9 Kinetics and Mechanism of Polymerisation By K. J. IVlN The Queen’s University of Belfast Belfast B T9 5AG IT IS six years’ since Annual Reports contained a section dealing specifically with polymerisation and since that time perhaps 5000 papers have been published on the subject of this report. Here we can do no more than to refer to a few hundred of these mostly published in the last 18 months. We shall not deal specifically with solid-state polymerisation nor with grafting. In the last few years there has been a spate of new review series aimed at keeping the polymer chemist abreast of his subject.2 The publication of symposia on specific topics such as analytical gel permeation chromatography the com-puter in polymer science and block copolymers has been particularly ~ s e f u l .~ ‘Ablative’ polymers provide the protective heat shield when space capsules re-enter the earth’s atmosphere ; the first published symposium4 on this subject contains a very interesting historical account of the development of these materials. Papers presented at the IUPAC Macromolecular Symposia in Prague (1965) Tokyo-Kyoto (1966) and Brussels-Louvain (1967) have now appeared ; 5 also the plenary and main lectures of the 1967 meeting.6 A new series of quarterly reports on polymers giving a critical appraisal of those papers which are considered the most significant should prove very help-f ~ l . ~ The publication of a Polymer Handbook has also filled a long-felt need.7a General.-Reference has already been made to the symposium on gel permeation chr~matography.~ This technique has now advanced to the where C.H. Bamford and G. C. Eastmond Ann. Reports 1963 91. “Progress in High Polymers’ ed. J. C. Robb and F. W. Peaker Heywood London, vol. 1 1961 ; ‘Advances in Macromolecular Chemistry’ ed. W. M. Pasika Academic Press London vol. 1 1968 ; Advances in Polymer Science (Fortschritte der Hochpoly-meren-Forschung) Springer-Verlag Berlin vol. I 1958 ; ‘Reviews in Macromolecular Chemistry’ ed. G. B. Butler and K. F. O’Driscoll Marcel Dekker New York vol. 1 , 1967. J. Polymer Sci. Part C Polymer Symposia Nos. 21 25 26. J . Macromol. Sci. 1969 A3 326. J . Polymer Sci. Part C Polymer Symposia Nos. 16 23 and 22 respectively. ’ International Symposium on Macromolecular Chemistry Brussels-Louvain 1967, Plenary and Main Lectures Butterworth London 1969.Quarterly Literature Reports Polymers Kogan Page (Volume I No. 1 covered the literature published 0ct.-Dec. 1968). 7 “ Polymer Handbook ed. J. Brandrup and E. H. Immergut Interscience New York, 1966. R. E. Jentoft and T. H. Gouw J. Polymer Sci. Part B Polymer Letters 1969 7 81 1 . * W. Heitz B. Bomer and H. Ullmer Makromol. Chem. 1969 121 102 122 K . J. Ivin distinct peaks can be obtained for oligomers up to a degree of polymerisation (DP) of 15 and beyond. Furthermore by the use of partitioning substrates which are bonded to the solid support and by operating under supercritical fluid condi-tions the whole operation can be speeded up by a factor of 10. Thus 20 mg of a supposedly monodisperse low molecular weight polystyrene (E = 6) was re-solved into 18 components in 60 min using 5 % methanol in n-pentane at 205 "C as the e l ~ t a n t .~ Isotactic and syndiotactic poly(methy1 methacrylate) have been separated by thin layer chromatography using silica as stationary phase." The solvent is very critical out of four solvents tried only ethyl acetate gave widely different elution rates. Isotactic and syndiotactic polymers appear to form a 1 1 complex which is only broken down in ethyl acetate. Racemic monomers in the presence of suitable catalysts may give rise to racemic polymers which can then be separated on a column of optically active material. One case where this has been achieved" is the polymerisation of dl-l-methyl-propyl vinyl ether catalysed by Al(OPri)3-H2S04. Partial resolution was effected by elution on a column containing the linear polymer of L-lactide, f OCH(CH,)CO+.N.m.r. provides detailed information about the structural sequences of mono-mer units in polymer chains and hence about the relative rates of competing propagation processes (see below). In the past such information could only be obtained for soluble polymers but a pulse method has now been developed which will give a spectrum for a solid sample which is comparable in appearance with that obtained from a solution in the conventional way.' An especially interesting new group of polyelectrolytes has been prepared by Rembaum. l3 These are formed by the Menschutkin reaction of di-t-amines with dihalides and have been termed ionenes for example : If the reactants are taken in stoicheiometric proportions the reaction is first order when the solvent is dimethylformamide (DMF) and second order when the solvent is 20% H20-80 % DMF.The rate increases with the polarity of the solvent (MeOH < MeOH-DMF < H,O-DMF) but is practica!ly independent of the number of methylene groups in the reagents. Bromides react more rapidly than chlorides. Molecular weights of up to 40,000 have been achieved and the products can be combined with negative polyelectrolytes such as poly(styrene sulphonic acid) to form a variety of membranes. l o H. Inagaki T. Miyamoto and F. Kamiyama J . Polymer Sci. Part B Polymer Letters, l 1 E. Chiellini G. Montagnoli and P. Pino J . Polymer Sci. Part B Polymer Letters 1969, l 2 D. Ellett U. Haeberlen and J.S. Waugh J. Polymer Sci. Part B Polymer Letters, l 3 A. Rembaum J. Macromol. Sci. 1969 A3 87. 1969 7 329. 7 121. 1969 7 71 Kinetics and Mechanism of Polymerisation 123 It has been generally believed that interfacial polymerisation e.g. between sebacoyl chloride and hexamethylenediamine occurs in a thin region of the oil phase close to the interface and that the interface has no specific effect other than permitting a controlled diffusion of water-soluble monomer into the oil phase and removing by-product acid from the reaction zone. However this model explains neither the rapidity of the reaction compared with that in solution nor the high molecular weight and narrow distribution of the polymer. Measure-ments of interfacial pressures lead to the conclusion that polymerisation probably occurs in a mixed monolayer of adsorbed monomers.' One of the assumptions of the Smith-Ewart theory of emulsion polymerisation is that the number of particles per unit volume of latex N remains constant during the constant-rate period. This assumption has been tested by electron-microscope observations on latex particles which have been embedded in a poly(viny1 alcohol) layer formed at an air-water interface and in fact N increases during the constant-rate period.' There have been comparatively few investigations of the kinetics of poly-condensation reactions. In the reaction between aliphatic diols and aliphatic dicarboxylic acids the rate increases considerably with the chain length of the diol but not with that of the acid.16 Free-radical Polymerisation.-There have been numerous investigations on the kinetics of free-radical polymerisation of vinyl compounds such as methyl methacrylate (MMA) styrene (Sty) and acrylonitrile (AN).These have frequently been aimed at elucidating the effectiveness and mode of action of catalyst systems, such as iron(n) bromate l 7 (MMA) &a'-azobisisobutyronitrile (MMA) aroma-tic halides with silver or mercury '' N-bromosuccinimide with reduced nickel2' (MMA) tetramethyltetrazene with benzyl chloride2' (AN) Fe" acaczZ2 (Sty etc.), Cu" acac2 with ammonium trichl~roacetate~~ (MMA) bis( - )ephedrine CU" witb cc1424 (Sty) Co"(CN) and Co"'(CN),(PhCH2) with organic halides25 (MMA Sty AN etc.) various tin compounds,26 dimethylaniline oxide with metal salts such as cobalt(@ nitrate2' (MMA) trichloroacetyl chloride with water2* (MMA) triethylboron with di-t-butyl peroxide2' (MMA) cerium(1v) salts with l 4 F.MacRitchie Trans. Furuduy SOC. 1969 65 2503. l 5 I. D. Robb J. Polymer Sci. Purr A-1 Polymer Chem. 1969 7 417. l 6 E. Makay-Bodi and I. Vancso-Szmercsanyi European Polymer J. 1969 5 45. I ' D. Pramanick and S. R. Palit Kolloid-Z. 1969 229 24. l 9 M. Kinoshita N . Yoshizumi and M. Imoto Makromol. Chem. 1969,127 185. 2 o T. Otsu and M. Yamaguchi J. Mucromol. Sci. 1969 A3 177. 2 1 T. Nakaya Y. Maki and M. Imoto Mukromol. Chem. 1969 125 161. 2 2 P. E. M. Allen and T. H. Goh European Polymer J. 1969,5 419. 2 3 P. E. M. Allen and T. H. Goh European Polymer J. 1969 5 335. 2 J J . Barton and M. Lazar Makromol.Chem. 1969 124 38. 2 5 S. Aoki C. Shirafuji and T. Otsu Mukromol. Chem. 1969 126 1 . 2 6 S. Aoki C. Shirafuji Y. Kusuki and T. Otsu Mukromol. Chem. 1969 126 8. 2 i T. Sat0 and T. Otsu Makromol. Chem. 1969 125 1 . 2 8 N . Sakota H. Nakamura and K. Nishihara Mukromol. Chem. 1969 129 47. 2 q I . Contreras J. Grotewold E. A. Lissi and R. Rozas J. Polymer Sci. Part A-1 Polymer W. Vogt and L. Dulog Mukromol. Chem. 1969 122 223. Chem. 1969,7,2341 124 K . J. Ivin pinaco13' (MMA) and the donor-acceptor system poly(2-vinylpyridine)-sulphur dioxide3 ' (MMA). In the last case it was shown that the derived values of kp/k,3 were consistent with free radical propagation. [ ''C]Tetramethylthiuram monosulphide can act as both initiator and re-tarder in the radical polymerisation of styrene; the system is very sensitive to light.32 The effect of various metal salts has been studied for the radical poly-merisation of ~inylpyridines,~~ N-t-b~tylacrylamide,~ N-vinyl-i m i d a ~ o l e ~ ~ methyl metha~rylate,~ and a~rylonitrile.~~ For N-t-butylacryla-mide in methanol at 25 "C the rate of polymerisation is inversely proportional to the concentration of iron(rrr) chloride indicating linear termination.The rate constant for propagation relative to that for termination by iron(1rr) chloride is considerably smaller than for the polymerisation of acrylamide in water pre-sumably as a result of the steric effect of the t-butyl s ~ b s t i t u e n t . ~ ~ 2,6-Dichlorophenol indophenol is a powerful retarder in both its dissociated and undissociated forms for the polymerisation of styrene (k = 17,700 1 mol-' s-') and methyl methacrylate (k = 8200 1 mol- ' s- ') at 60 "C; between one and two radicals are removed for each dye molecule consumed.39 lY3,5-Triphenyl-verdazyl (1) absorbs at 720nm and is superior to diphenylpicrylhydrazyl as a scavenger for polymer radicals in that the products do not react with further Ph radical^.^' I Solvent effects have been studied for the polymerisation of methyl methacry-1ate41v42 (in PhCl and PhBr) vinyl acetate43 (in HOAc) vinyl chloride44 (in PhCl) and a~rylonitrile~~ (in ethylene carbonate and dimethyl formamide).The fractional rate of polymerisation of methyl methacrylate is increased in the presence of aromatic halides the effect increasing in the order PhF < PhCl < PhBr < PhI.This has been interpreted in two ways first that the halide (S) 'O H. Narita S. Okamoto and S. Machida Makromol. Chem. 1969 125 IS. 3 1 M. Matsuda and Y. Ishioroshi Makromol. Chem. 1969 126 16. 3 2 J. C. Bevington and F. S. Rankin European Polymer J . 1969 5 437. 