Document 0q2po43qQbM323B8Ox0XwEByx

URNAL OF APPLIED POLYMER SCIENCE VOL. 13, PP. 377-391 (1969) Analysis of the Volatile Combustion Products of Vinyl Plastics E. A. BOETTNER, GWENDOLYN BALL, and BENJAMIN WEISS, School of Public Health, University of Michigan, Ann Arbor, Michigan 48104 Synopsis The extensive use of plastics for insulation end building materials has created interest in the possible toxicity of their combustion products. Three poly(vinyl chloride) homo- polymers, a vinyl chloride-vinyl acetate copolymer, and formulations of two of the homo- polymers and the copolymer have been examined to determine the composition and tox icity of their combustion products. Differential thermal analysis and thermal gravi metric analysis were used to study the breakdown process which occurred in several steps, the largest and first being the release of hydrogen chloride from the polymer at about 300C. Approximately 50 products of combustion were then determined qualitatively by using either infrared spectroscopy or a combination of gas chromatography and mass spectroscopy. Quantitative analyses were carried out on 22 of the combustion products, uantitative changes with varying air supply, temperature, and heating rate were ained. given. A comparison of the products of the polymers and their formulations is INTRODUCTION The use of plastics for wire insulation and other materials that may be intentionally or accidentally subjected to high temperatures has made it necessary for the environmental health scientist to know something about the nature and toxicity of their combustion products. The pyrolysis prod ucts of various polymers have been reported by others,1'* but much of this work has been carried out in inert atmospheres for the purpose of gaining information about polymer structure. Workers concerned with the com bustion of vinyl chloride polymers examined these primarily to determine the amounts of HCi, C0-, CO, and other gross products given off and the temperatures at which they were produced. Five such papers were pub lished recently in German journals.1-* In 1963, we began a program to determine the volatile combustion prod ucts of various plastics9 and to cany out animal exposure studies to deter mine the toxicity of these products. The combination of analytical data md acute toxicity studies is used to determine whether the compounds hown to be present can account for the overall toxicity of the decomposition products or whether synergistic effects may also be a factor to be taken into onsicleration. A prior knowledge of the physical nature and chemical 377 37S BOETTNER, BALL, WEISS composition of these pyrolytic products provides a necessary guide animal exposures, both as to the extent of exposure and to the possible na; of the toxic effects. This paper discusses the results of qualitative and quantitative anal, of the combustion products of polyfviuyl chloride) polymers and plast This particular group was selected, first, because of its wide applicatio: electrical insulation and construction materials, second, because of the tensive work done in the pyrolysis of the basic polymers by others, third, because only limited knowledge is available on the quantities of c pounds formed in combustion of plastic products of this type under van conditions.1 The portion of the study concerned with determining toxicity of the combustion products will be presented in another paper.11 EXPERIMENTAL The analytical portion of the study was carried out in three phases: temperatures at which chemical and physical changes take place in a c trolled air supply were determined; (2) compounds liberated at these t peratures were identified or characterized; (3) quantities of the prod were determined either at the temperatures covering each decompose step or for a sample collected over the entire combustion process. The temperatures at which changes take place were determined by di: ential thermal analysis (DTA) and thermogravimetric analysis (TC In the second phase of the analysis, which involved identification of c pounds liberated at