Document 4JGjKnYEV3OJ5DY0Bwv7mgapN

F ~ Cdoid O 3 . 0 EXHIBIT /?/rr^ Ur Chairman, ladioe and gontlomon. *) * < < C In honor to our Brittleh fcoet I will try to hold this lecturo in English. .... As the tltlo of this locture statoe, I am today going to toll about tho diocoYory of bocio hlthorto unoboerrod chlorinated hydrocorbono having up to eight chlorlno in tho nolooulo and found in reoiduo ana- lyoio. Tha chonlonl none of polychlorinated bifeaylB ( In tho following callod PCB). To got fcniliar with PCB I will start with tho chemistry and .toxioologi. f) flhonistry . . Th . The Bain-characteristic of PCB is 1. Their very high stability. Ab an oxnnplo they can be boiled with nitrin acid without being destroyed. 2. They aro tyardly metabolised in living organism. 3. if more than 4 chlorine are^pre^Bent they are non inflamable* It is cloar that theee threo characteristics doee It eaoy to undoretand that whon they havo entorod the giving organism.- the will have a low persistence . But it in difficult to explain how they find their way into the living organism. Ono thing seems to be clear, they don't come from agricultural ueo, but from a technical one and most probable it domes to the nature Yia waetoe that aro tried to bo burnt up, beoause then we have them at .onco in tho air, beoauae of thoir non inflamability. . . .T oxicologj J Tho PCB woro introduced in 1929 hod as early as 1936 Jones and Alden reported that 23 out of 24 men employed in manufacturing of PCB eufferod from ah acno form oruption nf the skin. Acne did not appear until 6 to B months after the material was first usod. In 1937 Drinker reported that rato expooed to chlorinated biphenylb in concentration of appXoxi- 3 . . *` *' mately 1 mg/ni for 16 hours a day for 6 weeks chowdd damage of tho livor. After that time tho allowod 'concentration of PCB in air le.0,5 (Por DDt the same value ie 0.5 -- 1 mg/m^). Tho nano authors finished thoir experiments in 1938, and related that those compounds have cn injouriouo effoct, manifeoted solely in the liver. Chlorinated bipohnylol appeared to.be the.moot injouriouo chlorinated compounds of all tested. [ WATER PCB-SD0000024860 TRAN 0 5 6 0 3 6 Girecnburg, Mayer and Smith 1939 reported that PCB and polychlorlnetcd napht&lcnoa are blamed for the death of three young workers, and that""" pregnant women and persona who have at tmy time had any liyor disea bob nro particularly suspectible. . ............... ............................ .. ............................................... T7odol, Hallor and Benton gave 1942 animals PCB including adminiotration by inhalation, ingootion and slcin absorbtion. Jliotological examination of tho Yiscora showed important toxic effect only in tho akin and liver, and tho dogonoration effects in the liver are ODBentially the cone whatovor was'tho method for ths administration, Paribok (1955/ found ae an occupational poison in the eleotricnl industry, mixed "totra and ponta chlorobiphenyl cauoos folliculitis, comodo, pyodermia and other okln affootions, and that its principal toxic effect is fatty degeneration of the liver. ' Miller (1944) injected 69 mg PCB (4 and 5 chlorine) eubcontnneouoly in 32 guinea pigB. Eight to ten days after injection, fat droplets were noted in tho liver colls, and after 16 dayB they were preDent in modernto or very largo numbers. Rabbits and rats were also tested in this invooti*>f > gation, ae well ao the PCB was administated both continouBly, eubcuntin- ouely or ingested in the food. In the feeding experiment 0 guinea pigs received 2'doses of 69 mg of the chlorinated biphenyl 1 week apart. Leath occurred in 11 to 29 days. Pinnlly Me Laughlln 1964 reported a method to test the chemical toxicity and teratogenic effect by injection into the yolk sac of fertile eggs prior to incubation. PCB was found between the eight compounds among 100 teoted having tho highest order of toxicity. Ho hatch was found at a level of 25 mg pn egg. At a level of ]0 mg per egg, one chick hatched out of 20 injected eggs, but died 2 days later. Some embryoB which were examined after they died, showed weak deformities (often a short upper brok) and growth retardation. Lead acetate resulted as an example in no hatch at a level of 1 mg per egg. Autopsy of the dead embyos have showed exton 8ive brain damage. Mercuric chloride showed no hatch even at a level of . 