3 3 N . N . Dass and M. H. George J . Polymer Sci. Part A - I Polymer Chem. 1969,7 269. 34 S. Tazuke K. Shimada and S. Okamura J . Polymer Sci. Part A - I Polymer Chem., 35 E. A. S. Cave11 and I. T. Gibson J . Polymer Sci. Part A - I Polymer Chem. 1969,7 1307. 3 6 S . Tazuke and S. Okamura J . Polymer Sci. Part A - I Polymer Chem. 1969 7 851. 3 7 F. D. Williams J . Macromol. Sci. 1968 A2 459. 3 8 R. G . Jones Polymer 1969 10 89. 39 I . Kar B. M. Mandal and S. R. Palit Makromol.Chem. 1969,127 195. 40 M. Kinoshita and Y . Miura Makromol. Chem. 1969 124 21 1 . 4 1 G. M. Burnett G. G. Cameron and B. M. Parker European Polymer J. 1969 5 231. 4 2 W. I. Bengough N . K. Henderson and D . Patsavoudis European Polymer J . 1969, 43 S. P. Potnis and A. M. Deshpande Makromol. Chem. 1969 125 48. 44 G. S. Park and D. G. Smith Trans. Faraday SOC. 1969,65 1854. 4 5 G. Vidotto A. Crosato-Arnaldi and G. Talamini Makromol. Chem. 1969 122 91. 1969 7 879. 5 463 Kinetics and Mechanism of Polymerisation 125 catalyses the initiation of chains through the sequence R. + S-+RS. RS. + M - + P - this being more efficient than R. + M-+P. (R. = primary radical, M = monomer) ; second that polymer radicals form charge-transfer complexes with both M and S and addition of S disturbs the propagation The latter theory predicts that the effect should be independent of catalyst.In fact benzoyl peroxide and &a'-azobisisobutyronitrile give very similar results for PhCl and PhBr as solvent but not for PhI.41 Other work4' indicates that not only k but also k is affected the latter by the viscosity of the medium whereas the initiation rate is not affected. Related effects are observed in the homogeneous polymerisation of vinyl chloride in 1,2-dichloroethane catalysed by a,a'-azobisisobutyronitrile. The rate is increased by the addition of small amounts of substances such as CBr, or C12Hz5SH.48 This is attributed to the sequence of reactions P- + M-+ X + Q. Q. + CBr -+ QBr + cBr3 the radical Q. formed by chain transfer to monomer being a less efficient initiator of new chains than cBr3.For the homo-geneous polymerisation of vinyl chloride in PhCl the catalyst exponent is 0-56 which suggests the participation of a degradative reaction with the solvent. The molecular weight data favour copolymerisation rather than chain transfer as the degradative reaction., The polymerisations of 0- rn- and p-hydroxy-styrenes give catalyst exponents of 0.72 0-52 and 0.50 re~pectively.~~ The high value for the o-compound is attributed to the ease of transfer of the hydrogen atom from the o-hydroxy group to the carbon radical resulting in an unreactive radical. Bromobenzene acts as a degradative chain transfer agent in the poly-merisation of vinyl acetate.42 Chain transfer agents for which quantitative measurements have been made include methyl oleate and stearate5' (the former is always 5.8 times more reactive towards both styrene and methyl methacrylate radicals at 60°C) the series SiCl,Me (n + rn = 4),51 the series HCF,(CF,),-,CH,OH ( n = 2 4 6),52 sub-stituted ben~aldehydes~~ and p-substituted ~umenes.~ In the last two cases the results have been correlated by means of the modified Hammett equation, log kfko = pa + y E R where yER represents the resonance term.The effects of p-substituents in the substrate cumene depend on contributions from both polar and resonance (1) 46 G . Henrici-Olive and S. Olive Z. phys. Chem. (Frankfurt) 1965 47 286; 1966,48 35, 4 ' C. H . Bamford and S. Brumby Makromol. Chem. 1967 105 122. 4M J. W. Brietenbach 0. F. Olaj H.Reif and A. Schindler Makromol. Chem. 1969, 49 M. Kato J . Polymer Sci. Part A - I Polymer Chem. 1969 7 2175. 5 0 E. F. Jordan B. Artymyshyn and A. N. Wrigley J . Polymer Sci. Part A-1 Polymer 5 ' Y. Minoura and H. Toshima J. Polymer Sci. Part A-1 Polymer Chem. 1969 7 2837. 5 2 I. Kar B. M. Mandal and S. R. Palit J . Polymer Sci. Part A - I Polymer Chem. 1969, 53 T. Yamamoto M. Hasegawa and T. Otsu Bull. Chem. SOC. Japan 1969 42 1364. 5 4 T. Yamamoto and T. Otsu J . Polymer Sci. Part A-1 Polymer Chem. 1969 7 1279. 51 ; Makromol. Chem. 1966,96 221. 122 5 1 . Chem. 1969,7 2605. 7 2829 126 K. J. Ivin factors but the effects of p-substituents in the attacking polystyryl radical can be accounted for solely in terms of polar End-group analyses of the pro-ducts of polymerisation of ethyl acrylate in acetic and n-butyric acids show that the solvent radicals formed as a result of chain transfer are not very efficient at initiating new chains.In the cyclopolymerisation of diallyl cyanamide the catalyst exponent for t-butyl benzoate as catalyst at 80 "C is reduced from 1.0 at normal pressures to 0.5 at 3000 atm,56 indicating a change with pressure in the predominant termina-tion mechanism. The absorption spectra of certain diallyl compounds show that there can be chromophoric interactions even when the double bonds are un-conjugated ; the structures which depict these interactions resemble the transi-tion states postulated for intra-inter-cyclopolymerisation. 5 7 The successful cyclopolymerisation of ally1 methacrylate is reported.* Progress has been made towards a better understanding of the mechanism of thermal initiation of the polymerisation of The principal dimeric products at 130 "C are cis- and trans-l,2-diphenylcyclobutane with relatively minor amounts of 1-phenyltetralin 1,3-diphenylbutene and l-phenyl-1,2-dihydronaphthalene. Mayo's basic postulate of radical-pair production by reaction of a Diels-Alder type dimer with styrene is the starting point of a more extended reaction scheme in which cage reactions form minor dimeric and trimeric products in competition with chain i n i t i a t i ~ n . ~ ~ Ring-fluorinated styrene C6F5CH=CH2 also undergoes thermal initiation.62 When poly(viny1 trichloroacetate) is dissolved in a monomer such as styrene, and irradiated in the presence of Mn,(CO),,- as photosensitiser all the initiating radicals are attached to the pre-p~lymer.~~ The growing side-chains of poly-styryl radicals combine in pairs and the system eventually gels.By comparing gel times of eight other monomers with that for styrene under conditions of known rate of initiation and allowing for chain transfer where necessary,64 a direct estimate can be made of the ratio of disproportionation to combination for each type of radical. Contrary to earlier conclusions it is found that both poly(methy1 acrylate) and poly(viny1 acetate) radicals combine exclusively. 1,l-Disubstituted monomers in general give radicals which tend to dispropor-tionate rather than combine. When nickel derivatives are used as initiators some unattached radicals are formed in addition to those attached to the pre-PO 1 y mer .5 5 P. V. T. Raghuram and U. S. Nandi J . Polymer Sci. Part A - I Polymer Chem. 1969, 5 6 J. P. J. Higgins and K. E. Weale. J . Polymer Sci. Part B Polymer Letters 1969,7 153. " G. B. Butler and B. Iachia J . Macromol. Sci. 1969 A3 803. " J. P. J. Higgins and K. E. Weale J . Polymer Sci. Part A - I Polymer Chem. 1968,6,3007. s 9 W. G. Brown Makromol. Chem. 1969 128 130. " K. Kirchner Makromol. Chem. 1969 128 150. 6 1 K. R. Kopecky and S. Evani Canad. J . Chem. 1969,47,4049. 6 2 W. A. Pryor and T-L. Huang Macromolecules 1969 2 70. 6 3 C. H. Barnford R. W. Dyson and G . C. Eastmond Polymer 1969 10 885. '4 C. H. Bamford R. W. Dyson G. C. Eastmond and D. Whittle Polymer 1969,10,759. " C . H.Bamford G. C. Eastmond and D. Whittle Polymer 1969 10 771. 7 2379 Kinetics and Mechanism of Polymerisation 127 A very neat experimental verification of Benson and North's theory66 of diffu-sion control in the mutual termination of polymer radicals has been d e ~ c r i b e d . ~ ~ In the case of two radicals of equal degree of polymerisation P the theory leads to equation (2) for the termination rate constant k,, K K k =-+- ' PLb Pb where K is a constant and PL corresponds to the value of P at which the segmental mobility becomes independent of P . The exponent b relates P to the hydro-dynamic radius Y (r cc Pb) and is given by (1 + a)/3 where a is the exponent of the intrinsic viscosity equation ([q] cc Pa). Experimentally fraction of poly(ethy1ene oxide) having P between 2 and 3000 are dissolved in water and submitted to pulse radiolysis.Hydroxyl radicals are generated and abstract atoms from the polymer chains. The decay of the polymer radicals so formed is then followed by rapid recording of the absorbance at 235 nm. For this system a = 0.60 giving b = 0.53. In accordance with equation (2) a plot of k us. P-0'53 is a good straight line the slope and intercept giving PL = 1300 rather higher than the arbitrary value of 100 assumed by Benson and North. The theory of diffusion control of k is further discussed by Itqb8 and experiments on solvent-viscosity effects have been d e s ~ r i b e d . ~ ~ . ~ ~ The gel effect in vinyl polymerisation has been investigated by differential scanning ~alorimetry.~' The generally accepted view that it is governed by diffusion effects is confirmed but a new inflection in the rate curve is reported and explained in terms of free-volume theory.Anionic Polymerisation of Vinyl Monomers.-Great strides have been made in the understanding of the kinetics of anionic polymerisation and related processes in the last six years particularly through the works of Szwarc and Schulz. An excellent summary can be found in Szwarc's book7' published in 1968. A very wide range of rates is observed in the anionic polymerisation of a given monomer which can be interpreted in terms of the following equilibria : (P-Na') P-Na' P - IINa' P - + Na+ ion pair contact (tight) solvent-separated (loose) free dimer etc. (n) ion pair (c) ion pair(s) anion (-) P- denotes the polymer anion which is usually associated with a cation such as Na+ either in intimate contact or separated from P- by a solvent shell.The 6 6 S. W. Benson and A. M. North J . Amer. Chem. SOC. 1962 84 935. 6 7 U. Borgwardt W. Schnabel and A. Henglein Makromol. Chem. 1969 127 176. '' K. Ito J . Polymer Sci. Part A - I Polymer Chem. 1969 7 827 2247 2707 2995. 6 9 K. Yokota and M. Itoh J . Polymer Sci. Part 8 Polymer Letters 1968 6 825. 