(690c) Evolution and Distribution of MoOx Species in Mo-Oxide-Impregnated H-ZSM-5 Catalyst for Methane Dehydroaromatization (MDA)

Methane dehydroaromatization (MDA), a non-oxidative direct process for producing benzene and other aromatics from methane, has attracted great attention in recent decades. Mo-oxide-impregnated H-ZSM-5, the most widely studied catalyst for MDA, has demonstrated effective catalytic properties toward benzene formation. Nonetheless, rapid catalyst deactivation is a great challenge for MDA commercialization. Among many unsettled questions underlying MDA, understanding the speciation of the Mo-oxide catalyst precursors that reduce into the catalytically active oxycarbide and carbide species is of great importance. Through previous work, we have demonstrated that Mo-oxide precursors exist in several stoichiometries including MoO₂OH⁺, MoO₂²⁺, and Mo₂O₅²⁺, and the distribution may be influenced by catalyst synthesis technique (incipient wetness impregnation (IWI) or physical mixing (PM)), metal loading, and the zeolite acidity. Each unique motif anchors differently at the Brønsted acidic sites (BAS), and the anchoring temperature is dependent on the method of impregnating the metal oxide into the zeolite. In complement to our endeavors, we investigated the distributions of each Mo-oxide species as a function of zeolite acidity and metal loading using thousands of models of Mo/H-ZSM-5 with varying acidity and Mo loading using Monte Carlo methods. We have compared the results of our statistical models with water desorption quantification from temperature-programmed oxidation (TPO) experiments. The mechanism of anchoring Mo-oxides at the BAS and the transformation of anchored Mo-oxo species (MoO₂OH⁺) to other types of Mo-oxide species, MoO₂²⁺ and Mo₂O₅²⁺, has also been simulated using climbing image nudged elastic band methods. We have further investigated anchoring of Mo-oxides (MoO₂²⁺ and Mo₂O₅²⁺) at Al site pairs at various separations using DFT, identifying numerous anchoring motifs not conventionally studied, and considered these models in MC simulations. Through this work, we have built an understanding at the atomistic level that facilitates interpretation of our experimental characterizations of these systems.