Document pmOkdz4YQqg22ZGONnx7RR8k6

B. F. Goodrich Chemical Company A DIVISION OF THE B. F. GOOONtCM COMPANY DEVELOPMENT CENTER "6 7y Degradation and Cottbus Cion Products From Poly(vinyl chloride) II - Description of Pyrolysis Unit snd Initial Pyrolysis of PVC Dace Completed: by M.M. O'Hara December 2, 1968 Date Issued: December 6, 1968 Distribution: Akron General Chemical Plant P.T. Whitmire - R.S. Reynolds Akron Legal Dept. E.K. Bean Avon Lake General Chemical Plant R.N. Rylands-R.W. McKay-J.M. Whitney BreckavlUe Research Center R.J. Fawcett C.F. Gibbs Calvert City *C.L. Woods - D.8. Schrock Cleveland V.F. Bixby W.E. Brodlne G.H. Metzger P.H. Lawrence E.W. Harrington M.W. Larson 0 C.D. Segner (7) R.D. Scotc - J.L. Nelson B.M. Zwlcker - J. Valentine Development Center L.F. Arnold L.A. Benne c c *R. R. Bloor L.H. Conklin B.A. DiLiddo C.E. Fleming C.H. R.M. *C.H. D.P. Kotheimer K re a ge r Luf ter Knecht ge s Development Center continued R.J. Meyer B K. Mikofalvy C.E. Parks N.H. Sherwood J.A. TePas G.L. Wheelock Technical Council R.J. Wolf C.T.F. (3) Henry C.B. Cooper - J.P. Piers Long Beach L.G. Crunkleton - S.S. Michels ra fan* T.R. Llnak - J.A. Klupar Modified Report 20055001 I - MAJOR OBJECTIVE: The major objective in this work is the development of the necessary analytical tools for the rapid determination of the degradation and combustion products frcrr. PVC. In addition to studies of PVC, the developed analytical-pyrolysis tool will be used in the following areas: 1. toxicity, 2. flammability, 3. polymer structure and stability, 4. efficiency of PVC stabilizers, 5. polymer identification (fingerprinting). II - LIMITED OBJECTIVES A pyrolysis unit that afforded the following capabilities was desired: 1. the temperature of pyrolysis accurately known, 2. wide range of pyrolysis temperatures possible, 3. all volatile pyrolysis products be admitted to the gas chromatograph for analysis, 4. capabilities for pyrolysis in an oxygen rich and oxygen poor atmosphere, 5. reproducibility. Ill - SUMMARY OF CONCLUSIONS; 1. A radiant heat pyrolyzer having the necessary capabilities for degradation, toxicity, flammability and polymer characterization studies has been constructed. 2. Initial pyrolyses of PVC (Geon 103EP F-7) in both a nitrogen and helium atmosphere have been carried out. The result of these studies showed that at 600*C., PVC degrades to 21 major (identifiable) and 25 minor components. 3. The major degradation products have been identified by mass spectromecrlc techniques. These products are: l. HCi 2. CO2 (and N2) 3. e thane 4. ethylene 5. propane 6. propylene 7. butane b. butene 9. butadienc 10. diacetylene i i . h# nt# rw. 20S55002 Page 2 12. toluene l3'\14. >3 xylenes J15. 16. chlorobenzene 17. ethyltoluene 18. indane 19. naphthalene 20. 21. Isomers of methylnaphchalene 4. A quantitative analysis encompassing volatiles and non-volatiles has been completed. The results of that determination are contained in the following cable. t -Uc --58 (Theoretical) -10 -32 Composition of Fraction HC1 Carbonaceous Ash(non-volacile) Volatile Aliphatic and Aromatic Hydrocarbon Fraction IV - FUTURE ACTIOS; The next phase of this research problem is: 1. carry out the pyrolysis of PVC in varying levels of oxygen and compare and correlate the results with those pyrolysis results obtained under inert conditions. 2. pyrolyze copolymers of PVC and other polymers in order to obtain a catalog of reference pyrograms (polymer characterization) 3. a study of the flammability characterlstics of PVC-AcrylatePVDC-Phosphonate-Polymers has been initiated. (Preliminary results are very encouraging.) 4. initiate a program in the thermo-oxidative degradation of Hydrin 100 and 200. This work will be in support of current studies on the Hydrin process. 20655003 Pago 3 V - INTRODUCTION AND DISCUSSION In a recent outline of object Ives,^ it was stated that a pyrolysisgas chroma tograph-mass spoct rome ter tandem system was being developed for purposes of carrying out polymer degradation studies. The gas chro-o. :ogr ;ph:c technology required for separating the typical degradation products fro-. ?7C has been developed and detailed in a status report.