Polymer International 46 (1998) 289È297 Kinetics of Thermal Degradation of Thermotropic Poly(p -oxybenzoate-co ethylene terephthalate) by Single Heating Rate Methods Xin-Gui Li,1* Mei-Rong Huang,1 Gui-He Guan2 & Tong Sun2 1 Department of Polymer Materials Science Engineering, College of Materials Science Engineering, Tongji University, 1239 Siping Road, Shanghai 200092, PeopleÏs Republic of China 2 Department of Polymer Materials Science Engineering, China Textile University, Shanghai 200051, PeopleÏs Republic of China (Received 24 July 1997 ; revised version received 4 November 1997 ; accepted 20 January 1998) Abstract : The kinetics of decomposition of thermotropic liquid crystalline poly(p-oxybenzoate-co-ethylene terephthalate) (poly(B-co-E), BE polymer) with di†erent monomer ratios in both nitrogen and air were studied by dynamic thermogravimetry (TG) from ambient temperature to 800¡C. The kinetic parameters, including the activation energy E@, the reaction order n, and the pre-exponential factor Z, of the degradation of the BE polymers were evaluated by the single heating rate methods of Friedman, FreemanÈCarroll and Chang. The BE polymers which degraded in two distinct stages in nitrogen and air, were stable under nitrogen, while almost completely burned in air. The weight losses in the Ðrst stage in nitrogen and air were dominated by the thermal degradation of both B and E segments, but the weight losses in the second stage were governed by the thermal degradation of B segment in nitrogen and by the oxidative degradation of both B and E segments in air. The maximum rate of weight loss increased linearly with the increase of E content and heating rate, but as E content increased, the char yield at 800¡C in nitrogen decreased linearly. The E@, n and ln Z values of the BE polymers in the Ðrst stage of thermal pyrolysis are higher in nitrogen than in air, indicating that the degradation rate is slower in an inert atmosphere. The E@ value increased with increasing heating rate but varied irregularly with the variations of B/E ratios and molecular weight. The n and ln Z values for the BE polymers in nitrogen were found to be in the wide ranges 1É5È5É6 and 12È48 min~1, respectively, suggesting a complex degradation process. The estimated lifetimes of the BE polymers at 250¡C were calculated to be at least 33 days in nitrogen and 3 h in air. ( 1998 SCI. Polym. Int. 46, 289È297 (1998) Key words : thermotropic polymer ; liquid crystalline polyester ; thermal degradation ; thermogravimetry ; decomposition kinetics ; thermostability have been studied extensively,1h3 but only very few investigators have attempted to study the kinetics of the thermal degradation of BE polymer, which is mostly used as high-tensile Ðbre, self-reinforced plastics and advanced Ðbre-reinforced composite matrix.3h6 Thermogravimetry (TG) is a technique widely used because of its simplicity and the information a†orded INTRODUCTION The thermal properties of thermotropic liquid crystalline poly(p-oxybenzoate-co-ethylene terephthalate) (poly(B-co-E) designated as BE polymer for brevity) * To whom all correspondence should be addressed. 