Document ymbod3bxJKadNZo9XvMvBR62r

n0 t iM-a NMintcM. CMtMsntv SEP 1981 nppe | Regulatory history and wDq | Analytical Problems Syntbaaiaof pofycbtomatad biphanyla (PCBa) naa flrat daaeribad by Schmidt and Scbull* <i> in INI. but tha induatrial apphcaUoa of that dineowary nan not fbtlly daoalapad until SO yim inter by tba Swan Carpora- PCBi aa induathal Hwicala far oaa iwclion nilh ibcrtdil twfamm, 1990. Chlorination of i)m patant meiaty, biphenyl, raauha in a aariaa of chlorinated produelo nbiah oahibit uninua charactarintka ofbath tbnrnal and chemical atebUKy. Tfcaeo odituraa of PCBa (author than indlriduri iacuinaai quietly pained aide atcap- Prom the nritiaJ tntraductioe af PCBa m iidwiriii chemioale in 1990, (hair widaapraad popularity in uncontnilad appikaliona (bundrada of bjIItooa of pouada) ant the anauing ao yaaia filially raaullad in their bacoa- inc a nataiataM and ubaouitoua anvi- mawnaatfal eontoainaal. In 1999, Jan aan (2). a Snadiab erientiet. naa ihr drat to draw alianlioa to tbc Iart that PCBa naan found in flab and bndn Hv lata 1971. it had bacoaw gtoinaiy ubvioua that ceatamiaelMo af tha awn. ronaunt bad lad la gMoi contannno Uan of ntldbfa >. In laaancn. K'Ba rmhk Afodaa, (Septan. Phanadar, Kaula, and Phyralewe. to Naa jwt a few af the major product Unae. la theory. tba total number ofpaaaibia pruducm laanUtaa from ehioriaaUan aTMahcnytie * (rtfwa I). Mapad In 13S inam to tha deaahr related pelyrhltrinated toiainfana (PCDFblpraaMt aa fcapurMao eommardal PCBa. 11m phyniaal pray- ftront from thaneafindlridual toecna.Arecieranil! aet react with adda.albatt, or neto under normal u pemtt'toArad tollfrchiarhmlodyoaducttoahaiet 490 *C to 99% ihieriMiid pradnet PCBa can ba dlaUPad al atoeapbaric praeourt nithoal carbantoticu or da- HONS 002271 laining PCBs. In 1972. this survey cul minated in a "Notice of Proposed Rule Making'* (5) to limit the level of PCBe in food* containing unavoidable PCB residues from environmental or industrial sources. Analysis of animal feed indicated less than 5% of the samples contained PCBs (0-0.6 ppm). Isolated accidents, however, had oc curred in pasteurization equipment, where PCBs (torn heat exchanger fluids had directly contaminated animpl feed end subsequently poultry and eggs intended for human con sumption. The use uf PCB-containing coatings on the inner walls of silos had also resulted in the contamination of silage, which had in turn caused PCB residues to appear in the milk of dairy cows (6). In a survey on food peckaging. 67% of the packaging samples contained PCBs, with the highest level observed being 338 ppm. Of these samples, only 19% of the actual food contained in the packaging contained PCBs, with s maximum level of 0.1 ppm. In the caee of pockaged infant food cereals, 75% of the samples contained PCBs, with an average concentration of 0.3 ppm; the maximum found was 1 ppm. At the time of this survey, knowledge of the toxicological effects of PCBs was somewhat limited. The FDA conclud ed that it would be prudent to reduce human iow-level exposure to PCBs by limiting the ways in which PCBs might enter the food chain, and to limit the levels of PCBs in food con taining unavoidable PCB residues from environmental or industrial sources by establishing temporary tol erances (Table I) until such time as additional information warranted change (7). The temporary tolerances estab lished in 1973 ware a result of a vigor ous evaluation of the date baac on PCBs prevailing at that time. Concern about the ubiquitous distribution of PCBs in the ecosystem end the "Yueho" accident led to a maseive sci entific probe into the toxicity of PCBe, as well ae their reported pres ence in moet wildlife (8). In 1975, the Environmental Protection Agency (EPA) sponeored e national confer ence on PCBs, at which the FDA an nounced it had initiated review of the appropriateness of the 1973 tem porary tolerances. After reviewing the wealth of data generated by the scien tific community and government on PCBs, FDA proponed to reduce the temporary tolerances (9) for unavoid able residue* of PCB* in several desses of food. More than 100 cornmenu on this proposal were received from interested parties. In considering these commenU, the FDA commis sioner issued a final order reducing PCB tolerances (Table I) (10). Since the food and feed industries of PCBs in human fat for both males were particularly vulnerable to acci and females from Tokyo within the dental contamination from PCBs, range 1-2 ppm. while Mood samples EPA, which has regulatory control of contained 1-5 ppb PCBs. Recently, PCBs and other toxic substances the eloquent demonstration by Poland under the Toxic Substance Control (14) that the toxicity of certain PCB Act of 1976, has issued rules governing congeners together with other polycy their continued use. As of Nov. 1, clic aromatics (PCDFs, ssobenxenes. 