Document x1LnwL4xqvMrNZKy0w9xM0jp1

l: tv -H / -a-.:. *4 JDI'AfNAl, Of POLYMER RCIBNC1: Polymer Clwmiltry KrliMom \Ol*. Iff. 71? ?49<I97}> Investigations on Poly(vinyl Chloride). I. Evolution of Aromatics on Pyrolysis of. Poly (vinyl Chloride) and Its Mechanism TAKEO IIDA, MOTOHARU NAKANLSHI, and KUNIO GOTO, Osaka Institute 0/ Technology, Department 0/ Applied Chemistry, Omiya 5-chome, Asahi-ku, Osaka Synopsis Pelyt vinyl chloride) (PVC) alone or mixed with 10 wt-7 and (iO TiOi, SnO,. ZnO, and AliO, wore pymlyxed by lining a pyrolysis gna chromatograph Bennone, tol uene, ethylbenzene, o-xylene, styrene, naphthalene, and various chlorobenzenes were identified No hydrocarbons could be detected in pyrolysis products of anv samples at `iOG'T More aromatic hydrocarbons than aliphatic hydrocarbons are released from the PVr-TiO, system and in preheated PVC. The contrary result is observed in the PVC-ZnO and PVC-SnOj systems. Aromatic* having methyl endgmupa arc easily re leased from the FVC-ZnO and PVC-SnOt systems and at. elevated pyrolysis temperature, because methylene groups are easily isolated along the chain by ZnO, SnOi and the heal ing The release of ethylbenzene o-xylene, and chlombensenes suggest* a repeated de- hydmchlorination and recombination of HCI and Clj to double bonds along the chain. Possible decomposition mechanisms of PVC are discussed. INTRODUCTION The thermal decomposition of poly(vinyl chloride) (PVC) has usually been represented as involving the dehydrochlorination of the polymer backbone. On the other hand. Stromberg1 assumed the release of chlorine molecules in his kinetics of the decomposition of PVC. The isolation of small amounts of Hi and Cli was reported by Tsuchiya* on pyrolysis of PVC in an inert gas. Ohta1 suggested that a part of the hydrogen chloride released from PVC may recombine with the double bonds along the chain introduced by the dehydrochlorination. After pyrolysis of PVC, many kinds of hydrocarbons consisting of aliphatic* and aromatics in addition to HCI were detected by Strombcrg1 and other workers.4-' In the present work, PVC, preheated PVC, and PVC containing various metal oxides were pyrolyzed. The pyrolysis products were identified by using gas chromatography and included aromatic hydrocarbons such as benzene, toluene, ethylbenzene, or various aryl chlorides. Different proportions of volatile products were obtained from the various pyrolyscs of PVC. Therefore, it was believed that the pyrolysis products did not depend on the original structure of PVC but on the final structure of polymer releasing their small molecules. In particular, aryl chlorides 737 g 1974 oy John Wiley A Son*, Inc. RECEIVED may 211974 R. N. WHEELER, JR. ; r ^ ` (. UCC 0374 IIOA, NAKAMSHI, Af.D GOTO may bo derived from the recombination of double bonds with chlorine as HC1 or Cli, assuming that no secondary reactions of the volatile decom position products take place. EXPERIMENTAL Materials The PVC used in this work was Geon-l03EP. Metal oxides were ob tained from Kanto Chemical Co., Inc. Instruments A Yanagimoto Model GP-1000 pyrolyzer was used to decompose PVC. It was directly attached to the inlet |>ort of a gas chromatograph, Yanagimoto Model GCG-oflOKP, with double flame ionization detectors. The furnace and stick used to introduces the samples into the furnace were made of quartz. The precut part in the pyrolyzcr was filled with 30 wt-% sodium hy droxide coated on Diasolid H to absorb HC1 during the decomposition of PVC and to avoid contaminating or damaging the instruments. 