J. MATER. CHEM., 1994, 4(8), 1167-1172 Curing Reactions in Acetylene-terminated Resins Part 5. t-Cyclotrimerization versus Linear Polyene Formation in the Catalysed Cure of Ethynylaryl-terminated Monomers William EmDouglas** and Andrew S. Overend Department of Industrial, Organic and Polymer Chemistry, Kingston Polytechnic, Penrhyn Road, Kingston-upon-Thames, Surrey, UK KT1 2EE Structural differences in the polymers obtained by catalysed cure of the ethynylaryl-terminated monomer [1,4-(1,3-HC-CC6H4CH,0)C6H4]2CMe2 (1) could not be detected by IR or NMR spectroscopy. However, a study of the solvent-free catalysed polymerization of the model monomer, phenylacetylene, showed that the extent of cyclotrimeriz-ation decreased in the order (y-Cp)Co(CO),>>(y-Cp),Ni z(PPh,),NiCI,>> (PPh3)2PdCI, with an accompanying iicrease in the degree of linear polyene formation, suggesting that this is so also for the catalysed cure of 1.High-performance resins formed from acetylene-terminated monomers (ATMs) are currently being developed to replace, for example, epoxies, which suffer from an undesirable reduction in glass-transition temperature when used under hot, wet This paper is part of a series describing the effect of catalysts on the curing reactions of the two main groups of ATMs, those terminated by aryl prop-2-ynyl ether and those by ethynylaryl groups. In Part 4,4 we have shown that, in the case of aryl prop-2- ynyl ether ATMs, the use of catalysts gives rise to quite different crosslink structures than in the absence of catalysts, and that by judicious choice of catalyst cyclotrimerization and/or linear polymer formation can be obtained. Here, we give the results of an IR and NMR study into the catalysed polymerization of ethynylaryl-terminated ATMs: the thermal properties of the fully cured resins were reported previ~usly.~ Some of the preliminary results concerning polymerization of phenylacetylene in the presence of nickelocene have been communicated.6 The uncatalysed cure of ethynylaryl-terminated ATMs takes place via a free-radical reaction resulting in a network whose crosslink sites are linear conjugated polyenes formed from six to eight ethynyl groups7 together with some cyclotrimerization of acetylene end-group~.~?~ Many transition-metal complexes are known to catalyse acetylene reactions such as cyclo-trimerization, cyclotetramerization and linear polyene forma- tion," and their use in the cure of ATMs might therefore be expected to affect the structure and properties of the resulting resins.However, few such studies have been reported. (PPh,),NiCl, has been found to lower the cure temperature of ethynylaryl-terminated monomers sufficiently (< 177 "C) for the resins to be used as adhesives for A1 alloys." The effect of high concentrations (20 mol% ethynyl group) of (PPh,)2Ni(C0)2 on the processing of ethynylaryl-terminated polyimides has also been examined.12 Two of the catalysts which we had originally reported as being active13 have been used since in the cure of aryl prop-2-ynyl ether ATMs, namely cobaltocene (which was found to give rise to cyclotrimeriz- ation)14 and bis( tripheny1phosphine)palladium dich10ride.l~ Experimental IR spectra (CsI discs) were recorded by use of a Perkin-Elmer 728 spectrometer connected to a Perkin-Elmer 360 data -f Part 4:W.E. Douglas and A. S. Overend, J. Muter. Chem., 1993, 3, 1019. $ Present address: CNRS UM 44, Case 007, USTL, Place E. Bataillon, 34095 Montpellier CCdex 5, France. station. 