Journal of Research of the National Bureau of Standa rds Vo!' 61, No.2, August 1958 Research Paper 2888 Thermal Degradation of Polyacrylonitrile, Polybutadiene, and Copolymers of Butadiene With Acrylonitrile
and Styrene 1 Sidney Straus and Samuel 1. Madorsky Polybutadiene and a co polymer consisting of 76.5 percent butadiene and 23 .5 percent of styrene were investigated as to t he rates of t heir t hermal degradation. The rates for polybu tadiene indicated an activ:ttion energy of 60 kilocalories per mole. The copolym er had very high init ial rates of degradation, followed by a rapid cirop, so t hat it wa~ not possible to obtain a reliable activation energy. Polyacrylonitrile and a copolymer consisting of 31 percent of acrylonitrile a nd 69 percent of butadiene were investigated with regard to t he nature and distribut ion of volatile products, as well as to t he rates of t heir thermal degradation. The rates were very high in itially, but dropped rapidly 50 t hat it was not possible to determine accurately the activation energy. Thp, acrylonitrile-butadiene copolymr r s howed a rlecomposit ion pattern similar to t hat of polybutadiene. Here, too, t he acLivation energy could !lot be rlec!uced accurately from the rate curves. Comparati\Oe thermal ~tab ili ty in terllls of temperature, at which half of the polymer sample is evaporated in 35 m inutes of heating, is as follows: -!07°C for polybutadienc, 374°C for SBll, 36+° C for polystyrene, 360° C for NEll, and3JGo C for polyacr ylon itrile. 1. Introduction more stable than the copolymer. Tbe ratios of small molecular fragments to large ones were 14 :84 for poly The copolymcrs, styrene-bu tadiene (GR- S) and bu tadiene and 12 :88 for SBR. acr ylonitrile-butadiene (niLrile rubber), curren tly In this inves tigation a study was made of rates of designated in the technical litcrature as SBR and thermal degradation of these two subs Lances b.\~ the NBR, respectivel.\', rcpresent two of the more im loss-of-weight method, using a very sensitive tungsten portant synthetic rubbers. This paper describes tbe spring balance enclosed in a vacuum [3]. Samplcs of results of an investigation of the relative thermal poly butadiene and of SBR (76 .5% of butadiene and stability and of the products of degradation obtained 23.5 % of st ~Te n e), weighing about 5 mg, were used. when these rubbers are pyrolyzed in a vacuum. A The materials were prepared b.\' purifying commercial similar description is given of the thermal behavior products usin g a method whiclt has been described of polyacr~T lonitril e and polybutadiene. previously [2]. D etails of the method employed in this investiga R esults of rate studies for polybu tadiene arc shown tion have been described previously [1 - 5].2 T he in figure 1. About 12 to 16 percen t of the samples method consists of two parts. In the first part, a was vaporized at zero time ; this vaporizaLioll Loo k sample weighing 25 to 50 mg is heated in a vacuum, place during the approxima lel.\' 15 min he ating-up first for 5 min dUTing a heating-up period, then for period. Zero time in these experime nts W itS reckon ed 30 min at a constant temperature; the residue is from the moment the sample had reached the operat- weighed, and the volatile products fractionated. The fractions are then weigh ed and analyzed in the 100 ,--.--r--,-----,,---,--,----.-.,--,.--,-,--,.--, 2.0 mass spectrometer or tested for average molecular weight by a micro cryoscopic m ethod. I n the second 1.6 part, a 4 to 5 mg sample in a platinum crucible w OJ suspended from a very sensitive tungsten spring o iii w balance [3] is heated in a vacuum, and the rate of a: loss of weight of the sample is observed at intervals. 1. 2 t5 The activation energy of degradation is calculated ;:: z w from the initial rates by using the Arrhenius equation. u a: lL w 040 . 80. 2. Thermal Degradation of Polybutadiene W lL ~ 0 ;:: 0 and SBR z ~ w 0 U ...J 520 .4 A study of th e relative tbermal stability of poly 0. butadiene and the copolymer SBR, and of yields and chemical nature of the products resulting from o 0 pyrolysis of these two materials, has been reported o 80 160 240 320 400 460 previously [2,4]. Polybutadiene was found to be TIME FROM START OF EXPERIMENT , MINUTES F IGURB 1. Thermal degradation of poly butadiene. 1 Presented before the 132d Meetlng of the American Chemical Society, New York, N . Y., September 7- 12, 1957. Solid lines represent percentage volatilization versus time and dashed lines , Figures in brackets indicate the literature references at the end of this paper. represent logaritbm of percentage residue versus time. 77 ing temperature. The fact that the logarithm of I.O,-----,----,------,---,------r--,----, percentage residue versus time curves deviate only sligh tly from straight lines indicates that the reaction w S.8 involved in the thermal degradation of polybutadiene z approaches first order. ~ a: w In figure 2 rate of volatilization is plotted versus a. w.6 amount volatilized. The apparent initial rates are ..J a. obtained by extrapolating the straight parts of the ::; rate curves to the ordinate. These rates, in percent '"Ul age of sample per minute, are 0.21, 0.31 , 0.42, and "° .4 0.63 for temperatures 380°, 385°, 390°, and 395° C, respectively. On plotting logarithm of the initial rates versus the inverse of absolute temperature, a straight line is obtained, whose slope indicates an activation energy of 62 kcal/mole.