1 33 -V ' j j. 33 * Fig. l. Pyrogram obtained fct 700#C of untreated polyfvinyl chloride). VM*, ^ v* * , * A .Zhkm v1', A... POLYfVINYl, CHLORIDE). I 73V Pyrolysis * Samples (powder), 1-3 mg, were pyrolyzed in helium gns (flow rate of helium gas, 25 ml/min) at 200-800C. Separation columns (3 mm id X 2.17 m) containing 20 wt-% PEG-20M coated on Celite 54,> were used at a programed increase in temperature from 60C to 200'C at a rate of 8C/min. The flow rate of carrier gas (He) was maintained at. 25 ml/min at G0C. Columns of 20 wt-% SE-30 coated on Celite 545 were used only to identify the pyrolysis products under the same conditions os the separation columns. PVC and the metal oxides were mixed on an agate mortar at room temperature. RESULTS AND DISCUSSION PVC samples with or without 10 wt-% and 50 wt-% TiOj, SnOj, ZnO, and AiiOj were pyrolyzed at 200-800C. A sample of PVC preheated at 200C for 40 min in air was also pyrolyzed similarly. Gas chromato grams of pure PVC and PVC with 10 wt-% SnOj pyrolyzed at 700aC shown in Figures 1 and 2 illustrate the effect of active metal oxide on the course of decomposition of PVC. l Fig 2 Fymgram obtained at 700*C of poly(vinyl chloriJe)-SnOt system (10 wt-7e SnO,). ucc 037491 740 UDA, NAKAN1SIII, AiND GOTO TABLE IA y Values of Pyrolyeie Products Peak Component Untreated 500*C 700*C y Preheated 500'C 700*C I Aliphatic hydrocarbon* 3.GO 31.87 5.07 33.26 2 Benzene 70.23 36.74 71.68 38.35 3 Toluene 4.98 9.87 5.40 8.14 4 Ethylbenzene 0,7ti 0.81 0.80 0.50 5A 0.48 0.70 0.00 0.51 6 o-Xylene 1.12 1.48 1.33 1.33 7 > (onochlorobenzene 0.37 0.14 0.30 0.00 8B 0.22 0.27 0.32 0.13 9 Styrene 1.63 1.85 1.52 2.08 10 C 0.08 0.14 0.12 0.12 11 Vinvltoluene 0.47 0.50 0.30 0.55 12 D 0.64 0.40 0.67 0.11 13 E 0.59 0.33 0.57 0.14 14 p-Dichlorobenzone- 0.09 0.06 0.08 0.00 15 o-Dichlorobenzene 1.58 2.46 1.40 3.19 Indene 16 F 0.40 0.29 0.53 0.00 17 1,3,3-Trichlorobenzene 0.25 0.11 0.39 0.09 18 G 0.52 0.39 0.38 0.45 19 1,2,4-Trichlorobenzene 1.35 0.66 1.26 0.70 20 H 0.13 0.06 0.20 0.00 21 I 0.70 0.27 0.26 0.28 22 Naphthalene 6.22 7.28 5.39 7.06 23 a-Methylnaphthalene 0.94 1.40 0.99 1.16 24 S-Mcthylnaphthalene 0.73 1.14 0.66 1.03 25 J 0.18 0.19 0.14 0.09 26 K 0.50 0.48 0.45 0.50 27 L 0.00 0.00 0.08 0.13 10% TiOi 500*C 700*C 9.35 61.53 6.98 1.09 0.87 1.93 0.42 0.42 2.06 0.16 0.76 0.90 0.71 0.20 2.15 42,39 31.87 10.57 0.91 0.97 1.78 0.12 0.34 1.32 0.19 0.42 0.38 0.31 0,06 1.68 0.74 0.33 0.85 1.33 0.25 0.55 3.82 0.97 0.63 0.15 0.40 0.00 0.71 0.19 0.22 0.55 0.07 0.32 3.46 0.77 0.62 0.00 0.00 0.00 At 2001,C, no hydrocarbons could be detected in pyrolysis products in any of the samples. There are 11 peaks at 300C, 20 peaks at 4008C, 26 peaks at o00C and more at the higher temperatures in the pyrograms of pure PVC. The percentage ratio y of each peak height to the sum of all