Attempted catalytic hydroboration of (4-methoxyphenyl)ethene 1 with R,R-3-isopropyl-4-methyl-5-phenyl-1,3,2-oxazaborolidine 6 proceeded extremely slowly relative to the 3-methyl analog 2 derived from rho-pephedrine when diphosphinerhodium complexes were employed. With phosphine-free rhodium catalysts, especially the 4-methoxyphenylethene complex 7, the reaction proceeded rapidly and quantitatively to give only the corresponding (E)-vinylborane 9 and 4-methoxyethylbenzene 8 in equimolar amounts. Isotopic labeling and kinetic studies demonstrated that this reaction pathway is initiated by the formation of a rhodium hydride with subsequent reversible and regiospecific H-transfer to the terminal carbon, giving an intermediate which adds the borane and then eliminates the hydrocarbon product. Further migration of the secondary borane fragment from rhodium to the beta-carbon of the coordinated olefin occurs, followed by Rh-H beta-elimination which produces the vinylborane product and regenerates the initial catalytic species. When the same catalytic reaction is carried out employing catecholborane in place of the oxazaborolidine, an exceedingly rapid turnover occurs. The products are again 4-methoxyethylbenzene and the (E)-vinylborane 23 but accompanied by the primary borane 24 in proportions which vary with the experimental conditions. None of the secondary borane, which is the exclusive product when pure ClRh(PPh(3))(3) is employed as catalyst, is formed. The product variation as a function of initial reactant concentration was fitted to a model in which the rhodium-borane intermediate in the catalytic cycle undergoes two competing reactions-beta-elimination of Rh-H versus addition of a further molecule of catecholborane. The model demonstrates that a kinetic isotope effect of 3.4 operates in the beta-elimination step, but none is evident in the addition of catecholborane B-D to rhodium. A similar analysis was successfully applied to the catalytic hydrosilylation of 4-methoxystyrene, with HSiEt(3), again employing the phosphine-free rhodium catalyst 7; the product distribution between primary silane 29 and vinylsilane 28 was successfully predicted. The results intimate that silation (i.e., the formation of vinylsilanes under the conditions of catalytic hydrosilylation) can best be explained by a Rh-H based mechanistic model rather than the commonly assumed variant on the Chalk-Harrod catalytic cycle. They provide an explanation for the ''oxygen effect'' on the rate of Rh-catalyzed hydrosilylations.
|Number of pages||13|
|Journal||Journal of the American Chemical Society|
|Publication status||Published - 9 Feb 1994|