using m/z 254.986 as a reference. Documentation of the mass resolution must then be accomplished by recording the peak profile of the PFK reference peak m/z 318.979 (or any other reference peak at a mass close to m/z 320/322). The format of the peak profile representation must allow manual determination of the resolution, i.e., the horizontal axis must be a calibrated mass scale (amu or ppm per division). The result of the peak width measurement (performed at 5 percent of the maximum) must appear on the hard copy and cannot exceed 31.9 mmu or 100 ppm. 8.3 Initial Calibration Initial calibration is required before any samples are analyzed for 2,3,7,8-TCDD. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.6. 8.3.1 All concentration calibration solutions listed in Table 1 must be utilized for the initial calibration. 8.3.2 Tune the instrument with PFK as described in Section 8.2.2. 8.3.3 Inject 2 uL of the column performance check solution (Section 7.7) and acquire SIM mass spectral data for m/z 258.930, 319.897, 321.894, 331.937 and 333.934 using a total cycle time of ^ 1 second (Section 8.3.4.1). The laboratory must not perform any further analysis until it has been demonstrated and documented that the criterion listed in Section 8.1.2 has been met. 8.3.4 Using the same GC (Section 8.1) and MS (Section 8.2) conditions that produced acceptable results with the column performance check solution, analyze a 2-uL aliquot of each of the 5 concentration calibration solutions in triplicate with the following MS operating parameters. 8.3.4.1 Total cycle time for data acquisition must be ^ 1 second. Total cycle time includes the sun of all the dwell times and voltage reset times. 8.3.4.2 Acquire SIM data for the following selected characteristic ions: D-10 �m/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD 321.894 Unlabeled TCDD 331.937 13 333.934 13 C12-2, 3, 7,8-TCDD, 13C12-1,2,3,4-TCDD C,, -2, 3, 7,8-TCDD,13C,,-1,2,3,4-TCDD 8.3.4.3 The ratio of integrated ion current for m/z 319.897 to m/z 321.894 for 2,3,7,8-TCDD must be between 0.67 and 0.90. 8.3.4.4 The ratio of integrated ion current for m/z 331.937 t to 13 m/z 333.934 for 13C12-2,3, 7,8-TCDD and 1 3C12-1, 2,3, 2,3,4nd C - 1 , TCDD must be between 0.67 and 0.90. 8.3.4.5 Calculate the relative response factors for unlabeled 2,3,7,8-TCDD [RRF(I)] relative to 13C,2-2,3,7,8-TCDD and for labeled C12-2,3,7,8-TCDD [RRF(II)] relative to 13C12-1,2,3,4-TCDD as follows: A RRF(I) - — * QIS — Qx * AIS A RRF(II) » is ^IS * *RS where Ax AIS « sum of the integrated ion abundances of m/z 319.897 and m/z 321.894 for unlabeled 2,3,7,8-TCDD. * 8Um of the integrated ion abundances of m/z 331.937 and m/z 333.934 for 13C12-2,3,7,8-TCDD. ARS • sum of the integrated ion abundances for m/z 331.937 and m/z 333.934 for 13C12-1,2,3,4-TCDD. QIg - Quantity of 13C12-2,3,7,8-TCDD injected (pg). QRg - quantity of 13C12~1,2,3,4-TCDD injected (pg). Qx - quantity of unlabeled 2,3,7,8-TCDD injected (pg). D-ll �RRF is a dimensionless quantity; the units used to express QXS> QRS must be the same. 8.4 an( * QX Criteria for Acceptable Calibration The criteria listed below for acceptable calibration must be met before analysis of any sample is performed. 8.4.1 The percent relative standard deviation (RSD) for the response factors from each of the triplicate analyses for both unlabeled and 13C12-2,3,7,8-TCDD must be less than +20 percent. 8.4.2 The variation of the 5 mean RRFs for unlabeled 2,3,7,8-TCDD obtained from the triplicate analyses must be less than +_20 percent RSD. 8.4.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise ratio of ^2.5 for m/z 258.930 and MO for m/z 321.894. SIM traces for 1"iCj2-2,3,7,8-TCDD must present a signal-tonoise ratio 2*0 for 333.934. 8.4.4 8.4.5 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within the allowed range. NOTE: If the criteria for acceptable calibration listed in Sections 8.4.1 and 8.4.2 have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The mean RRF from 5 triplicate determinations for unlabeled 2,3,7,8-TCDD and for 13C12-2,3,7,8-TCDD will be used for all calculations until routine calibration criteria (Section 8.6) are no longer met. At such time, new mean RRFs will be calculated from a new set of five triplicate determinations. 8.5 Routine Calibrations Routine calibrations must be performed at the beginning of a 12-hour period after successful mass resolution and GC column performance check runs. 8.5.1 8.6 Inject 2 uL of the concentration calibration solution which contains 5.0 pg/uL of unlabeled 2,3,7.8-TCDD, 10.0 pg/uL of 13C12-2,3,7,8-TCDD and 5.0 pg/uL 13C12-1,2,3,4-TCDD. Using the same GC/MS/DS conditions as used in Sections 8.1, 8.2 and 8.3, determine and document acceptable calibration as provided in Section 8.6. Criteria for Acceptable Routine Calibration The following criteria must be met before further analysis is performed. If these criteria are not met, corrective action must be taken and the instrument must be recalibrated. D-12 �8.6.1 8.6.2 The measured RRF for C12~2,3,7,8-TCDD must be within +20 percent of the mean value established by triplicate analysis of the concentration calibration solutions (Section 8.3.4.6). 8.6.3 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within the allowed range. 8.6.4 If one of the above criteria is not satisfied, a second attempt can be made before repeating the entire initialization process (Section 8.3). NOTE: 9. The measured RRF for unlabeled 2,3,7,8-TCDD must be within ^20 percent of the mean values established (Section 8.3.4.6) by triplicate analyses of concentration calibration solutions. An initial calibration must be carried out whenever any HRCC solution is replaced. QUALITY CONTROL PROCEDURES See Exhibit E for QA/QC requirements. 10. SAMPLE PRESERVATION AND HANDLING 10.1 Chain-of-custody procedures — see Exhibit G. 10.2 Sample Preservation 10.2.1 When received, each soil or sediment sample will be contained in a 1-pint glass jar surrounded by vermiculite in a sealed metal paint can. Until a portion is to be removed for analysis, store the sealed paint cans in a locked limited-access area where the temperature is maintained between 25° and 35°C. After a portion of a sample has been removed for analysis, return the remainder of the sample to its original container and store as stated above. 10.2.2 Each aqueous sample will be contained in a 1-liter glass bottle. The bottles with the samples are stored at 4°C in a refrigerator located in a locked limited-access area. 10.2.3 To avoid photodecomposition, protect samples from light. 10.3 Sample Handling CAUTION: Finely divided soils contaminated with 2,3,7,8-TCDD are hazardous because of the potential for inhalation or ingestion of particles containing 2,3,7,8-TCDD. Such samples should be handled in a confined environment (i.e., a closed hood or a glove box). 10.3.1 Pre-extraction sample treatment D-13 �10.3.1.1 Homogenization — Although sampling personnel will attempt to collect homogeneous samples, the contractor shall examine each sample and judge if it needs further mixing. NOTE: Contractor personnel have the responsibility to take a representative sample portion; this responsibility entails efforts to make the sample as homogeneous as possible. Stirring is recommended when possible. 10.3.1.2 Centrifugation — When a soil or sediment sample contains an obvious liquid phase, it must be centrifuged to separate the liquid from the solid phase. Place the entire sample in a suitable centrifuge bottle and centrifuge for 10 minutes at 2000 rptn. Remove the bottle from the centrifuge. With a disposable pipet, remove the liquid phase and discard it. Mix the solid phase with a stainless steel spatula and remove a portion to be air-dried, weighed and analyzed. Return the remaining solid portion to the original sample bottle and store it as described in 10.2.1. CAUTION: The removed liquid may contain TCDD and should be disposed as a liquid waste. 10.3.1.3 Weigh between 9.5 and 10.5 g of the air-dried soil sample (+0.5 g) to 3 significant figures. Dry it to constant weight at 100°C. Allow the sample to cool in a desiccator. Weigh the dried soil to 3 significant figures. Calculate and report percent moisture on Form B-l. 11. SAMPLE EXTRACTION 11.1 Soil Extraction 11.1.1 Immediately before use, the Soxhlet apparatus is charged with 200 to 250 mL benzene which is then refluxed for 2 hours. The apparatus is allowed to cool, disassembled and the benzene removed and retained as a blank for later analysis if required. 11.1.2 Accurately weigh to 3 significant figures a 10-g (9.50 g to 10.50 g) portion of the wet soil or sediment sample. Mix 100 uL of the sample fortification solution (Section 7.8) with 1.5 mL of acetone (500 pg of 13Cj2-2,3,7,8-TCDD) and deposit the entire mixture in small portions on several sites on the surface of the soil or sediment. 11.1.3 Add 10 g anhydrous sodium sulfate and mix thoroughly using a stainless steel spoon spatula. D-14 �11.1.4 After breaking up any lumps, place the soil-sodium sulfate mixture in the Soxhlet apparatus using a glass wool plug (the use of an extraction thimble is optional). Add 200 to 250 mL benzene to the Soxhlet apparatus and reflux for 24 hours. The solvent must cycle completely through the system at least 3 times per hour. 11.1.5 Transfer the extract to a Kuderna-Danish apparatus and concentrate to 2 to 3 uL. Rinse the column and flask with 5 mL benzene and collect the rinsate in the concentrator tube. Reduce the volume in the concentrator tube to 2 to 3 uL. Repeat this rinsing and concentrating operation twice more. Remove the concentrator tube from the K-D apparatus and carefully reduce the extract volume to approximately 1 mL with a stream of nitrogen using a flow rate and distance such that gentle solution surface rippling is observed. NOTE: 11.2 Glassware used for more than one sample must be carefully cleaned between uses to prevent cross-contamination (Note on page D-10). Extraction of Aqueous Samples 11.2.1 Mark the water meniscus on the side of the 1-L sample bottle for later determination of the exact sample volume. Pour the entire sample (approximately 1 L) into a 2-L separatory funnel. 11.2.2 Mix 100 uL of the sample fortification solution with 1.5 mL of acetone (500 pg of 13C12-2,3,7,8-TCDD) and add the mixture to the sample in the separatory funnel. NOTE: 11.2.3 A continuous liquid-liquid extractor may be used in place of a separatory funnel. Add 60 mL methylene chloride to the sample bottle, seal and shake 30 seconds to rinse the inner surface. Transfer the solvent to the separatory funnel and extract the sample by shaking the funnel for 2 minutes with periodic venting. Allow the organic layer to separate from the water phase for a minimum of 10 minutes. If the emulsion interface between layers is more than one-third the volume of the solvent layer, the analyst must employ mechanical techniques to complete the phase separation. Collect the methylene chloride (3 x 60 mL) directly into a 500 mL Kuderna-Danish concentrator (mounted with a 10 mL concentrator tube) by passing the sample extracts through a filter funnel packed with a glass wool plug and 5 g of anhydrous sodium sulfate. After the third extraction, rinse the sodium sulfate with an additional 30 mL of methylene chloride to ensure quantitiative transfer. D-15 �11.2.4 Attach a Snyder column and concentrate the extract until the apparent volume of the liquid reaches 1 tnL. Remove the K-D apparatus and allow it to drain and cool for at least 10 minutes. Remove the Snyder column, add 50 mL benzene, reattach the Snyder column and concentrate to approximately 1 mL. Rinse the flask and the lower joint with 1 to 2 mL benzene. Concentrate the extract to 1.0 mL under a gentle stream of nitrogen. 11.2.5 Determine the original sample volume by refilling the sample bottle to the mark and transferring the liquid to a 1000-mL graduated cylinder. Record the sample volume to the nearest 5 mL. 11.3 Cleanup Procedures - Column Cleanup 11.3.1 Prepare an acidic silica column as follows: Pack a 1 cm x 10 cm chromatographic column with a glass wool plug, a layer (1 cm) of Na£804/^003(1:1), 1.0 g silica gel (Section 7.1.2) and 4.0 g of 40-percent w/w sulfuric acid-impregnated silica gel (Section 7.1.3). Pack a second chromatographic column (1 cm x 30 cm) with a glass wool plug, 6.0 g acidic alumina (Section 7.1.1) and top with a 1-cm layer of sodium sulfate (Section 7.4). Add hexane to the columns until they are free of channels and air bubbles. 11.3.2 Quantitatively transfer the benzene extract (1 mL) from the concentrator tube to the top of the silica gel column. Rinse the concentrator tube with two 0.5-mL portions of hexane. Transfer the rinses to the top of the silica gel column. 11.3.3 Elute the extract from the silica gel column with 90 mL hexane directly into a Kuderna-Danish concentrator. Concentrate the eluate to 0.5 mL, using nitrogen blow-down as necessary. 11.3.4 Transfer the concentrate (0.5 mL) to the top of the alumina column. Rinse the K-D assembly with two 0.5-mL portions of hexane and transfer the rinses to the top of the alumina columns. Elute the alumina column with 18 mL hexane until the hexane level is just below the top of the sodium sulfate. Discard the eluate. Columns must not be allowed to reach dryness (i.e., a solvent "head" must be maintained.) 11.3.5 Place 30 mL of 20-percent (v/v) methylene chloride in hexane on top of the alumina and elute the TCDDs from the column. Collect this fraction in a 50-mL Erlenmeyer flask. 11.3.6 Certain extracts, even after cleanup by column chromatography, contain interferences which preclude determination of TCDD at low parts-per-trillion levels. Therefore, a cleanup step is included using activated carbon which selectively retains planar molecules such as TCDD. The TCDDs are then removed D-16 �from the carbon by elution with toluene. Proceed as follows: Prepare a 18-percent Carbopak C/Celite 545* mixture by thoroughly mixing 3.6 grams Carbopak C (80/100 mesh) and 16.4 grams Celite 545* in a 40-mL vial. Activate at 130°C for 6 hours. Store in a desiccator. Cut off a clean 5-mL disposable glass pipet at the 4-mL mark. Insert a plug of glass wool (Section 7.3) and push to the 2-mL mark. Add 340 mg of the activated Carbopak/Celite mixture followed by another glass wool plug. Using two glass rods, push both glass wool plugs simultaneously towards the Carbopak/Celite mixture and gently compress the Carbopak/Celite plug to a length of 2 to 2.5 cm. Preelute the column with 2 mL toluene followed by 1 mL of 75:20:5 methylene chloride/methanol/benzene, 1 mL of 1:1 cyclohexane in methylene chloride, and 2 mL hexane. The flow rate should be less than 0.5 mL min.~l. While the column is still wet with hexane, add the entire eluate (30 mL) from the alumina column (Section 11.3.5) to the top of the column. Rinse the Erlenmeyer flask which contained the extract twice with 1 mL hexane and add the rinsates to the top of the column. Elute the column sequentially with two 1-mL aliquots hexane, 1 mL of 1:1 cyclohexane in methylene chloride, and 1 mL of 75:20:5 methylene chloride/ methanol/benzene. Turn the column upside down and elute the TCDD fraction with 6 mL toluene into a concentrator tube. Warm the tube to approximately 60°C and reduce the toluene volume to approximately 1 mL using a stream of nitrogen. Carefully transfer the residue into a 1-mL mini-vial and again at elevated temperature, reduce the volume to about 100 uL using a stream of nitrogen. Rinse the concentrator tube with 3 washings using 200 uL of 12 toluene in CH2C12* Add 50 uL tridecane and store the sample in a refrigerator until GC/MS analysis is performed. 12. ANALYTICAL PROCEDURES 12.1 Remove the sample extract or blank from storage, allow it to warm to ambient laboratory temperature and add 5 uL recovery standard solution. With a stream of dry, purified nitrogen, reduce the extract/blank volume to 50 uL. 12.2 Inject a 2-uL aliquot of the extract into the GC, operated under the conditions previously used (Section 8.1) to produce acceptable results with the performance check solution. 12.3 Acquire SIM data according to 12.3.1. Use the same acquisition time and MS operating conditions previously used (Section 8.3.4) to determine the relative response factors. 12.3.1 Acquire SIM data for the following selected characteristic ions: D-17 �m/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD 321.894 Unlabeled TCDD 331.937 13 13 333.934 13 13 C12-2,3,7,8-TCDD, C12-2,3,7,8-TCDD, C12-1,2,3,4-TCDD C12-1,2,3,4-TCDD 12.4 Identification Criteria 12.4.1 The retention time (RT) (at maximum peak height) of the sample component ra/z 319.897 must be within -1 to +3 seconds of the retention time of the peak for the isotopically labeled internal standard at m/z 331.937 to attain a positive identification of 2,3,7,8-TCDD. Retention times of other tentatively identified TCDDs must fall within the RT window established by analyzing the column performance check solution (Section 8.1). Retention times are required for all chromatograms. 12.4.2 The ion current responses for m/z 258.930, 319.897 and 321.894 must reach maximum simultaneously (_+_ 1 scan), and all ion current intensities must be 2 2.5 times noise level for . positive identification of a TCDD. 12.4.3 The integrated ion current at m/z 319.897 must be between 67 and 90 percent of the ion current response at m/z 321.894. 12.4.4 The integrated ion current at m/z 331.937 must be between 67 and 90 percent of the ion current response at m/z 333.934. 12.4.5 The integrated ion currents for m/z 331.937 and 333.934 must reach their maxima within +_ 1 scan. 12.4.6 The recovery of the internal standard be between 40 and 120 percent. 13. 3 Cj2-2,3,7,8-TCDD must CALCULATIONS 13.1 Calculate the concentration of 2,3,7,8-TCDD (or any other TCDD isomer) using the formula: A X' A JS • W • RRF(I) D-18 �where: GX * unlabeled 2,3,7,8-TCDD (or any other unlabeled TCDD isomer) concentration in pg/g for soil/sediment and pg/L for aqueous samples. AX " sum of the integrated ion abundances determined for m/z 319.897 and 321.894. Ajg • sum of the integrated ion abundances determined for m/z 331.937 and 333.934 of r3C12-2,3, 7,8-TCDD (IS - internal standard). QIS * quantity (in picograms) of Cj2~2,3, 7,8-TCDD added to the sample before extraction (Qjg * 500 pg) . W * weight (in grams) of dry soil or sediment sample or volume of aqueous sample (in liters). RRF(I) - 13.2 calculated mean relative response factor for unlabeled 2,3,7,8TCDD relative to C,2~2, 3 , 7,8-TCDD. This represents the grand mean of the RRF(I)'s obtained in Section 8.3.4.5. Calculate the recovery of the internal standard 13C12~2» 3, 7,8-TCDD, measured in the sample extract, using the formula: Internal standard percent recovery • A,g * QRg —- x 100 ARS ' RRF(II) • where Ajg and Qjg have the same definitions as above (Section 13.1) ARg " sum of the integrated ion abundances determined for m/z 331.937 and 333.934 of C12-l ,2,3,4-TCDD (RS - recovery standard). QRS » quantity (in picograms) of ^C, 2~1 ,2,3,4-TCDD added to the sample residue before HRGC-HRMS analysis. (QRS - 500 pg). RRF(II) - calculated mean relative response factor for labeled C12-2, 3, 7,8TCDD relative to C,2-l ,2,3,4-TCDD. This represents the grand mean of the RRF(II)'s calculated in Section 8.3.4.5. 13.3 If the calculated concentration of unlabeled 2,3,7,8-TCDD exceeds 200 pg/g for soils or sediments, or 2000 pg/L for aqueous samples, the linear range of response vs. concentration may have been exceeded and a smaller portion of that sample must be analyzed. Accurately weigh to three significant figures a 1-g portion of the wet soil/ sediment. Add the sample fortification solution (Section 11.1.2), extract and analyze as discussed for the 10-g sample. Similarly, add the sample fortification solution (Section 11.2.2) to 100 mL of the aqueous sample, extract and analyze. D-19 �13.4 Total TCDD concentration — all positively identified isomers of TCDD must be within the RT window and meet all identification criteria listed in Sections 12.4.2, 12.4.3 and 12.4.4. Use the expression in Section 13.1 to calculate the concentrations of the other TCDD isomers, with Cx becoming the concentration of any unlabeled TCDD isoraer. Total TCDD 13.5 - Sum of the concentrations of the individual TCDDs. Estimated Detection Limit — For samples in which no unlabeled 2,3,7,8-TCDD was detected, calculate the estimated minimum detectable concentration. The background area is determined by integrating the ion abundances for m/z 319.897 and 321.894 in the appropriate region of the selected ion monitoring trace, multiplying that area by 2.5, and relating the product area to an estimated concentration that would produce that product area. Use the formula: (2.5) • (Ax) ' (QIS) (AIS) * (RRF(I)) ' (W) where G£ - estimated concentration of unlabeled 2,3,7,8-TCDD required to produce Ax. Ax - sum of integrated ion abundance for m/z 319.897 and 321.894 in the same group of >5 scans used to measure AI§. AIS » sum of integrated ion abundance for the appropriate ion characteristic of the internal standard, m/z 331.937 and m/z 333.934. Qlg, RRF(I), and W retain the definitions previously stated in Section 13.1. Alternatively, if peak height measurements are used for quantification, measure the estimated detection limit by the peak height of the noise in the TCDD RT window. 13.6 The relative percent difference (RPD) is calculated as follows: I Si - S2 | RPD - I Si - S2 | - Mean Concentration x 100 (Si + Si and S2 represent sample and duplicate sample results. References 1. "Carcinogens - Working with Carcinogens", Department of Health, Education and Welfare, Public Health Service, Center for Disease Control, National Institute for Occupational Safety and Health, Publication No. 77-206, Aug. 1977. D-20 �2. "OSHA Safety and Health Standards, General Industry" (29 CFR1910), Occupational Safety and Health Administration, OSHA 2206 (Revised January 1976). 3. "Safety in Academic Che'nistry Laboratories", American Chemical Society Publication, Committee on Chemical Safety, 3rd Edition 1979. D-21 �TABLE 1. COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS Recovery Standard Analyte 13 C12-1,2,3,4-TCDD 2,3,7,8-TCDD Internal Standard 13 C12-2,3,7,8-TCDD HRCC1 2.5 pg/uL 2.5 pg/uL 10.0 pg/uL HRCC2 5.0 pg/uL 5.0 pg/uL 10.0 pg/uL HRCC3 10.0 pg/uL 10.0 pg/uL 10.0 pg/uL HRCC4 20.0 pg/uL 20.0 pg/uL 10.0 pg/uL HRCC5 40.0 pg/uL 40.0 pg/uL 10.0 pg/uL Sample Fortification Solution 5.0 pg/uL of 13C12-2,3,7,8-TCDD Recovery Standard Spiking Solution 100 pg/uL 13 C12-1,2,3,4-TCDD Field Blank Fortification Solutions A) 5.0 pg/uL of unlabeled 2,3,7,8-TCDD B) 5.0 pg/uL of unlabeled 1,2,3,4-TCDD Internal Standard Spiking Solution 100 pg/uL of 13C12-2,3,7,8-TCDD (Used only in Section 4.2.1.A.2, Exhibit E) D-22 �TABLE 2. RECOMMENDED GC OPERATING CONDITIONS Column coating SP-2330 CP-SIL 88 Film thickness 0.2 urn 0.22 ura Column dimensions 60 m x 0.24 mm 50 m x 0.22 mm Helium linear velocity 28-29 cm/sec 28-29 cm/sec at 240eC at 240°C Initial temperature 70°C 45°C Initial time 4 min 3 min Temperature program 2,3,7,8-TCDD retention time Rapid increase to 200°C Rapid increase to 190°C 200CC to 250CC at 4°C/min 190CC to 2408C at 5°C/min 24 min 26 min D-23 �TABLE 3. 1. TYPICAL 12-HOUR SEQUENCE FOR 2,3,7,8-TCDD ANALYSIS Static mass resolution check 10/20/84 0700 hrs. 2. Column performance check 10/20/84 0730 hrs. 3. HRCC2 10/20/84 0800 hrs. 4. Sample 1 through Sample "N" 10/20/84 0830 hrs. 5. Column performance check 10/20/84 1800 hrs. 6. Static mass resolution check 10/20/84 1830 hrs. D-24 �2371 CT-SIl I (50 •) U7I M 12*7 If 4M Figure 1. •M Selected ion current profile for m/z 320 and 322 produced by MS analysis for performance check solution using a 50-m CP Sil-88 fused silica capillary column and conditions listed in Table 2. �IM 2371 I2J4 1271 K> If . «*• Figure 2. •0 • • M Selected ion current profile for m/z 320 and 322 produced by MS analysis of performance check solution using a 60-w 5P-2330 fused silica capillary column and conditions listed in Table 2. �APPENDIX B PROPOSED ANALYTICAL PROTOCOL for the Determination of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total TCDDs in Soil/Sediment and Water by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry December 1, 1985 This analytical protocol has been written in the format used in the Superfund program, as "Exhibit D" of a Statement of Work which in turn is part of an Invitation-for-Bid package under the Superfund Contract Laboratory Program. Also included are other exhibits listed below for the Statement of Work which have been tailored to meet the specific requirements of this protocol: EXHIBIT B: Reporting Requirements and Deliverables EXHIBIT C: Sample Rerun Requirements EXHIBIT D: Analytical Method EXHIBIT E: Quality Assurance/Quality Control Requirements �This protocol (Protocol B) is a modification of the protocol presented as Appendix A (Protocol A). Examination of the results from the single-laboratory evaluation of Protocol A had shown that the minimum amount of 2,3,7,8-TCDD that could be quantified under the conditions specified in Protocol A was 5 pg. However, a requirement existed to lower the quantisation limits to 2 ppt for soil and sediment samples and to 20 ppq for aqueous samples. The sample size should stay at 10 g for soil and sediments and at 1 L for aqueous samples, since the effect of larger sample sizes on the extract cleanup efficiencies is not known. Also, the range of the method should overlap with the 1-ppb lower limit of the low-resolution analytical method for TCDD used in the Superfund Contract Laboratory Program without necessitating second extractions for samples containing higher levels of TCDDs. After careful evaluation by EMSL-LV of the requirements and the options, the following protocol changes were made: o In Protocol B, the following calibration solutions will be used: HRCC1: 2 pg/yL 10 pg/yL HRCC2: 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD 13 C12-2,3,7,8-TCDD 13 2,3,7,8-TCDD and 10 pgM HRCC3: 10 pg/yL 13 50 pg/vL 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD 10 pg/yL 13 C12-1,2,3,4-TCDD C 12 -2,3,7,8-TCDD Cl2-2,3,7,8-TCDD �HRCC4: 100 pg/pl_ 10 pg/pL 2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD 13 C12-2,3,7,8-TCDD o In Protocol B, the final extract volume will be 10 pL. The decision to select a final volume of 10 pL was necessary in order to comply with the above requirements. It is realized that such a small volume may pose technical difficulties for the analyst. 0 In Protocol B, the fortification level of the internal standard 13 C12-2,3,7,8-TCDD was raised from 500 pg/sample to 1,000 pg/sample. This allows analysis of soil and sediment samples containing between 100 ppt and 1.2 ppb of any TCDD isomer and of water samples containing between 1 ppt and 12 ppt of any TCDD isomer by diluting a 2-pL aliquot of the remaining extract concentrate by a factor of 12 with a solution of the recovery standard (100 pg/pL of 13C12-1,2,3,4-TCDD in tridecane), Recoveries will be reported using the data generated from the first injection. Thus, the decision to dilute an aliquot of the 10-pL final extract will not be based on the concentration of 2,3,7,8-TCDD or total TCDD in the sample, but on the concentration of the most abundant TCDD isomer in the 10-pL final extract volume. This will eliminate unnecessary dilutions of the sample extract and analyses for samples containing between 100 ppt and 250 ppt for soil and sediment and 1 ppt and 2.5 ppt for water samples of a TCDD isomer, but for which the recoveries were low. �EXHIBIT B Reporting Requirements and Deliverables �1. SCOPE AND APPLICATION The Contractor shall provide reports and other deliverables as specified in the Contract Reporting Schedule. These reports are described below. All reports shall be submitted in legible form or resubmission shall be required. All reports and documentation required, including selected ion current profiles (also called selected ion monitoring traces), shall be clearly labeled with the Sample Management Office Case number and associated Sample/Traffic Report number(s). If documentation is submitted without the required identification, as specified above, resubmission shall be required. The Contract Reporting Schedule (Section 2) specifies the numbers of copies required, the delivery schedule and the distribution of all required deliverables. 1.1 Sample data package — Hard copy analytical data and documentation are required as described below. NOTE: This analytical protocol is designed for the receipt and analysis of samples by batches. Therefore, it is desired that sample data from samples in the same batch be reported together, i.e., on the same reporting form. However, contract accounting and billing are based on the sample unit. 1.1.1 Case narrative: Contains the Case number, Dioxin Shipment Record numbers, Contract number and detailed documentation of any quality control, sample, shipment and/or analytical problems encountered in a specific Case. Also included should be documentation of any internal decision tree process used along with a summary of corrective actions taken. The Case narrative must be signed in original signature by the Laboratory Manager or his designate. 1.1.2 Results of initial triplicate analyses of four (4) concentration calibration solutions (Form H-2), routine calibration solutions, (Form H-3), including all selected ion current profiles or selected ion monitoring (SIM) traces, calculated relative response factors (RRF), and computer-generated quantification reports (or manual calculations). 1.1.3 Completed data reporting sheets (Forms H-l, H-4, and H-5, H-8 and H-9) with appropriate SIM traces (including the lock mass SIM traces). Data results for levels less than 10 ppt but above the quantitation limit (Section 1.1, Exhibit D) attained for that sample shall be reported to two (2) significant figures; results greater than 10 ppt shall be reported to three (3) significant figures. Apply the rounding rules found in Section 7.2.2, "Handbook for Analytical Quality Control in Water and Wastewater Laboratories," EPA-600/4-79-019. Each SIM trace shall include computer-generated header information indicating instrumental (GC and MS) operating parameters during data B-l �acquisition. When samples are analyzed more than once, all sample data shall be reported. Rejected sample runs nust be separated and attached to the back of the data package and marked on the SIM trace as "Rejected," with an explanation of the reasons for the rejection. 1.1.A SIM traces generated during each GC column performance check analysis; peak profile outputs of the reference signal used to document the nass resolution. 1.1.5 Documentation of acceptable MS calibration (Section 8, Exhibit D, and Exhibit E) for each confirmatory analysis. As applicable, submit peak matching box settings and calculations for accurate nass assignments and any other related printouts. State, in ppn, the level of mass accuracy achieved (Section 8.2.2, Exhibit D). 1.1.6 A chronological list of all analyses performed (Form H-6). If more than one GC/MS system is used, a chronological list is required for each system. The list must provide the Data System File name, the EPA sample number, and (if appropriate) the contractor laboratory sample number for each sample, blank, concentration calibration solution, performance check solution, or other pertinent analytical data. This list shall specify date and time of beginning of analysis. All sample/ blank analyses performed during a 12-hour period must be accompanied by two GC column performance check solution analyses, one preceding and one following the sample/blank analyses. If multiple shifts are used, the ending GC column performance check sample analysis from one 12-hour period shall serve as the beginning analysis for the next 12-hour period; see Exhibit D, Section 8, for system performance criteria. The same schedule applies to the mass resolution check analysis. See Section 8.2.2, Exhibit D. 1.1.7 Verification of recovery of TCDDs from cleanup columns (Section 11.3, Exhibit D, and Section 4.2.1.2.2, Exhibit E). 1.2 Sample extracts and unused sample portions — Unused portions of samples and sample extracts shall be retained by the Contractor for a period of six months after receipt. When directed in writing by the Project Officer (PO) or Sample Management Office (SMO), the Contractor shall ship (not at Contractor's expense but in accordance with Department of Transportation Regulations) specific samples and/or extracts to specified locations and persons. After six months, upon obtaining PO or SMO clearance, remaining samples and extracts shall be disposed of by the Contractor at Contractor's expense, in accordance with applicable regulations concerning the disposal of such materials. 1.3 Document Control and Chain-of-Custody Package — The Document Control and Chain-of-Custody Package includes all laboratory records received B-2 �or generated for a specific case, that have not been previously submitted to EPA as a deliverable. These items include but are not limited to: sample tags, custody records, sample tracking records, analysts logbook pages, bench sheets, chromatographic charts, computer printouts, raw data summaries, instrument logbook pages, correspondence, and the document inventory (Exhibit G). NOTE: Pages from logbooks or bench sheets kept exclusively in a highhazard area (containment facility) need not be copied. 1.4 1.5 2. Monthly Sample Status Report — The Monthly Sample Status Report shall provide the status of all samples the Contractor has received or has had in-house during the calendar month. Required status information includes: samples received, samples extracted, samples analyzed, samples rerun, and samples which required special cleanup. All samples shall be identified by the appropriate EPA sample, case and batch/shipment numbers. Daily Sample Status Report — In response to a verbal request from the Sample Management Office or the Project Officer, the Contractor must verbally provide sample status information on a same-day basis. Should written confirmation be requested, the Contractor must send the daily sample status information in a written form that same day using first-class mail service. The required Daily Sample Status information shall include the items noted for the Monthly Sample Status Report and, in addition, shall require information on sample analysis reports in progress and analysis reports submitted/mailed. In accordance with applicable delivery requirements, the Contractor shall deliver specified items per the following Contract Reporting Schedule (Section 2.1). Recipients include the CLP Sample Management Office, the EMSL/LV QA Division, the appropriate Regional Technical Officer and NEIC. 2.1 Contract Reporting Schedule CONTRACT REPORTING SCHEDULE Item No. 1 2 Report No. Copies Sample Data Package Sample Extracts 3 Delivery Schedule 3 0 days after validated sample receipt date -OR1 0 days after initial data due date Within 180 days after analysis, 7 days after request by Project Officer or SMO SMO X X Report Distribution EMSL/LV Region NEIC X X X X As directed (Continued) B-3 �CONTRACT REPORTING SCHEDULE (Continued) Item No. 3 Report Delivery Schedule No. Copies Report Distribution SMO days after request by Project Officer or SMO Document 1 Control & Pkg Chain-ofCustody Package Monthly 2 Sample Status Report 5 7 EMSL/LV Region X NEIC X 5 days following end of each calendar month Daily Sample Status Report NOTE: Verbal and/or written upon request by SMO or PO; maximum frequency is daily. As directed All results shall be reported total and complete. 2.2 Addresses for distribution EMSL-LV SMO CLP Sample Management Office P. 0. Box 818 Alexandria, VA 22313 US EPA EMSL-LV QA Division Box 15027 Las Vegas, NV 89114 Attn: Data Audit Staff For overnight deliveries, use street address: 300 N. Lee St., Suite 200 Alexandria, VA 22314 NEIC US EPA NEIC Bldg. 53 Box 25227 Denver Fed. Center Denver, CO 80225 For overnight deliveries, use street address: 944 E. Harmon Ave. Executive Center Las Vegas, NV 89109 Regional Technical Officer — Following contract award and prior to Contractor's receipt of the first batch of samples, the Sample Management Office will provide the Contractor with the list of Technical Officers for the ten EPA Regions. SMO will provide the Contractor with updated Regional address/name lists as necessary throughout the period of the contract. 3. FORM INSTRUCTION GUIDE This section includes specific instructions for the completion of all required forms. These include instructions on header information as well as specific details to the bodies of individual forms. Instructions are arranged in the following order: B-4 �Data Summary (Form H-l) Initial Calibration Summary (Form H-2; 2 pages) Routine Calibration Summary (Form H-3) GC and Mass Resolution Check Summary (Form H-4) Quality Control Summary (Form H-5) Chronological List of All Analyses Performed (Form H-6) GC Operating Conditions (Form H-7) HRMS TCDD Calibration Report Form (Form H-8) High-Resolution MS TCDD Data Report Form (Form H-9) 3.1 Data Summary (Form H-l)— This form is used for summarizing the results from all samples in the batch. The detailed results are available on Form H-8 for each sample. Complete the header information at the top of the page, including laboratory name, case number and batch/shipment number (from the dioxin shipment record), and matrix (soil, sediment, water). Complete the form using one horizontal row for each sample. The SMO sample number should be suffixed with the appropriate letter code as needed. The TCDD retention time should be reported in minutes and seconds. TCDD levels are reported as parts per trillion (ppt) regardless of the matrix. Total TCDD concentration (in ppt) is the sum of the concentrations of all TCDDs reported on Form H-9. The S/N criteria apply to m/z 259, 320, 322 (for unlabeled TCDD) and m/z 322 and 334 (internal and recovery standards). The symbols used are: (+) all S/N ratios are 2.5 or greater including all TCDDs present, (-) S/N ratio for native 2,3,7,8-TCDD, the internal or the recovery standard are less than 2.5, (0) other suspected TCDDs are present but did not meet the S/N criteria. The file name is the HRGC/HRMS file name and is used for tracking results and raw data. The comments column should be used for any remarks specific to a particular sample. 3.2 Initial Calibration Summary (Form H-2): Page 1 — The header information should be filled in. The column headings are similar to those on Form H-l. A RRF(I) •= x ' QlS QX * AIS B-5 �RRF(II) - IS (Section 8.3.4.5, Exhibit D) A QlS ' RS Page 2 — The header information should be filled in. For each RRF, the mean, percent relative standard deviation (%RSD) and number of runs (N) are reported; N must be at least three (3) for each HRCC solution. The grand means (RRFs) are the mean of the individual means and are reported with their %RSD and N. The routine calibration relative response factor permissible ranges are also reported (Section 8.3.4.8, Exhibit D). 3.3 Routine Calibration Summary (Form H-3) — The header information includes case and batch numbers in addition to the laboratory and instrument identification. The columns are the same as on Page 1 of Form H-2. The results reported are for the routine calibration runs rather than the initial calibration. The calculated RRF(I) and RRF(II) must be within the routine calibration relative response factor permissible ranges (Section 8.3.4.8, Exhibit D) and other criteria listed in Section 8.6, Exhibit D must be met before further analysis is performed. 3.4 GC and Mass Resolution Check Summary (Form H-4) — The header information should be filled in. The TCDD isomer resolution (% valley) is measured from the column performance check solution (Section 8.1.2, Exhibit D). The resolving power and mass measurement error are measured using PFK (or equivalent) (Section 8.2, Exhibit D). 3.5 Quality Control Summary (Form H-5) — The items should be completed as indicated. The "other interferences" should be included even if they only occur at one mass. Form H-5 in conjunction with Form H-9 Is used to report results relative to the fortified field blank pair and rinsate analyses. The total TCDD retention time window is a window that includes all of the TCDD isomers and Is based on the first and last eluting isomers in the GC column performance check solution using the conditions summarized in Form H-7. All materials used should be recorded in the standard/reagent QC table. Standards provided by EPA should be listed, however, the QC columns may be left blank as these are reference materials. 3.6 Chronological List of All Analyses Performed (Form H-6) — The header information should be filled in. If more than one instrument is used, use one form per instrument. The "Analysis Identification" column should contain enough information for the data user to clearly identify the analysis, I.e., HRCC 2 Routine Calibration, Fortified Field Blank A, Fortified Field Blank B, B-6 �Reanalysis of Sample #1, 2, 3, 4, etc. The "SMO #" colunn should be used only for samples etc. which have an assigned SMO sample number. 