peak heights is shown in Table I. Pyrolysis of each sample at each tem perature was repeated at least 5-20 times. No variation of peak height ratio y was observed with pyrolysis periods of 15 sec and 20 min, respec tively. The peaks corresponding to lettered components A-L represent unidentified materials. The quantities of these are small and even though these unknown materials arc shown to be present, they are not considered in the conclusinns*of this paper. The lower aliphatic hydrocarbons in the products hove .been collected into one peak, because FEG-20M was used in the separation columns. In pyrolyaing PVC, an elevated temperature favors formation of ali phatic hydrocarbons, while aromatics are more easily released than aliohatics at lower temperatures. ucc 03749 f I'01jY(VINYL CIIL0K1DK). 1 741 Peak 1 2 :i 4 5 0 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 TABLE IB * y Vslues of Pyrolysis Product* 10% SnO, Component 500*C 700*C Aliphatic hydrocarbons Benzene Toluene Ethylbenzene A o-Xylene Monochlorobenzene B Styrene C Vinyltoluene D E p-Dichlorobenzene o-Dichlorobenzene Indene F 1,3,5-Trichlorobenzene G 1,2,4-Trichlorobenzene H I Naphthalene o-Methylnaphthalene 8-Methylnaphthalene J K L 32.70 34.57 7.24 2.04 1.38 2.46 0.57 1.04 2.06 0.37 0.49 0.73 0.79 0.45 0.70 64.46 18.49 6.79 0.44 0.92 1.10 0.69 0.30 1.45 0.24 0.60 0.23 0.09 0.14 1.25 1.10 0.59 0.25 0.36 0.18 2.50 4.79 1.11 0.78 0.00 0.00 0.00 0.30 0.05 0.23 0.60 0.13 0.15 4.26 0.86 0.88 0.00 0.00 0.00 y 10% ZnO . to' AUOi 500*C 700*0 , soot 700*0 24.26 41.16 6.03 1.94 0.00 1.46 0.69 1.11 0.97 0.42 3.60 2.15 0.28 1.94 2.91 GO.27 15.20 6.67 1.12 1.40 1.13 0.18 0.89 0.75 0.48 0.36 1.20 0.68 0.13 1.70 7.70 61.02 7.67 1.20 0.85 1.75 0.47 0.47 2.14 0.15 0.70 0.96 0.83 0.16 2.15 37.25 32.04 10 40 0.90 0.86 1.57 0.00 0.33 2.02 O.lfl 0.58 0.35 0.38 0.04 2.56 0.97 1.11 1.39 0.83 0.44 1.18 3.19 1.94 0.00 0.00 0.00 0.00 1.25 0.37 0.48 0.75 0.50 0.36 2.04 1.22 0.50 0.22 0.17 0.34 0.75 0.31 0.82 1.31 0.21 0.78 4.51 1.06 0.78 0.18 0.45 0.00 0.10 0.00 0.44 0.76 0.13 0.36 5.75 1.22 1.01 0.13 0.47 0.00 An important investigation in this work is the effect of added metal oxide on the formation of aromatics. Although preheated PVC and PVC mixed with metal oxides Bhow a temperature effect, aromatic hydrocarbons are more easily released from the polymer than aliphatics in pyrolysis of pure PVC PVC-TiOt, or preheated PVC. On the other hand, aliphatic hydrocarbons are more easily released during pyrolysis of PVC with ZnO and SnO". In particular, a peak of aliphatics can be found in the pvrograms of PVC-ZnO and PVC-SnO, systems at 300C, while it appears above 400 C in the other systems. . Decomposition yields of some aromatics obtained by the pyrolysis of PVC alone at each temperature are shown in Table II. These data were obtained by gas chromatography with p-dichlorobonzcne as an internal standard. With increasing temperature, formation of naphthalene in creases in comparison with benzene. More aromatics, futhermon;. are released from the preheater! PVC. These results indicate that the initial i } > 1 712 MDA. NAKAMS1II, AND GOTO TAB1-E II Yields of llecomposilion Product* (Raw Polymer) with