Following the synthesis of the first stable disilene, tetramesityldisilene (1), in 1981,1 many reactions of disilenes have been chronicled.2 However, information about the mechanisms of these reactions is still rather scarce. One of the most studied mechanisms is that of the 1,2-addition reaction of alcohols to the SidSi bond, but also this mechanism is not fully understood.2 Addition of alcohols to (E)-1,2-di-tert-butyl-1,2-dimesityldisilene gives a 1:1 mixture of the two diastereoisomeric alkoxysilanes, suggesting a stepwise mechanism involving a zwitterionic intermediate.3 In contrast, theoretical calculations (at the RHF/6-31G* level) predicted a four-center concerted transition state, leading to syn-addition, for the gas-phase addition of water to disilene.4 In agreement, Sekiguchi, Maruki, and Sakurai (SMS) found recently that the addition of 2-propanol and of tert-butyl alcohol to the transient disilenes (E)and (Z)-1,2-dimethyl-1,2-diphenyldisilene occurs with high syndiastereoselectivity.5 With ethanol the addition occurs with high syn-diastereoselectivity at low ethanol concentrations, but at a high ethanol concentration a 1:1 mixture of the addition diastereoisomers was obtained. No significant kinetic deuterium isotope effect was found when ethanol-d1 was used. SMS concluded that the rate-determining step of the addition reaction involves a nucleophilic attack of the alcoholic oxygen on the coordinatively unsaturated silicon center, forming a fourmembered intermediate which reacts further, either intramolecularly to give syn-addition products or intermolecularly (at high alcohol concentrations) to give anti-addition products.5 This contrasts with the common addition mechanism to olefins, which generally involves a rate-determining electrophilic step. To gain further insight into the mechanism of alcohol addition reactions to disilenes, we have carried out the first Hammetttype6 study for compounds with multiple bonds to silicon. We studied the kinetics of the addition reactions of seven paraand meta-substituted phenols to 1. The resulting Hammett plot has a concave shape with a minimum for phenol, indicating a change in mechanism from a rate-determining nucleophilic step for electron-rich phenols, to a rate-determining electrophilic step for electron-poor phenols. Kinetic isotope effect measurements support this interpretation. The kinetics of the addition of phenols to 1 was followed spectroscopically by measuring the decay rate of the absorption at 420 nm (corresponding to the SidSi bond in 11).7 A large access of the phenol (i.e., 50-6000-fold larger than that used for 1) was used to enforce pseudo-first-order kinetics. Plots of the pseudo-first-order rate constants (kobs) vs phenol concentrations are shown in Figure 1.8 The excellent correlations between the experimental measurements and first-order kinetics and the linear correlations in Figure 1 imply that the addition reactions are indeed first order in both the disilene and the phenol.9 The second-order rate constants k for the addition of the various phenols to 1 derived from Figure 1 are summarized in Table 1.10 Comparison with the results of SMS shows that the rate of addition of phenols to 1 (k ≈10-4-10-2 M-1 s-1) are much slower compared with the rates of addition of alkyl alcohols to (E)and (Z)-1,2-dimethyl-1,2-diphenyldisilene and 1,2,2-trimethyl-1-phenyldisilene (k ≈ 107-108 M-1 s-1).5 The larger steric bulk of the mesityl substituents in 1 may account for this large reactivity difference. A Hammett plot of log(kX/kH) vs σ (eq 1),6 where kX and kH are the second-order rate constants for the addition to 1 of the substituted and the nonsubstituted phenols, respectively (Table 1), is shown in Figure 2.
Read full abstract