Key points are not available for this paper at this time.
The discovery of high-performing catalysts for the oxygen reduction and oxygen evolution reactions (ORR/OER) and related oxidative reactions is critical for the development of energy-efficient hydrogen fuel cells and electrolyzers. To achieve this goal, affordable and abundant alternatives to precious metal catalysts with sufficient stability and catalytic activity must be discovered. Ab-initio catalytic simulations combined with machine learning (ML) methods can play an important role in screening a larger search space of potential electrocatalysts materials of higher structural and compositional complexity, such as class of transition metal oxides (TMOs). In this talk, we will show how electronic and structural descriptors derived from bulk DFT calculations can improve ML models for the OER and ORR on transition metal oxides. In our previous work 1 we demonstrated that the electronic structure of the bulk TMOs, obtained as the crystal orbital Hamiltonian populations (COHP) of the metal-oxygen bond, is an accurate descriptor for O and OH adsorption. Here, we will discuss our extended model for O* and OH* adsorption across crystal structures and oxidation states, utilizing a new dataset of adsorption on binary (AxOy) oxide surfaces spanning the entire transition metal series. Building on our recent work, we extend the COHP-based model to a ML-based prediction of adsorption across multiple oxidation states to obtain a MAE < 0.2 eV for the O-OH and OH adsorption energy descriptors. These results can enable a more efficient screening of catalysts on the bulk level of DFT computation to significantly reduce the computational cost 2. Furthermore, we will discuss how the COHP analysis can be directly applied to oxide surfaces to understand the impact of metal valency, structural reorganization, and magnetization on adsorption energetics 3. Our prediction models of stability of bulk phases from COHP and oxidation state beyond binary alloys will also be highlited 4. Lastly, we will discuss recent progress on collecting and accessing computational and underlying experimental data on catalysis-hub.org. This research was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Chemical Sciences, Geosciences, and Biosciences Division, Catalysis Science Program to the SUNCAT Center for Interface Science and Catalysis. 1 Comer, B. M., Li, J., Abild-Pedersen, F., Bajdich, M., Winther, K. T. (2022). Unraveling Electronic Trends in O∗ and OH∗ Surface Adsorption in the MO2Transition-Metal Oxide Series. , JPCC 2022 126 (18)(2022) 2 Comer, B. M., Bothra, .N, Lunger, J., Abild-Pedersen, F., Bajdich, M., Winther, K. T., Generalized Prediction of O and OH Adsorption on Transition Metal Oxide Surfaces from Bulk Descriptors (under review) 3 Bothra, .N, Comer, B. M., Winther, K. T., Bajdich, M., Understanding the Effects of Surface Bonding in Metal-Oxides Beyond the Limit of the Active Site (in preparation) 4 Craig, M. J., Kleuker, F., Bajdich, M., & Garcia-Melchor, M. (2023). FEFOS: A method to derive oxide formation energies from oxidation states. Catalysis Science & Technology.
Bajdich et al. (Fri,) studied this question.