Molecular Dynamics Simulation of Co–Fe-Based Perovskite Oxide/Water Interfaces

Abstract

Perovskite oxides have shown promising catalytic performances in the oxygen evolution reaction (OER), yet the intricate reconstruction at the oxide/electrolyte interfaces has complicated the understanding of the catalytic mechanism and hindered the rational design of catalysts. In this work, molecular dynamics (MD) simulations of perovskite oxide/water interfaces were performed based on on-the-fly machine learning potentials to provide atomistic insights into the influence of solvent on surface structure of the prototypical oxides Ba0.5Sr0.5Co0.75Fe0.25O3 (BSCF) and Rb0.25Sr0.75Co0.5Fe0.5O3 (RSCF). Results show that water plays important roles in many structural phenomena, in particular on transition metal-terminated surfaces. The presence of water facilitates the dynamic oxygen exchange between that in the lattice and that in water via the H–O bonding. Formation of surface peroxo species and gas O2 molecules was observed, leaving tetragonal [CoO4H] surface units in each of which one oxygen is stabilized by forming the O–H bond. In addition, we found that the oxygens bonded to Co atoms are responsible for nearly all of the observed lattice-oxygen-related events due to the weaker bonding of O with Co than with Fe. In comparison to BSCF, no peroxo species and gas O2 molecules were observed within the same simulation time length on RSCF because of the lower content of Co. These findings highlight the roles of water in surface reconstruction and the importance of Co for the reactivity of lattice oxygen.