J. MATER. CHEM., 1991, 1(4), 655-661 Aromatic Ether-Ketone-'X' Polymers Part 2.-EK-lmide Copolymers Christopher J. Borrill and Richard H. Whiteley* Raychem Ltd., R & Q Faraday Rd., Dorcan, Swindon, Wiltshire SN3 5HH, UK A modified Friedel-Crafts synthesis has been used to make two series of aromatic ether-ketone-imide copolymers. All the copolymers were semicrystalline and, in some cases, useful combinations of glass-transition temperature and melt temperatures were found. Some properties of the copolymers are reported and discussed. Keywords: Poly(ether-ketone); Friedel-Crafts; Aromatic polymer; Polyimide In the first paper of this series the synthesis and properties of a variety of high-molecular-weight aromatic ether-ketone-')(' (EKX) polymers were reported.' The polymers were made using a modified Friedel-Crafts reaction.2 One of the X groups that could be incorporated was the imide group, and because of the potential utility of EK-imide polymers as high Tg thermoplastic materials with very good thermo- oxidative stability, we have carried out further studies of this class of materials.Experimenta1 Two series of copolymers were made. One, series A, was made by polymerizing the imide monomer EIKIE (I) with mixtures of isophthaloyl chloride (IPC) and terephthaloyl chloride (TPC) to give polymers with the general structure 11. -1" The other series of copolymers, series B, was made by poly- merizing terephthaloyl chloride with mixtures of the ether- ketone monomer EKE I11 and the imide monomer I to give copolymers which contained the structural units IV and V.t IV J The ether-ketone polymer, PEKEKK, obtained by poly- merizing 111 with TPC is now being made commercially, and is marketed by BASF under the registered trademark 'ULTRAPEK'.2-4 Reagents and Solvents Terephthaloyl chloride (TPC), isophthaloyl chloride (IPC), aluminium chloride, 4,4'-diphenoxybenzophenone, and 4-phenoxybenzophenone were all supplied from the research laboratories of Raychem Corporation, Menlo Park, Cali- fornia. The imide monomer I was made as described pre- viously.' N,N-dimethylfomamide (DMF) and benzoyl chloride were obtained from Aldrich. Dichloromethane (DCM), 1,2-dichloroethane (DCE), sulphuric acid and meth- anol were obtained from Fisons.Synthesis and Characterization of Polymers The polymers were all made using the general procedure as described previously. Series A copolymers were made using a stoichiometric excess of diacid chloride(s) with a correspond- ing amount (i.e. twice the molar excess) of 4-phenoxybenzo- phenone as an end-capper, and with DCE as the solvent. With the exception of polymer 1, series B copolymers were made using a stoichiometric excess of dinucleophilic mono- mer(s) with a corresponding amount of benzoyl chloride as an end-capper7 and with dichloromethane as the solvent. DMF was used as the Lewis base in all the polymerizations except in the case of polymer 8 when dimethyl sulphone was used. The amounts of reagents used are listed in Table 1.Polymerization yields were between 93 and 100%. Polymer structures were confirmed using high-resolution NMR spec- troscopy, and detailed assignments for polymers 7 and 8 have been reported.' Because the general method of synthesis was similar for all the polymers, full details are given below for only two represen tat ive examples. Polymer 4 DCE (90cm3) was stirred and cooled under nitrogen at -30 "C in a 250 cm3 resin kettle. Aluminium chloride (24.6 g, 0.184 mol) was added, followed by DMF (4.71 cm3, 0.0612 mol), keeping the temperature below -10 "C. EIKIE (9.850 g, 0.01 50 mol) and 4-phenoxybenzophenone (0.165 g, 0.0006mol) were then added, followed by a mixture of IPC (1.553 g, 0.00765 mol) and TPC (1.553 g, 0.00765 mol), all added at a temperature of ca.