Consequences of Confinement for Alkene Epoxidation with Hydrogen Peroxide on Highly Dispersed Group 4 and 5 Metal Oxide Catalysts

Ti, Nb, and Ta atoms substituted into the framework of zeolite *BEA (M-BEA) or grafted onto mesoporous silica (M-SiO2) irreversibly activate hydrogen peroxide (H2O2) to form pools of metal-hydroperoxide (M-OOH) and peroxide (M-(n(2)-O-2)) species for alkene epoxidation. The product distributions from reactions with Z-stilbene, in combination with time-resolved UV-vis spectra of the reaction between H2O2-activated materials and cyclohexene, show that M-OOH surface intermediates epoxidize alkenes on Ti-based catalysts, while M-(n(2)-O-2) moieties epoxidize substrates on the Nb- and Ta-containing materials. Kinetic measurements of styrene (C8H8) epoxidation reveal that these materials first adsorb and then irreversibly activate H2O2 to form pools of interconverting M-OOH and M-(n(2)-O-2) intermediates, which then react with styrene or H2O2 to form either styrene oxide or H2O2 decomposition products, respectively. Activation enthalpies (AIM for C8H8 epoxidation and H2O2 decomposition decrease linearly with increasing heats of adsorption for pyridine or deuterated acetonitrile coordinated to Lewis acid sites, which suggests that materials with greater electron affinities (i.e., stronger Lewis acids) are more active for C8H8 epoxidation. Values of Delta H-double dagger for C8H8 epoxidation and H2O2 decomposition also decrease linearly with the ligand-to-metal charge-transfer (LMCT) band energies for the reactive intermediates, which is a more relevant measure of the requirements for the active sites in these catalytic cycles. Epoxidation rates depend more strongly on the LMCT band energy than H2O2 decomposition rates, which shows that more electrophilic M-OOH and M-(n(2)-O-2) species (i.e., those formed at stronger Lewis acid sites) give both greater rates and greater selectivities for epoxidations. Thermochemical analysis of Delta H-double dagger for C8H8 epoxidation and adsorption enthalpies for C8H8 within the pores of *BEA and SiO2 reveal that the 0.7 nm pores within M-BEA preferentially stabilize transition states for C8H8 epoxidation with respect to the 5.4 nm pores of M-SiO2, while H2O2 decomposition is unaffected by the differences between these pore diameters due to the small Stokes diameter of H2O2. Thus, the differences in reactivity and selectivity between M-BEA and M-SiO2 materials is solely attributed to confinement of the transition state and not differences in the identity of the reactive intermediates, mechanism for alkene epoxidation, or intrinsic activation barriers. Consequently, the rates and selectivities for alkene epoxidation reflect at least two orthogonal catalyst design criteria-the electronegativities of the transition metal atoms that determine the electronic structure of the active complex and the mean diameters of the surrounding pores that can selectively stabilize transition states for specific reaction pathways.