the decomposition temperatures, two different proaches were taken. Gases collected from either the TGA apparatus combustion furnace were analyzed directly for major compounds by frared absorption spectroscopy or mass spectroscopy. For positive idt fication of the minor constituents, it was necessary to collect the gases f a larger sample than the 200 mg that can be handled by the TGA appara To do this, a combustion furnace with controlled temperature and air su; was used to carry out the combustion on a tenfold scale (2 g) but under same conditions as in the TGA apparatus. The furnace, controlled 1 commercial programmer-controller, has a Yycor gas-tight core whic supplied air from a tank of purified air. Products from the combus furnace can be collected in several temperature or time fractions, depen on the collection method. The outlet of the core can be connected U-tube submerged in a cooling bath of Methyl cellosolve-Dry Ice folio by a plastic bag to collect the gaseous products in two phases (compoi boiling below --75C and above --75C), or it can be connected t plastic bag and a single sample collected to be separated by gas chroma raphy. To establish that the products from the TGA apparatus combustion furnace were the same, a gas chromatogram of the finance c bustion products was compared with a chromatogram of the TGA con: tion products over the same weight-loss range. Individual chrorr graphic peaks were collected by condensation at liquid air temperature analyzed by mass spectroscopy. TENO 652 VOLATILE COMBUSTION PRODUCTS 379 The third phase, quantitative analysis of the identified products, was carried out by infrared spectroscopy for the principal constituents and by gas chromatography for the minor constituents. Four polymers produced by two companies were analyzed, as were three formulations using three of these polymers. They are listed as follows with the approximate molecular weight (weight average) and approximate mesh size of each polymer sample: polymer A, a poly(vinyl chloride) homopolymer, MW 109000, mesh size 80; polymer B, similar to A but a later production, MW 111000, mesh size 80; polymer C, a po!y(vinyl chloride) homopolymer produced by another company, MW 250000, mesh size 60; copolymer D, a 86:15 copolymer of vinyl chloride and vinyl acetate, MW 55000, mesh size 100; plastic E, a formulation commonly used for wire insulation, containing 57% polymer C; plastic F, a formulation used for floor tiles, containing 35% copolymer D; plastic G, a wire insulation formulation, containing 51% polymer B. With one exception, polymers and formulations were standard commer cial materials furnished by the polymer manufacturers. The last product, plastic G, was formulated for us in a development lab of one manufacturer, a "typical" wire insulation formula being used. RESULTS AND DISCUSSION Breakdown Process Differential thermal analysis (DTA) was used to determine the tempera tures at which chemical and physical changes take place. With this apparatus a record as shown in Figure 1 was obtained for polymer C when heated at 10C/min in an excess of air. The polymer was mixed 1:1 with Fig. ]. Differential thermal analysis record of polymer C. 380 BOETTNER, BALL, WEISS Fig. 2. Differential thermal analysis record of plastic E. SiC and a 300 mg sample ran, SiC being used as a reference. This record shows the heat absorbed (from an endothermic change) or liberated (from an exothermic change) as AT plotted against the temperature of the sample. Full-scale sensitivity, AT, is 1.82C up to 450C and 3.30C from 450 to 6008C. Of significance is the endothermic peak with a maximum at about 300C, which results from release of chlorine atoms in the polymer. Un fortunately it is here obscured by the start of the exothermic peaks which began at 300C and continued to 600C, at which temperature the sample is completely combusted. The dehydrochlorination peak varies in position Fig. 3. Thermograviroetric analysis record of polymer C heated at 3"C/nnn in TENO 654 VOLATILE COMBUSTION PRODUCTS 381 by =fc20C and, as this variation seems