First Step of the Oxygen Reduction Reaction: A Computational Study of O2 Adsorption at Electrified Metal/Water Interfaces

Abstract

The oxygen reduction reaction (ORR) is a key limiting factor in fuel cell technology, driving extensive research efforts over the past few decades. Studies based on experiments and theoretical calculations on model single-crystal electrodes have helped establish fundamental trends across transition metal catalysts. In heterogeneous electrocatalysis, local electric field effects at the electrical double layer significantly influence the energies of reaction intermediates and impact catalyst performance in acidic or alkaline environments. However, these local field effects are challenging to model computationally and are frequently omitted. First, this work focuses on O2 adsorption as the initial step of the ORR to understand the role of the elec- tric field. The first part of the study focuses on the weak binding Au(111) metal catalyst surface, which favors the formation of hydrogen peroxide over water, with its activity strongly dependent on the (absolute) electrode potential. The underlying microscopic mechanisms remain unclear, likely involving key elementary reaction steps. We systematically enhance the double-layer model to clarify and compare the physical effects of the local field on O2 adsorption. This progression includes an applied saw-tooth potential in vacuum, an implicit solvent model, and explicit water modeling via ab initio molecular dynamics (AIMD). Two main contributions are identified to the potential dependence of O2 adsorption. Firstly, a dominant dipole- field interaction favors O2 binding going to reducing conditions across all models. Additionally, we observe stabilization from explicit H-bonding that can only be ob- served in AIMD, leading to a peroxo-O2* and a significant field response near the ORR onset. Since the O2* adsorption becomes favorable close to experimental ORR onsets and can explain experimental SHE-driven ORR activity, we predict that O2 adsorption is a potential-dependent, potentially rate-determining step of the ORR on the weak binding Au(111). These findings highlight the necessity of incorporating local electric field effects and explicit water in electrochemical interface models. Secondly, we draw a comparison to the more reactive Pt(111) surface. We conduct AIMD simulations to analyze and compare the properties of the metal/water interface Au(111) vs. Pt(111). Unlike Au(111), Pt(111) exhibits negligible potential dependence under realistic ORR conditions due to the inherently different reactivity of the two metals. We find a closely adsorbed peroxo O2* state with a relatively constant number of H-bonds, irrespective of the potential or surface coverage on the Pt(111). In contrast, the interfacial water structure on Pt(111) undergoes significant changes due to the inclusion of more realistic surface coverages and potential variations. In our set-up, we observe O2* dissociation determining the ORR selectivity towards H2O. We find an indirect effect of the potential through surface coverage: The O2* dissociation is promoted by desorption of H2O* at reducing conditions. Our overall results emphasize the importance of accounting for local field effects, which i) can directly impact reaction steps but ii) can also indirectly impact the reaction mechanism through a more complex interplay between potential, surface coverage, and water structure.