Percentage volatilization of SBR versus time is 10 20 30 40 50 60 70 shown plotted in figure 3 (solid lines). H ere again AMOUNT VOLATILIZED, PERCENT there is considerable loss of weight during the heating FIGURE 4. Rates of volatilization of SEll. up period. Plots of logarithm of percentage residue versus amount volatilized (dashed lines) deviate 3. Thermal Degradation of Polyacrylonitrile considerably from straight lines. The reaction involved in the thermal degradation of SBR appears and NBR to be more complicated than that for polybutadiene and the reaction is far from first order. R ate of The polyacrylonitrile 3 of high purity and in the volatilization versus amount volatilized is shown form of a fine white powder, had a number-average in fi gure 4. The rates, very high initially, drop molecular weight of about 40,000 . The NBR rapidly with extent of volatilization. It is not rubber was Goodrich H ycar 1042 consisting of about possible to determine with any accuracy the activa 30 percent of acrylonitrile and 70 percent of buta tion en ergy from these rate curves. diene. The rubber was extracted for 24 hI' with ethanol-toluene azeotrope in order to remove im purities. It was then kept in a vacuum at 60 to 70° C for 5 hr. The purified material analyzed 8.2 per cent of nitrogen, which corresponds to 31 percent of acrylonitrile. Specimens of polyacrylonitrile and NBR were heated prior to pyrolysis, in a vacuum at 110° C to a co nstant weight. R esults of pyrolysis are shown in table l. Frac tion V _ 1904 represented less than 0.1 percent of the total volatilized and is not shown in this table. As in the case of many other polymers the ratio V 25 : Vp yr remained constant tlll'oughout the temperature 10 20 30 40 50 60 70 80 90 AMOUNT VOLATILIZED ,PERCENT range. On the average, this ratio is 12 : 88 in the case FIGURE 2. Rates of volatilization oj polybllladiene. of polyacl'ylonitrile, and 14: 86 in the case of NBR. In the latter case, in calculating the average ratio, results of experiment 1 were excluded, because vola tilization was only 9 percent. As seen from table 1, volatilization of polyacry 1.6w lonitrile stabilizes at about 70 to 73 percent. T his :::> Q Ul is in agreement with the work of Kern and Fernow W a: [6], who found that dry distillation of this material r.2~ was accompanied by carbonization of the residue. A Z possible explanation of this stabilization, as given by W u M cCartney [7] and Houtz [8], is the formation, along 5 .8a. the chain, of ring structures involving nitrogen from "o the CN groups. NBR, which con tains only 31 per o ~ o cent of acrylonitrile, does not show any stabilization • 4..J through the entire temperature range . Fractions Vpyr from polyacrylonitrile and NBR had a tan waxlike appearance. They were not o 80 120 160 200 240 280 readily solu ble in ordinary organic solvents, including TIME FROM START OF EXPERIMENT, MINUTES
FIGURE 3. Thermal degradation of SEll. 3 The anthors are indebted to O. A. Sperati of the Polychcmicals D epartment, E. 1. duPont de Nemours & 00., for supplying them with this polymer, which Solid lines represent percentage volatilization versus time and clashed Jines is LI sen in the production of orIon fiber. represent logarithm of percentage residue versus time. , For defi nition of fractions sec footnote to tahle 1. 78
- ~~------TABf"E 1. P yrolysis of polyacrylonarile and NBR The main constituents in V25 are HON, acryloni trile, and vinylacetonitrile. The composition of V25 fl Volatile fractions a seems to depend on the time interval elapsing based on total vola· Experiment number T empera· VolaWiza· tilized beLween pyrolysis and analysis. This suggests t.he tnre tbn 1-----.----1 possibility that some of the component, parLicularly HON and acrylonitrile, react and become non volatile on standing and, as a result, the remain in g POLYACRYLONrTHILE volatile constituents become accentuated in the °C % % % analysis. This could explain why Houtz [8] r e L ...... 