70 K. Horie I. Mita and H. Kambe J . Polymer Sci. Part A-1 Polymer Chem. 1968 6, ' ' M. Szwarc 'Carbanions Living Polymers and Electron-transfer Processes' Inter-2663. science New York 1968 128 K. J . Ivin equilibria shift to the right with increasing solvent polarity benzene z dioxan < tetrahydropyran (THP) < methyltetrahydrofuran (MTHF) < tetrahydrofuran (THF) < 1,2-dimethoxyethane (DME) < diglyme < triglyme < tetraglyme.This shift and the facts that (a) the rate constants for propagation of the different species lie in the order k (zero) << k << k < k - ; (b) the rate of propagation by contact ion pairs increases with size of cation Li+ < Na+ < IS+ < Rb+ z Cs' ; and (c) the equilibria shift to the right with decreasing temperature (exothermic dissociation) are responsible for the great diversity of behaviour in these systems. It is not usual to have all four of the above species present simultaneously in any one system. In benzene and dioxan for example the free-ion contribution to polymerisation is negligible and association of the ion pairs in hydrocarbon solvents becomes less significant as the cation size is increased.72 With lithium as cation association can be exceptionally strong both for initiator (n = 6 ) and for the polymer ion pair (usually n = 2)73 leading to a relatively slow rate of initiation.In such cases it is politic to employ a pre-initiation technique in order to measure propagation constants. An interesting study of the variation of n with catalyst and solvent has been published.74 At the other end of the polarity spectrum it is possible to have systems in which propagation is thought to proceed exclusively through free ions." In solvents such as THP and THF it is possible to suppress the dissociation into free anions by the addition of electrolytes which are stronger than the polymer ion pair for example sodium tetraphenylborate.In the absence of added electrolyte the apparent second-order rate constant k, defined as RJLE] [MI where R is the rate of polymerisation and [LEI is the so-called (total) living end concentration, is given by equation (3) provided that the concentration of free ions is very small compared with that of ion pairs. k, = k* + k-K*[LE]-* (3) It has been shown in a number of cases that as required by equation (3) k is a linear function of [LEI-*. The rate constant for ion pairs k * may be determined from the intercept while k- can be found from the slope if K the electrolytic dissociation constant of the ion pairs has been determined from conductance measurements. In the presence of added electrolyte the contribution of free ions becomes very small compared with k* and a more reliable value for k* can then be obtained.On plotting log k against T-' there is sometimes a marked departure from the usual straight-line Arrhenius plot to such an extent that there may be a maximum as in the systems styrene-THF-Na+,76 a-methylstyrene-THF-Na+,7 72 J. E. L. Roovers and S. Bywater Trans. Faraday SOC. 1966 62 701. 7 3 H. S. Makowski M. Lynn A. N. Bogard J . Macromol. Sci. 1968 A2 665 683. 7 4 J . G. Carpenter A. G. Evans C. R. Gore and N. H. Rees J . Chem. SOC. (B) 1969,608. 7 5 C. E. H. Bawn A. Ledwith and N. McFarlane Polymer 1969,10 653. '' T. Shimomura K. J. Tolle J . Smid and M. Szwarc J . Amer. Chem. SOC. 1967 89 796. " F. S. Dainton G. A. Harpell and K. J. Ivin European Polymer J.1969,5 395 Kinetics and Mechanism of Polymerisation 129 styrene-DME-Na',78 and a-methylstyrene-THP-Li' .79 The reason for this is the shift in equilibrium from the more reactive solvent-separated ion pairs at low temperature to the less reactive contact ion pairs at high temperature. Kinetically the most well-characterised system from this point of view is styrene-THP-Na+ where experiments have been carried out over an exceptionally wide range of temperature (50 to - 40 "C). In this case the Arrhenius plot is sigmoidal approach-ing an asymptote at high temperature corresponding to kc.80 A summary of rate constants is given in Table 1. Small quantities of solvating agents such as triglyme and tetraglyme exert a powerful catalytic effect when added to less polar solvents.81 From the variation of the rate of polymerisation with concentration of additive (G) it is possible to deduce the number of molecules rn taking part in the solvation process as well as the equilibrium constant for solvation and the propagation rate constant k, for the glymated species (see Table 1).P-Na' = rnG P-rnGNa' for styrene-THP-Na+ at 25" rn = 1 and k is ca. 300 times k, but still an order of magnitude less than k for THF or DME. The activation energy for k is unexpectedly low and to explain this it has been suggested that there is an equilibrium between two types of glymated ion pair.81 The general notion of two types of ion pair in ether solvents has been amply confirmed by a wealth of physical evidence from non-polymer systems.As recent examples we may cite (a) the effect of temperature solvent and solvating agents on the absorption spectra of the metal f l ~ o r e n y l s ~ ~ . ~ ~ and alkali 43-methylenephenanthrenides ;84 (b) the effect of temperature and cation on the electrolytic dissociation constant of alkali anthracenides ;85 (c) the effect of tem-perature solvent and solvating agents on the alkali-metal splitting constant in the e.s.r. spectra of alkali naphthalenides and anthracenides,86 including evidence for two types of glymated ion pair ;87 and (d) the effect of cation temperature and solvent on the n.m.r. spectrum of the fluorenyl anion.88 Physical evidence for two types of ion pair in polymerising systems is harder to come by. However, there is distinct evidence from absorption spectra in the system or-methylstyrene-THF-Na+ though not in the corresponding styrene system.89 The varia-tion with temperature of the electrolytic dissociation constant of the alkali 7 8 T.Shimomura J. Smid and M. Szwarc J . Amer. Chem. SOC. 1967,89 5743. l9 F. S. Dainton K. M. Hui and K. J. Ivin European Polymer J. 1969 5 387. L. Bohm W. K. R. Barnikol and G . V. Schulz Makromol. Chem. 1967,110 222. 8 1 M. Shinohara J. Smid and M. Szwarc J . Amer. Chem. SOC. 1968 90 2175 Chem. Comm. 1969 1232. 8 2 T. E. Hogen-Esch and J. Smid J . Amer. Chem. SOC. 1969 91,4580. 8 3 T. Ellingsen and J. Smid J. Phys. Chem. 1969 73 2712. 84 D. Casson and B. J. Tabner J . Chem. SOC. ( B ) 1969 572. 8 5 D. Nicholls C. Sutphen and M. Szwarc J . Phys. Chem. 1968,72 1021.8 6 N. Hirota J . Amer. Chem. SOC. 1968 90 3603. " K. Hofelmann J. Jagur-Grodzinski and M. Szwarc J . Amer. Chem. SOC. 1969 91, 8 8 R. H. Cox J . Phys. Chem. 1969 73 2649. 8 9 J. Comyn and K. J. Ivin European Polymer J. 1969 5 587. 4645 System Styrene-THF-Na + Styrene-THP-Na +-tetraglyme Styrene-DME-Na + a-methylst yrene-THF-Na+ a-methylstyrene-THF-K +-trigly me Table 1. Arrhenius parameters for anionic polymerisation Rate constant (1 mol- ' s- ') at 25 "C Activation energy (kcal mol- ') k c k S k - k E Es E - E Refs. 30,000 65,000 a > 4000 d > 4700 e 12 55* (65,000) 3900 8.0 5.5 5.9 1.2 b c 200+ 830t 6.7 7.2 f g 120' 6.9 f * Now believed to be much higher (private communication from the authorsb). t Extrapolated from measurements at lower temperatures.Refs. T. Shimomura K. J. Tolle J. Smid and M. Szwarc J. Amer. Chem. SOC. 1967,89,796; L. Bohm W. K. R. Barnikol and G. V. Schulz, Makromol. Chem. 1967,110,222; M. Shinohara J. Smid and M. Szwarc Chem. Comm. 1969 1232; T. Shimomura J. Smid and M. Szwarc, J. Amer. Chem. SOC. 1967 89 5743; G. Lohr and G. V. Schulz Makromol. Chem. 1968 117 283; f J . M. Ginn and K. J. Ivin unpublished results; J. Comyn F. S. Dainton G. A. Harpell K. M. Hui and K. J. Ivin J. Polymer Sci. Part B Polymer Letters 1967 5 965 Kinetics and Mechanism of Polymerisation 131 a-methylstyrene oligomers has been measured both in THF and THP as solvents ; with the Li salts there are indications of a change in the type of ion pair with tem-perature.” The heat change AH for the conversion ofcontact ion pairs to solvent-separated ion pairs is generally in the range - 3 to - 7 kcal mol- ’ for a variety of ion pairs in ethereal solvents.The increase of k with increasing cation radius and the much higher value of k may be interpreted in terms of the decreasing amount of electrostatic work required to form the transition state. These variations are reflected in the magni-tude of the activation energies. In a ‘living’ polymerisation system in which the polymerisation is propagated by a single species the resulting molecular-weight distribution has a Poisson form provided that all chains are initiated simultaneously are allowed to grow for the same length of time and are terminated simultaneously by the addition of a suitable reagent. In a flow system these conditions are not precisely met unless the flow is turbulent.” If polymerisation is propagated by two species with different rate constants such as ion pairs and free ions the distribution is broadened.From the extent of broadening the rate constants for interconversion of the two species can be determined.92 For styrene-THF-Na’ the relaxation time for the interconversion of ion pairs and free ions is of the order of a few microseconds at the concentrations normally used. A common initiator for anionic polymerisation is sodium naphthalenide in THF. However it is troublesome in that it is not stable and must be freshly prepared. It has now been shown that two species are present in such solutions, both of which are effective as initiators for styrene in the absence of NaBPh but one of which is completely deactivated in the presence of NaBPh4.93 The inactive product has an extinction coefficient which is nearly the same as that of the living ends at 350nm.This explains the earlier discrepancies between values of k* obtained by Szwarc and Schulz for styrene-THF-Na’. The correct value at 20 “C appears to be 180 1 mol-’ s-’. One of the basic problems in the kinetics of anionic polymerisation is the accurate determination of the living end concentration at levels of 10- to M. Chemical methods are not easy because of the necessity to preclude traces of air and water. This problem has been overcome by the electrochemical generation of living ends whose concentration may be determined from Faraday’s law.94 Electrochemically generated PhNOz - ions initiate the polymerisation of acrylonitrile (AN) but not of MMA or ~tyrene.~’ Initiation is a relatively slow process and occurs by proton transfer from AN to PhN02-.For a series of sub-stituted nitrobenzene ions the rate constant for the initiation process shows a 90 J. Comyn F. S. Dainton and K. J. Ivin European Polymer J . 1970 6 349. 9 1 G. Lohr and G. V. Schulz Z . phys. Chem. (Frankfurt) 1969 65 170. 92 R. V. Figini G. Lohr and G. V. Schulz J . Polymer Sci. Part B Polymer Letters 1965, 93 B. J. Schmitt and G. V. Schulz Makromol. Chem. 1969,121 184. 94 B. L. Funt S. N. Bhadani and D. Richardson J . Polymer Sci. Part A - I Polymer Chem., 9 5 G . Mengoli G. Farina and E. Vianello European Polymer J . 1969 5. 61. 