1 15 18 https://www.nal.usda.gov/exhibits/speccoll/files/original/3822038d9ceab5efc202dd6661e0ed1f.pdf 9dbb381561ce5af93ecedcda02b02a03 PDF Text Text Item D Number 05557 D Author MacLeod, Colin M. CorporatB Author Panel on Herbicides, President's Science Advisory Com Report/Article Title Re NotScannBfl P°rt on 2,4,5-1 Journal/Book Title Year 197 Month/Day Ju| Color ° v D Number of Images ° Desorlpton Notes See item 882 for a later version Tuesday, March 19, 2002 - Page 5557 of 5611 �RA k /I\A /,FT R E P O R T ON 2, 4, 5-T • PANEL ON HERBICIDES i* PRESIDENT'S SCIENCE ADVISORY COMMITTEE c - ll �MEMBERS OF "THE PANEL **• Dr. Colin M. ^MacLeod, Chairman , New York University Medical Ce-nter,-New Y o r k , New York Dr. John D. Baldeschwieler, Stanford |University, Standford, California ;i '• ; i i • Dr. Nyle-C.' Brady, Cornell University, Ithaca, New York " D r . Emmamfel Farber, University of [Pittsburgh, Pittsburgh, "Pennsylvania| Dr. 'Paul Kotin, National Institute of Environmental Health • •• -~''~ Sciences, Durham, Nor£h;Carolina Dr. Brian Ma.cMahoh, Harvard School of Public. Health, Boston, ~~ Massachusetts Dr. Norton Nelson, New York University Medical Center, ' ' ' New York, New York Dr. L. Dale Newsom, Louisiana State University, Baton Rouge, . Louisiana •• Dr. John W. Tukey, Princeton University, Princeton, New Jersey G. Wilson, Children's Hospital Research Foundation, • :"C_incinnati, Ohio Dr. Edward J. Burger, Jr. , Staff,-' Technical Assistant, Office of Science and Technology, Executive Office • . '~ "of the President, Washington, D. C. �REPORT ON 2,4, 5-T Contents I. "~ Summary and Recommendations II. Introduction III. Chemistry -. ..— A. B. ,C. D. E. '-'"/ F. IV. • . V. C. D. E. F. G. - Analytic Methods Chemical Stability of 2,4, 5-T ' " ' Production .. ; ' Uses • 1. Domestic 2. Military •• ' _ Cost Sharing Program i ' . ' . . Acute Toxicological Investigations " Occupational Experience National Cancer Institute Screening Study ' " Reports of Birth Defects Toxicity of Dioxin ' . ." Teratogenesis Testing' ' Summary of Experiments Residues of 2, 4, 5-T in the Environment A. B. VII. • Synthesis ' . . . Formulation Solubilities Purity of Technical Grade Material Toxicology A. B. C. VI. ' Uses and Significance A._ 3.' • . Residues of 2, 4, 5-T resulting from. recommended uses in the United States Residues of 2, 4-D and 2, 4, 5-T resulting from, recommended uses by the military for defoliation. Some Ecological Effects �SECTION I SUMMARY AND RECOMMENDATIONS A. . Summary B. Recommendations- �I. A. SUMMARY AND RECOMMENDATIONS Summary This review of the herbicide, 2 , 4 , 5 - T , began with an examination of the r e s u l t s f r o m an experimental screening study which implicated it as a potential teratogen. It quickly became evident that examination f r o m such a r e s t r i c t e d basis was inadequate. T h e r e f o r e , the Panel' decided to study m o r e broadly important aspects of 2 , 4 , 5 - T , including details .of_its c h e m i s t r y and purity, its domestic uses and their relative importance,' the military significance attached to 2 , 4 , 5 - T as a defoliant, residue levels (in o r d e r to estimate probabilities of human e x p o s u r e ) , general e f f e c t s on the environment, as well as its toxicity. • In examining the toxicity of 2 , 4, 5-T, 'the Panel reviewed the information available from the literature (as well as some unpublished documents) which had been considered in the past in establishing policy decisions for the various uses to which' this pesticide had been put. ' . Selection of 2, 4, 5-T as an example for detailed examination has had a.number of advantages. There has been an extensive history of use and experience. 