Document aBDwrwggBKb7qMq9865Q7JgyN

Vinyl Chloride Formation from the Thermal DegnMiou of Poly(Vinyl Chloride) I. B. WAKEMAN ami H. R. JOHNSON Tenneco Chemicals, Inc. Corporate Research and Development Department Piscataivay, New Jersey 08854 The volatile products from the thermal degradation of polyvinyl chloride) (PVC) resins find compounds are shown to contain trace amounts of vinyl chloride. Data presented show tiie effect oftemperature and resin type on the amount ofvinyl chloride formed. At the maximum temperatures involved in PVC processing which may reach 210*C, vinyl chloride monomer (VCM) evolution amounts to less than 1 ppm (resin bads). A technique employing a thermogravimetric balance and charcoal adsorption of volatiles is described for studying thermal degradation of PVC. The volatiles are analyzed for vinyl chloride by gas chromatography. Peak identity was confirmed by mass spectrometry. INTRODUCTION Numerous studies and surveys have been published describing the thermal degradation and pyrolysis products of poly(vinyl chloride) (PVC). These studies have shown that at elevated temperatures essentially quantitative dehvdrochlorfnation ofPVC occurs, result ing in the formation ofa complex mixture ofaromatic and aliphatic hydrocarbons with benzene being the largest single component (1-8). Stromberg, et al. (1) studied the decomposition of PVC under vacuum at temperatures up to 400C and measured the variations in the decom position products as a function oftemperature. They did not indicate the formation of any vinyl chloride monomer (VCM) under these conditions. Boettner, et al. (2) investigated the thermal degradation ofPVC in air up to 6006C. Using infrared spectroscopy and gas chromatography-mass spectroscopy, they identified ap proximately 50 volatile degradation products and mea sured the rate of formation of the major products as a function of temperature. Their data indicate that up to BOOppm ofVCM is formed from PVC boroopolymerand up to 3000 ppm from PVC compounds. Although no mention is made of residual VCM in the polymer, the bulk of the VCM was detected in the 280-43CPC range and it is inferred that VCM is produced during the thermal decomposition. Woolley (3) reported low levels ofVCM in PVC pyrolysis products but did not measure residual VCM, and it is not dear whether any VCM was actuallyformed during the decomposition. Lewis (4) has recently reported degradation studies ofa homopolymer having a residual VCM content of< 0.5ppmand founds maximum VCM formation of 35 ppm at 350C. The recent findings (9) of carcinogenicity on pro longed exposure of test animals to elevated concen- trations of vinyl chloride in the air and of uses dW giosarcoma In vinyl chloride workers have caused seri ous concern about the effects on workers Involved fed phases of PVC manufacture and fabrication. Ahhoaeh present day manufacturing processes have reduced residual VCM content of PVC resins and compoundsto veiy low levels (10), the possibility of forming VCM hr degradation of the polymer during calendering, extru sion, and thermoforming operations Is still a earned concern, '' There is very little likelihood ofVCM formation Aw ing thermoforming of PVC for food packaging apphotion since this operation is carried out under anId 1 conditions(90-120eC for a few seconds). It has ka demonstrated previously that no detectable VCM is formed in up to 2 h exposure at 130*C, using an anabtv cal method capable of detecting 1 part per billion rf VCM (10). During calendering and extrusion