(520h) Combined Density Functional Theory and Molecular Dynamics Study of Aqueous Reforming of Glycerol on Pt (111) Via Competitive C-H and O-H Dehydrogenation

Glycerol is the most abundant byproduct in biodiesel production, and it can be catalyzed into H₂ and valuable liquid phase products such as 1-propanol, lactic acid, and ethylene glycol through aqueous phase reforming (APR). In this work, we use computational methods to study reactions in Pt (111)-catalyzed glycerol dehydrogenation pathways under aqueous conditions. Studying the reaction in aqueous conditions adds complexities to the analysis, since liquid systems involve large numbers of configurations that contribute to the system energies, meaning that multiple configurations must be sampled for each reaction intermediate. Given that glycerol decomposition can include dehydrogenation, dehydration, decarboxylation, hydrolysis and other processes, to study all of the reactions along all the pathways for glycerol APR would require significant amount of effort, while only a fraction of them are of importance to the overall mechanism. Because of the large reaction network and the need to study it in liquid conditions, we are seeking ways to rapidly estimate energies of reaction intermediates under aqueous conditions, in order to identify the most important intermediates and reactions. We present a two-step process aimed at rapidly estimating the energies of intermediates as well as their interaction energies with liquid water. Firstly, the binding energies of intermediates are estimated in vacuum by a linear scaling relationship obtained from density functional theory (DFT) results. Secondly, the structures are simulated in an explicit liquid water environment using molecular dynamics (MD). In this step, Lennard-Jones plus Coulombic (LJ+C) potentials are used to calculate the water/adsorbate interaction energies. Aqueous phase reaction energies are then calculated by summing the vacuum phase reaction energies and the difference in interaction energies between products and reactants. Using this approach, we evaluate intermediates and reaction steps in the aqueous phase glycerol decomposition reaction network, in an attempt to begin to understand the mechanism of glycerol APR. Our estimations are validated using a fully-DFT procedure previously reported by Bodenschatz, Sarupria, and Getman (J. Phys. Chem. C, 2015). Future work aimed at rapidly estimating the kinetics of these reactions is discussed.