289 ( 1998 SCI. Polymer International 0959È8103/98/$17.50 Printed in Great Britain X. G. L i et al. 290 from a simple TG thermogram. In this paper, TG and di†erential thermogravimetry (DTG) measurements of BE polymer are reported ; the kinetics of the thermal degradation of BE polymer in nitrogen and air from ambient temperature to 800¡C are investigated using Friedman, FreemanÈCarroll and Chang techniques which allow us to calculate every point in a TG curve and a DTG curve, and the variations with temperature are discussed. Friedman method8 ln(da/dt) \ ln Z ] n ln(1 [ a) [ E@/RT (1) where a is the weight loss of the polymer undergoing degradation at time t, R is the gas constant (8É3136 J mol~1 K~1) and T is the absolute temperature (K). The plot of ln da/dt as a function of 1/T should be linear with a slope equal to E@/R. Additionally, the E@/(nR) value could be determined from the slope of a linear plot of ln (1 [ a) versus 1/T . FreemanÈCarroll method9 EXPERIMENTAL BE polymer with the structural formula shown in Scheme 1 was synthesized following the procedure described elsewhere.7 The intrinsic viscosity of the BE polymers was measured at 0É5% concentration in phenol/sym-tetrachloroethane 50/50 (vol) at 30¡C. The TG and DTG thermograms were obtained using a Du Pont Instrument 9900 computer/thermal analyzer under a dynamic nitrogen atmosphere Ñowing at 50 ml min~1 and static air. Degradation experiments under nitrogen were conducted at the heating rates of 1, 2, 5, and 10¡C min~1 using 21È32 mg samples, and in every case a thermally stable residue was formed. The char residues were obtained from the TG curve by dividing the weight of the sample at 500È800¡C by the initial weight of the sample. The degradation in air was performed at the heating rates of 5 and 20¡C min~1 using a sample of 1È8 mg. The TG and DTG curves were then printed. *ln(da/dt)/*ln(1 [ a) \ n [ (E@/R)*(1/T )/*ln(1 [ a) The *ln(da/dt) and *ln(1 [ a) values are taken at regular intervals of 1/T , in this case 1/T \ 0É000 01. By plotting *ln(da/dt)/*ln(1 [ a) against *(1/T )/*ln(1 [ a), a straight line was obtained with the slope and intercept equal to [(E@/R) and n, respectively. Chang method.10 Equation (1) can be rewritten in the following form : ln[(da/dt)/(1 [ a)n] \ ln Z [ E@/RT (3) A plot of ln[(da/dt)/(1 [ a)n] against 1/T will yield a straight line if the decomposition order n is correctly selected. The slope and intercept of this line will provide the E@ and ln Z values, respectively. Determination of half-life time t .10 The following 1@2 equation is integrated : Determination of kinetic parameters of thermal degradation da/dt \ Z(1 [ a)n exp([E@/RT ) There are several methods (proposed by Friedman,8 Freeman and Carroll,9 Chang,10 Flynn and Wall,11 Chaterjee and Conrad,12 Horowitz and Metzger,13 Kissinger,14 Coats and Redfern,15 Van Krevelen,16 Reich,17 and Ozawa18) for calculating kinetic parameters which depend not only on the experimental conditions but also on the mathematical treatment of the data. We have used the Friedman, FreemanÈCarroll and Chang methods to evaluate the activation energy E@, reaction order n, and frequency factor Z based on a single heating rate measurement without making any assumptions. Detailed descriptions of the three methods are not given because the techniques for evaluating the kinetic parameters from TG/DTG traces are easily available from the literature.8h10 The equations employed in the methods are listed below. (2) (4) The half-life time t may be calculated from eqns (4) 1@2 and (5) when n is di†erent from unity : t 1@2 \ M(1 [ 0É51~n)/[Z(1 [ n)]N exp(E@/RT ) \ (1 [ 0É51~n)/[k(1 [ n)] (5) In this expression k is the rate constant of thermal decomposition (min~1). When n equals unity : t \ (0É693/Z)exp(E@/RT ) \ 0É693/k 1@2 (6) Determination of lifetime t .10 The lifetime t of polymer f f to failure is generally deÐned to be when the weight loss reaches 5%. Therefore : t \ M(1 [ 0É951~n)/[Z(1 [ n)]Nexp(E@/RT ) f \ (1 [ 0É951~n)/[k(1 [ n)] (n D 1) t \ (0É0513/Z)exp(E@/RT ) \ 0É0513/k f (7) (n \ 1) (8) Scheme 1 POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 T hermal degradation kinetics of thermotropic poly(B-co-E) 291 RESULTS AND DISCUSSION Effect of B /E ratio on thermal decomposition in nitrogen The typical dynamic TG and DTG thermograms of B/E (60/40) and (75/25) polymers in a dynamic nitrogen atmosphere are shown in Figs 1 and 2. Figures 3 and 4 Fig. 3. Friedman plots of Ln (da/dt) or Ln (1 [ a) vs. reciprocal temperature for the direct calculation of E@ or n value of thermal degradation of B/E (60/40) polymer in nitrogen at heating rates of (L) 1¡C/min ; ( x ) 2¡C/min ; (…) 5¡C/min ; (|) 10¡C/min. present their plots of ln(da/dt) versus 1/T and ln (1 [ a) versus 1/T . The decomposition temperature T , the d temperature at the maximum rate of decomposition T , E@, n, ln Z, and char yield are presented in Table 1 dm and Fig. 5. It is evident that :19,20 Fig. 1. Dynamic TG thermograms of thermotropic liquidcrystalline polymer with B/E monomer ratios of (a) 60/40, and (b) 75/25 in nitrogen at the heating rates of (ÈÈ) 1¡C/min ; (- - - -) 2¡C/min ; (È I È I) 5¡C/min ; (I I I) 10¡C/min. Fig. 2. Dynamic DTG thermograms of thermotropic liquidcrystalline polymer with B/E monomer ratios of (a) 60/40, and (b) 75/25 in nitrogen at the heating rates of (ÈÈ) 1¡C/min ; (- - - -) 2¡C/min ; (È I È I) 5¡C/min ; (I I I) 10¡C/min. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 (a) There is deÐnitely an increase in T and T as the B d dm unit content increases in the BE polymer. The B unit content in the BE polymer plays an important role in enhancing the thermostability of the polymer. Fig. 4. Friedman plots of Ln (da/dt) or Ln (1 [ a) vs. reciprocal temperature for the direct calculatin of E@ or n value of thermal degradation of B/E (75/25) polymer in nitrogen at heating rates of (L) 1¡C/min ; ( x ) 2¡C/min ; (…) 5¡C/min. X. G. L i et al. 292 TABLE 1. Kinetic parameters of thermal degradation of thermotropic BE polymer under nitrogen calculated by the Friedman method at 10ÄC min—1 B/E (mol/mol) 100/0a 100/0b 80/20a 80/20b 75/25c 60/40 60/40a 30/70a 0/100a 0/100b ÍiË (dl gÉ1) HMW LMW HMW LMW 0·87 0·60 0·57 0·72 ¿0·62 0·62 T /T /T d dm1 dm2 (¡C) 512/530/– 499/515/– 438/490/– 414/460/492 439/477/– 425/453/– 423/460/– 412/450/– 406/438/– 398/423/– E¾ (kJ molÉ1) 253 282 185 201 257 211 204 102 83 172 n 3·0 2·6 4·7 3·7 5·1 3·1 2·8 1·1 0·5 0·9 ln Z (minÉ1) 37 40 29 30 42 34 33 16 13 26 Char yield (%) (¡C) 36 40 28 30 28 25 22 16 6 12 800 600 800 600 600 600 800 800 800 600 a TG curves were obtained from ref. 19. b TG and DTG curves were obtained from ref. 20. c Kinetic parameters and char yield were obtained from the TG and DTG curves at 5¡C minÉ1. (b) There is a linear increases in the char yields as the B unit content increases because an increase of the B content will decrease the number of hydrogen atoms and retard the formation of more volatile degraded products, thus resulting in an increase in char yield. The formation of the char residues is probably due to branch formation and crosslinking of the product obtained mostly from B units during the thermal degradation under nitrogen. (c) There is a linear increase of the maximum rate of decomposition as the E content increases from 20 to 100 mol%. This phenomenon can be explained as being due to the heat instability of E units as compared with B units in the BE polymer. The degradation of a polymer containing more hydrogen atoms might lead to the formation of more volatile