Microkinetic Modeling of the Oxygen Evolution Reaction on Oxide Surfaces

Abstract

The oxygen evolution reaction (OER) is an important anodic reaction that enables the sustainable production of H2 from water at ambient conditions via the hydrogen evolution reaction or provides protons for cathodic COxreduction reactions. Extensive density functional theory (DFT) calculations have been previously carried out to determine the theoretical limiting potential (where all steps are exergonic) of the OER as a function of binding energy descriptors. This work extends those analyses by feeding DFT calculated energies into a mean field, steady state microkinetic model of the OER. This model demonstrates that a simple mechanism consisting of four coupled proton electron transfers can be used to predict activity trends in relevant catalyst surfaces such as rutile and perovskite oxides. A potential dependent barrier is assumed for each coupled proton electron transfer step. The location of the peak of the volcano at OER relevant potentials agrees with thermodynamic predictions. A sensitivity analysis shows that the volcano shape and peak location is dependent on the assumed barrier and how it scales with potential. The peak location is robust for reasonable barrier assumptions at OER relevant potentials. A degree of rate control analysis is carried out to determine how the rate constants for each step affect the overall rate of reaction as a function of catalyst binding energies. Changes in tafel slope for a given surface have been observed experimentally for the OER. The model presented here predicts these changes and demonstrates that changes in tafel slope can be due to switches in reaction mechanism, coverage, and rate controlling step. Surface Pourbaix diagrams are traditionally used to predict equilibrium coverages. Surface kinetics must be considered to determine the surface coverage at the potentials where OER (a non-equilibrium process) is occurring. This research presents a method to determine the steady state coverage of catalyst surfaces under OER relevant conditions. In this method, a coverage is initially assumed based on a surface Pourbaix analysis. Intermediate free energies are calculated at the specified coverage using DFT and fed into the microkinetic model. If the steady-state coverage predicted by the microkinetic model is in agreement with the coverage used in the DFT calculations, the coverage is self consistent. If the kinetic model predicts a different coverage than that used to determine the energies, the calculations must be redone at the coverage predicted by the kinetic model until a self-consistent coverage is found. Generally, we find the steady state coverage corresponds to the precursor of the thermodynamically limiting step for potentials above the limiting potential. This analysis demonstrate that qualitative trends can be predicted by a simple model and assumed proton transfer barriers and demonstrates how factors such as barriers, potential scaling, and coverage affect reaction rate. Future work in determining electrochemical barriers and their potential dependence is necessary to quantitatively predict rates. Figure 1