Data and code availability
All original data and code are available in the online repository at GitHub (https://github.com/XuhaoWan/HEA-ORR) and at Zenodo (http://doi.org/10.5281/zenodo.6666342).
Benefiting from huge chemical space, high-entropy alloys (HEAs) show great potential as heterogeneous catalysts for different reactions. However, vast chemical space makes it extremely difficult to comprehensively study HEAs by traditional trial-and-error experiments. Therefore, a machine-learning-assisted theoretical method is proposed to investigate the oxygen reduction reaction (ORR) catalytic activity of millions of reactive sites on HEA surfaces. The well-performed gradient boosting regression (GBR) model with high accuracy, generalizability, and simplicity is constructed by reasonable data extraction and feature engineering, which can accurately predict the catalytic activities of millions of reactive sites on HEA surfaces. Finally, one strategy to engineer the HEA surface structure by tuning the metal element component ratio is proposed, which doubles the amount of high-activity sites.
All original data and code are available in the online repository at GitHub (https://github.com/XuhaoWan/HEA-ORR) and at Zenodo (http://doi.org/10.5281/zenodo.6666342).