Document EvQM8N1QeJLzbNJpJnNarjxmx

Chamoephere No, 1, pp 109 - 119, 1973. PergMOn Preea. Printed in Ort Britain. c * b \ pohhatios of polycklorimatsb BiBaizonnum (rcir*) FROM THB PYROLYSIS OF PCBa Hana Rudolf Bueer and Hane-Paul Boaehardt Swiae Federal Reaearch Station CH - 6020 Wadenawl1, Bui tMiland and Chriatoffar Rappa Dapartnent of Organlo Cheaietry, Oniverelty of Data S - 901 87 tlnaa, Sweden (Sacalvad in UX 3 January 1973; accaptad for publication 5 January 1978) Introduction Polychlorinated blphenyla (PCBa) ara lnduatrlal ohealoala of vlda uaa and now fcnovn to ba widely diatributed In tha anvironment. Tha comarolal producta ara coaplex 1iturea of at laaat 80 dlffarent aubatanoaa, Due to thalr axtraaa atabillty in eoolegical ayataaa, tha uaa of PCBa la now raatrlotad to aaalad ayataaa like oapaoitora and tranaforaara. However, prior to 1972, PCBa due to tbeir good themal and chaaloal atabillty were uaed in a variety of other applloatlona; in particular tha uae aa heat tranafar flwlde and caatlng waxaa waa aaaoclatad with tha themal atabillty. Vary cloaely related to tha PCBa are tha polybroalnated blphenyla (PBBa), eoaa of which have been uaad aa flaaa retariante. Of apaclal intareat la Fire Mae ter BP-6 or Fire fleeter fT-1, which wara ahown to conelet nainly of 2,4,3,2*,4',5'-hexabroobiphenyl.* Intoxication with thla PBB hae bean directly aaaoclatad with a wlde-apread agricultural calamity In Michigan (OSA) In 1973-74.2 109 MOMS 213802 S.10 Do. 1 Part of the PCBe found in tho environment has escaped open applications and closed systems> Borne probably during incineration and burning. Karl seen and Rosen ^ have studied the thermal stability of PCBe; laboratory experiments and theoretical calculations showed that at temperatures higher than flOOC PCBe are thermodynamically unstable. The eiain decompoeition product* were found to bo C(a), CO, CO^, HC1 and Cl^. In 1970 Voa at al.^ identified polychlorinated dibentofurane (PCUFs) as toxlo impurltiss in European PCBs at the ppm-level, and the asm* has alBO been reported for Amerioan and Japanese PCB*,^'^ The toxlo effeots of the PCBFa are very alBllar to thoa* reported for the polyohlorinated dib*nzo-p-dloxlne (PCDDs), In both cases, ths 2,5,7,0-tetraohloro compounds are ths moat toxlo laomera, having LD^-valusa in the range of 1-10 jig/kg (guinea pig)7/' S like PCDDs, the PCDP* have been found to lncreaee cytoohxoa* P-450 activity and a number or 89 drug metabolizing enzyme* ' Korita *t al,10 have reported inoreaeed amount* of PCDFa in a eample of POO eealed with air in a glace tubs and heated at J00C for two week*. In this paper, w* report on the formation of PCDFe fro the pyxolyei* of a coi**rcial PCB (Aroclor 1254) in the presenoe of air in a eealed quarts tube. In order to study the msohsniam of the reaotion, we have included individual PCB icorner* ae model compound*. Experimental EfctjrJglq 2,6,2',6'-T*tra- , 2,4,5,2',4'.5'- and 2,4,6,2',4',6'-h*xachlorobiphenyl were prepared aooording to loiown method*.11 Aroclor 1254 mnd Fireaaater BP-6 were from commercial souroea. Analyala of the two hexachlorobiphenyls on glaee capillary column* showed single peak*, pentaohloroblphenyla being present ae impurltiea at levels of 1)6 in 2,4>5,2',4'>5'- and of 0,J)( in 2,4,6,2',4'l6'-hsxahlorobiphanyl, Mona of tha individual PCB iaoaera or Aroclor 1254 did oontain dstsotabls quantities (< 0,001-0,01)6) of PCDFs, polychlorinated benzenas, dlphanylathera or blphenylols. Standard solutions of ths PCBe vers prepared at concentrations of 1 and 10 mg/ml in n-hexane or benzene. t t No. 1 Micros of the e micr' placed a alig' heat ti, appro x 1)7.5 1 Tt. to temp in an 1 ampoul" heat th the tlm IS broken dilutio affect* and 50^ GC-H3 J An electro, 25 and ' tha ion 145. 