Olefin Metathesis by Supported Metal Oxide Catalysts

Abstract Olefin metathesis is considered to be a green route to production of olefins due to its high selectivity towards desired products. Due to their ease of preparation and catalyst lifetimes, heterogeneous supported metal oxide catalysts such as ReOx, MoOx and WOx are used at large scale industrial applications. Despite decades of catalysis research, the exact nature of catalytic active sites, reaction intermediates and kinetics are not well understood because of lack of modern characterization techniques in the past and absence of detailed knowledge at the molecular level. Extensive in situ and operando spectroscopy (Raman, UV vis, XAS and IR) experiments, theoretical DFT calculations, steady-state kinetics and temperature programmed surface reaction (TPSR) studies were undertaken for the first time to obtain unprecedented insights about the catalytic active sites (their anchoring sites, coordination and oxidation states), reaction intermediates, olefin adsorption/desorption/reaction and kinetics to unravel the fundamental molecular structure-reactivity relationships. The supported ReOx/Al2O3 system is the most reactive among the heterogeneous supported metal oxide catalysts. The long standing debates surround the nature of ReOx species, number of reactive intermediates and kinetics of this catalytic system. In situ 18O-16O Raman experiments along with in situ XAS and theoretical DFT calculations of the initial catalyst show that rhenia exists on the Al2O3 support as two distinct isolated surface ReO4 species with dioxo coordination. The two structures are related to their anchoring at different surface hydroxyl sites on the alumina support. The surface ReO4-I species on basic alumina sites were found to be stable and difficult to activate with propylene while the surface ReO4-II species on acidic alumina sites were found to be easily activated with propylene. This information allowed for the first time the use of acidic promoters to block formation of the inactive surface ReO4-I species and design catalysts with only active surface ReO4-II species.During activation with propylene, in situ UV-vis and XAS spectroscopy revealed that the surface ReO4-II species become partially reduced, mostly to Re+5 species, by forming the oxygenated CH3CHO and HCHO products (pseudo-Wittig mechanism). Subsequent reaction of the partially reduced Re+5 species with propylene oxidizes rhenia back to reactive Re+7-carbenes (Re=CH2, Re=CHCH3, etc.). The surface Re+7-carbenes are reactive at room temperature and in equilibrium with the gas phase olefins. Consequently, removal of the gas phase olefins significantly diminishes the concentration of reactive surface Re+7-carbenes by about an order of magnitude. This accounts for the low number of reactive intermediates reported in earlier studies that evacuated the catalysts prior to titrating the reactive intermediates.Two types of surface intermediates were found to be present: weakly adsorbed that reacts at room temperature and strongly adsorbed π-complexes that reacts at high temperatures (>100oC). The weakly adsorbed Re-carbenes are dynamic and in equilibrium with the gas phase. The strongly adsorbed π-complexes are not dependent on the gas phase composition and only react with olefins at elevated temperatures. TPSR studies showed that the weakly bound Re-carbenes follows a unimolecular reaction mechanism while the strongly bound π-complexes follow a bimolecular reaction mechanism. The olefin metathesis steady-state kinetics is affected by these intermediates: first-order in propylene partial pressure at low temperatures (<70oC) and second-order in propylene partial pressure at high temperatures (>140oC). C3H6/C3D6 TPSR studies also demonstrated that the rate-determining-step does not involve C-H bond breaking and all olefins share the same rate desorption rate.These new unprecedented insights are able to resolve the confusing claims that existed for decades in literature.