Document 9JYL5w3Y73Kng0X88J8VaeJDp

Dtral rt of (1957) 1927) tatukky, :t. To*., spt. No. 42, (1915) Mats Spedrometric Identification of Mircx Residues in Crude Extracts and in the Presence of Polychlorinated Biphenyls * by Solanc Uk, Chester M. Himel, and Toria* Dnuts PfMrlmfKl of Entomology .. Vniveruiy of Georgia Athena, Ca. JMOJ Abstract Maes spectrometry using 10 eV electron impact was employed in the analysis of Mirex residues in the pres ence of PCB's. Low voltage MS is a useful complementary tool for positive identification of both Mirex and PCB residues In organic extracts. The MS patterns of pure 2,4,5,2 *,5'-pentachloro and 2,5,2',5'-tetrachlorobiphenyl and of mixtures of PCB isomers in Aroclor 1254 and 1260 are simple and distin guishable from the MS spectrum of Mirex. There is no overlapping of ions (molecular or fragment) of PCB's with those of Mirex. Each pure chlorobiphenyl has its base peak at the molecular ion (M) in contrast to Mirex which exhibits the intense M/2 ion (m/e 270) as base peak. In mixtures of PCB's, peak clusters, 34 mass units apart, correspond to C1jHjq_)(C1x where x 4 to 9. A mixture of 500 yg of Mirex and 500 yg of Aroclor 12S4 added to 5 g of pork fat was detected by low voltage MS in the crude recovery extract. A probe temperature of 60 or 100C was satisfactory for identification of each component in the mixture. However, when the probe temperatures reached 1U0C interference bv extraneous peaks became more serious and most of Mirex and PCB's were depleted. The presence of these extraneous peaks indicated that the treatment of fat extracts by either 24-hr stirring in concentrated H2SO4 or by a single florisil column chromatography did not remove fatty resi dues satisfactorily. Thus, "standard" techniques of Mirex residue cleanup may be inadequate. Introduction The world-wide industrial use of polychlorinated biphenyls (PCB's) (4) has made them major potential pollutants. They concern environmental scientists be cause their toxicological properties are of a chronic nature. In addition, their chemical characteristics a Supported by a grant from the Agricultural Research Service 660B (PP-ARS 539). k Correspondence should be addressed to CMH. 97 I'ull. iin of ConttrniattioB A ToniwWy. VL C. No. X, Iff* <) fry SpriBger-Vtrbg ftw Yosfc Int. MONS 085477 create substantial difficulties in pesticide residue analysis (2, 6, 9). For example, some of the GLC peaks of Aroclor 12SM and 1260 overlap or coincide with Mirex peak when analyzed on QF-1 4 DC-200 column O) or on QF-1C or on 0V-210 (our laboratory). Mass spectrometry is a means of positively distinguishing Mirex from PCB's. Mass spectral patterns of Mirex and Kepone have been reported (10). The present paper reports the progress of our study of analytical method ology pertinent to the confirmation of Mirex residues in biological samples containing PCB's. Experimental Sample preparation One, 5, 20, and 500 Mg of Mirexe and of Aroclor 1254* were added to 5 g of pork fat. The sample was ground in a mortar' and pestle with the aid of purified sand in 4 portions of 5 ml of benzene. The benzene fractions were combined and treated with an equivolume of concentrated H^SO^. A modified Craig countercurrent distribution celled; was used for the HjSOji treatment by phase partition. Overnight stirring of the benzene extract with HjSOu was also tested. The benzene layer was then separated* filtered over glass wool-Na2S04 into a conical-bottom centrifuge tube, and carefully concentrated with a stream of dry air. When the volume si abuut C.5 ml, the benzene was carefully washed into a Pasteur pipette that had the tip cut and sealed off to allow the concentration to less than 5 yl. The residue was transferred into a melting point capillary tube which was then cut to 20 mm to fit the mass spectro meter probe. The tube was plugged with spectroscopically inert, fine mesh glass wool. Other samples of pure Mirex, pure chlorobiphenyl isomers, or mixtures of Mirex and Aroclors were put directly into glass capillary tubes cut to 20 mm length, and plugged with spectroquality glass wool. Mass Spectrometry A Bell-Howell/CEC model 21-490 mass spectrometer was used. All samples were introduced into the ionization c 3% QF-1 on Chromosorb G, AW-DMCS, 80-100 mesh. ^ 3% OV-210 on Supelcoport, 100-120 mesh. Poly Science Corp., Evanston, 111. 