Document reqZNN2ZKGr8zgGwxYMg6Vzv7

X) OtAVt s: f- rJ sisenic .J67. Acta _ * IP70. h I -7 S' Pesticide residue analysis in the presence of Polychlorobipht nyls (PCB's) By L. M. Reynolds* Conten s L Introduction............................................................ 27 II. Presence of PCB's in the environmenl..................................................... 31 III. PCB interference with pesticide resid le analysis.............................. 32 IV. Uses and properties of PCB's....................................................................... 34 V. Possible modes of entry of PCB's int > the ecosystem........................ 34 VI. Toxicity of PCB's......................................................................................... 36 VII. Analysis of pesticide residues in the p: esence of PCB's............................ 37 VIII. Confirmation of pesticide identity................................................................. 38 IX. Identification of PCB's in samples.................................... 40 X. Estimation of PCB's......................................................................................... 44 XI. Separation of PCB's from organo1ihosphorus compounds by the Florisil technique. ...................................................... 45 XII. Residues of organochlorine pesticides and PCB's in Canadian wildlife 45 Summary ..................................................................................................................52 Resume.............................. 53 Zusammenfassung..................................................................................................... 54 References................................................................................................................. 54 I. Introduction Although there has been so much discussion on the pros and cons of pesticide use, it would be unthinkable to start this review without a few comments on the controversy. The recent banning of DDT by some countries and the widespread and increasing animosity to wards the use of the organochlorine pesticides in general are due to a number of factors including: a) The large annual additions of the pesticides to the environ ment coupled with their apparent persistence. b) Their known acute toxicides. c) Their widespread distribution in biological materials and their reported buildup in the food chains. ' Ontario Research Foundation, Sheridan Park, Ontario, Canada. 27 DSW 373588 STLCOPCB4098963 S8 L. M. Reynolds d) Their more recent implication in the derangement of calcium metabolism (Peakall 1967, Srsncts 1967, Welch et al. 1969) possibly causing the decrease in eggshell thickness and the population decline of certain raptorial species of birds (Ratcliffe 1967, Hickey and .Anderson 1968, Stickel 1969). e) Their adverse effects on fish reproduction (Burdick et al. 1964) and high residue levels found ir Coho salmon in Lake Michi gan in 1969. f) Certain "fears" involving chronic toxidties and sublethal doses with possible deleterious effects on man and wildlife and the ultimate survival of man. The above factors are very compelling and certainly warrant man's concern about the threat to his environment and his very existence. Nevertheless, one cannot deny that these pesticides have been imporant in ensuring the supply of man's rabidly increasing food require ment and in the protection of his hea'th, as well as the saving of millions of lives. Surely, this is a case where the benefit-risks equation rs well as the alternatives to the organochlorine pesticides should 1 e carefully studied and evaluated to ar oid the possibility of getting into a worse predicament. It was quite ironic when recently a large city publicised that, because of an impending ban on DDT, the garbage collector would pick up separately all home garden sprays and other materials which contained even traces of DDT. The material that had been used more than generously and cherished for almost 30 years had suddenly b>!Come monstrous even in trace quantities. Wasn't there much greater danger posed to the collectors, children, and animals by such collec tion than if the home supplies had been allowed to run out gradually with no chance of replenishment? In fact, the announced bans on DDT so far are generally not total, and hence serve the same purpose as merely tightening the control on the uses of such pesticides. In other words, when it is necessary, and until suitable and satisfactory replacements are avail able, limited and judicious use will be made of these pesticides. With regard to the use of substitutes, caution is essential to avoid premature use of any compounds. It is now known that potentiation (or antagonism) of toxicity is possible with certain combinations of pesticides and with combinations of pesticide and drugs or environ mental chemicals (Dubois 1968). A compound might be fairly safe by itself yet harmful in the presence of others. It is a fact that even if suitable substitutes were found and the use of the organochlorine pesticides was banned completely, a need would still exist for at least a number of years to have reliable analyti cal methods for the determination of their residues. The present re view discusses the status of organochlorine pesticide residue analysis ___________ _________ _ ________ ________ ____________________ --~~ r DSW 373589 ... - ... ----------- ----------- .--- tmi . Polychlorobiphenyi* 29 derangement of calcium 967, Welch et al. 1969) gshell thickness and the tonal species of birds son 1968, Stickel 1969). ion (Burdick et dL 1964) o salmon in Lake Michi- rities and sublethal doses nan and wildlife and the d certainly warrant man's it and his very existence, sticides have been imporf increasing food requireas well as the saving of the benefit-risks equation hlorine pesticides should the possibility of getting irge city publicised that, j garbage collector would md other materials which irial that had been used ost 30 years had suddenly Wasn't there much greater d animals by such collecjwed to run out gradually with particular reference to serious inaccuracies caused by the pres ence of polychlorobiphenyls (PCB's) in some samples. As would be expected, in order to cope with the numerous new pesticides that have been put into use, pesticide residue methodology has changed drastically over the years. Zweig (1963) pointed out that until about 1940, the life of an analytical chemist working in the pesticide field was a relatively serene one in which the chemist had to be familiar only with analytical methods for arsenic, lead, fluoride, pyrethrins, rotenone, and a few others. Just about ten years back, DDT and other organochlorine compounds were estimated by colorimetric methods (Schecter et aL 1945) which obviously did not give precise information on the amounts and identities of the individual compounds present Today's residue methodology is much more complex, and in order to cope with the new pesticides being introduced, the techniques are constantly being modified and improved. The wide variations in toxicity and persistence of pesticides, new or old, dictate refine ments in methods in order to differentiate and determine individual residues. Bearing in mind the complexities mentioned above and the fact that the spray history of most samples (excluding controlled spraying experiments) is unknown to the analyst, the availability of chromato graphic techniques which afford separation of a number of compounds in a mixture has made the multi-residue screening approach1 ( U. S. Food and Drug Administration 1969 revision) the method of choice. The gas liquid chromatographic (GLC) separation with electron capture (EC) detection is presently the most popular analytical com bination for organochlorine pesticide residues. Multi-residue analysis involving GLC generally involves five Drin- dpal steps: so far are generally not a) Sampling to procure a representative aliquot of the sample. as merely tightening the b) Extraction of the residue with a solvent (usually organic since . other words, when it is most of the organochlorines (OC's) and many of the organo- ry replacements are avail- phosphates (Op's) are fat soluble). ade of these pesticides, c) Cleanup of the extract to remove interfering materials such as lution is essential to avoid fats, waxes, pigments, etc. v known that potentiation d) Analysis of the cleaned-up extract, generally by GLC-EC h certain combinations of technique. Ide and drugs- or environ- e) Confirmation of the pesticide identity. ound might be fairly safe t s. ; Certainly there are a great number of variations in carrying out tutes were found and the these steps, depending on the nature of the pesticide and its substrate, tanned completely,, a need | us to have reliable analyti residues. The present re, pesticide residue analysis 1 | j `Editor's note: See Burke, J. A.: Development of the Food and Drug Ad ministration's method of analysis for multiple residues of organochlorine pesti cides in foods and feeds. Residue Reviews, this volume. I I 8s DSW 373590 STLCOPCB4098965 I H l i l 'r 1 *'*30* 90 . ; L. M. Reynolds the limit of detectability required, and the laboratory staff and equip ment available. It is beyond the S' 'ope of this review to discuss any of these methods in detail. Furthermore, Gunther (1962) and Samuel and Hodces (1967) have reviewed this topic adequately. Accordingly, this review will deal mainly with problems involving PCB's and re lated compounds. No attempt will :e made to review gas chromatog raphy for pesticide residue analysis. Burke (1965) has discussed some Eractical aspects of this topic, while Thompson et al. (1969 and 1969 a) ave evaluated different gas chromatographic columns for chlorinated pesticides. Before discussing the PCB's and their interference with pesticide residue analysis, some1 key references in the English language on residue methodology are appropriate. These are listed alphabetically as follows: 1. Advances in pest control research, vols. 1-8. New York: Inter science (1957 et seq.) 2. A guide to the analysis of pesti aides by gas chomatography, 2nd ed. S. T. Preston, Jr. Evanston, J11.: Polyscience (1968). 