'H NMR spectra were run on a Bruker WP-80-SY spectrometer operating at 80.13 MHz or a Perkin-Elmer R32 instrument at 90 MHz. Mass spectra were obtained by use of an AE1 MS9 mass spectrometer. Molecular weights were determined by VPO in toluene at 45°C using a Knauer instrument calibrated with benzil.Elemental analyses were carried out by Butterworth Laboratories, Twicker, ham. The catalysts (~-cp)~Ni,(y-Cp)Co(CO), (technical grade in dicyclopentadiene; this particular batch assayed by Aldrich at 50%), and (PPh,),RhCl were commercial samples (Aldrich) and were used as received except nickelocene, which was resublimed before use. (PPh,),NiCl, was prepared from NiC1,-6H20 and PPh, in glacial acetic acid. The ethynylaryl ATM [I$( 1,3-HC-CC6H4CH,0)C6H~],cMe,(1) and samples of fully cured 1 for IR studies were prepared as pre- viously described5 (the IR absorbances of 1 are listzd below). v PjyQo1 1 Phenylacetylene (Aldrich) was redistilled under reduced pressure.When only a proportion of the reaction mixture was worked up, the overall percentage yield is given icalculated from the isolated yield for the sample taken), All reaction products were recovered in essentially their entirety, the sum of the individual yields approximating to 100% iii all cases. 1,2,4-Triphenylbenzene exists in two forms melting at 100"C and 119-120 OC;16 the mp of 1,3,5-triphenylbenzene is 172-173 OC.I7 IR Spectrum of 1 vmax/cm-': 3290s, 3060w, 3030w, 2980m, 2970111, 2950m, 2920m, 2870m, 2100w, 1885w( br), 1800vw, 1605m, 1580m, 1505s, 1480m, 1460m, 1440w, 1430m, 1410vw, 13&0s, 1360w, 1310w, 1285m, 1235s, 1180s, 1155w, 1115w, 1095w, 1085w, 1030s, 1020s, 970vw, 960vw, 945w, 930vw, 910w, 895m, 885m, 840s, 820s, 810m, 785m, 765m, 735w, 690m, 640m, 6:30m, 600m. (q-Cp)Co(CO),-Catalysed Polymerization of Pheny lacetylene To degassed phenylacetylene (3 g, 29.4 mmol) under N2 was added (y-Cp)Co(CO), (40 p1 of 50% solution in dicyclo- pentadiene, 0.11 mmol) and the mixture was heated at 140 "C for 4 h.A portion (0.89 g) of the resulting polymerized material was chromatographed on silica gel. Elution with 80 :20 v/v n-hexane-dichloromethane gave unchanged phenylacetylene (0.16g, 18%) followed by a white solid (0.68 g, 76%), the latter being a mixture of 1,2,4- and 1,3,5-triphenylbenzene with mp 132-136°C. (Found: C, 94.04; H, 5.96%. Calc. for C24H18: C, 94.08; H, 5.92%.) IR spectrum vmax/cm-' : 3080m, 3060m, 3025m, 1950w(br), 1880w(br), 18lOw(br), 1755w(br), 1595m, 1575w, 1495m, 1470m, 1440m, 1410w, 1390w, 1180w, 1155w, 1075m, 1030w, 1010m, lOOOw, 990w, 920w, 910vw, 900m, 875w, 845m, 780m, 765s, 750m, 740s, 715m, 705s.'H NMR, 6, (solvent CDCI,; standard TMS): 7.10-7.25 (1.00 H, m, arom.), 7.25-7.95 (1.73 H, m, arom.). (q-Cp),Ni-Catalysed Polymerization of Phenylacetylene To degassed phenylacetylene (3 g, 29.4 mmol) under N, was added (q-Cp),Ni (4.5 mg; 2.38 x lop5 mol) and the mixture was heated at 115 "C for 6 h. A solution of the crude product in 30 ml of dichloromethane was added dropwise to 300 ml of methanol, thus precipitating an orange polymer which was separated by filtration, washed thoroughly with methanol and dried. The precipitation pro- cedure was repeated and the product was dried under vacuum at 60°C to constant weight (1.17 g; 39%).The 'H NMR [Fig. l(u)] and IR spectra [Fig. 2(a)] showed the polymer to be trans-cisoidal poly( phenylacetylene) (see Discussion). M, 1600 g mol-', DP, 15.7. After removal of solvent from the combined filtrates, a portion (0.78 g) of the methanol-soluble material was chroma- tographed on silica gel. Elution with 80:20 v/v n-hexane-dichloromethane afforded unchanged phenylacetylene (0.10 g; I 8 I 7 I 6 L 8 Fig. 1 'H NMR spectra (90 MHz) of trans-cisoidal poly(pheny1ace-tylene) from polymerization of phenylacetylene in the presence of (a) (q-Cp),Ni, (b)(PPh,),NiCl, and (c) (PPh,),PdC12 J. MATER. CHEM., 1994, VOL. 4 1 6 3000 2000 1400 800 wavenumberkm-' Fig.2 IR spectra of trans-cisoidal poly( phenylacetylene) from poly- merization of phenylacetylene in the presence of (ti) (q-Cp),Ni, (h) (PPh,),NiCl, and (c) (PPh,)2PdC12 8%) followed by white crystals (0.31 g, 24%), the latter being a mixture of 1,2,4- and 1,3,5-triphenylbenzene with mp 122-126°C. (Found: C, 94.20; H, 5.80%. Calc. for CZ4Hl8: C, 94.08; H, 5.92%.) IR spectrum vmax/cm-' : 3080m, 3060m, 3020m, 1950w(br), 1880w(br), 18lOw(br), 1755w( br), 1595m, 1575w, 1495m, 1470m, 1440m, 1410w, 1390w, 1180w, 1155w, 1075m, 1030w, 1010m, 990w, 920w, 900m, 875w, 845m, 780m, 765s, 750w, 710m, 715m, 705s. 