1979, the use of PCBe in new heat end TCDD) paralleled their ability to transfer syeUms in plants manufac compete (in vitro) for specific binding turing or processing food, drugs, and sites in the cytosol carrier prtAeina has cosmetics was no longer authorised. provided e useful tool for structure- The EPA rule, however, did permit ectivity and mechanistic studies. the continued use of PCBe in electro To complicate the issues, the wealth magnets, transformers, heat transfer of data concerning the metabolism uf and hydraulic systems until July 1, PCBs has given rise to Mother set of 1964. Issuance of an interagency alert analytical problems. It is now clearly notice was prepared under the spon established that the major metabolic sorship of EPA to urge voluntary re route of PCBe is hydrosyletion. Me moval of equipment and its replace tabolism in rat microsomal systems ment with non-PCB units to prevent (15) of selected tri-, tetrs-, end penta- food contamination (11). chiorobiphenyls resulted primarily in Toxictty Evaluation of the toxicity of PCBs the formation of monohydroxylated PCBs with the number of metabolites decreasing with inrraaaing degree of has been complicated by the heteragr-- chlorination. Similar studies with neity of the commercial preparations PCBs in the rhesus monkey < 16) have and the analytical difficulties in sepa ration, identification and quantita tion. Acute oral LDao values of PCBs in mammals vary from 2-10 g/kg, with an apparent increase in mammalian toxicity with a decrease in chlorine confirmed this pathway, and also demonstrated dechlorination and arena oxide formation. Four lactating Holstein cows fad single doaea of Aroclors 1242 and 1254 (1.5 g) excreted in their milk 10 monohydroxy PCB* end content In humane, chronic ingestion ' four monohydroxy PCBs, respectively of small amount* (10 mg/kg) for more (17). Jensen et el. < 18) have also re than 50 days cataes chloracne. The ported in study of seal Mubber that fear that prolonged, continuous expo sure to low doees might well result in serious human health problems has promoted a maesive scientific inquiry methyl suifone metabolism is clearly evident The seal blubber contained approximately 150 ppm total PCB*. with 16 ppm of total PCB suifone*. into the toxicological and biological ef fects of PCBs. A recent study, involv ing the feeding of male broiler chick ens (12) low levels of Aroclor 1254 The concentration of PCBe stored in the fat tissue increases sad is then transferred st increasing concentra tions to the food chain. from hatching to eight weeks of ege. demonstrated marked reduction in Analytical ConaMaraffona growth with concurrent accumulation It should now be dear that any ai of PCBs in all tissues. Fukano at al. tempt to extract, separate, identify, (13) have illustrated the accumulation and quentifv a mixture of PCBs re- 11S4A ANALYTICAC CHEMISTRY, VOL. 93, NO. 11. SEPTEMBER 1091 MGNS 002272 quires'lmowledge of three major prob lem areas: Conpoiitlos. in spite of theoretical calculations establishing 209 poeeible congeners, Welti (19) has demon strated experimentally via capillary chromatography that not all possible compounds are present in Aroclor 1254. Of the 69 eluting peak profiles observed, 19 represented major componente while 50 represented minor contributions to the mixture. Certain positions on the biphenyl nucleus ere not favored for chlorine substitution; 2-, 4 and 6- substitutions are rare, end the occurrence of 3-, 3,5-, end 2,3isomers is likewise infrequent Parallel studies using picked columns have shown elution profiles with fewer than 20 discernible peaks. The complica tions induced by analyzing mixtures of PCBs rather than any single specific isomer have contributed a serious im pediment to quantitation, since most analyses ultimately involve estimation of compounds in which the degree of chlorination can run from mono- to decs-, i.e., a range of molecular weights (186-494) with varying chemi cal and physical properties. Interfering Organochlorine Pes ticides. The initial confirmation of the occurrence of PCB residues in wildlife by Jensen (2) was reported after repeated and frequent encoun