-30 "C. 30 cm3 of DCE were used to ensure complete addition of the reagents. The mixture was stirred and allowed to warm to -5 "C over 2 h. A tough rubbery gel separated from a clear, dark brown-yellow liquid. The mixture was left to stand for 15 h at 8 "C,then the liquid was decanted (ca.80 cm3) and the gel was worked up in two portions, each with 400 cm3 of methanol in a Waring blender. The combined decomplexed polymer portions were then blended again with a further 400cm3 of methanol. The resulting lemon-yellow fibrous polymer was dried at 140 "C for 2 h under vacuum. The yield was 12.02 g (100%). J. MATER. CHEM., 1991, VOL.l Table 1 Amounts of reagents used in polymerisation reactions" polymer solven t/cm EKE/mmol EIKIE/mmol TPC/mmol IPC/mmol AlCl,/mmol ~~ 1 340 - 137.3 133.3 - 1627 2 110 - 30 26.52 4.68 375 3 120 - 15 10.71 4.59 184 4 120 - 15 7.65 7.65 184 5 120 - 15 4.59 10.71 184 6 800 - 100 15.30 86.70 1228 7 120 - 15 - 15.30 184 8 330 309.1 - 303.0 - 1658 9 375 18'4.0 20 200.0 - 1641 10 400 164.0 40 200.0 - 1729 11 475 120.0 50 166.6 - 1514 12 500 103.3 66.7 166.6 - 1588 13 510 95.0 75 166.6 - 1624 14 620 86.7 83.3 166.6 - 1661 15 520 56.9 100 153.8 - 1635 16 520 31.4 1 14.3 142.9 - 1612 a Dimethyl sulphone was used as the Lewis base for the preparation of polymer 8.The mole ratio of dimethyl sulphone to TPC was 1.5: 1.DMF was used for the preparation of all the other polymers with a mole ratio of DMF to diacid chloride(s) of 4: 1. The stoichiometric imbalance was ca. 2%, except for polymers 1 and 2 when it was 3 and 4%, respectively. Polymer 11 DCM (300 cm3) was cooled under nitrogen to -20 "C in a 1 dm3, jacketed reaction vessel. The vessel was cooled using circulating ethylene glycol-water from a thermostatically con- trolled reservoir. Aluminium chloride (202 g, 1.5 14 mol) was added to the stirred solvent, followed by DMF (51.3 cm3, 0.666 mol) at such a rate that the reaction mixture was maintained at a temperature below -10 "C. TPC (33.824 g, 0.1666 mol) and benzoyl chloride (0.956 g, 0.0068 mol) was then added using 80 cm3 of DCM to ensure complete addition.The reaction mixture was cooled to below -15 "C and a mixture of 4,4'-diphenoxybenzophenone (43.97 g, 0.120 mol) and EIKIE (32.833 g, 0.050 mol) was added, again using DCM (95 cm3) to ensure complete addition. The nitrogen gas inlet was replaced with an outlet to a bubbler to monitor the evolution of hydrogen chloride, and the reaction mixture was warmed to room temperature over a period of ca. 1 h. Stirring was stopped when the solution became viscous, and the mixture was then left overnight at room temperature. The polymer gel was removed from the vessel as a 'lollipop' on the stirrer rod [the internal diameter of the vessel (100 mm) was the same as the flange opening], and was then cut up into four pieces.These were each worked up separately in 1 dm3 of cold (-18 "C) methanol in a 4 dm3 Waring blender. The batches of decomplexed polymer were then filtered and blended again, in one batch, with the minimum amount of cold methanol required to cover the polymer in the blender. The polymer was again filtered, then left to soak in methanol overnight before being boiled under reflux for 3 h. Finally, the polymer was filtered, washed with methanol and dried under vacuum at 200 "C overnight. The yield of fibrous yellow polymer was 97 g (98%). Equipment and Procedures Inherent viscosities were measured at 25 "C on 1.0 g dm-3 solutions of polymer fluff in 98% analytical-grade sulphuric acid. All other measurements were made on samples cut from compression-moulded plaques.DSC measurements were made on a DuPont 1090 instru- ment using a 910 cell. Indium and zinc were used for tempera- ture calibration. Scans were made in nitrogen at 10 "C min-'. 'H and 13C NMR spectra were recorded using a Bruker AM300 Fourier transform instrument. Solutions were obtained by soaking ca. 100 mg of polymer in 2 cm3 of CDC13 for 30 min, followed by the addition of trifluoroacetic acid dropwise, with stirring, until the polymer dissolved. TG measurements were made using a Perkin-Elmer TGS- 2 instrument with a System 4 controller. A scan rate of 10 "C min-' was used. Dynamic mechanical thermal analysis (DMTA) measure- ments were made using a Polymer Laboratories Mark 1 machine in single cantilever and auto-strain modes.A scan rate of 4 "C min-' was used, at an oscillation frequency of 1 Hz. Unless stated otherwise, the glass-transition temperature of a given polymer corresponds to the maximum in the DMTA plot of loss modulus uersus temperature. Wide-angle X-ray diffraction spectra were recorded using a Philips