'H NMR, BH (solvent CDC1,; standard TMS): 7.10-7.25 (1.00 H, m, arom.), 7.25-7.95 (1.14 H, m, arom.). Finally, elution with 60:40 v/v n-hexane-dichloromethane gave a yellow solid (0.38 g; 30%), this being a mixture of low- molecular-weight linear oligomers of phenylacet ylene. ( Found: C, 93.98; H, 6.02%, Calc.for C24Hl,: C, 94.08; H, 5.92%.) IR spectrum vmax/cm-' :3080m, 3060m, 3020m, 2925w, 2855w, 2250vw, 1950w( br), 1870w( br), 1800w( br), 1750w( br), 1625w( br), 1595m, 1570w, 1490s. 1445m, 13XOvw, 1330vw, 1265w, 1220w, 1 NOW, 1160w, llOOw, 1075w, 1030w, lolow, ~~OVW,910~, 885~, 850~, 755~, 740m, 695s. 'H NMR, 8, (solvent CDCl,; standard TMS) :6.5-6.9 ( 1 H, m, olefinic), 7.0-7.9 (5 H, m, arom.). M, 460 g mol-', DP, 4.5. (PPh,),NiCl,-Catalysed Polymerization of Phenylacetylene To degassed phenylacetylene (3 g, 29.4 mmol) under N, was added (PPh3)2NiC12 (15.6 mg; 2.38 x mol) and the mix- ture was heated at 140°C for 4 h.A solution of the crude product in 30 ml of dichloromethane was added dropwise to 300 ml of methanol, thus precipitating a pale-brown polymer, which was separated by filtration, washed thoroughly with methanol and dried. The precipi- tation procedure was repeated and the product was dried under vacuum at 60°C to constant weight (0.88 g; 29%). The 'H NMR [Fig. l(b)] and IR spectra [Fig. 2(h)] showed the polymer to be trans-cisoidal poly ( phenylacetylene) (see Discussion). M, 1100 g mol-', DP, 10.8. After removal of solvent from the combined filtrates, a portion (0.72 g) of the methanol-soluble material was chroma- tographed on silica gel. Elution with 80: 20 v/v n-hexane-dichloromethane afforded unchanged phenylacetylene (0.22 g; J.MATER. CHEM., 1994, VOL. 4 22%) followed by white crystals (0.26 g, 26%), the latter being a mixture of 1,2,4- and 1,3,5-triphenylbenzene with mp 124-128°C (Found: C, 94.32; H, 5.68%. Calc. for C24H18: C, 94.08; H, 5.92%.) IR spectrum identical to that for the corresponding material isolated in the case of the (q-Cp),Ni- catalysed polymerization of phenylacetylene. 'H NMR, 6, (solvent CDCI,; standard TMS) :7.10-7.25 (1.00 H, m, arom.), 7.25-7.95 (1.22 H, m, arom.). Finally, elution with 60 :40 v/v n-hexane-dichloromethane gave a yellow solid (0.18 g; l8%), this being a mixture of low- molecular-weight linear oligomers of phenylacetylene. (Found: C, 94.56; H, 5.44%. Calc. for C24H18: C, 94.08; H, 5.92%.) IR spectrum identical to that for the corresponding material isolated in the case of the (q-Cp),Ni-catalysed polymerization of phenylacetylene. 'H NMR, 6, (solvent CDC1,; standard TMS): 6.5-6.9 (1 H, m, olefinic), 7.0-7.9 (5 H, m, arom.).M, 490 g mol-', DP, 4.8. (PPh3),PdC1,-Catalysed Polymerization of Phenylacetylene To degassed phenylacetylene (3 g, 29.4 mmol) under N, was added (PPh,),PdCl, (16.7 mg; 2.38 x lop5mol), and the mix- ture was heated at 130°C for 5 h. A solution of the crude product in 30 ml of dichloromethane was added dropwise to 300 ml of methanol, thus precipitating a brown polymer, which was separated by filtration, washed thoroughly with methanol and dried. The precipitation pro- cedure was repeated and the product was dried under vacuum at 60 "C to constant weight (2.59 g; 86%). The 'H NMR [Fig.l(c)] and IR spectra [Fig. 2(c)] showed the polymer to be trans-cisoidal