5 2/mln tively. HONS 213803 Ho. 1 111 MloropvTolvsle Quart* mint-ampoules were Dade from quartz tubing using a gas-oxygen burner; the length* of these ampoules were 25-3O mm, the Internal disasters },4 n and the volumes 0,} ml. being a Dloro-eyrlnge, 10 jil of a standard solution (corresponding to 10 or 100 tig of coapound) was placed into the tip of an ampoule. After couplets evaporation of tha aolvant with the aid of a alight vacuum, the ampoules were cloeed uelng a very asall flama; oare wae taken not to heat the tips containing the oompounda, The volume of 0,3 si air in an ampoule was kept approximately constant, ttw molar ratios of PCD to oxygen In an ampoule wars about Ii75 and It7' 5 for amounts of 10 and 100 tig of PCB, respectively. The quarts mini-ampoules ware placed In an electrical oven (sax, temp, 1000C) preheatsd to temperatures ranging from 550 to S50C, The temperature vaB controlled with a thermocouple In an ampoule placed next to tha samples. Heating periods (time between placing a aaaple ampoule into the oven and removing It) were kept conet ant at GO aec. The time required to o heat tha mini-ampoules up to a temperature within 20 C of the preset value was around 55 sec, tha time of a sample at aotu&l pyrolysis temperature thus around 5 sec. After removing and immediate cooling to room temperature, the tips of the ampoules were broken end 200 til of benzene added, A 2-pl aliquot wee weed directly for analyele. Further dilution was required for saoplee with low decomposition of PCBs to overcome overloading effeote of the GC-KS system. Some samplee were cleaned-up on an alualna-microcoluan, uelng 2}t 12 and 50jt methylenechlorlde in n-hexane to separately elute PCBe and PCDFb, Analyees were carried out on a Finnigan 4000 quadrupole GC-KS Instrument equipped with elsotron-lmpaot ion-source and a Finnigan Gill data-Byetea, OV-17 glass oapillary columns of 25 and 50 a length (0,36 mm IP) were coupled via a platinum Interface leading directly Into the Ion-source, The oolumn conditions were as follows) 100, 2 sin Isothermal, 15/"ln to 145, 5/mln to 230C for the 25 OV-17 column; 100, 2 min leothermal, 20/mln to 160, 2/min to 230 for the 50 OV-17 column; helium carrier gae at 0,60 and 1-40 at, reepec- .o tlvely. Sample* were eplltlasely introduced with the vaporizer at 26O C, MONS 213804 I 112 Ho, 1 All samples were directly analysed on the 25 m OV-17 column by repetitive scanning of complete meee epectra (m/e 55-460 In 1,4 eec) ualng the data system. After data acquisition, the mase epectra (up to 1000) wore searched for FCBe, PCDFe and othor possible reaction producte by running and plotting maeB ohromatogr&me For optimal isomer separation some of the eamplee were reanalysed on the 50 OV-17 glass capillary column. Maes specific detection (mase fragmentogrephy) of PCDFs was carried out by oontlnuouely monitoring molecular Ions at m/e 202 , 236, 270, 304, JJ8 and 372 for the mono- to hexachloro compounds. Results pyrolvals of 2,4.5.21.41.5'- and 2.4,6.2*.4'.6'-hexachloroblphenyl In Tables 1 and 2 ws report the rsBults of the pyrolysis of two hsxachloroblphsnyls at temperatures ranging from 550 to 850C. As indioatsd, 37j( of the 2,4,5,2',4',5'- and 22JC of ths 2,4,6,2',4',61-hexachloroblphenyl dscompOBed at 550C. Both compounds were decomposed to 