3 985.1966 4 287 I 132 K . J. Ivin correlation with the Hammett Q value. The subject of electrolytically initiated polymerisation is reviewed by Y a m a ~ a k i . ~ ~ The polymerisation of AN has been studied with toluene as solvent at - 75 "C, using a Grignard initiator in conjunction with catalytic amounts of DMSO hexa-methylphosphoramide (HMPA) or DMF.97 Two effects can be detected an increase in the number of growing ends (for HMPA) and an increase in the propagation rate constant (for DMSO) ; for DMF both effects are important. With Bu,Mg BuMgC1 or Bu3Mg21 as initiator the kinetics indicate that neither termination nor transfer processes are significant but whereas the first two ini-tiators give a polymer with a narrow molecular weight distribution Bu3Mg21 gives a polymer with a broad distribution suggesting the presence of two types of growing species.98 In the polymerisation of P-cyanopropionaldehyde in THF initiated by Ph,CO-Na+ there is a marked influence of initiator concentration on the nature of the product.98" It is postulated that at high concentration the propaga-tion is mainly through ion pairs leading to insoluble stereoregular polymer whereas at low concentration propagation is mainly through free anions leading to soluble amorphous polymer.The effect of an electric field (15 kV cm- ') on anionic polymerisations has been extensively studied by Ise who has reviewed the findings to date.99 There is an enhancement in rate corresponding to an increase in k-K* in equation (3) but not in k* . It is not at present clear whether the enhanced rate is entirely due to the effect of field on K .The observed increase in k - K* is much bigger than predicted by the Onsager theory of the second Wien effect. Hence either the theory is at fault or there is a field effect on k - perhaps connected with the tendency for the solvent sphere round the anion to become disoriented as the anion is drawn through the solution. 0 ther vinyl monomers which have recently been studied include b ~ t a d i e n e ~ ~ nitroethylene,'" u- and p-methoxystyrene,"' p-dimethylaminostyrene lo2 and styrene in toluene in the presence of aromatic ethers ;Io3 also N-(p-viny1)phenyl-acrylamide' O4 which polymerises anionically only through the styrene-type double bond to give soluble polymers. 9 6 N.Yamazaki Fortschr. Hochpo1ym.-Forsch. 1969 6 377. 9 7 B. L. Erussalimsky and I. Krassnoselskaya Makromol. Chem. 1969 123 80. S. E. Bresler B. L. Erussalimsky and I. V. Kulevskaya J . Polymer Sci. Part A-1, Polymer Chem. 1968 6 2795. 98a H. Sumimoto and K. Hashimoto J . Polymer Sci. Part A - I Polymer Chem. 1969 7, 1331. 99 N. Ise Fortschr. Hochpo1ym.-Forsch. 1969 6 347. l o o H. Yamaoka H. Mori K. Hayashi and S. Okamura J . Polymer Sci. Part B Polymer Letters 1969 7 371. J. Geerts M. van Beylen and G. Smets J. Polymer Sci. Part A-1 Polymer Chem., 1969,7 2859. M. Fontanille D. Meimoun and P. Sigwalt European Polymer J. 1969 5 553. l o 3 J. Geerts M. van Beylen and G. Smets J . PolymerSci. Part A - I Polymer Chem. 1969, l o 4 H. Kamogawa J . Polymer Sci. Part A-1 Polymer Chem.1969 7 725. 7 2805 Kinetics and Mechanism of Polymerisation 133 The use of living anionic systems to prepare block copolymers is well known. One of the two cross-propagation reactions will sometimes not occur ; for example poly(methy1 methacrylate) anions will not add to styrene,"' nor will poly-isocyanate anions add to the common vinyl monomers,"' so limiting the types of block copolymer which can be obtained in these cases to types AB and ABA. Again the poly(radica1-anions) which can be obtained by treating the copolymer of styrene and p-vinyl-trans-stilbene with sodium in THF will not add AN MMA, or styrene to give graft polymers although they will initiate the polymerisation of these monomers by electron transfer.lo7 Chain transfer is not usually of importance in the systems described above.However transfer to toluene has been detected by a tracer methodlog; it also occurs in copolymerisations involving trans-stilbene. ' O9 The mechanism of anionic polymerisation outlined above is the one generally accepted. For a different interpretation see the papers by Korotkov.' "7' '' Anionic Polymerisation of Cyclic Monomers.-Increasing interest has been shown in the polymerisation of ring compounds particularly the episulphides which have been reviewed by Sigwalt. ' Numerous initiators have been found for the polymerisation of episul-phides,' 1 3 ' l 4 all ionic in nature. The enantiomers of propylene sulphide have been prepared' 5 9 1 l 6 and when polymerised by certain catalysts such as cadmium tartrate or sodium metal give polymers with a high constant optical activity.Other catalysts such as BF30Et2 ,l l6 give polymer with a lower optical activity. The former catalysts are believed to act through an anionic mechanism in which only the CH2-S bond is broken whereas the latter act through a cationic mechanism with opening of either the CH2-S or the CH-S bond giving rise to a certain amount of head-head tail-tail structure. Copolymerisation experiments show that the reactivity ratios depend very much on the catalyst used"7,"8 indicating a wide variation in the nature of the chain carrier. l o 5 C. G. Overberger and N. Yamamoto J . Polymer Sci. Part B Polymer Letters 1965,3, l o 6 R. A. Godfrey and G. W. Miller J. Polymer Sci. Part A-1 Polymer Chem. 1969 7, lo' D.Braun F-J. Q. Lucas and W. Neumann Makromol. Chem. 1969 127 253. lo' A. L. Gatzke J . Polymer Sci. Part A-1 Polymer Chem. 1969 7 2281. l o 9 Y . Okamoto M. Kato and H. Yuki Bull. Chem. SOC. Japan 1969 42 760. ' l o A. A. Korotkov and A. F. Podolsky J. Polymer Sci. Part B Polymer Letters 1969,7, ' ' I A. F. Podol'skii E. P. Skvortsevich and A. A. Korotkov Polymer Sci. (U.S.S.R.), 'I2 P. Sigwalt in 'Ring-opening Polymerization' ed. K. C. Frisch Marcel Dekker New ' I 3 D. R. Morgan and R. T. Wragg Makromol. Chem. 1969 125,220. 'I5 N . Spassky and P. Sigwalt Tetrahedron Letters 1968,32 3541. 'I6 N . Spassky and P. Sigwalt Bull. SOC. chim. France 1967 4617. '"S. Boileau and F. Borsali Compt. rend. 1969 268 C 590. ' I 8 M. F. Bouvier and N . Spassky Compt. rend.1969 268 C 681. 569. 2387. 85. 1969,11 A 295. York 1969 ch. 4 pp. 191-217. W. Cooper D. R. Morgan and R. T. Wragg European Polymer J. 1969,5 71 134 K. J. Ivin Catalysts made from mixtures of diethyl zinc and an optically active compound, such as ( -)leucir~e,"~ ( +)borneol,"g~'20 or ( -)menthol,'20 show a tendency to select one or other isomer if used to polymerise racemic propylene sulphide. The sign of rotation of the polymer is opposite to that of the alcohol and opposite to that of the residual monomer ; it is the same in chloroform and benzene unlike the case of optically active poly(propy1ene oxide).' l 9 The anionic polymerisation of propylene sulphide in THF is of the 'living' type as shown by (a) the increase in molecular weight with time,'21''22 (b) the formation of polymers with very sharp molecular weight distributions,' 2 1 and (c) the formation of block copolymer^.'^^ Despite the fact that poly(propy1ene sulphide) ion pairs in THF have dissociation constants which are a factor of 100 or so smaller than those of polystyrene ion pairs in THF12 (owing to the much greater localisation of the negative charge) the free ions nevertheless play a dominant part in the propagation reaction in THF k - and k (for Na+) being estimated as 12 and 0.01 1 mol- ' s- ' respectively at - 30 0C.125 In THP as solvent the ionisation can be suppressed by the addition of NaBPh, k (for Na') being 0.058 1 mol-' s-' at 20°C.126 Association of the ion pairs to ion-pair dimers becomes significant at high concentrations (> M) as shown by viscosity measurements before and after reaction of the living ends with a terminating agent.It has been known for some time that the physical properties of poly(propy1ene sulphide) depend on the catalyst used to initiate polymerisation.' 2 7 * 1 2 8 An X-ray investigation of the hard crystalline polymers formed using certain cad-mium catalysts shows that they have an isotactic structure and a slightly distorted planar zig-zag configuration129 of the main chain. N.m.r. spectra of -fCH2CD(CH3)S+ show two overlapping AB quartets arising from isotactic and syndiotactic dyads in the ratio 2 1 for CdC03 as catalyst and nearly 1 1 for ZnC03 as ~ a t a 1 y s t . l ~ ~ The directing influence of CdC03 may be attributed to the co-ordination of the cadmium ions to the terminal sulphur atoms of the growing anion and the sulphur atom of the approaching monomer molecule.In the radiation-induced polymerisation of cyclohexene sulphide the rate is proportional to the intensity whereas the degree of polymerisation is independent J . Furukawa N . Kawabata and A. Kato J . Polymer Sci. Part B Polymer Letters, 1967 5 1073. N. Spassky and P. Sigwalt Compt. rend. 1967 265 C 624. S. Boileau G. Champetier and P. Sigwalt Makromol. Chem. 1963 69 180. S. Boileau and P. Sigwalt Compt. rend. 1965 261 C 132. I" H. Ito S. Sakai and Y. Ishii Kogyo Kagaku Zasshi 1968 71 288. I z 4 S. Boileau and P. Sigwalt European Polymer J. 1967 3 57. 1 2 5 J. C. Favier S. Boileau and P. Sigwalt European Polymer J. 1968 4 3. G. Tersac S. Boileau and P.Sigwalt European Polymer J . 1968 65 1141. 12' S. Adamek B. B. J. Wood and R. T. Woodhams Rubber and Plastics Age 1965 56. J. P. Machon and P. Sigwalt Compt. rend. 1965 260 C 549. H. Sakakihara Y. Takahashi H. Tadokoro P. Sigwalt and N . Spassky Macro-molecules 1969 2 51 5. I 3 O K. J. Ivin and M. Navratil J . Polymer Sci. Part B Polymer Letters 1970 8 51 Kinetics and Mechanism of Polymerisation 135 of intensity ; an ionic mechanism has been proposed.' 31 This behaviour is similar to that for the radiation-induced polymerisation of cyclohexene oxide.' * Ethylene Oxide. This is polymerised by potassium t-butoxide in anhydrous DMSO and at low conversion the rate is proportional to the concentration of monomer and ~atalyst,~' with k = 0.10 1 mo1-l s- at 25 "C and E = 6-4 kcal mol- '.It is thought that the reaction is propagated largely by the free alkoxide ion the rate being cu. 1000 times greater than in methanol-dioxan as solvent.'33 Amorphous zinc dimethoxide obtained by refluxing diethyl zinc under methanol is a highly active catalyst for the living polymerisation of propylene oxide the rate being proportional to the monomer concentration and dependent on the co-ordination strength of the solvent.134 In a very comprehensive paper Vandenberg13' has described how the stereo-chemistry structure and mechanism of formation of epoxide polymers obtained using modified trialkylaluminium catalysts may be elucidated from the structure of the monomer dimer and trimer glycol fragments formed when the polymers are cleaved by Group IA organometallic compounds.The crystalline polymers obtained from cis- and truns-2,3-epoxybutane are respectively racemic and rneso-di-isotactic ; the amorphous polymer from the cis-oxide is disyndiotactic. The amorphous fraction of poly(propy1ene oxide) contains substantial amounts of head-to-head tail-to-tail structure a finding which is supported by the n.m.r. ~ p e c t r u m . ' ~ ~ It has been established that epoxides polymerise with inversion of configuration of the ring-opening carbon atom and that monosubstituted epoxides polymerise largely by attack on the primary carbon atom. In order to explain the inversion of configuration it is postulated that two or more metal atoms are involved in the propagation process. The anionic polymerisation of luctarns can be initiated at the comparatively low temperature of 9&150" by means of disubstituted carbamoyl lactams and lactam-N-carboxylic acid esters.The polymerisation of E-caprolactone initiated by dibutylzinc and tri-isobutylaluminium shows some of the features of a living polymerisation the molecular weight of the polymer being inversely proportional to the catalyst concentration and increasing with percentage conversion. 