poly(phenylacety1ene) (see Discussion). [Found: c, 93.95; H, 6.05%. Calc. for (C8H6),: C, 94.08; H, 5.92Y0.1 M, 1850 g mol-', DP, 18.1. After removal of solvent from the combined filtrates, the brown methanol-soluble material was chromatographed on silica gel. Elution with 80 :20 v/v n-hexane-dichloromethane afforded as the only other products unchanged phenylacety- lene (0.2 g; 7%) followed by white crystals (0.11 g, 4%) of 1,3,5-triphenylbenzene, mp 170-171 "C. (Found: C, 94.03; H, 5.97%. Calc. for C24H18: C, 94.08; H, 5.92%.) IR spectrum v,,,/cm-': 3080m, 3060m, 3025w, 1950w( br), 1880w(br), MOW( br), 1755w(br), 1595m, 1575w, 1500m, 1410m, 1310w, 1155w, 1075w, 1030w, 910w, 890vw, 875m, 765s, 750s, 700s.'H NMR, 6, (solvent CDC1,; standard TMS) :7.25-7.90 (m; arom.). MS (70 eV): m/z 306 (Mf). Results and Discussion IR and NMR Spectra of Polymers from 1 [1,4-(1,3-HC~CC6H4CH20)C6H,],CMe,(l),the ethynyla- ryl ATM, and samples of 1 fully cured in the presence of various catalysts were prepared as previously de~cribed.~ The polymerization enthalpies suggest that crosslinking occurs by cyclotrimerization in the presence of (y-Cp)Co(CO),, but predominantly by non-aromatic conjugated linear polyene formation with the other catalyst^.^ In principle, the IR spectra should reflect these differences.Thus, from the known pos- itions of the CH out-of-plane deformations for 1,2,4- and 1,3,5-triphenylbenzene,ls it can be expected that 1,2,4-cyclotrimerization would give rise to new absorbances at 900 and 845 cm-' and 1,3,5-cyclotrimerization to a new band at 875 cm-'. On the other hand, from IR data for cis- and trans- poly( phenylacetylene)," conjugated linear polyenes with a cis structure would be expected to show absorbances at 740, 895 and 1380cm-' whereas those with a trans configuration should possess bands at 922, 970 and 1265 cm-'. In fact, the IR spectra (Fig. 3) of the crosslinked resins are all very similar because absorbances originating in the monomer at 1380, 1235, 970,930, 895, 885, 840 and 735 cm-' (see Experimental) obscure any new bands arising in these regions.However, the absence of v(-C-H) and v(C=C) stretches in all cases shows that the resins contain no unreacted acetylene groups. Unlike the case of aryl prop-2-ynyl ether ATMs rn here the propargyl CH20 group could be used as a probe,4 the NMR spectra of both the partly and fully cured resins formed from 1 were of no help in determining crosslink structures since all new resonances fall in regions of the spectra where peaks are already present in the monomer. Previously, in an F'TIR and solid-state CPMAS NMR study of a fully cured ethynylaryl ATM,,' it proved difficult to detect different crosslink struc- tures. However, in a more recent CPMAS NMR investigation with resins obtained from isotopically labelled monomers of the ethynylaryl type, resonances from new aromatic groups were observed.' The investigation was therefore pursued by studying the solvent-free catalysed polymerization of phenylacetylene as a model monomer for 1.Unlike 1, phenylacetylene contains only a single acetylene group, and hence the polymerization products are soluble and can be separated and characterized. Catalysed Polymerization of Phenylacetylene in the Absence of Solvent The solution polymerization of phenylacetylene in the pres- ence of a wide variety of catalysts has been extensively studied previously. Here, we report the results of an investigation into the catalysed bulk polymerization of phenylacetylene under I 4 10 3000 2000 1400 800 wavenumbedcm-' Fig.3 IR spectra (CsI discs) of 1 fully cured in the presence of (a)no catalyst, (b) (~pCp)Co(C0), (0.4 mol% ethynyl group), (c) (q-Cp),Ni (0.08 mol% ethynyl group), (d) (PPh,),NiCI, (0.08 rn01':/~ ethynyl group) and (e) (PPh,),PdCI, (0.08 