2, 4, 5_-Tjwas f i r s t r e g i s t e r e d on March 2, 1948,. by the Amch.em Products Company^ Ambler, Pennsylvania. Since that • time considerable information on its properties and uses has accumulated. Furthermore, it has been the subject of reviews by others in the past. Finally, the pr^pblem^which brought it to the Panel's attention, suspicion of teratogenicity, appears to be a relatively manageable problem in contrast to many other biological e f f e c t s , notably tumor production and genetic alter. This is important since the r e c o m m e n d a t i o n s which follow can be" made with a degree of confidence that cannot be applied to carcinogenic or mutagenic effects. For example, the d o s e - r e s p o n s e characteristic of . ( teratogens is generally r e s t r i c t e d to a relatively small range of dosage. Accordingly, a threshold below which no e f f e c t v/ould be expected can be. • ; assigned with more certainty. Experiments to determine this range of ( v a l u e s can be p e r f o r m e d in a relatively short time and do not require very large numbers of animals. i The Panel is gratified that some of its recommendations are already being carried out, expecially further experiments to confirm and extend'the results of the original screening that indicated 2 , 4 , 5 - T to be tera'togenie. . -. �In c o n s i d e r i n g the c h e m i s t r y of 2, 4, 5-T, our attention was d r a w n to impurities which can result f r o m the manufacturing process. Particular attention was f o c u s e d on a single impurity, 2 , 3 , 7 , 8 tetrachlorodibcnzo-p-dioxin, which o c c u r s in commercial p r e p a r a t i o n s of the . . . „ }J herbicide in highly variable amounts unlcs.s particular care is taken . . ..-. ... to exclude it. This impurity is'extremely toxic. Its p r e s e n c e depends upon variations in the reaction conditions. Other dioxins (some more toxic than 2, 3, 7, 8-tctrachlorodibenzo-p~dioxin) can be 'formed f r o m . various impurities in the starting materials. The dioxin impurity came . to particular attention when the U.S. herbicide industry was asked to . produce larger quantities of 2,4, 5-T during the middle 1960's. However, ^itsp r e s e n c e as an il'npurity and certain of its acute toxic e f f e c t s had been known since 1957. Its concentration in commercial 2, 4, 5-T has been greatly • .-reduced'"in the pa.st year or so. * Analytic methods available for 2 , 4 , 5-T*are accurate and reliable., > '•• With the possible exception of citrus f r u i t s , determination of residues ;.• in-food has not presented a serious .analytic-problem. . .'., .£..'-. 2, 4, 5-T is relatively labile in nature. Residues in soils and water are not persistent except under unusual conditions. The herbicide. • • is not stored in'plants or animals to a'significant extent. Production of 2, 4, 5-T in the United States rose rapidly between I960 and 1968, Civilian use, most of which is for clearing of range land and rights-of-way and for treatment of p a s t u r e s , declined about 50% ' •: between .1964 and 1966. Military use of 2, 4, 5-T as a defoliant, expressed :, as number of acres sprayed, rose sharply between 1964 and 1967 but has declined since then. Although it accounts for only a small amount of the ,- ;1 total usage of 2, 4, 5-T, its place in control of aquatic weeds is significant.. There is a small but important list of agricultural uses where 2, 4, 5-T is applied to food crops. Potential human exposure i's recognized in • .-. this direct application to food crops, in range and pasture lands grazed by domestic meat and dairy animals, and possibly, in water supplies derived f r o m treated waterways and streams. The economic importance of the various uses is considerable, but is very much less than that of 2, 4-D. Substitution of 2, 4-D for 2, 4,.5-T can be made for certain uses. Defoliation, using mixtures of 2, 4-D and 2, 4, 5-T, has been "employed in Vietnam since 1962 , ' more intensively since 1967. Although not rigorously demonstrated, its military usefulness has been considered to be very high. �- 3 - • The background of toxicological information on 2, 4, 5-T is thin. Most of the animal studies have been concerned with acute toxicity (single doses or repeated doses for short periods of time). Based on__tho s c cxpc rime n ts . the acute toxicity of 2, 4, 5-T was found to^ be low. . Little is. known 'of the details"of the metabolic handling of the material although rapid excretion in the u r i n e seems to be the rule. / I I 7 A / / The screening study supported by the I\T a t i onal^Cj-jTi eg r I nstitute on the toxicity of certain pesticides and other important industrial chemicals m a r k s an important advance in toxicological testing in that the tests w e r e d e s i g n e d to detect carcinogenic, teratogenic and mutagenic potential. TJh^jprj^2g.^a/dojri_of_ 2 ,' 4 . 5-T us^_,in_thQse_te-s_ts_v.