13' However the molecular weight is 3-5 times larger than expected and the distribution is considerably broader than Poisson. Cationic Polymerisation of Vinyl Monomers.-For the polymerisation of iso-butene in hydrocarbon or alkyl chloride solvents at - 70 to - 30 "C with AlEt,Cl 1 3 ' S. Boileau and J. C. Miiller Compt. rend. 1969 268 C 2284.D . Cordischi A. Mele and R. Rufo Trans. Faraday Soc. 1968 64 2794. 1 3 3 G. Gee W. C. E. Higginson and G. T. Merrall J. Chem. SOC. 1959 1345. M. Ishimori G . Hsiue and T. Tsuruta Makromol. Chem. 1969 128 52. 1 3 5 E. J. Vandenberg J. Polymer Sci. Part A - I Polymer Chem. 1969 7 525. 1 3 6 H . Tani N. Oguni and S. Watanabe J. Polymer Sci. Part B Polymer Letters 1968, 13' G . Falkenstein and €3. Dorfel Makromol. Chem. 1969 127 34. 1 3 8 R. D. Lundberg J . V. Koleske and K. B. Wischmann J. Polymer Sci. Part A - I , 6 577. Polymer Chem. 1969 7 2915 136 K. J. Ivin as catalyst the effectiveness of cocatalysts lies in the order HCl HBr >> HF, H 2 0 > CCI3CO2H >> MeOH > MeCOMe. A similar investigation for styrene shows that with alkyl or aryl chlorides RC1 as cocatalyst the efficiency is apparently determined by the relative stability and/or concentration of the initiating carbonium ions provided by the RCl.14' Thus whereas Bu"C1 PriC1, and BUT1 exhibit low efficiency because of the resulting low ion concentration, Ph3CCl is poor because the stability of Ph3Cf is much higher than that of the propagating polystyryl cation.Chain transfer constants have been determined for the BF30Et2-initiated poly-merisation of styrene in benzene solution at 30°C in the presence of various p01yethers.l~~ The transfer constants for the entities -CH20CH2-, -CH20CH(CH3)- and -CH20H were 6 x lop3 7 x and 6 res-pectively. Thus for low-molecular-weight poly(ethy1ene oxide) chain transfer occurs mainly at the end groups.In the BF,OEt,-initiated polymerisation of the hydroxystyrenes the rate and molecular weight increase in the order m < o << p.14' With this catalyst the p-monomer gives a much higher molecular weight product than that obtained with free radical catalysts. With the 0- and m-monomers there is a considerable amount of reaction through the phenol nucleus. Substitution by a P-methyl group in styrene decreases the reactivity towards cationic catalysts but with vinyl ethers the reverse is true. 13C N.m.r. spectra of the vinyl ethers show that the increase in rate caused by a P-methyl substituent does not arise from the change in n-electron density at the P-carbon atom.'43 The polymerisation of N-vinylcarbazole with various catalysts appears to be cationic in nature.144 The apparent polymerisation by p-chloranil is due to an acid impurity 3,5,6-trichloro-2-hydroxy- 1,4-benzoquinone.14' Copoly-merisation experiments demonstrate that in 1,2-dichloroethane as solvent the catalysts BF30Et and tetracyanoethylene produce copolymer of essentially the same comp~sition.'~~ However it appears that the catalysts LiC1 LiBr and LiI, whose activities increase in that order initiate polymerisation through a molecular mechanism stimulated by co-ordination of the lone pair electrons on the nitrogen atom to the lithium ~ a t i 0 n . I ~ ' For a discussion of this question see the review by Tazuke.I4* 1 3 9 J. P. Kennedy J . Polymer Sci. Part A-1 Polymer Chem. 1968 6 3139. 140 J. P. Kennedy J . Macromol. Sci. 1969 A3 861. 1 4 ' Y .Minoura and M. Mitoh Makromol. Chem. 1969 128 41. 14' M. Kato J . Polymer Sci. Part A-1 Polymer Chem. 1969 7 2405. 1 4 3 T. Higashimura and S. Okamura J . Polymer Sci. Part B Polymer Letters 1969 7 23. 144 S. Tazuke M. Asai and S. Okamura J . Polymer Sci. Part A-1 Polymer Chem. 1968, 145 T. Natsuume Y . Akana K. Tanabe M. Fujimatsu M. Shimizu Y . Shirota H. Hirata, 146 J. M. Barrales-Rienda G. R. Brown and D. C. Pepper Polymer 1969 10 327. 14' L. P. Ellinger Polymer 1969 10 531. 14' S. Tazuke Fortschr. Hochpo1ym.-Forsch. 1969 6 321. 6 1809. S. Kusabayashi and H. Mikawa Chem. Comm. 1969 189 Kinetics und Mechanism of Polymerisation 137 The kinetics of cationic polymerisation of substituted vinylcyclopropanes 14' cyclohexa-l,3-diene,' 5 0 and cis,cis-cyclo-octa-l,3-diene1 have also been studied.Cationic Polymerisation Cyclopolymerisation and Cyclic Monomers.-The cyclopolymerisation of o-phthalaldehyde occurs readily in'the presence of cationic catalyst^.'^^ For example with BF30Etz in CH,Clz at - 78 "C the rate constant is 0.18 lmol-' s-l which is considerably faster than for the corresponding polymerisations of tetrahydrofuran and 1,3-dioxolan at 0 "C. The polymerisa-tion is reversible the equilibrium concentration of monomer being 1M at -43 "C. \ 3 /o\ CHO CHO CH CH The polymerisation of tetrahydrofuran is catalysed by Ph3CSbC16. 14C-Labelled catalyst does not enter the polymer but results in the formation of [ ''C]-Ph3CH indicating that one method of initiation is by hydride-ion abstrac-tion from the monomer.'53 Epoxides act as cocatalysts for this reaction, presumably by increasing the rate of initiation.' 5 4 Simultaneous conductance and dilatometric measurements of the poly-merisation of 1,3-dioxolan by Et,OBF show that the initial conductance declines considerably before the onset of polymeri~ation.'~~ On reversing the polymerisation and then repeating the measurements the conductance again declines but the polymerisation begins much more quickly.It is concluded that the formation of active centres which are stable when formed is slow and accom-panied by competitive hydride-ion abstraction from the monomer. Similar results were obtained for 1 3 - d i o ~ e p a n ' ~ ~ ~ ' ~ ~ which shows reversible poly-merisation in the temperature range - 65 to 5 OC.15' 1,3,6-Trioxocan has also been investigated.' 5 8 Comparison was made in a previous section between the anionic and cationic polymerisation of propylene sulphide.l1 Rather surprisingly the BF30Et,-catalysed reaction shows the characteristics of a 'living' polymerisation the molecular weight increasing with conversion and block copolymers being formed 1 4 9 T. Takahashi J . Polymer Sci. Part A-I Polymer Chem. 1968,6 3327. Y. Imanishi T. Yamane S. Kohjiya and S. Okamura J . Macromol. Sci. 1969 A3, 223. 15' Y. Imanishi K. Matsuzaki S. Kohjiya and S. Okamura J . Macromol. Sci. 1969 A3, 237. ' 5 2 C . Aso S. Tagami and T. Kumitake J. Polymer Sci. Part A - I Polymer Chem. 1969, 5 3 W. M. Pasika and J. W. Wynn J . Polymer Sci. Part A - I Polymer Chem. 1969,7 1489.1 5 4 I. Kuntz and M. T. Melchior J . Polymer Sci. Part A - I Polymer Chem. 1969,7 1959. 1 5 5 F. R. Jones and P. H. Plesch Chem. Comm. 1969 1230. ' 5 6 M. Okada S. Kozawa and Y . Yamashita Makromol. Chem. 1969 127 271. 15' P. H. Plesch and P. H. Westermann Polymer 1969 10 105. 15' M. Okada S. Kozawa and Y. Yamashita Makromol. Chem. 1969 127 66. 7 497 138 K. J . Ivin on adding ethylene sulphide to previously polymerised propylene sulphide.' 59 The rate increases with the polarity of the solvent Et20 < EtCl < EtNO,. The same effect of polarity is found in the polymerisation of N-phenyl ethyleneimine, catalysed by formic acid.'60 In solvents with similar relative permittivity the rate decreases with increase in nucleophilicity of the solvent. Co-ordinated Polymerisation.-Some of the systems already described under the headings of anionic and cationic polymerisation are undoubtedly of the co-ordinated type.These are distinguished most clearly by the greater degree of stereoregularity of the polymer in comparison with systems where there is no co-ordination. The extent of stereochemical control may vary from very little to very complete and a given catalyst may contain a variety of sites of different regulating power particularly if it is heterogeneous. As a result the product can frequently be separated into an amorphous soluble fraction and a crystalline insoluble fraction. ' 28 However the solubility and crystallinity of polymers depend to some extent on molecular weight and thermal history ; a more reliable guide is the n.m.r.spectrum as discussed in a later section of this report. In the polymerisation of ethylene with Ziegler catalysts of the type AlEt or AlEt,Cl with TiC1 or Ti& experiments with added deuterium indicate the presence of three kinds of Ti site Ti" sites at which P-hydrogens of the growing chain can exchange with deuterium; Ti"' sites at which CH2D groups can be formed by a transfer reaction; and inactive titanium hydride sites at which ex-change can result in the formation of deuteriated ethanes.161 Soluble Ziegler-type catalysts have been increasingly used in an effort to avoid the complications which arise with heterogeneous catalysts. The subject has been reviewed with particular reference to soluble catalysts of the type Cp2TiR'C1-R2A1C12 where Cp = cyclopentadienyl and R',R2 = alky1.16 For the poly-merisation of ethylene there is no induction period and the rate is proportional to the catalyst c ~ n c e n t r a t i o n .' ~ ~ * ' ~ ~ The catalytically active species is formu-lated as structure (2). 14C-Labelling of the catalyst shows that the polymerisation of styrene occurs exclusively by insertion of the monomer between the Ti-R bond. Co-ordination of monomer to the free site in (2) has been shown experi-mentally to enhance destabilisation of this bond.162 For styrene the kinetics indicate that the active sites are formed by a reaction between the catalyst (2) (R' = R2 = Et) and monomer.165 The order of reactivity of deuteriated styrenes with these catalysts is a-,H > PP-'H2 = undeuteriated; the i.r.spectrum of the 2H2-compound shows the polymer to be highly isotactic. 166 The combination L. A. Korotneva G. P. Belonovskaya N. A. Korol' and B. A. Dolgoplosk Doklady Akad. Nauk S.S.S.R. 1968 178 1084. 1 6 0 T. Kagiya T. Kondo K. Nakao and K. Fukui Bull. Chem. SOC. Japan 1969,42 1094. 1 6 1 A. Schindler Makromol. Chem. 1968 118 I . 1 6 * G. Henrici-Olive and S. Olive Fortschr. Hochpo1ym.-Forsch. 1969 6 421. 164 K. H. Reichert and E. Schubert Makromol. Chem. 1969 123 58. 1 6 5 K. H. Reichert J. Berthold and V. Dornow Makromol. Chem. 1969 121 258. 16' K. H. Reichert and J. Berthold Makromol. Chem. 1969 124 103. G . Henrici-Olive and S. Olive Kolloid-Z. 1969 228 43 Kinetics and Mechanism of Polymerisation 139 Ti acac3-Et,AlCl also gives a catalyst which is soluble in hydrocarbons; it polymerises ethylene but not propylene.' 67 The opposite approach to using a soluble catalyst is to use a single crystal of a-TiC13 which has been exposed to the vapour of A1Me3.168 Such a surface polymerises propylene and electron microscopy shows that within a few seconds the polymer appears as granular units on the lateral crystal faces.After a longer time continued growth transforms the polymer units into fibrils of cu. 40nm diameter displaying striations perpendicular to the fibre axis. The width of the striations corresponds to a repeat distance of 7 nm which may represent the folded polypropylene chain. The rate constant for propagation is estimated to be lo3-3 x lo6 1 mol- s - ' depending on how many titanium atoms are assumed to be called into play on the surface.Poly(N-alkyliminoalanes) f AlH-NR+, with n = 6-12 in conjunction with TiCl, polymerise isoprene to a gel-free cis-1,4-polymer of high molecular weight.16' Catalyst systems consisting of an aluminium compound with a cobalt complex are capable of polymerising penta-1,3-diene to cis-1,4- or 1,2-syndiotactic polymer depending on the aluminium compound and solvent.' 