to be erratic and independent of the sample, it is assumed that it is associated with some experimental parameter such as rote of sample melting, etc. Figure 2 shows a DTA record of plastic E which was formulated from polymer C. The only gross difference is a steeper slope in the exothermic peak at 325C. Because of the limited amount of interpretation that could be extracted from these curves, information derived from the DTA technique did not prove as helpful to our study as did thermogravimetric analysis, where a recording of weight change with increasing temperature is obtained. Figure 3 shows a TGA record for polymer C heated at 3C/min in air. Here a rapid weight loss at 300C is observed, corresponding to the endo thermic peak in the DTA record. On collecting the combustion gases up to 300C and analyzing them by infrared absorption, it was found that the 60% weight loss was almost entirely due to removal of chlorine in the form of hydrogen chloride gas and, by calculation, it appeared that the amount corresponded to almost all the chlorine atoms in the polymer. Above 300C, the TGA record shows a slower but gradually increasing rate of weight loss up to 600C, corresponding tothe exothermic peaks in the DTA record. This weight loss appears to take place in five stages. First is the rapid loss up to 280C; second, a decreasing rate of loss up to 350C; third, a slow constant rate of loss to 430C; fourth, a more rapid rate of loss to 51QC; and finally, a faster rate of loss for the remainder of the sample. IThese particular temperature ranges are used later (Table IV) in describing he change in composition of combustion products with temperature. Figure 4 shows a similar record for plastic E, which contains polymer C. The height of the 300C step, attributed in Figure 2 to removal of chlorine, is as high here, although this plastic contains only 57% of the polymer. The remainder of the weight loss is due to boiling-off of the plasticizer, a phthalate which has a boiling point of about 280C. Fig. 4. Thermogravimetric analysis record of plastic E heated at 3C/mm in air. 3S2 BOETTXER, BALL, WEISS Fig. 5. Thennogravimetric analysis record of copolymer D heated at 3'C/min in ai Figure 5 is the TGA record of copolymer D, and it can be seen that i weight loss proceeds in the same manner as the previous polymer, exce; that here the initial weight loss is due to the removal of both chlorine atoir and acetate groups. Figure 6 is of plastic F, the floor tile formulation th: contains about 35% of polymer D. This formulation contains a conside able amount of inorganic minerals, such as asbestos, which is the reason f' the limited amount of weight loss up to 600C. Qualitative Analysis This study was primarily concerned with the "volatile products" < combustion, those compounds boiling up to about 150C being defined c Fig. 6. Thennogravimetric analysis record of plastic F heated at 3*C/min in air. TENO 656 VOLATILE COMBUSTION* PRODUCTS 3S3 volatile. As a result, although the TGA record of plastic E (Fig. 4) shows a weight loss due to boiling off of the phthalatc plasticizer, this compound is not listed in our results. Compounds that remained in the combustion boat or condensed in the sample bag or at the exit end of the combustion tube were not analyzed, although an attempt was made to determine the weight of this remainder to assure that the volatile fraction was all ac counted for. When the products of combustion were analyzed by infrared spectros copy, collecting the total products in gas form and introducing them into long-path gas cells, absorption bands of HCI, CO;, CO, and benzene were found, along with other bands in the 3000 cm-1 region indicating the presence of compounds containing C--H groups. It was evident that identifi cation of other compounds should be carried out by separating them from the mixture and collecting them in individual quantities that would per mit identification by infrared or mass spectroscopy. For this gas chro matography was employed, using sufficiently large columns to permit the introduction of up to 