Predicted Current Density Response Versus Applied Potential for Five Crystallographic IrO2 Facets: Comparison with Thin Film Experiment

Iridium oxide (IrO2) stands out as the most effective catalyst for the oxygen evolution reaction (OER), a pivotal component of proton exchange membrane (PEM) electrolyzers. Despite its exceptional catalytic prowess, the kinetics of this reaction still impose limitations on the efficient production of green H2, thereby inflating costs and impeding broader commercialization. The catalytic activity of IrO2 for OER is influenced by its local chemical and atomic environment, wherein distinct crystallographic facets with unique atomic arrangements can induce significant variations even with the same bulk composition. In this study, we examine the OER performance across five distinct rutile IrO2 facets. Through DFT calculations, we predict cyclic voltammetry (under the water oxidation potential) and compare directly to the results to experiments conducted on IrO2 thin films for validation. Unlike pure metallic systems characterized by degenerate active sites resulting in broad single redox peaks, or homogeneous systems where redox reactions occur at metallic centers, active sites on metal oxide surfaces exhibit localization yet remain interconnected via various types of oxygen bonds. This unique attribute introduces additional considerations that must be included to interpret redox behavior, because interactions between adjacent intermediates alter their identities differently for different applied potentials. We find that multiple redox features can overlap while a single type of redox couple can be associated with multiple redox features. To elucidate the OER activity under operational conditions (> 1.23 VRHE), we conducted grand canonical calculations to explicitly capture the dependence of reactions on the applied potential, with particular emphasis on non-electrochemical steps such as O-O coupling. Through the construction of various possible OER pathways and amalgamating their energy landscapes into a unified microkinetic scheme under steady-state assumptions, we successfully replicate the experimental OER current density for each of the five facets. Our analysis unveils mechanistic disparities between nucleophilic attack and direct oxo coupling across facets, predicting rates as a function of applied potential and identifying the rate-limiting steps. Our approach provides a thorough comprehension of the OER mechanism at the atomic level, offering valuable insights into the structural characteristics that enhance or inhibit the OER activity of IrO2. We expect these insights to have implications for understanding the structural effects on other heterogeneous catalysts which should aid the development of more efficient electrocatalysts.