Pyrolysis at Varioun Temperatures Product 300*C 400*C Yield, wt-% 500"C 600*C 700'C Benzene Toluene Ethylbenzene o-Xylene Styrene Naphthalene 3.50 -- -- -- -- -- 4.67 0.21 -- -- 0.13 0.47 5.00 0.67 0.16 0.16 0.22 0.52 5.23 0.82 0.18 0.18 0.37 0.73 6.30 1.55 0.18 0.23 0,74 1.50 800C 9.95 2.10 0.23 0.25 1.06 3.00 reaction it* a dohydrochlorination and then eyclization of polyene in poly mer chain begin*. As shown in Table I, frisubstituted benzene or naphthalene is not de tected except for aryl chloride, and all the disubstituted aromatics appear ing in the main peaks are ortho compounds. Accordingly, intemiolecular dehydrochlorination or cyclization is not expected during the decom position of PVC. This supports the results of infrared spectra reported by Olitani.* A secondary reaction of pyrolysis products with each other, e.g., FriedelCrafts reaction of toluene with ethylene, might not be involved, because only ortho substituents arc detected, even in materials obtained from pyrolysis of the PVC-A10 system. 'In this system, AljOi would be changed to AlGlj, so ortho and para substituents would be expected to be produced. However, compounds having para substituents cannot be detected. The aromatic pyrolysis products may be classified in three groups: group I contains substances with a molecular structure entirely constructed of conjugated double bonds such as benzene, styrene, and naphthalene; group II contains aromatic hydrocarbons with methyl endgroups, such as toluene, ethylbenzene, xylene, vinyltoluenc, 0-methylstyrene, and methylnaphthalene; group III contains various and chlorides such as mono-, di*. ami trichlorobenzene, chiorostyrcne, chlorotoluene, and chloronaphthalene. (yt) is defined as Oyi)# =" where hi is the sum of peak heights of individual substances belonging to group I, is the sum of peak heights of all the pyrolysis products. (71)0 is the ratio of group I peak heights to all peak heights obtained from the pyrolysis of raw .PVC. Similarly there are obtained group I ratios (ti)mo from the pyrolysis of the preheated PVC and of PVC containing metal oxide*. Ai, then, is expressed by (7i)no/(ttV The effect of metal oxides on the evolution of group I hydrocarbons from PVC may be estimated from At value. An and Am, similarly, will offer information on evolution of group II and group III pyrolysis products. Furthermore, values of 7 ucc 037494 POI.Y(VI\YI. CIILOIUOIC). I 7i:i and A have been obtained for certain individual pyrolysis products in the same way as for groups I, II, and III. Figures J-l 1 show plots of y or A as a function of temperature for each PVC composition pyrolyzed, specifically Ai in Figure 3. yn/m in Figure 4, Aii/i in Figure f>, y values for ethylbenzene (KH) and toluene (TL) in Figure G, the ratio of A values for the same compounds in Figure 7. y and A value ratios for o-xylone (OX) and toluene in Figures S and !), and y and A ratios for group III and group I in Figures 10 and 11. PYROLYSIS TBPERMURE PC) Fig. 3. Plots of Ai vs. pyrolysis temperature; () preheated PVC; (9) 10 wt-% TiOiPVC, 0)10 wt-%SnOr-PVC; (9) 10 wt-% ZnO-PVC; <) ll)f-% AMV-PVC. PYROLYSIS TtWCMTURE PC) Fig. 4. Pints of yu/yt