Document 2RYo5pbb3XbrE3VaGzd8x54Na

7- s A/r,. C O R P O R A T E O/y Jflh T ' *4yr* RESEARCH A.J U tr*jc GENERAL FLcCTRIC COMPANY. RESEARCH ANC DEVELOPMENT CENTER. ? O. O* il SCHENECTADY, NEW YORK 12301, Phon* (516) 3^5-2211 DEVSLOPMSNT If.. Building K-l, Room 3B35 h f . W\ January 15, 1979 Irving J. Selikoff, M.D. Professor of Community Medicine and Director - Environmental Sciences Lab Mt. Sinai School of Medicine 5th Avenue and 100th Street New York, New York 10029 vT . Subject: Analysis of PCB's for PCDF 's Dear Dr. Selikoff, The attached copy of my latest letter to Jim McKinney presents answers to some of the concerns you addressed in your letter of i December 19. As best I can determine, what actually happened down there was that The PCB samples got passed down the chain of command to a technicia who determined the molecular weights of a large number of cons; \tuents, got out one of those mass spectrometry tables that list several dozen reasonably common chemicals corresponding to each molecular weight value, made some selections that he thought plausible, and passed his report back up the chain. McKinney, with out realizing that the chemical identifications were purely suggc-a- tive rather than defin.if , got quite excited over their chemical, 'en vironmental, and medical implications, and wrote some extensive ys on the subject which he was understandably reluctant to swallr/w pletely after learning of the absence of supporting data. As best we can make out from the fairly extensive pile of raw data and in ternal reports he sent us, however, the only definite conclusion" that can be drawn at this time are these: 1. The "used transformer oil" sample analysed has the gross composition of a 1953-63 era transformer dielectric, which was a blend of commercial grade trichlorobenzene, tetrachlorobenzene, and Aroclor 1260 with a little tetraphenyl tin as a stabilizer. 2. The sample contained trace levels of a considerable num ber of other materials, probably mostly hydrocarbons, phthalatos, plasticizers, etc., along with a family of C ^ ^ i o - x ^ x 0 species which could represent either polychlorinated diphenylethers, p o ! chlorinated biphenols, or perhaps something else completely. Th GENP 010332 783143 Irving j. Selikoff - 2- January 15, 1979 is not'yet any analytical evidence to support the totally unpre cedented chlorine redistribution, oxidative coupling of chloro benzenes, or low temperature PCB oxidation which was hypothesized. 3. The sample did not contain significant levels of terphenyls or quaterphenyls (which were major constituents of Yusho PCB), nor more than trace levels of dibenzofurans * (The analytical inter ference by Ci2HlO-xClxO species really doesn't look like a serious problem unless one is willing to postulate that every single iso meric dibenzofuran gives a GC peak that falls directly on top of one of the Ci2lIlQ-xClxO peaks.) You are quite right, however, in point ing out that we cannot exclude the possible presence of a few ppm PCDF, such as have been reported for the French and German PCB's* Until we have definite quantitative PCDF analyses, we won't know'for sure what the actual level of PCDF hazard amounts to. i So where do we go from here? At present, I am confronting rapidly growing internal disenchantment with the idea .that there are any simple, low cost solutions to problems of PCB/PCDF analysis. GE has now been burned three times by sloppy work in this area: first by the young analyst at Woodson-Tennant who dry-labbed his fish data in order to meet a reporting deadline during the ISY5 hearings in 1975; second, by the guy at DuPont who sat on the usvd transformer oil sample during most of 1977 without ever getting his PCDF analytical method worked out; and third, by the internal com munication problems and confusion at UIEHS in 1978. It is evident that the accomplishment of an analysis for PCDF's in either capacitorgrade PCB (primarily trichlorobiphenyls) or in used transformer dielectric fluid (a more complicated blend) which is accurate, de fensible, and publishable, and hence capable of answering more questions than it raises, will require considerable care and effort. Accordingly, we have decided not to send any more samples of PCB dielectric fluids outside the company until we can be assured that suitable levels of care and effort can be expended on them. At the same time, we are actively exploring the possibilities of doing the analyses internally. As I indicated in my letter to McKinney, we now have the high resolution mass spec equipment needed to do un ambiguous qualitative analyses. Thus, the only remaining technical problem is to devise procedures, possibly involving perchlorinaLion, <TK nr> 783144 <C O oUi Irving J. Selikoff - 3 - January 15, 1979 that will allow us to quantitate the levels of all such species as phenols, biphenols, diphenyl ethers, dibenzofurans, terphenyls, quaterphenyls, etc. in a statistically significant group of indus trial PCB specimens. In addition, I would hope that a few laboratory PCB specimens could be examined. These are generally stored in glass containers, and hence subject to photochemical oxidation to PCDF's, which the industrial specimens are not. (The PCDF's themselves are believed to be photochemically unstable in the presence of water, and hence don't form under environmental conditions; but they might be formed in dry PCB on a laboratory shelf.) Hopefully, this would elucidate the problem posed by the frequent reports of highly variable PCD^like bioeffects coming from biological testing of PCB's. There one report by Ax and Hansen showing that even specimens that wereL supposed to be pure, individual PCB isomers, as well as most of the Aroclor specimens examined, actually contained PCDF's at the ppm level. However, I am aware of no other inv'estigator of PCB bio effects, other than Vos, who ever did a PCDF analysis on the PCD specimens that he fed his animals. As you may recall, Jim A l i o ; - . , who reported a level of Aroclor 1248 toxicity in monkeys that in creased nearly 100-fold during the years that he was studying i-1problem, got rather severely chided on this point from the fc'lou. the 1975 PCB conference. In short, your statement, "Finally, si;.?--? the biological effects of PCB's have generally been investigated without such admixtures, we need as much information as possiM.to sort out interactions," may seriously understate the real p lem, which is that we do not actually know what PCB's do in th* absence of such admixtures. At any rate, that's where we seem to be today. I'll keep y . posted on further progress. Sincerely John F. Brown, Jr. Mgr - Life Sciences Ba\u -\\ PHYSICAL CHEMISTRY LA BO!:'" GENP 010334 783145