pro- Fig. 5. E†ect of E unit content in B/E polymer on the maximum weight-loss rate and char yield in nitrogen at 10¡C/min heating rate. ducts, thus resulting in a decreasing amount of carbonaceous mass and faster degradation. (d) The liquid crystalline BE polymer containing 60È 80 mol% B unit exhibits E@ and ln Z values of 185È 257 KJ mol~1 (mean value 212 KJ mol~1) and 29È42 min~1 (mean value 34 min~1), respectively, which are higher than those of pure E homopolymer but lower than those of pure B homopolymer. (e) The n values of 2É8È5É1 (mean value 3É9) for the BE polymer exhibiting thermotropic liquid crystallinity are higher than those of pure B homopolymer and much higher than those of pure E homopolymer,21 polyvinyl chloride,8 polystyrene,22 polycarbonate23 and cellulose,24 which generally show a Ðrst-order decomposition kinetics. Generally, a decomposition order of zero characterizes the most rapid decomposition reaction. As the order increases, the decomposition reaction becomes slower. Therefore, the kinetic order of 5 for the B/E (75/25) polymer indicates a much slower degradation reaction, as shown in Fig. 2. It is seen from Table 2 that the highest order of 5É6 was observed for the B/E (75/25) polymer at the lowest heating rates of 1 and 2¡C min~1, indicative of the slowest degradation kinetics. Additionally, the kinetic order of the BE polymers seems to be dependent on the sample weight because the B/E (75/25) polymer exhibiting the highest order of thermal degradation has the greatest sample size, as shown in Table 2. A detailed study of the dependence of degradation kinetics on the sample weight needs to be done in the future. (f ) The decomposition behavour of the BE polymer under nitrogen is quite di†erent from that of the respective B or E homopolymers. The TG results obtained and discussed so far could be taken as POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 T hermal degradation kinetics of thermotropic poly(B-co-E) 293 TABLE 2. Effect of heating rate on kinetic parameters of thermal degradation of BE polymer under nitrogen calculated by the Friedman method B/E (mol/mol) ÍiË (dl gÉ1) Sample size (mg) Heating rate (¡C minÉ1) T /T /T d dm1 dm2 (¡C) E¾ (kJ molÉ1) n ln Z (minÉ1) 60/40 60/40 60/40 60/40 75/25 0·60 0·60 0·60 0·60 0·87 26·7 27·7 22·2 21·2 31·0 1 2 5 10 1 387/408/462 398/414/475 403/437/– 425/453/– 390/414/468 75/25 0·87 32·5 2 401/435/470 75/25 0·87 29·5 5 417/446/482 196 201 203 211 236 (171)a 248 (108)a 257 2·9 3·5 3·0 3·1 5·6 (2·6)a 5·6 (1·7)a 5·1 32 33 33 34 39 (24)a 41 (14)a 42 a The data in parentheses are the kinetic parameters of the second stage of thermal degradation. proof of the presence of a random sequence distribution in the polymer backbone25 because no distinct peaks representative of thermal decomposition of individual B and E homopolymers are observed during the thermal degradation of the BE polymer. In the case of random copolymer, generally stepwise degradation of individual B and E homopolymer segments merge into single steps located in between the maximum degradation temperatures of the corresponding homopolymers.26 Effect of molecular weight on thermal degradation Tables 1È4 list the kinetic constants of the thermal degradation of BE polymer with di†erent intrinsic viscosities or solid-state polymerization times, where a longer solid-state polymerization time means higher molecular weight. No clear relationship was found between the kinetic parameters and molecular weight/ sequence distribution for the same B/E ratios of 80/20 and 60/40. The thermal