vs. pyrolysis temperature: (O) untreated PVC; () pre heated PVC; (fl) 10 wt-' ,', TiOi-PVC; O) >0 wt-Vr rtnOi-l'VC; (*) 10 wt-'e ZnOPVC; () 10 wt-bi AliOi-PVC. IIDA, NAKANISHl, AND COTO i Fig. n. Plots of Au/i vs. pyrolysis temperature. Symbols as in Fig. 3. ft The temperature dependence of At is illustrated in Figure 3. In the pyrolysis of the PVC-ZnO or the PVC-SnOi system, less of the group I products in the evolved materials is found with increasing temperature. t It may be that ZnO or SnO is effective in breaking the polymer chains, u~it ;-i as evolution of aliphatic hydrocarbons is the result of an extreme rupture of chains. The relation of yu/yi to temperature of pyrolysis is shown j in Figure 4. Group II is more easily released than group I at elevated ! ft temperature, tlie relation of Au/i to pyrolysis temperature is shown in Figure X, where An/r is the ratio of ywhi for the polymer-metal oxide -H system to that of PVC alone. Ah/i, therefore, measures the effect of i addition or preheating upon the rate of formation of group II compounds relative to group I compounds. An/1 values for all experiments are equal to units at higher temperature! t ucc 037496 LsiUu POLY(VINYL CHLORIDE). I mOLTSIS TtKMTWE CO Fig. 6. Plot* of tn/yrk pyrolyni* temperature. Symbols m in Fig. 4. Fig. 7. Plots of Cbb/tt. vs. pyrolysis temperature. Symbols as in Fig. 3. Some results for VEs/m. and Aeb/tl are illustrated in Figure 0, Tabic III, and Figure 7 respectively, veb/vtl decreases with increasing temperature, but the curve for ZnO has the steepest slope. The curve for SnOt has upward curvature and a maximum at 500C. Aeb/tl has a constant value. For ZnO and SnO., the tendency is the same for teb/ ttl- The temperature dependences of Vox/ytl and of Aox/tl are illus trated in Figures 8 and 9 and Table IV. The rate of evolution of oxyionc is unaffected by addition of metal oxides, as shown in Figure 9. Tox/vtl decreases with increasing temperature. Aox/tl is unity through ucc 037497 1IDA, \AKAMSIII, AND GOTO Kin, S. Plots of tox/ttl vs, pyrolysis temperature. Symbols as in Fig. 4. PYROLYSIS TBKMTWE (*C) Fig. 0. Plots of Aox/n vs. pyrolysis temperature. Symbols as in Fig. 3. for all experiments, as shown by the horizonal line. The release of ethyl benzene is accelerated by ZnO and SnOi, but that of o-xylene is not. The ratio Aqx/tl is constant through the various experiments. ZnO is effective in producing release of group II compounds. Generally, dehydrochlorinatinn of 1JVC is completed below 400C, as reported by Matlack* and Salovey.10 ZnO is converted to ZnCh by reaction with HC1. This, it is suggested, accelerates dehydrochlorination.11 At same time, hydrochlorination is also advanced. Therefore, ZnO contributes ucc 037498 AtUUtKJA0.-t l*OLY(VI\YL CHbORlDK). I rtr Fig. 10. Plots of yin/n va. pyrolysis temperature. Symbols a* in Fig. 4. PYMLYSIS TBVEMTtME <*C) Fig. 11. Ptota of Ai/in vs. pyrolysis temperature. Symbols ns in Fig. 3. much to the isolation of methylene. It has been pointed out by Strombcrg and coworkcrs11 that the methylene content in PVC which has under gone a thermal treatment is high. Their conclusion is in good agreement with the results described in this paper. The pyrolysis temperature de pendence of yiu/yi and of im/i is shown in Figure 10 and Figure 11. 