60201 ^ Monsanto Co., St. Louis, Mo. 63166 98 MONS 085478 sidue -d with . (3) ; :Mng 'x and 'r nethod:idues in slor was ;rifled ene .volume current etment anzene layer SOu Jlly volume ra shed jg*led T ~The illary i spectrocopies))y ' *jre Mirex, ; x and tubes .lity rieter ionization chamber by means of a probe that can be heated independ ently of the source. Mass scans were taken at various probe temperatures for reasons discussed below. The source temperature was at 240C and the sample pres sure at about 1.5 x 10"* Torr. The bombarding energy was set at 10 eV since it gives simple mass spectra of Mirex suitable for identification (10). The mass spectral pattern of Mirex was used as reference in combination with isotopic ratios due to 37C1 to facil itate peak identification. Results and Discussion The fragmentation pattern of Mirex has been dis cussed previously (10). The mass spectral property of 2,4,5,2',S*-pentaehlorobiphenyl is presented in Figure 1. Unless otherwise specified, ion clusters due to chlorine isotopes C35C1 and 3'Cl) are identified in the figures and discussion only by the m/e of parent peaks (P), i.e., peaks due to *5C1, and molecular ions are referred to as M. Under 10 eV electron impact, 2, S,2',5'-tetrachlorobjphenyl showed breakdown process similar to that of 2,4,5,2',5'-pentachlorobiphenyl. Both compounds produced 2 fragments by losing 1 Cl (M-Cl fragment) and 2 C3 (M-2C1 fragment). Whether the molecule lost 1 and 2 Cl simultaneously or the M-Cl ion lost its second Cl ([M-Cll - Cl) is not clear. No other ion of significance was observed. Other cMoic.tiphcr.yl isomers (hexachloro, hcptochlorcbiphcnyl, etc.) apparently have similar fragmentation property as mass scans of the Aroclor 1254 or 1260 produced no peaks other than those corresponding to the molecular ion, the M-Cl, and the M-2C1 ions of the chlorobiphenyls. Biros, Walker, and Medbery (2) reported comparable results of trichloro, hexachloro, and heptachlorobjphenyl if peaks other than the aforedescribed ones are disregarded. Though not mentioned, the authors obviously used much higher electron energy. The relative abundance of molecular ions of various isomers produced by successive mass scans with a gradual increase of probe temperatures indicated that the major components of Aroclor 1260 are tetrachloro, pcntachloro, hexachloro, heptachloro, octachloro, and a small amount of nonachlorobiphenyle. Under similar conditions, Aroclor 1254 produced similar mass spectra but negligible amounts of heptachloro and the absence of octachloro and nonachlorobiphenyls. When a mixture of Mirex and Aroclor was analyzed, their mass spectral patterns (produced by 10 eV electron impact) were completely separated from each other as 99 HONS 085479 00| ofofH 70|- 55 z UJ 60 > *0 _J 30 Ut OC 20 M-2CI M-C1 o r in |Miniiii|iiinnir|rm|iin|iul|lllljini|iiii|ini|irn|mi^iiljlMHnii|iinftnijHMptT|fTTHIHI|MiTfffn^ 220 230 240 250 260 2*0 260 290 300 310 320 330 o> m /o * *> c Figure 1. Mass spectrum of 2,4,S,2*,S'-pentachlorobiphenyl under 10 eV electron bombardment. M is the molecular ion. there are no other peaks of significance. c. gg fcc % 1 g3 *w S OH 9 N 8 s N shown in Figure 2. There were no peaks of importance pertaining to any of the components below m/e 200. Further differences in fragmentation behavior between Mirex and PCB's are that the former has its characteristic base peak at M/2 cluster (m/e 270) and quite small molecular ion peak CM), while PCB's have intense molecular ions as base peaks; the latter property is due to the 2 benzene rings (8). Recovery of a mixture of 500 Mg of Mirex and 500 pg of Aroclor 1254 added to 5 g pork fat was easily detected by low voltage MS in the crude extract after it was treated with concentrated H?S0y . The most favorable probe temperatures for obtaining good, clearly recognizable mass spectra were at 90-100C as illustrated in Figure 3. Despite numerous extraneous peaks from the fat particularly at m/e below 200, mass scans in this temperature range showed minimum interference with ions of