mol% ethynyl group) the same conditions as we used for the preparation of fully cured resins from 1.' The reaction conditions and results are summarized in Table 1. (4(r7-CP)CO(C0)2 The reaction mixture was separated by chromatography into unchanged monomer (18%) and a white solid (76%) which was found to be a mixture of 1,2,4- and 1,3,5-triphenylbenzene with mp 132-136°C. The 'H NMR spectrum showed only aromatic proton resonances at 6 7.10-7.95. In particular, there were no signals characteristic of linear alkenes in the 6 6.5-7.0 region.Comparison of the IR spectrum with those of 1,2,4- and 1,3,5-triphenylben~ene'~confirms that the material is a mixture of the two isomers. Thus, benzoidal out-of-plane C-H bending absorbances are present both at 845 and 900 cm-l for 1,2,4-triphenylbenzene and at 875 cm-' for the 1,3,5 isomer. No bands were observed in the 1610-1630 cm-' region for linear poly( phenylacetylene). The proportion of the two isomers may be estimated by 'H NMR from the relative intensities of the aromatic proton resonances in the S 7.1-7.25 and 6 7.25-7.9 regions. 173,5-Triphenylbenzene exhibits reson- ances only in the latter region (intensity 18H),," whereas the 1,2,4 isomer shows signals in both regions with intensities of 10H and 8H, Hence, from the measured integration ratio of 1:1.73 the ratio of 1,2,4- to 1,3,5-triphenylbenzene is 2: 1 (a 3: 1 ratio would be expected on statistical grounds).Thus, (y-Cp)Co(CO), gives rise to exclusive cyclotrimeriz- ation of phenylacetylene in the absence of solvent. In solution, (y-Cp)Co(CO), is known to cyclotrimerize acetylenes.22 Therefore, cyclotrimerization probably occurs also in the (q-Cp)Co(CO),-catalysed cure of 1, although thermal studies indicate5 that after gelation other non-catalysed crosslinking reactions also take place because of reduced chain mobility. (b)(V-CP),Ni Addition of methanol to a dichloromethane solution of the reaction mixture precipitated a methanol-insoluble orange polymer (39%) which elemental analysis showed to be pure poly( phenylacetylene) with M, = 1600, corresponding to a degree of polymerization of 15.7.The 'H NMR spectrum [Fig. l(a)] showing a broad resonance in the 6 6-8 region is typical of that for trans-cisoidal poly (phenyla~etylene).'~ Furthermore, the IR spectrum [Fig. 2(a)] [with no v(C-C) stretch] is essentially identical to that for trans-cisoidal poly( phenyla~etylene),'~ both showing bands characteristic of trans-poly(phenylacety1ene) at 912, 970 and 1265 cm-', and also an absorbance at 885 cm-' which is specific to cis-poly(phenylacetylene). The methanol-insoluble polymer was therefore trans-cisoidal poly( phenylacetylene). The methanol-soluble material was separated into three fractions by column chromatography.Unchanged phenyl- acetylene (8%) was eluted first followed by a white crystalline solid (24%) which, as discussed above for the (y-J. MATER. CHEM.. 1994, VOL. 4 Cp)Co(CO),-catalysed system, was shown by elemental analy- sis and IR and 'H NMR spectroscopy to be an 84 : 16 mixture of 1,2,4- and 173,5-triphenylbenzene. The final fraction was a yellow solid obtained in 30% overall yield (not 21% as previously incorrectly reported6) which elemental analysis showed to be poly(phenylacetylene), the M, of 460 corre- sponding to a degree of polymerization of 4.5. The IR spectrum showed a v(CzC) stretch at 2250cm-' [but no v(-C-H) absorbance] as well as bands characteristic of both cis-(885 and 740 cm-') and trans-poly(phenylacety1ene) (970 and 912 cm-').19 The 'H NMR spectrum exhibited resonances in the 6 6.5-6.9 alkene region as well as aromatic signals, the integration ratio being 1:5.The yellow solid was therefore a mixture of low-molecular-weight linear oligomers of phenylacetylene. Acetylenes