3.7 GC Operating Conditions (Form H-7) — This form must be filled out to describe the GC operating conditions used to analyze a batch of samples and to analyze the GC performance evaluation check solution. 3.8 HRMS TCDD Calibration Report (Form H-8) — This form is to be filled in for each initial and routine calibration analysis made. It will be the first page of the chrotna tog rams and calculations for that analysis. It is suggested that this form be used as a worksheet for completing Forms H-2 and H-3. S/N ratios greater than five (5) may be reported with a (+); S/N ratios of five or less must have a numerical value reported with accompanying chromatograms scaled so that the measurements may be checked by the data user. 3.9 High-Resolution MS TCDD Data Report (Form H-9) — This form contains the details of the data reported in summary on Form H-l. It will be the first page of the chromatograms and calculations for each sample including the fortified field blank pair samples. All data presented (retention times, areas, and S/N ratios) must also be available on the accompanying chromatograms. The chromatograms must be scaled so that the data user may check any S/N ratios that are near or below five ( ) 5. It is suggested that this form be used as a worksheet for completing Form H-l. REPORTING REQUIREMENTS SUMMARY: Items that must be included with the data package: 4.1 Complete identification of the samples analyzed (sample numbers and type). 4.2 The dates and times at which all analyses were accomplished. This information should also appear on each selected ion current profile included with the report. 4.3 Raw mass chromatographic data which consist of the absolute peak heights or peak areas of the signals observed for the ion masses monitored. 4.4 The calculated ratios of the intensities of the M+0 to (M+2)+0 molecular ions for all TCDD isomers detected. 4.5 The calculated concentrations of native 2,3,7,8-TCDD and other TCDD isomers for each sample analyzed, expressed in picograms TCDD per gram of sample (that is, parts per trillion), as determined from the raw data. If no TCDDs are detected, the notation "Not Detected" or "N.D." is used, and the minimum detectable concentrations (or detection limits) are reported. B-7 �HIGH RESOLUTION FORM H-l DATA SUMMARY HRGC/HRMS DIOXIN ANALYSIS Batch/Shipment* SMO Sample Number TCDD 2.3.7.8 (IS) PPt 2378TCOO Total Meet. DL TCDD 320 322 Abundanca Ra«M 332 332 %Rac. S/N hiat. 334(18) 334(R8I (IS) Critaria ID Anatval* Data Time Ffc Nam* Commantt i oo Not Detected N- UnlabeledTCDDSpfce D- Duplicate FB- FtaWMank DtMB Hec - tj j „ ,j ^a^^t^ Namm oivnii laoova^f Matrix: S - SOU ^^^ O - OtfMf S/N Critaite: rapotl (+)MalS/l1 > 2.5 raport ( — I tf 2,3..8-TCDD. J ijf» 2.3.7.8-TCDDor 1.2.3.4 TCOD S/N < 2.8 raport»0)ifo*ar TCDO>ataauajMctad �HIGH RESOLUTION FORM H-2 INITIAL CALIBRATION SUMMARY Lab: CaMwation Standard Contract*: Rte Nama Data Tbna m/s m/i m/i 320/322* 332/334(15)* 332/334|RS|* Instrument ID: S/N Critaria RRRI)b w i VO * km ratio* mutt to in tfw ranga of 0.67 to 0.90 " 2.3.7.8-TCDO wnut "C12 2.3.7.8 TCDD e "C12-2.3.7.8-TCOD wnut '»C,2-1.2.3.4-TCDD 1of2 S/N Criteria: raport (+) if gtaatar than 2.B report (-) if tow than 2 B RRFfNf Commants �FORM H-2 HIGH RESOLUTION INITIAL CALIBRATION SUMMARY 2of2 Contract Instrument*: Data of IntrtW CfKbution: RRF (I) NWMI RRF (It) NWMI % RSD %RSD HRCC1 HRCC2 HRCC3 HRCC4 RRF (I) Grand NWMI: KRSD: I fc_* RRF (N) Grand MOT: %RSD: N: N: O Routfci RRF (l| = 2.3.7.8-TCOO va «Cw-2.3.7.8-TCOO Routfn RRF (ll| - «C -2.3.7.8-TCDO w» a Cll-1.2.3.4-TCDD N �HIGH RESOLUTION FORM H-3 ROUTINE CALIBRATION SUMMARY Cm«#: CaNbration Standard Rto Nama Data Batch*: m/i m/i m/z Thna 320/322* 332/334(15)* 332/334(RS|* * ton ratio* muat ba in the ranga of 0.67 to 0.90 • 2.3.7.8-TCOO vamn "C,2-2.3.7.8-TCDD c "C,, 2.3.7,8-TCDD warw* "C12-1.2.3.4-TCDD S/N Criteria Instrument ID: RRFU)" S/N Criteria: report (+) H graatarthan 2.5 report (-)H law than 2.S RRF^nl1 Comrrwnts �HIGH RESOLUTION FORM H-A GC AND MASS RESOLUTION CHECK SUMMARY Lib D«M Batch # Case* Intt. ID Sol. ID Tim* Flto N«m« 'M«M ut*d for ITWM mMiuffntnt vrror calculation B-12 TCDD Isomar R»«olution (SVtll^r) RaaoMng Pow*r MlOSValtoy *Mau M«MUi*awnt Error (PPM) �HIGH RESOLUTION FORM H-5 QUALITY CONTROL SUMMARY Lab: Batch # Case* Number of camples in bitch: _ MMn S of recovery for the I.S.: # of data points: Fortified field blank A, S recovery (1»C12-2.3.7.8-TCDD): SMO Sample «: Contamination by 1,2.3,4 TCDD Estimated Concentration (ppt) "C12-1,2.3,4-TCDD Retention times: Other interferences: Estimated concentrations (ppt): Fortified field blank B, % recovery C»C.,-1.2.3.4-TCDD|: "C12-2.3.7.8-TCDD Other interferences: SMO Sample *: B Estimated Concentration (ppt) Retention times: Estimated concentrations (ppt): Rlnaate. S recovery: Other TCDD Duplicate >n»ly»i». SMO Mmpl* #: e SMO sample #: Ves L_J Estimeted Concentration (PO/mL) "C.,-2.3,7.B-TCDD MMn Recovery: Percent Relative difference "C^^.S^.B-TCDD (Recovery) Percent relative difference "C12-2,3.7,8-TCDD (Concentration) Percent relative difference Total TCDD (Concentration) Method blank file name: Standard/Reagent Type Total TCDD retention time window from column performance check: HRMS Lab Number or Mfg.* Origin Date of QC OC File Name Results of QC Continue as needed B-13 �HIGH RESOLUTION FORM H-6 CHRONOLOGICAL LIST OF ALL ANALYSES PERFORMED Lab: Batch* Case* Instrument ID: File Name SMO Number Analysis Identification B-14 Oat* Time �HIGH RESOLUTION FORM H-7 GC OPERATING CONDITIONS Lab: Instrument ID: GC Column: Film Thickness: Column Dimensions: Initial Column Temperature: Temperature Program: Injector Temperature: Interface Temperature: Injection Mode: Injection Volume: Splrdess Valve Closed Time: Septum Purge Flow: Injector Sweep Flow: Carrier Gas Flow Rate (ml/min or cm/sec): B-15 �HIGH RESOLUTION FORM H-8 HRMS TCDD CALIBRATION REPORT FORM Lab: Calibration Solution: Case*: OC Column: — Batch /Shipment #: Date of Initial Calibration: Instrument ID: Calibration: Initial Routine 2.3,7,8-TCDD Time: Analysis Date: File Name Retention Time Ratios Area m/z 258.930 319.897 320 322 321.894 "C^.S.T.B-TCDD m/z 331.937 .222 334 333.934 "C12-1,2.3.4-TCDD m/z 331.937 m/z 333.934 <•) If S/N It grMttr than 6, wtttr ( +(; M MM than S, «nt»r th* nwMurad r*tk> B-16 S/N"> �HIGH RESOLUTION FORM H-9 HIGH RESOLUTION MS TCDD DATA REPORT FORM Lab: File Name: Batch/Shipment #: Instrument ID: Matrix: Water Circle _fine_ Aliquot Time: QC Column Percent Moiature Limit 2,3.7.8 TCDD: Report Date: Measured ppt 2.3.7.8-TCDD: Estimated Total TCDD (ppt): 2.3.7,8-TCDD m/z 268.930 319.897 321.894 Other. Soil Ratantion Time Ratios Araa 320 322 "C12-2.3.7,8-TCDD m/z331.937 333.934 "C12-1.2.3,4-TCDD m/t 331 .937 m/z 333.934 m •ercent Recovery "C12-2 ,3,7.8-TCDD: Other TCDDs Estimated Ratantion Time 320 322 S/N* 269 S/N* 320 S/N* 322 (PPt) •H S/N it greater than 6. enter (+); H leaf than 6 enter the meeaured ratio B-17 �4.6 The same raw and calculated data which are provided for the actual samples will also be reported for the duplicate analyses, the method blank analyses, the fortified field blank pair and rinsate analyses, and any other QA or performance sample analyzed in conjunction with the actual sample set(s). 4.7 The recoveries of the internal standard ( C^2~2,3,7,8-TCDD) in percent. 4.8 The calibration data, including relative response factors calculated from the calibration procedure described in Section 8.3, Exhibit D. Data showing that these factors have been verified at least once during each 12-hour period of operation must be included (Section 8.5, Exhibit D). Exact mass measurement error. Include peak matching box settings and calculations as appropriate. 4.9 The calculated dry weight of the original soil or sediment sample portion based on the dry weight determination of another sample portion of approximately equal wet weight. The exact volumes of the water and rinsate samples analyzed. 4.10 Documentation of the source of all TCDD standards used and available specifications on purity. 4.11 In addition, each report of analyses will include the following selcted ion current profiles: 1) those obtained from all samples analyzed, 2) those from each GC column performance check, and 3) those from the calibration solutions. The peak profile from each mass resolution check must also be part of the data package. 4.12 Identify which HRGC/HRMS system was used for the analyses (manufacturer and laboratory identification number of system - 01, 02, 03, etc.). 4.13 GC operating conditions such as type of GC column, film thickness, column dimensions, initial column temperature, temperature program, injector temperature, interface temperature, injection mode and volume, valve time (valve flush), septum purge flow, flow rate, and total injector flow should be provided (Form H-7). B-18 �EXHIBIT C Sample Rerun Requirements �1. SCOPE AND APPLICATION The Contractor shall be required to reextract and/or perform additional cleanup and reanalyze certain samples or batches of samples in a variety of situations that may occur in the process of contract performance. (For purposes of this contract, the term "sample rerun" shall indicate sample extraction of a fresh 10-g soil or sediment portion or 1-L aqueous sample, followed by cleanup and analysis, and the term "extract reanalysis" shall indicate analysis of another aliquot of the final extract. In situations where the sample rerun is required due to matrix effects, interferences or other problems encountered because of very complex samples, the Government will pay the Contractor for the sample reruns. Such sample reruns shall be billable and accountable under the specified contract allotment of automatic reruns. In situations where the sample rerun or the extract reanalysis is required due to Contractor materials, equipment or instrumentation problems, or lack of contractor's adherence to specified contract procedures, the sample rerun or extract reanalysis shall not be billable under the terms of the contract. Contractor's failure to perform any of the sample reruns or extract reanalyses specified herein, either billable or non-billable, shall be construed as Contractor nonperformance and may result in termination of the contract for default by the Contractor. 2. Required Sample Reruns and Extract Reanalyses 2.1 Automatic sample reruns and extract reanalyses that may be billable as such under the contract. 2.1.1 2.1.2 13 If the percent recovery for the internal standard Ci2~2,3,7,8TCDD was outside of the acceptance limits of MO percent and <120 percent, the Contractor shall reextract and reanalyze the sample. If the percent recovery for the sample rerun is still outisde the acceptance limits, then both analyses can be billed if the recoveries from both analyses are either <40% or >120%. If, however, the percent recovery for the sample rerun is within the acceptance limits, or if it is still outside the acceptance limits but the percent recoveries from the original analysis and the sample rerun are not both either <40% or >120%, then the sample rerun may not be billed. If the internal standard was not found upon monitoring m/z 331.937 and 333.934, the Contractor shall reextract and reanalyze the sample. If the internal standard is not found in the sample rerun, the sample rerun is billable. If the internal standard is found in the sample rerun, then the sample rerun is not billable. C-l �2.1.3 If either one of the isotope abundance ratios for m/z 319.897/ 321.894 or for 331.937/333.934 is less than 0.67 or greater than 0.90 and all other criteria contained in Section 12.4 of Exhibit D are met, then the extract shall be reanalyzed. If both ion abundance ratios now meet the criterion, these values shall be reported as the isotope abundance ratios, and the Contractor shall not bill the Government for the extract reanalysis. If the ratio in question is still outside the criterion, the Contractor shall rerun the sample (Section 7.2, Exhibit E). If either one of the ratios determined from the sample rerun is still outside the acceptance limits, then both runs and the extract reanalysis can be billed if the corresponding isotope abundance ratios from both runs are either <0.67 or >0.90. If, however, both isotope abundance ratios from the sample rerun meet the criteria, or if both corresponding isotope abundance ratios from the original run and the sample rerun are not both either <0.67 or >0.90, then the extract reanalysis and the sample rerun may not be billed. 2.1.4 If the recoveries of 2,3,7,8-TCDD (Section 4.2.1.1.3.1, Exhibit E) and/or 1,2,3,4-TCDD (Section 4.2.1.2, Exhibit E) in the fortified field blank pair are <40% or >120%, the Contractor shall reextract and reanalyze a second portion of the field blank sample (Section 4.2, Exhibit E). If the percent recoveries for the sample rerun are still outside the acceptance limits, then both analyses can be billed as long as the recoveries from both analyses are either <40% or >120%. If, however, the percent recoveries for the sample rerun are within the acceptance limits, or if they are still outside the acceptance Units but the percent recoveries from the original run and the sample rerun are not both either <40% or >120%, then the sample rerun may not be billed. NOTE: 2.2 Fortified field blanks as described in Sections 4.2.1.1.4 and 4.2.1.2.2, Exhibit E, can never be billed. Automatic sample extract dilution and HRGC/HRMS analysis, billable as such under the Contract. If any individual or group of coeluting TCDD isomer concentrations in the 10-uL final extract exceeds 100 pg/uL, the analyst will perform a dilution as specified in Section 13.3, Exhibit D, and reanalyze the diluted portion using HRGC/HRMS. 2.3 Sample reruns and/or extract reanalyses to be performed at Contractor's expense (i.e., not billable under the terms of the contract). 2.3.1 If the method blank contains any signal in the TCDD retention time window at or above the method quantltation limit (2 ppt C-2 �for soil and sediment and 20 ppq for aqueous samples), the Contractor shall rerun all positive samples in the batch of samples (Section 4.1.2, Exhibit E). 2.3.2 If the system performance using the GC column performance check (PC) solution does not meet specified criteria, the Contractor shall take corrective action, demonstrate acceptable GC column performance, and reanalyze the extracts from all positive samples run during the time period between the last acceptable PC run and the unacceptable PC run (Section 2.4, Exhibit E). 2.3.3 If a false positive is reported for an uncontaminated soil (blind QC) sample, upon notification by the Sample Management Office the Contractor shall reextract and reanalyze all samples reported as positive in the associated batch of samples (Section 8.1.1, Exhibit E). 2.3.4 If the analysis results for a performance evaluation blind QC sample fall outside of EPA-established acceptance windows, upon notification of the Sample Management Office the Contractor shall reextract and reanalyze the entire associated batch of samples (Section 8.4.1, Exhibit E). 2.3.5 If the isotope abundance ratio for m/z 319.897/321.894 or for 331.937/333.934 is less than 0.67 or greater than 0.90, and all other criteria contained in Section 12.4 of Exhibit D are met, then the extract shall be reanalyzed. If the ion abundance ratio in question now meets the criterion, this value shall be reported as the isotope abundance ratio, and the Contractor shall not bill the Government for the extract reanalysis. 2.3.6 If the system performance mass resolution check does not meet the specified criterion, the Contractor shall take corrective action, demonstrate acceptable mass resolution and reanalyze the extract from all positive samples analyzed during the time period between the last acceptable mass resolution check and the unacceptable mass resolution check (Section 2.4, Exhibit E). C-3 �EXHIBIT D Analytical Method 2,3,7,8-Tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and Total TCDDs in Soil/Sediment and Water by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry �EXHIBIT D Section Subject Page 1 Scope and Application. D-l 2 Summary of Method D-l 3 Definitions D-2 4 Interferences D-3 5 Safety D-3 6 Apparatus and Equipment D-4 7 Reagents and Standard Solutions. ...... D-6 8 System Performance Criteria D-9 9 Quality Control Procedures . D-14 10 Sample Preservation and Handling D-14 11 Sample Extraction D-15 12 Analytical Procedures D-18 13 Calculations D-19 �1. SCOPE AND APPLICATION 1.1 This method provides procedures for the detection and quantitative measurement of 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD; CAS Registry Number 1746-01-6; Storet number 3475) at concentrations of 2 pg/g (2 parts per trillion) to 100 pg/g (100 parts per trillion) in 10-g portions of soil and sediment and at 20 pg/L (20 parts per quadrillion) to 1000 pg/L (1 part per trillion) in 1-L samples of water. Dilution of an aliquot of the final extract permits measurement of concentrations up to 1.2 ng/g (1.2 parts per billion) or 12 ng/L (12 parts per trillion), respectively. This method also allows the estimation of quantities of total TCDD present in the sample. Samples containing concentrations of any individual TCDD isomer or group of coeluting TCDD isomers greater than 1.2 ng/g or 12 ng/L must be analyzed by a protocol designed for such concentration levels, with an appropriate instrument calibration range. 1.2 The minimum measurable concentration is estimated to be 2 pg/g (2 parts per trillion) for soil and sediment samples and 20 pg/L (20 parts per quadrillion) for water samples, but this depends on kinds and concentrations of interfering compounds in the sample matrix. 1.3 This method is designed for use by analysts who are experienced in the use of high-resolution gas chromatography/high-resolution mass spectrometry. CAUTION: 2. TCDDs are assumed to be extremely hazardous. It is the laboratory's responsibility to ensure that safe handling procedures are employed. SUMMARY OF METHOD One thousand pg of 13Ci2-2,3,7,8-TCDD (internal standard) are added to a 10-g portion of a soil/sediment sample (weighed to 3 significant figures) or a 1-L aqueous sample, and the sample is extracted with 200 to 250 mL benzene using a Soxhlet apparatus for soils and sediments with a minimum of 3 cycles per hour, or with methylene chloride using a continuous liquidliquid extractor for aqueous samples for 24 hours. A separatory funnel and 3 x 60 mL methylene chloride may also be used for aqueous samples. After appropriate cleanup, 10 uL of a tridecane solution of the recovery standard ( C12~l,2,3,4-TCDD) are added to the extract which is then concentrated to a final volume of 10 uL. One to three uL of the concentrated extract is injected into a gas chromatograph with a capillary column interfaced to a high-resolution mass spectrometer capable of rapid multiple ion monitoring at resolutions of at least 10,000 (10 percent valley). Identification of 2,3,7,8-TCDD is based on the detection of the ions in/z 319.897 and 321.894 at the same GC retention time and within -1 to +3 seconds GC retention time of the internal standard masses of m/z 331.937 and 333.934. Confirmation of 2,3,7,8-TCDD (and of other TCDD isomers) is D-l �based on the Ion m/z 258.930 which results from loss of COCL by the parent molecular ion. 3. DEFINITIONS 3.1 Concentration calibration solutions — solutions containing known amounts of the analyte (unlabeled 2,3,7,8-TCDD), the internal standard 13 C12-2,3,7,8-TCDD and the recovery standard C,2~1,2,3,4-TCDD; they are used to determine instrument response of the analyte relative to the internal standard and of the internal standard relative to the recovery standard. 3.2 Field blank — a portion of soil/sediment or water uncontaminated with 2,3,7,8-TCDD and/or other TCDDs. 3.3 Rinsate — a portion of solvent used to rinse sampling equipment; the rinsate is analyzed to demonstrate that samples have not been contaminated during sampling. 3.4 Internal standard — 13C12-2,3,7,8-TCDD, which is added to every sample (except the blank described in Sections 4.2.1 of Exhibit E) and is present at the same concentration in every method blank and quality control sample. It is added to the soil/sediinent or aqueous sample before extraction and is used to measure the concentration of each analyte. Its concentration is measured in every sample, and percent recovery is determined using an internal standard method. 3.5 Recovery standard — 13ci2-1»2,3,4-TCDD which is added to every sample extract (except for the blank discussed in Sections 4,2.1, Exhibit E) just before the final concentration step and HRGC-HRMS analysis. 3.6 Laboratory method blank — this blank is prepared in the laboratory through performing all analytical procedures except addition of a sample aliquot to the extraction vessel. 3.7 GC column performance check mixture — a mixture containing known amounts of selected standards; it is used to demonstrate continued acceptable performance of the capillary column, i.e., separation (<^ 25% valley) of 2,3,7,8-TCDD isomer from all other 21 TCDD isomers, and to define the TCDD retention time window. 3.8 Performance evaluation sample — a soil, sediment or aqueous sample containing a known amount of unlabeled 2,3,7,8-TCDD and/or other TCDDs. It is distributed by the EMSL-LV to potential contractor laboratories who must analyze it and obtain acceptable results before being awarded a contract for sample analyses (see IFB Pre-Award Bid Confirmations). It nay also be included as an unspecified ("blind") QC sample in any sample batch submitted to a laboratory for analysis. 3.9 Relative response factor — response of the mass spectrometer to a known amount of an analyte relative to a known amount of an internal standard. D-2 �3.10 Mass resolution check — standard method used to demonstrate static resolution of 10,000 minimum (10% valley definition). 3.11 Positive response for a blank — defined as a signal in the TCDD retention time window, at any of the masses monitored, which is equivalent to or above the method quantitation limit (2 ppt for soil and sediment, and 20 ppq for aqueous samples). 3.12 Sample rerun — extraction of another 10-g soil or sediment sample portion or 1-L aqueous sample, followed by extract cleanup and extract analysis. 3.13 Extract reanalysis — 4. analysis of another aliquot of th final extract. INTERFERENCES Chemicals which elute from the GC column within +10 scans of the internal and/or recovery standard (m/z 331.937 and 333.934) and which produce within the TCDD retention time window ions at any of the masses used to detect or quantify TCDD are potential interferences. Most frequently encountered potential interferences are other sample components that are extracted along with TCDD, e.g. PCBs, chlorinated methoxybiphenyls, chlorinated hydroxydiphenylethers, chlorinated benzylphenylethers, chlorinated naphthalenes, DDE, DDT, etc. The actual incidence of interference by these chemicals depends also upon relative concentrations, mass spectrometric resolution, and chronatographic conditions. Because very low levels of TCDDs must be measured, the elimination of interferences is essential. High-purity reagents and solvents must be used and all equipment must be scrupulously cleaned. Blanks (Exhibit E, Quality Control, Section 4) must be analyzed to demonstrate absence of contamination that would interfere with TCDD measurement. Column chromatographic procedures are used to remove some coextracted sample components; these procedures must be performed carefully to minimize loss of TCDDs during attempts to increase their concentration relative to other sample components. 5. SAFETY The toxicity or carcinogen!city 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 file of current 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. Additional references to laboratory safety are identified (1-3) (page D-21). 2,3,7,8-TCDD has been identified as a suspected human or mammalian carcinogen. The laboratory is responsible for ensuring that safe handling procedures are followed. D-3 �6. APPARATUS AND EQUIPMENT 6.1 High-Resolution Gas Chromatograph/High-Resolution Mass Spectrometer/Data System (HRGC/HRMS/DS) 6.1.1 The GC must be equipped for temperature programming, and all required accessories must be available, such as syringes, gases, and a capillary column. The GC injection port must be designed for capillary columns. The use of splitless injection techniques is recommended. On-column injection techiques can be used but this may severely reduce column lifetime for nonchemically bonded columns. When using the method in this protocol, a 2-uL injection volume is used consistently. With some GC injection ports, however, 1-uL injections may produce improved precision and chromatographic separation. A 1- to 3-uL injection volume may be used if adequate sensitivity and precision can be achieved. NOTE: If 1 uL or 3 uL is used at all as injection volume, the injection volumes for all extracts, blanks, calibration solutions and the performance check sample must be 1 uL or 3 uL. 6.1.2 Gas Chromatograph-Mass Spectrometer Interface The GC-MS interface may include enrichment devices, such as a glass jet separator or a silicone membrane separator, or the gas chromatograph can be directly coupled to the mass spectrometer ion source. The interface may include a diverter valve for shunting the column effluent and isolating the mass spectrometer ion source. All components of the interface should be glass or glass-lined stainless steel. The interface components should be compatible with 300°C temperatures. The GC/MS interface must be appropriately designed so that the separation of 2,3,7,8-TCDD from the other TCDD isomers which is achieved in the gas chromatographic column is not appreciably degraded. Cold spots and/or active surfaces (adsorption sites) in the GC/MS interface can cause peak tailing and peak broadening. It is recommended that the GC column be fitted directly into the MS ion source. Graphite ferrules should be avoided in the GC injection port since they may adsorb TCDD. Vesper" or equivalent ferrules are recommended. 6.1.3 Mass Spectrometer The static resolution of the instrument must be maintained at a minimum 10,000 (10 percent valley). The mass spectrometer must be operated in a selected ion monitoring (SIM) mode with total cycle time (including voltage reset time) of one second or less (Section 8.3.4.1). At a minimum, the following ions which occur at these masses must be monitored: m/z 258.930, 319.897, 321.894, 331.937 and 333.934. D-4 �6.1.4 Data System A dedicated hardware or data system is employed to control the rapid multiple ion monitoring process and to acquire the data. Quantification data (peak areas or peak heights) and SIM traces (displays of intensities of each m/z being monitored as a function of time) must be acquired during the analyses. Quantifications may be reported based upon computer-generated peak areas or upon measured peak heights (chart recording). NOTE: Detector zero setting must allow peak-to-peak measurement of the noise on the base line. 6.2 GC Columns For isomer-specific determinations of 2,3,7,8-TCDD, the following fused silica capillary columns are recommended: a 60-m SP-2330 (SP2331) column and a 50-m CP-Sil 88 column. However, any capillary column which separates 2,3,7,8-TCDD from all other TCDDs may be used for such analyses, but this separation must be demonstrated and documented. Minimum acceptance criteria must be determined per Section 8.1. At the beginning of each 12-hour period (after mass resolution has been demonstrated) during which sample extracts or concentration calibration solutions will be analyzed, column operating conditions must be attained for the required separation on the column to be used for samples. Operating conditions known to produce acceptable results with the recommended columns are shown in Table 2 at the end of this Exhibit. 6.3 Miscellaneous Equipment 6.3.1 6.3.2 Balance capable of accurately weighing to ^0.01 g. 6.3.3 Centrifuge capable of operating at 2,000 rpm. 6.3.4 Water bath — equipped with concentric ring cover and capable of being temperature-controlled within jf2°C. 6.3.5 Stainless steel spatulas or spoons. 6.3.6 Stainless steel (or glass) pan large enough to hold contents of 1-pint sample containers. 6.3.7 Glove box. 6.3.8 6.4 Nitrogen evaporation apparatus with variable flow rate. Drying oven. Glassware 6.4.1 Soxhlet apparatus — all-glass, Kontes 6730-02 or equivalent; D-5 �90 mm x 35 mm glass thimble; 500-mL flask; condenser of appropriate size. 6.4.2 Kuderna-Danish apparatus — 500-mL evaporating flask, 10-mL graduated concentrator tubes with ground-glass stoppers, and 3-ball macro Snyder column (Kontes K-570001-0500, K-5030000121 and K-569001-0219 or equivalent). 6.4.3 Mini-vials — 1-mL borosilicate glass with conical-shaped reservoir and screw caps lined with Teflon-faced silicone disks. 6.4.4 Funnels — glass; appropriate size to accommodate filter paper used to filter jar extract (volume of approximately 170 mL), 6.4.5 Separatory funnel -- 2000 mL with Teflon stopcock. 6.4.6 Continuous liquid-liquid extractors equipped with Teflon or glass connecting joints and stopcocks requiring no lubrication (Hershberg-Wolf Extractor - Ace Glass Company, Vineland, NJ; P/N 6841-10 or equivalent). 6.4.7 Chromatographic columns for the silica and alumina chromatography — 1 cm ID x 10 cm long and 1 cm ID x 30 cm long. 6.4.8 Chromatographic column for the Carbopak cleanup — disposable 5-mL graduated glass pipets, 6 to 7 mm ID. 6.4.9 Desiccator. 6.4.10 Glass rods. NOTE: Reuse of glassware should be minimized to avoid the risk of cross contamination. All glassware that is reused must be scrupulously cleaned as soon as possible after use, applying the following procedure. Rinse glassware with the last solvent used in it then with high-purity acetone and hexane. Wash with hot water containing detergent. Rinse with copious amounts of tap water and several portions of distilled water. Drain, dry and heat in a muffle furnace at 400°C for 15 to 30 minutes. Volumetric glassware must not be heated in a muffle furnace, and some thermally stable materials (such as PCBs) may not be removed by heating in a muffle furnace. In these two cases, rinsing with highpurity acetone and hexane may be substituted for muffle-furnace heating. After the glassware is dry and cool, rinse with hexane, and store inverted or capped with solvent-rinsed aluminum foil in a clean environment. 7. REAGENTS AND STANDARD SOLUTIONS 7.1 Column Chromatography Reagents D-6 �7.1.1 Alumina, acidic — extract the alumina in a Soxhlet with methylene chloride for 6 hours (minimum of 3 cycles per hour) and activate it by heating in a foil-covered glass container for 24 hours at 190°C. 7.1.2 Silica gel -- high-purity grade, type 60, 70-230 mesh; extract the silica gel in a Soxhlet with methylene chloride for 6 hours (minimum of 3 cycles per hour) and activate it by heating in a foil-covered glass container for 24 hours at 130°C. 7.1.3 Silica gel impregnated with 40 percent (by weight) sulfuric acid — add two parts (by weight) concentrated sulfuric acid to three parts (by weight) silica gel (extracted and activated), mix with a glass rod until free of lumps, and store in a screw-capped glass bottle. 7.1.4 Sulfuric acid, concentrated — 7.1.5 Graphitized carbon black (Carbopack C or equivalent), surface of approximately 12 m^/g, 80/100 mesh — mix thoroughly 3.6 grams Carbopak C and 16.4 grams Celite 545® in a 40-mL vial. Activate at 130°C for six hours. Store in a desiccator. 7.1.6 Celite 545*, reagent grade, or equivalent. ACS grade, specific gravity 1.84. 7.2 Membrane filters or filter paper with pore size of <25 urn; rinse with hexane before use. 7.3 Glass wool, silanized — extract with raethylene chloride and hexane and air-dry before use. 7.4 Desiccating Agents 7.4.1 Sodium sulfate — granular, anhydrous; before use, extract it with methylene chloride for 6 hours (minimum of 3 cycles per hour) and dry it for >4 hours in a shallow tray placed in an oven at 120°C. Let it cool in a desiccator. 7.4.2 Potassium carbonate—anhydrous, granular; use as such. 7.5 Solvents — high purity, distilled in glass: methylene chloride, toluene, benzene, cyclohexane, methanol, acetone, hexane; reagent grade: tridecane. 7.6 Concentration calibration solutions (Table 1) — four tridecane solutions containing Cj2~l»2»3,4-TCDD (recovery standard) and unlabeled 2,3,7,8-TCDD at varying concentrations, and ^c^-2,3,7,8TCDD (internal standard, CAS RN 80494-19-5) at a constant concentration must be used to calibrate the instrument. These concentration calibration solutions must be obtained from the Quality Assurance Division, US EPA, Environmental Monitoring Systems Laboratory (EMSL-LV), Las Vegas, Nevada. However, additional secondary standards may be obtained D-7 �from commercial sources, and solutions may be prepared in the contractor laboratory. Traceability of standards must be verified against EPAsupplied standard solutions. Such procedures will be documented by laboratory SOPs as required in IFB Pre-award Bid Confirmations, part 2.f.(4). It is the responsibility of the laboratory to ascertain that the calibration solutions received are indeed at the appropriate concentrations before they are injected into the instrument, NOTE: Serious overloading of the instrument may occur if the concentration calibration solutions intended for a low-resolution MS are injected into the high-resolution MS. 7.6.1 7.6.2 7.7 The four concentration calibration solutions contain unlabeled 2,3,7,8-TCDD and labeled ' »2 ,3 ,4-TCDD at nominal concenl trations of 2.0, 10.0, 50.0, and 100 pg/uL, respectively, and labeled ~2 ,3 , 7 ,8-TCDD at a constant nominal concentration of 10.0 pg/uL. Store the concentration calibration solutions in 1-mL minivials at 4°C. Column performance check mixture — this solventless mixture must be obtained from the Quality Assurance Division, Environmental Monitoring Systems Laboratory, Las Vegas, Nevada, and dissolved by the Contractor in 1 mL tridecane. This solution will then contain the following components [including TCDDs (A) eluting closely to 2,3,7,8-TCDD, and the first- (F) and last-eluting (L) TCDDs when using the columns recommended in Section 6.2] at a concentration of 10 pg/uL of each of these isomers: Approximate Amount Per Ampule Analyte Unlabeled 2,3,7,8-TCDD 10 ng 13 10 ng 1,2,3,4-TCDD (A) 10 ng 1,4,7,8-TCDD (A) 10 ng 1,2,3,7-TCDD (A) 10 ng 1,2,3,8-TCDD (A) 10 ng 1,3,6,8-TCDD (F) 10 ng 1,2,8,9-TCDD (L) 10 ng C12-2,3,7,8-TCDD 7.8 Sample fortification solution — an isooctane solution containing the internal standard at a nominal concentration of 10 pg/uL. D-8 �7.9 Recovery standard spiking solution — a tridecane solution containing the recovery standard at a nominal concentration of 10 pg/uL. Ten uL of this solution will be spiked into each sample extract (except for the fortified field blank A) before the final concentration step and HRGC/HRMS analysis. It is also used for the dilution of the extracts from samples with high TCDD levels (Section 13.3, Exhibit D). 7.10 Internal standard spiking solution — a tridecane solution containing the internal standard ( C^~2,3,7,8-TCDD) at a nominal concentration of 10 pg/uL. Ten uL of this solution will be added to a fortified field blank extract (Section 4.2.1.1, Exhibit E). This is the only case where Cj22,3,7,8-TCDD is used for recovery purposes. 7.11 Field blank fortification solutions — the following TCDD isomers: Solution A: Solution B: 8. isooctane solutions containing 10.0 pg/uL of unlabeled 2,3,7,8-TCDD 10.0 pg/uL of unlabeled 1,2,3,4-TCDD. SYSTEM PERFORMANCE CRITERIA System performance criteria are presented below. The laboratory may use any of the recommended columns described in Section 6.2. It must be documented that all applicable system performance criteria specified in Sections 8.1, 8.2, 8.3 and 8.5 have been met before analysis of any sample is performed. Table 2 provides recommended conditions that can be used to satisfy the required criteria. Table 3 provides a typical 12-hour analysis sequence. The GC column performance and mass resolution checks must be performed at the beginning and end of each 12-hour period of operation. 8.1 GC Column Performance 8.1.1 Inject 2 uL (Section 6.1.1) of the column performance check solution (Section 7 . 7 ) and acquire selected ion monitoring (SIM) data for m/z 258.930, 319.897, 321.894, 331.937 and 333.934 within a total cycle time of <l second (Section 8.3.4.1). 8.1.2 The chromatographic peak separation between 2,3,7,8-TCDD and the peaks representing any other TCDD isomers must be resolved with a valley of ^25 percent, where Valley Percent - (x/y)(100) x « measured as in Figure 1 y - the peak height of 2,3,7,8-TCDD. It is the responsibility of the laboratory to verify the conditions suitable for the appropriate resolution of 2,3,7,8-TCDD D-9 �from all other TCDD isomers. The column performance check solution also contains the TCDD isomers eluting first and last under the analytical conditions specified in this protocol thus defining the retention time window for total TCDD determination. The peaks representing 2,3,7,8-TCDD and the first and the last eluting TCDD isomer must be labeled and identified as such on the chromatograms (F and L, resp.). Any individual selected ion current profile or the reconstructed total ion current (m/z 259 + m/z 320 + m/z 322) constitutes an acceptable form of data presentation. 8.2 Mass Spectrometer Performance 8.2.1 The mass spectrometer must be operated in the electron (impact) ionization mode. Static resolving power of at least 10,000 (10 percent valley) must be demonstrated before any analysis of a set of samples is performed (Section 8 2 2 . Static ..) resolution checks must be performed at the beginning and at the end of each 12-hour period of operation. However, it is recommended that a visual check (i.e., not documented) of the static resolution be made using the peak matching unit before and after each analysis. 8.2.2 Chromatography time for TCDD may exceed the long-term mass stability of the mass spectrometer and thus mass drift correction is mandatory. A reference compound [high-boiling perfluorokerosene (PFK) is recommended] is introduced into the mass spectrometer. An acceptable lock mass ion at any mass between m/z 250 and m/z 334 (m/z 318.979 from PFK is recommended) must be used to monitor and correct mass drifts. NOTE: Excessive PFK may cause background noise problems and contamination of the source resulting in an increase in "downtime" for source cleaning. Using a PFK molecular leak, tune the instrument to meet the minimum required resolving power of 10,000 (10% valley) at m/z 254.986 (or any other mass reasonably close to m/z 259). Calibrate the voltage sweep at least across the mass range m/z 259 to m/z 334 and verify that m/z 330.979 from PFK (or any other mass close to ra/z 334) is measured within ^ ppm (i.e., 5 1.7 mmu, if m/z 331 is chosen) using m/z 254.986 as a reference. Documentation of the mass resolution must then be accomplished by recording the peak profile of the PFK reference peak m/z 318.979 (or any other reference peak at a mass close to m/z 320/322). The format of the peak profile representation must allow manual determination of the resolution, i.e., the horizontal axis must be a calibrated mass scale (amu or ppm per division). The result of the peak width measurement (performed at 5 percent of the maximum which corresponds to the 10% valley definition) must appear on the hard copy and cannot exceed 100 ppm (or 31.9 ramu if m/z 319 is the chosen reference ion). D-10 �8.3 Initial Calibration Initial calibration is required before any samples are analyzed for 2 ,3,7,8-TCDD. Initial calibration is also required if any routine calibration does not meet the required criteria listed in Section 8.6. 8.3.1 All concentration calibration solutions listed in Table 1 must be utilized for the initial calibration. 8.3.2 Tune the instrument with PFK as described in Section 8.2.2. 8.3.3 Inject 2 uL of the column performance check solution (Section 7.7) and acquire SIM mass spectral data for m/z 258.930, 319.897, 321.894, 331.937 and 333.934 using a total cycle time of ^ 1 second (Section 8.3.4.1). The laboratory must not perform any further analysis until it has been demonstrated and documented that the criterion listed in Section 8.1.2 has been met. 8.3.4 Using the same GC (Section 8.1) and MS (Section 8.2) conditions that produced acceptable results with the column performance check solution, analyze a 2-uL aliquot of each of the 4 concentration calibration solutions in triplicate with the following MS operating parameters. 8.3.4.1 Total cycle time for data acquisition must be ^ 1 second. Total cycle time includes the sum of all the dwell times and voltage reset times. 