Stable Carbenium Salts

As we have mentioned in Chap. II, it is possible to prepare stable, fully ionised crystalline salts of carbocations provided the positive charge is sufficiently delocalised over the cationic moiety and the nucleophilicity of the anion is sufficiently low. Within this family of compounds, we will limit our attention to two major groups of salts with which most of the fundamental studies on cationic polymerisation of olefins has been carried out the triphenylmethylium (trityl) and the cycloheptatrienylium (tropylium) salts. It has been a common practice to assume that if these salts are fully ionised in the solid state, they will also be completely dissociated in solution. Thus, the equilibria 

As for the equilibria between ion pairs and free ions in solutions of these salts, we refer to the values collected by Ledwith and Sherrington in their recent review  

Their Tables 2 and 3 show that except for the work of Lichtin and Pappas, carried out in liquid sulphur dioxide, the values of the disarciation constant for reliable studies in chlorinated hydrocarbons are quite independent of the counterion. Thus, K j = (1 — 2) 

X 10 M at 25°C for the trityl salts, and = 0.5 x 10 M at 0°C for the tropylium salts. Note that those Tables contain some misprints and the ordinal references should be consulted for the correct values. A recent study by Gogolczyk et al. is particularly interesting in this context because while it confirms the above value of for trityl salts in dichloromethane, it also reports for the first time values of the dissociation constant in nitromethane and in mixtures of the two solvents. The value in pure nitromethane is more than a hundred times higher than that in dichloromethane, as expected from the considerable increase in polarity. Obviously then, the proportion of free ions in CH3NO2 is very high and at salt concentrations below 10 M they will be the predominant species. It must be emphasised that these ionic salts are only sparingly soluble in solvents of low dielectric constant such as carbon tetrachloride and therefore polymerisations caimot be carried out in these media. 

Having established some general criteria concerning the state of these salts in the solvents suitable for cationic polymerisation, it is advisable briefly to survey the most relevant unwanted reactions of these q ecies which could arise in the course of their use as initiators. 

---

Velichkova, R. S., Toncheva, V. D., and Panayotov, I. M., Macromonomers fi om vinyl and cyclic monomers prepared by initiation with stable carbenium salts, J. Polym. Sci., Part A Polym. Chem.. 25. 3283-3292, 1987. 

Unravelling the mechanism(s) of initiation, whereby chain carriers are formed in a system. This topic includes such studies as the mode of proton transfer frran a Br nsted acid to a monomer, the steps involrred in the interaction of a Lewis acid with an olefin (in the absence and/or in the presence of a cocatalyst), the reaction of stable carbenium salts with variais vinylic moieties, etc. 

The above observations are obviously not intended as a rejection of this very useful technique. We shall have numerous opportunities to underline the important contributions made by its judicious exploitation. Perhaps the most interesting application of conductivity in cationic polymerisation relates to measurements in model systems such as the study of the self-ionisation of initiators, the interactions of Lewis acids with cocatalysts, and of course the extent of dissociation of stable carbenium salts. 

The application of ultraviolet and visible spectroscopy to the identification and measurement of carbenium ions derived from aromatic and dienic monomer has already been discussed (see Sect. II-G-2). The use of this technique to monitor stable carbenium salts is also well known. We have finally stressed in a preceding section that the fate of certain anions could be followed spectrophotometrically during a cationic polymerisation. The limits of detection allowed by the values of the extinction coefficients of all these species and by the sensitivity of present-day instruments is 10 to 10 M. 

Ledwith and Sherrington have already discussed some of the basic aspects of polymerisation processes involving stable carbenium salts as initiators and reviewed the literature in this field up to 1974. Since then, some interesting and pertinent material has been published which makes it necessary to reassess the interpretation of the older work. We therefore cover this whole area of cationic polymerisation from its beginning. 

Depending on its nucleophilicity, when an alkenyl monomer is added to a solution containing a stable carbenium salt, diffo ent situations can arise, namely ... 