250 21. 5 86.3 13.7 2 ..•...... ••..•.•. 250 20.7 89.1 10. 9 ported that pyrolysis at 400 0 0 yielded only a trace 3 ...... 250 18.0 89.0 11. 0 4 ..•.•...... •••••••... 2i5 34.9 88.0 12.0 of HON. On the otber hand, Hideo Nagao and co 5 .•.••...•...•.•...•.•.• 300 46.6 86. 3 13.7 6 __ •• ______workers [9] reported the evolution of a considerable 7 ______• ______325 53.4 88.4 11. 6 8. ______325 52.3 886 11. 4 amount of HON when polyacrylonitrile was heated 9. ______325 49.7 87.4 12.6 0 0 32.> 50.6 88.8 11.2 at temperatures from 200 Lo 350 0 in an aLmos 10 ______350 57.7 88.7 11. 3 phere of air or niLrogen. Burlant and Parsons [10] 11 ______350 57.9 87.8 12.2 12 ______13 ______400 67.7 90.1 9.9 found ammonia as one of the decomposition products 14 ______425 70.2 87.2 J2.8 455 72.8 883 11. 7 from pyrolysis of polyacrylonitrile. We could no t find ammonia in the volatiles. A '·crage. ______88.1 11.9 Results of mass spectromeLer analy is of V 25 from NHH NBR are shown in table 3. Two of Lbe analyses 1. ______were made shorLly after the fractions were collecLed 2 ______310 9. 1 71 0 26.0 3 ______3.,0 29.4 8 1. 5 15.5 and kept at Lhe temperature of liquid N 2, a third 4 ______3G5 G~ ..1 87.8 12.2 one alter an interval of 3 days at room temperature, 5 ______380 80. () 8~ . 7 J5 3 400 95.0 84.8 15.2 and a fourth one aILer an interval of 8 days at the Avg (2 to 5 inc1.) ______85.5 ] I. 5 same LemperaLure. All the anal~T s e s were similar and the average results are shown in Lable 3. The results a T he following \'olatilc rractions were collected: V pyr, \'olatile at tbe Irmpera tnro of pyrolysis, bu t not at 25° C; \.", ,·olatile at 25° C, hUL not at -J90°. In the case of othcr polymers it was usually found necessary to separate t his frac- T A B IJE 3. jlJass-speclrometer analysis of I '25 from 7Jyrol ysis oj tion into two parts to facilitate m ass spectrometer analysis. H ero it was fo und nitrile ,·ubber unnecessary to do so; \'-10., ,·olatile at -1900 c. pyridine. Average molecular weight by tbe micro Formula. of co mponent I A " cl'agc of cryoscopic method in diphenylamine was 330 ± 9 for , 4 analyses polyacrylonitrile and 401 ±3 for NBR. .1 role % Fraction V -190 from pyrolysis of polyacrylonitrile (',H , ______. ______Y. I was analyzed in the mass spectrometer. This frac (', ll . ______4.7 C,11 , . ______3.0 tion consisted of hydrogen. Tbe Y-190 fra cLion 2. [) from NBR was too small for analysis. H . l Fraction V25 from pyroly is of pol.raerylollitrile ~ : ~: :- :~~:~:::~~~:~:::~~J 8. :l C,ll". ______, 3. 3 when first collecLed was a milky-whiLe liquid. The C,II, ______2.5 C,ll, ______9. I color of the liquid changed arLer standing, firsL Lo c,u,, ______4.8 tan and tben to brown. ResulLs of mass spec C SITI 2 ______. ______1.6 trometer analysis of Lhjs fracLion arc shown in table C,J[,. ______.!.:l C.ll, ______6. () 2. V 25 fractions LhaL collecLed in experiments 1, 4, C, H ,. ______5. 4 5, 11, 13, and 14 were analyzed within 1 to 2 hours C 6U I2 ______2.7 after they were collected. The Lime interval C, to C" inc1ush·e ______17.6 between collection and analysis was 1 day for ex TotaL ______100.0 periments 2 and 7, and 1 week for experiments 3 J and 8.
TABLE 2. 11fass-Sp ectrometer analysis of Y25 from pyrolysis of polyacrylonilrile
Time intrrval betwecn end of cxperiment and analysis
Component 1 to 2 homs 1 day 1 week ------.,----- ,---,----1----;----1---,----1 Expt. l' Expt.4 Expt.5 Expt. 11 Expt.13 Expt. 14 Expt.2 Expt. 7 Expt.3 Expt. 8 1------M ole % Mole % }.Iole % JIoie % }.fole % Mole % j\Iole % ;lIole % AIole % j\Iole % H ydrogen cyanide ______~4 22 16 ,l8 66 63 28 Aerylonitrile ______36 45 49 10 7 6 Vinylacetonitrile ______. _ ------30 30 32 13 13 12 88 63 86 59 Pyrrole __ . ______------3 3 12 11 Aeetonitrilc ______3 6 10 Butyronitrile ______2 3 Propionitrile ______4 8 6 9 19 TotaL ______100 100 100 100 100 100 100 100 100 100
u Expcriment numbers are the same as those giwn in table 1.
