Bayesian Learning Aided Theoretical Optimization of IrPdPtRhRu High Entropy Alloy Catalysts for the Hydrogen Evolution Reaction

Linke Huang,Zachary Gariepy,Ethan Halpren,Li Du,Chung Hsuan Shan,Chuncheng Yang,Zhi Wen Chen,Chandra Veer Singh

DOI: https://doi.org/10.1002/smtd.202401224

IF: 12.4

2024-11-13

Small Methods

Abstract:High‐entropy alloys (HEAs) can offer cost‐effective and enhanced catalytic performance for hydrogen evolution reaction (HER) through compositional optimization. This study combines density functional theory and Bayesian machine learning to elucidate unique catalytic mechanisms of IrPdPtRhRu HEA. A novel Bayesian approach improves composition optimization, achieving superior catalytic efficiency and reducing synthesis costs compared to equiatomic HEAs and Pt benchmarks. The complex compositional space of high entropy alloys (HEAs) has shown a great potential to reduce the cost and further increase the catalytic activity for hydrogen evolution reaction (HER) by compositional optimization. Without uncovering the specifics of the HER mechanism on a given HEA surface, it is unfeasible to apply compositional modifications to enhance the performance and save costs. In this work, a combination of density functional theory and Bayesian machine learning is used to demonstrate the unique catalytic mechanism of IrPdPtRhRu HEA catalysts for HER. At high coverage of underpotential‐deposited hydrogen, a d‐band investigation of the active sites of the HEA surface is conducted to elucidate the superior catalytic performance through electronic interactions between elements. At low coverage, a novel Bayesian learning with oversampling approach is then outlined to optimize the HEA composition for performance improvement and cost reduction. This approach proves more efficacious and efficient and yields higher‐quality structures with less training set bias compared with neural‐network optimization. The proposed HEA optimization theoretically outperforms benchmark Pt catalysts' overpotential by ≈40% at a 15% reduced synthesis cost comparing to the equiatomic ratio HEA.

materials science, multidisciplinary,chemistry, physical,nanoscience & nanotechnology