30 ml of sample. To separate the low-boiling com pounds, i.e., those boiling at temperatures less than 0C, a Porapak Q column was used; to separate higher-boiling compounds a column of 5% squalane on Chromosorb P was used. Columns were temperature pro grammed at 10C/min to 110C after an initial period of 5-15 minutes at A stream splitter on the column outlet of the chromatograph one hundredth of the gas through a hydrogen flame detector. The remaining 99 parts emerged from the column through a two-way valve and were diverted into a U-tube when a particular compound was eluted. The U-tube, filled with glass beads, was cooled with liquid air to condense out everything except the helium carrier gas. The U-tube and its cooling flask were then transferred to the inlet system of the mass spectrometer. Figure 7 shows the separation of the first eight compounds. These are primarily parafins and olefins through butane, and are identified in Table I according to peak number. These peaks were identified by using the Fig. 7. Chromatogram of low-boiling combustion products of polymer A on a Porapak Q column. TENO 657 Kc. BOETTNER, BALL, WEISS TABLE L Identification of Chromatogram Peaks Air 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 43 43 44 45 46 47 4S 49 50 51 Identification - Methane Ethylene Ethane Propylene Propane (methyl chloride) Vinyl chloride I-Butene (isobutene-butadiene) Butane frasi$-2-Butene cis-2-Butene 3-Methyl-1-butene Isopentane (1,4-pentadiene) 1-Pentene Pentane inww-2-Pentene ew-2-Pentene 2-Methyl-2-butene eis or tran*-l,3-Pentadiene Cis or trtr/U-2-Penten-4-yne Cydopeotene Cyclopentane 2-Methylpentane 1-Hexene (3-methylpeotane) Hexane 2-Hexene Methylcyclopentane 1-Methylcyclopentene Benzene Cyclohexane 1-Heptene 1,4-Dimetbylcydopentene* Heptane 3-Ethylcydopentene Methylcyclohexane Ethylcydopentane 1,2-Dime thylcyclopentene* 1- or 4-Ethylcyclopentene 1-Methylcydohexene Toluene Octane Ethylbenzene p-Xylene m-Xylene o-Xylene * Tentative. TENO 658 VOLATILE COMBUSTION PRODUCTS 385 -$1 vfA. -1-r Fig. 3- Chromatogram of the combustion products of polymer A on a column of 5% squalane on Chromosorb P. collection technique described to transfer the compound to the mass spec trograph. Mass spectrographic identification was then cross-checked by. injecting the pure compound on the chromatograph to establish that its elution time was the same as the collected peak.' A chromatogram on. the squalane column is shown in Figure 8. Here, in addition to straight-chain paraffins and olefins, both branched and cyclic compounds are found. These are again identified in Table I. There are minimally 59 volatile products of combustion, totaling the compounds identified by direct infrared analysis and 51 chromato- gKc peaks, at least four of which represent more than one compound. OfThis minimum figure of 59, we have identified 52 compounds. Because of the similarity in molecular weights and limited mass spectral reference data, positive identification of the remaining peaks has been difficult. These are higher-boiling compounds and are present in very small quanti ties, i.e., less than 30 pg/g, and therefore are of limited interest from the standpoint of toxicity. The qualitative picture was similar for all PVC polymers tested, with only a few differences found in the higher-boiling products and the formulations. Quantitative Analysis All the quantitative work was done on the 22 compounds present in the greatest quantities. As a uniform condition for intercomparing the poly mers and formulations of the polymers, the following combustion conditions were adopted: The rate of air supply was 60 cc/min which, when integrated over the entire run, would result in about twice the amount of oxygen necessary to convert all of the carbon to carbon dioxide. However, this was not true for any particular time in the combustion process, in that first, the carbon was not available to form carbon dioxide until higher temperatures were reached and second, the air flow pattern was such that all of the oxygen was.not in the immediate vicinity of the