^as shown r to be t e r a t o R c n i c in both j-n.t^ajifLnii.c e . There was no carcinogeiiic. While this study had a number of limitations which qualifTellTts u s e f u l n e s s , the teratogenic results we're sufficiently convincing '• • so that the Panel u r g e d , early in its discussions, that they be repeated and extended using better characterized preparations of 2, 4, 5-T, Analysis of a sample of the 2, 4, 5-T preparation u s e d in the original teratogenesis study revealed a dioxin level of about 27 pprn. _Such a considerable con-_ tamination by this highly toxic material r a jjgjsj-jh e^ques. t ion_a_s to wh ether ^ 2,_4_,_5^T itself, by the dioxin ___ impurity or by other imp_urities in. the commercial preparation testea. The Pa-nel v/as aware of press reports of increased birth ' defects in Vietnam attributed to the use of defo_liants. The lack of accurate epidemiological data on the incidence and kinds of birth defects in the Vietnamese population b e f o r e or since the military use of . defoliants precludes any estimate as to whether an increase in birth defects has occurred. Calculations of potential human exposures f r o m sources such as drinking water or direct fallout make this "appear unlikely (though theoretically possible). • " A r ej/i_ejA^j:>f_j^hj^^^ or p;anj. s ms_^jej/<iaj^__£ejar_ h_ar^mf ul c QHs_eo^u£nc e s_jif its r e c^mrnenQe d Indviced changes in vegetation are followed by alteration in numbers of wild animals. Accelerated crosioai' of 'soil may follow the killing of brush with herbicides but mechanical removal causes greater erosion. • . • . �- 4 B. Re c o m m on d a li o n s ' i' * * 1. Fxirthcr s t u d i e s . a . The animal e x p e r i m e n t s which r a i s e d the qv.c-stion of the te_ra_togcnic potential of 2,,4,5-T should be extended to include a. v.idgjr. rajTg_e of d o s e s ad minis tor cd to non-i'nbred strains of animals and to l a r g e r n u m b e r s 01 a n i m a l s . _ b . The i m p o r t a n c e of the impurities in 2 , 4 , 5 -T as potential health h a z a r d s s h o u 1 d b e a s c e r tain e d . R e c e n t experiments d e s i g n e d to distinguish between 2, 4, 5 - T and the dioxin impurity have s u g g e s t e d that both the herbicide and we dioxin are potential t e r a t o g e n s in some experimental animals. Hc>\vej/£r_.__£2\perirnents n e c e s s a r y to e s tab li s hthi^ arts \v er have Q i ^ X 3 » ~ - ' l c ^ as In addition, there may be additional impurities in commercially p r e p a r e d phenoxy herbicides which may be biologically active. c. The metabolism of Z, 4. 5-T in humans should be d e t e r m i n e d • and conaparcd to that in e x p e r i m e n t a l animals. '" " ' " ' I i * d. The level of dioxin, a recognized impurity in 2,4, 5-T should be rigorously controlled and -limited to not |more than 0. 5 ppm. A t reduction to not more than 0. 1 ppm should be urged. The family of ' compounds, known as polychlorinated dioxi'ns, have been found to be highly toxic and capable of eliciting teratog'enic e f f e c t s . The several dioxins vary widely in toxicity. F u r t h e r , it appears that they may arise in the environment f r o m multiple sources. Therefore, control over known sources should be exercised to the extent possible. • 2. R e s t r i c t i o n on use of 2, 4, 5-T. . • • A decision to r e s t r i c t the use of 2 , 4 , 5-T should not be bas ed on the isolated find ing of toxicity but on the expected e x p o s u r e follov.'ing recomrnencled u s e _ m relation to dose r e s ponse e f f e c t s . In general, the imposition of restrictions on the use of a pesticide would appear to be a function of two f a c t o r s , the potential for human exposure and the nature of the toxic e f f e c t s . For example, if c a r e f u l l y documented r e s ^d^ejinf_or_maj:^^ osu r e , j^e^j^k of ^acrvej^se eff ecjs__wc^d_b^lcss9'i gnificant than if exposvire were widespread. • ~~~ "™ ~~ ~ ' tha t s i gni,f ic ant r e_sidue s _ would result f r o m recommended u s e s of Z, 4, 5-T on food crops. It is �possible for residues to occur in tissues of animals grazing on recently treated pastures and range land. In fact, the only residues which have been identified in the total diet studies have o c c u r r e d in meat and dairy products. ' However, the few cases in which r e s i d u e s have been