operations, ht ever, temperatures may reach the I75-210#C ranged brief time periods. We know of no published inform*tion involving VCM formation under these conditk**. Our present work involved the investigation of the thermal degradation of PVC over the 130>500*C tem perature range with special emphasis on VCMformatfe* j at 210*C, the upper limit used in PVC fabrication- ^ EXPERIMENTAL Thermal degradation studies were carried out usia* duPont Model 950 thermal balance which provided * convenient method for accurately controlling the !*** peruture and heating rates. Samples of80-100 mgwe placed in a platinum foil boat suspended on the thennd liataiKv sample holder. The furnace tube surrounds^ 404 *HQLYMiK*UUMtUNG AND SOtNCI, AML. J979, V9i. I*. *** AP00007630 Vfity! Chloride Formalitm from the Thermal Degradation of Polt/(Vint/l Chloride) . r wtiple *p* with dean air at 60 ml/min. Prior degradation run, the cample was l^ai 13lfChr30 min with air flowing in the furnace ^g. TWi jmdure has been demonstrated to coinliirfj rrmorrriff residual VCM in the sample without ^oa any decomposition (10). After this conditioning ^fiotCtheeastside ofthe furnace was fitted with a short ^0s U-tmbe immersed In a cracked Ice bath. Ibis vedtocoof(he sweep gas prior to passing it through a ^aiTinl KIOSH approved, charcoal personnel yritming tube (SKC 226-01) to adsorb any VCM rafting odd the degradation of the sample. jtt the completion of the run the contents of the jfrrrniT tube were transferred to a vial capped with a gfefea* Uned septum (SKC 226-02-100). The vial was fall ii m auriuue-dry ice bath and 1 ml of carbon tiaTfiiT~ ftnj***^ **' i \ '^ the septum. "The vials were rieaalbmd lomndat room temperature for 30 min to *-jort> the VCM. if*1 w-f; ns.' ncr dtv jih*. Y.) Atru Car Chromatography A Perkin Elmer Model 3920 gas chromatograph ^pped with dual.flame ionization detectors was used t& work. I> view of the large number of products femed daring the thermal degradation ofPVC, and (he additional degradation products formed from additives orporated in FVC compounds, a variety of columns were screened to obtain optimum separation of the K3I peak. A combination column of a 10 ft x 1/8 is. fcapak P(Waters Associates)followed by a 10 in. X 1/8 m, Cbromosorb 104 (Johns Manville) was finally dected for the separation. This column operated at VC unth a kefim center gas flow of25 ml/min gave wd sepaialkM of the VCM peak (retention time--6 -fcrtiJIy, In order to definitely establish the identity of irif ke~VCM" peak, exposed charcoal tubes were submit- fcd to a commercial laboratory (11). Using the same CC p'**r u:^ rfwan, samples of die carbon disuIGde solutions were hr **cted into a gas chromatograph-masi spectrometer Mr fcrtxuraent. Mass spectrometer scans made at three jhl F*1-* on the VCM peak (peak maximum and leading ,m d portions) identified the peak as being en- ^VCM. The VCM peak from the thermal degrada- h" j. r certain PVC compounds prepared with sulfurrilhfceir. however, showed the presence of ' ' dioadde. This interference was eliminated by plac- hihe packed with Mallcosnrb (Mallinckrodf d VeHn) immediately before the carbon ad- tube. The MaRcosorb tube was shown to have "feet on VCM recovery bypassing known amounts of v jY* through the tube and analyzing the effluent gas. microliter altyuots of the carbon disulfide solu- ^ were injected into the gas chromatograph. Peak' i *j*vere determined using a Spcctra-Physfcs Afodel ^?)Uli[iegmtor. The VCM produced was c-alcu- *** Lt- comparing tin* VCM peak area