(2) The present paper deal< with the pyrolysis apparatus that has been constructed for these studios, with initial pyrolysis of PVC and with the identification of the major pyrolysis products from PVC. It was deemed necessary that the pyrolysis unit have a number of capabilities. These were: 1. that the temperature of pyrolysis be accurately known, 2. chat a wide range of pyrolysis temperatures be possible, 3. Chat all pyrolysis products be admitted to the gas chromatograph for analysis, 4. chat pyrolysis could be carried out under an oxygen rich and oxygen poor atmosphere, and 5. Chat some degree of reproducibility exist. In ail of Che areas in which this pyroly s-gas chromacographichii spectrometric tool will be used (Sec I, Major Objectives), it will be important that all volatile degradation products be analyzed. For studies in which polymer structure is to be correlated with degradation products, the temperature of pyrolysis must both be accurately controlled and known. Similarly, the capability of a wide range of pyrolysis temperatures la important. The need for being able to control pyrolysis atmosphere is important with reapect to simulating usual combustion conditions. Finally without some degree of reproducibility, correlations of degradation products and toxicity, or flammability, for example, would be impossible. It is felt that the pyrolysis unit constructed and described in detail below meets the above requirements. A. Description of the Pyrolysis Unit The unit basically consists of a quartz chamber which is heated by an external furnace. The furnace is thst which is provided with the P and E commercial unit and temperatures of up to 1000*C. are possible. The quartz chamber contains an inlet l ir.e through which carrier gas (He) can be admitted to the pyrolysis chamber and an effluent (exit) line through which the degradation products (and carrier gas) are carried. These two lines have been connected to an F and M Cas Sampling Valve in such a way so as to replace the noccvil "samplIna loop". A carrier gas line from the chromatograph leading to the valve and ;'n effluent line leading from the valve to the injection p'rt of the chror.itogrjph have alao been constructed. o era Cl Col o 20655005 O^'l'V 1 Page 4 With the valve in the "in" position, carrier gas by-passes the pyrolysis chamber (l.e., the "sample locp") tnrough the valve and travels directly to the injection port of the gas chron.a tograph. With the valve in this position, the pyrolysis unit is then loaded with a porcelin boat containing sar-plc, the hot zone of the charier is brought to temperature and the entire unit is purged with N'2 or N2/02- The boat is then moved into the hot zone where degraditlon is carried out. Once the degradation period has been completed, the gas sanpllog valve is switched to the "out" position. Carrier gas is then admitted to the pyrolysis unit and the degradation products are carried from the unit, through the heated sampling valve to the injection port of the chromatograph. (See accompanying block dia gram). All effluent lines through which the pyrolysis products pass, as well as the gas sampling valve Itself, are heated. Provisions were made during the construction of the effluent line to the injection port of the chromatograph for "on-column" injection. A calibration was made between the temperature read-out on the pyrolysis unit and the actual temperature Inside the hot zone of the pyrolysis chamber so that the true temperature of pyrolysis could be known. B. Volatilization of an Analytical Sample in the Pyrolysis Chamber. An analytical sample of benzene, toluene and naphthalene was prepared and analyzed on the gas chromatograph by two methods. The first was the usual direct injection procedure (Method A) while the second Involved placing the mixture in a sample boat, volatilizing the sample in the hot zone of the pyrolyzer and sweeping the vapor from the pyrolyzer to the chromatograph (Method B) for analysis. The results of these two runs are contained in Table I. It is believed that the 31 difference In analysis is a result of the differences in the two methods and not due to a "selective extraction" of one of the products while in transit from the pyrolysis chamber to the gas chromatograph. If this were happening, one would expect the naphthalene content to be low in the analysis obtained by Method B. However, the opposite of this results: Naphthalene is 31 higher when the analysis Is carried out by Method B. TABLE I - C.C. Analysis of Analytical Mixture. Method A: Direct Injection; Method B: Sample Volatilized in Pyrolysis Chamber. Method A Benzene Toluene Naphthale r.e X 41 48 11 Method B Jl. 38 48 14 20655006 Page 5 It should bo noted that while in Method A the chromatographic peak due to benzene is