470~83- 58-- 2 79 resemble those obtained previously for butadiene and 2 .0 UJ.... SBR (formerly GR- S) [2]. It is rather surprising :::> z that none of the products characteristic of polyacryl ::1 1.8 0: onitrile pyrolysis showed up in these analyses, al W "- 1.6 though two of the analyses were made shortly after w --' the fractions were prepared and were kept in liquid "- :;;
040 5o ,-----,-----,-----,------,-----,-----,------w N :::; ~ o--' > 30 w =====~~-----o------o 2 80 0 --' Q. ~
°0L-~==~---4~0----~60~--~8~0 ----~10=0~---1~20~--~,40 50 100 150 200 FROM STAR T OF EXPE RIMENT, MINUTE S TIME FROM START OF EXPERIMENT, MINUTES FIGURE 6. Thermal degradati on of polyacrylonitrile. FIGU RE 8. Thermal degradation of NER. 80 L_ ---- 1 versus time from star t of experiment, and in figure 9, where Tate of volatilization in percent of original sam ple pel' minute is plotted versus percen tage volatili zation. The two sets of curves Tesemble closely those for SBR shown above (fi gs. 3 and 4), and no definite conclusion as to initial rates and activation energy can be based on these curves.
0 .8 ~------Y------r------r----r-----,
H H H H ~ 0 .7 :::> C c· C C z "H/ U"II/ 1I J[ ]I II II IT/ IT" rI/Il" ~ C C +c= c C= C+·C C a: I I I II I II I I w ON CN CN eN Gl. ON (2) "- 0.6 w -' This is followed by an unzipping reaction yielding :;;"- monomers, as shown in eq (2).
AMOUNT VOLATILIZED, PERCENT double bonds in the chain, since these double bonds tend to enhance the thermal sLability of Lhe chain. FIGURE 9. Rates of volatilization of NER. Polyvinyl chloride [12] and po l yviJ1~Tlid eJ1e fluoride [13] may be cited a.s examples of a behavior similar to that of polyacrylon itrile. 5 . References [1] S. L. Madorsky and S. Straus, J . Research NBS !l0, 417 (1948) RP1886. [2] S. L. Madorsky, S_ Straus, D. Thompson, and L. William 4. Discussion soo, J. Research NBS !l2, 499 (1949) RP1989. [3] S. L. Madorsky, J . Polymer Sci. 9, 133 (1952); 11, 491 The main constituents of the light volatile fraction (1953) . [4] S. Straus and S. L. Madorsky, J. Research rBS 50, V 25 obtained in the pyrolysis of polyacrylonitrile, are 165 (1953) RP2405. hydrogen cyanide, acrylonitrile, and vinylacetoni [5] S. Straus and S. L. Madorsky, Ind. Eng. Chem. !lS, 1212 trile. The hydrogen cyanide most likely forms (1956) . by splitting-off the chain in the same manner [6] W. Kern and H. Fernow, J . prakt. Chem. 160, 281 (1942); Rubber Chem. and Technol. 17,356 (1944). as CH3COOH in the pyrolysis of polyvin yl acetate [7] J. McCartney, Poi.vmer degradation mechanisms, N BS [11], H Cl in tbe pyrolysis of polyvinyl chloride [12], Circ. 525, 123 (1950). and HF in the pyrolysis of hydroAuoroethylene [8] R Houtz, T extile Research J . 20, 786 (1950) . polymeTs [13]. [9] Hideo Nagao, Moriya Uchida, and Teruo Yamaguchi, K6gy6 Kagaku Zasshi 59, 698 (1958). II IT n II J[ IT [10] W. J . Burlant and J . L. Parsons, J . Polymer Sci. 22, c C C C IT C C 249 (1956). " n/ ""c/ u" / IT" ~IV-I II I" ?/H"g/II\ .. ··-c C +HCN [11] N. Grassie, Trans. Faraday Soc. !lS, 379 (1952); !l9, 835 (1953) . 6N CNI 6N 6N 6N (1) [12] R . R. Stromberg, S. Straus, and B. G. Achhammer (unpublished paper). Formation of monomer can be visualized as follows. [13] S. L. Madorsky, V. E. Hart, and S. Straus, J . Research A random scission of a C- C bond in the chain re NBS 51, 327 (1953) RP2461. sults in two free radical ends. VVASHINGTON, March 5, 1958. 81