3. Analysis of insecticides and acaricides, Gunther and Blinn. New York-London: Interscience (1955). 4. Analytical methods for pesticides plant growth regulators and food additives, vols. I-V, Zweig. Ne v York: Academic Press (1963). 5. Guide to the analysis of pestici< e residues, vols. 1 and 2, Burch field and Johnson. U. S. Public Health Service, Office of Pesticides, Washington, D. C. (1965). 6. Manual of methods for the determination of residues of Shell pesticides, Shell Chemical Company, Agricultural Chemicals Divi sion, New York (1969). 7. Official methods of analysis. Association of Official Analytical Chemists, 10th ed. Washington, D. C. (1965). 8. Pesticide analytical manual, vols. I and II, 17. S. Department of Health, Education, and Welfare, Food and Drug Administration, Washington, D. C. (1968). 9. Pesticide residue analysis handbook, Bonelli. Wilkins Instrument and Research, Inc., Walnut Creek, Calif. (1965). 10. Residue Reviews, vols. 1-34. Gunther. New York: Springer-Verlag (1962 et seq.) It should be emphasized that these references were all written prior to the knowledge that there was a possibility of PCB interference with 'Editor's note: There are now more than 95 books concerned with pesticide residues including 60 that discuss analytical methodology in some respects; eight countries are represented by the authors. Guntheb, F. A.: Pesticide residues in the total environment-Reliable detection and determination, mitigation, and legis lative control and surveillance programs. Pure and Applied Chem. 21. 355 (1970). TM DSW 373591 STLCOPCB4098966 ; . *> 1v'.^i't) "* ' '''*!#***' `<t. - ^ i .Sr ft V lLr_ :-^; ^.; **'* SL ^ i''**^" -,'r "$*&&. i5 --vj----- ... - ------- 0. Polychlorobiphenyls 31 certain pesticide residue analyses. In general, the methods described in these references can be modified to cope with PCB interference. II. Presence of PCB's in the environment Following the development of the EC detectors and their use with GLC, most residue chemists have observed unidentified peaks on their gas chromatograms- Generally, there was no specific pattern to these peaks, but analysts concerned with pesticide residues in wild life, especially in Europe (Roburl 1965, Harrison 1966, Holden and Marsden 1966), sometimes observed a particular series (with as many as ten or more peaks) of the unilentified peaks in their sample ex tracts- The materials giving rise to these unidentified responses occurred most frequently and in the largest proportions in extracts from aquatic raptorial species of b rds and fishes. The general supposi tion was that these responses w< re from condensation products of the metabolites of the organochlor.ne pesticides, since they were gen erally observed when large amounts of pesticides were also present. Roburn (1965) reported them to ire organochlorine compounds with fairly high chlorine contents, which no doubt accounts for their high sensitivities to the EC detector. However, Jensen (1966) was the first to state that these unidentified peaks corresponded to PCB com pounds, based on the analysis of a number of pike samples, an eagle, and human hair. Initially, the GLC and thin-layer chromatographic (TLC) patterns and chemical inertness were used to identify the unidentified peaks as PCB's (Jensen and Widmark 1967). Other workers in Great Britain (Holmes et al. 1967, Holden and Marsden 1967) and the Netherlands ( Koeman et al. 1967) also made early reports on the presence of PCB's in their fish and wildlife samples. At the Organisation for Economic Co-operation and Development (OECD) Conference on "The Unintended Occurrence of Pesticides in the Environment," held in Scotland in September, 1967, the Euro pean chemists in particular showed much concern regarding the pres ence of PCB's in their wildlife samples and the possibility of inter ference with their organochlorine pesticide residue analysis. Up to that time, there had been no report on the occurrence of these mate rials in North American samples. However, in a December, 1967, report on organochlorine pesticides in seals and porpoises, Holden and Marsden included a comparison of the amounts of PCB-type materials found in Scottish and Canadian samples. They reported that the levels of these non-hydrolysable materials were generally much lower in the Canadian specimens. Since that time Risebrough et al. (1968) have reported that the PCB's are widely dispersed in the global ecosystem, while Reynolds (1968, 1969, and 1969 a) has found them in some Canadian wildlife samples. 4'` ;f.-fV*'4`ri;';.-^V'\'H:' W* .f-s. i. ? & I &$ v$ & DS\N 373592 STLCOPCB4098967 32 L. M. Reynolds Positive confirmation of tl3 PCB's by gas chromatography-mass spectrometry (GC-MS) has nt v been reported in Sweden ( Wedmahl 1967), the Netherlands (Koeman et al. 1969), and in the United States (Reichel 1969). HI. PCB interference with pesticide residue analysis The polychlorobiphenyls ( X indicates the possible chlorine posi tions) and related compound (polychlorinated triphenyls, naph thalenes, terpenes, etc.) have nu merous important industrial uses but are not used as pesticides. Because of their similarities in structure and properties to the DDT pesticide group, the PCB's--if present in a substrate--are carried through the usual pesticide extraction and . creening procedures, and since they possess electron absorbing prooerties, will interfere with GLC-EC analysis of organochlorine as veil as a number of the organophosphorus pesticides (Reynolds 19(9 b). ' The reported GLC pattern* indicating PCB interferences have shown marked resemblances with `Aroclor' 1254 and 1260 (Jensen and Widmark 1967, Holmes ef al. 1967, Reynolds 1968 and 1969, and Koeman et al. 1969). This type of interference with pesticide residue analysis is demonstrated with a standard mixture of organo chlorine pesticides and a commercial PCB mixture (Aroclor 1254) in the chromatograms of Figure 1. The results confirm that the pres ence of PCB's in the sample extracts will cause interference with pesticide residue analysis under these or similar operating conditions. It should be emphasized that throughout this review when the term "PCB" is used, this includes the polychlorobiphenyls and also the other closely related compounds mentioned above. However, the polychlorobiphenyls are the most important from the point of view of interference with pesticide residue analvsis. Under the GLC operat ing conditions, the chlorinated polyphenyls and other related com pounds with molecular weights greater than the chlorinated biphenyls will generally not be separated and observed on the gas chromato grams. Modification of the operating parameters such as higher col umn temperatures or changes in the liquid stationary phases are usu ally necessary to detect the presence of such compounds. It must be borne in mind that when high molecular weight com pounds are present, their peaks can appear on subsequent chromato grams and cause "secondary interference." This occurrence can be recognised easily by the early emergence of unusually broad interfer ing peaks on the chromatograms. "Secondary interference" is a term used by the author to imply that the interfering peak does not have DSW 373593 STLCOPCB4098968 SBretet&tfMwfcxx O Polychlorobiphenyls 33 the same retention time as that of the compound of interest under the same operating conditions. The interfering peak often originates from a previous injection and cannot be simply repeated. Of course, if the elution time is very long the peak might be broad enough to escape detection and in such a case would not interfere. In contrast XU XIV XU XIV lllllll 11111111111111111 M .li I il. 11 111 llllll ill Ml 1111 III 111111111111 nh 10 15 20 25 30 35 Minutes Fig. 1. PCB interference with organochlorine pesticide residue analysis on four percent SE-30/six percent QF-1 on 60/80 mesh Chromosorb W, X 6' borosilicate column. Chromatogram A: standard mixture of organochlorine pesticides; peak numbers: 1 = 0.08 ng. of lindane, 2 = 0.10 ng. of neptachlor, 3 = 0.10 ng. of aldrin, 4 *= 0.14 ng. of heptachlor epoxide, 5 " 0.20 ng. of DDE, 6 = 0.20 ng. of dieldrin, 7 = 0.30 ng. of DDD, and 8 = 0.50 ng. of DDT. Chromatogram B: S ng. of Aroclor 1254 (the 14 major peaks are numbered I to XIV). Chromatogram C: combina tion of above