8.3.4.2 Acquire SIM data for the following selected characteristic ions: m/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD 321.894 Unlabeled TCDD 331.937 13 C12-2,3,7,8-TCDD, 13C12-1,2,3,4-TCDD 333.934 13 C12-2,3,7,8-TCDD, 13 C12-1,2,3,4-TCDD 8.3.4.3 The ratio of integrated ion current for m/z 319.897 to m/z 321.894 for 2,3,7,8-TCDD must be between 0.67 and 0.90. 8.3.4.4 The ratio of integrated ion current for m/z 331.937 to m/z 333.934 for 1JC12-2,3,7,8-TCDD and 13C12-1,2,3,4TCDD must be between 0.67 and 0.90. D-ll �8.3.4.5 Calculate the relative response factors for unlabeled 2,3,7,8-TCDD [RRF(I)] relative to 13C12-2,3,7,8-TCDD and for labeled 13C12-2,3,7,8-TCDD [RRF(II)] relative to 13C]2-1,2,3,4-TCDD as follows: Ax ' QIS RRF(I) = Qx ' AIS A RRF(II) = IS QlS 'ARS where Ax AJS - sun of the integrated ion abundances of m/z 319.897 and m/z 321.894 for unlabeled 2,3,7,8-TCDD. * sum for of the integrated ion abundances of m/z 331.937 and m/z 333.934 C12-2,3,7,8-TCDD. 13 sum of the integrated ion abundances for m/z 331.937 and n/z 333.934 for 13C12-1 ,2,3,4-TCDD. Q1S - quantity of QRS Qx C12-2,3,7,8-TCDD injected (pg). • quantity of 1 3Cj2-l ,2,3,4-TCDD Injected (pg). - quantity of unlabeled 2,3,7,8-TCDD injected (pg). RRF is a dimensionless quantity; the units used to express QIS> QRS must be the same. anc * QX 8.3.4.6 Calculate the four means (RRFs) and their respective relative standard deviations (%RSD) for the response factors from each of the triplicate analyses for both unlabeled and C12-2 ,3,7,8-TCDD (Form H-2). 8.3.4.7 Calculate the grand means RRF(I) and RRF(II) and their respective relative standard deviations (%RSD) using the four mean RRFs (Section 8 3 4 6 (Form H-2). ...) 8.3.4.8 Calculate the routine calibration permissible range for RRF(I) and RRF(II) using a +_20% window from the grand weans RRF(I) and RRF(II) (Section 8.3.4.7) (Form H-2). 8.4 Criteria for Acceptable Calibration The criteria listed below for acceptable calibration must be met before analysis of any sample Is performed. D-12 �8.4.1 8.4.2 The variation of the 4 mean RRFs for unlabeled and Cj2~ 2,3,7,8-TCDD obtained from the triplicate analyses must be less than 20 percent RSD. 8.4.3 SIM traces for 2,3,7,8-TCDD must present a signal-to-noise ratio of ^2.5 for m/z 258.930, m/z 319.897 and, m/z 321.894. 8.4.4 SIM traces for 13C12~2,3,7,8-TCDD must present a signal-tonoise ratio >2.5 for m/z 331.937 and m/z 333.934. 8.4.5 Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within the allowed range. NOTE: 8.5 The percent relative standard deviation (RSD) for the response factors from each of the triplicate analyses for both unlabeled and 13C^2-2,3,7,8-TCDD must be less than 20 percent. If the criteria for acceptable calibration listed in Sections 8.4.1 and 8.4.2 have been met, the RRF can be considered independent of the analyte quantity for the calibration concentration range. The mean RRF from 4 triplicate determinations for unlabeled 2,3,7,8-TCDD and for 13C12-2,3,7,8-TCDD will be used for all calculations until routine calibration criteria (Section 8.6) are no longer met. At such time, new mean RRFs will be calculated from a new set of four triplicate determinations. Routine Calibrations Routine calibrations must be performed at the beginning of a 12-hour period after successful mass resolution and GC column performance check runs. 8.5.1 8.6 Inject 2 uL of the concentration calibration solution which contains 10 pg/uL of unlabeled 2,3,7.8-TCDD, 10.0 pg/uL of C12-2,3,7,8-TCDD and 10 pg/uL C12-l,2,3,4-TCDD. Using the same GC/MS/DS conditions as used in Sections 8.1, 8.2 and 8.3, determine and document acceptable calibration as provided in Section 8.6. Criteria for Acceptable Routine Calibration The following criteria must be met before further analysis is performed. If these criteria are not met, corrective action must be taken and the instrument must be recalibrated. 8.6.1 The measured RRF for unlabeled 2,3,7,8-TCDD must be within 20 percent of the mean values established (Section 8.3.4.8) by triplicate analyses of concentration calibration solutions. 8.6.2 The measured RRF for 13C12-2,3,7,8-TCDD must be within 20 percent of the mean value established by triplicate analysis of the concentration calibration solutions (Section 8.3.4.8). D-13 �8.6.3 8.6.4 If one of the above criteria is not satisfied, a second attempt can be made before repeating the entire initialization process (Section 8.3). NOTE: 9. Isotopic ratios (Sections 8.3.4.3 and 8.3.4.4) must be within the allowed range. An initial calibration must be carried out whenever the HRCC 2 solution is replaced by a new one from a different lot. QUALITY CONTROL PROCEDURES See Exhibit E for QA/QC requirements. 10. SAMPLE PRESERVATION AND HANDLING 10.1 Chain-of-custody procedures — see Exhibit G. 10.2 Sample Preservation 10.2.1 When received, each soil or sediment sample will be contained in a 1-pint glass jar surrounded by vermiculite in a sealed metal paint can. Until a portion is to be removed for analysis, store the sealed paint cans in a locked limited-access area where the temperature is maintained between 25° and 35°C. After a portion of a sample has been removed for analysis, return the remainder of the sample to its original container and store as stated above. 10.2.2 Each aqueous sample will be contained in a 1-liter glass bottle. The bottles with the samples are stored at 4°C in a refrigerator located in a locked limited-access area. 10.2.3 To avoid photodecomposition, protect samples from light. 10.3 Sample Handling CAUTION: Finely divided soils and sediments contaminated with 2,3,7,8-TCDD are hazardous because of the potential for inhalation or ingestion of particles containing 2,3,7,8-TCDD. Such samples should be handled in a confined environment (i.e., a closed hood or a glove box). 10.3.1 Pre-extraction sample treatment 10.3.1.1 Homogenization — Although sampling personnel will attempt to collect homogeneous samples, the contractor shall examine each sample and judge if it needs further mixing. NOTE: Contractor personnel have the responsibility to take a representative sample portion; this responsibility D-14 �entails efforts to make the sample as homogeneous as possible. Stirring is recommended when possible. 10.3.1.2 Centrifugation — When a soil or sediment sample contains an obvious liquid phase, it must be centrifuged to separate the liquid from the solid phase. Place the entire sample in a suitable centrifuge bottle and centrifuge for 10 minutes at 2000 rpn. Remove the bottle from the centrifuge. With a disposable pipet, remove the liquid phase and discard it. Mix the solid phase with a stainless steel spatula and remove a portion to be weighed and analyzed. Return the remaining solid portion to the original sample bottle (which must be empty) or to a clean, empty sample bottle which is properly labeled, and store it as described in 10.2.1. CAUTION: The removed liquid may contain TCDD and should be disposed as a liquid waste. 10.3.1.3 Weigh between 9.5 and 10.5 g of the soil or sediment sample ( 0 5 g) to 3 significant figures. Dry it to +. constant weight at 100°C. Allow the sample to cool in a desiccator. Weigh the dried soil to 3 significant figures. Calculate and report percent moisture on Form H-9. 11. SAMPLE EXTRACTION 11.1 Soil/Sediment Extraction 11.1.1 Immediately before use, the Soxhlet apparatus is charged with 200 to 250 mL benzene which is then refluxed for 2 hours. The apparatus is allowed to cool, disassembled and the benzene removed and retained as a blank for later analysis if required. 11.1.2 Accurately weigh to 3 significant figures a 10-g (9.50 g to 10.50 g) portion of the wet soil or sediment sample. Mix 100 uL of the sample fortification solution (Section 7.8) with 1.5 mL acetone (1000 pg of 13C12-2,3,7,8-TCDD) and deposit the entire mixture in small portions on several sites on the surface of the soil or sediment. 11.1.3 Add 10 g anhydrous sodium sulfate and mix thoroughly using a stainless steel spoon spatula. 11.1.4 After breaking up any lumps, place the soil-sodium sulfate mixture in the Soxhlet apparatus using a glass wool plug (the use of an extraction thimble is optional). Add 200 to 250 mL benzene to the Soxhlet apparatus and reflux for 24 hours. The solvent must cycle completely through the system at least 3 times per hour. D-15 �11.1.5 Transfer the extract to a Kuderna-Danish apparatus and concentrate to 2 to 3 mL. Rinse the column and flask with 5 mL benzene and collect the rinsate in the concentrator tube. Reduce the volume in the concentrator tube to 2 to 3 mL. Repeat this rinsing and concentrating operation twice more. Remove the concentrator tube from the K-D apparatus and carefully reduce the extract volume to approximately 1 mL with a stream of nitrogen using a flow rate and distance such that gentle solution surface rippling is observed. NOTE: 11.2 Glassware used for more than one sample must be carefully cleaned between uses to prevent cross-contamination (Note on page D-6). Extraction of Aqueous Samples 11.2.1 Mark the water meniscus on the side of the 1-L sample bottle for later determination of the exact sample volume. Pour the entire sample (approximately 1 L) into a 2-L separatory funnel. 11.2.2 Mix 100 uL of the sample fortification solution with 1.5 mL acetone (1000 pg of ^Cj2-2 ,3 ,7,8-TCDD) and add the mixture to the sample in the separatory funnel. NOTE: A continuous liquid-liquid extractor may be used in place of a separatory funnel. 11.2.3 11.2.4 Add 60 raL methylene chloride to the sample bottle, seal and shake 30 seconds to rinse the inner surface. Transfer the solvent to the separatory funnel and extract the sample by shaking the funnel for 2 minutes with periodic venting. Allow the organic layer to separate from the water phase for a minimum of 10 minutes. If an emulsion interface between layers exists, the analyst must employ mechanical techniques (to be described in the final report) to complete the phase separation. Collect the methylene chloride (3 x 60 mL) directly into a 500-mL Kuderna-Danish concentrator (mounted with a 10-mL concentrator tube) by passing the sample extracts through a filter funnel packed with a glass wool plug and 5 g of anhydrous sodium sulfate. After the third extraction, rinse the sodium sulfate with an additional 30 mL of methylene chloride to ensure quantitative transfer. Attach a Snyder column and concentrate the extract until the apparent volume of the liquid reaches 1 mL. Remove the K-D apparatus and allow it to drain and cool for at least 10 minutes. Remove the Snyder column, add 50 mL benzene, reattach the Snyder column and concentrate to approximately 1 mL. Rinse the flask and the lower joint with 1 to 2 mL benzene. Concentrate the extract to 1.0 mL under a gentle stream of nitrogen. D-16 �11.2.5 Determine the original sample volume by refilling the sample bottle to the mark and transferring the liquid to a 1000-mL graduated cylinder. Record the sample volume to the nearest 5 mL. 11.3 Cleanup Procedures 11.3.1 Prepare an acidic silica column as follows: Pack a 1 cm x 10 cm chromatographic column with a glass wool plug, a layer (1 cm) of Na2804/K2003(1:1), 1.0 g silica gel (Section 7.1.2) and 4.0 g of 40-percent w/w sulfuric acid-impregnated silica gel (Section 7.1.3). Pack a second chromatographic column (1 cm x 30 cm) with a glass wool plug, 6.0 g acidic alumina (Section 7.1.1) and top with a 1-cm layer of sodium sulfate (Section 7.4.1). Add hexane to the columns until they are free of channels and air bubbles. 11.3.2 Quantitatively transfer the benzene extract (1 mL) from the concentrator tube to the top of the silica gel column. Rinse the concentrator tube with two 0.5-mL portions of hexane. Transfer the rinses to the top of the silica gel column. 11.3.3 Elute the extract from the silica gel column with 90 mL hexane directly into a Kuderna-Danish concentrator. Concentrate the eluate to 0.5 mL, using nitrogen blow-down as necessary. 11.3.4 Transfer the concentrate (0.5 mL) to the top of the alumina column. Rinse the K-D assembly with two 0.5-mL portions of hexane and transfer the rinses to the top of the alumina column. Elute the alumina column with 18 mL hexane until the hexane level is just below the top of the sodium sulfate. Discard the eluate. Columns must not be allowed to reach dryness (i.e., a solvent "head" must be maintained.) 11.3.5 Place 30 mL of 20-percent (v/v) methylene chloride in hexane on top of the alumina and elute the TCDDs from the column. Collect this fraction in a 50-mL Erlenmeyer flask. 11.3.6 Prepare an 18-percent Carbopak C/Celite 545* mixture by thoroughly mixing 3.6 grams Carbopak C (80/100 mesh) and 16.4 grams Celite 545® in a 40-mL vial. Activate at 130°C for 6 hours. Store in a desiccator. Cut off a clean 5-mL disposable glass pipet (6 to 7mm ID) at the 4-mL mark. Insert a plug of glass wool (Section 7.3) and push to the 2-mL mark. Add 340 to 600 mg of the activated Carbopak/Celite mixture (see NOTE) followed by another glass wool plug. Using two glass rods, push both glass wool plugs simultaneously towards the Carbopak/Celite mixture and gently compress the Carbopak/Celite plug to a length of 2 to 2.5 cm. Preelute the column with 2 mL toluene followed by 1 mL of 75:20:5 methylene chloride/methanol/benzene, 1 mL of 1:1 cyclohexane in methylene chloride, and 2 mL hexane. The flow rate should be less than 0.5 mL/min. While the D-17 �column is still wet with hexane, add the entire eluate (30 mL) from the alumina column (Section 11.3.5) to the top of the column. Rinse the Erlenmeyer flask which contained the extract twice with 1 mL hexane and add the rinsates to the top of the column. Elute the column sequentially with two 1-mL aliquots hexane, 1 mL of 1:1 cyclohexane in methylene chloride, and 1 mL of 75:20:5 methylene chloride/ methanol/benzene. Turn the column upside down and elute the TCDD fraction with 6 mL toluene into a concentrator tube. Warm the tube to approximately 60°C and reduce the toluene volume to approximately 1 mL using a stream of nitrogen. Carefully transfer the concentrate into a 1-mL mini-vial and, again at elevated temperature, reduce the volume to about 100 uL using a stream of nitrogen. Rinse the concentrator tube with 3 washings using 200 uL of 1% toluene in CH2Cl2« Add 10 uL of the tridecane solution containing the recovery standard and store the sample in a refrigerator until HRGC/HRMS analysis is performed. NOTE: The amount of activate Carbopak/Celite mixture required to form a 2-to 2.5-cm plug in the column depends on the density of the Celite being used. 12. ANALYTICAL PROCEDURES 12.1 Remove the sample extract or blank from storage and allow it to warm to ambient laboratory temperature. With a stream of dry, purified nitrogen, reduce the extract/blank volume to 10 uL. 12.2 Inject a 2-uL aliquot of the extract into the GC, operated under the conditions previously used (Section 8.1) to produce acceptable results with the performance check solution. 12.3 Acquire SIM data according to 12.3.1. Use the same acquisition and MS operating conditions previously used (Section 8.3.A) to determine the relative response factors. 12.3.1 Acquire SIM data for the following selected characteristic ions: in/z Compound 258.930 TCDD - COC1 319.897 Unlabeled TCDD t Unlabeled TCDD 321.894 331.937 13 C12-2,3,7,8-TCDD, 13C12-1,2,3,4TCDD 333.934 13 C12-2,3,7,8-TCDD, 13C12~1,2,3,4TCDD D-18 �NOTE: The acquisition period must at least encompass the TCDD retention time window previously determined (Section 8.1.2, Exhibit D). 12.4 Identification Criteria 12.4.1 The retention time (RT) (at maximum peak height) of the sample component m/z 319.897 must be within -1 to +3 seconds of the retention time of the peak for the isotopically labeled internal standard at m/z 331.937 to attain a positive identification of 2,3,7,8-TCDD. Retention times of other tentatively identified TCDDs must fall within the RT window established by analyzing the column performance check solution (Section 8.1). Retention times are required for all chromatograms. 12.4.2 The ion current responses for m/z must reach maximum simultaneously current intensities must be _>. 2.5 positive identification of a TCDD isomers. 258.930, 319.897 and 321.894 (+ 1 sec), and all ion times noise level for or group of coeluting TCDD 12.4.3 The integrated ion current at m/z 319.897 must be between 67 and 90 percent of the ion current response at m/z 321.894. 12.4.4 The integrated ion current at m/z 331.937 must be between 67 and 90 percent of the ion current response at m/z 333.934. 12.4.5 The integrated ion currents for m/z 331.937 and 333.934 must reach their maxima within _+_ 1 sec. 12.4.6 The recovery of the internal standard be between 40 and 120 percent. 13. 13 C12~2,3,7,8-TCDD must CALCULATIONS 13.1 Calculate the concentration of 2,3,7,8-TCDD (or any other TCDD isomer or group of coeluting TCDD isomers) using the formula: A X ' QlS AIg * W ' RRF(I) where: * unlabeled 2,3,7,8-TCDD (or any other unlabeled TCDD isomer or group of coeluting TCDD isomers) concentration in pg/g. * sum of the integrated ion abundances determined for m/z 319.897 and 321.894. * sum of the integrated ion abundances determined for m/z 331.937 and 333.934 of r3C12-2,3,7,8-TCDD (IS = internal standard). D-19 �QIS - W quantity (in picograms) of C12-2,3,7 ,8-TCDD added to the sample before extraction (Qjg = 1000 pg). • weight (in grams) of dry soil or sediment sample or volume of aqueous sample converted to grams. RRF(I) - calculated mean relative response factor for unlabeled 2,3,7,8-TCDD relative to Cj2~2,3,7 ,8-TCDD. This represents the grand mean of the RRF(I)'s obtained in Section 8.3.4.5. 13.2 Calculate the recovery of the internal standard 13C12~2 ,3 ,7 ,8-TCDD measured in the sample extract, using the formula: Internal standard percent recovery * Y — - • 100 ARS * RRF(II) Where: - sum of the integrated ion abundances determined for m/z 331.937 and 333.934 of 1JC12-2,3,7,8-TCDD (IS - internal standard). " sum of the integrated ion abundances determined for m/z 331.937 and 333.934 of C12-l ,2,3,4-TCDD (RS - recovery standard). Y » 0.1 for the "10-yL extract" injection (to be reported on Forms H-l, H-5 and H-9). and Y - 1.2 for the "24-yL extract" injection (Section 13.3) (to be reported on Form H-9 used for reporting the diluted extract analysis). 1O ___, RRF(II) « calculated mean relative response factor for labeled C^2-2, 3,7,8TCDD relative to C12~l ,2 ,3,4-TCDD. This represents the grand mean of the RRF(II)'s calculated in Section 8 3 4 5 .... 13.3 13.4 C If the concentration of the most abundant TCDD isomer (or group of coeluting TCDD isomers) exceeds 100 pg/uL in the 10 uL final extract, the linear range of response vs. concentration may have been exceeded, and a diluted aliquot of the original sample extract must be analyzed. Accurately dilute 2 uL of the remaining original extract with 22 uL of the tridecane solution containing 10 pg/uL of the recovery standard (Section 7.9, Exhibit D). Total TCDD concentration — all positively identified isomers of TCDD must be within the RT window and meet all identification criteria listed in Sections 12.4.2 and 12.4.3. Use the expression in Section 13.1 to calculate the concentrations of the other TCDD isomers, with Cx becoming the concentration of any unlabeled TCDD isomer or group of coeluting TCDD isomers. Total TCDD - Sum of the concentrations of the individual TCDDs including 2,3,7,8-TCDD. D-20 �13.5 Estimated Detection Limit — For samples in which no unlabeled 2,3,7,8-TCDD was detected, calculate the estimated minimum detectable concentration. The background area is determined by integrating the ion abundances for n/z 319.897 and 321.894 in the appropriate region of the selected ion current profiles, multiplying that area by 2.5, and relating the product area to an estimated concentration that would produce that product area. Use the formula: (2.5) ' (Ax) * (QIS) (AIS) ' (RRF(I)) * (W) where CE = estimated concentration of unlabeled 2,3,7,8-TCDD required to produce Ax. Ax = sum of integrated ion abundances for m/z 319.897 and 321.894 in the same group of >5 scans used to measure AIS. Aig - sum of integrated ion abundances for the appropriate ion characteristic of the internal standard, m/z 331.937 and m/z 333.934. QJS, RRF(I), and W retain the definitions previously stated in Section 13.1. Alternatively, if peak height measurements are used for quantification, measure the estimated detection limit by the peak height of the noise in the 2,3,7,8TCDD RT window. 13.6 The relative percent difference (RPD) is calculated as follows: I RPD s l ~ S2 I I sl ~ S2 I = x «= Mean Concentration 100 (8j + BI and 52 represent sample and duplicate sample results. References 1. "Carcinogens - Working with Carcinogens", Department of Health, Education and Welfare, Public Health Service, Center for Disease Control, National Institute for Occupational Safety and Health, Publication No. 77-206, Aug. 1977. 2. "OSHA Safety and Health Standards, General Industry" (29 CFR1910), Occupational Safety and Health Administration, OSHA 2206 (Revised January 1976). 3. "Safety in Academic Chemistry Laboratories", American Chemical Society Publication, Committee on Chemical Safety, 3rd Edition 1979. D-21 �TABLE 1. COMPOSITION OF CONCENTRATION CALIBRATION SOLUTIONS Recovery Standard Analyte 13 C12-1,2,3,4-TCDD 2,3,7,8-TCDD Internal Standard 13 C12-2,3,7,8-TCDD HRCC1 2.0 pg/uL 2.0 pg/uL 10.0 pg/uL HRCC2 10.0 pg/uL 10.0 pg/uL 10.0 pg/uL HRCC3 50.0 pg/uL 50.0 pg/uL 10.0 pg/uL HRCC4 100.0 pg/uL 100.0 pg/uL 10.0 pg/uL Sample Fortification Solution 10.0 pg/uL of 13 C12-2,3,7,8-TCDD Recovery Standard Spiking Solution 10.0 pg/uL 13 C12-1,2,3,4-TCDD Field Blank Fortification Solutions A) 10.0 pg/uL of unlabeled 2,3,7,8-TCDD B) 10.0 pg/uL of unlabeled 1,2,3,4-TCDD Internal Standard Spiking Solution 10 pg/uL of 13C12-2,3,7,8-TCDD (Used only in Section 4.2.1.1, Exhibit E) D-22 �TABLE 2. RECOMMENDED GC OPERATING CONDITIONS Column coating SP-2330 (SP-2331) CP-SIL 88 Film thickness 0.2 urn 0.22 Column dimensions 60 m x 0.24 mm 50 m x 0.22 mm Helium linear velocity 28-29 cm/sec at 240°C 28-29 cm/sec at 240°C Initial temperature 150°C 200°C Initial time 4 min 1 min Temperature program Rapid increase to 200°C (15°C/min) 200°C to 250°C at 4°C/min Program from 200°C to 240°C at 4°C/min Approximate 2,3,7,8-TCDD retention time 27 min 22 min TABLE 3. 1. urn TYPICAL 12-HOUR SEQUENCE FOR 2,3,7,8-TCDD ANALYSIS Static mass resolution check and mass measurement error determination 10/20/84 0700h Column performance check 10/20/84 0730h 3 . HRCC2 10/20/84 0800h 4. Sample 1 through Sample "N" 10/20/84 0830h 5. Column performance check 10/20/84 1800h 6. Static mass resolution check 10/20/84 1830h 2. D-23 �Relative Intensity W e n re H (A CA u o m <T i— — 1368 r- C ft o a a a. B H- ca o n C/l ft "0 O I nt N> ^. U> p. i-n O * 1234; 1237; 1238 o 3 •s §• H- f> »- * p- a. 01 1 9 X X 2 * O CO 1-^ C 01 ii 2. • a. Ho- o a" 1 « 2 O ^ 3 O' •• O o en r> a. r> >" if a n N)• •• O o 1289 �EXHIBIT E QA/QC Requirements �SUMMARY OF QC ANALYSES o Initial and periodic calibration and instrument performance checks. o Field blank analyses (Section 4.1); a minimum of one fortified field blank pair shall be analyzed with each sample batch; an additional fortified field blank pair must be analyzed when a new lot of absorbent and/or solvent is used. o Analysis of a batch of samples with accompanying QC analyses: Sample Batch —_<24 samples, Including field blank and rinsate sample(s). Additional QC analyses per batch: Fortified field blanks 2 Method blank (1*) Duplicate sample 1 TOTAL 3(4) * A method blank is required whenever a fortified field blank shows a positive response as defined in Section 3.11, Exhibit D. o "Blind" QC samples may be submitted to the contractor as ordinary soil, sediment or water samples included among the batch of samples. Blind samples include: Uncontaminated soil, sediment and water, Split samples, Unidentified duplicates, and Performance evaluation samples. QUALITY CONTROL 1. Performance Evaluation Samples — Included among the samples in all batches will be samples containing known amounts of unlabeled 2,3,7,8-TCDD and/or other TCDDs that may or may not be marked as other-than-ordinary samples. 2. Performance Check Solutions 2.1 At the beginning of each 12-hour period during which samples are to be analyzed, an aliquot each of the 1) GC column performance check solution and 2) high-resolution concentration calibration solution E-l �No. 2 (HRCC2) shall be analyzed to demonstrate adequate GC resolution and sensitivity, response factor reproducibility, and mass range calibration. A mass resolution check shall also be performed to demonstrate adequate mass resolution using an appropriate reference compound (PFK is recommended). These procedures are described in Section 8 of Exhibit D. If the required criteria are not met, remedial action nust be taken before any samples are analyzed. 2.2 To validate positive sample data, the GC column performance check and the mass resolution check must be performed also at the end of each 12-hour period during which samples are analyzed. 2.2.1 If the contractor laboratory operates only during one period (shift) each day of 12 hours or less, the GC performance check solution must be analyzed twice (at the beginning and end of the period) to validate data acquired during the interim period. This applies also to the mass resolution check. 2.2.2 If the contractor laboratory operates during consecutive 12-hour periods (shifts), analysis of the GC performance check solution at the beginning of each 12-hour period and at the end of the final 12-hour period is sufficient. This applies also to the mass resolution check. 2.3 Results of at least two analyses of the GC column performance check solution and the mass resolution check must be reported with the sample data collected during a 12-hour period. 2.4 Deviations from criteria specified for the GC performance check or for the mass resolution check (Section 8, Exhibit D) invalidate all positive sample data collected between analyses of the performance check solution, and the extract from those positive samples shall be reanalyzed Exhibit C). The GC column performance check mixture, concentration calibration solutions, and the sample fortification solutions are to be obtained from the EMSL-LV. However, if not available from the EMSL-LV, standards can be obtained from other sources, and solutions can be prepared in the contractor laboratory. Concentrations of all solutions containing unlabeled 2,3,7,8TCDD which are not obtained from the EMSL-LV must be verified by comparison with the unlabeled 2,3,7,8-TCDD standard solution (concentration of 7.87 ug/mL) that is available from the EMSL-LV. When a lower-concentration standard solution becomes available from the EMSL-LV, it will be substituted for the 7.87 ug/mL standard. E-2 �4. Blanks 4.1 A method blank is required whenever a positive response (Section 3.11, Exhibit D) is obtained for a fortified field blank. To that effect, perform all steps detailed in the analytical procedure (Section 11, Exhibit D) using all reagents, standards, equipment, apparatus, glassware, and solvents that would be used for a sample analysis, but omit addition of the soil, sediment or aqueous sample portion. 4.1.1 The method blank must contain the same amount of 13C^2~2,3,7,8TCDD that is added to samples before extraction. 4.1.2 An acceptable method blank exhibits no positive response (Section 3.11, Exhibit D) for any of the characteristic ions monitored. If the method blank which was extracted along with a batch of samples is contaminated, all positive samples must be rerun (Exhibit C). 4.1.2.1 4.1.2.2 4.2 If the above criterion is not met, check solvents, reagents, fortification solutions, apparatus, and glassware to locate and eliminate the source of contamination before any samples are extracted and analyzed. If new batches of reagents or solvents contain interfering contaminants, purify or discard them. Field blanks — Each batch of samples contains a field blank sample of uncontaminated soil/sediment or water that is to be fortified before analysis according to Section 4.2.1, Exhibit E. In addition to this field blank, a batch of samples may include a rinsate, that is a portion of solvent (usually trichloroethylene) that was used to rinse sampling equipment. The rinsate is analyzed to assure that the samples have not been contaminated by the sampling equipment. 4.2.1 Fortified field blank pair 4.2.1.1 Fortified field blank A: 2,3,7,8-TCDD 4.2.1.1.1 Weigh a 10-g portion or use 1 liter (for aqueous samples) of the specified field blank sample and add 100 uL of the solution containing 10.0 pg/uL of 2,3,7,8-TCDD (Table 1, Exhibit D) diluted in 1.5 mL of acetone (Section 11.1.2, Exhibit D). 4.2.1.1.2 Extract using the procedures beginning in Sections 11.1 or 11.2 of Exhibit D, as applicable, add 10 uL of the internal standard solution (Section 7.10, Exhibit D) and analyze a 2-uL aliquot of the concentrated extract. E-3 �NOTE: This is the only case where the recovery standard is used for other than recovery purposes. 4.2.1.1.3 Calculate the concentration (Section 13.1, Exhibit D) of 2,3,7,8-TCDD and the percent recovery of unlabeled 2,3,7,8-TCDD. If the percent recovery at the measured concentration of 2,3,7,8-TCDD is <40 percent or >120 percent, report the results and repeat the fortified field blank extraction and analysis with a second aliquot of the specified field blank sample (Exhibit C). 4.2.1.1.4 Extract and analyze a new fortified simulated field blank whenever new lots of solvents or reagents are used for sample extraction or for column chromatographic procedures. When a fortified simulated field blank produces a positive response (Section 3.11, Exhibit D) for any m/z being monitored at the retention time of 1 ,2,3,4-TCDD, a method blank (Section 4.1, Exhibit E) is required. NOTE: For this purpose only, the Contractor will simulate field blanks by using clean sand or distilled water. 4.2.1.2 Fortified field blank B: 1,2,3,4-TCDD 4.2.1.2.1 4.2.1.2.2 4.2.2 Repeat steps 4.2.1.1.1 to 4.2.1.1.3 using un .... to .... usng unlabeled 13 1,2,3,4TCDD (instead of 2,3,7.8-TCDD) and 13C C 13 -1,2,3,4-TCDD (instead of C12-2,3,7,8-TCD TCDD) as 12 recovery standard. Extract and analyze a new fortified simulated field blank whenever new lots of solvents or reagents are used for sample extraction or for column chromatographic procedures. When a fortified simulated field blank produces a positive response (Section 3.11, Exhibit D) for any m/z being monitored at the retention time of 2,3,7,8-TCDD, a method blank (Section 4.1, Exhibit E) is required. Rinsate sample 4.2.2.1 The rinsate sample must be fortified as a regular sample. 4.2.2.2 Take a 100-mL aliquot of sampling equipment rinse solvent (rinsate sample), filter, if necessary, and add 100 uL of the solution containing 10.0 pg/uL of I3 C12-2,3,7,8-TCDD (Table 1, Exhibit D). E-4 �4.2.2.3 4.2.2.4 Transfer the 5-mL concentrate in 1-mL portions to a 1mL mini-vial, reducing the volume as necessary with a gentle stream of dry nitrogen; see Exhibit D, Section 11.1.5 for volume reduction procedures. 4.2.2.5 Rinse the container with two 0.5-mL portions of hexane and transfer the rinses to the 1-mL mini-vial. 4.2.2.6 Just before analysis, add 10 uL tridecane recovery standard spiking solution (Table 1, Exhibit D), and reduce the volume to a final volume of 10 uL (no column chromatography is required). 4.2.2.7 Analyze an aliquot following the same procedures used to analyze samples (Section 12, Exhibit D). 4.2.2.8 5. Using a Kuderna-Danish apparatus, concentrate to approximately 5 mL. Report percent recovery of the internal standard and the level of contamination by any TCDD isomer (or group of coeluting TCDD isomers) on Form H-5 in pg/mL of rinsate solvent. Duplicate Analyses 5.1 Laboratory duplicates — in each batch of samples, locate the sample specified for duplicate analysis and analyze a second 10-g soil or sediment sample portion or 1-L water sample. 5.1.1 The results of laboratory duplicates (percent recovery and concentrations of 2,3,7,8-TCDD and total TCDD) must agree within 50 percent relative difference (difference expressed as percentage of the mean). If the relative difference is >50 percent, the Contractor shall immediately contact the Sample Management Office for resolution of the problem. Report all results. 5.1.2 Recommended actions to help locate problems: 5.1.2.1 5.1.2.2 If possible, verify that no error was made while weighing sample portions. 5.1.2.3 6. Verify satisfactory instrument performance (Section 8, Exhibit D). Review the analytical procedures with the performing laboratory personnel. Percent Recovery of the Internal Standard ^C, 2~2 ,3,7,8-TCDD — For each sample, method blank and rinsate, calculate the percent recovery (Section 13.2, Exhibit D) of the measured concentration of C12-2,3,7,8-TCDD. If E-5 �the percent recovery is <40 percent or >120 percent for a sample, analyze a second portion of that sample and report both results (Exhibit C). NOTE: 7. A low or high percent recovery for a blank does not require discarding analytical data but it may indicate a potential problem with future analytical data. Identification Criteria 7.1 7.2 8. If either of the two identification criteria (Sections 12.4.1 and 12.4.2, Exhibit D) is not met, it is reported that the sample does not contain unlabeled 2,3,7,8-TCDD at the calculated detection limit (Section 13.5, Exhibit D). If the first two initial identification criteria are met, but the third, fourth, fifth or sixth criterion (Sections 12.4.3 through 12.4.6, Exhibit D) is not met, that sample is presumed to contain interfering contaminants. This must be noted on the analytical report form and the sample must be rerun or the extract reanalyzed. Detailed sample rerun and extract reanalysis requirements are presented in Exhibit C. Blind QC Samples — Included among soil, sediment and aqueous samples may be QC samples that are not specified as such to the performing laboratory. Types that may be included are: 8.1 Uncontaminated soil, sediment or water. 8.1.1 If a false positive is reported for such a sample, the Contractor shall be required to rerun the entire associated batch of samples (Section 2.3.3, Exhibit C). 8.2 Split samples — one laboratory. 8.3 Unlabeled field duplicates — sample. 8.4 Performance evaluation sample — soil/sediment or water sample containing a known amount of unlabeled 2,3,7,8-TCDD and/or other TCDDs. 8.4.1 composited sample portions sent to more than two portions of a composited If the performance evaluation sample result falls outside the acceptance windows established by EPA, the Contractor shall be required to rerun the entire associated batch of samples (Exhibit C). NOTE: EPA acceptance windows are based on previously generated data. 9. Records - At each contractor laboratory, records must be maintained on E-6 �site for six months after contract completion to document the quality of all data generated during the contract performance. Before any records are disposed, written concurrence from the Contracting Officer must be obtained. 10. Unused portions of samples and sample extracts must be preserved for six months after sample receipt; appropriate samples may be selected by EPA personnel for further analyses. 11. Reuse of glassware is to be minimized to avoid the risk of contamination. LABORATORY EVALUATION 1. PROCEDURES On a quarterly basis, the EPA Project Officer and/or designated representatives may conduct an evaluation of the laboratory to ascertain that the laboratory is meeting contract requirements. This section outlines the procedures which may be used by the Project Officer or his authorized representative in order to conduct a successful evaluation of laboratories conducting dioxin analyses according to this protocol. The evaluation process consists of the following steps: 1) analysis of a performance evaluation (PE) sample, and 2) on-site evaluation of the laboratory to verify continuity of personnel, instrumentation, and quality assurance/ quality control functions. The following is a description of these two steps. 2. Performance Evaluation Sample Analysis 2.1 The PE sample set will be sent to a participating laboratory to verify the laboratory's continuing ability to produce acceptable analytical results. The PE sample will be representative of the types of samples that will be subject to analysis under this contract. 2.2 When the PE sample results are received, they are scored using the PE Sample Score Sheet shown in Figure 1. If a false positive (e.g., a PE sample not containing 2,3,7,8-TCDD and/or other TCDDs but reported by the laboratory to contain it and/or them) is reported, the laboratory has failed the PE analysis requirement. The Project Officer will notify the laboratory immediately if such an event occurs. 2.3 As a general rule, a laboratory should achieve 75 percent or more of the total possible points for all three categories, and 75 percent or more of the maximum possible points in each category to be considered acceptable for this program. However, the Government reserves the right to accept scores of less than 75 percent. 2.4 If unanticipated difficulties with the PE samples are encountered, the total points may be adjusted by the Government evaluator in an impartial and equitable manner for all participating laboratories. E-7 �Number of PE Samples Maximum Possible Score Recommended Passing Score (75%) 1 290 218 2 475 356 3 660 495 4 845 634 5 1030 773 On-Site Laboratory Evaluation 3.1 An on-site laboratory evaluation is performed to verify that (1) the laboratory is maintaining the necessary minimum level in instrumentation and levels of experience in personnel committed to the contract and (2) that the necessary quality control/quality assurance activities are being carried out. It also serves as a mechanism for discussing laboratory weaknesses identified through routine data audits, PE sample analyses results, and prior on-site evaluation. Photographs may be taken during the on-site laboratory evaluation tour. 3.2 The sequence of events for the on-site evaluations is shown in Figure 2. The Site Evaluation Sheet (SES) (Figure 3) is used to document the results of the evaluation. E-8 �PERFORMANCE EVALUATION SAMPLE SCORE SHEET Laboratory Date False Positive I. False Positive - If a laboratory reports a false positive on any PE sample, the laboratory may be disqualified, i.e., rendered Ineligible for contract award based on the failure to pass the PE sample analysis requirement. 2,3,7,8-TCDD ( ) Yes ( ) No Other TCDD(s) ( ) Yes ( ) No Possible Score II. Calibration Data 1. Method Blank: a. Results properly recorded on Forms H-l, H-5 and H-9. b. No native TCDD isomers at/or above method quantitative limit. 5 Results documented by selected ion monitoring (SIM) traces for m/z being monitored to detect TCDDs. 5 c. d. 2. Percent recovery of MO and £120%. 5 13 C12-2,3,7,8-TCDD 5 Initial Concentration Calibration: a. Results properly recorded on Forms H-2 and H-8. b. The percent relative standard deviation (RSD) for the response factors for each of the triplicate analyses for both unlabeled and 13C12-2,3,7,8-TCDD less than 20%. 5 The variation of the 4 mean RRFs for both unlabeled and labeled 2,3,7,8-TCDD obtained from the triplicate analyses less than 20% RSD. 5 For unlabeled 2,3,7,8-TCDD the abundance ratio must be X>«67 and £0.90 for m/z 319.897 to 321.894. 5 c. d. Figure 1. 5 Performance evaluation sample score sheet. E-9 Score Achieved �Possible Score e. f. The abundance ratios must be X>.67 and £0.90 for 331.937 to 333.934 for 13C12-2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD. 5 Results must be documented with appropriate SIM traces, labeled with the corresponding EPA sample numbers, and calculations. 5 Performance Checks: a. b. c. d. GC resolution and MS resolution checks performed at the beginning and end of each 12-hour period. 5 Results of performance checks properly recorded on Form H-4. 5 MS Resolution: PFK (or alternate) tune shows appropriate mass resolution (Section 8.2, Exhibit D) with mass assignment accuracy within _+_5 ppm. 5 GC Resolution: chromatograms meet the criteria specified in Section 8.1, Exhibit D. 5 Routine Calibration: a. b. c. d. e. f. g. Performed each 12 hours, after MS and GC resolution checks, using HRCC2. 5 Results of routine calibrations properly reported on Forms H-3 and H-8. 5 For unlabeled 2,3,7,8-TCDD: abundance ratio must be X)«67 and £0.90 for m/z 319.897 to 321.894. 5 Abundance ratio correct for isotopically labeled standards (e.g., 331.937/333.934 must be M>*67 and £0.90 for C,2-2,3,7,8-TCDD and 13C12-1,2,3,4-TCDD). 5 Response factors [RRF(I) and RRF(II)] are within ^202 of the mean of the respective initial calibration response factors. 5 Signal-to-Noise (S/N) Ratio: SIM traces for 2,3,7,8-TCDD demonstrate S/N of ^2.5. 