Novel Metal-Incorporated Mesoporous Silicates as Catalysts for Liquid Phase Ethylene Epoxidation with H2O2 Oxidant

Ethylene Oxide (C2H4O, EO), is one of the essential building blocks for a variety of consumer products, including plastics, fibers, sports gear, paints, coolants and detergents. Currently, the worldwide production of EO exceeds 20 Mt/yr, making it one of the high volume chemical intermediate in the industry. At present, EO is manufactured by catalytic processes using alumina supported Ag catalysts at relatively high temperature (T = 200-260 °C) and pressure (P = 10-30 bar). This process converts approximately 10-15% feedstock to CO2, which represents a monetary loss of approximately ;Researchers at the University of Kansas Center for Environmentally Beneficial Catalysis (CEBC) reported a liquid phase ethylene epoxidation process that eliminated ethylene burning to CO2. In this process, the ethylene gas is compressed under pressure and dissolved in a liquid reaction medium containing 50 wt.% H2O2/H2O as oxidant, pyridine N-oxide as promoter and methyl trioxorhenium (MTO) as homogeneous catalyst. This process produces EO with near complete selectivity with no CO2 detected in either the liquid or gas phases under relatively mild conditions (T = 35 °C, P = 50 bar). Further, the H2O 2 is fully utilized toward EO formation. Preliminary economic analysis suggests that the Re-based EO process has the potential to be competitive with the conventional process if the MTO catalyst remains active, selective and stable for at least one year at a leaching rate of approximately 0.11 lb MTO/h. This stringent performance metric coupled with the low abundance and high cost of Re metal pose significant challenges for the commercialization of Re-based technology. This motivated further investigation on alternative ethylene epoxidation catalysts that are relatively inexpensive. Literature survey identified tungsten (W) and niobium (Nb) based catalysts as possible alternatives. Indeed, W, Nb based catalysts catalyzed ethylene epoxidation with significant activity with aqueous hydrogen peroxide (H2O 2) as the oxidant and methanol as solvent under mild operating conditions (35 °C and 50 bar) where CO2 formation is avoided. The ethylene productivity on these catalysts (up to 4,304 g EO/h/kg Nb) either matches or surpasses that observed on the conventional Ag-based heterogeneous catalyst (with O2 as oxidant) as well as a Re-based homogeneous catalyst (with H2O2 as oxidant). The measured EO productivity over Nb-TUD-1 materials (342-4304 g EO/h/kg Nb) spans a greater range than those observed with Nb-KIT-6 (234-794 g EO/h/kg Nb), Nb-KIT-5 (273-867 g EO/h/kg Nb) and Nb-MCM-48 (71-219 g EO/h/kg Nb) materials at similar operating conditions. However, significant H2O2 decomposition (H2O 2 utilization to form EO < 78.7%) and Nb leaching (up to 88.7% after a 5 h batch run) were observed in all cases. Complementary DFT-based computational studies by collaborators in the KU Department of Chemistry, using minimalist models of the catalytically active metal site and their reaction with H 2O2, suggest that a reaction mechanism involving Bronsted acid sites, particularly niobium hydroxide sites, could contribute to H 2O2 decomposition and metal leaching. Indeed, Nb-TUD-1 catalysts prepared with lower metal loadings show significantly reduced Bronsted acidity and display markedly higher H2O2 utilization and EO productivity. These results provide valuable guidance for future computational and experimental investigations aimed at developing stable and practically viable catalysts.