organochlorine standard pesticide mixture and Aroclor 1254. Injector 250C., column 200C., detector base 250C., Ni at 20 30 ca/minute an- to this type of interference, materials which have identical retention times will cause "primary interference" which is easily repeated by reinjection of the extract. With this form of interference a single, sharp peak is obtained, and in the absence of confirmatory and other precautionary measures, misinterpretation can occur easily. PCB-type interference falls mainly in this group. tf+ DSW 373594 STLCOPCB4098969 I ^ I I l a "I 34 L. M. Reynolds IV. Uses and propeides on PCB's In order to understand how it is possible for PCB's to be present in wildlife and cause interference wi h pesticide residue analysis, it is necessary to give a brief description of some of their properties and uses. The PCB's are produced and ma-keted under a number of com mercial trade names, e.g., `Aroclor,' `Clophen,' and `Phenochlor.' Koeman et al. (1969) reported marked resemblance between Aroclor 1260, Clophen A60, and Phenochlor EP6. The `Aroclor' plasticizers, a series of chlorinated biphenyls and chic rinated polyphenyls, are typical of these compounds. As stated by the manufacturers (Monsanto Co. 1967), the `Aroclor' compounds are among the most versatile chemically produced materials available. They vary from mobile, oily liquids to white crystals and hard transparent resins. They are non-oxi dizing, inert, permanently thermopl istic, of low volatility, non corrosive to metals, insoluble in water, and resistant to alkalies, acids, and corrosive chemicals. The viscous, more highly chlorinated liquid and resin members do not support combustion and they impart fire retardance to other materials. The crystalline `Aroclor' compounds are relatively insoluble but the liquic and resinous compounds are soluble in most of the common organic solvents, thinners, and oils. These compounds are used in prolective coatings, as placticizers and extenders, as sealers in water-proofing compounds and putty, in asphaltic materials, printing inks, waxes, and synthetic adhesives. Liquid PCB's are used as dielectrics, as hydraulic fluids, in thermo stats, in cutting oils, as extreme pressure lubricants, as grinding fluids, and as heat transfer media. Solid PCB's are used to impregnate carbon resistors and as sealers in impregnating agents for electrical apparatus. The properties of many products can be varied and improved by the use of these-compounds either as primary' or secondary' plasticizers. Further details concerning the different `Aroclor' series, the grada tion of their properties, and the numerous possible applications can be found in the manufacturer's technical bulletin (Monsanto Co. 1967). In the system of designation for the `Aroclors,' the first two digits indicate the type of material, while the last two digits give the approximate weight percentage of chlorine in the product For example, `Aroclor 12o4 indicates a chlorinated biphenyl with approxi mately 54 percent chlorine, while `Aroclor' 5460 indicates a chlorinated triphenyl containing about 60 percent chlorine. 4 j.. .i ".-4 V, Possible modes of entry of PCB's into the ecosystem The means by which PCB's enter the ecosystem to contaminate fish and wildlife are still not clearly understood. There are however. ~J1:;'-t"`.-H,1--"ir. --fr -rf y ~ ,, ' DSW 373595 STLCOPCB4098970 I , v u*%* irt * CT m -;*ct "Tj> 4 4***J 'J'- r*S. i , In' i ~-n. i :ali* ( tan 'f Co. i ' j eve :* - ' For j I ; i ij( t Polych orobiphenyls 35 four most likely pathways by wl ich PCB's enter fish and wildlife: a) Because of their inert less and versatility and consequent numerous applications, it is quite conceivable that PCB's could be flushed as wastes into rivers, lakes, etc. to pollute fish and other wildlife. . b) There is a possibility th :t some contamination could proceed via the atmosphere when wastes containing these compounds are burnt c) Although it is unlikely, some type of Ullman reaction with condensation of aromafc c halides with the aid of metallic agents such as copper tc give rise to the formation of biaryls, cannot be completely ru! cd out. d) Lastly, but not necessar.ly the least likely, there is the possi bility that some compani js are using PCB's in some insecticide formulations to increase the kill-life. The idea of increasing the residual persistence of insecticides has been investigated for some tii ie and a brief resume on the more important background work is ir order. The early investigations by Lindquist et al. (1945) and Block (1948) indicated that the residual effectiveness of DDT was pro longed when the DDT was aj plied in resins and paints, van Tiel (1952) continued along these li.ies when he increased both the effec tive period and the toxicity of a DDT deposit on a glass surface by the addition of a small amount of coumarone resin to a DDT solution in kerosene. Subsequently, emphasis was placed on work involving the more volatile insecticides like lindane--a very effective pesticide but one which soon loses its activity because of volatilization. Sullivan and Hornstein (1953) in their experiments with the American cockroach, Periplaneta americana, found that lindane resi due remained toxic longer when it was compounded with `Aroclor' 5460. They suggested that the chlorinated polyphenyl prevents crystal lization of the lindane and lowers the vapor pressure of the lindane in the mixture, thereby preventing unsightly residues on surfaces as well as extending the effective kill-life of the insecticide. Tsao et aL (1953) obtained similar results against house flies, Musca domestica L. Continuing this work, Hornstein and Sullivan (1953) con firmed these findings and described a general method for prolonging the residual effectiveness of volatile insecticides. The method consisted of preparing concentrated solutions of the pesticides in the film-form ing polychlorinated polyphenyls. A Monsanto Co. Technical Bul letin (1965) states that because "`Aroclors' in formulations `trap' and hold more volatile ingredients, they make volatile insecticides and repellents "last longer' in residual activity." Attempts to determine whether this idea had been put into practice have so far been unsuccessful. However, if the PCB's are being used vu 1 L1' 1 DSW 373596 STLCOPCB4098971 i ^j ii inr^ii n lnmijmn;ii mi eh i i n( ii I' . f iin^ r.............. ~"r~i-- ------ --i 'I " r . .O [{ l-V, ***'-** * O ;. 38 L. M. Reynolds in pesticide formulation, then thi* coupled with the numerous other uses suggested in a later Monsai to Co. Bulletin (1967), might cer tainly explain their presence in wildlife tissues and other samples. There is an obvious need for clarification of the sources of PCB's in fish and wildlife, and this could probably be accomplished using tracer techniques. VI. Toxiei y of PCBs The PCB's were studied as early as 1881 (Schmidt and Schultz) and were in wide use by 1930 (Penning 1930). Later Jones and Alden (1936) reported that the compounds were toxic. Greenberg et al. (1939) reported that PCB's and polychlorinated naphthalenes caused the deaths of three worker.. Men employed in the production and use of PCB's developed acne-type skin eruptions ( Schwartz and Barlow 1942, Schwartz and Peck 1943, Schwartz 1943). Miller (1944) reported that I CB's caused pathological changes ; in laboratory animals. Brown (1947) warned about the toxicity dan ; gers of the 'Aroclors' when there was a suggestion that one of the `Aroclors' be used as the melting point bath liquid in preference to the customary sulfuric acid. Later, McLauchlin et al. (1963) found that Aroclor 1242, besides being toxic, produced teratogenic effects : in chick embryos at lower dosage levels. Risebhouch et al. (1968) ; reiterated that PCB's along with other clJorinated biocides, such as , DDT, could account for a large part of the aberration in calcium metabolism that has been observed in many species since the second world war. With high dosage of PCB's, hydropericards have been ; observed in quails (Koeman et al. 1969). ; It has been stated (Monsanto Co. 1967) that at ordinary tem peratures the `Aroclor' chlorinated polyphenyls do not present indus : trial toxicological problems. However, caution was advised in handling these materials. The hazard of potential toxic exposure varies with the volatility of the compounds: the lower chlorinated, more volatile ones present more of a potential problem from the standpoint of : both inhalation and skin contact. At elevatedtemperatures, the use i of PCB's requires very effective and efficient exhaustventilation < systems. It has also been stated by the manufacturers that "tests on animals ; . indicate that the maximum safe concentration of vapor is in the range of from 0.5 to 1.0 milligram of the lower chlorinated `Aroclor plasti cizers per cubic meter of air. The threshold limits (maximum allow able concentration for an 8-hour working day) set by the American Conference of Government Hygienists are 1.0 milligram of the lower chlorinated `Aroclor compounds per cubic meter of air and 0.5 milli gram of the more highly chlorinated compounds, such as `Aroclor' 1254, per cubic meter of air." t** -----\m*" ""* ....... ' ' '` ' ' T `-'51V 37,'3597 