Multiscale Computational Design of Core/Shell Nanoparticles for Oxygen Reduction Reaction

We propose a multiscale computational framework to design core/shell nanoparticles (NPs) for oxygen reduction reaction (ORR). Essential to the framework are linear scaling relations between oxygen adsorption energy and surface strain, which can be determined for NP facets and edges from first-principles and multiscale QM/MM calculations, respectively. Based on the linear scaling relations and a microkinetic model, we can estimate ORR rates as a function of surface strain on core/shell NPs. Employing the multiscale framework, we have systematically examined the ORR activity on Pd-based core/shell NPs as a function of their shape, size, shell thickness, and alloy composition of the core. Three NP shapesicosahedron, octahedron, and truncated octahedronare explored, and the truncated octahedron is found to be the most active and the icosahedron is the least active. Ni_<i>x</i>Pd_1–<i>x</i>@Pd NPs with high Ni concentrations and thin shells could exhibit higher ORR rates than the pure Pt(111) surface and/or Pt NPs. Ag_<i>x</i>Pd_1–<i>x</i>@Pd in the truncated octahedron shape and high Ag concentrations are predicted to be even more active than Ni_<i>x</i>Pd_1–<i>x</i>@Pd NPs under the same conditions. The highly active Ag_<i>x</i>Pd_1–<i>x</i>@Pd and Ni_<i>x</i>Pd_1–<i>x</i>@Pd NPs are thermodynamically stable