decomposition temperature T , d T and E@ depended slightly on the molecular weight of dm the BE polymer. T and T increased with higher d dm TABLE 3. Effect of intrinsic viscosity of B/E (60/40) polymer on the activation energy of decomposition in air at 20ÄC min—1a ÍiË (dl gÉ1) 0·67 0·71 0·74 0·79 0·82 0·88 Average T /T /T d dm1 dm2 (¡C) 424/463/587 429/460/601 431/468/601 431/468/600 433/470/604 432/471/602 430/467/599 molecular weight, but the E@ value appeared to decrease slightly in some cases. The dependence of T and T on d dm the molecular weight indicates that these polymers may involve end-group initiated thermal decomposition. Kinetic parameters of thermal degradation in air The TG and DTG curves of the BE polymers in air indicated two degradation stages, corresponding to thermal degradation and thermo-oxidative degradation respectively. The T , T , E@, n and ln Z values reported d dm in Tables 3 and 5 are lower than those obtained in nitrogen except that the T and T values listed in d dm Table 3 are slightly larger because of the higher heating rate. These indicate faster kinetics in air than in nitrogen. Similar behaviour was observed for the thermal degradation of wholly aromatic poly(ether ether ketone ketone),28 chlorinated atactic polypropylene, and polyvinyl chloride.29 This may be due to the faster rate of TABLE 4. Effect of solid-state polymerization and sequence distribution of B/E (80/20) polymer on its degradation under nitrogen at 10ÄC min—1 (ref. 20) n ln Z (minÉ1) BE polymer consisting of shorter B segment and shorter E segment 0 387/428/– 134 2 412/459/499 141 4 396/446/480 126 8 413/448/474 150 3·5 4·2 2·8 2·7 19 20 15 22 BE polymer consisting of longer B segment and longer E segment 0 384/420/– 142 2 410/452/487 111 4 400/448/486 101 8 418/455/486 146 2·4 2·5 2·6 3·5 21 14 12 21 Solid-state polymerization time (h) T /T /T d dm1 dm2 (¡C) E¾ (kJ molÉ1) Activation energy (kJ molÉ1) Thermal pyrolysis Thermo-oxidative pyrolysis 177 143 168 155 131 144 153 202 164 143 145 134 184 162 a The sample size is about 1 mg. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 X. G. L i et al. 294 TABLE 5. Kinetic parameters of thermal degradation of BE polymers under air calculated by the Friedman method B/E (mol/mol) 100/0 100/0a 70/30b 65/35b 60/40b 54/46b 0/100c 0/100 PMBa ÍiË (dl gÉ1) – – 0·79 0·74 0·72 0·81 – 0·67 DP ¼ 200 Heating rate (¡C minÉ1) T /T /T d dm1 dm2 (¡C) 20 10 5 5 5 5 5 20 10 530/570/597 475/–/525 362/404/495 350/378/487 345/390/490 347/405/510 385/428/520 390/453/538 412/430/575 Thermal pyrolysis Oxidative pyrolysis E¾ (kJ molÉ1) n ln Z (minÉ1) E¾ (kJ molÉ1) n ln Z (minÉ1) 440 210 114 131 135 119 201 157 260 – 2·2 2·1 1·8 2·6 2·9 2·0 – 3·5 – 31 18 22 22 18 30 – 43 146 469 180 180 184 168 142 213 257 – 0·7 0·9 0·6 0·6 0·6 1·3 – 1·1 – 69 28 27 28 25 20 – 35 a TG curves were obtained from ref. 27 ; PMB is poly(p -mercaptobenzoate). b Sample size for the TG measurements is about 8 mg. c From ref. 21. degradation of the Ñexible aliphatic moiety having a higher thermooxidative instability in air. In addition, the E@ and ln Z values of the thermooxidative degradation for BE polymer are larger than those of the thermal degradation, but the n value of the oxidative degradation is much smaller than that of the thermal pyrolysis as listed in Table 5. One can note that poly(p-mercaptobenzoate) shows di†erent kinetics of degradation in air as compared with B homopolymer because of the replacement of an oxygen atom in the main chain by a sulphur atom. Effect of heating rate on kinetic parameters