sample. The plastics were heated, (after an initial heating from room temperature) from 200C to 600C at a rate of 3C/min. Sft : 1 ~SWfll TENO 659 3S6 boettxer, ball, weiss During the study several parameters were examined that could affect the quantities of the various compounds of combustion, and these are reported as follows: (1) the variation in quantity within a class of compounds and between classes of compounds for the polymers; (2) changes with varying air supply; (3) changes with temperature; (4) changes with varying heating rates; (5) changes between the plastics and their polymers. Variation in Quantity within a Class of Compounds and between Classes of Compounds for the Polymers Table II shows the volatile products of combustion of polymer B. Quantites of these products are representative of the other two homopolymers. Also in this table are the same products from copolymer D. All quantities given are in mg of compound per gram of polymer. In polymer B, chlorin is almost completely accounted for in the form of HC1 gas (583 mg found, 584 mg possible). The only other chlorine compounds found, methyl chloride and vinyl chloride, are present in quantities of less than 1 mg. Some investigators have reported finding phosgene intermittently in very small quantities (0.1 ppm) but, using detector tubes of this sensitivity, we could not detect this compound. Carbon dioxide and carbon monoxide, which are formed at higher combustion temperatures, are present in quite large TABLE H Volatile Combustion Products of Polymer B and Copolymer D Combustion products, mg/g Compound Polymer B Copolymer D HC1 Acetic acid CO, CO Methane Ethylene Ethane Propylene Propane Vinyl chloride 1-Butene Butane Isopentane 1-Pentene Pentane Cyclopentene Cyclopentane 1-Hexene Hexane Methylcyclopentane Benzene Toluene 583. -- 729. r 73 ^ 442. T Ujf * 4.6 0.5S 2.2 0.47 0.S4 0.60 0.18 i! . # 0.28 0.02 = .. r 0.06 0.16 0.05 0.05 0.05 0.12 . 0.01 36. 1.3 500. 96. 923. 292. 4.4 0.60 2.3 0.56 0.88 0.72 0.22 0.29 0.02 0.09 0.21 0.05 0.06 O.OS 0.17 0.05 23. 0.96 TEN0 660 VOLATILE COMBUSTION PRODUCTS SS7 amount*. The most abundant hydrocarbon is benzene (35 mg/g). De creasing in quantitiy with increasing molecular weight are the straight chain aliphatic*, starting with methane. ' Olefins and side-branched aliphatic* are generally present in lesser quantities than the corresponding aliphatic*. Note that quantities of these compounds for the polymer and copolymer are very similar, except that some of the HCI is replaced by acetic acid. In these samples, volatile combustion products account for approximately 100% of the original polymers, and no residue remained in the combustion boats. Changes with Varying Air Supply One sample (polymer A) was combusted by use of three different condi tions of air supply. The first provided the minimum amount of oxygen for complete combustion of all carbon to CO. (30 cc air/minj. The second was about twice this amount (60 cc/min), and the third was with a large excess of oxygen. The results are shown in Table III. The data show that all of the hydro carbons except benzene decrease in quantity with increasing air. Although a change in the amount of air affects the quantity of hydrocarbons gener ated, surprisingly it has little effect on the quantity of CO and COj genera ted, which showed no trends using these air flows. However, it should be noted that the quantities of CO- varied erratically in most of our combu*- TABLE in Variation of the Combustion Products of Polymer A with Oxygen Combustion products, mg/g Compound Air, 30 cc/min Air, 60 cc/min Air, 25 cc/min + oxygen, 21 cc/min CO, CO Methane Ethylene Ethane Propylene Propane Vinyl chloride 1-Butene Butane Isopentane l-Pentene Pentane Cydopentene Cyclopentane 1-Hexene llexane M ethylcyclopent ane Benzene Toluene 861. 357. 6.7 0.76 2.6 0.80 1,3 0.51 0.25 0.53 0.02 0.10 0.26 0.07 0.08 0.07 0.16 0.06 33. 1.5 - 619. 429. 4.7 0.53 2.1 0.53 1-0 0.59 0.18 0.31 0.02 0.08 0.20 0.05 0.07 0.06 0.14 0.05 31. 1.1 814. 401. 3.8 0.28 1.7 0.28 0.66 0.66 0.06 0.15 0.01 0.04 0.11 0.03 0.03 0.03 0.09 0.03 32. 