The epoxidation of dienes, alpha,beta-unsaturated ketones, alpha,beta-unsaturated esters and vinyl esters by dimethyldioxirane

A study of the epoxidation of dimes (2--8), alpha,beta-unsaturated ketones (9--25), alpha,beta-unsaturated esters ( 26--33) and vinyl esters (34--40) by dimethyldioxirane (1) way accomplished via product studies, kinetic approaches and computational calculations. Product studies showed that the epoxidation by dimethyldioxirane in all cases gave the corresponding epoxides with the retention of stereochemistry in high yields. The second order rate constants were obtained employing UV techniques under pseudo-first order conditions. Kinetic studies for the epoxidation of the three conformation isomers ( cis,cis-, cis,trans- and trans,trans-) of 2,4-hexadienes showed they have similar k2 values. The epoxidation for cyclic alpha,beta-unsaturated ketones was shown to be 104 slower than that for the dienes. Within the series, the six-membered ring compounds had higher k2 value than the five-membered ring compounds. 2-Methyl-substituted compounds had higher k2 values than those with 3-methyl-substitution. Cisoid structures showed pronouncedly enhanced k2 values compared to those with transoid structures. The k2 values for vinyl esters were found to be the largest, substantially greater than the k2 values for the epoxidation of alpha,beta-unsaturated ketones and alpha,beta-unsaturated esters by dimethyldioxirane. Activation parameters for the epoxidation of selected cyclic alpha,beta-unsaturated ketones (9--11 and 13--15) were determined using Arrhenius method. All the compounds in this series were found to have very similar values for DeltaS≠ (-31.0 eu to -32.8 eu). The ab initio calculations with B3LYP at 6-31G basis set for the epoxidation of selected compounds (9--11 and 13--15) predicted the transition states to have a spiro geometry. The calculation results showed an excellent correlation between experimental activation energies and calculated energies. Since DeltaS≠ was shown to be essentially constant, one can successfully model the relative reactivity by normalizing the calculated energies in a series relative to the parent in the series.