ofdie sample to ^4,"1 pnrfhjced by injecting known gas standards Associates). The minimum amount of VCM ""*** could be detected was 0.1 nanogram, equivalent 4 Ppro of VCM from the 100 mg resin sample. Thermal Degradation Initial studies involved thermal degradation of PVC over the 130 to 500*C range using a heating rate of 10cC/min. A variety of PVC resins from different pro ducers were tested for total VCM evolution over this temperature range to determine if resin type or man ufacturing source has any effect on the amount of VCM produced. A few commercial compounds were similarly tested to determine the effect of stabilizers and other additives on VCM evolution. Thermal Degradation at PVC Fabrication Temperatures In ordertodetermine ifany exposure dangerexists for workers involved in PVC fabrication operations, a series ofpolymers, compounds and PVC sheets were tested for VCM formation during 5 and 30 min heating periods at 210C die maximum temperature which would be em ployed in PVC processing. Effect of Temperature on VCM Formation VCM formation as a function oftemperature was de termined on a homopolymer sample over the range of 200*450*0ata heating rate ofSVmlri. The'charcoal tubes were changed at 25* intervals during the heating cycle and analyzed individually for VCM content. The corre sponding thermogravimetric analysis (TGA) curve was recorded at the same time to provide a comparison of weight loss with VCM formation. RESULTS The total amounts ofVCM evolved from the thermal decomposition of a variety of PVC homopolymers and PVC-PVAC copolymers are listed in Table 1. Samples are included from six difTerentproducers. A consistently low level of VCM amounting to 15-30 ppm (based on Table 1. VCM Formation from Thermal Degradation of PVC (130-SOO'C) Reeln VCM, ppm (resin basts) Homopolymers Suspension low mol wt Suspension low mol wt Suspension medium molwt Suspension high mot wt Suspension high mol wt Dispersion Dispersion Dispersion Dispersion Dispersion Dispersion Blending Blending Blending Solution Polym. 19 23 22 18 20 21 20 19 18 22 19 19 15 31 21 Copolymers ' Type 1 Type 2 Type 3 20 25 20 'V-v5lr^ *WWKJWNG AND SCIfNCf, APRIL, 1979, Vl. 18, Mo. 5 40S AP00007631 1. B. Wofccman and H. li. Jahneon resin) uis found in the volatile decomposition products, frontal) ofthesamp!** tested. regardless ofre.si:> type or manufacturing source. To insure against . ; ssttl<* catalytic effect from the platinum bout, two homopolvmer and one copolymer samples were run using a glass boat. The levels of VCM produced were Identical to those obtained using the platinum boat. Results obtained from the degradation oftypical PVC compounds containing the most commonly used com* mercial stabilizers are shown In Table 2. Again, low levels ofVCM were found forall samples, indicating that the presence of stabilizers and other additives in the compounds does not Inhibit the formation of VCM. The effect of temperature on VCM formation from homopolymer is shown in Table 3 and in Fig. I. Pro gramming a 100 mg sample from 200 to 450*0 at 3/min resulted in the formation of a total of23.2 ppm ofVCM, the major portion being generated in the 275*350* re* glon. The corresponding weight loss curve shows that dehydrochlorination occurs most rapidly in the 250275C temperature interval. During this period only 2.3 ppm ofVCM is formed. Under the conditions employed in these studies, the major amount of VCM is formed after dehydrocblorination is essentially complete, leav ing a dark-colored residue. Programming the same