of thermal degradation The dependence of the kinetic parameters of thermal degradation on heating rate for BE polymer calculated by the Friedman, FreemanÈCarroll and Chang methods is shown in Figs 6È9 and Tables 2 and 6. The T , T , d dm E@ and ln Z values obtained by Friedman method all Fig. 6. E†ect of heating rate on the maximum decomposition rate of the polymers with the B/E ratios of (…) 60/40 and (L) 75/25 in nitrogen. Fig. 7. FreemanÈCarroll plots of [[*Ln(da/dt)]/ [*Ln(1 [ a)] vs. [*(1/T )/*Ln(1 [ a)] for the B/E polymer in nitrogen at heating rates of (L) 1¡C/min ; ( x ) 2¡C/min ; (…) 5¡C/min ; (|) 10¡C/min. Fig. 8. Chang plots of Ln[(da/dt)/(1 [ a)n] vs. reciprocal temperature for B/E (60/40) polymer in nitrogen at heating rates of (L) 1¡C/min at n \ 3 ; (…) 2¡C/min at n \ 3.3 ; (|) 5¡C/min at n \ 2 ; (>) 10¡C/min at n \ 2. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 T hermal degradation kinetics of thermotropic poly(B-co-E) Fig. 9. Chang plots of Ln[(da/dt)/(1 [ a)n] vs. reciprocal temperature for B/E (75/25) polymer in nitrogen at heating rates of (L) 1¡C/min at n \ 5.4 ; ( x ) 2¡C/min at n \ 5.4 ; (…) 5¡C/min at n \ 5. 295 BE polymer containing more than 70 mol% B unit,31 so it might be mainly ascribed to the degradation of the B unit. Additionally, there are some di†erences in the kinetic data calculated using the di†erent methods, as shown in Tables 2 and 6. The Friedman method gave the lowest E@ value but the highest n value of the three methods. As shown in Fig. 8, only the Chang method actually tends to form straight lines in the widest temperature range, which means a smaller error in the calculation of the kinetic parameters by this method. However, the temperature range used for the determination of the kinetic parameters by the Friedman and FreemanÈ Carroll methods is wide enough to obtain reliable results. Thermal lifetime prediction increase as the heating rate increases ; the n value behaves di†erently. At lower heating rates the di†usion of the degradation products apparently does not a†ect the kinetics of the decomposition process, so these values were found to be lower. On the contrary, the degradation of the BE polymers in nitrogen is probably faster than the di†usion of the degradation products through the polymer melt, so that at higher heating rate the kinetics of the degradation process are under di†usion control.28 This dependency of the kinetics on the heating rate is in agreement with those of E homopolymer,21 highly thermally stable aromatic poly(ether ether ketone), poly(ether ether ketone ketone),28 and poly[3-dimethyl(methacryloyloxyethyl)ammonium propanesulphonate],30 but is contrary to that of polyvinyl chloride.8 It should be noted that the kinetic data of the second stage of thermal degradation in nitrogen for B/E (75/25) polymer listed in Table 2 are much smaller than those of the Ðrst stage. The second stage of thermal degradation in nitrogen appears in the thermograms of the A major application of TG and DTG kinetic parameters is to predict the maximum useable temperature, the optimum processing temperature regions, and the estimated lifetime of the polymers. It o†ers a simple and convenient approach to use as an accelerated ageing process for quality control experiments. The decomposition kinetics at elevated temperature could be extrapolated back to the service conditions for which the lifetime prediction is required. The rate-constant k, halflife time t and estimated lifetime t for the BE poly1@2 f mers generated from the weight losses a of 0É5 and 0É05 