'p' 9Q?t at 650C. Complete deetructlon occurred at 700C and above. Ho isomerisation of these hermohlorobiphenyle or dechlorination to lower chlorinated biphenyls was observed In this temperature range. Table 1 Pyrolysis of 2.4.5.2'.4'.V-haxachloroblphenyl flOO ag) in quarti mini-ampoules Temperature Decomposition ) - PCDFs formed4, (jt) tetra-CDF penta-CDF 550 57 1.6 0.2 + 0.15 600 8J 0.5 0.35+ 0.1 650 97 0.2 1.1 + 0.1 700 > 99.99 < 0.01 < 0.01 850 > 99.99 < 0.01 < 0.01 e 1 tetrs- and 2 penta-CDF isomers formed Tetra- and psntachlorodibemofursnB (tetrm- and penta-CDFs) were formed at temperatures of 550-fi50C, The levels were 0.2-1.6jC and 0.1-0.5}tfor 2,4,5,2',4',5. and 2,4,6,2*,4' ,6'hexachloroblphenyl. At temperatures of 700C and above, complete destruction of these FCDFe occurred, No dl-, trl- or hexm-CDFB were formed at any of these temperatures. MONS 213805 5. 1 or i on, led the o at /i of .d to these ts ule ? i k -< j * v < < Ko. 1 113 Table 2 Pyrolysis of 2,4.6.2' ,4* .fi'-hexachlorobiphenyl (lOQ In quarts minl-anpoulea Temperature Decomposition ... (*> PCDFe formed* (jt) tetra-CHF oenta-CDF 550 22 0.5 0.1 600 90 0.5 0.3 650 94 0.35 0.35 700 > 99.99 < 0.01 < 0.01 850 > 99.99 < 0.01 < 0.01 + 1 tetra- and 1 penta-CDF leoaer fonaed The formation of pants- and hexacblorobensene was observed with both PCfi isomers at temperatures of 650C (0.05^) and 70OC (0.5-1^). The lower chlorinated bensenae would have escaped detection (elution from capillary column prior to starting of MS soanning). Both, penta- and hexaohlorobenrene were decomposed at 850C, tures 'Ws Figure 1 Partial mass fragmentograme (50 a OV-17, o/e J04 and 338) Bhowing elution of tetra- and penta-CBPs in a) pyrolyaed 2,4>5,2',4',5- and b) pyrolyaed 2,4,6,2*,4',6'-hexachlorobiphenyl. 1 - 2,3,7,8-tetra-, 2 preauaably 1,3,7.9tetra-CDF. HQNS 213806 114 lio, 1 In Figure 1 maBB frignentogi-ams of both hexachlorobiphenyls pyrolyied at 600 show the elution of tetra- and penta-CDFa on the 50 m 0V-17 glass capillary column. The pyrolysis of 2,4,5>2',4',5'-hexachlorobiphenyl did yield one tetra- and two penta-CDF isomers. By comparison with an authentic standard, it oould be shown that the Cl^-leooer is the very toxio 2,3,7,B-tetra-CDF.'Consequently, no rearrangeaent ie Involved in the cyolltatlon forming this lsooier. In the case of the two penta-CDFe, it is evident that a rearrangeaent oust be Involved. The nonrearrangad product should be the l,34,7.B-penta-CUF, but addi tional work is needed to establish the Identity of these penta-CDFs. 4 * * * ( 2 Isomers The pyrolysis of 2,4,0,2',4',6'-hexeohlorobiphenyl did yiold one tetra- and one pentaCSF laoaer. The Cl -isomer presumably is the 1,3,7,9-tetra-CLF, but this cannot be oonfirmed 4 due to laok of an authentic standard. The pente-CDF 1b forned in a rearrangeaent and its truoture le still unknown. Do additional PCDFe wars formed in a mixed pyrolysie of the two hexaohloroblphenyls Indiosting only intramolecular oyclisatlon processee. PytoIybIb of Aroolor 1254 Aroolor 1254 1* a commercial PCB containing tri-, tetra-. panta- and hexachloroblphenyle as the major conetituonte. CC-HS analysis using the 50 a 0V-17 glass capillary oolumn showed the presence of the 2,4,5,2'4*,5'- substituted isomer as one of the major HONS 213807 the t of i i en talinnad IB 1 two I i lor No, 1 115 hexachlorobiphenyle. No 2,4,6,2',4*,6'-hexachlorobiphenyl ti detected (< ljt relative to th 2,4,5.2',4*,5'-iBomer). Table J Pyrolyala of Aroclor 1254 in quartz mlnl-aapoulss Temp. Amt. Decomp. Jlsl ImI (*) mono* PCDFa formed* (*> dl- tri- tetre- 550 100 M 10 12 60 600 100 M 10 45 90 650 100 10 90 98 700 100 >99-9 M 10 >99.9 850 100 >99.9 0.25 0.02 0.10 < 0.01 < 0.01 0.01 < 0.01 < 0.01 < 0.01 0,65 0.40 0.40 0.10 0.02 0.12 < 0.01 < 0.01 < 0.01 0.90 0.70 0.70 0.25 0.16 0.25 < 0.01 < 0.01 <0.01 0.75 0.70 0.60 0-35 0.25 0.12 < 0.02 < 0.02 < 0.02 + oombined value* of all isomers of a PC DP oen ta- 0.20 0.15 0.12 0.05 0.12 < 0.02 < 0.02 < 0.02 < 0.02 In Table 3 the results of the pyrolysis of Aroolor 1234 In th temperature range of 550-fl50C and levele of 10 and 100 *ig are reported. At 550-6500, the decomposition at the 100 )ig-level ranged from 12^ to 90j(, at the 10 pg level from 60 to 98jt. Aroolor 1234 was oompletely deetroyed at temperatures of 700C and above. These doooapoeltlon values are based on the hexachlorobiphenyle present. The deooapotltlon of the lover chlorinated biphenyls was eoaevhat higher than the stated values. No penta- or hexachlorobenzena vaa detected at any temperature. Again, lover chlorinated benzenes vould have escaped detection. Mono- to penta-CDFe vers found at 55O-650C at levelB ranging froa 0.1-0.9?t. The maximum amounts found were at 550C. Taking Into consideration the amount of PCB recovered after pyrolysis, the total yield of PCDFe ranged from 3-25^. Xn addition to PCDFe, polyohlorlnated biphenylole ae possible precursors vers Identified at levels of approximately one fifth of the aaount of PCDFe present. In Figure 2 asss fragmentograms of pyrolyzed Aroolor 1234 (10O >ig at 600, cleaned-up sample) on * JO e OV-17 glass capillary column are shown. The elution of up to 30 major and more than 23 minor PCDFe can be observed. They range from the mono- (all 4 theoretically possible isomsrs) to the penta-CDFs. Since 2,4,5,2*,4',5'-hexachlorobiphenyl Is one of the HONS 21380 9U min 3S 25 15 5 Jlgura 2 Hut fragmeatograns (50 07-17* 202, 2J6 , 270 , 304 and J3fl) showing elution of (right to left) mono-, di-, tri-, tetra- and panta-CBPs from pj-rolfaed Iroclor 1254 (600C). Sensitivities* mono- and dl-: 2 V; tri- aad tatra-: 1 T{ penta-CDF: 0.5 T. Peak Identifications: 1- 2,3.7,8-tetra-; 3 and 4- penta-CDFa from 2*4,5,2' ,4' tS'-hexachlorohiphenj-l,) 2- position where tetra-CDP frou 2,4l6,2',4',6'-hex*clilorofclphenjrl would eluts. y <* 3 3 a <a. A 3a 8* 3* 3o> B \\Jj* ** U *<-* *c- o o HONS 213809 t r i - and te tr a - : 1 V ; penta-CSFs 0 .5 V. Peak Id e n tific a tio n s : 1 - 2, 3 ,7 ,8 - te tr * - ; 3 and < - penta-CDFs fr-ra 2 ,4 .5 ,2 ',4 ' ,5 '-h e x a e h lo ro b ip b e n y l; 2 - p o s itio n where tetra-C E F from 2 ,4 .6 ,2 * ,4" .fi'-hexaciilorotiipheTiyl No. 1 3 3 H> sj No. 1 117 main hexachlorobiphenyle in Aroclor 1234, it is of Interest to confirm the presence of 2,3,7,8- tetra-CDF end of the two psnta-CDFa in this oooplex nlxturs. The 2, J, 7,8-tetra-CHF ie not completely separated on this column from another tatra-CD? Isomer. The tstra-CDP formed from 2,4,6,2',4'6'-hexeohlorobiphenyl was not found to be present. In oaeo of the major leonera, the diaorete PCDfa could be identified by running eoaplete El mass epeotra. All mass spectra showed intense molecular ions with the expected ion clustering due to the chlorine isotopes and the characteristic fragmentation for FCDFa with formation of M+-C0C1 and M+-C0C1-C12. Pyrolysis of additional compounds Additional? ws have studied the pyrolysis of 2,6,2',6'-tstraehlorobiphenyl and of Flremaatar BP-6 (mainly 2,4,3,2',4*iS'-hexebromobiphenyl). Tatraohlorobiphenyl formed at 3500 ona di- and on# tri-CDF isomer