STLCOPCB4098972 "* f . -He range t plasti- allow ' American l_-ie lower : 0.5 milli- ' a Aroclor' Polychlorobip'lenyls However, as Risebrough et al. (1.H68) pointed out, only relatively small amounts of chlorinated hydroca rbons are required to cause en zyme-induced breakdown of steroids thus making irrelevant much of the p.p.m. approach to pollutant ecology based on toxicity data alone. There is no doubt that further research in this area is urgently needed. VH. Analysis of pesticide residues in the presence of PCB's PCB's are of interest to analysts or three main reasons: a) The compounds are toxic. b) They are widely distributed aid their uses are increasing. c) They interfere with GLC determinations of organochlorine and organophosphorus pesticic e residues. The first two reasons have been dealt with earlier and therefore need no further elaboration. Accordingly most of the remainder of this review wall be concerned with ihe analysis of pesticide residues in the .presence of PCB's. Also the present methods of estimating the amounts of PCB's will be discusse d. As mentioned earlier, it was emphasized at the 1967 OECD confer ence that PCB's were in fact being detected in fish and wildlife, especially in Europe. The conference among other tilings served to alert other chemists and scientists to the problems of PCB's and their interference with pesticide residue analysis. Since the discovery that PCB's are contaminating fish and wildlife and that they interfere with organochlorine pesticide analysis, very little has been published on how to cope with these interferences and how to differentiate the PCBs from pesticides. Based on preliminary work, Jensen and Widmark (1967) suggested a nitration technique based on the inert ness of the PCB's compared to the relatively easier nitration of the pesticides. Risebrough et al. (1968) reported the use of this method, but Reynolds (1969 b) pointed out that nitration is not a suitable approach since some pesticides (lindane, toxapene, Strobane, etc.) apparently will not nitrate, while some of the PCB's appeared to be nitrated under the suggested conditions. In addition, there is the problem of further complication from the interference of the nitroderivatives formed, especially when, the DDT-group pesticides are present in large amounts. To avoid the above complications and at the same time facilitate the chromatographic interpretations and afford a means of estimating the PCB's and pesticides independently, Reynolds (1969 b) intro duced a Florisil separation scheme, Koeman et al. (1969) have also reported a similar Florisil column technique to separate the PCB's from the pesticides. DS\N 373598 STLCOPCB4098973 38 - L. M. Reynolds - In Reynolds' method, the pai tially cleaned-up sample extract con taining the pesticides and PCBY, is placed on a Florisil column and the PCB's are eluted first with n-hexane, followed by elution of the pesticides with an ether-hexane mixture. Originally, this technique was intended to serve as the firal cleanup step, as well as effecting separation of the PCB's from tie pesticides. As will be explained, the method has since been slightly modified (miniaturized) to serve mainly as a separation technique. Of the many pesticides investigated, only heptachlor, aldrin, and the DDT metabolite p,p'-DDE are eluted with the PCB's. With the exception of p.p'-DDE, these do not present difficult problems since chemical reactions (epoxidation, hydrobromination, etc.) can be used to differentiate heptachlor and aldrin from the PCB's. Furthermore, these two pesticides are more frequendy found as the more stable epoxy products which are separable from the PCB's by this scheme. p,p'-DDE on the other hand presents a more difficult problem on account of its chemical inert ness. Its greater stability resembles that of the PCB's, and the compound is not amenable to any simple and convenient method of struerure modification to give a product which is readily detected within he usual GLC-EC operating param eters. Of course the ideal situation would be a convenient addition of one mole of hydrochloric acid to the p./Z-DDE to form p,p'-DDT while-the PCB's are not affected. This would afford confirmation and quantitative estimation of p,p'-DDE in the presence of PCB's. Gen erally under such circumstances :t would not be necessary to do an additional Florisil separation. The amount of p.p'-DDE present could be estimated indirectly from the p,p'-DDT produced since, as will be discussed later, the p,p'-DDT (and p.p'-DDD) can be separated from the PCB's by the use of highlv polar liquid phases. It should be emphasized that although p.p'-DDD and p.p'-DDT are mostly separated by the polar phase columns, without introducing a Florisil separation the earlier erherging pesticides are superimposed on the PCB's and hence a separation is necessary to identify and to quantify these latter pesticides. After the Florisil separations of PCB's and pesticides are made, the two eluates are chromatographed separately on the SE-30/QF-1 column and compared with appropriate standards for quantitation. This is normally followed by confirmation of the identities of the VDX Confirmation of pesticide identity Despite the separation of the PCB's and their elimination as sources of interference, the non-specific nature of the EC detector SW 373599 STLCOPCB4098974 ? Polychlorobi] ihenyls 39 iii makes it necessary to confirm the identity of the pesticides by addi tional techniques. The sequence usually followed by our laboratory is: a) Injection of the pesticide eluate onto a more polar liquid phase column (e.g., polyester such as DEGA or DEGS). Typical t 1 i -nf-T l i ii < t '< i DDT tli f< i / t Fig. 2. Separation of PCB's and organochlorine pesticides on polar phase column (five percent DEGS/two percent HjPCh). Chromatogram A: standard mixture of organochlorine pesticides with same amounts as in Figure 1; peak numbers: 1 " heptachlor and aldrin, 2 *= lindane, 3 = heptachlor epoxide, 4 ** DDE, 5 = dieldrin, 6 = DDT, 7 = DDD. Chromatogram B: 5 ng. of Aroclor 1254 (the 14 major peaks are numbered I to XIV as in chromatogram B of Figure 1, assuming that the order of elution is unchanged). Chromatogram C: combination of organochlorine standard pesticide mixture and Aroclor. 1254. Other GLC parameters as in Figure 1 - separations of pesticides, PCB's, and a mixture of the two groups are shown in Figure 2. The reversal of p.p'-DDD and p,p'-DDT and their separation from most of the PCB's is note worthy and quite useful. One drawback is that peak No. 13 of Aroclor 1260 interferes partially with p,p'-DDT. Other use ful information can be obtained by study of the retention times DSW 373600 STLCOPCB4098975 flilfri. au4 * f ^ ifi 'Wv`~fa; ^ fitt^ - 'it^ iB iffilii^rfiiaWr^i Ai^f.ii.iilgiriiff *L. 40 L. M. Reynolds of individual pesticides. For example, this approach affords separation of the three main BHC isomers. The technique is especially useful for ru'ing out the presence of pesticides in dicated as possibly present by the SE-30/QF-1 column. How ever, the retention times on two or more stationary phases cannot be regarded as independent parameters of identity (Robinson 1967). b) Derivatization and use of characteristic GLC retention times of the derivatives. As reported earlier (Reynolds 1969 a), our laboratory has increasingly utilized chemical reactions for con firmations of pesticide residues because the samples to be analysed are generally snail and the amount of residue is in sufficient to allow effective application of infrared, mass, and other spectroscopic methods. The derivatization techniqu. has been used by a number of work ers (Gunther 1962, Klein et id. 1964, Hammence et al. 1965, Sans 1967, Duffy and Wong 1967, O adchuk and Wanless 1968, Cochrane and Chau 1968, Chau 1969, Clau and Cochrane 1969, Wiencke and Burke 1969) and is gaining ir popularity. Success of the technique usually depends on the specific derivative remaining lipophilic and therefore extractable with the organic solvent (usually hexane) for direct analysis by the usual GLC'-EC method. A number of chemical reactions is available, but two of the more fcroad-spcetrum types that are quite useful for the pesticides generally found in wildlife are: (i) Reaction with ethanolie potassium hydroxide (mainly dehy drochlorination) to give products with shorter retention times. This is effective for DDT, DDD, and their isomers, as well as a- and y-BHC; the /J-isomer is not affected. (ii) Reaction with hydrogen bromide/acetic anhydride reagent to give bromohydrin or bromoacetoxy derivatives with longer retention times. This reaction has been found useful for heptachlor, aldrin, heptachlor epoxide, dieldrin, and endrin. c) Thin-layer chromatography. In spite of certain drawbacks with TLC for confirmation of small amounts of pesticide residues (Reynolds 1969 a) the technique is still valuable for the con firmation of DDE. TLC can be quite useful also for some group separation, as an ancillary cleanup technique, and for semiquantitative estimation of some pesticide residues. IX. Identification of FCB's in samples Data obtained by our laboratory in the course of work conducted for The Canadian Wildlife Service have demonstrated that PCB's are present in some but not all species of Canadian wildlife. Undoubt- L.u^hitiwil^Si^: '^rT a *$*#*%$[* DSW 373601 STLCOPCB4098976 v. and _ rrlnique ? iic and i*c) for ir : . more | ; federally | 5.iy dehyfrtention o.