ters of similar elution patterns or pro files while analyzing for DDT and other chlorinated pesticides in gener al. Many earlier reported residue anal yses foiled to recognize this PCB in- 6 ft X 2 mm, 200 C 99, NO. 11. SEPTEMBER 1*91 MOWS 002273 terference, which muHt have led to an overeslimnlion of DDT in the environ ment. After thin historical break through in 1966, the roles were sud denly reversed and the important l'CB residue analyses were then deemed to have interferences by a whole host of common and ubiquitous organochlorine pesticides. Methodolo gies were then devised to efficiently separata PCBs from such organnchlorine pesticides as DDT and Toxaphene, Heptachlorepoxide, 1 andane, and Dieldrin. The stability of PCBs to treatment with acids and alkali made it poeaibie in some procedures to ei ther destroy or alter one or more of the interfering pesticides. On the other hand, greater attention has been paid to the use of various columns (e.g. silicic aod-celite and Florisi!) and soivanti (mixtures ofacetonitrile, hexane, and methylene chloride) In separata PCBs from lh other chlori noted pesticides (2W). Masomoto 12/) has pointed out, however, that such procedures can be flawed by variable column preparation and hence vari able recoveries. Additionally, those congeners with the lowest chlorination were held on the silicic column and could only be eluted by a more |xilar solvent than petroleum ether, leading to lower recoveries for the less chlori nated Aroclors. The analyst must "choose appropriate analytical prore dures, depending on which interfer ences are present . Metabolism. The elution pattern* or profiles observed by electron cap Lure (EC) gas chromatography lor PCB residues from biological samples (22) very often do not resemble the profile for one of the PCB standards selected for quantitation. The pres ence of metabolites and degradation products is likely to he primarily re sponsible for this experimental obser valion. Dechlorination, arene oxide formation, and hydroxylation have been identified as metabolic pathways for PCBs (23). Dechlorination to less er chlorinated congeners will tend to contribute to the production of a pro file perhaps suggestive of a less chlori nated Aroclor and will lead to confu sion in the choke of the correct Arm* lor standard for quantitation by K(\ More serious by far is the removal from the observed PCB residue profile of congeners that have undergone Ftfwe Electron capture responses for 1 no injected ol selected PCBs: 2% OV101,6 ft X 2 mm. 180 C, attenuation 16 x 10"11 ampt 1196 A ANALYTICAL CHEMISTRY. VOL. S3. NO. 11. SEPTEMBER 1901 MOWS 0022 74 \xii / iMtoMic hydmyiiUon. Such chloro- biphinyloh do not exhibit the same eh*iot> charnctarbtica as the parent euhatanc--,and therefore do not coe lute with the solvent lyeteme generally Metabolic processes can thus con fute the assignment of the original own of the PCB contamination and hence the correct choice of a euitable etandard for quantitation. Such PCB residue profiles are further complicat ed by the fact that extraction and cleanup can afcsr the compoeition of mixture! becehee of the unique prop ortion of those congeners. ApprochM It Quantitation The problems encountered in the quantitation of PCB raoiduae have focuatd the attention of the analytical chemist on methods by which the con centration of PCBs is related to the estimation of single species. Such at tempts have included: carbon-skele ton GC, where PCBe are reduced oncolumn (&% Pi at 180 *C with Hs as carrier gas) to biphenyl (24): perchlorinatkm of PCB* by antimony pentachlorkla to yield dacachlorobiphenyl (35); and microcouiometric determi nation of the total chlorine content as HCI (85). These methodologies have bean developed because EC-GC detec tor response la disproportionate, i.e., highly dependent on the degree of chlorination of the biphenyl nucleus. EC detector response has bean shown to vary aa much as two to four orders of magnitude between monochioro and polycWoro species (37). The pro liferation of analytical approaches to PCB analysis has resulted mainly from the existence of this dbproportKmelity factor (38). This issue has swtrshadcwad other problem*, such aa interferences resulting from the pres ides of other organochlorid* pesti cides and PCB metabolites. Of prims importance when using EC as Um detection system for chlori nated aromatic speciss such as the PCBe Is the problem of disproportion ate rsaponsss At the present time, Arador standards have bean charac terised