respectively, at di†erent temperatures in nitrogen and air have been calculated and are listed in Tables 7 and 8. It can be predicted theoretically that the lifetimes of the polymers with B/E ratios of 60/40, 75/25 and 80/20 heated at 250¡C in nitrogen could reach 33 days, 293 days and 69 days, respectively. When the polymers are heated at 250¡C in air, however, the lifetimes for the B/E 54/46, 60/40, 65/35 and 70/30 polymers are 11 h, TABLE 6. Kinetic parameters of thermal degradation of the polymers with B/E ratios of 60/40 ( [ g ] = 0·6 dl g—1) and 75/25 ( [ g ] = 0·87 dl g—1) under nitrogen calculated by the Freeman–Carroll and Chang methods B/E (mol/mol) Heating rate (¡C minÉ1) Freeman–Carroll Chang E¾ (kJ molÉ1) n ln Z (minÉ1) E¾ (kJ molÉ1) n ln Z (minÉ1) 276 276 240 279 264 (233)a 269 (154)a 299 3·0 3·3 2·0 2·0 5·4 (2·0)a 5·4 (1·5)a 5·0 45 45 39 47 43 (34)a 43 (22)a 48 60/40 60/40 60/40 60/40 75/25 1 2 5 10 1 288 309 308 228 215 3·0 3·2 2·0 1·5 4·0 48 51 51 36 34 75/25 2 265 4·6 43 75/25 5 275 5·3 40 a Data in parentheses are the kinetic parameters of the second stage of thermal degradation. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 X. G. L i et al. 296 TABLE 7. Rate constant k of thermal degradation, half-life time t and estimated lifetime t 1¿2 f for BE polymers at different temperatures in nitrogen Temp. (¡C) B/E (60/40) 200 250 300 350 400 B/E (75/25) B/E (80/20) lg k (minÉ1) lg t 1@2 (min) lg t f (min) lg k (minÉ1) lg t 1@2 (min) lg t f (min) lg k (minÉ1) lg t 1@2 (min) lg t f (min) É8·56 É5·94 É4·51 É2·69 É1·64 8·75 6·33 4·69 2·88 1·84 7·29 4·67 3·23 1·42 0·38 É10·4 É6·87 É5·27 É2·91 É1·66 11·0 7·55 4·91 3·57 2·31 9·17 5·63 4·03 1·67 0·41 É9·26 É6·25 É5·38 É3·16 É1·95 9·57 6·66 5·69 3·56 2·36 8·00 5·00 4·12 1·90 0·70 and estimated lifetime t for BE polyTABLE 8. Rate-constant k of thermal degradation, half-life time t 1¿2 f mers at different temperatures in air Temp. (¡C) 200 250 300 350 400 B/E (54/46) B/E (60/40) B/E (65/35) B/E (70/30) lg k (minÉ1) lg t 1@2 (min) lg t f (min) lg k (minÉ1) lg t 1@2 (min) lg t f (min) lg k (minÉ1) lg t 1@2 (min) lg t f (min) lg k (minÉ1) lg t 1@2 (min) lg t f (min) É5·26 É4·07 É2·95 É2·16 É1·36 5·42 4·23 3·12 2·32 1·52 3·99 2·80 1·68 0·89 0·08 É5·15 É3·93 É2·55 É1·77 É0·73 5·25 4·21 2·66 1·87 0·83 3·88 2·66 1·28 0·49 É0·52 É4·95 É3·53 É2·44 É1·43 É0·67 4·92 3·50 2·41 1·40 0·63 3·67 2·25 1·16 0·15 É0·70 É4·83 É3·57 É2·62 É1·74 É1·08 4·84 3·59 2·64 1·76 1·11 3·55 2·29 1·34 0·46 É0·22 8 h, 3 h and 4 h, respectively. In reality, besides thermal degradation, all sorts of other processes including photodegradation, mechanical and chemical degradations, will shorten service lifetimes. Therefore, the real lifetime of the polymers should be shorter than those listed in Tables 7 and 8. The static processing time for the BE polymers at 400¡C will last only for 2É4È5É0 min in nitrogen and 0É2È1É2 min in air. The thermostability of the BE polymer is certainly much higher in nitrogen than in air. These results suggest that the possible meltprocessing temperature range for the BE polymer in an inert atmosphere is about 200¡C (between 200 and 400¡C) in order to avoid any serious degradation and maintain the molecular weight during processing. However, the optimum processing temperature would range from 250 to 350¡C because of the rapid change in melt rheological properties at lower or higher temperature.32 Carroll and Chang methods. The kinetic parameters exhibit a dependence on B/E ratio, molecular weight, heating rate, testing atmosphere and method of calculation. The kinetic