react with nickelocene to form two types of air- stable compound; green binuclear acetylene-bridged 2 and red mononuclear 3 (Scheme l).23Formation of the latter complex is favoured by the presence of electron-withdrawing substitu- ents in the acetylene compound.23 The complexes are prepared by reaction in THF for 20-30 h at room temperat~re,,~ or in the case of the ethyne complex 2, in THF at 12 atm and 80 "C for 15 h.24 The ethyne complex 2 as well as nickelocene itself have been claimed to be active catalysts in the solution polymerization of acetylenes at 70 "C both at atmospheric pressure and under high pressure.25 The presence of an aromatic heterocyclic amine (e.g.pyridine), which forms a reactive complex with the nickel catalyst and can also act as solvent, is a necessary component in the process.25 Until the present study was undertaken,13 nickelocene alone had not been used as a catalyst for acetylene polymerization, although with F,CC=CCF3 it has been reported to give trace amounts of the cyclotrimer after 10 h at 358 K.26At 50°C in benzene for 25 h, 0.1-0.3 mol% (q-Cp),Ni.2A1Br3 catalyses the conversion of acetylenes into a mixture of cyclotrimers and linear polymer.27 However, under the same conditions no reaction was observed in the absence of A~BI-,.,~ We have found that under solvent-free conditions, nickelocene catalyses terminal-acetylene p~lymerization.'~ As reported here, with Q 2 Scheme 1 Table 1 Reaction conditions and products for the catalysed polymerization of phenylacetylene methanol-insoluble trans-cisoidal catalyst catalyst conc.(mol YO) polymerization T/OC t/h conversion (YO) poly (phen ylacetylene) yield (YO) DP," (rl-CP)CO(CO), 0.4 140 4 82 - - (Ph,Pj,PdCI, (rl-CPj,Ni (Ph3P)2NiCI, 0.08 0.08 0.08 130 115 140 5 6 4 93 92 78 86 39 29 18.1 15.7 10.8 methanol-soluble linear oligomers yield (YO) DP," 30 4.5 18 4.8 triphen ylbenzene ratio yield (YO) 1,2,4:1,3,5' 76 66134 24 84:16 26 81:19 4 0:100 A '--' entry indicates that no material of that type was isolated."Calculated from the value of M, determined by VPO. bDetermined by 'H NMR spectroscopy (see Discussion). J. MATER. CHEM., 1994, VOL. 4 phenylacetylene both cyclotrimerization and linear polyene formation occur, whereas in the case of aryl prop-2-ynyl ether ATMs only cyclotrimers and cyclotetramers are ~btained.~ We are at present investigating the reaction mechanism. The green colour observed initially may be due to the formation of the known complex 2 (R=H, R’=Ph).28 The product distribution found for phenylacetylene is con- sistent with the value of the enthalpy of polymerization of 1 measured by DSC,’ suggesting that crosslinking occurs pre- dominantly uia non-aromatic conjugated linear polyene for- mation, together with some cyclotrimerization to give new benzene rings. (c) (PPh3)2NiC12 Addition of methanol to a dichloromethane solution of the reaction mixture precipitated a methanol-insoluble pale brown polymer (29%) which elemental analysis showed to be pure poly(phenylacety1ene) with a value of M, of 1100, correspond- ing to a degree of polymerization of 10.8. As described above for the nickelocene-catalysed system, the ‘H NMR [Fig. l(b)] and IR spectra [Fig.2(b)] showed the methanol-insoluble polymer to be trans-cisoidd poly( phenylacetylene).” The methanol-soluble material was separated into three fractions by column chromatography.Unchanged phenyl- acetylene (22%) was eluted first followed by a white crystalline solid (26%) which, as discussed above for the (y-Cp)Co(CO),-catalysed system, was shown by elemental analy- sis and IR and ‘H NMR spectroscopy to be an 81 :19 mixture of 1,2,4- and 1,3,5-triphenylbenzene. The final fraction was a yellow solid (18% yield) which elemental analysis showed to be poly(phenylacetylene), the M, of 490 corresponding to a degree of polymerization of 4.8. As in the case of the nickel- ocene-catalysed system, the IR and ‘H NMR spectra showed the yellow solid to be a mixture of low-molecular-weight linear oligomers of phenylacetylene with the structure PhC=C-( PhC=CH),-H. In previous studie~,~~?