discovered have all been at levels well below those which would result in significant _ toxicity f o r man. . ' . - * r'"'^! ' The cxpg.rirnental finjding_of ter a to gene sis requires furthjvr c lab o r &ti o n 1 ve'f o <T-o- i t c a n be i n t e r p r e t e d as a human health h a z a r d . ; ^_ '/ • '. ' • . \ i The r i s k of t e r a t o g e n i c e f f e c t s should be placed in perspective. ' ' Teratogenesis i n d u c e d by chemicals is a f e t a l r e s p o n s e at a particularly sensitive p e r i o d in embryonic development to lower d o s e s of the chemical than .are acutely. to.xic_to the mother. Birth defects can be produced in j the embryo through many mechanisms of injury when the agents are ad' : ministered during-critical periods of organogenesis. ' It is generally • ' •• held that by c a r e f u l choice of dosage, which may be close to the acutely toxic dose for the mother, most chemicals might be 'shown to be teratogenic in animals. For a variety of r e a s o n s , it is not possible to . : translate directly 'the results of experiments in animals to man. There • are differences in sensitivity which arise f r o m differences in metabolism. Comparative metabolic studies in man and animals, therefore, are important in interpreting toxicity for 'man. . • The important consideration is not only the demonstration of teratogenicity, which" may occur with many chemicals'at selected dosages, but the estimation' of the likelihood of teratogenic effects with the ' amounts likely to be ingested incident to recommended use. To restrict or ban usage of chemicals on the basis of demonstration °^ teratg_genicity at do_se_Leyels which far ex_cjij(id_jui£lial_c1r__e^j:)ected exposures is u n ~ ^rea^onab le_an d p nu 1 d_well d e n y u s a g e o f chemicals whose benefits far outweigh risks. �-63. R e g i s t r a t i o n s of 2 , 4 , 5-T for u s e s on p a s t u r e s and r a n g e l a n d s should be t r e a t e d as r e g i s t r a t i o n s for food c r o p u s e s . It is possible for r e s i d u e s of 2, 4, 5-T to occur in milk and tissues . o f animals g r a z i n g 'on land recently t r e a t e d with 2, 4, 5-T. To date, meat and dairy p r o d u c t s have been the only food products in'the total diet studies that contained measurable amounts of 2, 4, 5-T. Use on r a n g e and p a s t u r e land should be included in registration for use onfood crops. - 4- M o n i t o r i n g of 2 , 4 , 5-T r e s i d u e s should be significantly expanded^ e s p e c i a l l y f o r meat and milk. In sampling meat and milk, special attention should be given to geographic a r e a s where, treatment of pastures and r a n g e lands with 2 , 4 , 5-T is most common. The 2, 4, 5-T r e s i d u e s that may o c c u r in meat and milk of animals a.llpwed to graze on p a s t u r e s and ra_nge lands t r e a t e d according to current recommenda- • tions' should also be r e s t u d i e d . ' • f 5. Prompt and expeditious r e l e a s e of information. As new information is developed on pesticides,• it should be d i s s e m i n a t e d promptly to individuals and or sanitations that a r e legitimately c o n c e r n e d as m a n u f a c t u r e r s , f o r m u l a t o r s u s e r s and s c i e nti f i c i nv e s ti g a t o r s . ' " The case of 2, 4, 5-T is illustrative of inordinate delay.in makingavailable new r e s e a r c h i n f o r m a t i o n as it became known. The screening study of pesticides which was c a r r i e d out by Bionetics Corporation under contract wi-th the National Cancer Institute was completed about August 1968. It was 14 months later when the Government announced its coordinated actions on restricting the use'of 2,4, 5-T. It-was only a f t e r an additional s e v e r a l months that the detailed data of the screening study were made publicly available. A centralized.mechanism for handling and disseminating-new information about'pesticides could help alleviate this problem. • . . ^• Interim imposition of r e s t r i c t i o n s on pesticide use while additional i n f o r m a t i o n is being c o l l e c t e d . A mechanism should be e s t a b l i s h e d for r e s t r i c t i n g the u s e _of_a_ r e g i s t e r e d pesticide ternporarily_p_n the basis of information which implicates the c h e m i c a l as a_possibJ_e_JT£jij.t}i_ha.zard pending the