AI-driven design of multiprincipal element alloys for optimal water splitting

• 成果类型: Article
• 署名作者: Kim, Jihoon; Kim, Dong Won; Choi, Jong Hui; Goddard III, William A.; Kang, Jeung Ku
• 署名单位: Korea Advanced Institute of Science & Technology (KAIST); Korea Advanced Institute of Science & Technology (KAIST); California Institute of Technology
• 刊物名称: PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
• ISSN/ISSBN: 0027-12198
• DOI: 10.1073/pnas.2504226122
• 发表日期: 2025-07-15
• 关键词: entropy; STABILITY; strain

摘要

Water splitting for hydrogen production is essential in advancing the hydrogen economy. Multiprincipal element alloys offer promising opportunities for optimizing this process, yet their vast compositional space and the presence of local minima pose significant challenges for experimental and AI-driven exploration. To overcome these challenges, an AI framework is developed by integrating Gaussian Process Regression with a configuration entropy-based acquisition function for screening and a design of experiments (DoE) for data-efficient overpotential mapping. Through Bayesian optimization across 16.2 million chemical compositions, this entropy-screened and DoE dataset-trained AI identifies Fe12Co28Ni33Mo17Pd5Pt5 as the best composition for water splitting within its search space. The alloy exhibits ultralow overpotentials of 24 mV for hydrogen evolution and 204 mV for oxygen evolution at 10 mAcm-2 with robust stability, surpassing state-of-the-art non-noble and noble metal electrocatalysts including Pt/C+IrO2, Pt35Ru65, and Ru-VO2-demonstrating remarkable performance beyond reach by contemporary experimental and AI frameworks.