-aiers, as v reagent * ' di longer i for hep- f "-drin. f *'-*cla with ' residues the con i * for some and for ascj. ' conducted ,, 1 that PCB's 4 Undoubt- -V o -.'i Polychlorobiphenyls 41 edly the presence of PCB's hi. ; complicated the pesticide residue methodology, but this complication cannot be avoided if meaningful results are to be obtained. Since not all samples contain PCBs, it would be unwise to do a Florisil separation on all extracts before having some indication whether or not the separation is necessary. Accordingly our laboratory uses the approach described below. The sample to be analysed is extracted, cleaned-up, and assayed using the SE-30/QF-1 column. Depending on the EC results, a deci sion is made whether or not to make a Florisil separation and check further for the presence of PCL's. This decision is based on the as sumption that if there are no apparent p,p'~DDD and p,p'-DDT peaks, then the presence of PCB's would not be expected. The DDD-DDT combination is chosen as the criterion because peaks eight and ten, two of the larger peaks in Aroclor 1254 and 1260 have identical reten tion times with p.p'-DDD and ;,p'-DDT, respectively. Also, the PCB components giving rise to peaks eight and ten are less likely to be metabolized than the earlier emerging compounds. Hence, if PCB's (1254 or 1260 pattern) are present, peaks should show up giving apparent p^-DDD and p,p'-DDT values with the EC detection system. Other useful information regarding the presence of PCB's can be obtained from the initial gas chromatography. If the CLC pattern (prior to PCB separation on Florisil) shows high "apparent" p,p*-DDE with little or no p,]/-DDD and p.p'-DDT present, then all or most of the "apparent" DDE is probably "true" DDE. This is because peak five (the peak which interferes with p.p'-DDE) in the common PCB commercial mixtures (Aroclor 1254 and 1260) is relatively small compared to the DDD- and DDT-interfering peaks (eight and ten) as shown in Figure 1. This rule has been borne out by TLC confirmation of DDE in a number of such cases. As mentioned above, the two most commonlv found PCB types in wildlife samples resemble-Aroclor 1254 and 1260. Although these two Aroclors are quite similar in many respects, certain differences in their GLC patterns, as shown in Figure 3, can be used to identify one from the other. With a higher chlorine content, Aroclor 1260 shows about 17 major peaks (SE-30/QF-1 column) compared to 14 with Aroclor 1254. Besides this, one other obvious major difference is the peak-height ratio of peaks ten and 13. This ratio has rough values of 9.3 and 1.0 for Aroclors 1254 and 1260, respectively. Additional information regarding the PCB mixture present in a sample can be obtained bv comparisons of other peak-height ratios of the PCB standards and the PCB's in the sample. However, it must be borne in mind that in practice one might be faced with complex mixtures of PCB's rather than a single, specific mixture. In such cases an average value seems to be the best compromise. The method of estimating PCB's described below attempts to take this into account DSVM 373602 STLCOPCB4098977 w..t-iii>* * a.nrian--.n itf*i iito*' irtaaifriW itinaa-Mri6yln>it i Ur^ 42 V . . .. L. M. Reynolds : Although most of the early reports on the presence of PCB s in samples indicated the Aroclor 1254 GLC pattern, more recent work by Koeman et al. (1969) and the author's laboratory has indicated that a proportion of the speonnens does contain the Aroclor 1260 GLC pattern. This apparent predominance of the Aroclor 1254 type might have been due to its availability as a standard and lack of the Aroclor 1260 type in the early developmental stages of PCB methodology. The GLC profiles of a number of commercially available PCB mixtures were determined using the SE-30/QF-1 column. These are shown in Figure 4 and indicate that the more highly chlorinated w' ry-'-V;-vn - -v*-^ rw vpw r * * * r j * i * M u l'in i* *i ir i Fig. 3. Comparison of Aroclors 1254 and 1260. Chromatogram A: 5 ng. of Aroclor 1254; the 14 major peaks are numbered I to XIV as in chroma togram B of Figure 1. Chromatogram B: 5 ng. of Aroclor 1260; the 17 major peaks are numbered I to XVII (the early peaks correspond to those in Aroclor 1254). GLC parameters as in Figure 1 biphenyls (Aroclors 1254 and 1260) are easily detected with the usual operating parameters. On the other hand, the compounds of the lower chlorinated mixtures (Aroclors 1221, 1232, and 1242) and the higher molecular weight mixtures (e.g., Aroclor 5460) are less responsive and are, therefore, more likely to go undetected under normal operat ing conditions. Apparently, the lowered sensitivity is due to low chlo rine content in die first group and to high molecular weight with resultant long retention times in the second group. In both cases, therefore, higher concentration would be a prerequisite to ensure detection under normal conditions. This, as well as the possibility of metabolism of the less highly chlorinated biphenyls (Koeman et DSW 373603 STLCOPCB4098978 V feaifjjg*p&wiKSj *~Av*ii u ^VjVtg^^v'i'wl f^^aaswa^jaiiittfa. s ,,f A t !* ; x " < 3* ~Wr &* 4r- fr. Polychlorobiphenyls 43..- aL 1969), might explain the general absence or rarity of these com pounds in wildlife samples. It must not be overlooked, therefore, that in the presence of large amounts of the lower chlorinated biphenyls (e.g., Aroclors 1221, 1232, j iii *i ; t ng, of t Aromae (he 17 j .r.pood to .# i j * usual j ie lower higher >- -sponsive l -J operat- - low chlof~-At with ith cases, -to ensure possibility CtMAN et ( Fig. 4. GLC profiles of the more popular Aroclor mixtures under normal analy tical conditions. Note: All peak numbers correspond to those of Aroclor . 1254. Chomatograms: A = 5 ng. of Aroclor 1221, B = 5 ng. of Aroclor 1232, C 5 ng. of Aroclor 1242, D = 5 ng. of Aroclor 1254, E * 5 ng. of Aroclor 1260, and F -- 5 ng. of Aroclor 5460. GLC parameters as in Figure 1 and 1242) and negligible quantities of the higher chlorinated bi phenyls (e.g., Aroclors 1254 and 1260), different criteria would be necessary to detect PCB interference. To date, this condition has not been encountered in Canadian wildlife samples. DSW 373604 STLCOPCB4098979 ...;-,< I m< 44 L. M. Reynolds As might be expected from 'he complex nature of the PCB mix tures and the chemical inertness of the individual compounds, the confirmation of identities of these materials is much more difficult than with the organochlorine pesticides. Thus, in the absence of a mass spectrometer, confirmation of the PCB's is mainly by their GLC retention times and patterns on the mixed and polar phase columns in conjunction with their non-reactivities with the two general reagents used for the organochlorine pesticides. The prior separation of the PCB's from the organochlorine pesti cides by the Florisil technique also aids in the identification of the PCB's. X. Estima ion of PCB's Since the PCB's are themselves toxic, attempts have been made to estimate the amounts present in samples. It should, however, be recognized that only rough estimates can be made since pure stan dards of the individual compounds in the mixtures are to date unavail able. The situation resembles that of toxaphene estimation, but is of greater complexity. Koeman et al. (1969) based their PCB estimation on the peak of phenoclor DP6 having rz (relative retention time with dieldrin = 1 under his operating conditions) equal to 1.45. Risebrough et al. (1969) estimated the PCB's on the assumption that they have similar EC responses to p.p'-DDE and applied a factor to fit the assumed 54 percent chlorine content of the PCB's. Jexsen et al. (1969) re ported estimates as the sum of all the PCB components. The author's laboratory uses a method very similar to Koeman's with the exception that an average based on two peaks is usually employed. Should the two values before averaging vary to any extent, additional estimations are made using other peaks, and all the results are averaged. In the case of a few samples, the PCB level was ob tained by averaging results from all the major peaks in the mixture; however, the use of peaks eight and ten only has been found to give equally satisfactory results. Note that results are reported as being based on Aroclor 1254 or 1260 depending on the overall GLC pattern. Peaks eight and ten were chosen because of their stability (see section on identification). In a later section the above points are illustrated by analysis of resin powder taken from a fish hatchery trough. The total PCB results obtained by the method described may be slightly higher than the actual values because the early emerging PCB peaks (mainly one, two, and three) are usually absent from the chromatograms of sample extracts. However, this should not affect the overall results greatly since these peaks represent minor compo nents in the mixtures. Furthermore, there might be compensation, since no attempt has been made to apply recovery correction factors. . . ;; DSW 373605 STLCOPCB4098980 % * ** i1 ii $ -; f I* k I ~*. I^ i .!