collection o f rn or e d c f i n i ti v e i n f o r _ma id o n. If a p e s t i c i d e is a l r e a d y in established use, the decision is! p a r t i c u l a r l y difficult. Long established use inevitably implies a d e p e n d e n c e upon it by the consumer and a c o r r e s p o n d i n g reluctance by the m a n u f a c t u r e r to withdraw it f r o m the market. . .. �At the present time, a registration may be held in abeyance only by cancellation or suspension. .Because of the serious import of these actions they are put into e f f e c t with considerable reluctance. They \vere_not designed for situations such as the present with respect to 2 4 5-T where te m PO °"5e' or s us pc n sion p f r o gist ra ti on_,rm_gh_t_ha 1 ve b c e n a^mo r e _ajjpj;o pr i a t e a c t i o n . Such an alternative c o u r s e of action is not possible under p r e s e n t regulations. There is need for a mechanism w h e r e b y the use of a pesticide or other chemical that may a f f e c t human health can be t e m p o r a r i l y r e s t r i c t e d or held in abeyance. Such action would permit the gathering ' • of m o r e . d e f i n i t i v e information in time for sufficient consultation to permit a decision that would protect the public health and not impose an undue economic burclenrcm the p r o d u c e r s , m a r k e t e r s , and u s e r s of a product. Coincident with the imposition of r e s t r i c t i o n s on a pesticide, a mechanism should be available for informing and educating pesticide u s e r s and applicators so as to make them more r e s p o n s i b l e agents. 7. There is an xir.gent need for a_focus of responsibility in D / H E W ' to coordinate and monitor the toxicitv 'and health activities related to • ———•--—• -- ---effects of pesticides. \ } i .i Information about the health effects of a pesticide derives from a variety of sources including occupational exposures, residue monitoring, toxicological investigations, clinical experience and epidemiological studies. In the past, there has' been no single focus within D / H E W which has been c o n c e r n e d with all of these s o u r c e s of data and, more important, which has had the authority and responsibility to coordinate new invest!\ gative initiatives. The new Advisory Committee on" Pesticides to the \ Secretary of D / H E W can be expected to serve as a source of e x p e r t advice but cannot fill the essential need for a focus of responsibility and authority at the level of the Office of the Secretary. Consideration of the functions to be fulfilled and the r e s o u r c e s available s u g g e s t s that this responsibility should be assigned to the A s s i s t a n t S e c r e t a r y for Health and Scientific A f f a i r s because the various components of D / H E W ' concerned most with aspects of the health e f f e c t s of pesticides r e p o r t directly to him (National Institutes of Health, Food and D r u g Administration, National Communicable D i s e a s e Center, Environmental Health Service). 8. • Information provided in applications for registrations of pesticides^ should take into account not only the PejUcidc for which registration is^ s ought but E; hould id cntify other substanc e s inc. luding vehicles used in i o r m u 1 a ti o n s . V i . n e r t n i n g r p d i e n t s , a n d i m p u r i t i c_3 . �o- Investigation of the synthesis of 2, 4, 5-T and examination of the m a n u f a c t u r i n g p r o c e s s r e v e a l e d that an e x t r e m e l y toxic impurity, 2, 3, 7, S - t e t r a c h l o r d d i b e n z o - p - d i o x i n , p r e s e n t in variable .amount in commercial preparations of 2, 4, 5-T, may account for 'some of the toxicological characteristics assigned to 2, 4, 5-T itself . The presence of this impurity was r e c o g n i z e d as early as 1957. However, the importance of this impurity was not generally r e c o g n i z e d in the United States until after 1964. It a p p e a r s logical that g r e a t e r specificity in identifying the components and properties of thej. mixture of materials which areregistered u n d e r - a single name would increase the probability of identification of potentially toxic substances. • p r o c e d u r e s should include toxicological studies on r e l a t e d c o m p o u n d s where j u d g m e n t i n d i c a t e