7111/71 increases at the elevated temperatures. The curve for ZnO has upward curvature and a maximum at GOO'C. Am/i has tin- same trend as yui,'li- The initial structure of polymer is independent of evolution of arvi chlorides (Ani/i is not equal to unity, as shown in Fig. 11). 74# Pyrolysis temp, *C 300 400 300 600 700 800 hda, yakamshi, and c<rro TABLE III -i*/TL for Thermal Decomposition Aki/ti. Untreated Preheated 10% TiO, 10% SnO, 10% ZnO 10% AliOi _ _ _ __ 1.00 1.00 1.00 1.00 0.00 0.97 0.06 0.73 0.79 1.02 1.12 1.05 0.89 1.83 1.28 0.70 2.70 2.11 1,84 2.05 0.91 1.03 1.00 1.12 Pyrolysis lenip. "C 300 400 .300 600 700 800 TABLE IV Aox/tl for Thermal Decomposition doX/Tt, Untreated Preheated 10% TiO, 10% SnO, 10% ZnO 10% Aid), 1.00 1.00 1.00 1.00 1.1X1 1)6 1.10 1.00 1.09 1.10 0.78 1.23 1.22 1.12 0.90 0.63 1.51 1.43 1.14 1.32 1.05 1.08 0.97 1.13 1.16 0.97 1.01 0.99 1.00 0.97 Chlorinat'd polybutadicnc, chlorinated polyethylene, vinyl chloridevinylidenc chloride co|xilymer. and chlorinated PVC were pyrolyzed pre viously and their pyrolysis products were described in detail by Tsuge.1,~1* Chlorinated polyethylene, containing more than 0.71 Cl for every two carbons, was required to release mono- and dichlorobenzene. For chlo rinated polybutadiene, it was 1.57 Cl per four carbons, and it was 2.32 Cl for release of tetrachlorobenzene. On pyrolysis of chlorinated PVC, tri- and tetrachlorobenzene were increased. Monochloro-, m-dichloro-, and l,3,5-trichiorobenz<nie were released from vinyl chloride-vinylidene chloride copolymer. The recombination of chlorine during the pyrolysis of PVC may be a possible mechanism. The author* wish to thank Mr. K. Suetsugu for hia asaiatance in the experimental work ami Geon Co. Ltd. (Japan) for the PVC. Reference* 1. It. It. Htmmbcra, S. Straus, and B. G. Achhammer, J. Priym. Set., 35, 353 (19.30). 2. F. Tsnctiiya and K. Sumi, J. Appl. Chem,, 17, 364 (1967). !l, M, Ohta, Kogi/n Kngoliu Ztuthi, 55, 31 (1052). 4. M. M. O'Maru. J. Polym. Sri. A~I, S, 1887 (1070). 5. E. A. Boettner, G. Ball, and B. Wefcn, /. Appl. Polym. Sri., 13,377 (1069). 6. 1). NofTi, W. Beni, and W. l*fab, Z. Altai. Chtm., 235,121 (1068). ucc 037500 ,.V> A.V; ,S. POLY(VINYL CHLORIDE). I 7-. 7. S. A, Liebman, D. H. Ahlslrom, E. J. Quinn, A. G.'Geigley, and J. T. Meluskey, J. Polym. Set. A-l, 9. 1021 (1971). 8. S. Ohtani, Kogye Kagaku Ztuxhi, (1,447 (1968). 9. J. P. Matlock and A, P. MeUger, J.Appl. Polym. Set., 12,1746 (1908). 10. U.SalovcyandH.E, Bnir,y.Apj>/.Po/vw.Sct., 14,713(1970). 11. II. Nagatomi and Y. Saoki, Kogyo Kagaku Zattht, 65.303 (1902). 12. K. 11. Stromberg, S. Straus, and B. G. Aohhammer, J. Ret. Mai. litir. Stand., 60, 147 (1968). 13. H. Ito, S. Tsuge, T. Okumoto, and T. Takeuchi, Makromol. Chem., 138, 111 (1970). 14. S. Tsuge, T. Okumoto, and T. Takeuchi, Bull. Chem. See. Japan, 42,2870 (1909). 15. S. Tsuge, T. Okumoto, and T. Takeuchi, Makromol. Chem., 123,123 (I960). 16. S. Tsuge, T. Okumoto, and T. Takeuchi, Maeromoleeulte, 2,277 (1909). Received June 12,1972 Revised January 3,1974 t t). t, ucc 037501 `Y