(223ag) A Comparison of Explicit Solvation Models for Density Functional Theory Simulations of Heterogeneously-Catalyzed Systems

Density Functional Theory (DFT) simulations have been used to measure catalytic activities for various homogenously- and heterogeneously-catalyzed systems. DFT is especially useful for calculations relating to fuel cell applications, which currently utilize expensive transition metal catalysts. DFT can be used to screen many possible catalyst material candidates, allowing for an analysis comparing cost and activity while saving time and money associated with experimental testing of many various candidate materials and compositions. DFT simulations have been used to successfully model gas-phase reactions on solid catalysts, such as those used in hydrogen fuel cells (HFCs), but modeling of liquid phase reactions on solid catalysts, such as those relevant to direct methanol fuel cells (DMFCs), is hindered by a lack of knowledge regarding the interactions between solvent molecules and surface-bound species. Thus, in order to effectively use DFT simulations to screen catalyst materials for fuel cell applications with liquid reactant streams, these interactions need to be understood more intimately. Our goal is to use DFT screening to reduce the use of transition metals such as platinum, which is heavily used as a catalyst and currently costs ~$1450 per ounce, to increase the cost-effectiveness, and therefore utilization, of fuel cells.

In this work, we aim to understand the effects of solvation on the oxidation of CO by OH on Pt(111), which has been shown to be the rate-limiting step in DMFCs[1]. Our long-term goals include using the knowledge gained from this work to effectively screen novel, cheaper catalyst materials for DFMCs. We compare various explicit solvation models at several adsorbate coverages of CO and OH to determine the optimal reaction environment for DMFC simulations. Solvation models considered include 2D and 3D ice crystal lattices, as well as 3D liquid-like and vapor-like models. We consider adsorbate coverages up to 1/3^rdof a ML. Our results show that adsorption site, solvation model, water configuration, adsorbate coverage, and inclusion of dispersion forces all have important influences on the water-adsorbate interaction energies. Thus, it is essential to understand these effects in order to effectively screen potential catalyst candidates for heterogeneously catalyzed liquid phase systems such as those relevant for DMFCs.

References:

[1] H.A. Gasteiger, N. Markovic, J.P.N. Ross, E.J. Cairns, J. Phys. Chem.,1994, 98:617.