at lavala of 1.6 and 2.5)t, respectively. The tri-CDF must be formed in a rearrangement. At 700C, the tetrechloroblphenyl was completely destroyed and no PCDFa ware found. Flremaster BP-6 was also completely destroyed at 700C, but at 600C new compounds ware formed, one of which could be a tatrabromodibensofuran (M+-480, Br.) 4 major fragment H^-Br^i other fragments H+-Br and M+-COBr), In analogy with the corresponding PCB isomer, it seems most likely that it ia the 2,3,7,8-tetra-BDF, The emount formed is ebout equal to the amount of tatra-CDF formed from the corresponding PCB. Conclusion! , The results reported in this peper suggest that uncontrolled burning of PCBe can be an important environmental eouroe of the hazardous PCDFs. Although PCDFs have not yet bean found in the environment, they have been identified in fly ash from munlcipel incinerators end industrial heating facilitlea.1^'1^ Moreover, Nagayama et sl.1^ have found highly inoreaeed levels of PCDFa In the PCB associated with the Yuaho accident in Japan in 1966, and on* of 16 the main components was sleo the 2,3,7,8-tetre-CDF. In spite of being partly banned, large volumes of PCBe are still In use, specially in trxneformera and capacitors. In Sweden the emount now in use in oapacltore is about 2300 a duller amount applies for Switaorlnnd aooording to a aurvay by tha Swlaa Federal "'** Public Health. It la evident that all disposal of PCBe should be carefully HONS 213810 na No. 1 controlled In Order to avoid acoldental fornatlon of hazardous PCDFs. Our investigation also Indicates that special processes May be associated with occupational health risks due to formation of FCDFe. One of these processes is the veiling of leaking syeteae containing PCBs. Another prooeee is the casting of iron using PCB oontainlng casting waxes. I Accidental burning of naterlei impregnated with 2,4,5,2'>4'.5'-hexabroaobiphenyl say yield gases containing the very toxlo 2,3,7,8-tetra-BDF. Consequently, the ues of this ooopound as a fleas retardant should be oarefully reoonsidered. References 1. K. Andersson, A. HorstrBn, C. Rappe, B. Rasauson and B, Swahlin, Biy. Quality and Safety. Supp. Tol. Ill, p. 798 (1975). 2. Anonyaoue, Chen. Ena. Neve. February 24, p. 7 (1975). 5. L.Karleson and . Rosen, Chsaloa Sorlota. 1, 6l (1971). 4. J.C. Voa, J,H. JCoeaan, H.L. van der Haas, H.C. tan Noever de Brauw end R.H. de Toa, Fd. Coeast. Toxlool.. 8, 625 (1970). 5. J.A.C. Roaoh and I.U. Poaerantx, Bull, l&tvlron. Con tan. Toxlool.. 12. 358 (1974). 6. C.W. Bowes, M.J. Mulvihill, B.R.T. Siaoneit, A,L. Burllngaaa and H.W. Rleebrough, Nature, 556, 305 (1975). 7. J.D. McKinney, K. Chae, B,B. Gupta, J.A. Moore and J.A. Goldateln, Toxlool. Aool. tbaraaaol.. 56, 65 (1976). 8. S,B. McConnell, J.A. Moore, J.K. Haeeaen and M.W. Barrie, Toxicol. Appl. Fhsrann^.. 52. 146 <1976). 9. A. Poland, I. Glower, A.8. Kende, J. Biol. Chee.. 251. 49J6 (1976). 10. M. Morlta, J. Nakagawa, X. Aklyaaa, S. Mimure and N. lsono. Bull. Environ, Cnnt- Toxigal., 26, 67 (1977). 11. 0. Hutslnger, S. Safe and Y. Zltko, The Chealetrr of PCBs. CRC Preee, Cleveland, Ohio (1974). 12. H.R. Bueer, J. Chroaatoar.. 107. 295 (1975), HONS 213811 No. 1 119 1J. K. 011a, P.L. Vantteulen and 0. Hutzingcr, Cheapaphere. 6, 455 (1977). 14. H.R. Busar end H.-P. Boaahardt, Mitt. Get. Labanaaltteluntera. u. Hyg. In prees (1979). 15. J. Nagayana, M. Kurataune and Y. Maauda, Bull, faivlron. Contaw. Toxicol.. JJ, 9 (1976). 16. C. Rappe, A. Cara, H.R. Buaar and H.-P. Boaahardt, Chaaoaphara. 5, 231 (1977). HONS 213812