* *r. . < Polychlor jbiphenyls 45 ; The method of estimation described above appears to give a rea sonable estimate of PCB contam nation based on commercial PCB mixtures. For absolute measurement of PCB content it would be necessary to prepare standards fi r the individual PCB components, and, even if these were available, the determinations would be complex and costly. Considering our present limited knowledge of total PCB toxicity and our even eater ignorance about the toxicity I of the individual components of the mixtures, it is doubtful whether . such accurate determinations world be meaningful at this time. As : stressed earlier, the most urgent reed is for toxicological investigation . in this area to indicate whether some of the PCB compounds are more toxic or have greater subh thal effect than others, as observed with the BHC isomers. At that point it would become essential to ; identify and estimate accurately specific individual compounds in the ; mixtures. | | XI. Separation of PCB's from organophosphorus compounds by the Floi isil technique ] A limited amount of work has been carried out in our laboratory f to determine whether the Florisil column technique is applicable to j the separation of the PCBs from the organophosphorus compounds. J Preliminary results on those tested (phorate, diazinon, ronnel, mala . thion, parathion, methyl Trithion, ethion) indicate that they are not eluted with the PCB's by hexane, and can therefore be separated. In comparison to the organochlorine pesticides, the organophosphorus compounds generally require larger volumes of the ether-hexane mix ture for elution from the Florisil column, most likely because of their more polar nature. As with the organochlorine pesticides, the presence of PCB's in a sample extract will interfere,with some of the more common organophosphates when the EC detection system is used. However, with the use of a dual detector (EC -(- phosphate or thermionic) connected to a dual-pen recorder, the effluent from a single injection containing PCBs and organophosphates can be split and differentiation made as shown in Figure 5. Like the organochlorine pesticides, the PCB's - are not detected on the phosphate detector. Nevertheless, in the ab sence of a detector which is specific for phosphorus, the Florisil col umn technique can be used to separate the PCB's from the organo1 phosphates to facilitate the analyses. XII. Residues of organochlorine pesticides and PCB's in Canadian wildlife Following the approaches outlined, the author's laboratory has analysed a number of Canadian wildlife specimens for residues of I DS\N 373606 STLCOPCB4098981 46 - L. M. Reynolds L Fig. 5. Differentiation of Aroclor 1254 and organophosphorus pesticides using dual detectors and effluent split. Chromatogram A: organophosphorus standard pesticide mixture; peak numbers 1 <= 1.75 ng. of phorate, 2 ' 2.0 ng. of diazinon, 3 = 3.0 ng. of Ronnel, 4 = 13.0 ng. of malathioc. 5 = 5.0 ng. of parathion, 6 = 15.0 ng. of methyl Trithion, and 7 = 10.0 ng. of ethion. Chromatogram B: combination of organophosphorus mix ture (chromatogram A) and 10 ng. of Aroclor 1254 numbered I to XIV as in chromatogram B of Figure 1 (a compromise of the E.C. response is necessary to obtain maximum response on the phosphorus detector). ~P " phosphorus detector and EC = electron capture detector. Column four percent OV-100/six percent OV-210 on 60/80 mesh Chromosorb V/ (AW); other GLC parameters as in Figure 1 organochlorine pesticides and PCB's. Some of the data on aquatic birds from Western Canada are given in Table I. The inclusion of these tabulated results, which represent only a minor portion of the total number of samples analysed, is intended to illustrate the types of analytical problems created by the presence of PCB's, and to indi cate potential errors due to high levels of PCB's in samples. Referring to Table I, some general remarks are in order: 1. The greater portion of the apparent DDE in samples is in fact DDE. This is predictable from the comparatively smaller amounts of apparent DDD and DDT, and has been verified by TLC. 2. A large proportion of the apparent DDD in the samples is due to PCB's. However, DDD is present in some samples. DSW 373607 STLCOPCB4098982 o 'iiy*' *f< i vi f '-' . Polychlo. obiphenyls 47 3. The effect o PCB contamination on DDT values is less marked ! than in the case of DDD, but it is still substantial. 4. The apparent HE values are in most cases due entirely to the pesticide. This is somewhat incongruous with the rest of the data since peak four of the commercial PCB mixtures does interfere with HE. However, this anomaly may be explained on the basis that the PCB component giving rise to peak four is degraded ex tensively in the bird's metabolic processes. 5. Based on all of the PCB est'mates in wildlife samples noted to date, including those in Table I, the following general trends are evident: (a) Aquatic raptorials generally contain higher residue levels of pesticides and PCB's than oth'r wildlife species. : (b) High levels of PCB's ar*. found only in the presence of high levels of organochlorine pest aides (the resin powder mentioned ; below is an obvious exception) but high levels of organochlorine : pesticides are not necessarily accompanied by high levels of PCB's. (c) The most abundant PCB types detected are those which show : Aroclor 1254 and 1260 patte ns and, therefore, fit well with the proposed method of quantitat on. Ilow'ever, it should not be over ; looked that with the present residue methodology some other PCB ; type compounds might go und elected. 6. It is -important to measure moisture and fat content of samples i in conjunction with residue levels. This allows for reporting on ; a wet-, dry-, or fat-weight basis. However, caution should be used when reporting residues in wildlife samples, for example, on a j fat basis when the fat content is very low. Such figures can be ! inaccurate as well as highly misleading. ; The gas chromatogram of a typical sample showing the presence ! of PCB's is given in Figure 6. It indicates the Aroclor 1260 pattern i and was obtained from a duckling found dead in the Toronto area j in the summer of 1969. - f A good example of PCB's showing up in unexpected places oc- j curred recently. A wildlife biologist was faced with the problem of i fish dying in a hatchery for no apparent reason. He scraped some I of the resin powder from the hatchery trough and found, surprisingly, that experimental fish were susceptible to even a dilute solution of 1 the powder. He submitted a sample of the resin powder to the author's { laboratory, and upon analysis a typical GLC pattern of PCB's (Aroclor ! 1254 type) wfas obtained. This is shown in Figure 7. Using the meth- odology described earlier in this review, the sample was found to | contain PCB's but no pesticides. Based on calculations from 11 indi- j vidual peaks of Aroclor 12-54, the PCB concentration was estimated I at 35 p.p.m (range 25 to 46). Using the average results based on j peaks eight and ten only, the PCB estimate was 29 p.p.m. This is f\ DSW 373608 STLCOPCB4098983 ... ,;;i>ytei< 1 : ---- ' ....... 48 ' L. M. Reynolds rw'fJji*nujy-r'- DSW 373609 STLCOPCB4098984  2 IS Polychlorob iphenyls 38"$" (OhhOhQO S|8| |83 0.12 Trace* 0.05 Trace* Trace* 0.01 0.01 After S ts? S 0 ^1 a a After Before (apparent) 52 2 o o 8 S dddddod SS8828S 2888838 ooooooo SSSS3SB' ddddddo Before (apparent) Table I. ( is ss*--< sss After Q Q SaSScgS O^HO^OO O 0. tC 8S882gg O O 1-* o o o Before (apparent) Dieldrin* i ssssss-s o .. S \q ocoNot^eo^ ^CS M-H CfOt Wr-< W OJ 49 Sample 1 area 6 Samplo 2 area 6 Sample 3 area 6 Sample 1 area 7 Sample 2 area 7 Samplo 1 area 8 Sample 2 area 9 . arjSgggV^-',' .:,yffr*,-vr DSW 373610 STLCOPCB4098985 Fig. 6. Typical sample indicating t ie presence of PCB's of the Aroclor 1260 type in a duckling from the Toronto area. Chromatogram A: 5 ng. of I Aroclor 1260 numbered I tc XVII as in chromatogram B of Figure 3. Chromatogram B: the equivilent of 0.33 mg. of duck sample (hexane portion of Florisil split). Note the DDE contribution to PCB peak V. CLC parameters as in Figure 1 ' Fig. 7. Resin powder extract indicating the presence of PCB's of the Aroclor 1254 type. Chromatogram A: 5 ng. of Aroclor 1254 numbered I to XIV as in chromatogram B of Figure 1. Chromatogram B: The equivalent of 0.02 mg. of resin powder extract (hexane portion of Florisil split), numbered I to XIV as in Aroclor 1254. CLC parameters as in Figure 1 DSW 373611 STLCOPCB4098986 I IS30 * of .