70 Similarly truns-2-methylpenta-1,3-diene can be polymerised either to cis-1,4-or truns-1,4-polymer according to choice of ~ata1yst.l~ ' Stereospecific poly-merisation of cyclic olefins can produce either linear or cyclised polymers. A mixture of WC16 and AlBr3 with no added alkyls is reported to convert various cyclic olefins and cyclic dienes to linear polymers.' Other monomers whose polymerisation with complex catalysts is reported include methyl methacrylate (with VOCl3-A1Et3),' 7 3 P-cyanopropionaldehyde (with A1Et3-monomer adducts),' 74 I-methylpropyl vinyl ether [with A1(OPri),-H2S0,],' ' propylene propylene sulphide" 5 9 1 l 6 and or-methylbenzyl isonitrile which appears to give a polymer with units linked solely through the carbon atoms of the isonitrile groups.'76 The last four monomers all contain 1 6 ' W.P. Watt F. H. Fry and H. Pobiner J . Polymer Sci. Part A-I Polymer Chem., 1968 6 2703. J. Y . Guthman and J. E. Guillet Macromolecules 1968 1 461. 1 6 9 A. Mazzei S. Cucinella and W. Marconi Makromol. Chem. 1969 122 168. L. Porri A. di Corato and G. Natta European Polymer J . 1969,5 1 . 1 7 ' D .Cuzin Y . Chauvin and G . Lefebvre European Polymer J . 1969 5 283. P. R. Marshall and B. J. Ridgewell European Polymer J . 1969 5 29. S. S. Dixit A. B. Deshpande L. C. Anand and S. L. Kapur J . Polymer Sci. Part A - I , Polymer Chem. 1969 7 1973. ' 7 4 K. Kobayashi and H. Sumitomo J . Polymer Sci. Part A - I Polymer Chem. 1969 7 , 1287. 1 7 5 H. Tani and N. Oguni J . Polymer Sci. Part B Polymer Letters 1969 7 769. 1 7 6 F. Millich and G . K. Baker Macromolecules 1969 2 122 140 K. J. Ivin asymmetric carbon atoms and can be polymerised with either complete or partial retention of optical activity. Stereoselection in which the isomers in the racemic monomer tend to polymerise separately but at the same rate has been demon-strated for the ether. Stereoelection in which one isomer polymerises preferen-tially has been demonstrated for the sulphide with certain catalysts.' 199120 Copolymerisation.-General.Copolymer composition studies provide a quick and easy method of determining the relative reactivity of two monomers for two types of chain carrier and much work continues to be done in this field. Analysis of composition data by computer has been described.'77,'78 The effect of additives in modifying the reactivity ratios even to the extent of producing alternating copolymers is a subject of increasing interest (see beiOw),179-191 the mechanical properties of such polymers often being superior to those of random copolymers. 192*193 In some cases n.m.r. spectra give detailed information about the proportions of various types of dyad triad tetrad and even hexad structures permitting a corresponding refinement of copolymerisation theory and allowance for pen-ultimate unit effects on the reactivity of the chain carrier^."^-'^* A number of 17' G.R. Brown and J. G. Byrne Polymer 1969 10 333. 17' Shu-Pei Chang T. K. Miwa and W. H. Tallent J . Polymer Sci. Part A - I Polymer Chem. 1969,7,471. '19 A. Kawasaki I. Maruyama M. Taniguchi and R. Hirai J . Polymer Sci. Part B, Polymer Letters 1969 7 613. J. Furukawa R. Hirai and M. Nakaniwa J . Polymer Sci. Part B Polymer Letters, 1969 7 671. S. Pasynkiewicz W. Kuran and T. Diem J . Polymer Sci. Part A - I Polymer Chem., 1969 7 241 1 . S. Yabumoto K. Ishii and K. Arita J . Polymer Sci. Part A - I Polymer Chem. 1969, 7 1577. N. G .Gaylord and A. Takahashi J . Polymer Sci. Part A - I Polymer Chem. 1969 7, 443. M. Tamiguchi A. Kawasaki and J. Furukawa J . Polymer Sci. Part B Polymer Letters 1969 7 41 1 . l a 5 N. G. Gaylord and H. Antropiusova J . Polymer Sci. Part B Polymer Letters 1969 7, 145. N. G . Gaylord and A. Takahashi J . Polymer Sci. Part B Polymer Letters 1968 6, 743 749. S. Inoue K. Kitamura and T. Tsuruta Makromol. Chem. 1969 126 250. T. Otsu and H. Inoue Makromol. Chem. 1969 128 31. T. Kokubo S. Iwatsuki and Y. Yamashita Makromol. Chem. 1969 123 256. Chem 1968 120. 154. 19"S. Iwatsuki T. Kokubo K. Motomatsu M. Tsuji and Y. Yamashita Makromol. 9 1 A. A. EI'Saied S. Y. Mirlina and V. A. Kargin Polymer Sci. (U.S.S. R . ) 1969,11 A 3 14. lY2 S. Yabumoto K. Ishii M. Kawamori K.Arita and H. Yano J . Polymer Sci. Part A - I Polymer Chem. 1969 7 1683. 1 9 3 J. Furukawa Y. Iseda K. Haga N. Kataoka T. Yoshimoto T. Imamura Y. Shido, A. Miyagi K. Tanaka and K. Sakamoto J . Polymer Sci. Part B Polymer Letters, 1969 7 561. 1 9 4 K-H. Hellwege U. Johnsen and K. Kolbe Kolloid-Z. 1966 214 45. 1 9 5 J. B. Kinsinger T. Fiscber and C. W. Wilson J . Polymer Sci. Part B Polymer Letters, 196 U. Johnsen and K. Kolbe Kolloid-Z. 1969 232 712. 197 C. E. Wilkes J . C. Westfahl and R. H. Backderf J . Polymer Sci. Part A - I Polymer 19' Ting Kai Wu J . Phys. Chem. 1969 73 1801. 1967 5 285. Chem. 1969 7 23 Kinetics and Mechanism of Polymerisatim 141 systems in which one or more propagation steps are reversible have been in-vestigated both t h e o r e t i ~ a l l y ~ ~ ~ .~ ~ ~ and experimentally.20 '-'06 Free-radical Copolymerisation. Reactivity ratios have been determined for the copolymerisation of styrene (M,) with ring-perfluorinated styrene207 ; p-isopropyl styrene;208 acrylic acid209 (rl = 0.29 1 = 0.075 at 65"C differing markedly from previous values) ; methacrylic acid ;'09 methyl methacrylate,' l o where the effect of solvent has been correlated with its proton-donating ability ; ethyl, isobutyl and n-nonyl methacrylates ;,' glycidyl /I-vinylacrylate and glycidyl sorbate seventeen 1,l-disubstituted ethylenes where it is concluded from an analysis of the trends of r and r2 with the Hammett 0 values that steric effects of the substituents are not significant compared with polar and resonance fac-tors ; l 3 diethyl fumarate2I4 and di(2-cyanoethyl) fumarate ;" and N-acryloyl-2-0xazolidone,~ l 6 CH2 CHCOkCH2CH20k0.Methyl methacrylate has been copolymerised with ring-perfluorinated styrene,207 p-isopropylbenzene,208 acrylic acid,209 and methacrylic acid.209 Vinyl acetate has been copolymerised with allilidene diacetate CH CHCH (OCOMe) di(2-~yanoethyl)fumarate and the 1,2-di~hloroethylenes,~ where it is found that the trans-isomer is about six times as reactive as the cis-isomer towards the poly(viny1 acetate) radical. In the last case there is good evidence that chain transfer occurs by a chlorine atom elimination reaction. Vinyl chloride (MI) has been copolymerised with seven vinyl esters,'19 and there is a direct correlation between r l and the Hammett 0 value; also with 2-methylpentyl brassylate.' 7 8 The n.m.r.spectrum of the copolymer of vinyl chloride with ethylene provides a direct measure of the relative concentrations of the three types of dyad M I M l M1M2 and M2M, and thence the value of 199 G. G. Lowry J. Polymer Sci. 1960 42 463. 'O0 M . H. Theil Macromolecules 1969 2 137. '01 K. F. O'Driscoll and F. P. Gasparro J . Macromol. Sci. 1967 A l 643. 'O' J. E. Hazel1 and K . J. Ivin Trans. Faraday SOC. 1962 58 342; 1965 61 2330. '03 K. J. Ivin and R. H. Spensley J . Macromol. Sci. 1967 Al 653. * 0 4 K. F. O'Driscoll and J. R. Dickson J . Macromol. Sci. 1968 A2 449. * 0 5 A. I. Kuzayev G. M. Komratov G. V. Korovina G. A. Mirontseva and S. G. EnteIis, ' 0 6 Y. Yamashita T. Asakura M. Okada and K . Ito Makromol.Chem. 1969 129 1. '07 W. A. Pryor and T-L. Huang Macromolecules 1969 2 70. '08 R. H. Wiley and Jung-I1 Jin J . Macromol. Sci. 1969 A3 835. '09 G. Smets and K. Van Gorp European Polymer J. 1969 5 15. ' l o T. Ito and T. Otsu J. Macromol. Sci. 1969 A3 197. '" A. K. Chaudhuri and S. R. Palit Makromol. Chem. 1969 121 33. "' Y. Iwakura F. Toda R. Iwata and Y. Torii Bull. Chem. SOC. Japan 1969,42 837. ' I 3 B. Yamada and T. Otsu J . Polymer Sci. Part A - I Polymer Chem. 1969 7 2439. '14 K. Horie I. Mita and H. Kambe J. Polymer Sci. Part A-I Polymer Chem. 1969 7, '" V. Arendt and S. Kaizerman J. Polymer Sci. Part A - I Polymer Chem. 1969 7 2741. ' I 6 T. Endo R. Numazawa and M . Okawara Makromol. Chem. 1969 123 46. ' I 7 M. Sadamichi and K. Noro J. Macromol. Sci. 1969 A3 845.' I 8 T. L. Dawson R. D. Lundberg and F. J. Welch J . Polymer Sci. Part A - I Polymer 'I9 K. Hayashi and T. Otsu Makromol. Chem. 1969,127 54. Pol-vmer Sci. ( U . S . S . R . ) 1969 11 A 502. 2561. Chem. 1969,7 173 142 K. J. Ivin r1r2 = 0.7 in agreement with that derived from composition data;"' similarly for the copolymer of ethylene with vinyl formate.' 98 Acrylonitrik (M 1) has been copolymerised with cyclopentene,220 which enters the copolymer to the maximum extent of one unit in three; also with four monomers for which r2 = 0, namely P-cyanoacrolein"' (rl = 3.2) and three monomers of formulae PhCH :C(CN)R,222 with R = H (rl = 6.4) C02Et (rl = 18) and CN (rl = 6.2). In contrast with styrene as M1 there is no penultimate unit effect. The copolymer of vinylidene chloride (M 1) with isobutene (M,) is excep-tionally interesting in that its n.m.r.spectrum is uncomplicated by spin-coupling or tacticity effects and at the same time the chemical shifts for the CH2 groups in the homopolymers are widely separated. As a result it is possible to make an accurate determination of the relative concentrations of different tetrad struc-tures and hence to determine the reactivity ratios for four types of growing radical : Radical M l M l . M2Ml. M2M2 * MlM2' r1 (SO0) 2.1 _+ 0.4 4.65 & 0-7 r2 (50") 0.22 f 0.15 0.08 f 0.05 E (kcal mol - I ) - 2.2 f 0.4 - 3.0 & 0.5 Hexad structure can also be detected. Anionic Catiolzic and Co-ordinated Ionic Copolymerisation. A comprehensive investigation of the copolymerisation of isoprene (M1)223 and other d i e n e ~ ' ~ ~ with 1 1-diphenylethylene (M,) using anionic initiators is reported.M2 does not homopolymerise so that r = 0. rl is independent of cation (Li' Na' K') for tetrahydrofuran as solvent but is strongly dependent on cation for benzene as solvent ranging from 37 for Li+ to 0.05 for K'. Smaller but substantial varia-tions with solvent and cation have also been for the copolymerisation of methyl and benzyl methacrylates r1r2 varying from 0.98 to 3.14. These dif-ferences show that the propagating species is an ion pair and not a free anion. A remarkable series of copolymers of the type fMRMj- have been made by reacting a-methylstyrene butadiene or isoprene (M) with a suspension of lithium metal in THF to which a dihalide e.g.Br(CH,),Br (n > 3) has been added.226-228 The polymers result from a coupling reaction between the dimeric dianion of M with the dihalide this reaction being faster than the further propagation of polymerisation of M. As a result the M units are joined in head-to-head fashion to R. Reactivity ratios have been determined for the copolymerisation of phenyl glycidyl ether (Ml) PhOCH2CHCH2d with its p-C1- m-MeO- p-Me- and 2 2 0 1. K. Hecht J . Polymer Sci. Part B Polymer Letters 1969 7 31. 