0.68 388 BOETTNER, BALL, WEISS tion runs, possibly because of sample ignition at higher temperatures. The amount of benzene seems to be independent- of air conditions. The amount of HC1, 580 5 mg/g (not listed in the table), is also independent of air conditions and accounts for nearly all the chlorine atoms of the polymer. Vinyl chloride Increases consistently with increasing air supply, but this increase is not enough to affect the HC1 concentration significantly. Changes with Temperature Table IV shows the variation in quantities of products of combustion as a function of temperature. The products were collected in five fractions during a single heating run. Fractions were selected on the basis of inflec tions on the TGA curve shown in Figure 3. During the first temperature fraction almost 80% of the benzene is formed along with a small amount of toluene and some unsatuiated hydrocarbons. Production of HC1 (not listed) roughly parallels that of benzene. TABLE IV Variation of Combustion Products of Polymer A with Temperature Combustion products, mg/g Compound 252S0'C 280350*C 350430"C 430510*C 510580*0 CO, CO Methane Ethylene Ethane Propylene Propane Vinyl chloride 1-Butene Butane Isopentane 1-Pentene Pentane Cyclopentene Cyclopentane 1-Hexene Hexane Methylclopentane Benzene Toluene -- -- -- 0.04 -- 0.06 -- 0.04 0.02 -- -- -- -- -- -- -- -- 24. 0.12 9.7 20. 0.20 0.33 0.12. 0.11 O.OS 0.25 0.04 0.03 -- 0.01 0.01 0.02 0.01 0.01 0.01 -- 6.6 0.18 181. 46. 1.3 0.39 0.94 0.31 0.44 0.17 0.08 0.20 0.005 0.03 O.OS 0.01 0.02 0.02 0.05 0.02 0.35 0.55 244. 151- 1.8 -- 0.41 -- 0.11 0.02 -- 0.02 0.001 -- 0.01 -- -- -- 0.01 -- 0.16 0.03 237. 181. 0.31 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 0.01 In the second fraction (280-350C), CCL and CO appear, toluene con tinues to increase, but HC1 and benzene are already decreasing and continue to decrease through the higher ranges. HC1 is present only in trace amounts after 300C. Additionally, in the second fraction the maximum amount of vinyl chloride is formed. As expected, CO* and CO reached their maxima at higher temperatures. TENO 662 VOLATILE COMBUSTION PRODUCTS 389 Methane reached its maximum in the fourth step, but was still generated at higher temperatures. The other straight-chain aliphatics reached their maxima in the third step and were present in the last step in very small quantities. Olefins began to form in the first stop, reached their maxima in the third, and did not appear in the last two fractions. Changes with Varying Heating Rates The data on Table V represent what happens when the heating rate is increased considerably. An increase in the heating rate results in increased hydrocarbon production, and of note is that unsaturated compounds increase by a greater amount than the saturates. As a result, with an exceptionally fast heating rate, olefins will be present in greater quantities than aliphatics. This is illustrated in the chromatogram shown in Figure 7, where the products were collected from a run in which the controlling thermocouple burned out, with the result that a heating rate greater than 80aC/min was attained. Also to be noted is that at higher heating rates production of CO* and CO decreases. The heating rate has no significant effect on the amount of HC1 produced. TABLE V Variation of Combustion Products of Polymer A with Heating Rate Combustion products, mg/g Compound 3*C/min 50C/min CO, CO Methane Ethylene Ethane Propylene Propane Vinyl chloride 1-Butene Butane Isopentane 1-Pentene Pentane Cyclopentene Cvclopentane 1-Hexenc Hexane Mcthylcydopentane Ben/.one Toluene 619. 429. 4.7 0.53 2.1 0.53 1.0 0.59 0.1S 0.31 0.02 0.08 0.20 0.05 0.07 0.06 0.14 0.05 31. 1.1 397. 269. 8.7 2.3 3.5 1.5 1.3 0.64 0.67 0.69 0.02 0.18 0.29 0.19 0.11 0.13 0.20 0.08 43. 