Getting Around the Intrinsic Errors of Density Functional Theory Calculations in Electrocatalysis

Abstract

This article shows that the intrinsic gas-phase error in density functional theory (DFT) calculations can be effectively eliminated by referencing adsorption energies and reaction barriers to surface-bound intermediates rather than to molecular species. In oxygen reduction reaction modeling, referencing OH adsorption energies simultaneously to Pt(111) and the Sabatier volcano apex cancels the gas-phase error and reduces uncertainty through error cancellation. The same principle applies to activation barriers, as demonstrated for nitrogen reduction reaction pathways, where different reference paths, although thermodynamically equivalent, propagate distinct uncertainties due to variations in the intermediate structures. Using Bayesian error estimation, we quantify remaining uncertainties and show that the proposed transformation yields internally consistent and reproducible energy diagrams, completely avoiding empirical gas-phase corrections. This approach offers a computationally efficient, transferable strategy for producing reliable, interpretable predictions. As the strategy is independent of specific DFT methods, the approach can become even more accurate and have lower uncertainty as DFT functionals are developed using updated experimental or computational benchmarks.