2 2 1 I. Takemura and H. Sumimoto Bull. Chem. SOC. Japan 1969,42 634. 2 2 2 S. H. Ronel and D. H. Kohn J . Polymer Sci. Part A - I Polymer Chem. 1969 7 2209. 223 H. Yuki and Y . Okamoto Bull. Chem. SOC. Japan 1969,42 1644. 224 H. Yuki K. Hatada and I. Inoue J . Polymer Sci.Part B Polymer Letters 1968 6, 2 2 5 K. Ito T. Sugie and Y . Yamashita Makromol. Chem. 1969 125 291. 2 2 6 D. H. Richards N. F. Scilly and F. Williams Polymer 1969 10 603. 2 2 7 D. H. Richards N. F. Scilly and S. M. Hutchinson Polymer 1969 10 611. 2 2 a D. H. Richards and N. F. Scilly J . Polymer Sci. Part B Polymer Letters 1969 7 99. 3333 Kinetics and Mechanism of Polymerisation 143 p-MeO- derivatives and with propylene oxide catalysed by potassium t-butoxide in DMSO solution.229 As expected for an anionic mechanism the reactivity is enhanced by electron-withdrawing groups ; the reverse behaviour is found when Et3Al-H20 is used as catalyst which is therefore believed to act as a Lewis acid and to cause propagation by a co-ordinated cationic mechanism.230 Episulphides copolymerised using sodium carbazyl as catalyst in THF show the following order of reactivity ethylene sulphide > propylene sulphide z iso-butene ~ulphide.~~' With cationic initiators cyclohexene sulphide is more reac-tive than propylene sulphide which is more reactive than both ethylene sulphide and isobutene s ~ l p h i d e .~ ~ ~ The following cationic copolymerisations of vinyl monomers have been studied N-vinylcarbazole (MI) with p-methoxy~tyrene'~" ( r = 21 r2 = 0.13, nearly independent of catalyst and temperature) ; vinyl ferrocene (MI) with styrene and with vinyl isobutyl ether only vinyl addition being detected in spite of the high cationic reactivity of M1;233 phenyl vinyl ether (MI) with ring-substituted-M' where in five cases out of six there is a good correlation of rl with the Hammett o-value unlike the corresponding reactions of styrene where there is no oxygen atom to suppress direct conjugation to the double bond.234 Although cis-fl-chlorovinyl ethyl ether is more stable than the trans-isomer it is eight times more reactive in cationic copolymerisation with vinyl isobutyl ether.235 Cationic copolymerisations of cyclic compounds include two studies of the system 1,3-dioxolan-1,3,5-trio~an~~~~~~~ and one on the system propylene oxide (M'ktetrahydrofuran (rl = 0.05 r2 = 0.5 at 10 0C.)205 Copolymerisations of vinyl and heterocyclic compounds are much more difficult to achieve than copolymerisations of two vinyl or two cyclic compounds because of the disparity in reactivity of carbonium ions and for example oxonium ions with the respective monomers.Many of the earlier reports of the formation of copolymers in such systems may have to be reinterpreted in the light of new work which shows for example that styrene may be grafted cationically to polyepi~hlorohydrin.~~~ In attempting to copolymerise styrene and epichloro-hydrin the epoxy-compound polymerises first and later develops grafts of poly-styrene ; likewise with propylene oxide and styrene.238 In the light of this it is not surprising to find that the composition curves for the cationic copolymerisation of 3,3-bischloromethyloxetan with various vinyl monomers are rather unusual 2 2 9 C. C. Price Y. Atarashi and R. Yamamoto J . Polymer Sci. Part A-1 Polymer Chem., 230 C. C. Price and L. R. Brecker J .Polymer Sci. Part A-1 Polymer Chem. 1969 7 575. 2 3 1 S. Boileau and F. Borsali Compt. rend. 1969 268 C 590. 232 M. F. Bouvier and N. Spassky Compt. rend. 1969,268 C 681. 2 3 3 C . Aso T. Kunitake and T. Nakashima Makromol. Chem. 1969,124 232. 234T. Fueno T. Okuyama I. Matsumura and J. Furukawa J . Polymer Sci. Part A-1, 2 3 5 T. Okuyama and T. Fueno J. Polymer Sci. Part A-1 Polymer Chem. 1969 7 2433. 2 3 6 M. Baccaredda M. Giorgini A. Lucchesi F. Morelli and R. Tartarelli J . Polymer 2 3 ' Y . Minoura and M. Mitoh Makromol. Chem. 1969 126 56. 2 3 8 Y . Minoura and M. Mitoh Makromol. Chem. 1969 124 241. 1969 7 569. Polymer Chem. 1969 7 1447. Sci. Part A-1 Polymer Chem. 1969 7 209 144 K . J. Ivin in shape and do not fit the Lewis-Mayo equation.239 Again n.m.r.spectra show that the copolymers of 1,3-dioxepan with isobutyl vinyl ether and of 1,3,6-trioxocan with 2-chloroethyl vinyl ether contain blocks with relatively long sequences.240 Copolymerisation studies with Ziegler-Natta catalysts include the following : acrylonitrile (M,) with methyl methacrylate using the soluble catalyst Cp2TiC12 , A1Et3 (rl = 0.075 r2 = 0.55);241 and propylene (M,) with but-1-ene using TiC14,AlR3 (rl = 2.4 r2 = 0.5).242 In the EtAlCl,-initiated copolymerisation of n-butyl vinyl ether (M,) with other ethers it is found that the reactivity of M, correlates with the Taft o* value for the R-group of M2 ; there is also a close correlation with the chemical shifts of the CH2 protons of the vinyl In the ethylene-propylene copolymer made with vanadium-based catalysts about one third of the propylene units in M1M2 dyads are inverted.244 The y-induced copolymerisation of vinyl chloride with styrene at - 78 to 50 "C appears to be partly radical and partly ?-Radiation also induces the copolymerisation of vinyl chloride and ethylene dissolved in liquid carbon dioxide at 30 0C.246 Alternating Copolymerisation.Catalysts have been described which result in the formation of alternating copolymers of butadiene with propylene ;l 79,1 acry-lonitrile with vinyl chloride,18' styrene,'82,'83 isoprene,'83 and butadiene;'83,'84 styrene with methyl methacrylate ; 1 8 5 9 1 8 6 phthalic anhydride with propylene oxide to give a polyester'87 +o-PhCO2CH2CH(CH3)0CO-f ; and maleic anhydride with ethyl vinyl sulphide,' 88 phenyl vinyl sulphide,' 88 1,2-dimethoxy-ethylene,lg9 p - d i 0 ~ e n e .l ~ ~ and acrylic acid.'" In the last case it is possible to direct the composition away from 1 l by the addition of naphthalene. It is generally supposed that the alternating behaviour is a result of complex formation between the additive and either the monomer or the growing chain in such a way as to reduce both reactivity ratios. In a number of cases spontaneous copolymerisation occurs with or without an additive. For the spontaneous reaction of methyl methacrylate with styrene' 85 in the presence of Et3A12C13 the molecular weight increases with conversion, suggesting a type of living polymerisation in which the growing end may be a polarised molecule rather than an ion pair free ion or free radical.Spontaneous copolymerisations are also reported for bicycloheptene with SO2 ;247 l-methyl-3 9 T. Higashimura T. Masuda and S. Okamura J . Polymer Sci. Part A - I Polymer Chem., 1969 7 1 1 15. 240 M. Okada and Y. Yamashita Makromol. Chem. 1969,126 266. 2 4 1 C. Simionescu N. Asandei I. Benedik and C. Ungurenasu European Polymer J . 1969, 2 4 2 R. Laputte and A. Guyot Makromol. Chem. 1969,129 234. 243 H. Yuki K. Hatada and M. Takeshita J . Polymer Sci. Part A - I Polymer Chem. 1969, 2 4 4 C. Tosi A. Valvassori and F. Ciampelli European Polymer J. 1969 5 575. 2 4 5 M. Ryska C. Schneider and D. 0. Hummel Makromol. Chem. 1969 126 23. 246 M. Hagiwara T. Miura and T. Kagiya J. Polymer Sci. Part A - I Polymer Chem., 247 N. L. Zutty C. W. Wilson G.H. Potter D. C. Priest and C. J. Whitworth J . Polymer 5 449. 7 667. 1969 7 513. Sci. Part A Polymer Chem. 1965 3 2781 Kinetics and Mechanism of Polymerisation 145 cyclopropene with SO2 ;248 acrylonitrile with styrene,'82*' 83 isoprene,'83 or b ~ t a d i e n e ' ~ ~ in the presence of zinc chloride; and maleic anhydride with 1,2-dimetho~yethylene,'~~ p-dioxene,lS9 ethyl vinyl sulphide' 8 8 and phenyl vinyl sulphide.' 88 Closely related in mechanistic type are the homopolymerisations of N-vinylcarbazole induced by the addition of lithium saltsI4' or fi-cyanoacro-lein,249 and of 2-vinylpyridine induced by quaternisation with methyl iodide. 190 Charge transfer interactions (donor-acceptor complexes) must play an important part in such reaction^.'^^ Tacticity of Polymers.-The pioneering work of Bovey2" showed what a tremendous amount of information concerning the configuration and conforma-tion of polymer chains was potentially available from n.m.r.spectra. In order to appreciate the likely future developments in this field the reader need only con-sult Bovey's latest review.251 Two valuable summaries have appeared of work up to 1967.252,253 A IUPAC report deals with the problems of nomenclature of stereoregular polymers.254 Here we may note that the isotactic (i) syndiotac-tic (s) and heterotactic (h) triads in vinyl polymers may also be designated as mm rr and mr (rn = meso dyad r = racemic dyad); the latter symbolism has the advantage that it may be readily extended to tetrads and pentad^.^' ' In some cases it is now possible to measure the proportions of tetrads quantita-tively from n.m.r.spectra and also to obtain estimates of pentads. This provides an even closer test of two-parameter models of the chain growth process and extension where necessary to four-parameter models. Triad tacticity data have been interpreted in terms of two types of two-parameter model first the enantio-morphic-sites model (EMS) in which the mode of placement is determined by the stereochemistry of the catalyst site X as well as that of the end unit of the chain ; second the polymer-end-control model (PEC) in which the stereochemistry of the end unit and penultimate unit determine the mode of p l a ~ e m e n t . ~ ~ ~ > The characteristics of these two models have been examined and it is shown that the EMS model allows only a rather restricted set of i and s values.Another paper257 develops the EMS model in terms of the energy differences associated with right ( I ) - and left ($-handed addition of monomer. These are assumed to be governed by the catalyst [E = E(Xrj) - &(Xsj)] and by the end-unit [U = &(riri- 1) - &(risi- I)]. A plot of E us. U can be divided into five areas 2 4 8 S. Iwatsuki T. Kokubo and Y . Yamashita J . Polymer Sci. Part A-I Polymer Chem., 249 H. Sumimoto and I. Takemura Bull. Chem. SOC. Japan 1969,42,631. 2 s 1 F. A. Bovey Amer. Chem. Soc. Polymer Preprints 1969 10 9. 2 5 2 P. R. Sewell Ann. Rev. N.M.R. Spectroscopy Academic Press London 1968 vol. 1 , 253 K. C. Ramey and W. S. Brey Macromol. Rev. 1967 C l 263. 2 5 4 M.L. Huggins G. Natta V. Desreux and H. Mark Pure and Applied Chem. 1966, 2 5 5 R. A. Shelden T. Fueno and J. Furukawa J . Polymer Sci. Part A-2 Polymer Phys., 2 5 6 R. A. Shelden J . Polymer Sci. Part A-2 Polymer Phys. 1969 7 1 1 1 1 . 2 s ' P. Luisi and R. M. Mazo J . Polymer Sci. Part A-2 Polymer Phys. 1969 7 775. 1968 6 2441. F. A. Bovey and G. V. D. Tiers Fortschr. Hochpo1ym.-Forsch. 1963 3 139. p. 165. 12 645. 