5 Results documented with appropriate SIM traces and calculations. Subtotal II Figure 1. (Continued). E-10 5 105 Score Achieved �Possible Score III. Performance Evaluation (PE) Sample Data (Scores to be determined for each sample in the PE set) 1. Forms H-l and H-9 properly filled out for sample. 2. Measured concentration of unlabeled 2,3,7,8-TCDD within acceptance window established by EPA. 40 Estimated concentration of total TCDDs within acceptance window established by EPA. 20 3. 4. I d e n t i f i c a t i o n Criteria for 2,3,7,8-TCDD: a. b. c. d. e. 5. 5 R e t e n t i o n time (RT) (at maximum peak height) of the sample component m/z 319.897 is within -1 to +3 seconds of the m/z 331.937 1 3 C 1 2 2,3,7,8-TCDD internal standard peak. 10 The ion current responses for m/z 258.930, 319.897 and 321.894 must reach a maximum simultaneously (_+! second) and must be >2.5 times noise level. 10 The m/z 319.897/321.894 r a t i o is X).67 and £0.90. 10 The m/z 331.937/333.934 r a t i o is X).67 and £0.90. 5 The S/N r a t i o for m/z 331.937 and 333.934 is ^2.5. 5 Identification Criteria for other TCDDs: a. b. c. Retention time must fall into window established by GC performance check. 5 The ion current responses for m/z 258.930, 319.897, and 321.894 reach a maximum simultaneously (_+! second) and are _>2.5 times noise level. 10 The m/z 319.897/321.894 ratio is 20.67 and £0.90. Figure 1. (Continued). E-ll 5 Score Achieved �Possible Score 6. 7. 8. 9. 10. Concentrations of unlabeled TCDDs are calculated according to D-13.1. 10 Duplicate analysis values agree within ±50%. 10 Estimated detection limits calculated according to D-13.5. 10 Percent recovery of >40 and <120%. 13 C,2~2,3,7,8-TCDD 10 Results documented with appropriate SIM traces and calculations. 20 Subtotal III Total Figure 1. 185 290 (Continued). E-12 Score Achieved �EVENT SEQUENCE FOR ON-SITE LABORATORY EVALUATION I. Meeting with Laboratory Manager and Project Manager Introduction; discuss purpose of visit; discuss problems with data submitted by the laboratory. II. Verification of Personnel Review qualification of contractor personnel in place and committed to project (Section I, SES). III. Verification of Instrumentation Review equipment in place and committeed to project (Section II, SES). The Contractor must demonstrate adequate equipment redundancy, as defined in SES, Section II.D., to ensure his capability to perform the required analyses in the required time. IV. Quality Control Procedures Walk through the laboratory to review: 1. 2. 3. 4. 5. 6. 7. 8. 9. Sample receiving and logging procedures, Sample and extract storage area, Procedures to prevent sample contamination, Security procedures for laboratory and samples, Safety procedures, Conformance to written SOPs, Instrument records and logbooks, Sample and data control systems, Procedures for handling and disposing of hazardous materials, 10. Glassware cleaning procedures, 11. 12. 13. 14. V. Status of equipment and its availability, Technical and managerial review of laboratory operations and data package preparations, Procedures for data handling, analysis, reporting and case file preparation, and Chain-of-custody procedures. Review of Standard Operating Procedures (SOPs) Review SOPs with the Project Manager to assure that the laboratory understands the dimensions and requirements of the program. VI. Identification of Needed Corrective Actions Discuss with the Project Manager the actions needed to correct weaknesses identified during the site inspection, PE sample analysis or production of Figure 2. Event Sequence for On-Site Laboratory Evaluation. E-13 �reports (hard copies and, if appropriate, manual calculations) and documentation. Determine how and when corrective actions will be documented, how and when improvements will be demonstrated, and identify the contractor employee responsible for corrective actions. VII. Previously Identified Problems Check the most recent SES to verify that all previously identified problems have been corrected. VIII. Identification of New Problems a. Discuss any weaknesses identified in the performance evaluation sample analyses and reports. b. Discuss any weaknessess identified in this site inspection. Figure 2. E-14 (Continued) �SITE EVALUATION SHEET Laboratory: Date: Location: EVALUATORS Name Organization 1. 2. 3. 4. 5. 6. 7. I. Laboratory Personnel Committed to Project: A. Project Manager (responsible for overall technical effort) Name: Title: B. GC/MS Operator: Experience:* (one year minumum) C. GC/MS Data Interpreter: Experience:* (two year minimum) D. Person responsible for sample exraction, column chromatography and extract concentration: Experience:* (one year minumum) E. Person(s) responsible for calculations and report preparation: Hardcopy Reports: F. Person responsible for handling, storage and (if appropriate) preparation of solutions of standard compounds: *Experience is deemed to mean "more than 50 percent of the person's productive work time." Figure 3. Site Evaluation Sheet. E-15 �G. Person responsible for standards preparation/storage: H. Person responsible for record keeping: I. Quality Assurance Officer: J. Personnel checklist ( ) Yes ( ) No 1. Do personnel assigned to this project have the appropriate level and type of experience to successfully accomplish the objectives of this program? 2. Is the organization adequately staffed to meet project requirements in a tmely manner? ( ) Yes ( ) No 3. Does the Laboratory Quality Assurance officer report to senior management levels? ( ) Yes ( ) No 4. Was the Quality Assurance officer available during the evaluation? II. ( ) Yes ( ) No Laboratory Equipment A. Gas chromatograph(s)* Manufacturer and Model: Installation Date: Type of Capillary Column Injection System: Capillary Column to be used (length, ID, coating, etc.): Necessary Ancillary Equipment (gases, syringes, etc.): B. High Resolution Mass Spectrometer(s)* Static Resolution Capability (10,000 min.): Peak matching system: Manufacturer and Model: Installation Date: Pertinent Modifications: Peak Matching System/Accuracy (Mfg. spec.): C. Data System(s)* Manufacturer and Model: If more than one GC/MS/DC, indicate system 1,2,3, etc., by numbering components with 1,2,3, etc. Figure 3. (Continued). E-16 �Installation Date: Software Version Identifier: Appropriate selected ion monitoring software/hardware ( ) Yes ( ) No Capability to produce hard copies of computergenerated information ( ) Yes ( ) No D. Evidence that at least one GC/MS/DS system can be reasonably expected to be operating acceptably at any given time: ( ) More than one adequate GC/MS/DS system is available in-house, (i.e.,meeting requirements specified in SOW Section 6.1, Exhibit D). ( ) Appropriate in-house replacement parts and trained service personnel are available. ( ) A service contract is in place with guaranteed response time (specify type of contract and limitations). ( ) Voltage control devices are used on major instruments; isolated circuits are used. ( ) Other (specify) III. Facilities Checklist A. Does the laboratory appear to have adequate workspace (120 sq. feet, 6 linear feet of unencumbered bench space per analyst)? ( ) Yes ( ) No B. Does the laboratory have a source of distilled/ demineralized water? ( ) Yes ( ) No C. Is the analytical balance located away from draft and areas subject to rapid temperature changes or vibration? ( ) Yes ( ) No D. Has the balance been calibrated within one year by a certified technician? ( ) Yes ( ) No E. Is the balance routinely checked with class S weights before each use and the results recorded in a logbook? ( ) Yes ( ) No F. Is the laboratory maintained in a clean and organized manner? ( ) Yes ( ) No Figure 3. (Continued). E-17 �G. Is the facility designed for hazardous organic chemical analysis? ( ) Yes ( ) No 1. Is ventilation provided in the sample preparation areas? ( ) Yes ( ) No 2. Are vented hoods available and adequately vented in the sample preparation areas? ( ) Yes ( ) No 3. Are the hoods equipped with charcoal and HEPA filters? ( ) Yes ( ) No 4. Are instruments, including GC/MS pumps, vented into hoods or control devices such as charcoal traps? ( ) Yes ( ) No H. Are adequate secured facilities provided for storage of samples, extracts, and calibration standards, including cold storage? I. Are the temperatures of the cold storage units recorded daily in logbooks? ( ) Yes ( ) No J. Are chemical waste disposal policies/procedures in place? ( ) Yes ( ) No K. IV. ( ) Yes ( ) No Is the laboratory secure? ( ) Yes ( ) No Analysis Control Checklist A. Do the project personnel have SOPs for the required activities? ( ) Yes ( ) No B. Is a logbook maintained for each instrument and is information such as calibration data and instrument maintenance continually recorded? ( ) Yes ( ) No C. Do the analysts record bench data in a neat and accurate manner? ( ) Yes ( ) No D. Standards 1. Are fresh analytical standards prepared at a frequency consistent with good QC? ( ) Yes ( ) No 2. Are reference materials properly labeled with concentrations, date of preparation, and the identity of the person preparing the sample? ( ) Yes ( ) No 3. Is a standards preparation and tracking logbook maintained? ( ) Yes ( ) No Figure 3. (Continued). E-18 �4. Are working standards traceable to EPA standards or validated against EPA standards? V. Documentation/Tracking ( ) Yes ( ) No Checklist A. Is a sample custodian designated? name of sample custodian. Name: If yes, ( ) Yes ( ) No B. Are the sample custodian's procedures and responsibilities documented? If yes, where are these documented? ( ) Yes ( ) No Are the chain-of-custody procedures documented? ( ) Yes ( ) No C. Are written Standard Operating Procedures (SOPs) developed for receipt of samples? If yes, where are the SOPs documented (laboratory manual, written instructions, etc.)? ( ) Yes ( ) No D. Are quality assurance procedures documented and available to the analysts? If yes, where are these documented? ( ) Yes ( ) No E. Are written Standard Operating Procedures (SOPs) developed for compiling and maintaining sample document files? If yes, where are the SOPs documented (laboratory manual, written instructions, etc.)? ( ) Yes ( ) No F. Are the magnetic tapes stored in a secure area? ( ) Yes ( ) No G. ( ) Yes ( ) No Are samples that require preservation stored in such a way as to maintain their integrity? If yes, how are the samples stored? Documentation/Notebooks Checklist A. Is a permanently bound notebook with preprinted, consecutively numbered pages being used? ( ) Yes ( ) No B. Is the type of work clearly displayed on the notebook? ( ) Yes ( ) No C. Is the notebook maintained in a legible manner? ( ) Yes ( ) No D. Are entries noting anomalies routinely recorded? ( ) Yes ( ) No Figure 3. (Continued). E-19 �E. ( ) Yes ( ) No F. Are inserts (i.e. chromatograms, computer printouts, etc.) permanently affixed to the notebook and signed across insert edge and page? ( ) Yes ( ) No G. Has the supervisor of the individual maintaining the ( ) Yes ( ) No notebook personally examined and reviewed the notebook periodically, and signed his/her name therein, together with the date and appropriate comments as to whether or not the notebook is being maintained in an appropriate manner? H. VI. Has the analyst avoided obliterating entries or the use of a pencil? Where applicable, is the notebook holder referencing reports or memoranda pertinent to the contents of an entry? ( ) Yes ( ) No Quality Control Manual Checklist Does the laboratory maintain a Quality Assurance/ Quality Control (QA/QC) Manual? ( ) Yes ( ) No Does the manual address the important elements of a QA/QC program, including the following: ( ) Yes ( ) No A. Personnel ( ) Yes ( ) No B. Facilities and equipment C. Operation of instruments ( ) Yes ( ) No D. Documentation of Procedures ( ) Yes ( ) No E. Procurement and inventory practices ( ) Yes ( ) No F. Preventive maintenance ( ) Yes G. Reliability of data ( ) Yes ( ) No H. Data validation ( ) Yes ( ) No I. Feedback and corrective action ( ) Yes ( ) No J. Instrument calibration ( ) Yes ( ) No K. Recordkeeping ( ) Yes ( ) No L. Internal audits ( ) Yes ( ) No Figure 3. ( ) Yes ( ) No (Continued). E-20 ( ) No �Are QA/QC responsibilities and reporting relationships clearing defined? Have standard curves been adequately documented? ( ) Yes ( ) No Are laboratory standards traceable? ( ) Yes ( ) No Are quality control charts maintained for each routine analysis? ( ) Yes ( ) No Do QC records show corrective action when analytical results fail to meet QC criteria? ( ) Yes ( ) No Do supervisory personnel review the data and QC results? VII. ( ) Yes ( ) No ( ) Yes ( ) No Data Handling Checklist Are data calculations checked by a second person? ( ) Yes ( ) No Are data calculations documented? ( ) Yes ( ) No Do records indicate corrective action that has been taken on projected data? ( ) Yes ( ) No Are limits of detection determined and reported properly? ( ) Yes ( ) No Are all data and records retained for the required amount of time? ( ) Yes ( ) No Are quality control data (e.g., standard curve duplicates) accessible for all analytical ( ) Yes ( ) No results? VIII. Summary Do responses to the evaluation indicate that project and supervisory personnel are aware of QA/QC and its application to the project? ( ) Yes ( ) No Do project and supervisory personnel place positive emphasis on QA/QC? ( ) Yes ( ) No Have responses with respect to QA/QC aspects of the project been open and direct? ( ) Yes ( ) No Has a cooperative attitude been displayed by all project and supervisory personnel? ( ) Yes ( ) No Figure 3. (Continued). E-21 �Does the organization place the proper emphasis on quality assurance? ( ) Yes ( ) No Have any QA/QC deficiencies been discussed before leaving? ( ) Yes ( ) No Is the overall quality assurance adequate to accomplish the objectives of the project? ( ) Yes ( ) No Have corrective actions recommended during previous evaluations been implemented? ( ) Yes Are any corrective actions required? ( ) Yes ( ) No list the necessary actions below. E-22 If so, ( ) No �TECHNICAL REPORT DATA (fltat ftta Instructions on the reverse before completing) 2. 3. RECIPIENT'S ACCESSION NO. 1. REPORT NO. 4. TITLE AND SUBTITLE 5. REPORT DATE PROTOCOL FOR THE ANALYSIS OF 2,3,7,8-TETRACHLORODIBENZOt 6. PERFORMING ORGANIZATION CODE p-DIOXIN BY HIGH-RESOLUTION GAS CHROMATOGRAPHY/HIGHRESOLUTION MASS SPECTROMETRY '. AUTHOR(S) 8. PERFORMING ORGANIZATION REPORT NO. J. S. Stanley and T. M. Sack 0. 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, Missouri 64110 11. CONTRACT/GRANT NO. Contract Number SAS 1576X 12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED Environmental Monitoring Systems Laboratory - LV, NV Office of Research and Development U.S. Environmental Protection Agency Las Vegas, NV 89114 14. SPONSORING AGENCY CODE EPA/600/07 16. SUPPLEMENTARY NOTES Project Officer - Werner F. Beckert, Environmental Monitoring Systems Laboratory Las Vegas, NV 89114 16. ABSTRACT An analytical protocol for the determination of 2,3,7,8-tetrachlorodibenzo-pdioxin (TCDD) and total TCDDs in soil, sediment and aqueous samples using highresolution gas chromatography/high-resolution mass spectrometry (HRGC/HRMS) was developed using the best features of several candidate methods and input from experts in the field. Preliminary tests led to refinements of the chromatographic cleanup procedures and corresponding changes in the protocol. A final single-laboratory evaluation of the refined protocol, consisting of triplicate analyses of five solid and five aqueous samples showed that the method is useful for the determination of 2,3,7,8-TCDD and total TCDDs at concentrations from 10 to 200 pg/g (ppt) in soils and 100 to 2,000 pg/L (ppq) in aqueous samples. Based on the data generated and on the evaluation of several options, parts of the protocol were modified at the EMSL-LV to lower the quantitation limit for TCDD to 2 ppt in soil/sediments and to 20 ppq in aqueous samples. KEY WORDS AND DOCUMENT ANALYSIS 7. IB. DISTRIBUTION STATEMENT b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group IB. SECURITY CLASS (This Report/ DESCRIPTORS 21. NO. OF PAGES UNCLASSIFIED 20. SECURITY CLASS (This page! RELEASE TO PUBLIC EPA Fwm 2220.1 (*.•*. 4-77) UNCLASSIFIED PNKVIOUS COITION is OMOLCTC 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. 196 Folder The folder containing the original item. 5521 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 Stanley, John S. Thomas M. Sack 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 1986-01-01 Title A name given to the resource Protocol for the Analysis of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin by High-Resolution Gas Chromatography/High-Resolution Mass Spectrometry with attached letter transmitting the report to Alvin L. Young, from Ronald K. Mitchum, Director, Quality Assurance Di ao_seriesVIII https://www.nal.usda.gov/exhibits/speccoll/files/original/12d389d8b4284bd4449487c863588d4a.pdf 2ac54d1125819827e8149b24fc3c8d5c PDF Text Text Item! Number °5463 Author Stanley, John S. Corporate Author United States Environmental Protection Agency (EPA), D RODOrt/ArtiCto TItto Methods of Analysis for Polychlorinated Dibenzo-pDioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) in Biological Matrices - Literature Review and Preliminary Recommendations: Task 6 Final Report Journal/Book TItto Year 1984 Month/Day February Gator Number of tongas DflSCrinton NOtRS * D 121 EPA Prime Contract No. 68-01 -5915 MRI Project No. 4901-A (6) EPA 560/5-84-001 Friday, March 15, 2002 Page 5463 of 5571 �United Slates Environmental Protection Agency Toxic Substances vvEPA METHODS OF ANALYSIS FOR POLYCHLORINATED DIBENZO-p-DIOXINS (PCDDs) AND POLYCHLORINATED DIBENZOFURANS (PCDFs) IN BIOLOGICAL MATRICES LITERATURE REVIEW AND PRELIMINARY RECOMMENDATIONS �METHODS OF ANALYSIS FOR POLYCHLORINATED DIBENZO-£-DIOXINS (PCDDs) AND POLYCHLORINATED DIBENZOFURANS (PCDFs) IN BIOLOGICAL MATRICES LITERATURE REVIEW AND PRELIMINARY RECOMMENDATIONS by John S. Stanley TASK 6 FINAL REPORT February 16, 1984 EPA Prime Contract No. 68-01-5915 MRI Project No. 4901-A(6) 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. David P. Redford, Task Manager Mr. Daniel Heggem, Task Manager �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. �PREFACE This report presents a literature review of the analytical methods used for the measurement of polychlorinated dibenzo-£-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in human adipose tissue. Also included in this report are recommendations from a meeting of scientists recognized for their efforts in PCDD and PCDF analyses held April 27th and 28th at MRI. This work was accomplished on MRI Project No. 4901-A, Task 6, "Planning Survey and Analysis Projects," for the U.S. Environmental Protection Agency (EPA Prime Contract No. 68-01-5915). The review was conducted and the document prepared by Dr. John S. Stanley, with assistance from Jerry Hurt, Barbara Mitchell, Kathy Funk, Lanora Moore, Cindy Melenson, Carol Shaw, Gloria Sultanik, Judy Daniels and Mary Walker. MRI would also like to thank the people listed in Appendix A for their cooperation, as well as David Redford, Madeline O'Neill-Dean and Daniel Heggem of FSB/OTS, EPA. MIDWEST RESEARCH INSTITUTE fthn E. Going Program Manager Approved: James L. Spigarelli, Director Analytical Chemistry Department �CONTENTS Preface Figures Tables List of Terms, Abbreviations, and Symbols 1. 2. Summary Introduction 3. Literature Acquisition and Review Procedure ii iv vi viii 1 2 Sources of information Review procedure 4. Analytical Methods - A Review Extraction Cleanup Instrumental analysis 5. Applicable Techniques - Recommendations Discussion meeting summary Discussion meeting recommendations 5 5 5 8 8 12 25 72 72 77 Appendices A. Invited participants B. Discussion meeting schedule of events C. Bibliography 111 82 89 92 �FIGURES Number 1 2 3 RP-HPLC fractionation chromatograms of (a) calibration standard and (b) European refuse incineration fly ash demonstrating the application for collection of PCDDs by homolog 15 Schematic of EPA sample preparation procedures for preparation procedures for preparation of biological matrices for TCDD analyses 17 Schematic presentation of the sample presentation used by FDA and laboratories and the New York State Department of Health in collaborative study of fish sample preparation and analysis 4 5 6 18 Flow diagram for enrichment and fractionation of PCDDs and PCDFs from tissue samples (FWS procedure) 19 Schematic for sample preparation for PCDD analysis by the Dow analytical approach 20 Glass chromatography columns for sample cleanups: A = silica, B = 22% sulfuric acid on silica, C = 44% sul- furic acid on silica, D = 33% 1 M sodium hydroxide on silica, E = 10% AgNOs on silica, and F = basic alumina . . 7 GC/ECD chromatograms of extracts from unfortified catfish (1/20 of the sample extract) 8 9 10 11 21 23 Range of application of some analytical techniques for dioxins 26 PGC/MS chromatogram of PCDD homologs extracted from incinerator fly ash sample 27 Comparative 2,3,7,8-TCDD PGC/MS mass chromatograms for electrostatic fly ash (a) after RP-HPLC, and (b) subsequent silica-HPLC 29 Separation of PCDD-isomers by GC/MS using a high resolution capillary column 30 IV �FIGURES (continued) Number 12 Page Mass chromatograms (m/e 320) of a composite pyrolyzate sample showing elution of all 22 TCDD isomers on HRGC columns 31 HRGC chromatogram of a mixture of the 22 TCDD isomers on glass and fused silica capillary columns (60 m) coated with SP-2330 and SP-2340, respectively, indicating isomer specified separation for 2,3,7,8-TCDD 32 Electron capture chromatograms of (a) entire nonphenolic fraction, (b) first microcolumn fraction containing chlorodiphenyl ethers (c) second basic alumina microcolumn fraction containing chlorodibenzo-£-dioxin and chlorodibenzofurans 40 15 HRGC/MS-SIM chromatogram of TCDD analysis 43 16 Method detection limit versus final extract volume and initial sample size assuming a GC/MS instrumental detection limit of 5 pg/pl on-column 52 Statistical treatment of validation data for 2,3,7,8-TCDD and OCDD in human milk samples 62 Statistical treatment of reported concentrations versus concentrations of TCDD actually added to standard solutions and beef adipose 63 Schematic of proposed analytical method using high resolution mass spectrometry (HRMS) 75 Schematic of proposed analytical method using low resolution mass spectrometry (LRMS) 76 Example of possible interlaboratory organization 81 13 14 17 18 19 20 21 �TABLES Number 1 Molecular Formula, Molecular Weight, and Number of Isomers of PCDD 3 2 Sample Types Analyzed for TCDD 3 3 Criteria for Rating Published PCDD Analytical Methods. . . . 6 4 Relative Efficiency of Various Methods Used at Each Stage of Analysis 7 Estimated Half-Lives ( ! of Several Dioxins in Refluxing t) KOH Solutions. . . . ? 9 5 6 Effect of Potassium Hydroxide Concentration, Time, and Temperatures on Polychlorodibenzo-£-dioxin Stability . . . 10 7 A Listing of Some Cleanup Procedures 13 8 Resources Required to Extract and Clean Up Fish Samples. . . 22 9 Summary of GC/MS-SIM Results of Study of TCDD ExtractionCleanup 24 10 Response From Possible Environmental Contaminants 34 11 EPA Phase I Dioxin Implementation Plan Beef Fat Samples Analyzed for TCDD 36 Some Compounds that may Interfere with the Determination of TCDD at m/z Values of 319.8966 and 321.8936 37 12 13 Interferences of Selected Chemical Families in MS Determination of PCDFs and PCDDs 41 14 Partial Scan Confirmation for TCDD 44 15 Range of Reported Percent Relative Abundances for Most Intense Ion in Isotope Clusters From Electron Impact Mass Spectra of the Chlorinated Dibenzo-£-dioxins 45 Area Response Factors of PCDDs Relative to 1,2,3,4-TCDD at m/z 322 47 16 vi �TABLES (continued) Number 17 Comparison of Relative Peak Ratios of PCDDs Through a Glass Jet and Silicone Membrane Separator 47 Exact Masses and Relative Isotope Abundances of Major Molecular Cluster Ions for PCDDs 50 Feasibility Study for the Quantitative Determination of TCDD in QA Tissue Samples 53 20 Detection Limits for TCDD in Various Samples 54 21 Percent Recovery of Internal Standard and Percent Accountability for Native Dioxins Spiked into Control Milk Homogenate 57 Summary of Some Published Method Validation Data for 2,3,7,8-TCDD Recovered From Fortified Biological Matrices. 59 Results of Recovery Tests Performed on the Analytical Procedure, or Its Single Parts 60 Interlaboratory Studies and Method Validations for the Analysis of Tetrachlorodibenzo-2-dioxins (TCDD) 64 25 Results of Analysis of TCDD in Human Adipose Tissue 66 26 Results of Interlaboratory Validation Studies 68 27 Concentration of 2,3,7,8-TCDD in Fish Samples From Interlaboratory Study 69 Percent Recoveries of Internal Standard TCDD in the Interlaboratory Study 70 18 19 22 23 24 28 VII �LIST OF TERMS, ABBREVIATIONS, AND SYMBOLS Accuracy Closeness of analytical result to "true" value. AOAC Association of Official Analytical Chemists. Congener One of 75 PCDDs or 135 PCDFs, not necessarily the same homolog. DDE 1,1, -Dichloro-2,2-bis (j>-chlorophenyl)ethylene. DDT 1,1,l-Trichloro-2,2,-bis(£-chlorophenyl)ethane. 2,4-D 2,4-Dichlorophenoxyacetic acid. ECD Electron capture detector. El Electron impact ionization (mass spectrometry). EIMS Electron impact mass spectrometry. FID Flame ionization detector. GC Gas liquid chromatography (column type unspecified). GC/MS Gas liquid chromatography/mass spectrometry (ionization mode unspecified). HCDD Hexachlorodibenzo-£-dioxin. HpCDD Heptachlorodibenzo-£-dioxin. Homolog One of the eight degrees of chlorination of PCDDs and PCDFs. HPLC High performance liquid chromatography. HRGC silica. High resolution gas chromatography, glass or fused Vlll �HRMS High resolution electron impact mass spectrometry. Internal standard Standards used expressly for quantitation added to sample extract immediately prior to the analytical determination. Internal standards are used for PCDD and PCDF analyses to accurately measure recoveries of spiked surrogate compounds. Isomer One of up to 22 PCDDs or 38 PCDFs possessing the same degree of chlorination (1,2,3,4-TCDD and 2,3,7,8TCDD are different isomers). KOH Potassium hydroxide. LOD Lower limit of detection (see also MDL). Lowest concentration at which an analyte can be identified as present in a 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. LRMS Low resolution mass spectrometry. MDL Method detection limit. Mean Arithmetic mean. MS Mass spectrometry. m/z Mass-to-charge ratio. NRCC National Research Council of Canada. OCDD Octachlorodibenzo-£-dioxin. PCB Polychlorinated PCDD Polychlorinated dibenzo-£-dioxin (including monochlorodibenzo-£-dioxins). PCDF Polychlorinated dibenzofuran monochlorodibenzofuran). PGC Packed column gas liquid chromatography. ppb Parts per billion (1 x 10~9 g/g, ng/g). ppm Parts per million (1 x 10 6 g/g, IX biphenyl. (including �ppt Parts per trillion (1 x 10 12 g/g, pg/g). Precision Reproducibility of an analysis, measured by standard deviation (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. RP-HPLC Reverse phase high performance liquid chromatography. RSD Percent relative standard deviation (SD/mean x 100). SD Standard deviation. Sensitivity The slope of instrument response with respect to the amount of analyte. Also used colloquially to refer to lowest detectable amount of analyte. SIM Selected ion monitoring (also mid or mass fragmentography). Surrogate Standard compounds added to the sample prior to any analytical manipulations for the express purpose of measuring recovery through extraction, cleanup, etc., and to provide true internal standard quantitation. TCDD Tetrachlorodibenzo-£-dioxin. 13 C12-TCDD Carbon-13 stable isotope labeled TCDD. 37 C14-TCDD Chlorine-37 stable isotope labeled TCDD. 2,4,5-T 2,4,5-Trichlorophenoxyacetic acid. x �SECTION 1 SUMMARY The published literature on polychlorinated dibenzo-£-dioxins (PCDDs) analyses for biological matrices is reviewed. The analytical methods are discussed for sample extraction, cleanup, and instrumental analysis. This report also presents a synopsis of a discussion meeting concerning the analysis of polychlorinated dibenzo-£-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) held at Midwest Research Institute (MRI) on April 27 and 28, 1983. The primary objective of this meeting was to define the needs of an analytical method for the analysis of PCDDs and PCDFs in human adipose tissue. This method will be used in the future for population studies. Several major programs were identified as necessary to achieve these goals, These included (a) the need for establishing a repository of PCDD/PCDF standards of known quality; (b) the organization and implementation of a strong quality assurance program; (c) the acquisition of sufficient human adipose tissue to generate a homogeneous sample matrix for the QA program; (d) independent studies of extraction procedures using bioincurred radiolabeled PCDDs; (e) intralaboratory ruggedness testing of a proposed analytical method; and (f) interlaboratory evaluation of the proposed method. Simultaneous activity in several of these areas is necessary in the coming months. �SECTION 2 INTRODUCTION Polychlorinated dibenzodioxins (PCDDs) are a series of compounds with varying chlorine atom substitution on the dibenzo-£-dioxin parent compound. Table 1 presents the 75 possible positional isomers distributed from monochloroto octachlorodibenzo-ja-dioxin. The dioxin considered to be most toxic is the 2,3,7,8-tetrachlorodibenzo-£-dioxin (TCDD). The potential long-term consequences of exposure to PCDDs, particularly 2,3,7,8-TCDD, are an issue of increasing public concern. Highly intense analytical and toxicological investigations have been conducted in recent years as a result of the presence of TCDD as an unexpected contaminant in the defoliant, Agent Orange, which is a formulation of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), 2,4-dichlorophenoxyacetic acid (2,4-D), and related ester herbicides. Also, the accidental release of TCDD from a factory near Seveso, Italy, the discovery of TCDD contaminated soil in Missouri, and the indication that PCDDs are emitted from numerous combustion sources have generated a demand for highly sensitive and specific analytical measurements for these contaminants in a wide spectrum of matrices. Table 2 presents some of the highly diverse sample matrices that have been analyzed for PCDDs, particularly for 2,3,7,8-TCDD. The need to determine PCDDs in these diverse matrices has resulted in the development of a number of well-documented approaches to analysis. Although the exact approaches vary between laboratories, the basic requirements of all methods include quantitative extraction, efficient cleanup and separation from the bulk of the sample matrix and chlorinated compounds that might act as interferences, and sensitive and specific methods of instrumental analysis. The early work of Baughman and Meselson (1973) has been refined and expanded to accommodate complex matrices and to achieve detection at parts per trillion levels in numerous samples. The overall objective of this review and preliminary method recommendation is to assist the EPA's Office of Toxic Substances (OTS) in proposing an analytical method for PCDDs in human adipose tissue in conjunction with the Veterans Administration's (VA) Agent Orange study. The Field Studies Branch of EPA/OTS has for many years been directly involved with the EPA's National Human Monitoring Network. The Network has adipose specimens archived which may provide evidence of exposure to Agent Orange. Part of the overall plan is (a) the identification of specimens for which exposure can be documented, and (b) the analysis of those specimens for evidence of exposure. The second part of the study is addressed in this document. �TABLE 1. MOLECULAR FORMULA, MOLECULAR WEIGHT, AND NUMBER OF ISOMERS OF PCDD Chlorinated dibenzo-£-dioxin Molecular formula Total number of isomers C12H7C102 Monochloro (MCDD) 2 10 Dichloro (DCDD) Trichloro (T3CDD) 14 ^12^501302 22 Tetrachloro (TCDD) Pentachloro (P5CDD) Hexachloro 14 0^2^301502 10 (HCDD) 2 C12HC1702 Heptachloro (HpCDD) 1 Octachloro (OCDD) TABLE 2. SAMPLE TYPES ANALYZED FOR TCDD Human milk Human adipose tissue' Beef liver o Beef adipose tissue Beef blood Wildlife samples - deer, elk, shrew, etc. Fish3 Source: a Water, soil and sediment Workplace air samples Fly ash samples Gasoline and diesel automobile exhaust Chemical products Chemical process streams rt Municipal incinerator Harless, R. L., and R. G. Lewis, "Quantitative Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin Residues by Gas Chromatography/Mass Spectrometry," in Chlorinated Dioxins and Related Compounds. Impact on the Environment, 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds.), Pergamon Press, 1982, pp. 25-36. 2,3,7,8-TCDD residues were confirmed and quantified. Presence of other TCDD isomers confirmed in various samples. �This report reviews methods used for analysis of PCDDs in biological matrices. Section 3 describes the literature review procedures. Analytical methods are reviewed in Section A in terms of sample preparation, extraction, instrumental analysis, quantitation, and quality assurance. The advantages of specific methodologies, the purpose of specific steps, and the limitations of the particular technology are discussed. Section 5 presents a synopsis of a meeting held at MRI to discuss analytical approaches to the analysis of human adipose for PCDDs and PCDFs. Section 5 also provides recommendations for identifying an analytical method and organization of major program areas for method validation and sample analyses. Appendix A provides a list of persons who provided peer reviews and were invited to attend the meeting held at MRI. Appendix B provides the schedule of discussion topics for that meeting. Appendix C is a bibliography of references compiled and reviewed for this literature review. �SECTION 3 LITERATURE ACQUISITION AND REVIEW PROCEDURE This section describes how the published literature on analytical techniques for PCDDs in biological matrices was reviewed and presents in tabular form some suggested criteria for rating published methods. SOURCES OF INFORMATION Computerized and manual searches and relevant references in recent articles were used. Also, many documents not available in the open literature were obtained from the working files of MRI scientists professionally involved in PCDD research. Recent issues of several key journals (Analytical Chemistry, Journal of Chromatography, Journal of the Association of Official Analytical Chemists, Environmental Science and Technology) were searched manually to pick up any recent references not yet in the computer data bases. In addition, several leading scientists (Appendix A) were called to discuss analytical approaches. 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 1978, printing all references containing "polychlorinated dibenzo-£-dioxin," "PCDD," "TCDD," CAS registry numbers, and synonyms and keywords beginning with the following notations: "anal," "detn," "quant," "measure," "tissue," "milk," "adipose" and "biol." A similar search was performed on the National Technical Information Service data base (including Smithsonian Science Information Exchange) and the Toxline data base. Once the primary search data had been reviewed, it became apparent that several authors were of primary interest and all of their recent (1980 to 1983) publications were retrieved by a CA name search. These authors included H. Buser, W. Grummet, A. Dupuy, M. Gross, R. Harless, L. Lamparski, T. Nestrick, C. Rappe, D. Stalling, T. Tiernan, H. Tosine, and A. Young. References contained in primary literature and review articles were also checked to assure that no important articles had been missed by the computer search. Several articles were added to the files by these searches. REVIEW PROCEDURE All articles cited in the bibliography of this document 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 applicable key subject areas such as extraction, cleanup, HRGC/MS, method validation, interlaboratory study, etc. �The analytical methodologies for the analysis of PCDDs have previously been reviewed by several authors (Harless, 1977; Firestone, 1978; McKinney, 1978; Hass and Friesen, 1979; Buser, 1980; Cairns et al., 1980; Esposito et al., 1980; Baker, 1981; NRCC, 1981; Fishbein, 1982; Karasek, 1982; Mahle and Shadoff, 1982; Tiernan, 1983). Although these reviews were directed principally toward the final measurements with mass spectrometry, they contained a wealth of information in terms of consolidated analytical results and method performance data. The National Research Council of Canada (NRCC, 1981) and Mahle and Shadoff (1982) have directed attention to complete analytical methods. The NRCC rated analytical methods current to 1981 by the criteria listed in Table 3. None of the techniques reviewed by NRCC received the highest point rating since no method had been fully evaluated through collaborative testing. Mahle and Shadoff (1982), on the other hand, rated methods from low to high with respect to the technical aspects of extraction and cleanup, separation of isomers, and detection and quantitation. Table 4 is an example of the rating scheme reported by Mahle and Shadoff (1982). TABLE 3. CRITERIA FOR RATING PUBLISHED PCDD ANALYTICAL METHODS Essential elements Point rating 1 (highest) Complete quality assurance as described by ACS (1980). An ideally developed, evaluated method including collaborative studies. Isomer specific, extensive recovery studies, interferences removed and separation achieved through extensive chemical workup; lacks collaborative evaluation and assumes confirmation. Incompletely isomer specific, some recovery studies, interferences partially removed and partial separation achieved through chemical workup; lacks collaborative evaluation and assumes confirmation. Essentially a screening method for most homologs, interferences partially removed and partial separation through limited chemical workup; lacks collaborative evaluation and assumes confirmation. 5 Same as 4, except inadequately documented for recovery, cleanup, etc. 6 Insufficient for the present state of the art. Source: National Research Council of Canada, "Polychlorinated Dibenzo-j>dioxins: Limitations to the Current Analytical Techniques," NRCC No. 18576, ISSN 0316-0114 (1981). 