PW1050 goniometer using Cu-Ka radiation with a nickel filter. Crystallinity was calculated using the formula where A, is the area under the I uersus 0 diffraction scan due to amorphous scattering, and A, is the area due to crystalline scattering, i.e. the total area minus A,. [N.B. This method yields an approximate value. For more accurate values s21 uersus s scans should be used, where s =2 sin0/A.'] Tensile testing was done according to BS2728 at a strain rate of 1 mm min-'.BS type 2 dumbell test samples were cut from 2mm thick plaques using a router. Values quoted are the averages of four tests. Volume resistivities were measured according to BS2782 method 202B. Permittivities and loss factors were evaluated according to BS2782 method 206B at 15.92 Hz. Smoke measurements were made using a small-scale dynamic smoke-measuring instrument developed at Raychem, Swindon.6 Samples of 0.2 g were used. Smoke values were evaluated by integration of absorbance uersus time plots, and values quoted are the averages of three measurements. Average total smoke values (s)were obtained from integration of smoke uersus temperature plots over the temperature range 473-1 173 K, using the formula (1173 K) (473 K) Limiting oxygen index measurements were made according to ASTM D2863-74.The solutions used for testing hydrolytic stability were 0.01 mol dm- (as.) HC1 (pH 2), 0.025 mol dm-3 (aq.) Na2HP04-0.025 mol dm-3 (aq.) KH2P04 (pH 7) and J. MATER. CHEM., 1991, VOL.l 0.01 mol dm-3 (aq.) NaOH (pH 12). The pH of each of the solutions was checked at regular intervals during the testing and adjusted when necessary. Results and Discussion Series A Copolymers Table 2 lists the copolymers that were made together with their solution inherent viscosities (IV), and some differential scanning calorimeter (DSC), dynamic mechanical thermal analysis (DMTA), and thermogravimetric analysis (TG) data. Polymer 1, made from EIKIE and TPC, was semicrystalline with a Tgof 247 "C and a T, of 435 "C.Polymer 7, made from EIKIE and IPC, was also semicrystalline with a Tgof 218 "C and a T, of 358 "C. These polymers were described in an earlier paper,' and it was partly because of their interesting properties that we extended our studies to the copolymers discussed in this paper. Polymers with nominally the same repeat unit as 1 and 7 have also been made by the amic acid route.7 We expected that the incorporation of mixtures of isoph- thaloyl and terephthaloyl units would disrupt the structure of the resulting polymers to such an extent that crystallization would be suppressed. However, to our surprise all the copoly- mers were semicrystalline.Fig. 1 shows the effect of the changing ratio of isophthaloyl to terephthaloyl units on both Tgand T,. With the exception of polymer 2 the Tgvalues increased with increasing terephthaloyl content. However, the changes in T, were more complex. The general trend was an increase in T, with increasing terephthaloyl content, but replacement of 15% of isophthaloyl units with terephthaloyl units (1 5 :85, TPC :IPC) caused a slight reduction of T, from 358 to 354 "C which was contrary to the trend. Also at 85: 15 TPC :IPC (polymer 2) two melting points were observed. Fig. 2 shows the DSC scan of a quenched (amorphous) sample of the polymer. In addition to the Tgtransition at 231 "C and a crystallization exotherm at 288 "C, there is a major melting endotherm at 395 "C and a minor melting endotherm at 432°C.The higher melting point corresponds with that of the 1OO:O TPC:IPC polymer and is presumably associated with blocks of that structure. The anomalous Tg of this polymer is probably related to its unusual morphology. Fig. 3 shows a wide-angle X-ray diffraction scan of an annealed sample of polymer 4 which contained a 50 :50 ratio of isophthaloyl and terephthaloyl units. There are two strong diffraction peaks corresponding to d spacings of 3.96 and 4.75 A. Integration of the areas under the crystalline peaks and under the amorphous halo indicated an approximate degree of crystallinity of 34%. All the copolymers showed good thermo-oxidative stability as measured by TG.A 1% mass loss occurred at an average temperature of 478 "C in air, and at