The first detailed study of a cationic polymerisation of vinyl ethers induced by stable carbenium salts was reported in 1971 by Bawn et Isobutyl vinyl ether was polymerised with trityl and trc ylium hexadiloroantimonates and trityl fluoroborate. From calorimetric measurements of the rate of polymerisation, it was concluded that all the initiator used was consumed roon after mixii and the assumption was made that an equal number of active species was formed in this fast initiation reaction. Propagation rate constants were thus obtained and attributed to the action of free ions. It was als) claimed that no significant termination took place during the polymerisation ce successive monomer additions produced polymerisations having the same propagation rate constant. Later work performed in the same laboratory on other vinyl ethers... 

The study of the polymerisation of cyclopentadiene by stable carbenium salts has been the exclusive domain of S walt and his group. In 1967, they reported the first observations on a system involving trityl hexachloroantimonate in methylene chloride between — 70 and 20°C Because of the interesting features of this polymerisation, the work was pursued by kinetic and mechanistic investigations. In I969 it was shown that initiation took place by direct addition and that transfer and termination were imimpor-tant particularly in the first stipes of the process. The kinetics of initiation were followed by visible spectroscopy and provided furthw evidence for a one-to-one reaction. Thus,... 

Isrprene has been polymerised by both tropylium and trityl salts in nitrobenzene by Gaylord and Svestka who also found that less polar media were ineffective for this reaction. The complex behaviour of these systems did not allow any fundamental conclusion to be reached on the medianism of initiation except that some incorporation of trityl groups (but not of tropylium groups) was detected in the polymers. Interestingly, isoprene seems to display a limiting nucleophilicity and sit at the boundary between monomers polymerisable and non-polymerisable by these stable carbenium salts. 

The extremly high susceptibility of N-vinylcarbazole towards cationic polymerisation is weU illustrated by the ease and rapidity with which this monomer is converted into polymer even at low temperatures under the action of very small concentrations of stable carbenium salts. We have already mentioned that oxonium salts are inactive for the polymerisation of olefins. Turchi et al. tried triethyloxonium hexafluorophos-phate as a possible initiator for the polymerisation of this monomer and proved in a series of elegant experiments that this salt is not an initiator. Polymerisation was observed but caused by traces of PFs present in the system. The importance of this paper stems from its conclusive evidence showing that even the most basic among the vinylic monomers is insensitive to oxonium salts. 

The work of Aso et al. on this monomer has been discussed at lei th by Ledwith and Sherrington in their review This investigation does not contribute substantially to the basic understanding of the mechanism of initiation by stable carbenium salts and no further comments are therefore necessary in the present context. 

The polymerisation of this monomer by stable carbenium salts was first reported by Sauvet et who only remarked that at —70°C in methylene chloride, about 30% yield was obtained in two hours with a 10 M concentration of trityl hexachloroanti-monate. In the same year, Higashimura et al. measured the rate of interaction between trityl pentachlorostannate and a-methylstyrene at 30-60°C in ethylene chloride and mixtures of this solvent with benzene. They monitored the disqipearance of the characteristic visible bands of the trityl ion as the reaction proceeded. Good first order plots were obtained and the external order in monomer was found to be unity so that kinetically the reaction was bimolecular with a kj of 1.3 min at 30°C in pure eth-... 

The major drawback concerning the use of these stable carbenium salts as initiators for the polymerisation of vinylic monomers, is of course their inabflity to attack certain olefins such as isobutene. The search for alternative catalysts within this famfly has resulted in the remarkable success obtained by Plesch already quoted. Hopefully more such new initiators will become avaflable after this breakthrough. 

Within the context of stable carbenium salts initiatron, we already examined a very interesting and successful study on the block copolymeriation of a-methylstyrene with cyclopentadiene performed by Vairon and Villesange (see Sect. V-A-4-b). The preparation of the product required three basic operations (i) the living anionic polymerisation of a-methylstyrene to give monodisperse macromolecules, (ii) transformation of their end groups into stable carbocationic moieties, and (iii) initiation of the polymerisation of cyclopentadiene from these active ends under conditions of minimal transfer and termination reactions. Thus, the macroinitiators in the second polymerisation were generated by a controlled anionic polymerisation and allowed tiie synthesis of a triblock near-isomolecular copolymer. 

---