6 �TABLE 4. RELATIVE EFFICIENCY OF VARIOUS METHODS USED AT EACH STAGE OF ANALYSIS Method Description Stage I: L M H sample preparation Chemical treatment and/or extraction without chromatography L + column chromatography M + HPLC Stage II: sample introduction L M H No gas chromatography (direct probe) Packed column GC Capillary column GC Stage III: mass spectrometry L M H Source: a Low resolution (300-2000) Medium resolution (> 2000-9000) High resolution (> 9000) Mahle, N. H., and L. A. Shadoff, "The Mass Spectrometry of Chlorinated Dibenzo-£-dioxins," Biomedical Mass Spectrometry, 9:45-60 (1982). L = Low, M = medium, H = high. �SECTION 4 ANALYTICAL METHODS - A REVIEW The analytical methods applicable to the measurement of PCDDs in biological matrices are discussed in this section. The quality and limitations of the applicable methods are frequently documented by referral to data published in the literature. Most of the methods reviewed in this section allow the simultaneous analysis of polychlorinated dibenzofurans (PCDFs) and PCDDs in biological matrices. EXTRACTION Reliable PCDD analyses begin with the quantitative extraction of the analytes from the sample matrix. In general, the extraction method is dependent on the sample type and the complexity of the matrix. Extraction methods used in preparing biological samples have included neutral extractions, alcoholic potassium hydroxide saponifications, and acidic digestions followed by transfer of the PCDDs into an organic solvent such as hexane, methylene chloride, or petroleum ether. Neutral extraction of fatty tissues, liver, and milk have been reported in several studies. The procedures begin with homogenization of the tissues with anhydrous sodium sulfate (Na2S04) in ratios of 1 part tissue to 4-10 parts Na2S04. The resulting dry mixture can then be Soxhlet extracted, packed into a chromatography column and eluted, or it can be blended directly with an organic solvent. Ryan et al. (1980), Albro and Corbett (1977), and Hass et al. (1978) have blended liver samples directly with chloroform and methanol, then subsequent back extracted with aqueous solutions. O'Keefe et al. (1978) have used an approach that consists of rendering the fatty sample and dissolving it in hexane. Shadoff (1980) has reported the used of a cellulose gauze to absorb the fat content of human milk samples as the first step in analyzing human milk samples for 2,3,7,8-TCDD. The cellulose gauze with the adsorbed milk sample was extracted with hexane under refluxing conditions. An additional neutral extraction procedure has been described by DeRoos et al. (1982). High pressure liquid carbon dioxide extraction of fish samples proved to be quantitative for samples (5 g) spiked at 20 to 200 parts per trillion of 2,3,7,8-TCDD. The saponification of fatty tissues with alcoholic KOH preceding the extraction of PCDDs from the matrix with an organic solvent evolved from the early work of Baughman and Meselson (1973). Modifications of this procedure have been used for preparation of most samples for analysis for 2,3,7,8-TCDD under the Dioxin Monitoring Program (BMP). The digestion carried out under the reflux conditions as presented by Baughman and Meselson (1973), however, �may lead to the destruction of the higher chlorinated homologs of the PCDDs. Table 5 presents the estimated half-lives (tj) of several PCDD compounds including the hexa-, hepta-, and octachloro-homologs, with no sample matrix in refluxing KOH solution. As indicated on Table 5, the concentrations of the octa- and heptachlorinated homologs are significantly reduced during the recommended 1.5- to 2-hr reflux step. TABLE 5. ESTIMATED HALF-LIVES (t, ) OF SEVERAL DIOXINS IN REFLUXING KOH SOLUTION3 Dioxin 1,2,3,6,7,8- and 1,2,3,7,8,9-HCDD 7 hr 1,2,4,6,7,9- and 1,2,3,4,7,8-HCDD 2 hr 1,2,3,4,6,7,8-HpCDD 23 min 1,2,3,4,6,7,9-HpCDD 16 min 1,2,3,4,6,7,8,9-OCDD 4.5 min Source: Firestone, D., JAOAC, 60:354-356, 1977. Report on Oils and Fats a Ten to 40 ng dioxin refluxed gently with 50 ml 32% aqueous KOH solution and 20 ml ethanol. b HCDD = hexachlorodibenzo-£-dioxin; HpCDD = heptachlorodibenzo-£-dioxin; OCDD = octachlorodibenzo-£-dioxin. Lamparski et al. (1978) have studied this effect in somewhat greater detail. Table 6 presents data for the decomposition of hexa- (HCDD) and octachlorodibenzo-£-dioxin (OCDD) based on the effects of KOH concentration, time, and digestion temperature. These data were generated during a study of the determination of pentachlorophenol, hexa- and octachlorodibenzo-£-dioxin in bovine milk. As can be seen from these data, lengthy periods of digestion at elevated temperatures will drastically reduce HCDD and OCDD concentrations. To avoid this problem, a less alkaline digestion matrix or shaking at room temperature rather than refluxing has been used to prepare samples for extraction. �TABLE 6. EFFECT OF POTASSIUM HYDROXIDE CONCENTRATION, TIME, AND TEMPERATURES ON POLYCHLORODIBENZO-E-DIOXIN STABILITY HCDD Initial concentration, Percent ppb decomposition OCDD Initial Percent concentration , ppb decomposition Temperature digestion, °C Digestion time, h 22 24 20 1 14 1 44 35 24 20 1 54 1 72 60 24 20 1 72 1 > 95 80 24 20 1 > 95 1 > 95 22 1 4 o.ia 10a O.la 10b 22 2 4 0.1 10b 0.1 iob Source: Percent KOH concentration Lamparski, L. L., N. H. Mahle, and L. A. Shadoff, "Determination of Pentachlorophenol, Hexachlorodibenzo-£-dioxin, and Octachlorodibenzo-£-dioxin in Bovine Milk," J. Agric. Food Chem., 26:1113-1116 (1978). a Lower detection limits are possible because no sample matrix is present, b These values are reported to one significant figure. �Tiernan and Taylor (1983, personal communication) have provided additional data reflecting that saponification at elevated temperatures also provided degradation of OCDD in beef adipose tissue. Aklaline conditions at room temperature (22°C) with shaking (12 hr) provided complete digestion of liver tissue with quantitative recovery of a chlorine-37 labeled OCDD internal standard. These researchers, however, point out that heating was necessary for complete digestion of the beef adipose tissue. Langhorst and Shadoff (1980) and Tosine et al. (1982, 1983) have provided the only published reports on the use of acid digestion of a biological sample matrix prior to the determination of PCDDs. Langhorst and Shadoff (1980) reported that 30-g samples of human milk were digested with 200 ml of concentrated HC1 prior to solvent extraction. The advantage of the extraction procedure is that it eliminates the caustic digestion that affects the stability of the higher chlorinated dioxins. The reported recoveries of stable isotopelabeled PCDDs from spiked milk homogenates ranged from an average of 36% for the tetra- to 78% for the hexachlorodibenzo-p_-dioxin. Validation data for reagent blanks were also presented and recoveries varied from an average of 34% for the tetra-, to 85% for the hexa-, to 31% for the octachlorodibenzo-pdioxin. However, it is not clear from the data presented what effect the concentrated HC1 digestion had on the recovery of these components. Regardless of the exact extraction procedure employed, the reliability of the data in roost of the studies has been supplemented by the repeated recovery of surrogate compounts spiked into the sample prior to extraction. Typically, carbon-13 or chlorine-37 stable labeled PCDDs were added at concentration levels 10 to 100 times higher than the analytical method limit of detection. In summary, three methods for the extraction of PCDDs from biological matrices have been reported, although there has been no study intended to address the advantages of one procedure over another. Brumley et al. (1981) have reported on six different extraction and cleanup procedures with one common instrument analysis approach for final analysis. However, this study lacks the specificity to identify differences arising from the various extraction techniques since all sample preparations were completed with different cleanup steps. Thus, the need remains to evaluate the three extraction procedures with a common sample source followed by a consistent cleanup procedure and final analysis. One possibility for determining the true extraction efficiency of PCDDs and PCDFs in adipose tissue with any of the three procedures will require the use of bioincurred radiolabeled compounds. Radiolabeled PCDD and PCDF compounds are used to provide a measurement independent from GC/MS techniques. This approach to study the extraction mechanism was proposed recently at MRI during a meeting to discuss approaches to the analysis of human adipose differences for PCDDs and PCDFs. 11 �CLEANUP The effective separation of PCDDs from materials coextracted from the sample matrix has required a combination of efficient cleanup techniques. The cleanup methods used for isolating PCDDs have been developed by several analysts. Table 7 is a summary of cleanup procedures used for biological matrices. The cleanup procedures reviewed include acid and base washes, liquid-liquid partition, column chromatography with alumina, florisil, silica gel, chemically modified silica, and carbon impregnated foam. Reverse phase (RP) and normal high performance liquid chromatography (HPLC) have been used to remove interferences that are chemically similar to the PCDDs and to improve isomer specificity with the final instrument determination. A large percentage of the lipid materials in tissue extracts are presumably sulfonated or saponified with the concentrated sulfuric acid or strong base washes. These procedures promote the degradation and hydrolysis of complex molecules including some pesticide residues. Many of the procedures listed in Table 7 used a concentrated sulfuric acid wash. Several of the methods followed the acid wash with a saponification step using a basic solution, typically IN KOH. As can be seen from the data presented in Table 6, there should be little or no adverse effect of the base at this concentration on the stability of the hexa- through octachlorinated PCDD homologs. Some samples however have a tendency to form emulsions with a wash procedure. The decision to use a chemically modified acid or base silica column is based on the analyst's experience. The advantages of using the impregnated column materials include less manipulation of samples, reduced exposure to active glass surface, and greater rate of sample turnover. The emulsion problem is not encountered with treated columns. However, the eluent flow from concentrated acid columns may become restricted due to impaction from precipitated or charred coextractives in samples with high concentrations of lipid and other oxidizable compounds. Langhorst and Shadoff (1980) have overcome this problem in human milk analyses by using a precolumn with a lower acid (22%) loading prior to the more concentrated acid (44%) column. The 22% acid column is a less effective reagent than the 44% acid column but is also less prone to plugging or reduced flow. The combination of these reagents was reported as quite successful. Column chroraatography following the acid/base extract treatment is used to separate PCDDs from chlorinated residues such as the organochlorine pesticides and polychlorinated biphenyls (PCBs). Alumina is the most widely used adsorbent material, as indicated in Table 7. Florisil and silica have been used in a few specific procedures as a means of separating bulk interferences preceding final separation with alumina columns. The final column chromatography step in many instances was accomplished using micro-columns of alumina (1.0 g) in disposable Pasteur pipettes. Harless et al. (1980) have used two such columns in sequence as the final cleanup step. A 10% silver nitrate impregnated silica column has been used by Lamparski et al. (1979) preceding the final column for the analysis of fish. The silver nitrate column is effective for the removal of DDE, chlorinated aliphatic hydrocarbons, and sulfides. The basic alumina column in this sequence is used primarily to separate PCBs from the PCDD-containing fraction. 12 �TABLE 7. A LISTING OF SOME CLEANUP PROCEDURES Column chromatography Wish Acid Base Acid Base Alumina Florisil Silica gel Foam charcoal RP HPLC HPLC Reference Harless et al. (1980) Harless et al. (1980) Mitchum et al. (1980) + Lamparski et »1. (1978) AgN03 Lamparski et al. (1979) O'Keefe et al. (1978) Firestone et al. (1979) Mahle et al. (1977) Baughraan and Heselson (1973) Phillipson and Puna (1980) Fanelli et al. (1980) + AgN03 AgN03 Langhorst and Shadoff (1980) Tosine (1981) Norstrom et al. (1981) Hummel (1977) Chess and Gross ( 9 0 18) Buser (1978) Baughman and Meselson (1971) Hummel (1977) Ryan and Pilon (1980) Haas et al. (1978) Haas et al. (1978) Tiernan et al. ( 9 0 18) DiDomenico et al. (1979) DiDomenico et al. (1979) TLC Levin and Nilsson (1977) Albro and Corbett (1977) Source: National Research Council of Canada (NRCC), "Polychlorinated Dibenzo-£-dioxins: to the Current Analytical Techniques," NRCC No. 18576, 1981, 172 pp. < + indicates used only as one step of the procedure, b ++ indicates two separate columns were used. 13 Limitations �The separation of PCDDs into fractions containing combinations of the various isomers prior to final instrumental analyses has been accomplished using reverse phase (RP) and/or normal HPLC technique. Lamparski et al. (1979) and Langhorst and Shadoff (1980) have used RP-HPLC cleanup to provide additional removal of contaminants (e.g., PCBs, DDE, phthalates) and to remove components that are very similar to dioxins, such as chlorinated benzylphenyl ethers. Specific fractions of the eluent from the RP-HPLC are collected for analysis of PCDDs by homolog. This approach is especially significant for studies that require data on low parts per trillion concentration levels for tetra- to octa-PCDD homologs. Typically, the low parts per trillion measurements require final concentration of sample extracts to 10-20 [Jl. Instrumental analysis for a specific PCDD homolog may consume a major portion of the extract, presenting difficulties if the need exists to include other PCDD homologs. The RP-HPLC separation of the sample extract as shown in Figure 1 allows collection of PCDDs by homolog, enabling the measurement of all PCDDs at low parts per trillion levels. This approach has been demonstrated by Langhorst and Shadoff (1980) for the analysis of tetra-, hexa-, hepta-, and octachlorodibenzo-£-dioxins in human milk. Langhorst and Shadoff (1980) have also used RP and normal silica HPLC for separation and identification of 2,3,7,8-TCDD from the other TCDD isomers in extracts from human milk. Regardless of the specific cleanup procedure, the analyst must take precautions to ensure that adsorbents are fully activated and method blanks do not yield extracts with high backgrounds. Huckins et al. (1976) have reported on some contaminants and limitations of silica gel for the chromatographic separation of polychlorinated aromatics and pesticides. The data presented in this paper implicated the presence of sulfuric acid in silica gel as responsible for producing contaminants that interfered with the analysis. It is our experience that sulfuric acid modified silica gels and batch extractions with concentrated sulfuric acid generate contaminants that appear to be oxygenated compounds with aliphatic moieties. These artifacts can be removed by base modified silica gel, batch extraction with a base and/or the use of fully activated basic alumina. As mentioned earlier, the approach to the determination of PCDDs in biological matrices is dependent on the experience of the analyst and the associated laboratory. The actual extraction and cleanup procedures practiced may differ markedly from one laboratory to another. In view of the variety of methods in use, a comparison of six different extraction and cleanup procedures was conducted by Brumley et al. (1981) with respect to the analysis of 2,3,7,8-TCDD in fish. The relative efficiency of the different methods was determined based on two criteria: (1) the relative number and amounts of undesired components present in the final extracts, and (2) the extent to which these components interfered with TCDD analysis. The objective of the study was to compare the overall efficiency of the six available analytical cleanup procedures using a common GC/MS (low resolution) analysis approach. Six fish samples were submitted to six participating laboratories including the Bureau of Foods, Food and Drug Administration (BF/FDA), Detroit District/ FDA, Dow Chemical Company, the Environmental Protection Agency (EPA), Fish and Wildlife Service (FWS), and the New York State (NYS) Department of Health. The samples were prepared for TCDD analysis according to the procedure routinely �? RP-HPLC Collection Zone Calibration Standard 1O CHCI3 Solvent *• O OCDD D O CM O i TCDDs 1 \ O • i H7DDs , I i. 1 a ISO = 2 2378 150=1 \\ » 1 & HCDDs i \ \ \ \\ \ » \ £ o £ £ > \_ A. AA w_ _y\ __A. i 2 4 6 8 10 12 14 16 18 20 22 24 (Minutes) 26 28 30 i ^ i i i i 32 34 36 38 40 42 44 42 44 RP-HPLC European Fly Ash (Municipal Refuse Incinerator) in n C"J D O OCDD CM O OCDF 0 Source: 2 4 6 8 10 12 14 16 18 20 22 24 (Minutes) 26 28 30 32 34 36 38 40 Lamparski, L. L., and T. J. Nestrick, "Determination of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-p-dioxin Isomers in Participate Samples at Parts per Trillion Levels," Anal. Chem., _52, 2045^2054 ( 9 0 . 18) Figure 1. RP-HPLC fractionation chromatograms of (a) calibration standard and (b) European refuse incineration fly ash demonstrating the application for collection of PCDDs by homolog. �used by each laboratory. All extracts were then analyzed by one laboratory using gas chromatography/mass spectrometry by selected ion monitoring (GC/MS-SIM), scanning GC/MS, and GC/ECD (electron capture detector). This study did not evaluate the overall analytical method used by any of the participating laboratories. The results of the evaluation of the cleanup efficiency did not necessarily reflect upon the validity of TCDD analyses performed by the participating laboratories using these combined cleanup and MS procedures. Figures 2 to 6 are schematic representations of the extraction and cleanup procedures used by the different participating laboratories. As part of this study, each laboratory was asked to specify the time and personnel requirements necessary for sample preparations. Table 8 is a summary of the resources required to extract and clean up fish samples. As indicated on Table 8, the EPA-neutral procedure and FWS carbon/dual column procedure provided the most rapid turnaround time. The methods that require the use of HPLC equipment were more labor intensive. Figure 7 presents data from representative sample extracts prepared by the six laboratories, as determined by GC/ECD at the BF/FDA laboratory. The results indicate that the BF/FDA, Detroit District/FDA, Dow, and FWS sample preparations provide extracts that are significantly less complex than the other approaches. Further analysis by BF/FDA of the sample extracts using low resolution GC/MS-SIM yielded the data presented in Table 9. Twelve ions were monitored, including eight ions representative of the molecular ion cluster and the loss of COC1, two ions representative of the internal standard [13Cj2] TCDD, and two ions representative of possible interferences arising from tetrachloromethoxybiphenyl. Analysis of all 12 ion chromatograms for all six of the participating laboratories indicated that only the NYS, Dow, and FWS cleanup procedures provided sample extracts with no interference at the retention time of TCDD. The summary of results (Table 9) obtained by GC/MS-SIM indicate whether TCDD was confirmed in the sample and quantitation of observed responses for the appropriate ions. Based on their findings, Brumley et al. (1981) placed the six extractioncleanup procedures into four categories. The Dow and FWS procedures were in the first category because TCDD was confirmed and quantitated and the ion currents for the 12 ions monitored indicated that the extracts were free of interferences. The NYS procedure was placed in a second category since the overall levels of coextractants appeared to be significant. The FDA and EPA procedures comprised the third and fourth categories, respectively, because of excessive amounts of coextractive, greater than 100% recovery of the surrogates, and interferences appearing for the monitored ions. 16 �Acid/Bose Procedure Neutral Procedure Alcoholic Potassium Hydroxide Saponification Homogenize with Dry Ice I I Hexane Extraction Acetonitrile Extraction I i Acid/Base Washes Partition with Acetonitrile Saturated Hexane 1 Alumina Microcolumn Fractionation (])CCI4 Florisil Column T)lO%CH 2 CI 2 ©CH2CI2 1 f Discard Discard Alumina Microcolumn Fractionation ©cc,4 Figure 2. Alumina Microcolumn Fractionation ©cc, 4 @CH2CI2 1 Discard (D 25 %CH 2 CI 2 /Hexane (2)CH2CI2 Discard Schematic of EPA sample preparation procedures for preparation of biological matrices for TCDD analyses. �FDA Acid-Base Cleanup New York State Deportment of Health Neutral Cleanup Alcoholic Potassium Hydroxide Saponificarion Neutral Extraction Blend Sample/CH2CI2/Na2SO4 I I Filter and Solvent Exchange to Hexane Hexane Extraction I 1 Magnesia -Celite 545 Column Acid/Base Washes * . (T)Eth Neutral Alumina Column Discard ©20% CCI4 Hexane i I 1 (|)CH2Cl2 Neutral Alumina i Discard 0cci4 rionsil Column (D 1 Discard (T)lO%CH2CI2/Hexane i Discard Florisil Column (f)CH2CI2 i HPLC Zorbax -CDS t (T) <T) Hexane **-/ r (|)CH2CI2 Discard HRGC/MS Figure 3. Schematic presentation of the sample preparation schemes used by FDA and laboratories and the New York State Department of Health in collaborative study of fish sample preparation and analysis. �PART I EXTRACTION ond ADSORPTION on CARBON -Solvent (C 6 H 12 /CH 2 CI 2 1 : 1 v/v) (500g 1 :4 w/w) -Potassium Silicate (30g) Remove acidics and other polar biogenic compounds that interfere with adsorption of PCDDs and PCDFs on carbon -Silica Gel (30g) 'Cesium Silicate (10g) -Silica Gel ( 6 g ) -Carbon (50mg) Glass Fibers Mixture PART II • Selective adsorption of PCDDs and PCDFs and similar residues FRACTIONATION of AROMATIC RESIDUES Cesium Silicate (0.54g) ] I Removal of residual biogenic f substances H2S04/S;iica Gel (0.47g) I -Alumina (3.65g) • •Fractionation of xenobiotic residues Fraction Compounds 0-23 mL 23-55 mL Source: Solvent 0-2%CH 2 CI 2 /C 6 H U 5-8% CH2Cl2/C6Hu PCBs, PCNs PCDDs, PCDFs Stalling, D. L., J. D. Petty, L. M. Smith, C. Rappe, and H. R. Buser, "Isolation and analysis or Polychlorinated Furans in Aquatic Samples," in Chlorinated Dioxins and Related Compounds; Impact on the Environment, 0. Hutzinger, R. W. Frei, E. Merian, F. Peschari, Eds., Pergamon Press, 1982. Figure A. Flow diagram for enrichment and fractionation of PCDDs and PCDFs from tissue samples (FWS procedure). 19 �Benzene Soxhlet Extraction Chemically Modified Classical Adsorbent Chromatography Classical Adsorbent Chromatography —(Higher Chlorinated CDDs)—i Silico-HPLC Refroctionation RP-2378/SIL-2378 2378-TCDD RP-2378/SII/1 1237-TCDD 1238-TCDD 1247-TCDD 1248-TCDD RP-2378/SIL*2 1278-TCDD RP-2378/SII/3 1246-TCDD (1249-TCDD) 1236-TCDD 1239-TCDD RP-2378/511*4 (1246-TCDD) 1249-TCDD Source: Lamparski, L. L., and T. J. Nestrick, "Determination of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-p-dioxin Isomers in Particulate Samples at Parts per Trillion Levels," Anal. Chem., 52, 2045-2054 ( 9 0 . 18) Figure 5. Schematic for sample preparation for PCDD analysis by the Dow analytical approach. 20 �20mm 20mm T 10 mm 22cm 9cm -A -B 9cm JHJ 6mm 36cm Source: Lamparski, L. L., and T. J. Nestrick, "Determination of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-£-dioxin Isomers in Particulate Samples at Parts per Trillion Levels," Anal. Chem., _52, 2045-2054 ( 9 0 . 18) Figure 6. Glass chromatography columns for sample cleanups: A = s i l i c a , R = 22% sulfuric acid on silica, C = 44% sulfuric acid on silica, D = 33% 1 M sodium hydroxide on silica, E = 10% AgN03 on silica, and F = basic alumina. 21 �TABLE 8. Cleanup method RESOURCES REQUIRED TO EXTRACT AND CLEAN UP FISH SAMPLES Analyst' s per set Number of samples per set Extraction- cleanup time, h, per set Extraction- cleanup time, h, per sample per analyst FDA acid/base HPLC 1 6 24 4 Dow dual-column/HPLC 2 4 16 8 EPA-A/B 2 4 8 4 EPA-Neutral 1 4 8 2 FWS carbon/dual column 1 6 20 3.3 NYS multi-column 1 2 16 8 Source: a Brumley, W. C., J. A. Roach, J. A. Sphon, P. A. Dreifuss, D. Andrzejewski, R. A. Niemann, and D. Firestone, "Low-Resolution Multiple Ion Detection Gas Chromatographic-Mass Spectrometric Comparison of Six Extraction-Cleanup Methods for Determining 2,3,7,8-Tetrachlorodibenzo-D-dioxin in Fish," J. Agric. Food Chem., 29:1040-1096 (1981). Time required for one or two analysts (see second column) to extract and clean up a set of samples. �16X 12 16 20 Minutes 24 28 32 12 16 20 Minutes 24 28 32 36 12 16 20 Minutes 24 28 32 36 12 36 16 20 Minutes 24 28 32 36 4X 12 16 20 Minutes 24 28 32 36 12 16 20 Minutes 24 28 32 36 4X 0 4 8 Source: Brumley, W. CL., J. A. Roach, J. A. Sphon, P. S. Dreifuss, D. Andrzejewski, R. A. Niemann, and D. Firestone, "LowResolution Multiple Ion Detection Gas Chromatographi-Mass Spectrometric Comparison of Six Extraction-Cleanup Methods for Determining 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Fish." J. Agric. Food Chem., 29, 1040-1046 (1981). Figure 7. GC/ECD chromatograms of extracts from unfortified catfish (1/20 of the sample extract). (A) BF/FDA; (B) Det/FDA; (C) Dow; (D) EPA-A/B; (E) EPA-Neut; (F) FWS; (G) NYS. The arrows indicate the retention time of 2,3,7,8-TCDD, as determined by GC of a 2,3,7,8TCDD standard solution. 23 �TABLE 9. SUMMARY OF GC/HS-SIH RESULTS OF STUDY OF TCDD EXTRACTION-CLEANUP3 Sample no. BF/FDA conf. quant. DET/FDA conf. quant. 1 no 5 no 6 2 no 67 no 3 . no 34 4 no 5 e 6 no Source: NYS conf. quant. EPA-A/B conf. quant. EPA-neut conf. quant. Dowb conf. quant . FWS conf. quant. no c no c no c no c no 9 89 yes 77 no c no c yes 67 yes 47 no 42 yes 57 no c no c yes 25 yes 22 188 no 99 yes 128 d d no c yes 113 yes 117 e no 53 yes 38 d d d d yes 45 yes 56 178 no 199 yes 107 d d d d yes 100 yes 96 Brumley, W. C. , J. A. Roach, J. A. Sphon, P. A. Dreifuss, D. Andrzejewski, R. A. Niemann, and D. Firestone, "Low-Resolution Multiple Ion Detection Gas Chromatographic-Mass Spectroraetric Comparison of Six Extraction-Cleanup Methods for Determining 2,3 ,7,8-Tetrachlorodibenzo-gdioxin in Fish ," J. Agric. Food Chem. , 29:1040-1046 (1981). a Confirmation of the identity of TCDD was obtained if the responses of the 12 monitored ions for the sample extract were consistent with the responses of the 12 monitored ions of the TCDD standard. Quantitation was based on the observed responses at m/z 322 and 334. Quantitation in nanograms per kilogram. b Quantitation by the external standard because of the [13C]TCDD carrier. c No entry in original data presentation. d Samples were not analyzed due to large amounts of coextractives. e Some or all of the sample was lost. �INSTRUMENTAL ANALYSIS The selection of the analytical methodology must take into account a wide concentration range of PCDDs and the possible interferences in different sample matrices. Figure 8 illustrates the detection ranges of analytical techniques that have been used for measurement of PCDDs. This figure presents techniques used in industrial quality control for relatively simple samples at the higher concentration range. Environmental and biological matrices require instrumental methods that have lower limits of detection to achieve parts per billion (nanograms/gram) and parts per trillion (picograms/gram) measurements. Although gas chromatography with electron capture detection (GC/ECD) is capable of low level measurements, the technique lacks the necessary specificity to positively identify PCDDs in a sample extract that contains other halogenated hydrocarbons, pesticides, PCBs, phthalates, etc. Radioimmunoassay (Luster et al., 1980, 1981) and GC/MS-SIM are comparable with respect to achievable limits of detection. However, radioimmunoassay does not yield the identification of individual dioxins and has been used primarily for the screening of a large number of samples for the presence or absence of PCDDs. Two alternate screening techniques for the presence of PCDDs based on biological or biochemical properties are the hydrocarbon hydroxylase induction assay (Bradlaw and Casterline, 1979) and the cytosol receptor assay (Hutzinger et al., 1981). Since the bioanalytical methods do not provide the specificity necessary for identification of PCDDs, these techniques are not discussed in detail below. For a thorough discussion, see National Research Council of Canada (1981). The analytical detection method most frequently reported for the measurement of PCDDs by homolog or by specific isomer in all sample types is gas chromatography combined with mass spectrometry (GC/MS). Gas Chromatography The requires The NRCC used for final separation of PCDDs from interferences in the sample extract gas chromatography with either packed or capillary columns (HRGC). (1981) has compiled a listing of column lengths and liquid phases specific and general PCDD analyses. Packed Column Gas Chromatography-Packed column gas chromatography (PGC) has been used primarily for screening applications to determine the presence of PCDDs and the range of occurring homologs. Figure 9 is an example of packed column gas chromatographic separation of PCDD homologs in an extract from an incinerator fly ash sample (Liberti et al., 1982). The packed column was a 2 m x l . 5 m m I D glass column packed with Supelcoport (100/120 mesh) coated with 1.5% SP-2250 and 1.95% SP-2401. The PCDDs in the sample extract were identified by high resolution mass spectrometry. The packed column chromatogram shown in Figure 9 indicates that tetra- through octachlorodibenzo-£-dioxins were identified in the sample. 25 �GC/MS GC/FID, LC/UV Detection GC/ECD GC/MS-SIM (Selected Ion Monitoring) Radioimmunoassay 10 pg/mL 100 pg/mL 1 ng/mL 10 ng/mL 100 ng/mL 1 /Ag/mL MO'14 Source: MO'13 MO-12 MO' 11 10'10 10'9 15/ig/mL 1.5-10' 8 Karasek, F. W. , and I. Onuska, "Trace Analysis of the Dioxins," Anal. Chem. , 54, 309A-324A (1982). Figure 8. Range of application of some analytical techniques for dioxins. The selection- ion monitoring (SIM) mode of GC/MS is the most applicable. 26 �2 m Packed Column 1.5% SP-2250/1.95% SP-2401 Supelcoport 100/120 Mesh Number of Chlorine Atoms I 20 I 30 I 40 50 Time (Minutes) Source: Liberti, A., P. Ciccioli, E. Brancaleoni, and A. Cecinato, "Determination of Polychlorodibenzo-p-dioxins and Polychlorodibenzofurans in Environmental Samples by Gas Chromatography-Mass Spectrometry," J. Chrom., 242, 111-118 (1982). Figure 9. PGC/MS chromatogram of PCDD homolops extracted from an incinerator fly ash sample. �Packed column gas chromatography columns lack the necessary resolution for isomer specific separation of PCDDs other than the hepta- and octachlorocompounds, as indicated in Figure 9. However, Nestrick et al. (1979) have demonstrated the isomer specific determination of 2,3,7,8-TCDD using a packed column following the fractionation of a mixture of the 22 possible TCDD isomers by RP-HPLC and normal HPLC, as discussed in the section on cleanup procedures. The packed GC column used for the specific analysis was a 210 cm x 2 mm ID glass column packed with a 0.6% OV-17/0.4% Poly S-179 on a specially deactivated Chromosorb W-AW (80/100) support. Lamparski et al. (1979), Langhorst and Shadoff (1980), and Lamparski and Nestrick (1980) have used this procedure for the determination of 2,3,7,8-TCDD at spike levels equivalent to 10 ppt in fish, 1 ppt in human milk, and 10 ppt in particulates (fly ash, industrial dust, urban dust, etc.). Figure 10 presents packed column gas chromatograms of fractions collected from the RP-HPLC and silica HPLC procedures allowing the isomer specific measurement of 2,3,7,8-TCDD by low resolution mass spectrometry (Nestrick and Lamparski, 1980). Quantitation of the peak corresponding to the RP-HPLC fraction for 2,3,7,8-TCDD yielded a value that was approximately four times the concentration found after the extract had been fractionated further with the silica HPLC system. The value obtained before the silica HPLC fractionation was qualified as being the concentration of 2,3,7,8-TCDD plus possibly four unseparated isomers. This demonstrates that PGC can be used for isomer specific PCDD analysis if extended efforts are made to isolate the desired component prior to gas chromatographic separation. High Resolution Capillary Chromatography— The current approach in many analyses for PCDDs by homolog or for specific isomers is the application of high resolution capillary gas chromatography (HRGC) (either glass or fused silica columns). High resolution glass capillary columns were first used by Buser (1975) for the analysis of PCDDs and PCDFs in chlorinated phenols. Since that time numerous studies have reported qualitative identification and quantitation of PCDDs using HRGC columns for separation. Liquid phases for the HRGC columns have ranged from low (SE-30, OV-17, OV-101) to high polarity phases (Silar IOC, SP-2330, SP-2340), and column lengths have ranged from 18 m for general analysis of PCDD to 60 m for isomer specific measurements. Figure 11 is a chromatogram depicting the elution of tetra- to octa- PCDDs on a HRGC column. Isomer specific measurements have been of prime importance in most studies (both environmental and biological), particularly for 2,3,7,8-TCDD. Figures 12 and 13 present chromatograms of the mixture of the 22 possible TCDD isomers yielding the isomer specific separation for 2,3,7,8-TCDD. Buser (1980) used the three liquid phases (Figure 12) Silar IOC, OV-17, and OV-101 to determine specific assignments for the 22 isomers. Figure 13 presents the separation of TCDDs on a glass column coated with SP-2330 and a fused silica column coated with SP-2340 that is currently recommended for 2,3,7,8-TCDD specific analyses (EPA, 1982, 1983). In addition to these columns, Harless (1980) has reported isomer specific determination with a 30-m SE-30 column, and the current EPA method for the determination of 2,3,7,8-TCDD in soils and sediments implies that 30-m Durabond DB-5 fused silica columns provide sufficient separation for specific 2,3,7,8-TCDD measurements. 28 �(a) 13 (b) C-2378-TCDD 13 C-2378-TCDD n/e 332 m/e 332 native 2378-TCDD native TCDDs >^ I/I m/e 324 i/e 324 ^ c o o m/e 322 m./e 320 i (Minutes) 3 Source: m/e 322 n/e 320 • 6 (Minutes) 3 4 • 6 Lamparski, L. L. , and T. J. Nestrick, "Determination of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-p_-dioxin Isoroers in Particulate Samples at Parts per Trillion Levels," Anal. Chem., _52, 2045-2054 (1980). Figure 10. Comparative 2,3,7,8-TCDD PGC/MS mass chromatograms for electrostatic fly ash (a) after RP-HPLC and (b) subsequent silicaHPLC. 29 �Peok No. PCDD Congener 1 2 3 4 5 6 7 .3.6,8-lerro-CDD .3.7,9.3.7.8.3,6.7.3.7.8.3,8.9,2,7,8- 15 16 17 18 19 20 21 8 9 10 11 12 13 M .2.4.6.8- (or 1. 2 4 7.9-)penlo-CDD .2.3,6.8.2.4.7.8.2.3.7.9.2,3.7.8.2.3,6.7.2,3,8.9- 22 23 24 .2.4.6.7.9- (or 1.2. 4. 6.7.9- )h«xa-CDD .2.3.4.6,8,2.3.6.8.9- (or 1.2. 3. 6. 7. 9-) .2,3,4,7.8,2.3,6.7.8.2.3,7.8.9.2.3.4.6.7,2.3.4,6,7-fiepla-CDD ,2,3,4,6,7.8oclo-CDD 8 15 + 16 9 17 22 ,23 10 " 12 L. UllUUIAOl) penta- tetra- 1 190° C Source: \J J^,^1I_J[ 1 200 1 210 1 220 230 240 Lustenhower, T. W. A., K. Olie, and 0. Hutzinger, "Chlorinated Dibenzo-p-dioxins and Related Compounds in Incinerated Effluents," Chemosphere, ^ 501-522 ( 9 0 . , 18) Figure 11. Separation of PCDD-isomers by GC/MS using a high resolution capillary column. �55M SIIAR 10c 1237/1238 Source: 50M O V - I O I I 1246/1249 Buser, H. R., and C. Rappe, "High Resolution Chromatography of the 22 Tetrachlorodibenzo-_p-dioxin Isomers," Anal. Chem.. 52, 2257-2262 (1980). Figure 12. Mass chromatograms (m/e 320) of a composite pyrolyzate sample showing elution of all 22 TCDD isomers on HRGC columns. �SP-2330 Glass Column 11 12 14 Minutes 13 15 17 SP-2340 Fused Silica Column 11 12 13 14 15 16 17 18 Minutes Source: "Rapid Separation of 2,3,7,8-TCDD from Other TCDD Isomers," The Supelco Reporter, H4) , I (1982). Figure 13. HRGC chromatogram of a mixture of the 22 TCDD isomers on glass and fused silica capillary columns (60 m) coated with SP-2330 and SP-2340, respectively, indicating isomer specific separation for 2,3,7,8-TCDD. 32 �The advantages of using HRGC columns over PGC columns include increased isomer specificity, resolution of interferences from analytes of interest, and increased sensitivity due to less band spreading. Fused silica HRGC columns allow the direct routing of the column into the ion source of the mass spectrometer, a procedure which leads to fewer problems resulting from dead volumes and to greater sensitivity. The major disadvantage of HRGC columns is the ease of overloading by coextractives. This problem has been overcome in most cases, however, by using effective and efficient cleanup procedures prior to HRGC separation of the sample extract. Gaps in PGC and HRGC Information-A major deficiency in the area of PGC and HRGC separation is the lack of information regarding the retention times of common interferences with respect to the PCDDs. This information would indicate whether polychlorinated biphenyls (PCBs), the common pesticides (e.g., DDE and DDT), polychloromethoxybiphenyls, or polychlorobenzylphenyl ethers actually elute within the retention windows required for the measurement of the PCDDs. This problem has been partially addressed by Hummel (1977), who considered possible interferences from pesticides and PCBs for the analysis of TCDD. Table 10 provides some of the information for relative retention times and responses for the ions characteristic of 2,3,7,8-TCDD. Mass Spectrometry The application of mass spectrometry for the analysis of PCDDs in biological matrices, commercial products, and environmental samples has been reviewed by Hass and Friesen (1979), Cairns et al. (1980), the National Research Council of Canada (1981), Mahle and Shadoff (1982), and Tiernan (1983). Mass spectrometry measurements have been reported for quadrupole (low resolution) and magnetic sector (high resolution) instruments. Electron impact is the most common method of ionization but chemical ionization mass spectrometry techniques have also been reported as a means of confirmation of the identity of PCDDs. As indicated in Figure 8, MS-SIM techniques are required to obtain the necessary sensitivity for measurement of PCDDs at the parts per trillion concentration range required for biological matrices. The sensitivity of the SIM method is enhanced as a result of making multiple measurements of a few selected ions characteristic of the PCDDs rather than scanning an entire molecular range in the same time frame. Most of the analytical studies reported in the literature have focused on the measurement of TCDDs. Langhorst and Shadoff (1980) have reported analytical methods for the analysis of tetra-, hexa-, hepta-, and octachlorodibenzo-£-dioxins in human milk based on the RP-HPLC fractionation scheme combined with PGC/MS. The alternative to this approach is computer-sequenced analysis of each PCDD homolog in a single analysis. Tiernan (1983) and Liberti et al. (1982) have emphasized the application of this procedure to provide data at the parts per trillion level for a wide range of PCDD homologs, 33 �TABLE 10. RESPONSE FROM POSSIBLE ENVIRONMENTAL CONTAMINANTS Compound Chlordane £,£'-DDE £,£'-DDD £,£'-DDT Dieldrin Endrin Endosulfan Mi rex PCBs Aroclor 1242 Aroclor 1254 Aroclor 1260 Q Toxaphene Source: a Retention time difference from 2,3,7,8-TCDD (sec) 2,3,7,8-TCDD . equivalent peak height m/e 320 m/e 322 -35 +15 +85 +187 -85 -38 +9 +57 NDd 1 0.07 0.01 0.005 0.060 0.0002 0.00028 ND 0.2 0.03 0.003 0.003 0.024 0.0009 0.00022 ND ND 0.001 0.001 ND 0.0005 0.000002 0.00001 0.00005 0.000005 -251, -194, -184 -158 -83 -16 -155 -120 -96 +257 ND ND 0.020 0.015 0.032 0.008 0.000002 0.00001 0.000005 ND R. A. Hummel, "Cleanup Techniques for the Determination of Parts per Trillion Residue Levels of 2,3,7,8-TCDD," J. Agric. Food Chem. , 25:1049-1053 (1977). 2,3,7,8-TCDD retention time 390 sec. Peak width at half-height = 30 sec. b The ratio of response of the compound at its retention time to the response of an equal weight of 2,3,7,8-TCDD measured at 390 sec. c Only those peaks near 2,3,7,8-TCDD are listed. d ND = not detected; no peaks were detected at m/e 320 or 322. 34 �Low Resolution versus High Resolution Mass Spectrometry— One of the major points of contention in the analyses of low level (ppt) PCDDs is the necessity of low resolution (M/AM = unit) versus high resolution (M/AM = 10,000) mass Spectrometry measurements. Many of the methods rely on efficient cleanup steps prior to low resolution mass Spectrometry to provide low level backgrounds. Other methods, however, utilize the mass resolving power of single or double focusing mass spectrometers to identify and quantitate low level PCDDs in the presence of other chlorinated compounds (Harless et al., 1980). The need for high resolution mass Spectrometry for various extraction and cleanup procedures has been demonstrated by Brumley et al. (1981) via the interference noted for electron capture detector and low resolution mass Spectrometry measurements for extracts prepared by six different laboratories. Hummel and Shadoff (1980) have directed attention to the need for high resolution confirmation of TCDDs in sample extracts, especially when the concentration approaches values of 20 ppt or less as measured by low resolution mass Spectrometry. Table 11 provides data presented by Hummel and Shadoff (1980) for the levels of TCDD in beef fat samples analyzed for Phase I of the EPA Dioxin Implementation Plan. The data presented in this table indicate that of 93 total samples analyzed by low resolution mass Spectrometry 37 were determined to contain TCDD. Further analyses of these positives by high resolution mass Spectrometry yielded that only 20 of the 37 samples contained TCDD. The two control samples identified as positive by high resolution mass Spectrometry present the additional problem of false positives for measurements near the detection limit. Additional studies of the extracts after a second cleanup also presented the possibility of false negatives by high resolution mass Spectrometry when sample extracts are dirty. Shadoff and Hummel (1980) concluded that analysis by low resolution mass Spectrometry is acceptable if suitable control samples demonstrate the absence of interferences. Otherwise, high resolution mass Spectrometry should be used for confirming positive results . Interferences— Some of the compounds identified as interferences in the analysis of TCDDs by mass Spectrometry are presented in Table 12. The alternate methods of resolution are the approaches that have been specifically addressed in the literature. The separation of PCBs, polychlorodiphenyl ethers and polychlorobenzyl phenyl ethers has been reported by Mieure et al. (1977) and Lamparski et al. (1979). 