an average temperature of 505 "C in nitrogen. However, the melt stability of the copolymers was not good. It was possible to make com-I I I 200 300 400 T/ "C Fig. 2 DSC scan of polymer 2 10 20 30 40 201" Fig. 3 Wide-angle X-ray diffraction scan of polymer 4 440 r p320 2801I 200 1OO:O Fig. 1 polymer I I I I I 7-I 80:20 60:40 40:60 20:80 0:lOO TPC:IPC ratio Series A copolymers; T, and T,,,data Table 2 Series A copolymers TPC:IPC ratio IV/ cm3 g- ' TI%(air)/ "C 100 :0 121 247 435 480 472 85: 15 132 227 395,432 487 543 70 :30 71 233 381 465 502 50 :50 109 229 382 469 507 30 :70 168 227 37 1 490 520 15:85 100 220 354 494 508 0: 100 112 218 358 459 486 J.MATER. CHEM., 1991, VOL.1 Table 3 Series B copolymers polymer EKE :EIKIE ratio IV/cm3 g-' Tg/"C TJC AHJJ g-I Tl%(air)/ "C Tl%(N2)/"C 8 1oo:o 115 170 38 1 75 495 520 9 90.2:9.8 108 180 369 65 470 505 10 80.4:19.6 134" 186 357 35 405 470 11 70.6:29.4 127 186 347,372 34 445 495 12 60.8 :39.2 141 196 33 5,3 72 32 475 495 13 55.9:44.1 173 204 33 1,374,406 29 450 515 14 5 1 .O:49.0 163 208 377,409 27 455 480 15 36.3 :63.7 152 220 382,420 17 465 500 16 21.6: 78.4 139 237 37 5,427 52 465 520 1 0: 100 121 247 435 57 480 472 " Some undissolved gel was filtered from the solution.pression-moulded plaques of fair quality by pressing at a temperature slightly above the T,, but attempts to measure the melt viscosity of the polymers by capillary rheometry showed that rapid and large increases in viscosity occurred when the polymers were held for short times above their melting points. We intend to report further work concerning melt stability in due course. Series B Table 3 lists the copolymers that were made together with IV, DSC, DMTA and TG data. Polymer 8 is the aromatic ether- ketone polymer PEKEKK which has a T, of 170 "C and a T, of 381 "C.* We have studied this series of copolymers in greater depth than the series A copolymers and, in addition to examining their thermal properties, we have investigated some mechan- ical, electrical, flammability, and hydrolytic properties.Thermal Properties As with series A we expected that the incorporation of monomer I into the backbone of the polymer would disrupt the structure and produce amorphous polymers. However, again we were surprised to find that all the copolymers were semicrystalline. Furthermore, although incorporation of I caused the T, of the polymers to rise steadily, from 170 to 247 "C, the effect at low levels of imide incorporation was to depress the melting point rather than to increase it as we had expected. We thus had the attractive opportunity of being able to increase the T, of the basic PEKEKK polymer whilst simultaneously lowering the T, and so potentially increasing the facility with which the polymer could be melt processed.Fig.4 shows the effect of increasing imide content on T, and T,. It can be seen that between ca. 30 and 80mol% imide? more than one melting point was observed, indicating the presence of different crystalline morphologies. At 40 mol% imide the T, has increased to 196°C whilst the higher T, was still slightly lower than for PEKEKK at 372 "C. The melting behaviour of the copolymer series was most unusual and suggested that three different morphologies could occur, represented by lines 1, 2 and 3 in Fig. 4. The line 1 melting points are presumably associated with a disrupted PEKEKK structure. The line 2 melting points are all similar and average 376 "C. This suggests that they are associated with a basically undisrupted PEKEKK structure. The line 3 melting points are all above 400 "C and are presumably associated with a disrupted PEIKIEKK structure.Fig. 5 shows the changing nature of the DSC endotherms with increasing imide content. The intensity of the melting endotherms decreased as the imide content increased, falling 7 By 30 mol% imide we mean a 70:30 mol ratio of monomer 111 (EKE): monomer I (EIKIE). 