1969 7 763 146 K. J . Ivin corresponding to the formation of predominantly isotactic (two areas) syndio-tactic stereoblock and atactic polymers; also see ref. 258. A matrix method for the complete description of tactic polymers and copolymers has been The problem of ‘line ordering’ i.e. the factors which determine whether the n.m.r. resonances for triads lie in the z order i < h < s or the reverse has been discussed by Ritchey and Knollz6’ and the effects interpreted in terms of bond anisotropy associated with polar groups in the side chains.However the 13C-spectrum of poly(methy1 methacrylate) in which the carbonyl carbon resonance exhibits fine structure due to pentads,251 indicates that electron density rather than anisotropic shielding is the important factor in causing different chemical shifts for different configurational sequences. The line ordering is not always easy to determine without reference to X-ray crystallographic data or model com-pounds. However the 220 MHz spectrum of poly(a-methylstyrene) permits the assignment of the CH2 peaks to ten tetrad structures and confirms the original assignment i < h < Information concerning the tetrad structures in polypropylene has been derived from a study of the deuterium-decoupled 100 MHz spectra of polymers made from cis- and trans-[ 1,2,3,3,3-2H5]propylene.262 There are six tetrad structures designated rrr (ido) rrm (gluco) mrm (manno) mmm (allo) mmr (altro) rmr (galacto) in the last three of which the central proton may be syn or anti with respect to the two nearest methyl Accordingly nine resonances are observed six in each amorphous polymer (with three in common) and a single line for each of the two di-isotactic polymers and the syndiotactic polymer.A partial assignment can therefore be achieved. The 220 MHz spectrum of free-radically and anionically initiated poly(methy1 methacrylate) may be interpreted in terms of tetrads and pentads.251 The free-radical polymer conforms to a single-parameter model for chain growth in which the probability of isotactic placement P = 0.24.However with the anionic polymer it is necessary to use a four-parameter (second-order Markov) model in which Pmmlm = 0-94 Pmrl = 0.56 Prml = 063 Prrl = 0-28 where P,,,,/ de-notes the probability of a monomer adding in m fashion to a chain ending in mm, etc. For a two-parameter (first-order Markov) model to apply P,,, = Prmlm and Pmrlm = Prr/,. The proportions of pentads may be interpreted equally well on the basis of the second-order Markov model and the Coleman-Fox two-site mechanism in which it is assumed that chain growth can occur at two distinct types of reaction site.Unlike methyl methacrylate the trityl ester gives predominantly isotactic polymer with free radical as well as with anionic catalysts ; the bulky CPh3 group evidently makes syndiotactic placements unfavourable even for free-radical 2 5 8 M. Farina Mukromol. Chem. 1969 122 237. 2 5 9 C. Tosi and G. Allegra Mukromol. Chem. 1969 129 275. 2 6 0 W. M. Ritchey and F. J. Knoll J . Polymer Sci. Part B Polymer Letters 1966 4 853. 26’ K. C. Ramey G. L. Statton and W. C. Jankowski J . Polymer Sci. Part B Polymer 2 6 2 A. Zambelli and A. Segre J . Polymer Sci. Part B Polymer Letters 1968 6 473. 263 A. Zambelli A. L. Segre M . Farina and G. Natta Malcromol. Chem. 1967 110 1 . Letters 1969 7 693 Kinetics and Mechanism of Polymerisation 147 addition.264 In poly(methy1 a-chloroacrylate) the methoxy protons are sensitive to triads.The free-radically initiated polymer is predominantly syndiotactic but the anionically initiated polymer is predominantly isotactic only if made at 0 "C (rather than - 78 0C).265 Three other acrylate derivatives have been studied:266 CH,:C(CH,)C02R with R = C6H5 C6F5 and C3F,CH2. The effect of additives on the tacticity of both free-radi~ally,~~ and anionically26 * initiated poly(methy1 methacrylate) has also been examined. The vicinal coupling constants between a and /I protons in poly(acry1ic acid) are unaffected by neutralisation of the polyion from which it is concluded that expan-sion of the polyion with increasing neutralisation results from long-range inter-actions and not from substantial conformational changes about every main-chain bond.However the relative chemical shifts for different configurations are markedly changed by hydrolysis or n e u t r a l i ~ a t i o n . ~ ~ ~ Tacticity in copolymers of methacrylic acid and methyl methacrylate has been explored through the reaction of the copolymer with diazomethane to yield poly(methy1 metha~rylate),~~ O and much more effectively by copolymerising deuteriated monomers [CD C(CD3)C02Me or CD :C(CD,)CO,H] with the complementary undeuteriated m~nomer.~ '' Six a-CH3 resonances are observed which may be assigned to the twenty possible triads (six i six s and eight h). The 13C1-resonance in polystyrene (C adjacent to the main chain) is of the same general form as the a-'H-resonance in p~ly[flfl-~H~]styrene and shows clear evidence for discrimination of pentad resonances.Free-radically initiated polystyrene is exceptional in being predominantly isotactic in contrast to other free-radically initiated polymers. Anionically initiated polyb-methoxystyrene) shows ten methoxy resonances, reflecting pentad structure.272 The tacticity is dependent both on the initiator and solvent. For polymerisation in THF the one-parameter model applies but for initiation by n-butyl-lithium in toluene a two-parameter model must be used, indicating a penultimate-unit effect. Benzyl vinyl ether is polymerised by BF30Et2 in toluene at - 78 "C to give highly isotactic polymer.273 The side-chain CH2 is sensitive to triads the line 264H. Yuki K. Hatada Y. Kikuchi and T. Niinomi J .Polymer Sci. Part B Polymer 2 b 5 K. Matsuzaki T. Uryu and K. Ito Makromol. Chem. 1969 126 292. 2 6 6 W. L. Lee B. R. McGarvey and F. R. Eirich J . Polymer Sci. Pqrt C Polymer Sym-2 6 7 S . Okuzawa H. Hirai and S. Makishima J . Polymer Sci. Part A-I Polymer Chem., 2ba M. Maruhashi and H. Takida Makromol. Chem. 1969 124 172. 2 6 9 Y. Muroga I. Noda and M. Nagasawa J . Phys. Chem. 1969 73 667. " O E. Klesper and W. Gronski J . Polymer Sci. Part B Polymer Letters 1969 7 661. 2 7 1 E. Klesper and W. Gronski J . Polymer Sci. Part B Polymer Letter? 1969 7 727. 2 7 2 H. Yuki Y . Okamoto Y. Kuwae and K. Hatada J . Polymer Sci. Part A-I Polymer 2 7 3 H. Yuki K. Hatada K. Ota I. Kinoshita S. Murahashi K. Ono and Y. Ito J . Polymer Letters 1968 6 753. posia 1967 22 1197.1969 7 1039. Chem. 1969,7 1933. Sci. Part A-I Polymer Chem. 1969 7 1517 148 K . J . Ivin ordering being i < h < s. The proportions of triads require a two-parameter model but when the solvent is changed to 50/50 toluene-nitroethane one para-meter suffices. In poly(methy1 propenyl ether) both the a-methoxy and P-methyl (decoupled from P-hydrogen) resonances are sensitive to dyad structure.274 The spectra of polymers made from various cis-trans monomer mixtures at - 78 "C using BF,0Et2 in toluene as initiator show that the pure trans-isomer would give the threo-meso configuration (3) about both OL and B carbon atoms. OCH3 H OCH3 H I C-I I I -c-c-c-I I I CH3 H CH3 I H (3) With this catalyst the trans-propenyl ethers give crystalline polymer whereas the cis-isomers usually give amorphous polymer.However with the hetero-geneous catalyst Al,(S04)3 ,H2S0 the situation is reversed and presumably it is then the catalyst site rather than the chain end which governs the tacticity of the p r ~ d u c t . ~ 75 trans-2-Chloroethyl propenyl ether gives only 80 % threo-meso structure instead of 100% for the methyl ether.276 Poly(viny1 chloride) prepared by Bu'MgCl-initiation in THF may be separated into two fractions one soluble and one insoluble in cyclohexanone at 25 "C. The 220 MHz spectra of both fractions in o-dichlorobenzene solution at 150 "C exhibit six CH2 triplets which may be assigned to the six tetrad structures.277 The fraction insoluble in cyclohexanone has a high proportion of syndiotactic triads.Free-radically initiated poly(viny1 chloride) contains a much higher proportion of syndiotactic triads when made in butyraldehyde solution com-pared with that made by bulk or emulsion p~lymerisation.~~~ The CH resonance, when decoupled from CH2 consists of three lines in the z order s < h < i of which h and i show further splitting into two and three lines respectively arising from pentad structure^.^ 79 Pyrolysis of copolymers of vinyl chloride and vinyli-dene chloride gives a mixture of benzene chloro benzene m-dichlorobenzene and 1,3,5-trichlorobenzene whose proportions give a quantitative measure of the triad structures in the copolymer.280 These agree with those calculated from the reactivity ratios. 2 7 4 T. Higashimura Y. Ohsumi S. Okamura R.ChiijS and T. Kuroda Makromol. Chem., 2 7 5 T. Higashimura S. Kusodo Y . Oshima A. Mizote and S. Okamura J . Polymer Sci., 7 6 T. Higashimura Y . Ohsumi S. Okamura R. Chiij6 and T. Kuroda Makromol. Chem., 1969 126 87. Part A - I Polymer Chem. 1968 6 251 1 . 1969 126 99. 2'7 Pham-Quang Tho J . Polymer Sci. Part B Polymer Letters 1969 7 103. 2 7 8 J. Millan and G. Smets Makromol. Chem. 1969 121 275. 279 U. Johnsen and K. Kolbe Kolloid-Z. 1967 221 64. 2 8 0 S. Tsuge T. Okumoto and T. Takeuchi Makromol. Chem. 1969 123,-123 Kinetics and Mechanism of Polymerisation 149 Other recent investigations include those on p~ly([a-~H]acrylonitrile),~~ ‘ polymethacrylonitrile,282 poly(N-vinyl c a r b a ~ o l e ) ~ ~ ~ poly(4methylhex- 1-ene ~ulphone),~ 84 pol y( 5-methylhep t- 1 -ene ~ulphone),~ 84 poly( [2-2H]propylene sul-phide) 30 poly( [2-2H]propylene ~ x i d e ) ~ 8 5 and poly(t-b~ty1[2-~H]ethylene oxide).286 In the last case the CH2 spectrum of the amorphous polymer consists of four AB quartets probably arising from isotactic syndiotactic and two heterotactic triads (which are not identical for a polymer with a three-atom main-chain repeat unit).Template Polymerisation.-The search continues for synthetic analogues of natural polymer reproduction. The first such template polymerisation pro-ceeding by a radical mechanism has been reported.287 Polyethyleneimine is the substrate on to which acrylic acid molecules are adsorbed and then polymerised. A number of cases have been reported of polymers such as poly(vinylimidazo1e) which catalyse the solvolysis of small molecules through attachment to the p ~ l y r n e r .~ ~ ~ - ~ ~ ’ 2 n 1 K. Matsuzaki T. Uryu M. Okada and H. Shiroki J . Polymer Sci. Part A-1 Polymer Chem. 1968,6 1475. 2 n 2 H. Hirai T. Ikegami and S. Makishima J . Polymer Sci. Part A-1 Polymer Chem., 1969 7 2059. 2 8 3 S. Yoshimoto Y . Akana A. Kimura H. Hirata S. Kusabayashi and H. Mikawa, Chem. Comm. 1969,987. ’ ~ 3 ~ R. Bacskai L. P. Lindeman D. L. Ransley and W. A. Sweeney J . Polymer Sci., Part A-1 Polymer Chem. 1969 7 247. H. Tani N. Oguni and S. Watanabe J . Polymer Sci. Part B Polymer Letters 1968,6, 577. 2 8 6 H. Tani and N. Oguni J . Polymer Sci. Part B Polymer Letters 1969,7 803. ’” C. H. Bamford and Z . Shiiki Polymer 1968 9 595. C. G . Overberger R. Covett J. C. Salomone and S. Yaroslavsky Macromolecules, 1968 1 331. 2 8 9 C. Aso T. Kunitake and F. Shimada J . Polymer Sci. Part B Polymer Letters 1968, 6 467. 290 T. Kunitake F. Shimada and C. Aso Makromol. Chem. 1969 126 276. ”‘ A. S. Lindsey Macromol. Rev. 1969 C3 1