-** X 5! i<l Polychloi ibiphenyls 51 4 an interesting example of PCB s causing contamination as a direct i j result of industrial application. With regard to the Florisil technique for the separation of PCB's from pesticides, one important modification to the original method (Reynolds 1969) is now used routinely in the wildlife studies. Origi i nally the scheme was designed to accommodate simultaneously both the cleanup of the sample extract and the separation of PCB's. How ever, as pointed out earlier in tills review, not all samples contain i It PCB's. Consequently, for general samples, to avoid unnecessary split ting of all sample extracts (each split doubles the number of eluates to be chromatographed), a normal cleanup including a Florisil col umn, but no differential elution, is conducted. The Florisil-PCB sepa \ j ration is then carried out on the cleaned-up extract, if it is required. it1 ( Under these conditions smaller columns have been found to be as efficient as the larger columns us id in the original method, and there is a substantial saving in the amount of reagents required; Details of the modified procedure are as follows: f A glass column (44 cm. X 1 cm. i.d. having a 50-ml. reservoir on top) is packed with 30 cm. (~10.7 g.) >f Florisil (60/100 mesh, Floridin Co., pesticide grade, stored at 130C. until ready for use) and topped with a 2-cm. layer of anhydrous Na,SO.. The sample extract of about 5 ml. (n-hexane) is added to the column. The first elution which removes PCB's, heptachlor, aldrin, and DI E, is effected with 60 ml. of n-hexane (Nanograde, Mallinckrodt). The remaining pesticides arc then eluted with 40 ml. of 50 percent ethyl ether in hexane. Note that in filling the columns, the Florisil is packed down by gentle tapping, and the sample extract is added to a dry column, i.e., no pre wetting. After the first elution the receiver is changed as soon as the last of the hexane meets the NaiSO,. It should be noted also that with one batch of pesticide grade Florisil, some PCB peaks (two, six, nine, 11) showed partial elution in the second eluate. These, however, are com paratively minor peaks and do not interfere with any of the common organochlorine pesticides. As there are differences in batches of Florisil, the procedure has to be checked out fully prior to application to actual samples. Modifications of volumes, etc., may be necessary to effect the desired separation. It is likely that adsorbents other than Florisil, e.g., . . silica gel or alumina, will give similar separation of PCB's from pesticides but these have not been investigated to date. j Acknowledgments | Work in connection with this review was supported by funds from ^ Atoclor j the Pesticide Section, Canadian Wildlife Service, Ottawa, and from '< to XIV j the Province of Ontario through the Department of Trade and Devel""timt of j opment. The writer gratefully acknowledges helpful suggestions of iplit), ! Dr. S. G. Reid, Director of Organic Chemistry Department, Ontario * tifuro 1 i DSW 373612 STLCOPCB4098987 J2O *t,'g<-ti -^itaiyirfi..;. ' 52 L. M. Reynolds ; Research Foundation in preparing this review for publication. The '. technical assistance of Terry Cooper 'md Sandra Takacs is gratefully acknowledged. i I. . <., Table II. Chemical designations of ; esticides mentioned in test ?- r Pesticide Chemical name aldrin p,p-DDD (TDE) p,p'-DDE p.p'-DDT diazinon dieldrin ethion heptachlor beptachlor epoxide lindane (y-BHC) mele.thion methyl Trithion parathion phorate (Thimet) Ronnel 1.2.3.4.10.10-hexachlor>-l,4,4a,5,8,8a-hexahydro-l,4-endo, ero-5,S-dimethanons phthalene 2,2-bis(p-chlorophenyly-l,l-dichioroethane 1.1-dichloro-2,2-bis(p-< hlorophcnyl)ethylene ' 1.1.1-trichIoro~2,2-bis( j-chIorophenyl)ethane 0,0-diethyl-0-(2-isopropyl-4-methyI-6-pyrimidyl)-phos- phorothioate 1.2.3.4.10.10-hexachlo: o-6,7-epoxy- l,4,4a-5,6,7,8,8a-octar hydro- 1,4-endo, exo-l, 8-dimethanonaphthalene 0,0,0',0'-tetraethyl-i,iS'-methylene bisphosphorodithioate 1.4.5.6.7.8.8-heptach]iro-3a,4,7,7a-tetrahydro-4,7-endo- methanoindene 1.4.5.6.7.8.8-heptachL ro-2,3-epoxy-3a,4,7,7a-tetrahydro-4, 7-mcthanoindane y-1,2,3,4,5,6-hexachlo -ocyclohexane S-[l,2-bis(ethoxy carbonyl)ethyl]0,O-dimethyl phosphorodi- thioate 0,0-dimcthyl S-(p-chlorophenylthio) methyl phosphorodi- thioate 0,0-diethyl 0-p-nitro jhenyl phosphorothioate 0,0-diethyl S-(ethyltaio) methyl phosphorodithioate dimethyl 2,4,5-triclilcrophenyl phosphorothionate Summary i l. P-* ' / SC *!. f :t : O* . lie : - C t >? v * * Since 1966, PCB's have been detected in fish and wildlife, espe cially in Europe and North America. Contamination from industrial wastes via the water ways appears to be the main source of PCB's in aquatic species. The PCB's have many important industrial uses but are not utilized as pesticides. Due to their similarities in structure and properties to the organochlorine pesticides, the PCB's tend to interfere with accurate GLC-EC determination of organochlorine and many organophosphorus pesti cides. In addition to their interference with pesticide residue analysis the PCB's are themselves toxic so that their quantitation is desirable. A method for recognising the presence of PCB's in samples, sep arating them from pesticides, and determining the pesticide levels accurately is described. Simultaneously an estimate of total PCB con tent is afforded. i ?: t - < jr r fir A: i '. !r - WTW-i' DSW 373613 STLCOPCB4098988 s' ~ r-.: ir.. 4 Polychlorobiphenyis 53 The residue data of some Ca: adian aquatic birds indicate that high levels of PCB's are found only in the presence of correspondingly high levels of the DDT pesticide group. On the other hand, high . organochlorine pesticide levels are not necessarily accompanied by , high PCB content Further research on toxicity and sublethal effects of PCB mixtures and their individual components is required. If particular components are shown to be more toxic than o'hers, refining of the PCB quantita tion methods will be necessary R4sum4* Analyse de r4sidus de pesticides en presence de diphenyls polychlores Depuis 1966 des diphenyls pclychlores (PCBs) ont ete detectds 1 dans les poissons et le gibier, particulierement en Europe et en Amerique du Nord. II apparait que la presence de PCBs dans la faune aquatique soit due principal ement a la contamination des corns d'eau par des dechets industriels. Les PCBs sont utilises pour un grand nombre de procedes industriels importants, mais ils ne sont pas appliques comme pesticides. Etant donne que la structure et les proprietes des PCBs sont semblables k celles des pesticides organochlores, ils ont tendance & fausser la determination exacte des substances organochlorees et organophosphorees lors d'analyses par chromatographie en phase gazeuse via capture d'electrons. En plus de leur interference lors de l'analyse des residus des pesticides, les PCBs sont eux-memes des composes toxiques dont la determination quantitative est souhaitable. ! Le present article contient une methode qui permet de reconnaitre la presence des PCBs dans des 4chantillons et, en mane temps, de les s4parer des pesticides dont Je taux de residus peut etre exactement ; determine. La methode permet egalement une estimation du contenu total en PCBs. Des resultats d'analyses portant sur des oiseaux aquatiques canadiens indiquent qu'un taux elev4 de PCBs est li4 k un taux elev4 de r4sidus de pesticides du groupe DDT. D'autre part, des taux eleves de r4sidus de substances organochlorees ne sont pas necessairement accompagn4s d'un taux 61eve de PCBs. Des recherches supplementaires sur la toxicite et les effets subletaux des melanges de PCBs et de leurs composes individuels seront nec4ssaires. Si on peut demontrer que des substances particulieres sont plus toxiques que d'autres, on aura besoin d'une methode plus sp4cifique permettant l'anal)'se individuelle des PCBs. * Traduit par H. GbssbUhler. SW 373614 STLCOPCB4098989 ti)V 1 ^ r " 1t - ^ 1* " - v i "rt'r'frH r'-'r h 'i :r~- r i f fr - i: H jrYinV^^.... -.y...iw^.^.,, 3- * AW.4-..--.fl 54 L. M. Reynolds Zusarjnenfassung Die Analyse von Pflanzenschutzmittel-Riickstanden in Cegenwari von Polychlordiphenylen Seit 1966 werden Polychlordiphenyle (PCB-Stoffe) in Fischen und Wildtieren, besonders in Europa und Nordamerika, nachgewiesen. Die Verschmutzung der Wasserwege durch Industrieabfalle scheint dabei die Hauptursache von PCB-Riickstanden in der aquatiscben Fauna zu sein. Die PCB-Sic le werden fiir eine Anzahl wichtiger Industrieprozesse gebraucht, aJs Schadlingsbekampfungsmittel werden sie jedoch nicht eingesetzt. Da die PCB-Stoffe in ihrer Struktur und in ihren Eigenschaften den chlorierten Kohlenwassc rstoff-Insektiziden sehr ahnlich sind, konnen sie die gaschromatog: aphische Bestimmung mit Elektroneneinfang-Detektor von chlorieiten Kohlenwasserstoffen und mancher Phosphorsaureester verfalschea. Die PCB-Stoffe konnen nicht nur Riickstandsanalysen beeintracl itigen, sondem sie sind selbst toxische Substanzen, deren Nachweis s ch aufdrangt. Im vorliegenden Artikel vird eine Methode beschrieben, die es erlaubt, das Vorhandensein vcn PCB-Stoffen in Proben nachzuweisen, sie von den Schadlingsbeka npfungsmitteln abzutrennen und die Rlickstande der letztem genau zu bestimmen. Dariiber hinaus gestattet die Methode, den Totalgehalt an PCB-Stoffen abzuschatzen. Riickstandsdaten einiger kanadischcr Wasservogel weisen darauf hin, dass hohe PCB-Konzentrationen nur in Gegenwart hoher Riickstiinde von Substanzen der DDT-Gruppe auftreten. Anderseits sind hohe Riickstande von chlorierten Kohlenwasserstoff-Insektiziden nicht unbedingt von hohen PCB-Konzentrationen begleitet Weitere Untersuchungen iiber die Toxizitat und die sublethale Wirkung von PCB-Mischungen und ihrer individuellen Komponenten sind angezeigt. Sollte es sich dabei ergeben, dass gewisse Einzelstoffe toxischer sind als andere, wird eine Verfeinerung der Nachweis- imd Bestimmungsmethoden fiir PCB-Stoffe notwendig. References Block, S. S.: Insecticidal surface coatings. Soap & Sanit. Chem. 24 (2),138 (1948). Brown*, R. M.: On the toxicity of the "Arodors". Chemist Analyst 36, 33 (1947). Burdick, G. E., E. J. Harris, H. J. Dean, T. M. Walker, J. Skea, and D. Colbv: The accumulation of DDT in lake trout and the effect on reproduc tion. Trans. Amer. Fisheries Soc. 93 (2), 127 (1964). Burke, J. A.: Gas chromatography for pesticide residue analysis; some practical aspects. J. Assoc. Official Anal. Chemists 48, 1037 (1965). Chau, A. S. Y.: Derivative formation for the confirmation 'of endosulfan by gas chromatography. J. Assoc. Official Anal. Chemists 52, 1240 (1969). * Ubersetzt von H. Geissbuhler. iillliiitfltrttittfT ifa-Vr i 11tfff f>''ey iWrrfiiVrrirt^Vn'>- 'H'iii^^'''T iifim" i DSW 373615 t- STLCOPCB4098990 lJ k * s. -> $- * Vi * -.