450 II I I I 400 350 9,300I--250 200 1OO:O 80:20 60:40 40:60 20:80 0:lOO EKE:EIKIE ratio Fig. 4 Series B copolymers; Tgand T,,,data from 75 J g-' for polymer 8, PEKEKK, to a minimum of 17 J g-' with polymer 15 which contained 63.7 mol% imide. As the imide content increased further the endotherms increased in intensity, reaching 57 J g-' with polymer 1, PEIKIEKK.Thus, although the incorporation of the imide unit into the PEKEKK structure did not completely destroy the crystallinity of the polymer, it did reduce it considerably. All the DSC scans showed a small endotherm at ca. 280 "C. We have found that all annealed samples show this minor melting endotherm, typically between 10 and 20 "C above the annealing temperature. Similar behaviour has been noted in PEEK.' It should be appreciated that our analysis of the melting behaviour is mainly conjectural at this stage and is no doubt an oversimplification. The relative intensities of the melting endotherms observed in DSC scans are sensitive to the thermal history of the polymers, and the morphology of the copolymers will be affected by how random or blocky the polymers are, and this in turn may be affected by the polymerization conditions.Work in this area is continuing. The copolymers showed good thermo-oxidative stability as measured by TG. 1% mass loss in air was at 461 "C, and in nitrogen was at 497 "C.However, as with the series A copoly-mers, melt stabilities were not particularly good, and they became worse with increasing imide content. J. MATER. CHEM., 1991, VOL.1 Table 4 Mechanical properties of series B copolymers ~ ~ ~~~~ polymer EKE :EIKIE ratio modulus/MPa tensile strength/MPa" EtJ (%) 8 1oo:o 3730 103.4 3.5 9 90.2 :9.8 3720 106.4 4.5 10 80.4:19.6 3760 96.1 3.3 11 70.6:29.4 3770 108.7 6.6 12 60.8:39.2 3600 103.4 6.5 13 55.9:44.1 3650 102.9 4.9 14 51.O :49.0 3655 64.4 2.0 15 36.3 :63.7 3575 109.1 8.3 16 21.6 :78.4 3700 63.0 1.8 1 0: 100 4205 63.2 1.8 The data given are ultimate tensile strengths except for polymers 11, 12 and 15 when they are yield stresses.in Table 5. The relative permittivity was unaffected by imide content and was 3.08 k0.09. The loss factor (tan 6) decreased with increasing imide content from 0.0024 at Omol% imide to 0.0018 at 100mol% imide. This was contrary to our expectations as the EIKIE unit contains more polar groups than the EKE unit and therefore we expected power loss to increase. Flammability and Smoke Production Fig. 6 shows the limiting oxygen index (LOI) of the copolymers as a function of imide content.All the copolymers were inherently fire resistant with LO1 values between 35.5 and 44.5. The slight fall in LO1 as the imide content increased to ca. 30 mol% may be related to the decrease in the melting temperature of the copolymers. Fig. 7 shows the plots of smoke production as a function of temperature for five of the copolymers. No smoke was detectable below 525 "C,but above this temperature smoke production rose rapidly until the self-ignition temperature (Tip)was reached at ca. 630 "C.Above Ti, smoke production was lower and approximately constant for a given polymer. Increasing imide content has a pronounced smoke suppressing effect, and Fig. 8 clearly illustrates this showing how the average total smoke production, S, fell from 549 cm2 g-' at 0 mol% imide to 175 cm2 g-' at 100 mol% imide.Hydrolytic Stability In our previous paper we suggested that the method of synthesis of these EK-imide polymers should result in good hydrolytic stabilities.' To test this hypothesis we boiled poly- mer fluff samples in distilled water and in pH 2, 7, and 12 aqueous solutions, and we monitored the changes in inherent viscosity of the polymer with time. Table 6 shows the results obtained for polymer 8, PEKEKK, which contained no imide groups, and Table 7 shows the data for polymer 11 which contained 29.4 mol% of EIKIE. In the case of PEKEKK all the measured inherent viscosities were in the range 108-121 cm3 g-' and they showed no significant Table 5 Electrical properties of series B copolymers 0% imide 200 300 /I560 9.8% 19.6% 29.4% 39.2% 44.I Yo 49.0% 63.7% 78.4% 100% 2z 300 400 500 Fig. 5 Series B copolymers; DSC scans Mechanical Properties Initial modulus, tensile strength, and elongation-to-break (Eb) data are listed in Table4. Because of the relatively poor quality of the compression-moulded test specimens, many of the samples broke before they yielded. However, data from the better quality specimens suggested that both modulus and tensile strength were unaffected by imide content. Tensile strengths were ca. 105 MPa and moduli were ca. 3.7 GPa; similar to the values quoted for both PEEK and PEKEKK.