s this to be desirable. • "~~ \ « Toxicological studies provided as information in behalf of • 2, 4, 5-T registration were performed on a Variety of related compounds (the f r e e acid, several types of e s t e r s and 3. variety of salts). Results of these, tests w e r e r e g a r d e d as being interchangeable and applicable to the related compounds. There is evidence to caution against this concept; For example, the National Cancer Institute study indicated carcinogenic propensities of some but not all 2 , 4 - D esters. The thorough 1 testing of i s o m e r s , e s t e r s , s a l t s j and 'related compounds is a very large and expensive task. Nevertheless, information about a potential health h a z a r d may be incomplete unless all of the compounds 1 to be used are tested. . . . �II. INTRODUCTION In 1964, The National Cancer Institute "undertook through a -contract a s c r e e n i n g ' s t u d y " of a number of pesticidal chemicals. . Among the r e s u l t s of this s c r e e n i n g study was the .finding that birth defects could be provoked experimentally in rats and mice by the • administration of relatively large doses of the herbicide, 2 , 4 , 5-T. By the time t h e a d results were reported, 2, 4, 5-T had been in common "use as'Sn herbicrd-e for more than 20 years. Further, it had been . employed along v.ith 2, 4-D as a defoliant in Vietnam since 1962 although in sizeable quantities only since 1967. In October. 1969 several agencies of Government moved in a coordinatedjocuiainer to bring about r e s t r i c t i o n of th'e use of 2 , 4 , 5-T both within the United States and abroad,, >. In terms of domestic agricultural use, restriction was placed on the use of 2, 4, 5-T on food crops pending the acquisition of f u r t h e r information that might permit the Food 'and D r u g Administration to grant a t o l e r a n c e . Use as a defoliant in Vietnam was restricted to non-populated areas. •For a number of reasons, it seemed wise to explore this issue in some detail. . The most important of these r e a s o n s , p e r h a p s , was the d e s i r e to examine the scientific evidence available to stand behind future policies governing the use of 2, 4, 5-T and to suggest 'directions for f u r t h e r experimental research. Accordingly, a panel of experts was assembled by the P r e s i d e n t ' s Science A d v i s e r to consider a. number of a s p e c t s of a variety of herbicides some of . which were used as defoliants in Vietnam. The present report represents~their review of 2, 4, 5-T. ' This review considers topics which are of concern to those who are f a c e d with policy decisions for 2, 4, 5-T. We hope that it can serve as an example for the consideration of the health effects and safety of pesticides and other chemical's purposefully placed in the environment. A number of issues are raised when the utility and safety of an already existing material is questioned. It is elementary but nevertheless t r u e to say that the issues are complex. In a way, their examination can b e - c o m p a r e d to following a seemingly endless and continuously branching program. The subject of how much assurance of safety should be afforded is important. �11-2 Teratogcncsis a p p e a r s to be a m o r e m a n a g e a b l e problem than some other health effects (such.as tumor production). Prediction of safety can p r o b a b l y be made with reasonable a s s u r a n c e . In addition, e x p e r i m e n t s to test a suspected substance are r e a s o n ably straightforward to conduct. % • r Among other problems, the purity of the chemical became an issue. In the case of 2, 4, 5-T, separation of biological e f f e c t s of the principal m a t e r i a l f r o m those of the impurities t u r n e d out to have unusual importance. For this reason, the resolutio.n and accuracy of analytic methods available and used to detect 2, 4, 5-T and its impurities had to be evaluated. In_a_s certain ing the probabj._e hajtjard to-jnar. of a n _ a g r i c u l t u r a l chcmical, its toxicity in absolute terms must be r e l a t e d to the probabilities of human exposure. Residue information on 2,4, 5-T therefore was explored. Finally, there remains a series of policy questions which are at least as philosophical as they are technical, the most crucial one being how wide a margin of safety should a society adopt for itself. The panel also touched on a n a r r o w e r aspect of-this question by posing an additional one. This is the practical problem, in the case of a material already in use, of how the Government should a c t . in the interim between the time of acquisition of preliminary experimental data which reveal a chemical suspect and the performance of more definitive experiments which establish the risk.. �II-3 REFERENCES (1) Press Release on 2, 4, 5-T. Office of Science and Technology,''October 29. 