35 �Sample TABLE 11. EPA PHASE I DIOXIN IMPLEMENTATION PLAN BEEF FAT SAMPLES ANALYZED FOR TCDD No. of apparent No. of samples positive results by analyzed by PGC/LRMS PGC/HRMS3 PGC/LRMS Grazed on treated land 64 19 9 Control 20 9 2 9 9 9 Fortified extracts and solutions Source: a R. A. Hummel and L. A. Shadoff, "Specificity of Low Resolution Gas Chromatography-Low Resolution Mass Spectrometry for the Detection of Tetrachlorodibenzo-p_-dioxin in Environmental Samples," Anal. Chem., 52:191-192 (1980). In this part of the study, only those extracts showing an apparent positive result or a limit of detection greater than 20 ppt were analyzed by PGC/HRMS. b The fortification level was 20 to 100 ppt TCDD in the beef fat. 36 �TABLE 12. SOME COMPOUNDS THAT MAY INTERFERE WITH THE DETERMINATION OF TCDD AT m/z VALUES OF 319.8966 AND 321.8936 Elemental composition Ion Mass lost Heptachlorobiphenyl C12H3 35C17 M+ -235C1 321.8678 0.0258 Nonachlorobiphenyl C12H 35C19 C12H 35C18 M* M -435C1 -335C1 37 C1 319.8521 321.8491 Tetrachloromethoxy biphenyl C13H8 35C140 M* C13H8 35C13 37C10 M Tetrachlorobenzylphenyl ether C13H8 35CL40 M* C13H8 35C13 37C10 M+ Pentachlorobenzylphenyl ether C13H7 35C14 C13H7 35C13 37 C120 DDT C14H9 35C13 37C12 C14H9 35C12 37C13 DDE C14H8 C14H8 Compound 35 37 C1 37 C10 M* M M* M C12 37C12 M* C1 37C13 M 35 m/z AM TCDD Mass resolution for separation M/AM Alternate means of resolution 12476 Alumina micro column, HPLC, HRGC 0 . 0445 0 . 0445 7189 7233 Alumina micro column, HPLC, HRGC 319.9329 321.9299 0.0363 0.0364 8805 8848 AgN03 ( 0 ) 1% impregnated silica, alumina, HPLC 319.9329 321.9300 0.0363 0.0364 8813 8843 Alumina micro column, HPLC, HRGC -H35C1 -H35C1 319.9143 321.91138 0.01773 0.01778 18043 18104 Alumina micro column, HPLC, HRGC -H35C1 -H35C1 319.9321 321.92917 0.03552 0.03557 9006 9050 Alcoholic saponification converts DDT to DDE 319.9321 321.92916 0.03550 0.03556 9011 9052 AgN03 (10%) impregnated silica (continued) �TABLE 12 (continued) Mass resolution for m/z AM TCDD separation M/AM M 319.8966 321.8936 0.00 0.00 a a NRb M 319.8966 321.8936 0.00 0.00 a NR 319.9143 321.9114 0.01773 0.01778 Elemental composition Ion Hydroxytetrachlorodibenzofuran C 12^^1402 Tetrachlorophenylbenzoquinone Ci2H4Cl402 Tetrachloroxanthene C13H60 35C13 C13H60 35C12 Compound 37 C1 M* M Mass lost Alternate means of resolution a 18043 18104 NR oo Source: Adapted from National Research Council of Canada, "Polychlorinated Dibenzo-g-dioxins: Limitations to the Current Analytical Techniques," NRCC Report No. 18576, ISSN 0316-0114, 1981. a Cannot be resolved by MS. b NR = not reported specifically in the literature. �Mieure et al. (1977) specifically reported that PCBs and polychlorobenzylphenyl ethers are separated from PCDDs on a microalumina column (basic, super grade 1). Figure 14 is an example of separation of chlorinated interferences using the microalumina column. The first chromatogram is representative of the total nonphenolic fraction from technical grade pentachlorophenol obtained after removing the phenolic components with a macroalumina column (Fisher A-540, 5% deactivated). The second chromatogram is the first fraction from the microalumina column and contains the polychlorodiphenyl ethers (hexa- to decachloro), chlorobenzenes, and PCBs. The third chromatogram represents the second fraction collected from the microalumina column that contains PCDDs and PCDFs (hexato octachloro). It is concluded from these chromatograms that the octachlorodibenzo-£-dioxin and octachlorodibenzofuran can be measured without possible interferences from the other chlorinated compounds. However, the micro column cleanup is necessary to isolate the lower chlorinated homologs of the PCDDs and PCDFs for specific analysis. Chlorinated methoxybiphenyls were reported as interferences in the analysis of PCDDs in fish extracts by Phillipson and Puma (1980). These compounds eluted in the retention window for tri- to pentachlorodioxins by PGC/ low resolution mass spectrometry and produced intense molecular ions having the same nominal masses and chlorine isotopic abundances as those observed for the PCDDs. These authors suggested the need for monitoring fragment ions in addition to ions representative of the molecular ion cluster for low resolution mass spectrometry. Alternately, high resolution mass spectrometry may be used as presented in Table 12 to differentiate the PCDDs from interfering chlorinated methoxybiphenyls. This interference might also be removed by the cleanup procedures as described by Nestrick et al. (1980). Smith and Johnson (in press) have presented detailed information on the potential of interferences to arise from selected congeners of seven families of polychlorinated aromatic compounds with the analytical method for part-pertrillion determinations of PCDFs and PCDDs used by the Fish and Wildlife Service (Stalling et al. 1982). The polychlorinated aromatic compounds evaluated as potential interferences included polychlorinated-biphenyls (PCBs), -naphthalenes (PCNs), -diphenyl ethers (DPEs), methoxy-PCBs (MeO-PCB), -hydroxy-PCBs (HO-PCB), -methoxy-diphenyl ethers (MeO-DPE), -hydroxy-diphenyl ethers (HO-DPE), -benzylphenyl ethers (BzPE) and -biphenylenes. The potential interferences from these compounds arise from the large number of congeners of each chemical family exhibiting chromatographic retention times that are similar to PCDDs and PCDFs and from mass spectral patterns that overlap to varying degrees with PCDDs and PCDFs. In addition, some of these potential interferences have the same nimonal masses and the same number of chlorine substituents as those of PCDDs and PCDFs, making the molecular ions indistinguishable by low resolution mass spectrometry. Also, at least five of the chemical families include compounds that under thermal conversions and/or conversion following ionization produce PCDDs and PCDFs. Table 13 presents the potentially interfereing chemical according to the degree of potential interference that can be encountered. Smith and Johnson (in press) concluded that the specific polychlorinated aromatic compounds used in this study did not produce a significant number of false positives with the particular analytical procedure. However, these authors do note that only a few of the large number of potential compounds were available for this study. 39 �a ) Total Non-Phenolic Fraction b ) Diphenyl Ether Fraction a a a. o U Time (Minutes) Source: Mieure, J. P., 0. Hicks, R. G. Kaley, and P. R. Michael, "Determination of Trace Amounts of Chlorodibenzo-p-dioxins and Chlorodibenzofurans in Technical Grade Pentachlorophenol," J. Chrom. Sci. , JL5, 275-277 (1977). Figure 14. Electron capture chromatograms of (a) entire nonphenolic fraction, (b) first microcolumn fraction containing chlorodiphenyl ethers, (c) second basic alumina microcolumn fraction containing chlorodibenzo-p_-dioxins and chlorodibenzofurans. 40 �TABLE 13. INTERFERENCES OF SELECTED CHEMICAL FAMILIES IN MS DETERMINATIONS OF PCDFs AND PCDDs Level of interference Family of polychlorinated compounds PCBs Overlap of fragmentation patterns PCDDs PCDFs Indistinguishable by LRMS PCDDs PCDFs Indistinguishable by HRMS PCDFs PCDDs ++/+++ PCNs + DPEs ++/+++ MeO-PCBs +++ HO-PCBs +++ HO-DPEs BzPEs ++ X X X X X +++ MeO-DPEs Biphenylenes +++ X X X X +++ X X +++ X ++ Source: L. M. Smith and J. L. Johnson, "Evaluation of Interferences from Seven Series of Polychlorinated Aromatic Compounds in an Analytical Method for Polychlorinated Dibenzofurans and Dioxins in Environmental Samples," in Chlorinated Dioxins and Dibenzofurans in the Total Environment, L. H. Keith, G.Choudry, and C. Rappe (Eds.), Pergamon Press, in press. NOTE: In the first two columns "+" indicates minor overlap, "++" indicates major overlpa, and "++•*-" indicates complete overlap. In the last four columns, an "X" indicates that a particular type of interference is observed. The abbreviations used for polychlorinated compounds are: PCBs-biphenyls; DPE-diphenyl ethers; PCNs-naththalenes/ BzPEs-benzylphenyl ethers. The prefixes Meo- designate methoxy and HO- hydroxy. �In summary, a number of chlorinated compounds have been noted to interfere with the mass spectrometry analysis of PCDDs, particularly 2,3,7,8-TCDD. The problems arising from these interferences have been overcome in part by using efficient cleanup procedures or high resolution mass spectrometry, or a combination of the two. Criteria for Positive Identification of PCDDs-Positive identification of PCDDs as a particular homolog or specific isomer requires the analyst to ensure that the instrumental response meets specific criteria. Most analysts to date have used the coincident response of a minimum <j>f three different ions from the molecular^ ion cluster (M , [M-2] , [M+2] ) and from fragment ions (e.g., [M-COC1] ). The retention time of the selected ions must fall within a designated or established retention window. In addition the selected ions must have the correct response ratios. Figure 15 is an example of a HRGC/MS-SIM analysis for 2,3,7,8-TCDD demonstrating these criteria. The ion at m/z 257 is representative of the fragment ion [M-COC1] , while m/z 320 and 322 are indicative of the molecular ion cluster for TCDD. Isomer specific measurement of the 2,3,7,8-TCDD was accomplished with the SP-2330 glass column discussed earlier. Documentation of the 2,3,7,8TCDD retention time is represented by m/z 332 for the carbon-13 labeled 2,3,7,8TCDD. Harless et al. (1980) have specifically designated the following criteria as essential to the final analysis for TCDD by HRGC/HRMS. 1. Correct HRGC-HRMS retention time for 2,3,7,8-TCDD. 2. Correct HRGC-HRMS multiple ion response for 37C1-TCDD and TCDD masses (simultaneous response for elemental composition of m/z 320, m/z 322, m/z 328). 3. Correct chlorine isotope ratio for the molecular ions (m/z 320 and m/z 322). 4. Correct responses for the co-injection of sample fortified with 37 C1-TCDD and TCDD standard. 5. Response of the m/z 320 and m/z 322 must be greater than 2.5 times the noise level. Supplemental criteria that Harless et al. (1980) suggested may be applied to highly contaminated sample extracts are: (A) COC1 loss indicative of TCDD structure, and (B) HRGC-HRMS peak matching analysis of m/z 320 and m/z 322 in real time to confirm the TCDD elemental composition. Other supplemental information for confirmation of TCDD in particular can be obtained from the partial scan of the TCDD peak (Table 14) when present at parts per billion concentrations (EPA, 1983). Table 15 provides a range of reported relative abundances for the most intense ion in the isotopes clusters. 42 �SCANS 758 TO 875 188288 257.87 8.501 248576 328.891 ± 8.58) 4 831 836 842 849 297472 322.891 ± 8.58) 114816 332.09: 8.50i 768 17i44 Source: 868 20:84 SCAN TIME MRI RC-693-A, "Analytical Chemistry Application of Isotopically Labeled Compounds, 1982. Figure 15. HRGC/MS-SIM chromatogram of TCDD analysis. Mass charge (m/z) ratios 257, 320, and 322 are representative of natural abundance TCDD isomers, while m/z 322 represents the level of -^Clabeled 2,3,7,8-TCDD internal standard. This chromatogram was obtained on a 60-m SP-2330 glass capillary column. Fifty picograms of the 13C-TCDD were injected. �TABLE 14. PARTIAL SCAN CONFIRMATION FOR TCDDa ra/z Ratios Response ratios 320/324 1.58 ± 0.16 257/259 1.03 ± 0.10 194/196 1.54 ± 0.15 Source: "Determination of 2,3,7,8-TCDD in Soil and Sediment," U.S. Environmental Protection Agency, Region VII, Kansas City, Kansas, February 1983. a All ions including 160 and 161 must be presented with at least 5% relative abundance to the ion at 322. 44 �TABLE 15. RANGE OF REPORTED PERCENT RELATIVE ABUNDANCES FOR MOST INTENSE ION IN ISOTOPE CLUSTERS FROM ELECTRON IMPACT MASS SPECTRA OF THE CHLORINATED DIBENZO-B-DIOXINS Number of chlorines 2 3 4 5 6 7 8 100 100 100 100 100 100 100 100 2 Ion 5-7 5 0-10 20 7-10 5-15 3-10 21-60 40 31-34 28-35 11-35 1 [M]+ [M-C1]+ [M-COC1]+ 14 20-24 24-36 [M-2C1]+ 0-3 1 0-5 15 4-6 0-4 2-5 [M-C202C1]+ -p- 0 9 0-3 2 0 5 0 0-3 1-5 [M-C202C12]+ 0 13-18 3-25 10-25 10-23 13-55 35 17-21 Ln Source: Mahle, N. H. , and L. A. Shadoff , "The Mass Spectrometry of Chlorinated Dibenzo-£-dioxins , " Biomedical Mass Spectrometry, 9 :45-60 (1982). �Quantitation Several variables have been reported for quantitation of PCDDs by mass spectrometry methods. These variables include electron impact versus chemical ionization mass spectrometry, selection and availability of standard compounds, and internal versus external standard calibration. Electron Impact Versus Chemical Ionization Mass Spectrometry-Although chemical ionization mass spectrometry, especially the negative chemical ionization mode, has the potential to enhance specificity and sensitivity for individual isomers (Hass et al., 1978; Mitchum et al., 1981; Rappe et al.), electron impact ionization has been used most often for quantitative analysis of PCDDs. The inconsistencies of response factors across a homolog of PCDDs noted with negative chemical ionization (NCI) mass spectrometry (Kuehl and Dougherty, 1980) and the scarcity of all the specific standard PCDDs are disadvantages to its use for routine analysis of PCDDs. Kuehl and Dougherty (1980) have reported that the relative sensitivity for 2,3,7,8-TCDD is roughly a factor of 50 less than that for other TCDD isomers or higher chlorinated dioxins. However, specific analysis for 2,3,7,8-TCDD has been reported for NCI methods (Hass et al., 1978). Hass et al. (1981) have also suggested that both negative chemical ionization and electron impact ionization are necessary to provide reliable measurements for PCDDs and PCDFs in the presence of PCBs and polychlorinated diphenyl ethers. Selection of Calibration Standards— There is concern regarding the need for standard compounds representing each homolog to provide appropriate assessment of the possible effects arising from trace levels of PCDDs in biological samples. Nestrick et al. (1982) addressed this problem as a systematic error that affects accuracy and reliability in the analysis of environmental samples for PCDDs. The source of error originated by assuming that the response factors for penta- through octachloro PCDDs were consistent with the response factor for TCDD. Table 16 provides response factors of several PCDDs relative to 1,2,3,4TCDD at the molecular ion (m/z) 322. These data illustrate the possible margin of quantitative error that could be introduced by the assumption of a constant response factor for all PCDD homologs. The data from Table 16 indicate differences of approximately 3 to 1 when comparing the response of 1,2,3,4-TCDD to the response for octachlorodibenzo-j>-dioxin (OCDD). Nestrick et al. (1982) also demonstrated the differences in response factors that arise when working with PGC/MS systems that rely on silicone membrane separators and jet separators for introduction of the PCDDs to the ion source of the mass spectrometer. Table 17 summarizes the data and indicates a significant difference for the response factors of hepta- and octa-PCDDs measured with a quadrupole mass spectrometer equipped with silicone membrane or jet separators. 46 �TABLE 16. AREA RESPONSE FACTORS OF PCDDs RELATIVE TO 1,2,3,4-TCDD AT m/z 322a Component m/z 1,2,3,4-TCDD 322 322 356 390 426 460 2,3,7,8-TCDD 1,2,3,7,8-PCDD HCDD mixture 1,2,3,4,6,7,8-HpCDD OCDD Source: a 1.00 0.89 0.52 0.44 0.46 0.32 No. of replicates 5 5 2 4 3 3 ± 0.03 ± 0.03 ± 0.02 ± 0.02 ± 0.01 ± 0.01 Nestrick, T. J., L. L. Lamparski, W. B. Crummett, and L. A. Shadoff, "Comments on Variations in Concentrations of Organic Compounds Including Polychlorinatd Dibenzo-j)-dioxins and Polynuclear Aromatic Hydrocarbons in Fly Ash from a Municipal Incinerator," Anal. Chem., 54:824-825 (1982). One hundred picograms of each component injected. TABLE 17. COMPARISON OF RELATIVE PEAK RATIOS OF PCDDs THROUGH A GLASS JET AND SILICONE MEMBRANE SEPARATOR3 Rel response (± rel std dev) Component 1,2,3,6,7,8-HCDD membrane jet 1,2,3,4,6,7,8-HpCDD membrane jet OCDD membrane jet Source: a Rel response (± rel std dev) No. of replicates 1.00 1.00 7 4 0.34 ± 0.04 0.63 ± 0.10 7 4 0.21 ± 0.05 0.38 ± 0.04 7 4 Nestrick, T. J., L. L. Lamparski, W. B. Crummett, and L. A. Shadoff, "Comments on Variations in Concentrations of Organic Compounds Including Polychlorinatd Dibenzo-£-dioxins and Polynuclear Aromatic Hydrocarbons in Fly Ash From a Municipal Incinerator," Anal. Chem., 54:824-825 (1982). All values normalized to HCDD response. 47 �Data summarizing the response factors for PCDDs by homolog or by isomer by electron impact ionization versus chemical ionization mass spectrometry do not appear in the primary literature. There is a need to determine the variability of the response factors for isomers within a homolog in order to evaluate the maximum systematic error that might be encountered in using a response factor for a single isomer within a homolog. Rappe et al. (in press) have recently reported some response factor data for PCDFs using electron impact and negative chemical ionization mass spectrometry. The data presented indicated the relative response factors for 13 TCDFs varied considerably less with electron impact than with negative chemical ionization. The range of response factors however, was not markedly different for higher chlorinated PCDFs when comparing the two ionization techniques, although the negative chemical ionization absolute response is considerably greater. Internal Versus External Standard Quantitation-Quantitation for PCDDs requires calibration of the instrument with standards bracketing the expected concentration range of any sample extracts. The internal standard quantitation method has been used by most analysts for measurement of the levels of PCDDs in biological and environmental matrices. This method requires response factors be determined for the internal standard versus an authentic analyte. Typically, the stable isotope labeled compounds, such as carbon-13 or chlorine-37 analogs of native PCDDs, are incorporated in calibration solutions and samples as internal standards. The level of the labeled internal standard is usually held constant and the native PCDD is varied for calibration purposes. If response factors are determined to remain constant over the expected concentration range, the true internal standard quantiation method is applicable for calculation of the PCDD concentration. On the other hand, if the response factor is not consistent across the calibration range, it becomes necessary to use external standards and calibration curves routinely for measurements of PCDD contamination. The EPA methods for analysis of 2,3,7,8-TCDD in water and wastewater (EPA, 1982) and soil and sediment (EPA, 1983) require measurement of the response factors over a designated concentration range at the initiation of any sample analysis event and the daily check of the response factor value. If the response factor does not agree within ± 10% of the value generated for the concentration range, a recalibration is necessary. True internal standard quantitation provides a correction of the reported value without a true measure of the recovery for each analysis. Recovery can be estimated by comparing the response of the labeled compound in a sample extract versus an external standard. More accurate measurements of method recovery are achieved by using a second internal standard added to the sample extract immediately before GC/MS analyses. For instance, carbon-13 (13Ci2) labeled 2,3,7,8-TCDD can be added as a surrogate prior to sample preparation to provide true internal standard quantitation for native TCDD and chlorine-37 (37C14) labeled 2,3,7,8-TCDD can be added to the sample extract prior to GC/MS analyses to provide accurate recovery measurement of the carbon-13 TCDD. 48 �The true internal standard quantitation is accomplished using the fol lowing equation: C = (As) (I )/(AT_)(RF)(W) x s 10 where CA = concentration of the PCDD in the original sample As = peak area response for the PCDD quantitation ion AT<, = peak area response for the internal standard quantitation ion !„ = amount of internal standard added to the sample W = weight or volume of the sample KF = response factor The response factor (RF) is calculated according to the equation where A , = peak area response for the standard PCDD quantitation ion A TC Xo = peak area response for the internal standard quantitation ion C ., = concentration of the standard PCDD std C,Q - concentration of the internal standard Stable isotope labeled compounds are commercially available for internal standard quantitation of tetra-, hepta-, and octachloro-PCDDs, KOR Isotopes, Division of ICN Pharmaceuticals, and Laraparski and Nestrick (1982) have presented details for the laboratory preparation of carbon-13 (13Cj2) labeled penta- through octachloro-PCDDs from the commercially available carbon-13 (13cl2) 2,3,7,8-TCDD. Bell (in press) has recently reported on the synthesis of carbon-13 labeled PCDFs. Regardless of the quantitation technique, the quantitation ion monitored for each PCDD isomer or homolog is selected from the molecular ion cluster. Table 18 provides the exact masses, relative isotope abundances, and the chlorine pattern for the major molecular cluster ions for the mono- through octachlorinated dibenzo-g-dioxins . Limit of Detection-The limit of detection is the lowest concentration of an analyte the analytical method can reliably detect. The limit of detection in PCDD studies is the concentration of the analyte that gives rise to a signal that is at least 2.5 times the background noise for the sample 49 that most response matrix. �TABLE 18. EXACT MASSES AND RELATIVE ISOTOPE ABUNDANCES OF MAJOR MOLECULAR CLUSTER IONS FOR PCDDs Relative abundance No. of chlorines Exact mass 0 184.0524 - 1 218.0135 220.0105 100.00 33.82 2 251.9746 253.9716 255.9686 100.00 66.45 11.43 3 285.9356 287.9326 289.9296 291.9266 100.0 99.07 33.10 3.86 4 319.8967 321.8937 323.8907 325.8877 327.8847 75.93 100.00 49.68 11.13 0.99 5 353.8578 355.8546 357.8518 359.8488 361.8458 60.86 100.00 65.96 21.91 3.70 6 387.8188 389.8158 391.8128 393.8909 395.8068 50.78 100.00 82.25 36.23 9.05 7 421.7799 423.7769 427.7709 429.7679 43.56 100.00 98.55 54.10 17.90 455.7410 457.7380 459.7350 461.7320 463.7290 33.21 87.08 100.00 65.76 27.10 425.7739 8 Source: Radolovich, G., Midwest Research Institute (personal communication) (1983). 50 �The limit of detection has been found to vary with each sample (Crumraett, 1979). The differences in reported limits of detection are dependent on initial sample size, final extract volume, volume of final extract analyzed, residual interferences from the sample matrix, extraction and cleanup procedures, chromatography and instrumental performances, purity of reagents used for preparation of samples, and absolute sensitivity obtainable with any particular mass spectrometer. Figure 16 presents the direct relationships of method limit of detection with respect to initial sample size and final extract volumes. The data generated for Figure 16 were calculated assuming a conservative GC/MS instrumental detection limit of 5 pg per 1.0 pi on-column injection. Based on these data, the instrumental detection limit required for measurement of 1 ppt levels of a PCDD in a 1-g sample concentrated to 10 pi would be 0.1 pg/pl assuming 100% recovery. The only study approaching this level of effort has been presented in part by Harless (1980). Table 19 provides the data presented for the feasibility study regarding the analysis of TCDD at the parts per trillion level in 250-mg samples of human adipose tissue (equivalent to a needle biopsy). These data suggest than an extremely clean and sensitive mass spectrometer was used to measure these levels of TCDD. Limits of detection have been presented in many of the studies dealing with the analysis of PCDDs in biological matrices. Table 20 is a summary of data presented in a review of TCDD analysis by Shadoff and Hummel (1978). The data presented in Table 20 generated by gas chromatography low resolution mass spectrometry show that the original biological sample matrix has little effect on the average detection limit that is obtainable. The lowest parts per trillion limits of detection were obtained for samples sizes of 10 to 20 g. Final extract volumes were taken to 10 to 20 pi, providing concentration (or enrichment) factors of 1,000 for the larger sample sizes. In comparison, the evidently higher LOD reported for a 1-g blood sample is in part due to the difference in achievable enrichment (100) of TCDD in the final extract. Thus, actual limits of detection vary with the sample extract and instrumental condition. If significant interferences prevent measurement of the desired LOD value with low resolution mass spectrometry, the alternative approach is to use high resolution mass spectrometry. 51 �100 g Method Detection Limit Versus Final Extract Volume 10, 40 50 60 Detection Limit (pg/g) Figure 16. Method detection limit versus final extract volume and initial sample size assuming a GC/MS instrumental detection limit of 5 pg/yl on-column. �TABLE 19. FEASIBILITY STUDY FOR THE QUANTITATIVE DETERMINATION OF TCDD IN QA TISSUE SAMPLES TCDD fortification level6 (PPt) (P8) i TCDD detection limit (ppt) TCDD detected (PPt)3 3 8 1 4 5 16 0 0 1 2 0.5 2C 1 3 2 8 1 6 16 6C Source: a Harless, R. L., "Analytical Methodology for 2,3,7,8-Tetrachlorodibenzo-£-dioxin and Its Application by the United States Environmental Protection Agency to Human and Environmental Monitoring," presented at the Assistant Administrators Program Review, U.S. EPA, Washington, D.C., April 1980. 37 C1-TCDD mean percent recovery - 75%. percent recovery losses. Values are not corrected for b Each 0.250 g sample was fortified with 0.5 ng 37C1-TCDD. c Standard solutions. 53 . �TABLE 20. DETECTION LIMITS FOR TCDD IN VARIOUS SAMPLES No. of determinations Arkansas and Texas Catfish Viscera Bass, walleyed pike Sunfish, etc. Flesh Viscera Liver Skin Eel Flesh Viscera Skin Shark liver Sea cucumber Flesh Viscera Crayfish Whole Muscle Viscera Tadpole Toad Rabbit Liver Pelt Beaver liver Opossum Liver Fat Deer Liver Fat Insects Insect larvae Diving beetles Snails Mice Liver Skin Whole Rat liver Shrimp Limit of detection (ppt) Range Average 57 2 52 11 6 2 5 8 10 7 2-22 2-3 2-8 5-15 2-10 1-14 3 4 10 7 1 1 1 1 7 5 4 11 1 1 1 1 1 1 1 1 1 4 4 7 20 3 1 1 11 8 2 9 3-17 1 1 3 1 1 1 1 1 1 3 10 1 4 10 10 4 4 3 8 30 2 4-5 8 20 5 20 1 10-40 3-8 1 (continued) 54 �TABLE 20 (continued) No. of Limit of detection (ppt) determinations Milk (40 g) Cream Human milk Rice Rat feed Sheep feed Cattle feed Grass Seed (grass) Sorghum Leaves Roots Soil Water Blood (1 g) Source: 3-10 2-7 6 4 7 15 12 5 Liver Kidney Muscle Bovine Average 60 7 Beef Fat Liver Sheep Fat Range 5-15 3-10 3-6 2-6 9 7 5 5 28 4 6 21 5 1 1 2 5 2 1 2 100 4 2 0.5-1 3-5 1-6 2-7 4-6 1 4 3 4 5 3 6 13 7 3 4 5 6 0.2 40 12-14 2-12 2-3 4-6 3-10 0.1-0.2 40 Shadoff, L. A., and R. A. Hummel, "The Determination of 2,3,7,8Tetrachlorodibenzo-£-dioxin in Biological Extracts by Gas Chromatography Mass Spectroraetry," Biomed. Mass Spectrom., 5:7-13 (1978) 55 �Quality Assurance All studies concerning the analysis of biological samples for PCDDs have included some form of a quality assurance (QA) program. The routine use of stable isotope labeled PCDDs as surrogates for internal standard quantitation and method recovery is practiced most frequently as a QA procedure. Studies undertaken by the Dioxin Monitoring Program (Harless et al., 1980) included the use of stable labeled surrogates, submission of blind samples, duplicates, and blanks to the analyst, establishing of criteria for the positive identification of PCDDs, and interlaboratory studies to bolster the significance and validity of TCDD data generated. The methods adopted by EPA for the analysis of TCDD in water, wastewater (EPA, 1982), soil and sediment (EPA, 1983) require a specified number of samples to be analyzed in duplicates or as spiked samples at levels near the detection limit. In addition, these methods specify routine performance evaluations with respect to isomer specificity by HRGC, consistency of response factors, evaluation of method blanks, qualitative criteria, and analysis of blind spiked samples. Other analysts (Gross et al., 1981; Langhorst and Shadoff, 1980; Kocher et al., 1978; Stalling et al., 1982; Tosine, 1981; O'Keefe et al., 1978; Mahle, 1977; Mitchum et al., in press) have also completed method validations as QA procedures with respect to the analysis of PCDDs. Stable Isotope Labeled Compounds— Chlorine-37 or carbon-13 labeled 2,3,7,8-TCDD were available for use as surrogate compounds for the analysis of 2,3,7,8-TCDD in most studies. The advantage of using these compounds is that they behave exactly as native TCDDs throughout extraction, cleanup, and gas chromatography separation. The mass spectra of the native and stable isotope labeled compounds vary enough to allow differentiation during the analysis. The surrogate compound added to the sample is used as a true internal standard for quantitation. The concentration calculated by the internal standard method provides a recovery correction. The method recovery can be determined by comparing area response of the quantitation ion for the internal standard in the sample extract versus the area response in an external standard. A more accurate measurement of method recovery can be achieved by adding a second internal standard to the sample extract prior to instrumental analysis. For example, TCDD can be measured in a sample by internal standard quantitation with accurate method recovery determination by combining the use of the carbon-13 and chlorine-37 labeled TCDD compounds. Stable isotope labeled compounds are also commercially available for the hepta- and octa-PCDDs. Nestrick and Lamparski (1982) have described techniques for synthesizing carbon-13 labeled penta- to octa-PCDDs from perchlorination of microgram amounts of carbon-13 labeled 2,3,7,8-TCDD. Langhorst and Shadoff (1980) have reported the analysis of tetra-, hexaand octa-PCDDs in human milk samples and have provided recovery data for each homolog determined by comparing external standards to the surrogate compounds. Table 21 is an example of the end use of surrogate recovery and internal standard quantitation as presented by Langhorst and Shadoff (1980). These data were generated while validating an analysis method for human milk. The values for percent recovery are the recoveries of the isotopically labeled surrogates . The percent accountability refers to the amount of observed native dioxin corrected for recovery of internal standard compared to the actual 56 �TABLE 21. PERCENT RECOVERY OF INTERNAL STANDARD AND PERCENT ACCOUNTABILITY FOR NATIVE DIOXINS SPIKED INTO CONTROL MILK HOMOGENATE Concn , PPt Added Found No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 avg std dev 1.0 1.0 1.3 1.3 1.3 2.0 2.0 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 3.9 3.9 12 12 0.2 0.2 0.7 0.8 0.8 1.6 2.2 2.0 2.4 2.2 1.9 1.9 2.5 1.9 1.4 2.3 3.3 10 12 OCDD HCDD 2,3 ,7,8-TCDD I Recovery Accountabili ty 42 65 56 33 49 38 96 25 25 32 26 21 11 23 34 34 25 33 33 37 ±19 20 20 54 62 62 80 110 77 92 85 73 73 96 73 54 59 85 84 97 77 ±16 Concn, ppt Added Found 5.3 5.3 6.6 6.6 6.6 9.9 9.9 13 13 13 13 13 13 13 13 20 20 60 60 4.5 3.7 7.0 7.2 7.2 5.7 8.0 9.0 9.9 8.9 8.1 13 13 13 10 17 11 42 42 % Recovery Account ability 70 67 64 57 57 70 86 76 68 85 74 47 38 53 52 59 61 81 85 67 ±11 Concn , PPt Added Found 84 70 106 109 109 58 81 68 74 67 61 98 98 98 75 86 56 70 70 81 ±17 21 21 27 27 27 41 41 53 53 53 53 53 53 53 53 80 80 240 240 21 38 23 40 37 49 41 41 33 42 38 38 40 45 57 51 98 360 270 Recovery Accountability 52 39 53 31 45 33 110 53 49 55 48 43 32 43 35 50 31 45 58 48 ±17 100 180 85 150 140 120 101 77 62 79 71 71 75 84 108 64 120 150 110 94 ±43 Precision data for eight replicates samples (no. 8-15) Com pound Cone added, ppt % Recovery % Accountability 2,3,7 ,8-TCDD HCDD OCDD 2.6 13 53 25 ± 7 65 ± 13 45 ± 8 78 ± 13 80 ± 16 78 ± 14 Source: a Langhorst, M. L., and L. A Shadoff, "Determination of Parts per Trillion Concentrations of Tetra-, Hexa-, Heptaand Octachlorodibenzo-g-dioxins in Human Milk Samples," Anal. Chem., 52:2037-2044 (1980). Corrected for internal standard recovery. �amount of the native PCDD that was added. The precision of the analysis was also demonstrated by the results of eight replicates (Table 21). These data show the usefulness and applicability of isotopically labeled compounds for producing analytical results of known quality. The data demonstrate the improvement of precision at concentrations much higher than the detection limit and also enlighten the analyst on the difficulties of measurements near the detection limits. Intralaboratory Validation of Method-Methods development for the analysis of any particular compound or compounds requires validation of the partial steps (extraction, cleanup, etc.) as well as the entire method. Table 22 is a summary of some of the published method vaidation data for PCDDs reported in the literature. Many of the methods reported were validated using replicate measurements of samples fortified with native PCDDs and/or the available isotopically labeled PCDDs. The mean percent recovery of the native compounds and the isotopic surrogates vary with respect to the validation experiments near the detection limit. The values summarized in Table 22 are indications of the total method performance. Intralaboratory validation requires a closer study of the individual method steps or procedures. This subject has already been demonstrated in this review with respect to extraction, cleanup and quantitation procedures in general. Table 23 is a specific example of intralaboratory validation of specific steps for a single method. The data in Table 23 were generated by diDomenico et al. (1979) while developing analytical methods for 2,3,7,8-TCDD in environmental samples near Seveso, Italy. The sample extracts were cleaned using a combination of the four procedures alluded to in Table 23 as cleanup steps A to D. Step A was a wash with concentrated sulfuric acid that did not introduce any appreciable losses. Procedure B involved a chromatographic cleanup with sulfuric acid treated Celite 545. Acetonitrile partitioning (Step C) of the extract from Step B proved to be an alternate to Step A but was also found to be more time consuming. Final cleanup (Step D) was accomplished using a micro alumina chromatography column. The data presented in Table 23 are representative of replicate analyses of spike recovery experiments for the individual steps and combination of procedures without the influence of a sample matrix. The recovery values for the extraction and cleanup of soil, grass, and cotton swabs are also compiled in Table 23 and are indicative of the entire method performance for samples spiked with 5 to 550 (Jg of 2,3,7,8-TCDD. Two other studies reported in the literature provided statistical evaluation of the method validation data. Langhorst and Shadoff (1980) and Gross et al. (1981) evaluated data for human milk and bovine fat samples, respectively. The methods of sample preparation and mass spectrometry analysis differed significantly between these two studies. 58 �TABU. 22. SUMMARY OK SOME PUBLISHED METHOD VALIDATION DATA FOR 2,3,7,8-TCDD RECOVERED FROM FORTIFIED BIOLOGICAL MATRICES Reference Bovine milk Langhorst and Shadoff (1980) Human milk Tosine (1981) Fish Gross et al. (1983) Human adipose TCDD Level of fortification Native Isotope ng-kg \ 1JC, (-"Cl) ng-kg 1 O'Keefe et al. (1978) Fish, liver Human milk 25 ± 7 37 ± 19 20 6 - 92 ± 4 0 1 1 1 1 ND 150 125 110 (37C1) 1000 0-125 0-5 250 Bovine milk Bovine milk O'Keefe et al. (1978) a 625 . 390 c -i+c Bovine fat 13 25 100 200 Baughman and Meselson (1973) Liver Kocher et al. (1978) Bovine fat +C 20 10 Source: ND 86 ± 15 68 ND 86 ± 17 100 ± 8 85 ± 9 71 ± 12 71 ± 12 87 i 21 _ 2 - 40 40 45 40 + 15 b - 38 17 17 4 4 4 4 66 66 66 _ 0.7 13 65 Mahle et al. (1977) Mean % recovery with s.d. Native Isotopes 8 166 2.6 6 16 38 Harless et al. (1980) Number of replicates 3 4 4 4 4 4 4 83.3 64 ND 100 80 85 88 ± ± ± ± ND 15 5 7 18 9 34 ± 7 7 1000 77 77 77 105 ± ± ± ± 18 18 18 9 76 ± 10 27 ± 5 Adapted from National Research Council of Canada, "Polychlorinated Dibenzo-j>-dioxins: Limitations to Current Analytical Techniques," NRCC No. 18576, ISSN 0316-0114, 1981. a Indicates publishing author's recovery data were converted from ng to ppt or from ppt to % by the Panel, b These data indicate the mean % accuracy for TCDD obtained with quality assurance samples, c Plus indicates fortified with isotope but amount not specified clearly. 59 �TABLE 23. RESULTS OF RECOVERY TESTS PERFORMED ON THE ANALYTICAL PROCEDURE, OR ITS SINGLE PARTS3 Operation and number of tests Minimum TCDD % Recovery Maximum Average Cleanup step B, 12 80 124 101 ± 12b Cleanup step C, 15 73 114 91 ± 13 Cleanup step D, 14 95 120 Cleanup steps B-D, 18 80 115 96 ± 12 Cleanup steps B-C-D, 8 58 93 76 ± 11 Soil, 28C 74 101 86 ± 7 Grass, 12C 72 98 85 ± 7 Cotton swabs , 46 68 112 86 ± 10 Source: 102 ± 7 diDomenico, A., et al., "Analytical Techniques for 2,3,7,8-Tetrachlorodibenzo-£-dioxin Detection in Environmental Samples After the Industrial Accident at Seveso," Anal. Chem., 51_: 735-740 (1979). a Cleanup step A did not introduce any appreciable 2,3,7,8-TCDD loss provided the operation was performed with the utmost care. This conclusion was reached after a number of recovery tests had been carried out by applying a sequence of cleanup steps including A. 2,3,7,8-TCDD quantity used: 0.1 and 0.01 |Jg/test in 1 ml solvent. b Standard deviation. c Values reported take into account 2,3,7,8-TCDD losses due to cleanup steps. 60 �The data generated by Langhorst and Shadoff (1980) for the analysis of tetra-, hexa- and octa-PCDDs for seven controls, seven replicate samples and spiked samples were presented in Table 21. The actual limitations of the method were defined by the dioxin level in the control sample and by statistical treatment of the data. Figure 17 is an example of the statistical data treatment for 2,3,7,8-TCDD and OCDD analysis. The heavy solid line is the actual level of native dioxin spiked. The dashed line represents the least squares fitted line for the dioxin concentration observed. The shaded area represents the total uncertainty of the determination including the error associated with the least squares fitted line and the error associated with the final recovery of dioxin for GC/MS analysis. The statistical validation of the method practiced by Gross et al. (1981) was generated from analytical results for 26 bovine fat samples and 26 standard solutions spiked at levels ranging from 0 to 81 ppt. The samples and standards were prepared and submitted simultaneously for TCDD analysis by PGC/ HRMS with blind sample codes. The sample identifications were decoded when all the analytical results (52 samples) were submitted for statistical analysis, The statistical analysis results for the standard solutions and beef fat samples are illustrated in Figure 18. The theoretical line y = x representing perfect extraction and quantitation was included for comparative purposes. Two sets of upper and lower 95% confidence limits were included for least squares regression of reported values (y) on spiking levels (x). The boundary lines closest to the regression lines represent the upper and lower 95% confidence limits for an infinite number of analyses under the same conditions. The outer boundary lines are indicators of the 95% confidence limits for a single analysis. Based on the results of the statistical analysis of the data, Gross et al. (1981) determined the lower limit of quantitation to fall between 5 and 9 ppt. Interlaboratory Studies— The review of analytical methods for PCDDs presented by the NRCC (1981) included an evaluation of the techniques with respect to applicability to matrix, specificity, method validation, and interlaboratory studies. None of the methods reviewed at the time was given the highest rating because evaluation through a collaborative study was not included. Since that time several interlaboratory studies have been completed or are still in progress. These studies are summarized in Table 24. The only study conducted for biological matrices with directions to follow a specific analytical method is with the pork adipose matrix (EMSL/LV). The participants were instructed to follow the procedures published by Harless et al. (1980) for parts per trillion measurements of 2,3,7,8-TCDD in pork adipose tissues. Samples that were analyzed in other studies were prepared according to Harless et al. (1980), while GC/MS measurements of 2,3,7,8-TCDD were conducted according to the practices of the individual laboratory. 61 �2 , 3 , 7 , 8 - Tetrachlorodibenzo-p-Dioxin 14 12 10 § 6 I 4 I 5 6 7 Amount Added (ppt) I 8 10 11 12 Ocfachlorocfibenzo-p-Dioxfn 350 300 250 a. a. 200 g 150 100 50 0 Source: 50 100 150 Amount Added (ppt) 200 250 Langhorst, M. L. , and L. A. Shadoff, "Determinations or Parts per Trillion Concentrations of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-p-dioxins in Human Milk Samples," Anal. Chem. . _52, 2037-2044 ( 9 0 . 18) Figure 17. Statistical treatment of validation data for 2,3,7,8-TCDD and OCDD in human milk samples. 