2025 AIChE Annual Meeting

(675h) Machine Learning for Unraveling Reactivity Origin of Oxygen Reduction at High-Entropy Alloy Electrocatalysts

Authors

Yang Huang, Auburn University
Shih-Han Wang, Virginia Tech
Noushin Omidvar, Virginia Polytechnic Institute and State University
Luke Achenie, National Science Foundation
Sara Skrabalak, Indiana University - Bloomington
Hongliang Xin, Virginia Tech
Machine learning interatomic potentials (MLIPs) have emerged as powerful tools in computational catalysis. In particular, graph neural network (GNN)-based MLIPs have demonstrated superior accuracy over traditional artificial neural networks (ANNs). Trained on high-fidelity electronic structure data, these models enable efficient and accurate simulations of complex systems over extended timescales. High-entropy alloys (HEAs) are a unique class of catalytic materials composed of five or more metal elements forming a single-phase solid solution. Their catalytic potential stems from the vast diversity of local atomic environments and multielement surface sites, stabilized by configurational entropy. However, this same complexity leads to an enormous design space, making conventional first-principles calculations computationally infeasible for thorough understanding.

In this work, we apply a graph neural network–based interatomic potential to investigate the electrocatalytic oxygen reduction reaction on a PdCuPtNiCo HEA nanoparticle. Using large-scale molecular simulations, we determine the most stable surface structure under reaction conditions, while pretrained machine learning models reveal key insights into local surface reactivity. Our results elucidate the mechanisms underlying the enhanced catalytic activity, with surface segregation and electronic structure modifications driving an optimal OH binding strength that agrees well with experimental observations. This study underscores the versatility and predictive power of advanced computational frameworks for rational catalyst design and marks a significant step toward the development of next-generation electrocatalysts for sustainable energy applications.

[1] Y. Huang, S.-H. Wang, X. Wang, N. Omidvar, L.E.K. Achenie, S.E. Skrabalak, and H. Xin, “Unraveling Reactivity Origin of Oxygen Reduction at High-Entropy Alloy Electrocatalysts with a Computational and Data-Driven Approach,” J. Phys. Chem. C Nanomater. Interfaces 128(27), 11183–11189 (2024).