*,'' The elongation to yield of the better quality polymer EKE :EIKIE ratio 8 1oo:o 9 90.2 :9.8 10 80.4:19.6 11 70.6:29.4 relative permittivity loss factor x I04 3.08 24 3.14 20 3.19 22 2.97 20 3.18 20 3.13 21 3.03 19 3.10 17 3.04 18 2.89 18 specimens was ca.6.5% 12 60.8:39.2 13 55.9:44.1 Electrical Properties 14 51.O :49.0 All the copolymers were good insulators with resistivities in 15 36.3 63.7 excess of 1.4 x 1OI6 0 cm (the maximum measurable on our 16 21.6 :78.4 1 0: 100equipment). Relative permittivities and loss factors are listed 660 48 I I I I 44 2 40 36 0 32 I I I I 0 20 40 60 80 100 imide (mol%) Fig. 6 Series B copolymers; LO1 us. imide content +-19.6% imide +-44.1% imide 1 +--78.4% imide +-100%imide 1 0' II I I 500 600 700 800 900 T/ "C Fig.7 Series B copolymers; smoke us. temperature. Imide content (Yo):0,0; 0, 19.6; A, 44.1; 0,78.4; 0, 100 trends with time. If one assumes therefore that no degradation took place, a mean IV of 114.5 cm3 g-' and a standard deviation of 3.4 cm3 g- ' can be calculated. This gives a useful measure of the precision of the IV determination. In the case of polymer 11 there were no significant changes in IV in distilled water, pH 7, and pH 2 solutions up to 167 h. After 982 h in distilled water and in the pH 2 solution the IV values had fallen from 113 cm3 g-' to 106 cm3 g-' and 105 cm3 g- ' rzspectively. These are small but possibly signifi- J. MATER. CHEM., 1991, VOL.l 100 0 20 40 60 80 100 imide (mot%) Fig.8 Series B copolymers; s us. imide content Table 6 Inherent viscosities of polymer 8 (PEKEKK) after boiling in water time/h distilled water PH 2 pH7 pH 12 0 115 115 115 115 1 116 118 115 116 7 108 117 109 113 31 116 119 116 121 167 114 116 114 115 479 112 114 108 113 982 110 116 109 115 Table 7 Inherent viscosities of polymer 11 after boiling in water IV/cm3 g- time/h distilled water PH 2 pH7 pH 12 0 113 113 113 113 1 I18 115 112 114 7 115 113 113 112 31 115 113 114 109 167 115 116 112 107 479 109 110 105 101 982 106 105 99 92 cant changes. After the same time in the pH 7 phosphate buffer solution the IV had fallen by slightly more to 99 cm3 g-'.Degradation was more evident in the pH 12 solution where there was a steady fall in IV with time to 92 cm3 g-' after 982 h. In general, as we had predicted, the hydrolytic stability of the EK-imide polymer was good. We believe that these results support the case that the reported hydrolytic instability of some polyimides made by the amic acid route is because of defects in the polymer chain rather than because of any inherent instability of the imide bond.' '-14 J. MATER. CHEM., 1991, VOL.l 661 We thank Bruce Fox for X-ray diffraction measurements, Tim Harding for the testing of tensile properties, John Harris for DSC, DMTA and TG measurements, Pat Horner and Ian Towle for their helpful discussions, and Raychem Ltd. for permission to publish this paper.5 6 7 8 J. F. Rabek, Experimental Methods in Polymer Chemistry, Wiley-Interscience, Chichester, 1980, p. 507. R. H. Whiteley, Int. Wire Cable Symp. Proc., 1982, 427. P. M. Hergenrother, N. T. Wakelyn and S. J. Havens, J. Polym. Sci., Polym. Chem. Ed., 1987, 25, 1093. BASF product brochure B607e, ‘ULTRAPEK’, 1989. 9 S-S. Chang, Polym. Commun., 1988, 29, 138. References 10 T. W. Haas, in Handbook of Plastic Materials and Technology, ed. I. Rubin, Wiley, New York, 1990, p. 287. P. J. Horner and R. H. Whiteley, J. Muter. Chem., 1991, 1, 271. V. Janson and H. C. Gors (Raychem Corp.), PCT Znt. Appl., WO 84 3891, 1984. P. Becker, P. J. Horner, S. Moore, L. J. White, L. M. Edwards, B. Macknick and R. J. Mosso (Raychem Corp.), Eur. Pat. Appl., 11 12 13 14 C. E. Sroog, J. Polym. Sci., 1965, 3, 1373. D. R. Askins, Hydrolytic degradation of Kapton film, Air Force Wright Aeronautical Laboratories, AFWAL-TR-83-4125, 1983. J. H. Hodgkin, J. Appl. Polym. Sci., 1976, 20, 2339. D. Kumar, J. Polym. Sci., Polym. Chem. Ed., 1980, 18, 1375. EP 314384, 1989. J. Koch, W. Stegmaier and G. Heinz, BASF AG, DE 3829520 Al, 1990. Paper 1/01065D; Received 6th March, 1991