homopolymer at 10/min and 6nishing with a 30 min hold at 450*C, also resulted in the formation of 23 ppm of VCM. A primary objective of this study was the determina- Tabie 2. Formation of VCM from Degradation of PVC : Compounds Healed from 130 to 500*C at 10*C/mln Compound typo Stabilizer type VCM, ppm (compound baste) Filled flexible Riled rigid Flexible Flexible Flexible Flexible Pb Pb Unknown Butyl tin Unknown Ba. Cd. Zn. P 27 13 40 27 27 52 Table 3. Effeet of Temperature on VCM Formation from Homopolymer Temperature, *C Total VCM formed, ppm (resin baels) Weight lose, percent by weight 200 225 250 275 300 325 350 375 400 . 425 450 N.D. N.D. N.O. 2.3 6.6 15.4 16.4 21.4 22.4 232 23.2 tempi* <* * 100 mg. TN*.Om.p*>Mumep*rte*gcimm4.m in0.o1 pp3m`M. n. N.O. 0.5 5 ' 46 67 60 62 63 64 66 72 406 TcMPtunat C. Fig. I. PVC degradation and VCM formation a a temperature. ' tion ofthe levels ofVCM to be expected at PVC E&tfr* tion temperatures. This was carried out by holdings plesat210*C, the upper limitofthe fabrication temper ture range, for periods of 5 and 30 mis. As showth Table 4, only traces of VCM are formed under be conditions amounting to a maximum of 0.5 ppm'(rab basis) after 5 min exposure and a maximum of 1.2ppm after 30 min exposure. This is particularly sfgni&M since it demonstrates that workers involved in nc fabrication are not exposed to dangerous oonceabatkas of VCM, and that VCM levels in fabrication areas a* well below the limits established by OSHA and EPl standards. - *. y:v - Table 4. Vinyl Chloride Formation at PVC Fabricsdoe Temperature (210*C) Sample VCM. nra (resin boa* 5 Minute 30 kttara heating heme* Homopolymors Suspension Low Mol. Wt Suspension Medium Mol. Wt. Suspension Medium Mot. Wl. Suspension High Mol. WL Blending Blending Dispersion N.D. N.O. N.D. ->< T` N.D. ../> as f. N.O. ...'sftS-0.5 1Jt` N.D. Copolymers Type 1' Type 2 Films Flexible Flexible Rigid Rigid Flexible N.D. i -am N.D. *^4& N.D. NJL`NO. NIL . * N.D. NIX. N.D. NIX N.D. NIX* Compounds Flexible Flexible Filled flexible Rigid N.D. o.t . N.D. 0.1 * N.D. N.D. N.D. t WH>. -pn* tirtrctad, *0.1 ppm. OtVMfR ENGINEERING AND SCIENCE, Anil, 197*. Vet. It. H+* lli.- ...a. ` An JiltVl uM i-i-UK V- t',r I* I'* i* th- TU ih' ai AP00007632 Vinyl Chloride Formation from the Thermal Degradation of PolyfVinyl Chloride) *llie amounts ofVCS( detected during thermal degra> _dation of PVC resins and compou nds in this study are in . ptod agreement with those published by Lewis (3). . A recent paper by Hoffman, f a!. (12) describes the ^ and determination of low nanogram levels of jVCM is the smoke from tobacco. Their data suggest that thetotalinorganic chloride in tobacco is the determining factor for the amount ofVCM in the smoke, the neces^ssyorganic radical being generated by the burning of tobacco products. A similar mechanism may account sjir the low levels of VCM observed in our work. ate-. ACKNOWLEDGMENTS authors wish to acknowledge the assistance ofM. - *<Stdey, N. Conro.and P. Heiney in carrying out much of ijbe analytical work in support of this project. REFERENCES 1. R. R. Stromberg. S. Strauss, and B. G. Achhamnu-r./. Ptilym. Sci., 39, 335 0959). 2. E. A. Boetnner, G. Ball, and B. Weiss,/. Appl. Polym. Sci., 13,377(1069). 3. W. D. Woolley, Srif. Polym.J., 3, 186 (1971). 4. J. W. Lewis, Paper presented at meeting ofAmerican Soci ety for Mass Spectroscopy, Houston, Texas (Mav 1975). 5. A. R. Berens, Poiym. Eng. ScL, 14, 318 (1974). 6. E. P. Chang and R. Salovey,/. Polym. Sci.. 12,2927(1974). 