1969. I \. �SECTION i l l CHEMISTRY A'. Synthesis B. Formulation C. Solubilities D. Purity of Technical Grade Material E. . .' - • . ' Analytic Methods 1. Standard Procedures 2. Bound Residues F. Chemical Stability of 2, 4 , 5 - T ' 1. • • ..- ._ . Photochemical Degradation- 2. : Hydrolysis 3. Thermal Stability 4. Biochemical Degradation �' S U M M A R Y ^ - Section III" - Chemistry 2 , 4 , 5 - t r i c h l o r o p h e n o x y acetic acid is produced commercially by a process which begins with tetrachlorobenzene as starting material. Technical g r a d e 2, 4, 5-T is 90-92% pure acid. One of the • important i m p u r i t i e s , a polychlorinated dioxin, results both f r o m impurities in the s t a r t i n g material and as side products of the desired reaction. A large number of e s t e r s and amine salts of 2,4, 5-T have been developed as well as a variety of formulations in o r d e r to derive specific p r o p e r t i e s of volatility and solubility. About one-half of this total production can be accounted for by the 2, 4, 5-T acid and its n-butyl ester. The f r e e acid is practically insoluble in water and, •generally, the e s t e r s are slightly soluble. The amine salts tend to • be more soluble. Among other things, penetration into the soil or leaching is a.function of water solubility. In general, 2, 4, 5-T residues can be expected to be relatively unstable materials in the environment. They are broken down by microbial action and by sunlight and esters are readily hydrolyzed to the f r e e acid. Available analytic methods for detecting residues are quite sensitive. The sensitivity o f - g a s chromatography with microcoulometric detection is about 0. 01 ppm (10 ppb). Detection of residues in plant material appears to be uncomplicated by bound or complexed residues with; the possible exception of residues in the peels of citrus fruits. • � 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. 5557 Series The series number of the original item. Series VIII Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource MacLeod, Colin M. et al. Description An account of the resource <strong>Corporate Author: </strong>Panel on Herbicides, President's Science Advisory Committee Date A point or period of time associated with an event in the lifecycle of the resource 1970-07-01 Title A name given to the resource Report on 2,4,5-T 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 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. 175 Folder The folder containing the original item. 5165 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 Schecter, Arnold J. John J. Ryan et al. Title A name given to the resource Chlorinated Dibenzodioxins and Dibenzofurans in Human Adipose Tissue from Exposed and Control Patients 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. 175 Folder The folder containing the original item. 5163 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 Weerasinghe, N.C.A. et al. Title A name given to the resource Levels of 2,3,7,8- Tetrachlorodibenzo -p- Dioxin (2,3,7,8 - TCDD) in Adipose Tissue of U.S. Vietnam Veterans Seeking Medical Assistance 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. 175 Folder The folder containing the original item. 5154 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 Rappe, Christoffer et al. Title A name given to the resource Identification of 2,3,7,8 -Substituted Ploychlorinated Dioxins (PCDDs) and Dibenzofurans (PCDFs) in Environmental and Human Samples 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. 5009 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 Ghassemi, Masood et al. Description An account of the resource <strong>Corporate Author: </strong>TRW Environmental Division, Redondo Beach, CA Date A point or period of time associated with an event in the lifecycle of the resource 1981-12-01 Title A name given to the resource Environmental Fates and Impacts of Major Forest Use Pesticides 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. 