High Throughput Experimental and Theoretical Predictive Screening of Materials − A Comparative Study of Search Strategies for New Fuel Cell Anode Catalysts

Abstract

A comparative study of experimental and theoretical combinatorial and high-throughput screening methods for the development of novel materials is presented. Both methods were applied to the development of new anode fuel cell alloy catalysts with improved CO tolerance. Combinatorial experimental electrocatalysis was performed on a 64-element electrode array. Sputter-deposited ternary thin-film electrocatalysts of composition PtRuM (M = Co,Ni,W) were screened in parallel for their methanol oxidation activity, and their individual geometric and specific chronoamperometric current density were monitored and evaluated against standard PtRu catalysts. Density functional theory calculations of a variety of model ternary PtRuM alloy catalysts yielded detailed adsorption energies and activation barriers. Feeding these thermodynamic and kinetic data into a simple micro kinetic model for the CO electro oxidation reaction, the relative activities of a number of PtRuM ternary alloys were calculated. The experimental and theoretical computational results reveal very similar trends in electrocatalytic activity as a function of alloy composition; they also point at similar ternary PtRuM alloys as candidates for improved anode catalysts for low-temperature fuel cells.

Bifunctional anode catalysts for direct methanol fuel cells

“…[17][18][19][20] Guided by first principle predictions of the CO tolerance of ternary Pt-Ru-M alloys, Strasser et al discovered an active class of Co-rich Pt-Ru-Co methanol oxidation catalysts. 21,22 The superiority of this system was later confirmed in an independent study by Cooper and McGinn. 23 The density functional theory (DFT) prediction suggested that Co exerts a ligand effect on Pt surface atoms, reducing the CO chemisorption energy.…”

mBEEF: An accurate semi-local Bayesian error estimation density functional

“…Density functional methods have over the past decade reached a level of maturity where they can be applied not just in detailed theoretical studies of a given material, but be used to search for novel materials for technologically relevant applications in materials science 4-10 and chemical engineering. [11][12][13][14][15][16] Such studies often take a screening approach where massive amounts of DFT calculations are performed using efficient semi-local approximations for the KS exchange-correlation (XC) energy and potential. These include generalized gradient approximations (GGAs) and recent meta-GGA (MGGA) functionals.…”

Molecular dynamics simulations of ternary PtxPdyAuz fuel cell nanocatalyst growth

“…Beside experimental investigations, molecular simulations in conditions matching experiments are also of particular interest [4,29]. While density functional theory (DFT) simulations are able to predict both structure and composition of nanocatalysts leading to higher ORR efficiency, molecular dynamics (MD) simulations are able to predict their growth mode and the final possible structure when simulations are carried out with initial conditions consistent with experimental ones [30,31].…”