data in nitrogen suggest that BE polymer exhibits good thermostability. In air the thermostability of BE polymer is greatly reduced because the thermal degradation is accompanied by thermooxidative degradation. The degradation seems to be a random scission process of the ester linkages. The optimum melt-processing temperature for BE polymer would range from 250 to 350¡C. ACKNOWLEDGEMENTS This project is supported by the National Natural Science Foundation of China and by the Science Technology Development Foundation of College of Materials Science Engineering of Tongji University of China. CONCLUSIONS REFERENCES On the basis of thermogravimetry and derivative thermogravimetry results obtained at a single heating rate, some important kinetic parameters of thermal degradation for the thermotropic liquid crystalline poly(poxybenzoate-co-ethylene terephthalate), such as activation energy, degradation order, and frequency factor, have been easily calculated by the Friedman, FreemanÈ 1 Li, X.-G., Huang, M.-R., Guan, G.-H. & Sun, T., J. Appl. Polym. Sci., 59 (1996) 1. 2 Mathew, J., Ghadage, R. S., Ponrathnam, S. & Prasad, S. D., Macromolecules, 27 (1994) 4021. 3 Brostow, W., Hess, M. & Lopez, B. L., Macromolecules, 27 (1994) 2262. 4 Sugiyama, H., Lewis, D. N., White, J. L. & Fellers, J. F., J. Appl. Polym. Sci., 30 (1985) 2329. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 T hermal degradation kinetics of thermotropic poly(B-co-E) 5 Muramatsu, H. & Krigbaum, W. R., J. Polym. Sci. Part B, Polym. Phys., 25 (1987) 2303. 6 Itoyama, K., J. Polym. Sci. Part B, Polym. Phys., 26 (1988) 1845. 7 Li, X.-G., Huang, M.-R., Guan, G. H. & Sun, T., J. Appl. Polym. Sci., 66 (1997), 2129. 8 Friedman, H. L., J. Polym. Sci. C., 6 (1964) 183. 9 Freeman, E. S. & Carroll, B., J. Phys. Chem., 62 (1958) 394. 10 Chang, W. L., J. Appl. Polym. Sci., 53 (1994) 1759. 11 Flynn, J. H. & Wall, L. A., J. Polym. Sci. B, 4 (1966) 323. 12 Chaterjee, P. K. & Conrad, C. M., J. Polym. Sci. A-1, 6 (1968) 594. 13 Horowitz, H. H. & Metzger, G., Anal. Chem., 35 (1963) 1464. 14 Kissinger, H. E., Anal. Chem., 29 (1957) 1702. 15 Coats, A. W., Redfern, J. P., Nature, 201 (1964) 68. 16 Van Krevelen, D. W., Van Heerden, C. & Huntjens, F. J., Fuel, 30 (1951) 11. 17 Reich, L., J. Polym. Sci., Polym. L ett. Ed., 2 (1964) 621. 18 Ozawa, T., Bull. Chem. Soc. Jpn., 38 (1965) 1881. 19 Garozzo, D., Giu†rida, M., Montaudo, G. & Lenz, R. W., J. Polym. Sci., Part A, Polym. Chem., 25 (1987) 271. 20 Shinn, T. H., Chen, J. Y. & Lin, C. C., J. Appl. Polym. Sci., 47 (1993) 1233. POLYMER INTERNATIONAL VOL. 46, NO. 4, 1998 297 21 Cooney, J. D., Day, M. & Wiles, D. M., J. Appl. Polym. Sci., 28 (1983) 2887. 22 Carrasco, F. & Pages, P., J. Appl. Polym. Sci., 61 (1996) 187. 23 Tagle, L. H. & Diaz, F. R., J. Macromol. Sci., Chem., A28 (1991) 397. 24 Calahorra, M. E., Cortazar, M., Eguiazabal, J. I. & Guzman, G. M., J. Appl. Polym. Sci., 37 (1989) 3305. 25 Li, X.-G., Huang, M.-R., Guan, G.-H. & Sun, T., Angew. Makromol. Chem., 227 (1995) 69. 26 Eroglu, M. S., Hazer, B., Guven, O. & Baysal, B. M., J. Appl. Polym. Sci., 60 (1996) 2141. 27 Kricheldorf, H. R. & Conradi, A., Macromolecules, 22 (1989) 14. 28 Abate, L., Calanna, S., Pollicino, A. & Recca, A., Polym. Eng. Sci., 36 (1996) 1782. 29 Mukherjee, A. K. & Patri, M., J. Macromol. Sci., Chem., A26 (1989) 213. 30 Lee, W. F., J. Appl. Polym. Sci., 37 (1989) 3263. 31 Li, X.-G., Huang, M.-R., Guan, G.-H. & Sun, T., Chinese J. Polym. Sci., 11 (1993) 230. 32 Li, X.-G., Zhou, Z.-L., Wu, X.-G. & Sun, T., J. Appl. Polym. Sci., 51 (1994) 1913.