154 Folder The folder containing the original item. 4574 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 Menges, Robert W. et al. Source A related resource from which the described resource is derived Veterinary Medicine Date A point or period of time associated with an event in the lifecycle of the resource 1970-06-01 Title A name given to the resource Arsanilic Acid Blindness in Pigs 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. 150 Folder The folder containing the original item. 4355 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 Levander, Orville A. et al. Description An account of the resource <strong>Corporate Author: </strong>Committee on Medical and Biologic Effects of Environmental Pollutants Source A related resource from which the described resource is derived Arsenic Date A point or period of time associated with an event in the lifecycle of the resource 1977 Title A name given to the resource Chapter 6: Biologic Effects of Arsenic on Man 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. 144 Folder The folder containing the original item. 4146 Series The series number of the original item. Series VI Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Facchetti, S. F. Geiss et al. Date A point or period of time associated with an event in the lifecycle of the resource 1982-07-01 Title A name given to the resource Isotopic Lead Experiment: Status Report 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. 129 Folder The folder containing the original item. 3653 Series The series number of the original item. Series VI Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Hluchan, E. et al. Source A related resource from which the described resource is derived Ceskoslovenska Hygiena Date A point or period of time associated with an event in the lifecycle of the resource 1968 Title A name given to the resource Presence of Arsenic Compounds in the Surroundings of an Aluminium Plant 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. 120 Folder The folder containing the original item. 3246 Series The series number of the original item. Series VI Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource di Domenico, A. et al. Source A related resource from which the described resource is derived Analytical Chemistry Date A point or period of time associated with an event in the lifecycle of the resource 1979-05-01 Title A name given to the resource Analytical Techniques for 2,3,7,8 -Tetrachlorodibenzo -p- dioxin Detection in Environmental Samples after the Industrial Accident at Seveso 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. 119 Folder The folder containing the original item. 3200 Series The series number of the original item. Series VI Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Camoni, I et al. Title A name given to the resource Analytical Techniques for 2,3,7,8-Tetrachlorodibenzo-p-dioxin Detection in Environmental Samples After the Industrial Accident at Seveso 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. 118 Folder The folder containing the original item. 3141 Series The series number of the original item. Series VI Subseries II Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Adamoli, P et al. Source A related resource from which the described resource is derived Ecological Bulletins Date A point or period of time associated with an event in the lifecycle of the resource 1978 Title A name given to the resource Analysis of 2,3,7,8-Tetrachlorodibenzo-para-dioxin in the Seveso Area 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. 111 Folder The folder containing the original item. 3077 Series The series number of the original item. Series V 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 Latalski, M. et al. Source A related resource from which the described resource is derived Annales Universitatis Mariae Curie-Sklodowska Lublin-Polonia Date A point or period of time associated with an event in the lifecycle of the resource 1980 Title A name given to the resource An Effect of Acute Poisoning with Herbicide 2,4-D on the Granulogenesis in Eosinophiles