(407b) First-Principles Design of Metal Alloy Catalysts for Electrocatalytic Methanol Oxidation

The direct methanol fuel cell (DMFC) may provide a lightweight power source for portable electronics and military applications. The design of improved anode electrocatalysts is needed to make the application of DMFCs a reality. Overpotential loses due to anode poisoning by the carbon monoxide intermediate of methanol oxidation as well as dissolution of anode alloy components are two factors limiting the performance of the DMFC. In this paper, we describe first principle theoretical results aimed at understanding the elementary steps that control the electrocatalytic oxidation of methanol at the anode of the DMFC. Elucidating the elementary steps and the potential inhibition by CO could help to offer potential solutions toward increasing catalytic activity and fuel cell performance. We have developed a periodic density functional theory based method to model the electrocatalytic environment that includes both the effects of solvation and an electric field on the elementary step energetics. This method is applied to clarify key parameters in the design of alloy materials for the DMFC anode. Once the elementary reaction kinetics are determined, including the complexities of the electrocatalytic environment, a simplified model system is used to test design improvements to the anode catalyst. First principle quantum mechanical calculations have proven to be a valuable complement to experimental efforts in the elucidation of surface chemistry. We carry out periodic gradient corrected density functional theoretical calculations to analyze the dual path of methanol oxidation over well-defined Pt surfaces. The potential dependent reaction energetics for formation of the CHxO or CO intermediate are illustrated. At potentials of interest for the DMFC, the path through a carbon monoxide intermediate dominates. A substantial overpotential is associated with CO oxidation on pure Pt surfaces. The addition of ruthenium to the Pt anode reduces the overpotential to CO oxidation. Model single crystal surfaces of Pt-Ru alloys are employed to understand the effect of ruthenium addition on the elementary step reaction energies and activation barriers for CO oxidation. The addition of surface Ru atoms reduces the overpotential of water activation, indicating a bifunctional mechanism, in which methanol is oxidized over Pt atoms and water is activated over Ru atoms, contributes to the reduction in overpotential. An overlayer model, in which a monolayer of Pt is added above a Ru substrate, shows a substantially reduced binding of carbon monoxide and reduces the equilibrium potential of the overall CO oxidation reaction. Furthermore, the activation barrier to the carbon monoxide-hydroxyl coupling step is significantly reduced in the overlayer model, indicating the ligand effect, in which the proximity of Ru weakens the binding of CO to Pt, also acts to reduce the overpotential in the alloy system. A combined model system, with a surface alloy of Pt and Ru over a Ru bulk substrate, combines the bifunctional and ligand effects to provide a maximum reduction in the CO oxidation overpotential. Following detailed consideration of methanol oxidation over Pt and Pt/Ru alloys, a simpler model examining only gas-phase reaction energetics over various alloy metal surfaces was employed for rapid screening of catalyst materials. Bulk alloys, surface alloys, overlayers, and ternary alloy systems are evaluated, with first principles methods used to determine their ability to oxidize methanol to carbon monoxide, activate water, and oxidize carbon monoxide from the surface. The reaction energies determined in the simplified model are extrapolated as a function of potential, and an Evans-Polanyi relationship is used to determine activation barriers. These results are used as input to a deterministic micro-kinetic model used to evaluate electrocatalytic performance and identify promising alloy compositions and configurations.