Mechanistic insight into the cyclohexene epoxidation with VO(acac)(2) and tert-butyl hydroperoxide

M. Vandichel, K. Leus, P. Van der Voort, M. Waroquier, V. Van Speybroeck
Journal of Catalysis
294, 1-18
2012
A1

Abstract 

The epoxidation reaction of cyclohexene is investigated for the catalytic system vanadyl acetylacetonate (VO(acac)2) with tert-butyl hydroperoxide (TBHP) as oxidant with the aim to identify the most active species for epoxidation and to retrieve insight into the most plausible epoxidation mechanism. The reaction mixture is composed of various inactive and active complexes in which vanadium may either have oxidation state +IV or +V. Inactive species are activated with TBHP to form active complexes. After reaction with cyclohexene, each active species transforms back into an inactive complex that may be reactivated again. The reaction mixture is quite complex containing hydroxyl, acetyl acetonate, acetate, or a tert-butoxide anion as ligands, and thus, various ligand exchange reactions may occur among active and inactive complexes. Also, radical decomposition reactions allow transforming V+IV to V+V species. To obtain insight into the most abundant active complexes, each of previous transformation steps has been modeled through thermodynamic equilibrium steps. To unravel the nature of the most plausible epoxidation mechanism, first principle chemical kinetics calculations have been performed on all proposed epoxidation pathways. Our results allow to conclude that the concerted Sharpless mechanism is the preferred reaction mechanism and that alkylperoxo species V+IVO(L)(OOtBu) and V+VO(L1)(L2)(OOtBu) species are most abundant. At the onset of the catalytic cycle, vanadium +IV species may play an active role, but as the reaction proceeds, reaction mechanisms that involve vanadium +V species are preferred as the acetyl acetonate is readily oxidized. Additionally, an experimental IR and kinetic study has been performed to give a qualitative composition of the reaction mixture and to obtain experimental kinetic data for comparison with our theoretical values. The agreement between theory and experiment is satisfactory.

Open Access version available at UGent repository