62 �Theoretical Line, Y = X Regression Line, Y = 0.98X - 1.30 90 80 70 95% Conf. Limits for Regression Line 95% Conf. Limits for Individual Analyses TJ 50 o o_ £ 40 o 5 30 20 10 0 10 20 30 40 50 60 TCDD Added (ppt) 70 80 70 80 Theoretical Line, Y = X Regression Line Y = 0.89X + 90 80 95% Conf. Limits: Regression Limits Individual 70 & 6° t o JU I 40 " O a ¥ 30 20 10 0 Source: 10 20 30 ' 40 50 60 TCDD Added (ppt ) Gross, M. L . , T. Sun, P. A. Lyon, S. F. Wojinski, D. R. Milker, A. E. Dupuy, Jr. , and R. G. Heath, "Method Validation for the Determination of Tetrachlorodibenzodioxin at the Low Parts per Trillion Level," Anal. Chem. , _53_, 1902-1906 (1981). Figure 18. Statistical treatment of reported concentrations versus concentrations of TCDD actually added to standard solutions and beef adipose. 63 �TABLE 24. Matrix Water Wastewater Soil Sediment INTERLABORATORY STUDIES AND METHOD VALIDATIONS FOR THE ANALYSIS OF TETRACHLORODIBENZO-E-DIOXINS (TCDD) Number of participating laboratories Method (s) TCDD concentration range Reference 13 EPA Method 613 EPA Region VII protocol for soil and sediment 20 - 200 ppt a 1 - 100 ppb EMSL/LVb 1 - 100 ppt EMSL/LVb McMillin et al. (1982) Pork adipose Earless et al. (1980)c Human adipose Harless et al. (1980)d 3 1 - 100 ppt Dioxin Monitoring Program Gross et al. (1981) Beef adipose Harless et al. (1980)d 4 1 - 100 ppt Dioxin Monitoring Program Gross et al. (1981) Fish e 6 105 - 321 ppt Fish e 13 1 - 100 ppt Ryan et al. (1983) Fish f 8 1 - 200 ppt O'Keefe et al. (1983) 2 1 - 100 ppt Dioxin Monitoring Program Gross et al. (1980) 1-100 ppt EMSL/LVb 1 - 10 ppb EMSL/LVb Beef adipose Soil Sediment Pottery clay d Harless et al. (1980) EPA Region VII protocol for soil and sediment EPA Region VII protocol for soil and sediment a Brumley et al. (1981) a These samples are used as performance evaluation samples for laboratories involved with analysis of 2,3,7,8-TCDD in soils and sediments. b Personal communicator] J. Donnelley (1983). c Participating laboratories were instructed to follow procedure as described by Harless et al. (1980). Some modifications to the method were reported. d Samples prepared by method described by Harless et al. (1980), but GC/MS conditions varied, e Each Participting laboratory used current in-house analytical method. f Sample extracts divided at specific steps of one protocol and submitted to participating laboratories for further analysis. 64 �Some examples of the data generated in these interlaboratory studies are presented in Tables 25 to 28. Table 25 illustrates the results of the analysis of human adipose samples for 2,3,7,8-TCDD. These samples were analyzed initially by PGC/HRMS. A subset of these samples were reextracted and/or reanalyzed at other laboratories to provide interlaboratory validation of reported detections at these low levels. The validation of the analyses was accomplished in two ways. Remaining extracts from PGC/HRMS were reanalyzed using HRGC/HRMS, and portions of the tissues were submitted for reextraction and cleanup followed by HRGC/HRMS. All samples were coded and their identities were not known to the analysts. Based on the interlaboratory validation study, it was confirmed that two of the three samples designated as having heavy exposures contained 2,3,7,8-TCDD at higher levels than those observed for other participants. In addition, 2,3,7,8-TCDD was detected in tissue from other exposed and nonexposed persons designated as controls that were also examined by the interlaboratory studies. Table 26 provides a comparison of the results obtained by the different methods, PGC/HRMS versus HRGC/HRMS. Gross et al. (1981) have discussed the differences in concentration as reflecting the relatively large uncertainties in quantitation techniques in the parts per trillion range. Some of the variations between sample extracts analyzed by PGC/HRMS and HRGC/HRMS may be due in part to differences in resolution of the 2,3,7,8-TCDD from the other 21 possible isomers. In addition, the results may indicate sample inhomogeneities since different portions of unhomogenized tissue were used in each experiment. The interlaboratory study reported by Ryan et al. (1983) involved 13 laboratories having experience in determination of low levels (parts per trillion) of 2,3,7,8-TCDD in biological samples. Each laboratory agreed to analyze four fish samples for 2,3,7,8-TCDD using their routine extraction, cleanup and detection procedures. Table 27 presents the data reported by 8 of the 13 laboratories. The relative standard deviation for samples A, C and D is surprisingly low (14.0, 18.4, and 25.3%, respectively) considering the picograms per gram levels in the original sample. This variation is significantly less than that predicted by Horowitz et al. (1980) for low level quantitation. The recoveries of the internal standard (either carbon-13 or chlorine-37) 2,3,7,8-TCDD are presented in Table 28 for six of the laboratories that used internal standard quantitation. The average recovery of the individual laboratories ranged from 57 to 82% with a relative standard deviation of approximately 25%. The range of all the individual measurements yielded 29 to 109% recovery of the internal standard. This difference in method performance indicated the needs and usefulness of the internal standard quantitation approach. The results indicated fish samples C and D (Table 27) contained similar levels of 2,3,7,8-TCDD. These data were statistically evaluated according to the methods of Youden to determine variations between laboratories (systematic error) and within laboratories (random error). The results indicated that the difference between laboratories (reproducibility) was somewhat greater than the variance within laboratories (repeatability), although the differences reported were not significant at the 95% confidence level. 65 �TABLE 25. RESULTS OF ANALYSIS OF TCDD IN HUMAN ADIPOSE TISSUE' VA code number Concentration (ppt) Detection limit (ppt) Percent recovery Ratio0 .85 .75 .77 - "Heavily Exposed Veterans" 10 10 19 26 26 23 35, h NDD 99 63 4 9 3 10 6 65 100+ 20 90 45 ND ND 7 8 5 5 2 4 6 3 50 80 50 40 100 5 5 ND3 3 9 4 ND 5 5 12 10 ND 13 ND 3 3 3 2 3 3 4 3 4 4 3 6 5 3 65 50 40 55 60 65 60 80 45 45 100+ 100 60 95 "Lightly Exposed Veterans" _ 1 13 28 28 34 .88 .78 .85 "Possibly Exposed Veterans" 6 8 9 11 12 14 16 24 24 25 25 27 29 30 .90 .90 .77 .88 .74 .71 .78 .88 - "Controls" 5 7 17 18 20 21 23 23 31 32 33 4 3 4,3 ND 5 6 8 6 7 4 14 66 4 2 3 4 4 3 2 3 4 4 7 (continued) 65 60 75 30 50 35 100 55 50 60 100 1.02 .92 .84 .86 1.07 .78 .98 .74 .94 �TABLE 25 (continued) VA code number Concentration (ppt)b Detection limit (ppt) Percent recovery Ratio0 50 85 50 .77 .94 .76 "USAF Scientists" 2 3 4 Source: a 5 4 6 2 1 2 Gross, M. L., J. 0. Lay, P. A. Lyon, D. Lippstred, N. Kangas, R. L. Harless, S. E. Taylor, and A. E. Dupuy, "2,3,7,8Tetrachlorodibenzo-£-dioxin Levels in Adipose Tissue of Vietnam Veterans" (personal communication). Sample sizes ranged from 2.2 to 11.6 g for each extraction. Internal standard amounts used varied from 2.0 - 2.6 ng/extraction. b ND = not detected. c Ratio of intensities of m/z 320 and m/z 322. Acceptable values are 0.78 ± 0.10. 67 �TABLE 26. RESULTS OF INTERLABORATORY VALIDATION STUDIES VA Code UN-L/UN-L3 TAC/RTPC TAC/RTPd 36 - 173 - — 86 20 3 - 10 - UNL/RTPb UN-L/UN-Le "Heavily Exposed Veterans" _ VA-26 VA-10 VA-19 63,99 23,35 ND(3)e — ND(29) USAF Researchers VA-3 VA-2 48 5 _ 24 Other Vietnam Veterans _ VA-13 VA-8 VA-9 VA-15 VA-34 ND(2) 5 ND(3) 7 5 ND(0.2) 3 3 - ND(7) 5 - — ND(7) 18 ND(5r Controls _ VA-17 VA-18 VA-21 VA-31 VA-20 Source: 4,3 f ND(4) 69 ND(4) 5 5 3 - 20 8 12 ND(3) "™ *™ 14 - 9 19 20 Gross, M. L., J. 0. Lay, P. A. Lyon, D. Lippstred, N. Kangas, R. L. Harless, S. E. Taylor, A. E. Dupuy, "2,3,7,8-Tetrachlorodibenzo-p_-dioxin Levels in Adipose Tissue of Vietnam Veterans" (personal communication). a Extracted at UN-L/analyzed at UN-L (University of Nebraska, Lincoln). The values given in parentheses are the detection limits. b Portion of the extract from UN-L/analyzed at RTP (Research Triangle Park). c Extracted at TAC (Toxicant Analysis Center)/analyzed at RTP. d Another portion of tissue shipped from UN-L, extracted at TAC/analyzed at RTP. e Extracted at UN-L/analyzed at UN-L. Results obtained with knowledge of the code. f Poor recovery of internal standard (< 40%). g Isotope ratio for m/z 320 and n/z 322 not correct. 68 �TABLE 27. CONCENTRATION OF 2,3,7,8-TCDD IN FISH SAMPLES FROM INTERLABORATORY STUDY Values are single determinations expressed in pg/g (ppt). Fish sample C D NDC(10) 35 45 58 ND(1.3) 37 33 4d 49 ND(2) 23 19 5e 58 ND(1) 34 38 ND(5) 51f 55 Lab No. la 104b 3 B A 6 ND(5)f 7 72 ND(2.3) 25 32 9 70 ND(5) 33 27 12 60 378 26 32 A Av. h 61.2 30.4 32.3 3.6 8.5 cv, % n Source: 0 5.6 8.2 14.0 - 18.4 25.3 6 SD 6 7 7 Ryan, J. J., J. C. Pilon, H. B. S. Conacher, and D. Firestone, "Interlaboratory Study for the Analysis of Fish for 2,3,7,8Tetrachlorodibenzo-£-dioxin," in press, 1983. a Also reported GC/ECD values of 103, ND(10), 39, 37 pg/g, respectively. b Value given judged to be an outlier by Dixon's test; recovery of this sample was judged by the analyst to be high (74%), so an average recovery (51%) was used to calculate value given. c Not detected followed by bracketed detection limits in pg/g. d Also reported higher values of 58, ND(2), 37, 38 pg/g for acid-base method; these values are closer to average than neutral method preferred by the analyst. e Confirmed by atmosphere pressure-negative chemical ionization GC/MS on same extract with values of 54, ND(2.3), 32, and 31 ppt, respectively, for samples A, B, C, D. f Value given judged to be outlier. g Value given judged to be outlier; subsequent analysis showed a value of ND(10) pg/g. h Does not include any outliers or values from laboratory 6. 69 �TABLE 28. PERCENT RECOVERIES OF INTERNAL STANDARD TCDD IN THE INTERLABORATORY STUDY Av.3 SD cv, % lb 57.0 11.6 20.4 3C 69.0 16.6 24.1 4d 80.0 12.3 15.4 5C 83.1 18.7 22.5 r 35.3 5.2 14.7 9C 74.6 18.4 24.7 12C 81.8 14.5 17.7 Lab No. A A 67.7 Range Source: 29-109 Ryan, J. J., J. C. Pilon, H. B. S. Conacher, and D. Firestone, "Interlaboratory Study for the Analysis of Fish for 2,3,7,8-Tetrachlorodibenzo-£-dioxin," in press, 1983. a Each value represents the average of 4 reported values. b Fortified duplicate with native 2,3,7,8-TCDD. c 13 d 37 C-2,3,7,8-TCDD. Cl-2,3,7,8-TCDD. 70 �The analytical results for samples C and D, although somewhat limited, were further evaluated to see if there were significant differences for the different analytical methodologies. No differences were determined with this treatment for methods that used digestion or extraction; high or low resolution mass spectrometry; and specific or nonspecific isomer separaton. Needs for Future Validation Studies-Although several interlaboratory studies have been conducted, there is need for further validation of specific procedures. The results from such studies presented by Ryan et al. (1983), Brumley et al. (1981), Gross et al. (1980), and O'Keefe et al. (1983) demonstrate that the available methodologies are comparable in performance and provide reasonably valid measurements with respect to other approaches. Critical assessments of specific steps of the methodologies have not been attempted. There is need for a single laboratory to compare the best approach, for example, for initial extraction of PCDDs from the sample matrix (acid digestion, alcoholic saponification, or neutral extraction). Likewise, cleanup procedures should be compared and evaluated to generate information on recovery of analytes and separation from specific contaminants such as PCBs, chlorodiphenylethers, chloromethoxybiphenyls, etc. In order to accomplish this evaluation of methodology, it is important to vary only one parameter at a time. Additional validation of the methods is required if it is necessary to measure other homologs of PCDDs other than TCDD. Another important aspect that must be evaluated when considering interlaboratory validation of a single method is the ease of individual analytical steps. In order to demonstrate and fully evaluate the validity of a method all participating laboratories should be able to manipulate all procedural steps with good precision and accuracy. 71 �SECTION 5 APPLICABLE TECHNIQUES - RECOMMENDATIONS Following the first submission of the literature review (Sections 1 - 4 , this report), MRI was requested to organize a meeting to discuss analytical approaches for the analysis of PCDDs and PCDFs. This Section presents a synopsis of a discussion meeting held at Midwest Research Institute, Kansas City, Missouri, on April 27 and 28, 1983. The specific purpose of this meeting was to discuss analytical methods that are applicable to the analysis of polychlorinated dibenzo-£-dioxins (PCDDs) and dibenzofurans (PCDFs) in human adipose tissues. The discussion meeting was attended by scientists (Appendix A) recognized as experts in the field of PCDD and PCDF analysis. The meeting served as an additional source of information pertaining to specific considerations for low parts-per-trillion measurements of PCDDs and PCDFs in human adipose tissue. The meeting followed the first draft of the written literature review with preliminary method recommendations for analysis of PCDDs in adipose tissue and a peer review (Stanley, 1982) of the initial document. DISCUSSION MEETING SUMMARY The meeting was organized to promote open and detailed discussion on the criteria that must be considered for an effective analytical method and study of PCDD levels in human adipose tissue. Scientists recognized as experts in the field of PCDD and PCDF analysis were invited to participate (Appendix A). Most of the participants had previously provided peer review comments to the literature review and preliminary recommendations. Representatives from EPA/OTS and the VA presented overviews on the design of a general population study to determine PCDD exposure using existing adipose sample repositories and an update of the VA involvement with Agent Orange studies. A summary of the primary issues identified from the peer reviews of the literature review and preliminary recommendations was presented. These issues included (a) the need for stating the primary objectives of the program, (b) the use of high resolution mass spectrometry (HRMS) versus low resolution mass spectrometry (LRMS), (c) the practical limitations of the proposed extract cleanup procedures, and (d) additional measures for the quality assurance program. The discussion of methods of analysis were held to four major subject headings. They were: primary objectives of the method, instrumental analysis, sample preparation, and method validation (Appendix B). 72 �Primary Objectives The primary objective of the method was defined as the need to accurately determine the level of 2,3,7,8-TCDD in human adipose tissue. However, higher chlorinated PCDDs and PCDFs including tetrachlorodibenzofurans are also of interest in the overall program. It was recognized that it may be difficult to achieve this additional data if sufficient sample sizes are not available. It was emphasized that if possible, a method should provide data on PCDDs and PCDFs with chlorine substitution in the 2,3,7,8-positions. The objectives of a method as expressed in the discussion were (a) isomer specific measurement of 2,3,7,8-TCDD, (b) determination of PCDDs and PCDFs with chlorine substitution in the 2,3,7,8-positions, and (c) measurement of total PCDDs and PCDFs by homolog. Instrumental Analyses It was a consensus that mass spectrometry is necessary for the identification and quantitation of PCDDs and PCDFs. The criteria for qualitative identification of PCDDs and PCDFs are similar regardless of whether low resolution or high resolution mass spectrometry is used for analysis. These criteria include (1) coincident response of at least two ions characteristic of the molecular ion cluster of a specific homolog, (2) the proper ion response ratio, and (3) the correct retention times. In addition, response of a fragment ion characteristic of the loss of COC1 is necessary to confirm the presence of a PCDD congener. Electron impact ionization mass spectrometry was presented as the most useful for analysis of PCDDs and PCDFs. It was pointed out, however, that other mass spectrometry methods, negative ion chemical ionization in particular, are applicable to the analysis of specific PCDD or PCDF congeners. These alternate mass spectrometry methods also provide additional sensitive confirmatory information. Method detection limits for analysis of 2,3,7,8-TCDD were estimated at 1 to 5 parts per trillion (ppt), providing that the original sample size is at least 1 to 3 g. It was recognized by the meeting participants that this small sample size may not be sufficient to allow analysis for other PCDDs and PCDFs. The only means of extending a small sample for the analysis of all PCDDs and PCDFs is to isolate the different chlorinated homologs using liquid chromatography techniques. Estimates for method detection limits of octachlorodibenzo£-dioxins and octachlorodibenzofurans ranged from 20 to 100 ppt. It was generally recognized that the use of high resolution rather than low resolution is based on the extent that potential interferences are removed from the sample extract. If sufficient extract cleanup is achieved, low resolution mass spectrometry is acceptable for the analysis of PCDDs and PCDFs at low parts per trillion. Compounds that are known to interfere with the analysis of 2,3,7,8-TCDD were presented in the literature review. A set of compounds that was not considered in the review was chlorinated benzoquinones. The need to study 73 �the potential interferences of these types of compounds was addressed and discussed. The potential interferences to the analysis of higher chlorinated PCDDs and PCDFs have not been identified. It was speculated that compounds similar to the interferences for 2,3,7,8-TCDD analysis, but with greater chlorine substitution, may interfere with the analysis of other PCDDs and PCDFs. These compounds include the polychlorinated biphenyls, benzoquinones, benzylphenyl ethers, and diphenyl ethers. Sample Preparation The procedures for sample preparation were discussed with respect to quantitative extractions of PCDDs and PCDFs from sample matrices and the degree of cleanup necessary for instrumental analysis. Several of the participants were asked to describe their analytical preparation schemes and to provide comments as to the advantages or purpose of the particular method steps. The cleanup procedures discussed were designed with final instrumental technique in mind. The procedure presented in the preliminary recommendation required less stringent cleanup and high resolution gas chromatography/high resolution mass spectrometry. Other procedures require mass extensive cleanup, fractionation of the sample extract with high performance liquid chromatography and analysis by packed column gas chromatography/low resolution mass spectrometry. Figures 19 and 20 are schematics of the two analytical schemes presented at the meeting following the discussion of sample preparation. These schemes represent routes to final analysis by HRGC/HRMS and HRGC/LRMS. A macro alumina column is recommended to provide additional separation of PCDDs and PCDFs from interferences. If it is necessary to separate PCDDs and PCDFs by homolog, an HPLC step may be necessary. Several of the meeting attendees commented on the advantages of activated charcoal for separating PCDDs and PCDFs from interferences. This step has been proposed as part of the overall scheme for low resolution mass spectrometry. Considerable discussion centered around the equivalency of extraction procedures. There have been some indirect comparisons of the recovery efficiencies of acidic digestions, basic saponifications, and neutral extractions with fish samples in previous interlaboratory studies. There is a need for a direct comparison of these procedures followed using a common rigorous cleanup procedure to fully evaluate the extraction efficiencies. A more definitive study could be performed by using adipose containing a bioincurred radiolabeled PCDD. Recovery of the radiolabeled PCDD versus recovery of a spiked stable isotope PCDD would provide detailed information on the actual recovery from adipose tissue for each specific technique. 74 �Initial Sample Preparation Spike with Stable Isotope Labeled PCDDs • • 37 CI-2,3,7,8-TCDD C-2,3,7,8-TCDD 13 • Other 13c- Labeled PCDDs/PCDFs I Extraction Neutral Extraction or Basic Saponification I Bulk Matrix Cleanup Macro- Column Acid/Base Modified Silica Gel Provides Cleanup of Oxidizable Compounds with Rapid Sample Turnaround, Improved Cleanup Efficiency and Recovery I Removal of Chemical Interferences Provide Separation of PCBs and Other Potential Interferences from PCDDs Alumina Macro-Column HRGC/HRMS Simultaneous Detection, Quantitation and Confirmation Figure 19. Schematic of proposed analytical method using high resolution mass spectrometry (HRMS). 75 �Initial Sample Preparation Spike with Stable Isotopes Labeled PCDDs •37ci-2,3,7,8-TCDD • 13c-2,3,7,8-TCDD • Other 13c- Labeled PCDDs/PCDFs Extraction Neutral Extraction or Basic Saponification 1 Bulk Matrix Cleanup Macro-Column Ac id/Base Modified Silica Gel Provides Cleanup of Oxidizable Compounds with Rapid Sample Turnaround, Improved Cleanup Efficiency and Recovery Carbon/Glass Fiber or Carbon/Celite Adsorption Column Provides Selective Adsorption of PCDDs/PCDFs and Similar Residues I Removal of Chemical Interferences Alumina Macro-Column Provide Separation of PCBs and Other Potential Interferences from PCDDs/PCDFs i HRGC/LRMS Simultaneous Detection, Quantitation and Confirmation Figure 20. Schematic of proposed analytical method using low resolution mass spectrometry (LRMS). 76 �Method Validation Validation of a primary analytical method will require participation of at least eight laboratories with a minimum of four samples prepared as Youden pairs at mid-level and lower concentration ranges. A few of the meeting participants felt that a comprehensive evaluation requires the analysis of multiple samples (7-10) at several spiked concentration levels to measure precision of the analytical procedures and to define the actual method detection limit. In addition, an interest was expressed to analyze the same set of samples used for the method validation by alternate analytical methods to independently verify the analysis. Other points that were presented regarding preparation for a full-scale method validation are presented below. Some samples prepared for the method validation should be spiked with potential interferences to define limitations of the analytical methods. Also, samples of known spiked PCDD concentrations should be provided to a small group of laboratories as a means of identifying potential problems with the written analytical method. One of the more significant contributions was the suggestion to include actual adipose samples with spiked quality control (QC) samples in the interlaboratory study. Actual adipose samples of sufficient mass would be selected from the repository. These samples would be split and supplied to different laboratories along with the QC samples. The resulting data from the paired laboratories should provide some preliminary information on general population exposure as well as method performance. Ideally, the method validation study should encompass analyses for tetrato octachloro-PCDDs and PCDFs. Realistically, this may not be possible because of the significant cost and time required to complete a validation of this magnitude in a single study. It must be kept in mind that the most important issue is analysis for 2,3,7,8-TCDD. DISCUSSION MEETING RECOMMENDATIONS The discussion meeting was beneficial in identifying several major programs necessary for the success of the primary analytical method validation and the proposed population studies. These programs include (a) the need for establishing a repository of PCDD/PCDF standards of known quality, (b) the organization and implementation of a strong quality assurance program, (c) the acquisition of sufficient human adipose to generate a homogeneous sample matrix for the QA program, (d) independent studies of extraction procedures using adipose with bioincurred radiolabeled PCDDs, (e) intralaboratory ruggedness testing of a proposed analytical method, and (f) interlaboratory evaluation of the proposed method. Simultaneous activity in several of these areas is necessary in the coming months. The participation of scientists experienced in analysis of PCDDs and PCDFs is needed in many of these programs to aid in designing solid approaches for a successful program. The major action items are discussed in more detail below. 77 �Intralaboratory Testing A draft of a method will be prepared. The individual steps of the method will be characterized using clean samples or spiked blanks. The total method will be evaluated using adipose tissue spiked with PCDDs and PCDFs. Carbon-14 radiolabeled PCDDs and PCDFs will be used if available to help define critical variables in a more rapid fashion than can be achieved with HRGC/MS. Ruggedness testing of the method will require varying sample sizes, quantities of adsorbent, volumes of solvent, etc., to help define the critical variables and limitations of the method. The total method including HRGC/MS will be challenged with potential interferences spiked in the sample matrix. A formal method will be written and will undergo peer review to identify uncertainties in the written instructions. Tissue Program A large pool of homogeneous adipose tissue is needed to prepare quality control (QC) samples for the overall QA program and interlaboratory validation studies. It is estimated that 40 to 50 kg of adipose tissue are needed to prepare a sufficient number of control samples at known spiked concentration levels with and without the addition of potential interferences. The adipose tissues will be collected through the National Human Monitoring Program network. A repository of the samples will be established. When sufficient samples are collected (40 to 50 kg total), the samples will be pooled and rendered to provide a homogeneous matrix that will be subdivided for spiking procedures. The timing of tissue collection is important since these activities will overlap with the design of the Quality Assurance Program, the Standards Program and needs of the intralaboratory testing and interlaboratory studies. The following parameters will be considered for collection of the pool of adipose tissues. The adipose tissues will be collected from male trauma victims within 24 hr after death. The specimens will be collected from males born between 1937 and 1952, which is coincident with birthdates for veterans serving in the Vietnam area. All adipose tissues will be frozen until composited for homogenization with other specimen. A background analysis of the homogenized tissue is necessary to provide information on the levels of PCDDs, PCDFs, and potential interferences. It is recognized that the assistance of laboratories (EPA/RTP; University of Nebraska; Wright State University; Health Protection Branch, Food Division, Canada; Fish and Wildlife Services) with experience in the analysis of PCDDs and PCDFs in adipose tissues will be of benefit in obtaining this information in the most expedient manner. These background analyses must be completed before proceeding with subdividing the homogeneous tissue for spiking purposes as designed under the QA program. 78 �Quality Assurance Program The Quality Assurance Program will influence the success of the overall program with respect to method validation and performance evaluations for the routine analysis of tissue samples for population studies. The quality assurance program plan will provide details for preparation of fortified tissue samples containing PCDDs and potential interferences. The tissue samples should be spiked with at least one isomer from each PCDD and PCDF homolog. A subset of QA tissue samples should be spiked with compounds known to interfere with the analysis of PCDDs and PCDFs. This type of performance evaluation sample will provide information on the potential for false positive results. The QA program will specify the procedures for sample handling, sample coding, frequency of the spiked QC samples, distribution of samples, data handling, and decoding. The design of the QA program must be initiated immediately to provide support to the intra- and interlaboratory method validations. Standards Program Procurement of a sufficient quantity of PCDD and PCDF congeners of known quality is essential to provide consistent results from interlaboratory studies, method validations, and actual analysis programs. There is a critical need to establish a repository of the PCDD and PCDF compounds. Currently, participants from the discussion meeting are being surveyed for inventories of PCDDs and PCDFs in specific laboratories. The information gathered from this survey will be useful in identifying needs for procurement or synthesis of specific congeners for the overall program. Labeled PCDDs are commercially available as carbon-13, chlorine-37, and carbon-14 labeled TCDDs, and carbon-13 labeled octachlorodibenzo-£-dioxin. These compounds will be used as surrogates or internal standards for sample analyses. Stable isotope labeled compounds are not currently available for penta-, hexa-, and heptachloro-PCDDs or any of the PCDFs. If the overall objective of the analysis program is to include tetra- through octachloro-PCDDs and PCDFs, there is a need to study the most cost-effective means to acquire these compounds. The standards program will also cover collection of potential interferences. Polychlorinated biphenyls and DDE are readily available for addition to samples as interferences. However, compounds such as the chlorinated diphenyl ethers, chlorinated benzylphenyl ethers, and chlorinated benzoquinones may be more difficult to obtain. Purity of the standard compounds, stable isotope labeled standards, and potential interferences must be known before these compounds can be used for spiking the homogenized tissues for the QA program. Once the purity of the compounds is documented and the repository established, distribution of the compounds to collaborators may occur. Distribution of the standards will be most effective by supplying solutions of accurately determined concentrations. 79 �Bioincurred Program The need to investigate the extraction efficiency of PCDDs and PCDFs from adipose tissue was discussed at the meeting. A feasible approach to study the extraction efficiency is through use of tissue with bioincurred compounds. The use of carbon-14 radiolabeled 2,3,7,8-TCDD in feeding studies will provide the necessary bioincurred matrix. The recovery of bioincurred carbon-14 labeled compound compared to recovery of spiked stable isotope labeled or native compounds will indicate the adequacy of sample spiking procedures and provide an absolute extraction efficiency. The bioincurred program will necessarily require several months for completion of the study. Again, there is need for the overlap of this study with other aspects of the total program. Interlaboratory Studies Interlaboratory studies are necessary for primary analytical method validation and background analyses of homogenized tissues. The Interlaboratory studies required for method validation include a preliminary study of three to four laboratories followed by a full-scale collaborative study with 10 to 12 participants. The preliminary method evaluation will be conducted with samples of known concentration. The purpose of the preliminary study is to familiarize the participants with the method and identify potential difficulties of the method. The analytical method will be refined if necessary based on the preliminary study. The full-scale method validation will require a significantly larger number of participants. The samples will include the samples prepared under the QA program and will be submitted to the participants under blind codes. The design of the interlaboratory studies should include adipose samples that are (a) spiked near the method limit of detection, (b) spiked with potential interferences, and (c) Youden pairs to determine accuracy and precision. Actual samples may possibly be included in the interlaboratory validation. These samples would be selected from the pool of samples identified by EPA/OTS and the VA as representative of the general population and Vietnam veterans. Figure 21 is an example of such a study. The TAC sample numbers are included only for illustration purposes. Each actual sample would be split between two laboratories to provide additional data on the accuracy and precision of interlaboratory measurements. Organization of the interlaboratory studies must begin several months before the actual study. The efforts for organization of the interlaboratory study will overlap with the quality assurance program, standard program, tissue collection, and intralaboratory studies. 80 �Sample Method Validation A F F F F F F F F X a a t t a t + Interf. a t + Interf. a t + PCDD a t + PCDD a t + PCDD + Interf. a t + PCDD + Interf. TAC 352 353 354 355 356 357 358 359 B X X X X X X X C Laboratory D E F G X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Figure 21. Example of possible interlaboratory organization. 81 H X �APPENDIX A INVITED PARTICIPANTS 82 �"METHODS OF ANALYSIS FOR POLYCHLORINATED DIBENZO-£-DIOXINS (PCDDs) IN BIOLOGICAL MATRICES" EPA/VA/MRI Meeting April 27-28, 1983 Invited Participants: Dr. Donald Barnes* Environmental Protection Agency 401 M Street, S.W. Mail Drop TS-788 Washington, DC 20460 FTS 382-2897 Dr. David Bayse Center for Environmental Health Center of Disease Control Atlanta, GA 30333 FTS 236-4111 Dr. Warren Bontoyan Environmental Protection Agency Building 402 ARC East Beltsville, MD FTS 344-2187 20705 Dr. Mike Dellarco Environmental Protection Agency Office of Exploratory Research RD 680 401 M Street, S.W. Washington, DC 20460 FTS 382-5730 Dr. Fred DeRoos* Battelle Institute Columbus Laboratories 505 King Avenue Columbus, OH 43201 (614) 424-4247 * Attendees. 83 �Dr. Joseph Donnelly* LEMSCO USEPA/Lockheed P.O. Box 15027 Las Vegas, NV 89114 FTS 545-2299 Dr. Ralph C. Dougherty* Department of Chemistry Florida State University Tallahassee, FL 32306 (904) 644-5725 Dr. Aubry Dupuy* USEPA Toxicant Analysis Center Building 1105, NSTL NSTL Station, MS 39529 FTS 494-3212 Dr. David Firestone* Food and Drug Administration HFF426 200 C Street S.W. Washington, DC 20204 FTS 245-1381 Dr. Michelle Flicker* Vetrans Administration Kansas City, MO (816) 861-4700 Dr. Jean Futrell* University of Utah Department of Chemistry Salt Lake City, UT 84112 (801) 581-7307 Dr. Michael Gross University of Nebraska Department of Chemistry Lincoln, NE 68586 (402) 472-2794 * Attendees. 84 �Mr. Robert Harless* Environmental Monitoring Systems Laboratory Environmental Protection Agency Research Triangle Park, NC 27711 FTS 629-2248 (919) 541-2248 Dr. Harry Hertz A113, Chemistry National Bureau of Standards Washington, DC 20234 Dr. Fred Hileman* Monsanto Research Center 1515 Nicholas Road P.O. Box 8, Station B Dayton, OH 45407 (513) 258-3411 Dr. Mike Hoffman* USDA, FSIS Building 318 ARC-East Beltsville, MD 20705 (301) 344-1846 Dr. Verne Houk Center for Environmental Health Center for Disease Control Atlanta, GA 30333 FTS 236-4111 Dr. Philip C. Kearney USDA Building 050 BARC-West Beltsville, MD 20705 FTS 344-3076 Dr. Lawrence H. Keith* Radian Corporation P.O. Box 9948 8501 MoPac Blvd. Austin, TX 78766 (512) 454-4797 * Attendees. 85 �Dr. Robert Kleopfer* Environmental Protection Agency Region VII 25 Fuston Road Kansas City, KS 66115 (913) 374-4285 Dr. Frederick W. Kutz Environmental Protection Agency Office of Toxic Substances, TS-798 401 M Street, S.W. Washington, DC 20460 FTS 382-3583 Dr. Lester L. Lamparski Analytical Laboratories Dow Chemical Company Building 574 Midland, MI 48640 (517) 636-6207 Dr. W. Ligon* General Electric Corporate Research and Development P.O. Box 8 Building K-l Schenectady, NY 12301 Dr. Willie May* A113, Chemistry National Bureau of Standards Washington, DC 20234 Dr. James D. McKinney* National Institute of Environmental Health Sciences P.O. Box 12233 Research Triangle Park, NC 27709 Dr. Larry Needham* Center for Environmental Health Center of Disease Control Atlanta, GA 30333 FTS 236-4111 Attendees. 86 �Dr. Ross J. Norstrom* Department of the Environment National Wildlife Research Center 100 Gamelin Boulevard Building 9 Hull, Quebec Canada (819) 997-1410 Dr. Patrick O'Keefe* Center for Laboratories and Research New York State Department of Health Empire State Plaza Albany, NY 12201 (518) 473-3378 Dr. Jim Petty* Columbia National Fisheries Research Laboratory U.S. Fish and Wildlife Service Department of the Interior Route 1 Columbia, MO 65201 FTS 276-5399; (314) 875-5399 Mr. David P. Redford* Environmental Protection Agency Office of Toxic Substances, TS-798 401 M Street S.W. Washington, DC 20460 FTS 382-3583 Dr. John J. Ryan* Health Protection Branch Food Division Tunney's Pa s ture Ottawa K1A OL2 Canada (613) 593-4482 Dr. Lewis Shadoff* Analytical Laboratories Dow Chemical Company Building 574 Midland, MI 48640 (517) 636-5584 Attendees. 87 �Dr. David Stalling" Columbia National Fisheries Research Laboratory U.S. Fish and Wildlife Service Department of the Interior Route 1 Columbia, MO 65201 FTS 276-5399; (314) 875-5399 Dr. Michael Taylor* Wright State University Department of Chemistry Brehm Laboratory Dayton, OH 45435 (513) 873-3119 Dr. Paul Taylor* California Analytical Laboratories 5895 Power Inn Rd. Sacramento, CA 95824 (716) 381-5105 Dr. Anthony Wong* California Analytical Laboratories 5895 Power Inn Rd. Sacramento, CA 95824 (716) 381-5105 Major Alvin Young* Veterans Administration 810 Vermont Avenue, N.W. Washington, DC 20420 FTS 389-5534 MRI Participants: Dr. Mitch Erickson Dr. John E. Going Dr. Clarence Haile Mr. Gil Radolovich Dr. Jim Spigarelli Dr. John Stanley (816) 753-7600 FTS 758-6781 * Attendees. 88 �APPENDIX B DISCUSSION MEETING SCHEDULE OF EVENTS 89 �DISSCUSSION OF "Methods of Analysis for Polychlorinated Dibenzo-p-Dioxins (PCDD) in Biological Matrices" at Midwest Research Institute Kansas City, Missouri April 27-28, 1983 SCHEDULE OF EVENTS 8:30 - 9:00 - Registration of Participants - Arthur Mag Conference Center 9:00 - 9:10 - J. S. Stanley (MRI) Opening Remarks and Introductions 9:10 - 9:25 - David P. Redford (EPA/OTS) Primary Objectives of EPA/OTS in Assisting the VA with the Sampling and Analysis Program 9:20 - 9:35 - Dr. M. Flicker (VA) Overview of Veterans Administration Agent Orange Programs 9:35 - 9:55 - J. S. Stanley (MRI) Recommendations for Analytical Method Identifying the Primary Issues from Peer Reviews 9:55 - 12:00 - Instrumental Analysis - Discussion - Low Resolution vs. high resolution mass spectrometry - Definition of high resolution mass spectrometry - Compromises between low resolution and high resolution mass spectrometry - Quantitation practices - Criteria for qualitative identification Low resolution mass spectrometry High resolution mass spectrometry Gas chromatography - Criteria for quantisation Limits of detection Limits of quantisation - Isomer specificity What degree of confidence necessary with any method Possible interferences Quality assurance/quality control procedures - Role of screening techniques 90 �12:00 - 1:15 - Lunch 1:15 - 2:30 - Instrumental Analysis Discussion (Concluded) 2:30 - 3:15 - Sample Preparation - Discussion - Surrogate spiking - Approaches to preparing spiked samples with native PCDDs - Extraction procedures—neutral, acid or base - Cleanup of extract Advantages and disadvantages of the proposed cleanup procedures - Quality assurance/quality control procedures - Other cleanup procedures 3:15 - 3:25 - Break 3:25 - 5:00 - Sample preparation - Discussion (Concluded) 5:30 - 7:30 - Social Hour (Hilton Plaza Hotel) April 28. 1983 8:30 - 8:35 - J. S. Stanley - Opening Remarks 8:35 - 9:00 - A. L. Young - VA Need for Primary Analytical Method 9:00 - 11:00 - Method Validation Studies - Intralaboratory validation of extraction procedure Ruggedness testing of method--intralaboratory approach Preliminary interlaboratory studies Full-scale collaborative study . . . . Number of participating laboratories Number of total samples Preparation of spiked tissue samples Availability of native and isotopically labeled standards . Needs for spiking samples with potential interferences 10:15 - 10:25 - Break 10:25 - 12:00 - Summary of Discussions and Recommendations 91 �APPENDIX C BIBLIOGRAPHY 92 �APPENDIX C BIBLIOGRAPHY Adamoli, P., E. Angeli, G. Bandi, A. Bertolotti, E. Bianchi, L. Boniforti, I. Camoni, F. Cattabeni, G. Colli, M. Colombo, C. Corradi, L. De Angelis, G. De Felice, A. Di Domenico, A. Di Muccio, G. Elli, R. Fanelli, M. Fittipaldi, A. Frigero, G. Galli, P. Grassi, R. Gualdi, G. Invernizzi, A. Jemma, L. Luciano, L. Hanaro, A. Marinella, F. Merli, S. Nicosia, F. Rizzello, C. Rossi, G. Rossi, G. Salvatore, A. Sampolo, G. P. Schmidt, F. Taggi, G. Tebaldi, E. Zaino, and G. Zapponi, "Analysis of 2,3,7,8-Tetrachlorodibenzo-para-dioxin in the Seveso Area," in C. Ramel (Ed.), Chlorinated Phenoxy Acids and Their Dioxins, Ecol. Bull. (Stockholm), 27:31-38 (1978). Albro, P. W., and B. J. Corbett, "Extraction and Clean-Up of Animal Tissues for Subsequent Determination of Mixtures of Chlorinated Dibenzo-£-dioxins and Dibenzofurans," Chemosphere, 7:381-385 (1977). Albro, P. W., M. L. Luster, K. Chae, S. K. Chaudhary, G. Clark, L. D. Lawsori, J.T. Corbett, and J. D. McKinney, "Radioimmunoassay for Chlorinated Dibenzo|>-dioxins," Toxicol. Appl. Pharmacol., 50:137-146 (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., 19J7:155-169 ( 9 0 . 