sharp with no tailing, the same peak by Method B is broad and tails considerably. This factor Is the most probable source of error in this analytical de ternlnat ior.. C. Initial Pyrolysis of PVC. Identification of Major Degradation Products Preliminary degradations of PVC were carried out at 600*C. for 5 minutes in an inert atmosphere (N2 or H). Initially, sample sizes in the 5 co 10 mg. range were used but due to the preponderance of products (other than HCl), amounts in the 20-30 mg. range were necessary. Since HCl has an adverse effect or. the chromatographic column used for separating the degradation products CSE52 on ABS), it was removed through reaction with copper powder. In a typical pyrolysis run, a 25 mg sample of PVC (Geon 103EP F-7) produced, upon thermal decomposition, 21 major and 25 minor products. The identification of the major products has been completed and the results of these identifications are contained in Table II. HCl CO2 Ethane Ethylene Propane Propylene Butane Butene Butadiene Diacetylene Benzene Toluene Xylene (3 Isomers) Chlorobenzene Ethyl toluene Indane NaphthaleneP-Me thylnaphthalenc a-Methylnaphthalene 20655007 Page 6 A typical pyrogram of PVC can be divided Into three regions: volatile gases (COj C^H.;, C2H6, etc.), low boiling saturated and unsaturated hydrocarbons (C^-Cfc) and high boiling aromatics (benzene, toluene, naphtha loin-). In order to more fully Investigate the volatile gas products (the low end of the pyrogran), gas chrcrotographic separation of these particular pyrolysis products on a Forapak QS column wjj necessary. Thus, In order to Identify all degradation products, analysis on a Porapak and on an SE-52 column had to be carried out. The minor degradation products from PVC could not be Identified due to sensitivity limitations In the mass spectrometer. However, the majority (18 of 25) of these minor products appear In Che "aromatic region" of the PVC pyrograa while the remaining minors (7) are In the light hydrocarbon region. Boettner and Keiss^^ have reported that both aliphatic and aromatic products are evolved when PVC is degraded In the presence of air (TABLE III). A more recent report(^) by Noffz indicates that the only degradation products from PVC (aside from HC1) are aromatic type compounds (TABLE IV). TABLE III - Weight Loss vs. in Air.(3) H55S Aromatic (4OX) CO HCl (97X), Benzene (3X) Aliphatic (30X' CO and CO, (?0X) CO2 ^OX 2 OX weight loss OX 1. Benzene 2. Toluene 3. Ethylbenzene 4. Xylene 5. Chlorobenzene 6. Xylene 7. Styrene 8. n-Propylbenzene 9. Allylbenzene 10. Ci-Methyl styrene ll. Azulenc 12. Indanc 13. Me thylIndanc 14. Indcnc 15. Tetralln 16. K: thylindane 17. Naphthalene 18. Methyl naphthalene 19. a* Me thy l naphtha ler.e 20. Diphenyl me thane 21. C*2Hl2 ** ers Cl Cl 0 0 00 Page 7 The results from the work described in the present paper correspond fairly closely to the findings of Boettner and Weiss. Although Noffz has reported that the only products arising from the thermal degradation of PVC are various aromatics, chc actual state of affairs appears to be that both aliphatic, olcfinic hydrocarbons and aromatics are formed. One of the more interesting aspects of the degradation products coming from PVC is the large amount of CO2. Direct probe experiments involving the degradation of PVC in the heated inlet system of the mass spectroiretcr have verified that a large amount of CO2 is released during degradation in vacuum. This result reflects one of two situations: a considerable amount of absorbed oxygen on the PVC resin or oxygen chemically bonded to the backbone of chc polymer. In order to further investigate this, a sample of the resin was placed in the cold zone of Che mass spectrometer inlet system and evacuated overnight (10~& nxn Hg). The sample was Chen moved to the hot zone of the inlet system where degradation proceeded. The amount of CO2 released from the degassed sample of PVC was identical to that released from a sample chat had not undergone overnight degassing. The results, however, cannot be construed as evidence for oxygen bonded into the PVC polymer. It may well be that the released CO2 results from absorbed ger. that is not effectively removed during the vacuum treatment. -s phenomenon of CO2 generation during Che pyrolysis of PVC in an inert atmosphere is being further investigated. Another factor that requires limited investigation is the effect of secondary reactions on Che types of degradation products. Since the