Water cocatalyst

To avoid high resin chloride content associated with the use of high concentrations of aluminum trichloride, a ttialhylalurninum—water cocatalyst system in a 1.0 0.5 to 1.0 mole ratio has been used in conjunction with an organic chloride for the polymerization of P-pinene (95). Softening points up to 120°C were achieved with 1—3 Gardner unit improvement in color over AlCl produced resins. 

The Lewis acids must be used with a protonic cocatalyst such as water or methanol which generates protons through the following kinds of equilibria ... 

Friedel-Crafts (Lewis) acids have been shown to be much more effective in the initiation of cationic polymerization when in the presence of a cocatalyst such as water, alkyl haUdes, and protic acids. Virtually all feedstocks used in the synthesis of hydrocarbon resins contain at least traces of water, which serves as a cocatalyst. The accepted mechanism for the activation of boron trifluoride in the presence of water is shown in equation 1 (10). Other Lewis acids are activated by similar mechanisms. In a more general sense, water may be replaced by any appropriate electron-donating species (eg, ether, alcohol, alkyl haUde) to generate a cationic intermediate and a Lewis acid complex counterion. 

Strong protonic acids can affect the polymerization of olefins (Chapter 3). Lewis acids, such as AICI3 or BF3, can also initiate polymerization. In this case, a trace amount of a proton donor (cocatalyst), such as water or methanol, is normally required. For example, water combined with BF3 forms a complex that provides the protons for the polymerization reaction. 

C ) using AICI3 cocatalyzed with a small amount of water. The cocatalyst furnishes the protons needed for the cationic polymerization ... 

A variety of initiators have been used for cationic polymerization. The most useful type of initiation involves the use of a Lewis acid in combination with small concentrations of water or some other proton source. The two components of the initiating system form a catalyst-cocatalyst complex which donates a proton to monomer... 

Fell and Bari (89) also studied the rhodium-catalyzed reaction. A rho-dium-N-methylpyrrolidine-water catalyst system was very effective for producing the propane-1,2-diol acetate directly. The best yields (>90%) of product of about 9 1 alcohol aldehyde ratio were obtained in the region of 95°-l 10°C. This range was very critical, as were other reaction parameters. Rhodium alone gave the best yield of aldehyde (83%) at 60°C. Triphenylphosphine as cocatalyst induced the decomposition of the aldehyde product. 

Co-catalysts other than water. Trichloro- and monochloro-acetic acids, when used as cocatalysts, induced instantaneous polymerisation at -140°. With the following co-catalysts the rate of polymerisation at -78° decreased in the order acetic acid > nitroethane > nitromethane > phenol > water [75a]. Since this is also the sequence of the acid dissociation constants of these substances in water, it appears that the catalytic activity , as shown by the rate of polymerisation, is correlated with the acidity of the cocatalyst in aqueous solution. Flowever, there are two reasons for questioning the validity of this correlation. 

Plesch, Polanyi, and Skinner [28] found that HC1, S02, C02, EtOH, and EtzO were not cocatalysts, and the last two substances were shown to be inhibitors in that the addition of moist air to a solution containing them did not induce polymerisation. The search for co-catalysts other than water led to the discovery that trichloroacetic acid, sulphuric acid, and 20 percent oleum would act as a co-catalyst to titanium tetrachloride in hexane at about -75°, though none of these acids alone showed any catalytic activity under these conditions [9, 71]. 

With chloroform as solvent without added co-catalyst (water being the putative cocatalyst), and with [TiClJ = (3 - 48) x 10"3 mole/1, and also with CC13C02H as cocatalyst ([TiCl4]/[CCl3C02H] 3-4, [TiClJ = (5-8) x 10 3 mole/1) Imanishi et al. [79] found that the Mayo monomer plots were linear at -20°, -50°, and -78°. The values of kjkp calculated from these are shown in Table 4. With CC13C02H as co-catalyst kjkp is almost the same as with water. ... 

It is useful to note here a fundamental distinction between cationic and anionic polymerizations (including Ziegler-Natta systems). In the latter, residual water merely inactivates an equivalent quantity of catalyst, whereas in the former water may be a cocatalyst to the metal halide catalyst in excess it may decrease the rate by forming catalytically inactive higher hydrates and in very many systems it, or its reaction product(s) with a metal halide, act as extremely efficient chain-breakers, thus reducing the molecular weight of the polymers (see sub-section 5.4). 

Thus, our experiments with isobutene show that for this monomer the Gantmakher and Medvedev theory is not applicable they also show that for isobutene CH2C12 is not a cocatalyst to TiCl4. However, it was still possible that the polymerisation of styrene at the lowest water concentration was due not to residual water, but either to co-catalysis by the solvent or to direct initiation by the Gantmakher and Medvedev mechanism. However, since we found the molecular weight to be independent of both the water and the TiCl4 concentration and the rate at low water concentration to be independent of the TiCl4 concentration, these alternatives appeared unlikely. 

Water, alcohols, ethers, or amines can cause inhibition of ionic polymerization. However, these substances can act in different ways according to their concentration. For example, in polymerizations initiated by Lewis acids (BF3 with isobutylene) or organometallic compounds (aluminum alkyls), water in small concentrations behaves as a cocatalyst, but in larger concentrations as an inhibitor (reaction with the initiator or with the ionic propagating species). 

---