Mirex and/or PCB's except at the following masses: 237.5, 241.5, 256.5, 270.5, 271.5, 360.5, and 361.5. Nevertheless, peak clusters due to Cl isotopes at the aforementioned m/e regions remained clearly recognizable. The presence of extraneous peaks proved that the treat ment of pork fat extracts with H2SO1J either by simple binary phase partition or by continuous overnight stirring did not eliminate organic impurities. One sample from florisil column chromatography showed similar incomplete cleanup of organic contaminants. The results presented here show the feasibility of concenii'atiug pebtiviuc itsidue extracts, to yl-velume for positive confirmation by mass spectrometry. For relatively stable compounds with low vapor pressure such as Mirex, phe sample in organic solvent such as benzene can be introduced into the glass capillary tube plugged with glass wool 1 to 2 weeks in advance before mass epectrometric testing. Thus, the sending of prepared samples to other laboratories where a mass spectrometer is available could be done without affecting the analysis. Thus, crude residues can be quantitated by EC-GLC then confirmed by mass spectrometry. Analytical sensitivity remains a problem since we were unable to detect the presence of Mirex and/or Aroclor 1254 in sample extracts containing 1, 5, or 20 vg of each chemical. However, a standard preparation of 3.5 vg of pure Mirex gave highly satisfactory mass spectra during 5 repeated scans for about 5 min. De pending on the types of biological materials, better techniques of sample cleanup and of concentration to Ml-volume would certainly improve the sensitivity of mass spectrometric analysis. Hutzinger and Jamieson (5) reported the analysis of 1 ppm of 2,6-dichloro-4nitro-aniline (apparently a total amount of 100 Mg) in 101 MONS 085481 MONS 0 8 5 4 8 2 Figure 2. Mass spectra of a mixture of Mirex and Aroclor 1260 under 10 eV electron impact. Tetrachloro, pentachlorobiphenyl, etc., are referred to as 4 CBP and 5 CBP, etc. Fragments are designated as M.W.-C1. The Hexachlorocyclopentalene ion of the Mirex fragment is called Mirex/2 at ra/e 270. There are no peaks of importance below m/e 200. OD numerous peaks below m/e 200, but none pertain to Mirex or Aroclor. crude peach extract with low resolution mass spectrometry. Low voltage mass spectrometry (10 eV or lower, i.e., slightly above ionization potential) should be a very useful complementary technique in the analysis of multiple pesticide residues as it produces, rather simple spectra. We also found that, under 10 eV, Dieldrin and DDE produced mass spectra easily distinguishable from each other as well as from those of Mirex and PCB's (unpublished data). Low voltage mass spectrometry in organic analysis has been briefly discussed by Robot (7) who gave references to further details. Acknowledgment s We are indebted to Dr. R. E. Lowins,'Dr. Fairwell Thomas, and John Craig, Department of Biochemistry, University of Georgia, for their kind cooperation in making available the mass spectrometer. Our appreciation is also addressed to Dr. R. G. Webb, Southeast Water Laboratories, Athens, Ga., for providing pure samples of tetrachloro, and pentachlorobiphenyls, and to Miss Carol Lord of our laboratory for her technical assist ance . References 1. RPR07.A, M., TNSCOr, M. H BDWMAM t * r Residue Reviews, Vo). 30, p. 25 (1967), Springer- Verlag, New York 2. BIROS, F. J., WALKER, A. C., and MEDBERY, A. Bull. Environ. Contam. Toxicol. 6, 385 (1970) 3. GAUL, J. and CRUZ-LAGRANGE, P. Private Communication. Food and Drug Administration, New Orleans, La. (1971) 4. GUSTAFSON, C. G. Environ. Sci. Tech. 10, 814 (1970) 5. HUT2INGER, 0. and JAMIESON, W. D. Bull. Environ. Contam. Toxicol. 5, 587 (1971) 6. REYNOLDS, L. M. Bull. Environ. Contam. Toxicol. 4, 128 (1969) 7. ROBOZ, J. Introduction to Mass Spectrometry, p. 321 (1968), Interscience Publishers, New York 8. S1LVERSTEIN, R. M. and BASSLER, G. C. Spectrometric Identification of Organic Compounds, p. 13 (1967), John Wiley 6 Sons, New York 9. SNYDER, D. and REINERT, R. Bull. Environ. Contam. Toxicol. 6, 385 (1971) 10. UK, S., HIMEL, C. M. and DIRKS, T. F. Bull. Environ. Contam. Toxicol. 7(4), (1972) 1M MUNS 085484