by EC (Figure 2) and are em ployed as reference materials. TTm daasat matchingprofile is used aa a standard for quantitation of environ mental samples. A good example of "ep^P ittt.y of ves.ponae b.y EC is that exhibited by the Isoroeric-mo- usahlarolohianm (Figure 3), in which dMUsrencea of several orders of magni tude ware observed. Although this ease b a relatively dramatic example, Uw prevailing situation within the congeners of the PCB family has been found to be similar but somewhat leas pfwumiad. The EC responses to 1 ng of five selected dichlorobiphenyb under similar conditions are ilhis- ^ i \cc^ trated in Figure 4. Depending on the (9) Fed. Regie., 42, 17467 (April 1.1977) location of the chlorine atoms within the biphenyl nucleus, varying detector responses ware observed. This dis (10) Fed. n"ctu., 4* 4` ,38330 (Jum 29.1979). 11) Fed. Regis.. 46,30990 (May 9.1980) 12 F. H Bird. C. B Chawan. and R W. Genty, Poult Set., 17,538 (1978). turbing dbpropurtkmality has bean (13) S. Fukano and M. Doguchi, Bud. Kn- the main criticism of PCB analysis with EC. viron. Contain. Toxicol., 17,81.1 (I977>. (14) W. F. Greenlee and A. Poland, J. H*oi Chcm., 284,9814 (1979). The analytical pitfalb created by (15) A. M. Ghbauddin, R. B. Msnser. and the lack of well-defined standards when dealing with mixtures has ulti mately led to the increased applica J. O. Nslaen, Toxicol. Appl. Pharmvcnl, 98,187 (1978). (18) D. H. Norback,J. Britt, M. T. Haia. and J. R. Allan, Fed. Prot., 37,299 tion of mast spectrometry in its vari (1978). ous forms to assist in resolving diffi (17) A. M. Gardner, H. P. Rightor, and l. cult problems concerning identifica tion and quantitation of these sys tems. A novel approach to routine A. Roach,*/. Assoc. Off. AnaL Chcm., SO, 273(1978). (18) S. Jensen and B. Jansson, Ambiu , S, 267(1978).. quantitative analysis by chemical ion (19) D. Simone and D. Wslti, J Chroma- isation mass spectrometry b described In the research section of this issue (89). togr, 10,15(1971). (20) J. Amour and J. Burke, J. Asaoc Off Anal. Chem., S3,761 (1970) (21) H. T. Maaumoto, J. Attoc Off. Anal. Chcm., IS, 1092 (1972). (22) J. Tetchman, A. Bevenue, and J. W. (1) H. Schmidt and G. Schultz, Ann. Hyiin,*/. Chrematogr., 161.156(t978). Cheat., JOT,338(1881). (23) J. Sparling, D. Fung, end 8. Safe. (2) S. Jensen, New Set., 32,812 (1966). Biomed. Mass Spectre*., 7,13 (1980). (3) J. H. KOsman, M. C. Debrauw, and (24) M. Cooke, G. Nicklsss, A. M. Prescott, R. H. DeVos, Nature, 221,1126 (I960). end D. J. Roberta, J. Chromatogr, III, (4) S. Ssski, A. Tsutsui, K. Oguri, H. Yoshimurs, sad M. Haraana, Fukuoka 293(1979). (25) A. L. Robbins and C. R. WiUhite. Attn Mod., 02,90 (1971). Bud. Environ Contain. Toxicol., 2t, 42H . (5) Frd. grp*.. 07.64 (March 18,1972). (1979). (6) R. P. Skrsntny, R. W. Hemken, and (26) P. Donkin, S. V. Mann, and & 1. H.W. Dorough, Bull. Environ. Contain. Hamilton, AnaL Chcm. Acta, 88,289 Toxicol., 0,409(1971). (1977). (7) Fed. Refit.. SO, 18008 (July 6.1973). (27) J. W. Rote and P. G. Murphy, Hull (8) 8. G. Winston and H. B. Gerstner, Environ Contam. Toxicol, 4, .777 '`Polychlorinated Biphenyls, Polybrotnh--~ 0971). ........... natod Biphenyls ana Their Contami (28) T. S. Krull, Residue Rev , 00,186 nants--A Literature Compilation, (1977). 1986-1977," ORNL/TIRC--78/2, Natu (29) T. Cairns and E. G- Siegmund, Anal. re Library of Madicine, March 1978. Chcm., S3,1599 (1981). Thomat Cairns, instrumentation npc! cialist in maxi spectrometry far the FDA since 1975, received his PhD de gree in chemical spectroscopy from the L/niversify of Glasgow, Scotland, in 1965. His strong interest in analyt ical toxicology developed during his tenure as acting director of the Na tional Center for Toxicological Re search, Jefferson, Ark., from 1977 to 1960. His present research is mainly concerned with pesticide residues, drug metabolism, and the btoavailability of glucocorticoids. Emil G. Siegmund (seated), research chemist in mass spectrometry with the Los Angeles District Office of the FDA since 1963, received his AS de gree in chemistry from the University of Massachusetts in 1969. His princi pal activities involve the application of GC/MS and LC/MS to regulatory I analysis in the identification of sus pected pesticide residues, as well as the detection of unregistered pesti cides, particularly on imported fruit and vegetables. He is currently in volved in developing analytical assay procedures for various glucocorti coids. HONS 002275 ANALYTICAL CHEMISTRY. VOL