*jk itmk^. , Polychlorobip'aenyls 55 -, and W. P. Cochrane: Cyclodiene chemistry. III. Derivative formation for the identification of heptachlor ef oxide, cis-chlordane, trans-chlordane, dieldrin, and endrin pesticide residues by gas chromatography. J. Assoc. Official Anal. Chemists 52, 1220 (1969) Cochrane, W. P., and A. S. Y. Chau: Note on gas chromatographic identification of heptachlor pesticide residues by derivative formation. J. Assoc. Official Anal. Chemists 51, 1267 (1968). Dubois, K. P.: Combined effects of pesticides. In: Current views on pesticides, p. 35. Symposium sponsored by Foot; and Drug Directorate, Department of National Health and Welfare, Ottawa, 1968; reprinted from Can. Med. Assoc. J. 100 (No. 4) (Jan. 25, 1969). Duffy, J. R., and N. Wong: Residues of organochlorine insecticides and their metabolites in soils in the Atlantic prosinces of Canada. J. Agr. Food Chem. 15, 457 (1967). . Gbeenburc, L., M. R. Mayers, A. R. S nra: The systemic effects resulting from exposure to certain chlorinated hydrocarbons. J. Ind. Hyg. ToxicoL 21, 29 (1939). Gunther, F. A.: Instrumentation in pesticide residue determinations. Adv. Pest Control Research 5, 191 (1962). Hammence, J. H., P. S. Hale, and D. J. Caverly: The identification and deter mination of chlorinated pesticides residues. Analyst 90, 649 (1965). Harrison, R. B.: Residues in wildlife with special reference to endrin. J. Sd. Food Agr. 17, 10 (1966). Hickey, J. J., and D. W. Anderson: Chlorinated hydrocarbons and eggshell changes in raptorial and fish-eating "lirds. Srience 162, 271 (1968). Holden, A. V., and K. Marsden: The exar. lination of surface waters and sewage effluents for organochlorine pesticides. J. Proc. Inst. Sewage Purif. p. 295 (1966). --------- ---------- Organochlorine pesticides in seals and porpoises. Nature 216, 1274 (1967). Holmes, D. C., J. H. Simmons, and J. O'G. Tatton: Chlorinated hydrocarbons in British wildlife. Nature 216, 227 (1967). Hornstein, I., and W. N. Sullivan: The role of chlorinated polyphenyls in improving lindane residues. J. Econ. Entomol. 46, 937 (1953). Jensen, S.: Report of a new chemical hazard. New Scientist 32, 612 (1966). --------- , A. G. Johnels, M. Olsson, and G. Otterlind: DDT and PCB in marine animals from Swedish waters. Nature 224, 247 (1969). --------- , and G. Widmark: Swedish report at the OECD pesticide conference on "Unintended residues in the environment" (1967). Jones, J. W., and H. S. Alden: An acneform dermatergosis. Arch. Dermatol. Syphilol. 33, 1022 (1936). Klein, A. K., and J. O. Watts: Separation and measurement of perthane, DDD (TDE), and DDT in leafy vegetables by electron capture gas chroma tography. J. Assoc. Official Anal. Chemists 47, 311 (1964). Koeman, J. H-, A. A. G. Oscamp, J. Veen, E. Brouwer, J. Rooth, P. Zwart, E. v, d. Broek, and H. van Genderen: Meded. Rijks-faculteit Landbouw- wetenschappen Gent. 32, 841 (1967). Koeman, J. H., M. C. ten Noever de Brau, and R. H. he Vos: Chlorinated biphenyls in fish, mussels and birds from the River Rhine and the Nether lands coastal area. Nature 221, 1126 (1969). Lindquist, A. W., A. H. Madden, H. G. Wilson, and E. F. Knipling: DDT as a residual-type treatment for control of house flies. J. Econ. Entomol. 38, 257 (1945). McLaughlin, J., Jr., J. P. Marliac, M. J. Verrett, M. K. Mutchler, and O. G. Fitzhuch: The injection of chemicals into the yolk sac of fertile eggs prior to incubation as a toxicity test. Toxicol. Applied Pharmacol. 5, 760 (1963). DSW 373616 STLCOPCB4098991 ii 'i 56 L. M. Reynolds Miller, J. W.: Pathologic ch rnges in animals exposed to a commercial chlorinated diphenyl. U. S. Public I. lalth Reports 59, 1085 (1944). Monsanto Co.: The Aroclor compounds, p. 17 (1965). ----------Aroclor plasticizers. Tech. Bull. O/PL-306 (1967). OsadchuCK, M., and E. B. Wanless: Identification and quantitative estimation of aldrin residues in the presence of interfering materials in samples of plant and animal origin. J. Assoc. Official Anal. Chemists 51, 1264 (1961). Peakall, D. B.: Pesticide-induced enzyme breakdown of steroids in birds. Nature 216,505 (1967). Pennixc, C. H.: Physical characteristics and commercial possibilities of chlori nated biphenyl. Ind. Eng. Chem. 22, 1180 (1930). Ratcuffe, D. A.: Decrease in eggshell weight in certain birds of prey. Nature 215,208 (1967). Reichel, W.: Personal communication (1969). Reynolds, L. M.: Canadian wildlife service reports (Unpublished data) 1968 1969). ---------- Pesticide residue a- alvsis in the presence of PCB's. 4th Annu J W. Can. Seminar on Pesticide Residue Analysis. Winnipeg, June 2--4 (19 >9 a). ----------Polychlorobiphenvls (PCB's) and their interference with pesticide residue analysis. Bull. Environ. Contamination Toxicol. 4, 128 (1969 b). Risebrouch, R. W., P. Rdeche, S. G. Herman, D. B. Peakall, and M. N. Kirven: Polychlorinated biphenyls in the global eco-system. Naturt 220, 1098 (1968). . ----------, P. Rieche, and H. S. Olcott: Current progress in the determination of the polvchlorinated biphenyls. Bull. Environ. Contamination Toxiiol. 4, 192 (1969 ). Robinson, J.: Residues of organochlorine insecticides in dead birds in the United Kingdom. J. Chem. Ind., p. 1974 (1967). -Roburn, J.: A simple concentration-cell technique for determining small amount'; of halide ions and its use in the determination of residues of organochlorine pesticides. Analyst 90, 467 (1965). Samuel, B. L., and H. K. Hcidces: Insecticide screening methods for organochlo rine and organophosphate insecticides in foods and feeds. Residue Reviews 17, 35 (1967). Sans, W. W.: Multiple insecticide residue determination using column chromatog raphy, chemical conversion, and gas liquid chromatography. J. Agr. Food - Chem. 15, 192 (1967). Schmidt, H., and G. Schultz: Patent for manufacture of pentachlorobiphenyl. Ann. 207, 338 (1881). Schwartz, L.: An outbreak of halowax acne ("cable rash") among electricians. J. Amer. Med. Assoc. 122, 158 (1943). ----------, and F. A. Barlow: Chloracne from cutting oils. Public Health Rept. 57, 1747 (1942). ----------, and S. M. Peck: Occupational acne. N. Y. State Med. J. 43, 1711 (1943). Schechter, M. S., S. B. Soloway, R. A. Hayes, and H. L. Haller: Colorimetric determination of DDT. Ind. Eng. Chem., Anal. ed. 17, 704 (1945). Simkiss, K.: Calcium in reproductive physiology. New York: Reinhold (1967). Stickel, L.: Wisconsin hearings. Science 163, 551 (1969). Sullivan, W. N., and I. Hornstein: Chlorinated polyphenyls to improve lindane residues. J. Econ. Entomol. 46, 158 (1953). Thompson, J. F., A. C. Walker, and R. F. Moseman: Study of the performance of gas chromatographic columns under severe injection loading. J. Assoc. Official Anal. Chemists 52, 1251 (1969 a). --------------------------------- Evaluation of eight chromatographic columns for chlon- nated pesticides. J. Assoc. Official Anal. Chemists 52, 1263 (1969 b). ....... .............- -------- ---- -------------- ---------- -- DSW 373617 STLCOPCB4098992 a _ ..._o.,.,........ . .* -**** -XA, . V. -t - Dfi. .i; --* *> Polychlorobiphenyls 57 Tsao, C. H, W. N. Sullivan, and I. Hornstein: A comparison of evaporation rates and toxicity to house flies of lindane and lindane-chlorinated polyphenyl deposits. J. Econ. Entomol. 46, 882 (1953). V. S. Food end Drug Administration: Pesticide Analytical Manual, VoL I and II, revised annually. Washington, D.C. (1969). Van Teel, N.: Improvement of the residual toxicity of DDT solutions by the addition of coumarone resin. Bill. Entomol. Research 43, 413 (1952). Vermeer, K-, and L. M. Reynolds: Organochlorine residues in aquatic birds in the Canadian Prairie Provinces. (In preparation, 1970). Welch, R. M., M. Levlve, and A. H Coxxey: Effect of chlorinated insecticides on steroid metabolism. In: Chemical fallout. Proc. 1st Rochester Conf. on Environmental Toxicity', Univ. of Rochester (1969). Widmakk, G.: IUPAC commission on methods of pesticide residue analysis. J. Assoc. Official Anal. Chemists $% 1069 (1967). Wiencke, W. W., and J. A. Burke Derivatization of dieldrin and endrin for confirmation of residue identity J. Assoc. Official Anal. Chemists 52, 1277 (1969). Zwnc, G. (ed.): Analytical methods for pesticides, plant growth regulators and food additives, vol. I, p. 1. Few York: Academic Press (1963). Manuscript received 20 January 1970; accepted 17 February 1970. ' -idane ^"minee } Assoc. I " chlori- DSW 373618 STLCOPCB4098993