~’ the (PPh,),NiCl,-catalysed poly-merization of phenylacetylene has been studied in refluxing benzene in air.After 12 h, the reaction mixture was found to contain 39% unchanged phenylacetylene, 24% cyclotrimers (83:17 mixture of 1,2,4- and 1,3,5-triphenylbenzene) and 4% methanol-insoluble linear poly( phenylacetylene) with M, =2090. (d)(PPh3)2PdCl, Addition of methanol to a dichloromethane solution of the reaction mixture precipitated a methanol-insoluble brown polymer (86%) which elemental analysis showed to be pure poly(phenylacety1ene) with a value of M, of 1850, correspond- ing to a degree of polymerization of 18.1. As described in (b) for the nickelocene-catalysed system, the ‘H NMR [Fig. 1 (c)] and IR spectra [Fig, 2(c)] showed the methanol-insoluble polymer to be trans-cisoidal poly( phenyla~etylene).’~ The methanol-soluble material in this case consisted of only two compounds, which were separated by column chromato- graphy; unchanged phenylacetylene (7%) was eluted first followed by a white crystalline solid (4%) identified as 1,3,5-triphenylben~ene.’~,’~ The present results are in good agreement with those from previous studies of the bulk polymerization of phenylacetylene in the presence of considerably higher concentrations of (PPh,),PdCl,; with 5 mol% catalyst for 5 h at 140 OC,I9 or in air with 5 or 0.2 mol% catalyst for 4 h at 140°C.31 In both cases, methanol-insoluble trans-cisoidal poly (phenylacetylene) was obtained together with a small proportion of 1,3,5-triphenylbenzene which, it was suggested, resulted from thermal cyclization and scission of the linear poly( phenyl- acetylene) chain rather than from Pd-catalysed cyclo-trimerization.” The value of the enthalpy of polymerization of 1 in the presence of (PPh,),PdCl, is consistent with crosslinking occurring predominantly via non-aromatic conjugated linear polyene f~rrnation,~ and the present results in the ciise of phenylacetylene support this mechanism.In a previous study into the polymerization of the aliphatic diethynylaryl-terminated ATMs HC-C(CH,),C -CH (n=2-5) in refluxing cyclohexane in the presence of (Ph,P)2Ni(CO), it was shown by using IR and other analytical techniques that depending on the value of n, formation of both aromatic cyclotrimers and conjugated linear polyalkene structures occurs.32 Conclusions The ability of the catalysts to promote cyclotrimerization of phenylacetylene in the absence of solvent decreases in the following order with a concomitant increase of linear poly( phenylacetylene) formation: z(Ph,P),NiCl, >>( PPh,),PdCl, The same sequence would thus appear to obtain fctr the crosslinking reactions occurring in the cure of 1, linear polyene structures predominating in the case of (PPh,),PdCl, and sub- stantial cyclotrimerization taking place with (y -Cp)Co( CO),.It is to be expected that these fundamental differences in chemical structure will affect the physical properties of the cured resins. In particular, resins cured in the presence of (y -Cp)Co( CO), should contain highly stable benzene crosslink sites.However, overall thermal stability of the resin may depend on the structure of the monomer itself rather than that of the crosslinks, as indeed is the case for the resins obtained by catalysed cure of l.5 Finally, note that although the catalysed polymerization of phenylacetylene is nearly complete for nickelocene and (PPh3),PdC1,, ca. 20% of the monomer remains unreacted in the cases of (y-Cp)Co(CO), and (Ph,P),NiCl, [for the (Ph,P),NiCl,-catalysed polymerization conducted in refl uxing benzene, as much as 40% of the phenylacetylene was recovered ~nchanged~~,~’].Therefore, for resins produced from 1 in the presence of the last two catalysts, at least 20% of the crosslinks may be expected to be of the linear polyene type resulting from the uncatalysed reaction of the remaining ace1 ylene groups in the post-cure stage.We thank the Royal Borough of Kingston-upon-Thamcs for a Research Assistantship awarded to A.S.O. References 1 C. Y-C. Lee, in Developments in Reinforced Plastics, ed. G. Pritchard, Elsevier, Barking, 1986, vol. V, pp. 121-150. 2 P. M. Hergenrother, in Encyclopaedia of Polymer Sciencx and Engineering, ed. H. F. Mark, N. M. Bikales, C. G. Overherger, G. Menges and J. I. Kroschwitz, Wiley, New York, 2nd edn. 1985, VO~.1, pp. 61-86. 3 V. A. Sergeev, Yu. A. Chernomordik and A. S. Kurapoj, Usp. Khim., 1984,53,518 (Russ. Chem. Rev., 1984,53,307). 4 W. E. Douglas and A. S. Overend, J.Muter. Chem., 1993,3,1019. 5 W. E. Douglas and A. S. Overend, Eur. Polym. J., 1993,29, 1513. 6 W. E. Douglas and A. S. Overend, J. Organomet. Chem.. 1993, 444, C62. 7 C. Y-C. Lee, 1. J. Goldfarb, T. E. Helminiak and F. E. Arnold, Natl. SAMPE Symp. Exhib. (Proc.) 28th (Muter. Pro( esses-Contin. Innovations), 1983,699. 8 M. D. Sefcik, E. 0. Stejskal, R. A. McKay and J. Scliaefer, Macromolecules, 1979,12,423. 1172 J. MATER. CHEM., 1994, VOL. 4 9 10 11 12 13 14 15 16 17 S. A. Swanson, W. W. Fleming and D. C. Hofer, Macromolecules, 1992.25, 582. J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke, Principles and Applicutions of Orgunotransition Metal Chemistry, University Science Books, Mill Valley, 2nd edn., 1987. L. G.Picklesimer, M. A. Lucarelli, W. B. Jones, T. E. Helminiak and C. C. Kang, Polym. Prep., 1981, 22, 97; M. A. Lucarelli, W. B. Jones Jr., L. G. Picklesimer and T. E. Helminiak, ACS Symp. Ser., 1982, 195, 237. T. Takeichi and J. K. Stille, Macromolecules, 1986, 19, 2093; 2108. W. E. Douglas and A. S. Overend, J. Organomet. Chem., 1986, 308, C14. V. A. Sergeyev, V. K. Shitikov, A. S. Kurapov and I. P. Antonova- Antipova, Vys. Soyed., 1989, A31, 1188 (Polym. Sci. USSR, 1989, 31,1300). A. P. Pigneri and R. S. Bauer, US Put. Appl., 1989,386,083(Chem. Abs., 1991, 115, 30129J). L. S. Meriwether, E. C. Colthup, G. W. Kennerly and R. N. Reusch, J. Org. Chem., 1961,26,5155. A. F. Donda, E. Cervone and M. A. Biancifiori, Recueil, 1962, 81, 585. 20 21 22 23 24 25 26 27 28 29 30 31 J. L. Koenig, Report AFWAL-TR-83-4064, 1983 (Chern. Ah., 1984,100,157487~). Sadtler Handbook of NMR Spectra, Sadtler Research Laboratories, 1978, (a) spectrum no. 8599M; (b) spectrum no. 5082M. K. P. C. Vollhardt, Angew, Chem., Int. Ed. Engl.. 1984,23, 539. H. Brunner and W. Pieronczyk, Bull. Soc. Chinz. Belg., 1977, 86, 725, and references therein. M. Dubeck, J. Am. Chem. Soc., 1960,82,502. M. Dubeck and A. H. Filbey, US Patent, 3256260 (Chem. Abs., 1966,65, 7307a). J. L. Davidson and D. W. A. Sharp, J. Chem. Soc., Dalton Trans., 1976,1123. V. 0.Reikhsfel'd, B. I. Lein and K. L. Makovetskii, Dokl. Akad. Nauk SSSR, 1970,190,125 (Proc.Acud. Sci.USSR, 1970,190,31). J. F. Tilney-Bassett, J. Chem. Soc., 1961, 577. A. Furlani, P. Bicev, M. V. Russo and M. Fiorentino, Gazz. Chim. Ital., 1977, 107, 373. G. Sartori, A. Furlani, P. Bicev, P. Carusi and M. V. Russo, Isr. J. Chem., 1976177, 15, 230. D. L. Trumbo and C. S. Marvel, J. Polym. Sci., Polym. Chem. Ed., 1987,25,1027. 18 R. C. Doss and P. W. Solomon, J. Org. Chem., 1964,29, 1567. 32 E. C. Colthup and L. S. Meriwether, J, Org. ChtJm., 1961,26,5169. 19 C. I. Simionescu, V. Percec and S. Dumitrescu, J. Polym. Sci., Polym. Chem. Ed., 1977, 15,2497. Paper 4/0034lA; Received 19rh Junuary,1994