3.5 Variations between Plastics and their Polymers A comparison of the combustion products of the three plastics with the combustion products of their polymers is given iu Table Vf. Except for 390 BOETTNEfi, BALL, WEISS the aromatic compounds, the hydrocarbons in plastics E and G- have all increased in quantity by factors ranging from 1.3 up to 8 times, tho average being about 4 times. Likewise, the amount of vinyl chloride appears to be about 5 times as great in these samples. Most of the increases arc attribu table to the breakdown of the phthalate plasticizer, which forms a series of hydrocarbons similar to those produced by PVC. The plasticiser, either dioctyl phthalate or diisodecyl phthalate, cannot be directly responsible for 'the increase in vinyl chloride. It is presently not clear if this increase is real or if there hus been some chromatographic interference with an uniden tified compound being eluted at the same time as vinyl chlorid . The plasticizers do form several oxygenated compounds such as ethylene oxide, acetone, and at least one alcohol which have not been fully investigated. TABLE VI Comparison of Combustion Products of the Plastics with the Combustion Products of their Polymers Combustion products, rog/g Compound Polymer B Plastic G Polymer C Plastic E Co polymer D Plastic F HQ Acetic acid 583. -- CO, 729. CO 442. Methane 4.6 Ethylene 0.58 Ethane 2.2 Propylene 0.47 Propane 0.84 Vinyl chloride 0.60 1-Butene 0.18 Butane 0.28 Isopentane 0.02 1-Pen tene 0.06 Pentane 0.16 Cydopentene 0.05 Cyclopentane 0.05 1-Hexene 0.05 Hexane 0.12 Methylcyclopentane 0.14 Benzene 36. Toluene 1.3 Residue* -- 273. -- 616. 67. 6.6 2.3 3.0 2.0 1.7 3.3 1.1 1.1 0.15 0.35 0.58 0.14 0.16 0.24 0.49 0.14 10. 0.94 159. 584. -- 730. 403. 5.8 0.33 2.5 0.56 1.1 0.52 0.28 0.39 0.02 0.11 0.27 0.58 0.07 0.09 0.25 0.07 29. 1.1 -- 333. -- 11S2. 90. 6.8 2.0 2.9 1.4 1.4 2.6 0.58 0.74 0.04 0.15 0.38 0.07 0.09 0.18 0.35 0.09 11. 1.0 61. 500. 96. 923. 292. 4.4 0.60 2.3 0.56 0.88 0.72 0.22 0.29 0.02 0.09 0.21 0.05 0.06 0.03 0.17 0.05 28. 0.96 -- 73. 20. 456. 31. 0.30 0.13 0.13 0.11 0.10 0.30 0.06 0.05 0.01 0.01 0.02 0.004 0.003 0.01 0.01 -- 0.86 0.04 709. * Residue b what remained in combustion boat. It does not include products which condensed st the end of the combustion tube. Different results are noted for plastic F, the floor tile formulation made from copolymer D. This product contains about l/j copolymer D, but the hydrocarbon products generated, especially saturated aliphatics and ben TEN0 664 * VOLATILE COMBUSTION PRODUCTS 391 zene. are considerably loss than /, the amount generated in copolymer D. Likewise, the amount of HC1 to CHjCOOH is only 20%, rather than the 33% expected. This product contains about 70% inert material, such as asbestos and calcium carbonate, which may play a part in inhibiting break down of the polymer and production of hydrocarbons. It should be stated that plastic F has three compounds with boiling points in the vicinity of 120<1C that have not been identified, which are present in quantities of 1-5 mg/g and probably arise from combustion of the plasticizer. This investigation was supported in part by Public Health Service Research Grant No. UI-0G4S5 from the National Center for Urban and Industrial Health. References 1. S. L. Madorsky, Thermal Degradation of Organic Polymen, Wiley, New York, 1964. 2. D. W. Levi, Literature Survey on Thermal Degradation, Thermal Oxidation, and Thermal Analysis of High Polymers, Picatinny Arsenal, Dover, N. J., June 1963. AD423646, Clearinghouse for Federal Scientific and Technical Information. 3. D. A. Teetsel and D. W. Levi, Literature Survey on Thermal Degradation, Thermal Oxidation, and Thermal Analysis of High Polymers, II, Picatinny Arsenal, Dover, N. J.f Jan. 1966, AD63I655, Clearinghouse for Federal Scientific and Technical Information. 4. K. Thinius, E. Schroder, and A. Gustke, Plaete Kautschvk, 11,67 (1964) 5. E. Hagen aud G. Friedrich, Plait* Kautschuk, 12,215 (1965). 6. H. Luther and H. Kruger, Kunststojfe, 56,74 (1966). 7. D. Braun aud M. Thallmaier, Kunststojfe, 56, SO (1966). Z. Yymazml and J. Stepek, KuntUioffe, 56,86 (1966). E. A. Boettner and B. Weiss, Amer. Ind. Hyg. Aeeoe. J., 28,535 (1967). 10. Undenerilen' Laboratories Bull. Res. No. 53 (July 1963). 11. E. A. Boettner, H. Cornish, and E. Aban, in preparation. Received October 7, 1967