pyrolysis products from PVC reside in the hoc zone of Che pyrolyzer for relatively long periods (5 minutes) of time, secondary degradations of primary products are quite possible. This important factor will be further investigated by varying the temperature and time of degradation. D. Quantitative Analysis Sumnsrv: A quantitative analysis of the degradation products from PVC at 600*C. reveals that approximately 58T (by weight) of the polymer is converted to HC1, 321 to aliphatic and aromatic hydrocarbons and 10X remaining as a carbonaceous ash. It has previously been stated chat the large amount of CO2 generated during the pyrolysis of PVC cannot be explained in the absence of further data . U^4 M 20S55009 Pago 8 The an.oun: of CO2 generated during a typical pyrolysis accounts for over half of tho volatile fraction of the degradation products. In view of this d isc r-: r s.icy, a precise quantitative analysis is not possible at this time. However, a fairly good approximotion of the quantity of volatile products in relation to the quantity of starting polymer can be rode. Since only 307. of the polymer is covcrted into a hydrocarbon and CC^C?) fraction, uncertainty in the origin of the CO2 does not result in too large of an error. In order to determine the amount of residue chat results from the 600*C. degradation of PVC in an inert atmosphere, a 21.7 mg. sample of the polymer was pyrolyzed at 600*C. for 5 minutes in the absence of copper powder. The residue (2.10 mg) thus accounts for approximately 107. of the sample, by weight. Since 58'. by weight of the polymer is HC1, this leaves approximately 327. of tit original polymer that is converted into the hydrocarbon and CO2 (?) fraction. [Boettner and Vciss have reported that 16T (TABLE III) of PVC is converted into aromatic and aliphatic hydrocarbons] The following analysis (TABLE V) is of ^his 3j!7 vol_<K_il_e ^rjc_tjkm that is produced during the pyrolysis of TvC at f>06*C. In each of the two quantitative analyses* that were carried out, two exact weight samples of PVC were degraded (for each analysis) under the same conditions. The products from one of the degradations were analyzed on the Porapak QS column while the products from the second degradation were analyzed on the SE52 column. In this way, both the volatile gases and the higher molecular weight aromatics could be precisely analyzed. This dual degradation-analysis procedure was carried out twice and the results of these two analyses are presented In Table V. The results demonstrate the excellent reproducibility that is obtained in these pyrolysis experiments. For example, in only one Instance do the two analyses differ by more than 2% absolute. In that case (butane, butadiene, butene, diacctylene), the calculation of the gas chromatographic peak area was difficult due to the dissymmetry of the peak (See Appendix). Approximately half of the volatile fraction consists of saturated and unsaturated hydrocarbons while the remaining products are the aromatics. Of these, 687. Is benzene. It should be noted that the CO2 content that results in each analysis is not Included in the calculations. Thus, the quantitative analysis involves only the hydrocarbon fraction. This analysis can be adjusted if It is found that the large amount of CO2 does indeed result from the degradation of PVC. 20555010 Page 9 I ! i TABLE V - Quantitative Analysis of the Volatile Fraction (32% bv weirht) fr.-m PVC.*___________________ Component Analysis 1 C.C. Peak Area 7. n2 cha i co2 127.1 58.2 367.4 18 - C2H4 21.7 7 c2h6 30.2 9 c3h6 c3h8 Butane Butadiene Butene Diacetylene (?) | V I J 25.7 9.5 23.2 8 3 7 Benzene 109.4 34 Toluene Xylenes ( 3 isomers) Chlorobenzene Xndane i Ethyl toluene 14.8 7.7 7.4 1.3 1.7 5 2 2 -l Napththalene Methyl naphthalene (cr-?) rMethyl naphthalene (3-?) _> 7.7 32 2 l Analysis 11 C.C.P.A. 7. 31.2 38.2 20 232.0 - 12.6 7 17.2 9 11.4 6 3.6 2 8.0 4 67.2 9.3 4.7 5.3 1.5 1.4 5.6 3.1 35 5 2 3 -1 -l 3 2 20855011 Z 321.7 100 189.3 100 Page 10 1. L.B. Crider, M.M. O'Mara - "A Study of the Degradation and Combustion Products from Polymers. Objectives and Problem Description" Inter-Organization Correspondence. B.F.C. Chemical Company, August 2, 1968. 2. M.M. O'Mara, "Degradation and Combustion Products From Poly(vinyl chloride) : part I, September, 1968. 3. Boettner and Weiss, American Industrial Hygiene Assoc. Journal 28. November-December, 1967. 4. Noffz, Benz, Pfag, Z. Anal. Chem. 235 #1 p. 121-37 (1968). o O CT CT O 20S55013