In the context of biomass valorization by heterogeneous catalysis, computational chemistry is key to provide guidance to establish the nature of the active sites in combination with experimental characterizations. Then, the reaction mechanism can be studied to determine the rate determining transition state and intermediate and further design in silico better catalysts. We implemented this approach in several reactions involving alcohols that are key in the shift from a petroleum chemical feedstock to a biomass-based feedstock. Firstly, we focused on liquid phase alcohol oxidation by oxygen, a reaction that generally requires an alkaline environment, which is detrimental to the atom economy of the process since it generates the carboxylate salt instead of the carboxylic acid. We proposed a model of metal/basic water interface that includes the adsorption of hydroxide anion.
It charges the metallic surface and modifies its catalytic activity. This model was first validated comparing the predicted activity of Au and Pt in presence and in absence of a base, and then used oxidation of alcohol ethoxylates by bimetals. Then, we switched to gas phase dehydration of C3 and C4 alcohols using phosphate-based catalysts. The modeling of the surfaces was based on experimental characterizations. The molecular coverage of water on the surface in function of the pressure and temperature was established using ab initio thermodynamic. The simulations of infrared spectra of CO, NH3 and C2H2 adsorption allowed us to identify the acido-basic sites which play an important role in the reaction mechanism investigation that followed.
Keywords: reaction mechanism, DFT, heterogeneous catalysis, alcohol oxidation, metal phosphate, bimetal, basic environment