0,5 mg por egg. TRAN 0 5 6 0 3 7 As the analytical chemistry is a pronounced service science I have boen in contact with many ecientiets from other fields during the work with residue analysis, and I have always found this contact -very stimulating for my cwn work. This co-operation often demands that wo ere talking the same scientific language. Because of this need I will today try to give a lecture in low level analytical chemistry for biologists,- illustrated by tho residue anolyeio of polychlorinated biphenyls. The lecture will be divided in the following three eub-diviBions: WATER PCB-SD0000024861 o 1. Chemistry of PCB and their tocicology. 2. Analytical methods for rooiduo analyBifl and proof of structurco. J. Behaviour of PCB in nature, differenciee in metabolising XQto ...... " of the PCB components, potenoation in an ecological oerie, con centration lovels and examples of samples which have boon proved to con tain PCB. A residue analysis can be divided ini ' ' 1. Extraction of the pesticides from the biological material, . followed by a careful cloaning-up to take away interferring substances, most often fats. *a 2. Identification analysis by moan of gaB chromatography. Thin- , layer chromatography and mass spectrometry. 3. Quantitative analysis. '. ' . At on ecological laboratory in fliksmuBfiet in Stockholm 1-2 g of a Bamplo is cut out of the biological material and transferred into a weighed and O carofully cleaned test tube, and stored at -20 until analysis. Smaller samplos have been used, min. 5 mg of body fat, and with dry materials Buch as hair, feathers, pine needles 100 mg are sufficient to reach the desired 10 ng/g level in residue analysis. In caBeB of water proofs 1 1. is used for reaching the 10 pg/g. level. B.1 (homog) In order to facilitate complete extraction of the fatty materials from tho biological sample, the doulbe amount of finely powdered anhydrous magnesium sulphate is added to the eampling tube, and the whole is'honogeniecd with an inBertable homogenizor. The resulting powder is transferred into a special Soxhlet extractor. After 4 hours of extraction the solvent is evaporatod, leaving the fat in a small weighed test tube at the bottom Sox. -tube) of the extractor. This fat is dissolved in methylene chloride in such a way that 100 ul (0,1 ml) contain 20 mg of fat. The 100 ul solution 1b now transferred to a little object glaBfe, 3 s 7 cm, covered with a silioagel layer 1 mm thick, in order to fora a line 0,7 cm from one end of the slide. Inserting this thin-layer plate into a vessel the bottom of which 1b covered by a few mm of methylene chloride, the Bolvcnt will be sucked up in the dry layer of silicagol,_and at least reach the upper end of the plate. The fact is that the fat has a greater affinity to the powder on the plate than the chlorinated hydrocarbon have. -- and we get a separation. The fat being more polar than the chlorinated ihydrocarbons will never go longer than 2 cm before thfi TRAN 056038 WATER PCB-SD0000024862 .1 2LDTI0H tubo CO g ) o 0 aolvont roaclioe tho upper part of the class* Tho front of tho fnt appoaro quite vioiblo against a lamp, and with tho aid of a rcsor blado tho zone above tho fat ia tranoferred to tho olutloa tubo and tho chlorinated bioeidou absorbed on the powdor can non bo........ oluted by ono ml of othor. The concentration in sufficient for detection of tho chlorinatod hydrocarbons down to tho 10-12 g levol. Tho noxt atop in tho analytical procedure concerns tho ooporation of tho dlfforont chlorinatod hydrocarbons that tho Dcunplo may contain. Ao a rmttor of fact, this is a broublosome task. It is easy to eotimato what is not proaont, but more difficult to say exactly one. is proaent* V/e ouffor from tho nogativo demonstration, as Trill bo shown lator. , At first a few words about the separation of the components preoont in the sample and their visualization. _ Tho separation is accomplished by mean of a gas chromatograph fitted to a dotoctor that tranofero itB impulse to a recorder. Tho Bystom is shortly .described: A spirally formed glass tube with an inner diameter of 2 mm and about 2 a in length is filled up by a support, covored with an thin layer of an oil. The tubo is heated in the chromatograph to about 200. Through tho tube a stream of nitrogen continously follows. When about 10 ul (l/lOO of 1 