7. D. Braun, Pure Appl. Chem., 26,173 (1971). 6. M. M. O'Mara,/. Polym. Sci., 8.1687 (1970). 9. OSHA Permanent Standard 39 Federal Register 35890. 10. 1. B. Wakeman, M. Sagsese, and F. Owens, Symposium Reprints, Chemical Marketing and Economics Division, ACS. Meeting New York (April, 1976). 11. Gollob Analytical Service Corp. 12- D. Hof&nann, C. Patrlanakos, K. D. Bmnnemann, and G. B. Goric, Anal. Chtm* 48,47 (1976). .! ). to- 5 J&B-YMtK tNQINtlRtNG AND SCIENCE, APRIL, 1978, Vof. 18, No. 5 407 AP00007633 / 2*. 1lSRI E-MED1ATn> MITAGENIOTY Ol VINYUDENC [ CHL-OKIDIilN SALMONEL1.A TSPHIMIKIIM 1A1535. ~ Jnncj BK, HF Central Taxievloo Lab.. Imperial (Itcmical InduMrlcs Ltd.. At* Jerky Park, Cheshire. Sk.lt 4TJ, tnuUnd Canetr Lett; 1478 The Ames* muiagenicitv assay, modified to assess the mutagenicity of gases and sapors, was used to assess the mutagenic potential of vinylidene chloride (VDC) when incubated with a fortified mammalian kidney or liver tis sue nosirriiiochondrial supernatant {S-9 mix) from various species. Seeded dishes of Salmonella typhimurium strain TA100 or TA1535 were exposed to an atmosphere of 5% VDC in air for 72 hr. TAI00 and TAI535 gave similar results. VDC was_strongly..mutagenic when mediated .by mgyse k*fiTncy (23-i'old increase in mutation frequency) and fiver (18-fold increase) S-9 mix from Aroclor 1254 induced animals. VDC was weakly mutagenic when mediated by S-9 mix from uninduced mouse kidney (2.3-fold increase) and liver (1.6-fold increase). VDC was weakly positive (5- fold) when mediated by liver 5-9 mix from similarly in duced rats, but it was not mutagenic (less than or equal to 1.2-fold) when mediated by kidney or liver S-9 mix from noninduced rats. VDC showed weak mutagenicity (3-fold) when tested in a system mediated by iiver S-9 mix from a human who had received long-term phenobarbetal medica tion. but it was not mutagenic when mediated liver S-9 mix from noninduced marmosets or a noninduced human. -Thus, the- mutagenic- potential of VDS depends considera bly on the degree of activation of relevant drug- metabolizing enzymes. The results agree with the greater availability in treated mice than in rats of the reactive VDC metabolites l,l-dichloroethv!ene oxide and chloroace- tyf chloride and with the VDC carcinogenicity found in mice but not in rats. The data suggest that the limited number of primates examined respond more like rau than mice with regard to generation of alkylating metabolites and their reaction with bacterial DNA. (10 Refs) increased with dose and length of exposure. Ruts were more resistant to the carcinogenic effects of VC and VDC. <26 Kefs) 28. ROLF. OF L1VF.R GLUTATHIONE IN 1.1D1CHLOROF.THYl.KNF MKT.Uim.lSM AM) HLPATOTOXICITY IN INTACT HATS AND ISOLATED PERFUSED KaT LIVER. Bek-hurt 0, Werner HW, lU-itschkr D Institut fur PharntatohKh', I'aitmlUt Wurzburg, Vetxbjuiiur Straxe 9, D-8700 Wurzburg, W. Germany Arefc Toxicol (Bcrl); 41(3): I6V-178 1978 The role of liver glutathione in the metabolism and hepatotoxicity of 1,1-dichloroethylene (vinylidene chloride: VDC) was studied after po administration of 1.000 mg/kg VDC to female Wistar rats and after perfusion of isolated rat livers with 26.0 ug/ml VDC. After po administration, VDC lowered the liver glutathione concentration in a dosedependent fashion. Glutathione levels dropped to approx 30% of control values after 4 hr; however, these levels returned to control values within 24 hr. Addition of VDC to the perfusate caused a 23% drop in the glutathione concentration in isolated livers after 90 and ISO