18) Albro, P. W. , "Validation of Extraction and Cleanup Procedures for Environmental Analysis," in J. D. McKinney (Ed.), Environmental Health Chemistry: The Chemistry of Environmental Agents as Potential Human Hazards, Ann Arbor Science Publishers, Inc. (1981), pp. 163-175. Baker, P. G., R. A. Hoodless, and J. F. C. Tyler, "A Review of Methods for the Determination of Polychlorodibenzo-£-dioxins and Polychloro-dibenzofurans in Phenoxyalkanoic Acid Herbicides," Pesticide Sci. , 12:297-304 (1981). Baughman, R., and M. Meselson, "An Analytical Method for Detecting TCDD (Dioxin): Levels of TCDD in Samples from Vietnam," Environmental Health Perspectives, 5:27-35 (1973). Baughman, R., and M. Matthew, "An Improved Analysis for Tetrachlorodibenzo£-dioxins," in E. H. Blair (Ed.), Chlorodioxins - Origin and Fate, Advances in Chemistry Series 120, American Chemical Society, Washington, D.C., 1973. Bell, R. A., "Synthesis of UL-13C Polychlorinated Dibenzofurans," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. 93 �Bertoni, G., D. Brocco, V. Di Palo, A. Liberti, M. Possanzini, and F. Bruner, "Gas Chromatographic Determination of 2,3,7,8-Tetrachlorodibenzodioxin in the Experimental Decontamination of Seveso Soil by Ultraviolet Radiation," Anal. Chem., 50:732-735, May 1978. Bowes, G. W., M. J. Mulvihill, B. R. T. Simoneit, A. L. Burlingame, and R. W. Risebrough, "Isolation and Identification of Chlorinated Dibenzofurans from Polychlorinated Biphenyls (PCB) and from Yusho Rice Oil Containing PCB," in F. Cattabeni, A. Cavallaro, G. Galli (Eds.), Dioxin: Toxicological and Chemical Aspects, Spectrum Publications, Inc. (1978), pp. 79-98. Bowman, M. C., (Ed.), in Handbook of Carcinogens and Hazardous Substances; Chemical and Trace Analysis, Marcel Dekker, Inc., (1982). Bradlaw, J. A., and J. L. Casterline, Jr., "Induction of Enzyme Activity in Cell Culture: A Rapid Screen for Detection of Planar Polychlorinated Organic Compounds," J. Assoc. Off. Anal. Chem., 62:904-916 (1979). Brumley, W. C., J. A. G. Roach, J. A. Sphon, P. A. Driefuss, D. Andrzejewski, R. A. Niemann, and D. Firestone, "Low-Resolution Multiple Ion Detection Gas Chromatographic-Mass Spectrometric Comparison of Six Extraction-Cleanup Methods for Determining 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Fish," J. Agric. Food Chem., 29:1040-1046 (1981). Buser, H. R., "Analysis of Polychlorinated Dibenzo-£-dioxins and Dibenzofurans in Chlorinated Phenols by Mass Fragmentography," J. Chromtogr., 107, 295-310 (1975). Buser, H. R., "Analysis of TCDDs by Gas Chromatography-Mass Spectrometry Using Glass Capillary Columns," F. Cattabeni, A. Cavallaro, G. Galli (Eds.), in Dioxin: Toxicological and Chemical Aspects, SP Medical and Scientific Books (1978), pp. 27-41. Buser, H. R., "Identification of Polychlorinated Dibenzo-g-dioxins and Dibenzofurans in Environmental and Industrial Samples Using High-Resolution Gas Chromatography and Mass Spectrometry," in A. Frigerio and M. McCamish (Eds.), Recent Developments in Mass Spectrometry in Biochemistry and Medicine, 6, Elsevier Scientific Publishing Company, Amsterdam (1980), pp. 515-521. Buser, H. R., "High-Resolution Gas Chromatography of the 22 Tetrachlorodibenzo-£-dioxin (TCDD) Isoraers," in 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds.), Chlorinated Dioxins and Related Compounds Impact on the Environment, Pergamon Press (1982), pp. 15-24. Buser, H. R., "The Seveso Accident - An Environmental Application of Mass Spectrometry," Trends in Analytical Chemistry, l.:318-321 ( 9 2 . 18) Buser, H. R., and C. Rappe, "Identification of Substitution Patterns in Polychlorinated Dibenzo-£-dioxins (PCDDs) by Mass Spectrometry," Chemosphere, 2:199-211 (1978). Buser, H. R., and C. Rappe, "High-Resolution Gas Chromatography of the 22 Tetrachlorodibenzo-£-dioxin Isomers," Anal. Chem., 52:2257-2262 (1980). 94 �Cairns, T., L. Fishbein, and R. K. Mitchum, "Review of the Dioxin Problem, Mass Spectroraetric Analyses of Tetrachlorodioxins in Environmental Samples," Biomedical Mass Spectrometry, 7:484-492 (1980). Camoni, I., A. Di Muccio, D. Pontecorvo, and L. Vergori, "Clean-Up Procedure for the Extraction of Soil Samples in the Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin," J. Chromatogr., 153:233-238 (1978). Cattabeni, F., Cavallaro, A., and Galli, G. (Eds.), Dioxin: and Chemical Aspects (1978). Toxicological Cavallaro, A., G. Bandi, G. Invernizzi, L. Luciani, E. Mongini, and G. Gorni, "Negative Ion Chemical lonization MS as a Structure Tool in the Determination of Small Amounts of PCDD and PCDF," in 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds.), Chlorinated Dioxins and Related Compounds Impact on the Environment, Pergamon Press (1982), pp. 55-66. Chess, E. K., and M. L. Gross, "Determination of Tetrachlorodibenzo-g-dioxins by Mass Spectrometric Metastable Decomposition Monitoring," Anal. Chem., 52:2057-2061 (1980). Choudhary, G., "Occupational Exposure to Chlorinated Dioxins and Dibenzofurans, A Perspective," in L. H. Keith, G. Choudry, C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Clement, R. E., G. A. Eiceman, F. W. Karasek, D. Bowers, and M. L. Parsons, "Rapid Analysis of Chlorinated Dioxins in Complex Organic Mixtures by Application of a Gas Chromatography-Mass Spectrometry-Calculator System with UserDeveloped Software," J. Chromatogr., 189:53-59 (1980). Crummett, W. B., "The Problem of Measurements Near the Limit of Detection," Annals of the New York Academy of Sciences, 320:43-47 (1979). Crummett, W. B., "Fundamental Problems Related to Validation of Analytical Data Elaborated on the Example of TCDD," Toxicological and Environmental Chemistry Reviews, 3:61-71 (1979). Cutie, S. S., "Recovery Efficiency of 2,3,7, 8-Tetrachlorodibenzo-j>-dioxin from Active Carbon and Other Particulates," Analytica Chimica Acta, 123:25-31 (1981). DeRoos, F. L., D. G. Aichele, J. E. Tabor, and M. Larson, "High Pressure Liquid Carbon Dioxide Extraction of Tetrachlorodibenzo-£-dioxins (TCDDs)," presented at the American Society for Mass Spectrometry 30th Annual Conference on Mass Spectrometry and Allied Topics, June 1982, Honolulu, Hawaii. Di Domenico, A., F. Meril, L. Boniforti, I. Camoni, A. Di Muccio, F. Taggi, V. Vergori, G. Colli, 0. Elli, A. Gorni, P. Grassi, G. Invernizzi, A. Jemma, L. Luciani, F. Cattabeni, L. De Angelis, G. Galli, C. Chiabrando, and F. Fanelli, "Analytical Techniques for 2,3,7,8-Tetrachlorodibenzo-£-dioxin Detection in Environmental Samples after the Industrial Accident at Seveso," Anal. Chem., 51:733-734 (1979). 95 �Di Domenico, A., G. Viviano, and G. Zapponi, "Methodological Problems in Assessing 2,3,7,8-TCDD Environmental Contamination at Seveso," in 0. Hutzinger, R. W. Frei, E. Merrian, F. Pocchiari (Eds.)> Chlorinated Dioxins and Related Compounds Impact on the Environment, Pergamon Press (1982), pp. 47-54. Dobbs, A. J., and C. Grant, "Octachlorodibenzo-£-dioxin in Wood Treatment Materials and Treated Wood," Chemosphere, IQ:1185-1193 (1981). Dolphin, R. J., and F. W. Willmott, "Separation of Chlorinated Dibenzo-£dioxins from Chlorinated Congeners," J. Chromatogr., 149:161-168 (1978). Egestad, B., T. Curstedt, and J. SjSvall, "Simple Procedures for Enrichment of Chlorinated Aromatic Pollutants from Fat, Water and Milk for Subsequent Analysis by High-Resolution Methods," Analytical Letters, ^5:293-307 (1982). Environmental Protection Agency, "Determination of 2,3,7,8-TCDD in Soil and Sediment," Region VII Laboratory, Kansas City, KS, February 1983. Environmental Protection Agency, "2,3,7,8-Tetrachlorodibenzo-£-dioxin— Method 613," in J. E. Longbottom and J. L. Lichtenberg (Eds.), Methods for Organic Chemical Analysis of Municipal and Industrial Wastewater, EPA-600/482-057, July 1982. Erk, S. D., M. L. Taylor, and T. 0. Tiernan, "Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin Residues on Metal Surfaces by GC-MS," Chemosphere, 4:7-14 (1979). Esposito, M. P., T. 0. Tiernan, and F. E. Dryden, Dioxins, U.S. Environmental Protection Agency Report No. EPA-600-2-80-197, November 1980, Industrial Environmental Research Laboratory, Office of Research and Development, Cincinnati, Ohio. Facchetti, S., A. Fornari, and M. Montagna, "Distribution of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in the Tissues of a Person Exposed to the Toxic Cloud at Seveso," Proceedings of the 8th International Mass Spectrometry Conference, Advances in Mass Spectrometry, 8:1405-1414 (1980). Fanelli, R., M. P. Bertoni, M. Bonfanti, M. G. Castelli, C. Chiabrando, G. P. Martelli, M. A. Noe, A. Noseda, and C. Sbarra, "Routine Analysis of 2,3,7,8Tetrachlorodibenzo-£-dioxin in Biological Samples from the Contaminated Area of Seveso, Italy," Bull. Environ. Contam. Toxicol., 24:818-823 (1980). Fanelli, R., M. G. Castelli, G. P. Martelli, A. Noseda, and S. Garattini, "Presence of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Wildlife Living Near Seveso, Italy: A Preliminary Study," Bull. Environ. Contam. Toxicol., 24: 460-462 (1980). Firestone, D., "Determination of Polychlorodibenzo-£-dioxins and Polychlorodibenzofurans in Commercial Gelatins by Gas-Liquid Chromatography," J. Agric. Food Chem., 25:1274-1280 (1977). 96 �Firestone, D., "The 2,3,7,8-Tetrachlorodibenzo-para-Dioxin Problem: A Review," in C. Ramel (Ed.), Chlorinated Phenoxy Acids and Their Dioxins, Ecol. Bull, 27:39-52 (1978). Firestone, D., "Report on Oils and Fats," J. Assoc. Off. Anal. Chem., 6 ( ) 02: 354-355 (1977). Firestone, D., M. Glower, Jr., A. P. Borsetti, R. H. Teske, and P. E. Long, "Polychlorodibenzo-£-dioxin and Pentachlorophenol Residues in Milk and Blood of Cows Fed Technical Pentachlorophenol," J. Agric. Food Chem., 27:1171-1176 (1979). Fishbein, L., "Halogenated Contaminants: Dibenzo-£-dioxins and Dibenzofurans," M. C. Bowman (Ed.), In Handbook of Carcinogens and Hazardous Substances: Chemical and Trace Analysis, Marcel Dekker, Inc., New York and Basel (1982), pp. 671-740. Foerst, D. L. , J. E. Longbottom, and R. J. Wesselman, "The Determination of 2,3,7,8-TCDD in Industrial and Municipal Wastewaters, Method 613, Part 1 Development and Detection Limits," in L. H. Keith, G. Choudry, C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press , in press. Freudenthal, J., "The Quantitative Determination of TCDD with Different Mass Spectrometric Methods," in F. Cattabeni, A. Cavallaro, and G. Galli (Eds.), Dioxin: Toxicological and Chemical Aspects, SP Medical and Scientific Books (1978), pp. 43-50? Gasiewicz, T. A., J. R. Olson, L. H. Geiger, and R. A. Neal, "Absorption, Distribution and Metabolism of 2,3,7,8-Tetrachlorodibenzobioxin (TCDD) in Experimental Animals," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London ( 9 3 , pp. 495-525. 18) Gross, M. L., 0. J. 0. Lay, Jr., P. A. Lyon, D. Lippstreu, N. Kangas, R. L. Harless, S. E. Taylor, and A. E. Dupuy, Jr., "2,3,7,8-Tetrachlorodibenzo-j>dioxin Levels in Adipose Tissue of Vietnam Veterans," personal communication. Gross, M. L., T. Sun, P. A. Lyon, S. F. Wojinski, D. R. Bilker, A. E. Dupuy, Jr., and R. G. Heath, "Method Validation for the Determination of Tetrachlorodibenzodioxin at the Low Parts-per-Trillion Level," Anal. Chem., 53:1902-1906 (1981). Gruninger, J. H., and A. Freedman, "Pattern Recognition Techniques Applied to Spectroscopic Detection of Trace Amounts of Chlorinated Dioxins," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Harless, R. L., "Analytical Methodology for 2,3,7,8-Tetrachlorodibenzo--£-dioxin and Its Application by the United States Environmental Protection Agency to Human and Environmental Monitoring," presented at the Assistant Administrators' Program Review, United States Environmental Protection Agency, Washington, D.C., April 1-2, 1980. 97 �Harless, R. L., and R. G. Lewis, "Quantitative Determination of 2,3,7,8Tetrachlorodibenzo-£-dioxin Residues by Gas Chromatography/Mass Spectrometry," in 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds,), Chlorinated Dioxins and Related Compounds Impact on the Environment, Pergamon Press, 1982, pp. 25-36. Harless, R. L., and R. G. Lewis, "Quantitative Capillary Column Gas Chromatography-Mass Spectrometry Methods of Analysis for Toxic Organic Compounds," presented at Symposium on "Practical Solutions to Quantitative Capillary Column Gas Chromatography," The Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Atlantic City, New Jersey, March 10-15, 1980. Harless, R. L., P. R. Ellis, and E. 0. Oswald, "A Brief Review of the 2,3,7,8Tetrachlorodibenzo-£-dioxin (TCDD) Problem, Analytical Clean-up Methodology, and Mass Spectrometric Methods of Analysis," presentation given at NIEHS, RTP, North Carolina, June 28, 1977. Harless, R. L. , and E. 0. Oswald, "Low- and High-Resolution Gas ChromatographyMass Spectrometry (GC-MS) Method of Analysis for the Presence of 2,3,7,8-Tetrachlorodibenzo-£-dioxin (TCDD) in Environmental Samples," in F. Cattabeni, A. Caveallaro, and G. Galli (Eds.), Dioxin-Toxicology and Chemical Aspects, Chapter 6 (1978). Harless, R. L., E. 0. Oswald, and M. K. Wilkinson, "Gas Chromatography/Mass Spectrometric Methods of Analyses for Toxaphene and Dioxins in Human and Biological Samples," 26th Annual Conference on Mass Spectrometry and Allied Topics, 1978. Harless, R. L., E. 0. Oswald, R. G. Lewis, A. E. Dupuy, Jr., D. D. McDaniel, and H. Tai, "Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Fresh Water Fish," Chemosphere, U:193-198 (1982). Harless, R. L., and E. 0. Oswald, "The Application of High Resolution Gas Chromatography Interfaced with Chemical lonization and High Resolution Electron Impact Mass Spectrometry for Environmental Analysis," 26th Annual Conference on Mass Spectrometry and Allied Topics, Washington, D.C., 1977. Harless, R. L., E. 0. Oswald, M. K. Wilkinson, A. E. Dupuy, Jr., D. D. McDaniel, and H. Tai, "Sample Preparation and Gas Chromatography-Mass Spectrometry Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin," Anal. Chem., 52:1239-1245 (1980). Harless, R. L., A. E. Dupuy, and D. D. McDaniel, "High Resolution Mass Spectrometry Methods of Analysis for Chlorinated Dibenzo-£-dioxins and Dibenzofurans," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Publishing Corporation (1983), pp. 65-73. 98 �Harrison, D. D., and R. C. Crews, "A Field Study of Soil and Biological Specimens From a Herbicide Storage and Aerial-Test Staging Site Following Long-Term Contamination with TCDD," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London (1983), pp. 323-339. Hass, J. R., and M. D. Friesen, "Qualitative and Quantitative Methods for Dioxin Analysis," Annals New York Acad. Sci., 320:28-42 (1979). Hass, J. R., M. D. Friesen, D. J. Harvan, and C. E. Parker, "Determination of Polychlorinated Dibenzo-£-dioxins in Biological Samples by Negative Chemical lonization Mass Spectrometry," Anal. Chem., 50:1474-1479 (1978). Hass, J. R., M. D. Friesen, and M. K. Hoffman, "Recent Mass Spectrometric Techniques for the Analysis of Environmental Contaminants," in J. D. McKinney (Ed.), Environmental Health Chemistry, The Chemistry of Environmental Agents as Potential Human Hazards, Ann Arbor Science (1981), pp. 219-243. Heath, R. G., "Interlaboratory Method Validation Study for Dioxin," an Interim Report, Human Effects Monitoring Branch, OPP, OTS, EPA, January 5, 1979. Horowitz, W., L. R. Kamps, and K. W. Boyer, "Quality Assurance in the Analysis of Foods for Trace Constituents," J. Assoc. Off. Anal. Chem., 63:1344-1354 (1980). Huckins, J. N., D. L. Stalling, and J. L. Johnson, "Industrial Chemicals: Silicic Acid Chromatographic Separation of Polychlorinated Biphenyls and Pesticides: Some Contaminants and Limitations," J. Assoc. Off. Anal. Chem., 59:975-981 (1976). Huckins, J. N., D. L. Stalling, and W. A. Smith, "Foam-Charcoal Chromatography for Analysis of Polychlorinated Dibenzodioxins in Herbicide Orange," J. Assoc. Off. Anal. Chem., 61:32-38 (1978). Huff, J. E., J. A. Moore, R. Saracci, and L. Tomatis, "Long-Term Hazards of Polychlorinated Dibenzodioxins and Polychlorinated Dibenzofurans," Environmental Health Perspectives, 36:221-240 (1980). Hummel, R. A., "Clean-Up Techniques for the Determination of Parts per Trillion Residue Levels of 2,3,7,8-Tetrachlorodibenzo-£-dioxin (TCDD)," J. Agric. Food Chem., 25:1049-1053 (1977). Hummel, R. A., and L. A. Shadoff, "Specificity of Low Resolution Gas Chromatography-Low Resolution Mass Spectrometry for the Detection of Tetrachlorodibenzo£-dioxin in Environmental Samples," Anal. Chem., 52:191-192 (90. 18) Hutzinger, 0. K. Olie, J. W. A. Lustenhouwer, A. B. Okey, S. Bandiera, and S. Safe, "Polychlorinated Dibenzo-£-dioxins and Dibenzofurans: A Bioanalytical Approach," Chemosphere, W:19-25 (1981). 99 �Jensen, D. J., R. A. Hummel, N. H. Mahle, C. W. Kocher, and H. S. Higgins, "A Residue Study on Beef Cattle Consuming 2,3,7,8-Tetrachlorodibenzo-£-dioxin," J. Agric. Food Chetn. , 29:265-268 (1981). Karasek, F. W., and I. Onuska, "Trace Analysis of the Dioxins," Anal. Chem., 54:309A-321A (1982). Keith, L. H., R. C. Hall, R. C. Hanisch, and A. E. Jones, "A New GC Detector for Screening TCDD Samples," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Kocher, C. W., N. H. Mahle, R. A. Hummel, L. A. Shadoff, and M. E. Getzendaner, "A Search for the Presence of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Beef Fat," Bull. Environ. Contam. Toxicol., 20:229-236 (1978). Korfmacher, W. A., and R. K. Mitchum, "Relative Retention Times of 39 Polychlorinated Dibenzo-£-dioxins Using SP2100 Fused Silica Capillary Chromatography," Journal of HRC and CC, 5:682-682 (1982). Kuehl, D. W., and R. C. Dougherty, "Screening of Human and Food Chain Samples for Contamination with Toxic Substances Using Negative Chemical lonization Mass Spectrometry," Proceedings of the 8th International Mass Spectrometry Conference, Advances in Mass Spectrometry, 8:1451-1459 ( 9 0 . 18) Kuehl, D. W., R. C. Dougherty, Y. Tondeur, D. L. Stalling, L. M. Smith, and C. Rappe, "Negative Chemical lonization Studies of Polychlorinated Dibenzo-£ dioxins, Dibenzofurans and Naphthalenes in Environmental Samples," in J. D. McKinney (Ed.), Environmental Health Chemistry: The Chemistry of Environmental Agents as Potential Human Hazards, Ann Arbor Science Publishers, Inc. (1981), pp. 245-261. Lacey, M. J. and C. G. MacDonald, "Ion Current Surfaces and the Determination of Tetrachlorodibenzo-£-dioxin by Tandem Mass Spectrometry," Anal. Chem., 54: 135-136 ( 9 2 . 18) Lamparski, L. L., and T. J. Nestrick, "Synthesis and Identification of the 10 Hexachlorodibenzo-£-dioxin Isomers by High Performance Liquid and Packed Column Gas Chromatography," Chemosphere, .10:5-18 (1981). Lamparski, L. L., and T. J. Nestrick, "The Isomer-Specific Determination of Tetrachlorodibenzo-£-dioxin at Part per Trillion Concentrations," in 0. Hutzinger, R. W. Frei, E. Merian, and E. Pocchiari (Eds.), Chlorinated Dioxins and Related Compounds: Impact on the Environment, Pergamon Press (1982), pp. 1-14. Lamparski, L. L., N. H. Mahle, and L. A. Shadoff, "Determination of Pentachlorophenol, Hexachlorodibenzo-£-dioxin, and Octachlorodibenzo-£-dioxin in Bovine Milk," J. Agric. Food Chem., 26:1113-1116 (1978). Lamparski, L. L., and T. J. Nestrick, "Determination of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-£-dioxin Isomers in Particulate Samples at Parts per Trillion Levels," Anal. Chem., 52:2045-2054 ( 9 0 . 18) 100 �Lamparski, L. L., and T. J. Nestrick, "Micro Chlorination Procedure for Synthesis of Higher Chlorinated Dibenzo-£-dioxins from [13C]-2,3,7,8-Tetrachlorodibenzo-£-dioxin," Anal. Chem., 54:402-406 (1982). Lamparski, L. L., T. J. Nestrick, and R. H. Stehl, "Determination of Partper-Trillion Concentrations of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Fish," Anal. Chem.. 51.: 1463-1482 (1979). Langhorst, M. L., and L. A. Shadoff, "Determination of Parts-per-Trillion Concentrations of Tetra-, Hexa-, Hepta-, and Octachlorodibenzo-£-dioxins in Human Milk Samples," Anal. Chem., 52:2037-2044 ( 9 0 . 18) Liberti, A., D. Brocco, I. Allegrini, A. Cecinato, and M. Possazini, "Solar and UV Photodecomposition of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in the Environment," The Science of the Total Environment, 10:97-104 (1978). Liberti, A., P. Ciccioli, E. Brancaleoni, and A. Cecinato, "Determination of Polychlorodibenzo-£-dioxins and Polychlorodibenzofurans in Environmental Samples by Gas Chromatography-Mass Spectrometry," J. Chromatogr., 242:111-113 (1982). Lustenhouwer, J. W. A., I. Olie, and 0. Hutzinger, "Chlorinated Dibenzo-£dioxins and Related Compounds in Incinerator Elluents: A Review of Measurements and Mechanisms of Formation," Chemosphere, 9^:501-522 (1980). Luster, M. I., P. W. Albro, K. Chae, L. D. Lawson, J. T. Corbett, and J. D. McKinney, "Radioimmunoassay for Quantitation of 2,3,7,8-Tetrachlorodibenzofuran," Anal. Chem., 52:1497-1500 (1980). Luster, M. I., P. W. Albro, K. Chae, S. K. Chaudhary, and J. D. McKinney, "Development of Radioimmunoassays for Chlorinated Aromatic Hydrocarbons," in J. D. McKinney (Ed.), Environmental Health Chemistry: The Chemistry of Environmental Agents as Potential Human Hazards, Ann Arbor Science Publishers, Inc. (1981), pp. 279-297. Mahle, N. H., H. S. Higgins, and M. E. Getzendaner, "Search for the Presence of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Bovine Milk," Bull. Environ. Contain. Toxicol., 18:123-130 (1977). Mahle, N. H., and L. A. Shadoff, "The Mass Spectrometry of Chlorinated Dibenzo£-dioxins," Biomedical Mass Spectrometry, 9_:45-60 (1982). Masuda, Y. H. Kuroki, T. Yamaryo, K. Haraguchi, M. Kuratsune, and S. T. Hsu, "Comparison of Casual Agents in Taiwan and Fukuoka PCB Poisonings," Chemosphere, 1.1:199-206 (1982). Masuda, Y. , H. Kuroki, and J. Nagayama, "Polychlorinated Dibenzofurans in the Tissues of Patients with Yusho and Their Enzyme Inducing Activities on Aryl Hydrocarbon Hydroxylase," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. 101 �Mazer, T., F. D. Hileman, R. W. Noble, and J. J. Brooks, "Synthesis of the 38 Tetrachlorodibenzofuran Isomers and Identification by Capillary Column Gas Chromatography/Mass Spectrometry," Anal. Chem., 55:104-110 (1983). Mazer, T., F. D. Hileman, R. W. Noble, and J. J. Brooks, "Characterization of Tetrachlorodibenzofurans," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. McKinney, J. D., "Analysis of 2,3,7,8-Tetrachlorodibenzo-para-dioxin in Environmental Samples," Ecol. Bull. , 27^:53-66 (1978). McKinney, J. D. P. W. Albro, R. H. Cox, J. R. Hass, and D. B. Walters, "Problems and Pitfalls in Analytical Studies in Toxicology," in S. K. Bandal, G. J. Marco, L. Golberg, and M. L. Leng (Eds.), ACS Symposium Series, No. 160, The Pesticide Chemist and Modern Toxicology (1981), pp. 439-460. McMillin, C. R., F. D. Hileman, D. E. Kirk, T. Mazer, B. J. Warner, J. Longbottom, and R. Wesselman, "Determination of 2,3,7,8-TCDD in Industrial and Municipal Wastewater, Method 613 - Part 2 - Performance Evaluation and Method Study Results," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. McNulty, W. P., K. A. Nielsen-Smith, J. 0. Lay, Jr., D. L. Lippstreu, N. L. Kangas, P. A. Lyon, and M. L. Gross, "Persistence of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Fat of a Rhesus Monkey (Macaca mulatta)," Food Chem. Toxicol., in press. Mieure, J. P., 0. Hicks, R. G. Kaley, and P. R. Michael, "Determination of Trace Amounts of Chlorodibenzo-£-dioxins and Chlorodibenzofurans in Technical Grade Pentachlorophenol," Journal of Chromatographic Science, 15^:275-277 (1977). Mitchum, R. K., J. R. Althaus, W. A. Korfmacher, and G. F. Moler, "Application of Negative Ion Atmospheric Pressure lonization (NIAP) Mass Spectrometry for Trace Analysis," in A. Quayle (Ed.), Advances in Mass Spectrometry, 8B, Heyden and Son Ltd. (1980), pp. 1415-1421. Mitchum, R. K., G. F. Moler, and W. A. Korfmacher, "Combined Capillary Gas Chromatography/Atmospheric Pressure Negative Chemical lonization/Mass Spectrometry for the Determination of 2,3,7,8-Tetrachlorodibenzo-p_-dioxin in Tissue," Anal. Chem., 52:2278-2282 (1980). Mitchum, R. K., W. A. Korfmacher, G. F. Moler, and D. L. Stalling, "Capillary Gas Chromatography/Atmospheric Pressure Negative Chemical lonization Mass Spectrometry of the 22 Isomeric Tetrachlorodibenzo-g-dioxins," Anal. Chem., 54:719-722 (1982). Mitchum, R. K., W. A. Korfmacher, and G. F. Moler, "Validation Study for the GC-Atmospheric Pressure lonization-MS Method for the Isomer-Specific Determination of 2,3,7,8-TCDD," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. 102 �Moler, G. G., R. R. Delongchamp, R. K. Mitchum, W. A. Korfmacher, and B. A. Pearce, "Confidence Limits for the Isotope Dilution/Gas Chromatographic-Mass Spectrometric Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Environmental Samples," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Nagayama, J., Y. Masuda, and M. Kuratsune, "Chlorinated Dibenzofurans in Kanechlors and Rice Oils Used by Patients with Yusho," Fukuoka Igaka Zasshi, (Fukuoka Acta Med.) 66(10):593-599 (1975). National Research Council of Canada, "Polychlorinated Dibenzo-£-dioxins: Limitations to the Current Analytical Techniques," NRCC No. 18576 (1981). National Research Council of Canada, "Polychlorinated Dibenzo-£-dioxins: Criteria for Their Effects on Man and His Environment," NRCC No. 18574 (1981). Nestrick, T. J., L. L. Lamparski, and R. H. Stehl, "Synthesis and Identification of the 22 Tetrachlorodibenzo-£-dioxin Isomers by High Performance Liquid Chromatography and Gas Chromatography," Anal. Chem., 51:2273-2281 (1979). Nestrick, T. J., L. L. Lamparski, and D. I. Townsend, "Identification of Tetrachlorodibenzo-£-dioxin Isomers at the 1-ng Level by Photolytic Degradation and Pattern Recognition Techniques," Anal. Chem., 52:1865-1874 (1980). Newton, M., and S. P. Snyder, "Exposure of Forest Herbivores to 2,3,7,8-Tetrachlorodibenzo-£-dioxin (TCDD) in Areas Sprayed with 2,4,5-T," Bull. Environ. Contam, Toxicpl., 20:743-750 (1978). Niemann, R. A., W. C. Brumley, D. Firestone, and J. A. Sphon, "Analysis of Fish for 2,3,7,8-Tetrachlorodibenzo-£-dioxin by Electron Capture Capillary Gas Chromatography," Anal. Chem., in press. Norstrom, R. J., and M. Simon, "Preliminary Appraisal of Tetra- to OctachloroDibenzodioxin Contamination in Eggs of Various Species of Wildlife in Canada," Proceedings of the 5th International Congress Pesticide Chemicals, Kyoto, Japan, August 29 - September 4, 1982, Pergamon Press, in press. Nowicki, H. G. C. A. Kleda, V. Current, and T. H. Scheefers, "Column Chromatography Fractionation of Complex Waste Water Sample Extracts for Measurement of ppt Levels of 2,3,7,8-Tetrachlorodibenzo-£-dioxin," Journal of HRC and CC, 4:178-179 (1981). O'Keefe, P. W. , "A Neutral Cleanup Procedure for TCDD Residues in Environmental Samples," in F. Cattabeni, A. Cavallaro, and G. Galli (Eds.), Dioxin: Toxicological and Chemical Aspects, SP Medical and Scientific Books (1978), pp. 59-78. O'Keefe, P. 0., C. Meyer, D. Hilker, K. Aldous, B. Jelus-Tyror, K. Dillon, R. Donnelly, E. Horn, and R. Sloan, "Analysis of 2,3,7,8-Tetrachlorodibenzo-£dioxin in Great Lakes Fish," in press. 103 �O'Keefe, P. W. M. S. Meselson, and R. W. Baughman, "Neutral Cleanup Procedure for 2,3,7,8-Tetrachlorodibenzo-£-dioxin Residues in Bovine Fat and Milk," J. Assoc. Off. Anal. Chem. , 61^:621-626 (1978). O'Keefe, P., C. Meyer, and K. Dillon, "Comparison of Concentration Techniques for 2,3,7,8-Tetrachlorodibenzo-£-dioxin," Anal. Chem., 54:2623-2625 (1982). O'Keefe, P. W., R. Smith, C. Meyer, D. Hilker, K. Aldous, and B. Jelus-Tyror, "Modification of a High Performance Liquid Chromatographic-Gas Chromatographic Procedure for Separation of the 22 Tetrachlorodibenzo-£-dioxin Isomers," J. Chromatogr. , 242:305-312 (1982). Petty, J. D., L. M. Smith, P. Bergqvist, J. L. Johnson, D. L. Stalling, and C. Rappe, "The Composition of PCDF and PCDD Residues in Sediments of the Hudson and Housatonic Rivers," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Phillipson, D. W., and B. J. Puma, "Identification of Chlorinated Methoxybiphenyls as Contaminants in Fish and Potential Interferences in the Determination of Chlorinated Dibenzo-p_-dioxins," Anal. Chem., 5_2:2332-2336 (1980). Porter, M. L., and J. A. Burke, "Separation of Three Chlorodibenzo-£-dioxins from Some Perchlorinated Biphenyls by Chromatography on an Aluminum Oxide Column," J. Assoc. Off. Anal. Chem., 54(6):1426-1428 (1971). Rappe, C., "Identification of Polychlorinted Dibenzofurans (PCDFs) Retained in Patients with Yusho," Chemosphere, 4:259-266 (1979). Rappe, C., H. R. Buser, and H. P. Bosshardt, "Identification and Quantification of Polychlorinated Dibenzo-p^dioxins (PCDDs) and Dibenzofurans (PCDFs) in 2,4,5-£-ester Formulations and Herbicide Orange," Chemosphere, 5:431-438 (1978). Rappe, C., H. R. Buser, D. L. Stalling, L. M. Smith, and R. C. Dougherty, "Identification of Polychlorinated Dibenzofurans in Environmental Samples," Nature, 292:521-526 (1981). Rappe, C., M. Nygren, H. Buser, Y. Masuda, H. Kuroki, and P. H. Chen, "Identification of Polychlorinated Dioxins (PCDDs) and Dibenzofurans (PCDFs) in Human Samples, Occupational Exposure and Yusho Patients," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London (1983), pp. 241-253. Rappe, C., M. Nygren, and G. Gustafsson, "Occupational Exposure to Polychlorinated Dioxins and Dibenzofurans," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Rappe, C., S. Marklund, M. Nygren, and A. Gara, "Parameters for Identification and Confirmation in Trace Analyses of Polychlorinated Dioxins and Dibenzofurans," in L. . Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. 104 �Robeson, R. K., J. R. Donnelly, A. E. Dupuy, Jr., R. L. Harless, and W. L. Budde, "Quality Assurance Samples for the Dioxin Monitoring Program," in L. H. Keith, G. Choudry, C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Ryan, J. J., and J. C. Pilon, "High-Performance Liquid Chromatography in the Analysis of Chlorinated Dibenzodioxins and Dibenzofurans in Chicken Liver and Wood Shaving Samples," J. Chromatogr. , 19_7:171-180 (1980). Ryan, J. J., J. C. Pilon, Henry B. S. Conacher, and D. Firestone, "Interlaboratory Study for the Analysis of Fish for 2,3,7,8-Tetrachlorodibenzo-£-dioxin," accepted for publication in J. Assoc. Off. Anal. Chem. (1983). Ryan, J. J., B. Lau, J. C. Pilon, D. Lewis, and H. McLeod, "A Survey of Lake Ontario Commerical Fish for 2,3,7,8-Tetrachlorodibenzo-£-dioxin Residues," L. H. Keith, G. Choudry, and C. Rappe (Eds.), In Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Safe, S., V. D. Jamieson, 0. Hutzinger, and R. E. Pohland, "Mass and Ion Kinetic Energy Spectra of Some Chlorinated Dibenzo-£-dioxins," Anal. Chem., 47:327-329 (1975). Shadoff, L. A., "The Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Human Milk," in J. Harvey, and G. Zweig (Eds.), Pesticide Analytical Methodology, ACS Symposium Series 136, Chapter 15, American Chemical Society, Washington, D.C. (1980), pp. 277-285. Shadoff, L. A., and R. A. Hummel, "The Determination of 2,3,7,8-Tetrachlorodibenzo-£-dioxin in Biological Extracts by Gas Chromatography Mass Spectrometry," Biomedical Mass Spectrometry, 5:7-13 (1978). Shadoff, L. A., R. A. Hummel, D. J. Jensen, and N. H. Mahle, "The Gas Chromatographic-Mass Spectrometric Determination of 2,3,7,8-Tetrachlorodibenzo-g-dioxin in Fat from Cattle Fed Ronnel Insecticide," Annali di Chimica, 6^:583-592 (1977). Shadoff, L. A., R. A. Hummel, L. Lamparski, and J. H. Davidson, "A Search for 2,3,7,8-Tetrachlorodibenzo-£-dioxin (TCDD) in an Environment Exposed Annually to 2,4,5-Trichlorophenoxyacetic Acid Ester (2,4,5-T) Herbicides," Bull. Environ., Contain. Toxicol. , 1.8:478-485 (1977). Smith, R. M., P. W. O'Keefe, K. M. Aldous, D. R. Hilker, and J. E. O'Brien, "2,3,7,8-Tetrachlorodibenzo-£-dioxin in Sediment Samples from Love Canal Storm Sewers and Creeks," Environ. Sci. Technol., 1/7:6-10 (1983). Smith, R. M., D. R. Hilker, P. W. O'Keefe, K. M. Aldous, C. M. Meyer, S. N. Kumar, and B. M. Jelus-Tyror, "Determination of Tetrachlorodibenzo-£-dioxins and Tetrachlorodibenzofurans in Environmental Samples by High Performance Liquid Chromatography, Capillary Gas Chromatography and High Resolution Mass Spectrometry," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London (1983), pp. 73-94. 105 �Smith, L. M., and J. L. Johnson, "Evaluation of Interferences from Seven Series of Polychlorinated Aromatic Compounds in an Analytical Method for Polychlorinated Dibenzofurans and Dioxins," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Stalling, D. L., J. D. Petty, L. M. Smith, C. Rappe, and H. R. Buser, "Isolation and Analysis of Polychlorinated Dibenzofurans in Aquatic Samples," in 0. Hutzinger, R. W. Frei, E. Merian, and F. Pocchiari (Eds.), Chlorinated Dioxins and Related Compounds, Impact on the Environment, Pergamon Press (1980), pp. 77-86. Stalling, D. L., J. D. Petty, L. M. Smith, and G. R. Dubay, "Contaminant Enrichment Modules and Approaches to Automation of Sample Extract Cleanup," in J. D. McKinney (Ed.), Environmental Health Chemistry: The Chemistry of Environmental Agents as Potential Human Hazards, Ann Arbor Science Publishers, Inc. (1981), pp. 177-193. Stalling, D. L., L. M. Smith, J. D. Petty, J. W. Hogan, J. L. Johnson, C. Rappe, and H. R. Buser, "Residues of Polychlorinated Dibenzo-£-dioxins and Dibenzofurans in Laurentian Great Lakes Fish," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Publishing Corporation (1983), pp. 221. Stanley, J. S., Polychlorinated and Preliminary No. 68-01-5915, "Peer Review and Author's Replies to Methods of Analysis for Dibenzo-£-dioxins in Biological Matrices - Literature Review Recommendations," Draft Interim Report No. 2, EPA Contract Task 6, April 1983. Taylor, M. L., R. 0. Yelton, C. F. Vanness, T. Mazer, J. G. Solch, S. D. Erk, and T. 0. Tiernan, "GC/MS Techniques for Determining Chlorodioxins in Complex Chemical-Environmental Matrices," presented at the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, March 13, 1980, Atlantic City, New Jersey. Tiernan, T. 0., "Analytical Chemistry of the Chlorinated Dibenzodioxins and Dibenzofurans - A Review of the Current Status," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Ann Arbor Science, Ann Arbor, Michigan in press. Tiernan, T. 0., J. H. Garrett, J. G. Solch, G. F. VanNess, and M. L. Taylor, "Capillary Column GC-Low Resolution MS Techniques for Monitoring Toxic Organic Compounds in Combustion Effluents," presented at the American Society for Mass Spectrometry, 30th Annual Conference on Mass Spectrometry and Allied Topics, June 1982, Honolulu, Hawaii. Tosine, H., "Method Used by the Ontario Ministry of the Environment Primarily for the Analysis of Fish Tissues and Raw and Treated Waters," Appendix V In Polychlorinated Dibenzo-p-dioxins: Limitations to Current Analytical Techniques, NRCC 18576 (1981). 106 �Tosine, H., "Dioxins: A Canadian Perspective," in L. H. Keith, G. Choudry, and C. Rappe (Eds.), Chlorinated Dioxins and Dibenzofurans in the Total Environment, Pergamon Press, in press. Tosine, H., D. Smillie, and G. A. V. Rees, "Comparative Monitoring and Analytical Methodology for 2,3,7,8-TCDD in Fish," in R. E. Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London (1983), pp. 127-139. Tucker, R. E., A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London ( 9 3 . 18) Williams, D. T., and B. J. Blanchfield, "Screening Method for the Detection of Chlorodibenzo-£-dioxins in the Presence of Chlorobiphenyls, Chloronaphthalenes, and Chlorodibenzofurans," J. Assoc. Off. Anal. Chem., 55:93-95 (1972). Williams, D. T. Williams, and b. J. Blanchfield, "An Improved Screening Method for Chlorodibenzo-]D-dioxins," J. Assoc. Off. Anal. Chem., 5_5:1358-1359 (1972). Wipf, H. K., and J. Schmid, "Seveso - An Environmental Assessment," in R. E Tucker, A. L. Young, and A. P. Gray (Eds.), Human and Environmental Risks of Chlorinated Dioxins and Related Compounds, Plenum Press, New York and London (1983), pp. 255-274. 107 �50272-101 REPORT DOCUMENTATION 1. REPORT NO. EPA-560/5-84-001 PAGE 4. Title and Subtitle Methods of Analysis for Polychlorinated Dibenzo-p_-Dioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) in Biological Matrices - Literature Review and Preliminary Recommendations 7. Author(s) J. S. Stanley 9. Performing Organization Name and Address Midwest Research Institute 425 Voiker Boulevard Kansas City, MO 64110 12. Sponsoring Organization Name and Address Office of Toxic Substances Field Studies Branch, TS-798 Environmental Protection Agency Washington...DC 20460 15. Supplement.try Notes Frederick W. Kutz, Project•Officer David P. Redford, Task Manager Daniel T. Heggem, Task Manager 3. Recipient's Accession No. 5. Report Date February 16, 1984 8. Performing Organization Rept. No. Final Report 10. Proiect/Task/Work Unit No. 4901-A(6) 11. Contract(C) or Grant(G) No. (0 68-01-5915 (G) Task 6 13. Type of Report & Period Covered Final 10/82 - 8/83 14. 16. Abstract (Limit: 200 words) . . . . i• • *., _, j 4. • The overall objective of this review and preliminary method recommendation was to assist the EPA's Office of Toxic Substances (OTS) in proposing an analytical method for PCDDs in human adipose tissue in conjunction with the Veterans Administration's (VA) Agent Orange study. The published literature on polychlorinated dibenzo-p_-dioxins (PCDDs) analyses for biological matrices was reviewed. The analytical methods are discussed for sample extraction, cleanup, and instrumental analysis. This report also presents a synopsis of a discussion meeting organized at the request of EPA/OTS concerning the analysis of polychlorinated dibenzo-£-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) held at Midwest Research Institute (MRI) on April 27 and 28, 1983. The primary objective of this meeting was to define the needs of an analytical method for the analysis of PCDDs and PCDFs in human adipose tissue. Several major programs were identified as necessary to achieve these goals. These included (a) the need for establishing a repository of PCDD/PCDF standards of known quality; (b) the organization and implementation of a strong quality assurance program; (c) the acquisition of sufficient human adipose tissue to generate a homogeneous sample matrix for the QA program; (d) independent studies of extraction procedures using bioincurred radiolabeled PCDDs; (e) intralaboratory ruggedness testing of a proposed analytical method; and (f) interlaboratory evaluation of the proposed method. 17. Document Analysis a. Descriptors 2,3,7,8-Tetrachlorodibenzo-p_-dioxin Polychlorinated dibenzofurans 2,3,7,8-TCDD PCDF Literature review Polychlorinated dibenzo-p_-dioxins Human adipose tissues PCDD Analysis b. ldentifiers/Op*n-Ended Terms Chromatography Analytical methods Mass spectrometry Recommendations Cleanup Extract-ion c. CO3ATI Fiiif'.l/Grnup IS. Availabii^y £t3tem»nt 21. No. of Pages 39. Security Clais (This Report) Release unlimited Unclassified | 119 I 20. Security Class (This ?Jxe) j 22. Focc i Unclassified ' (S-.-> ANSI-23J.1S) OPTIC .'.'.L FC:;M r/i >-'-77) � 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. 193 Folder The folder containing the original item. 5463 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 Stanley, John S. Description An account of the resource <strong>Corporate Author: </strong>United States Environmental Protection Agency (EPA), Office of Toxic Substances Date A point or period of time associated with an event in the lifecycle of the resource 1984-02-01 Title A name given to the resource Methods of Analysis for Polychlorinated Dibenzo-p-Dioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) in Biological Matrices - Literature Review and Preliminary Recommendations: Task 6 Final Report ao_seriesVIII Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 168 Folder The folder containing the original item. 5039 Series The series number of the original item. Series VII 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 Stanley, John S. John E. Going David P. Redford Frederick W. Kutz Alvin L. Young Date A point or period of time associated with an event in the lifecycle of the resource 1983-08-01 Title A name given to the resource Preprint Extended Abstract: Analytical Methods for Measurement of Polychlorinated Dibenzo-p-dioxins (PCDDs) in Human Adipose Tissues