ml) of tho purified sample is injected into the tubo, tho components of the eample will be eyaporized and go forward through the column with the gaB stream. As the constituents havo dtifferent affinity to tho column filling thoy will pass the column with different spood and it will take differont time for them to roach tho dotoctor at tho othor end of tho glass tube. If tho temperature and the nitrogon flow Eire held constant thiD timo, tho retontlon tin, haB a spocificvalue for a cortain compound. This is true, but unfortunatoly it is also a fact that two components can havo tho-samo retontlon timo. This is one of tho bigger problems in gas chromatographic analysis of unknown samples, as will soon be obviouo. To make it possible to ostimato tho retention tine it is necessary to visualize the chlorinated hydrocarbons. For that purposo more or loss specific detoctors are used. ThB dotoctor eh. at ofton usod in postlcldo analysis is the so called eloctron capture detector, which cm detect down to ono picogram (=> 10** g of lindan). Unfortunately this dotector ia not specific for chlorine, bu gives answer alco for oxygencontainlng compounds. The response hero is muc lower but can be counterbalanced if the concen tration of the oxygen containing is much higher. TRAN 056039 Tho principlo for tho electron capture dotector is shortly: At the end of the gas chromatographic tube ia placed a little tube con taining a foil tnnrto of tHnninm tritido. Thin in no-radiant. The o- WATER PCB-SD0000024863 ~TJ o 3 particle a ora reacting with the nitrogen molecules coming from the column. Then we get + Ng -- e~ + N*. Ovor the detector v/e have a tcnoion of 90 volt and by mean of the..electrons we will get a constant eloctrical current over the detector. This standing current is transferred to a ono-mV rocordor as a constant baseline. V/hon now a chlorinated hydrocarbon loaves tho oolumn this oonpound hao a high affinity to tho olcotrons and this moans that tho amount of electrons will diminish, and they v/ill diminish proportionally to the amount of chlorine. Tho clcctrichl current will also diminish and this is noted as a peak on the reoordor. Tho area of tho peak will be proportional to the amount of substance in the sample. By moan of a standard injection it is now possible to compare the roten- tion time and the aroa of an unknown component with the retention time and area of the known standard. As eaid before this detoctor is not speci fic for chlrine but anyhow very useful, because of its high conoitivity. Tho systom described has, as we have Been, two disadvantages: 1. Two different compounds can have the same retention time and bo dctocted as one peak. 2. A registrated peak does not need to be chlorinated , because the detector is not specific. If the sample is injected in two different columns with different chemical properties we have increased the chance for a good separation. If two compounds have the same retention time on one column they may not have it on another. When a result seems doubtful, - if the compound being responsible for a certain peak contains chlorine or not - it is possible to concontrate the sample and analyse /it on a less sensitive detector such as the microcloumetric one, which is specific for chlorine. The compound is burned in a furnace and teh generated chlorine titrated directly. - As is seen from the two last mentioned possibilities it is anyhow possible to get a rather high degree of certainty in residue analysis, but it is t a rather time-consumeing work.When using this method just described, we very often found that many chromatograms from residue analysis of most carefully purified samples still contain a large number of peako. Uany of these have retention times that do not agree with any known chlorinated pesticides, or their metabolites. This chromatogram con serve as an example. It was obtained by residue analysis of a sea-eagle found deed in the archipelago of Stockholm. In the range of the kr.ovm peaks, there are so many unidentified that there also must be an obvious risk of the known peako to be covered by unknown onen. Tf thd S remark lr. TrmnH tho y.r.nr+o^ rnwil + o n-r mn.... .-hwIimio . WATER PCB-SD0000024864 TRAN 056CH0 titabivo anal:, .As must be brought into queav in. In the precont investi gation it is shown that most of the unknown peak of chromatograms at rcciduo analysis of chlorinated peoticideo oro due to polychlorinated biphonyls. I will ohow a cliromatogram of human fat analysed on a so called 5k 96............ ...................................................... "................ ................................... ' '' ............... "` * column, the most ofton used typo in pesticide analyses. Early retention times wore in agreement with DEE, DDTop and DDTpp. Next slido shown the Bane sample analysed on a QE-1 column. Now the former 2 DDT peaks have divided into 4 peaks, and two of th*'m are siill in agreenont with DDTpp and op., tho two now wore unknown. Logically, those unknown components were at first thought to be netabolitos of the insecticides. Against that spoke that noigther treatment nor concentrated sulfuric acid in other. This treatment made it rather euro that tho compounds did not contain oxygen. In Sv/odon residues of organic mei cury have boen investigate;: rather intensively in the Swedish fauna. As those compounds give very high responses to the electron capture detec tor it was also investigated if the unknown peaks could have a mercuric origin. It was found that the water-ecological series had high residues of both mercury (v/eotermork,Johnelo) and the unknown ones, when the same indivi duals were analysed. Anyhow, the pheasant suffering most from mercury poisoning only contained low levels of electron capturing compounds and thece belonged ' ' to the normal insectisides. Therefore the unknown could hardly be mercu rials or metabolitos of `them. As the eagle sample giving the chromatogram shown in fig. 10. could be estimated to contain DDT and DDE up to 13 g/kg in extractable fat, the amount of unknown compounds clso were suggested to be in the same range, and tehn sufficiently high to do a run on the combined gas chromatograph mass spectrometer. If this could be done successfully it would be possible' to get very important informationsabout the chemical nature of the unknown, for ex. the molekular weight numbers of chlorine etc. This method is up to now the method giving the highest degree of certainty in the low level analytical chemistry, amounto of 100 ng substance being enough. TRAN 0 5 6 0 4 1 As thiG method for identification of totally unknown residues surely will be very important in the future (when f.cx. a biologist has found that fishes in a river die) it maybpossible by mean of this method to find out exactly what compounds are responsible for the death. For this reason, I will go into come details with this method. In the actual case we took the extract from 20 mg eagle and concentrated WATER PCB-SD0000024865 It as much as possible and made an injection on the gas chromatograph combined with the mass spectrometer. The result was the chromatogram shown on the next slide. Every time tho recorder showed that a compound is loaving the column,, the effluent is led to the rnaes epectromotor. Now Just a fow words about tho mass spec. '* Tho molecules leaving the column are bomded with electrons at E. Wo have now got tho molocule positive charged, but with the same maoo an before. :This M+ is accelerated in a vacuum and will then get a kinetic enorgi. where is the speed. Next comes the magnetic field that trios to bend the direction of the molecule. This v/ill be big for-a small molecule and less for . If we have a slove in the other end we can directly read the molecular weight. Added to this parent molecule M+ we v/ill also got addition in formations, because of the fact that hl+ may not be stisble, a part of thorn will be broken down before they reach the sieve in the other end. P.ex. U M DDT _ CC1^ Mass spectrograms from the different unknown peaks in tho eagle sample as shown. Tho mass numbers equal to the molecular weights of the unknowns could be read to 426,392, 358, 324. Astonishingly, the molecular diffe rences were constantly 34 mass unitB. This difference shows a familarity in origin of the unknown. Now the fact is that chlorine exsiats b.b a ............. mixture of two isotopes with atom weights 35 and 37 in proportion 75s25. If the molecule haB one chlorine, this will give two molecule peaks, one for and one for Cl^. 