min of perfusion. The decrease was more pronounced when die thyl malcatc, a mercanturic acid precursor, was also add ed to the perfusate. VDC metabolism was lowered by 18% after diethyl maleaie-induced depletion of the reduced liver f' lutathionc pool hr Hvers from fed, but not fasted, rats, ncreases in the extra-hepatic glutathione poo! did not produce a change in total VDC conversion. A VDCmduced reduction in liver glutathione did not increase liver SCOT or SGPT levels in the perfused liver, but the lactatc/pyruvate ratio was significantly increased. These findings indicate that there is no correlation between the liver glutathione level and the hepatotoxicity of VDC in fasted rats. (22 Refs) cv /it. carcinogenicity of vinyl chloride and vinyWwtIDESE CHLORIDE. Lee CC. Bhandarf JC, WJnrtoa JM, House \VB, Dlxoa RL. Woods JS Midwest Res. lost., 425 Volkcr Bl'd.. Kiikis City, MO. A41I0 J Toxicol Emiron Health: 4( l>:15-30 1978 The effects of exposure to 50. 250, or 1,000 ppm vinyl chloride (VC) or 55 ppm vinylidene chloride (VDC) for 6 hr/day. S/days/wk, were studied in albino CD-I mice and CD rats. The animals were killed after I, 2, 3, 6, 9, or 12 mo. Bronchioloalveolar adenomas developed in 12/63, 22/63. and 48/69 mice exposed to 50, 250. and 1,000 ppm Vl, respectively. Bronchioloalveolar adenomas developed in 6/35 animals exposed to VDC. Three of 29, 23/63, and 31/69 mice exposed to 50, 250, and 1,000 ppm VC. respec tively, developed hemanaiosarcomas (HS) In the liver. In the mice exposed to VDC. HS developed in the livers of two males and one female. Mammary gland tumors oc curred in 9/34, 3/34, and 13/36 female mice exposed to 50, 250, 1,000 ppm VC respectively, and they included ductubr adenocarcinomas and squamous and anaplastic cell carcinomas with metastasis to the lung. Malignant lymphomas were found in various organs of mice exposed to VC but not in any of the mice exposed to VDC. Twelve of 70 and 21/70 rats exposed to 250 and 1,000 ppm VC, respectively, developed HS in the liver; 3/34 female rats exposed to 250 ppm VC and 13/70 rats exposed to (.000 VC also developed HS in (he lung. Two of 36 male rats exposed to VDC developed HS in the mesenteric lymph node or sc tissue. It is concluded that VC is highly carci nogenic in mice; the incidence and severity of the tumors 29. STUDY OF COCARCINOGEMC EFFECT OF POLYVt NYLTHLORIDE AND 20-METHYLCHOLANTHRENE. Draganor |, Rakhe' R. Radtvi M Ret. Inti. Oncology. Sofia. Ruic>fit OnkologlU; 1S(3):1J2-137 1978 The cocarcinogenic effect of polyvinyl chloride (PVC) and 20-methylcholanthrene (20-MC) was studied in random-bred rats. PVC plates were implanted in the ani mals and, 30 days later, the animals were inoculated with 20-MC (10 ma in 0.5 ml of sunflower oil. sc). The av latent period oT tumor development was 2*5 mo. compared with 16-20 mo in animals exposed to PVC alone. The av latent period of tumor development in rats treated with 20MC alone was also 2-5 mo. Tumor incidence was 17.5%20% in rats exposed to P.VC alone. 42% in rats treated with 20-MC alone, and 68% in rats exposed to PVC 4- 20MC. (9 Refs) (In BUL) 30, IN 'VIVO TRAPPING OF A VINYL CHt-ORWE META RO UTE BY MEAN'S OF3.4-DICHLOROBENZENETHIOL. Duller G, Norpqth K.Ovtzarski W Inotitut fur Siaubluiigenrorschiing. Wt-slring 10, D-4400 Mun-aer, W. Germany lot Arch Occup Kmlreo H-sllh;42(2):137-l39 1978 A method for trapping metabolites of vinyl chloride (VC) and 2,2'-dichforodiethylciher (DDE) is reported. Rats treated with VC and 3,4-dichlorobcnzenethioi excrete S- 8 AP00007634