7f thore are two chlorine we have the possibility of one with only one with both Cl^ ^d 37 and one with 2 C1^ and therefore TRAN 056042 WATER PCB-SD0000024866 \jtr () O o 2ho relation of the poaks found on tho different mass epee woro: Molecular weight '' 324 358 392 426 Chlorino content 5 67 8 Ah erplahation of the familiarity of the compounds can be given if one eubotanco la built from the former by substituting a hydrogen with chlorine RH 14+ + Cl. R?1 + HC1 M + 34 Thon it is poesibD io calculate tho molocular weight' of the parent . hydrocarbon PHC. Mpj,^ * M - x +x , where l& 1b the molecular weight of the conponont having x chlorine atoms. F.ex. for m = 426 and 8 Cl we will get " 426 - 280 + 8 o 154 and equal with the other molekylB. Tho moot probable formula with carbon and hydrogen giving this molecular woight ie C^2 and `this can only be satisfied when the parent-hydro carbon is biphenyl, and the unknown being polychlorinated biphenyls. This explanation was later fully verified by injection of a eynthotic FBC on the mass spec. Furthermore extensive gas chromatographic investigations provod that the FBC standard gave peaks with the same retention time as the unknown poaks from the sea eagle. With the method just described I suppose that wo have a new possibility to study the residues in the air because the pine needles can allways bo . We have had great difficult; in quantifying the FC3, but when getting a little more time it will be possible. We have done a few calculations on a few BpecieB, and 1 suppose they are right within a factor 2. We have found the residue to be from It has been my statement here to-day to present this method for studies of defiling of the nature, and with this method a new typo of defiling agents has been found to be present in nature, and a few experiment have shown where they may bo found. Now this method is going to be used in the firct hadn to estimate how the situation is in nature as a whole, and in the other hand to find the leaks throug which they find its way to naturo. Soem maybe are present here today to get news about the leaks, and to them I want to oay come back in a year. TRAN 056043 WATER PCB-SD0000024867 So much I think I-can cay again that the PCB hardly can cone from agriculture. As support lor this auggeotion I can say that we have found PCB in oaglo foathero from Rikerauceet from 1944, where hardly any ................... chlorinated pesticides were used in ngricultrue. One moro thing that I find important to say is that in contrast to the mercury problem this dooo not ooom to bo a pure Swedish problem. I have Just studied chromatograms tokon from London air, and they dourly contain PCB, and dr. Holdon hun told mo that ho also find them in his fishsamples. But finally in waiting at moro rosults Inshould like to pointHSe morething. It is proved that PC comoo to naturo, wo dont know now where they are uood, but thoy nro vory persistant to chemicals and to fire. I third: the poison Jury Bhould try to state that a content of PCB shall always bo found in an open declara tion. TRAN 056044 / WATER PCB-SD0000024868 WATER PCB-SD0000024869 )zc> VtUJ c. o N N t\ ^s ^ Ki o\ >> <o *.b% K> Kt & i o>?S u *b *bs c> **\ V rv <b ^. II I 1Ii o ^) TRAN 056046 WATER PCB-SD0000024870 TRAN 0560<*7 WATER PCB-SD0000024871 0 V .Vs s V. _ Ov f* 'O la * ^ >*Jnj O . VAjnq ; -c- i ".. ........... '..... '............................. C ^ Cj ^ Z Sajd^ *. s . fa C Oi Vi ft * / SAjrt-j .q * '.' \ _. ` o *t ^ 4' 3 . -> > "b s *' $ A'/V O ltf |1 > ... ... o > - - TRAN 056048 .) 1 I i ii `j i ; O* WATER PCB-SD0000024872 WATER PCB-SD0000024873 L/W D fW t ty- ^ N ok Q> Q\ t k %k A: Q vl O c"> *N -Cuzx-d CL Cv N -J V. ~-&-ur+ r **o r*' -Crstr-rc .. o 6-ti^XLC. <j o K I *> > s #SJ C-> *s n k K \3 ^ <N TRAN 056050 WATER PCB-SD0000024874 L /A/D/?/Jf O Q w ) O 'I 1 *V 2* $V C -> o \_ c p. <Li '-i t 0 c <J 4.0 <0 ' V. ^- 15 V/ S V3 C U. o- O' *m. TRAN 056051 O WATER PCB-SD0000024875 t 0 I, li i !j li 11 O i! o WATER PCB-SD0000024876 WATER PCB-SD0000024877 1 WATER PCB-SD0000024878 WATER PCB-SD0000024879 f >-? )o 3 t is e ^* *\ o. TRAN 056056 WATER_PCB-SD0000024880 -s ft 0 $ *-. >-VV**.`**. TRW 056057 WATER PCB-SD0000024881 *6V. ~$ 'S. -t $! tK- > V .V. .Q & o o,,JL 't 0 * (X a \ 5 O o* *oJ- &' J-- ;$ / / Jl v>> TRAN 056058 s WATER PCB-SD0000024882 oo ) I I { A So ( *) o I 1 JL 'BZ TRAN 056059 WATER PCB-SD0000024883 o sA o rm TRAN 056060 WATER PCB-SD0000024884 OS ( / t.' I} * l .. cJ u --t w ) o o >1 TRAN 056061 WATER PCB-SD0000024885