Scholarlink

<p>High-entropy alloys are increasingly recognised as promising catalyst materials due to their highly tunable compositions. In particular, the Cantor alloy, CrCoFeMnNi, has attracted significant interest as a low-cost, earth-abundant alternative to Pt-group catalysts. However, the Cantor alloy is susceptible to degradation in acidic environments. To address this, Pt has been proposed as an addition to the Cantor alloy to leverage the high catalytic activity and surface stability of Pt. In this study, we identify and elucidate the catalytic activity trend for the oxygen reduction reaction (ORR) by screening CrCoFeMnNiPt alloys containing 50–100 at.% Pt and up to 50 at.% of each 3d element at 10 at.% resolution using a purely computational, machine learning-accelerated density-functional-theory workflow. Assuming the formation of a Pt skin through surface reconfiguration, a machine learning-accelerated simulation protocol infers the strain-corrected adsorption strengths of key reaction intermediates and uses them as descriptors for kinetic modelling. Our results show that Pt-modification of the Cantor alloy is a potentially viable strategy towards active and durable ORR catalysts. However, the simulations also predict binary Pt-rich alloys with approximately 80 at.% Pt to exhibit the highest ORR activity due to synergistic ligand and strain effects. As this model identifies known alloys with high ORR performance, we suggest that this framework contributes to a clear understanding of how bulk composition affects the reaction kinetics of complex surface alloys.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning-Accelerated Computational Screening of Pt-Modified Cantor Alloys as High-Entropy Electrocatalysts for Oxygen Reduction Reaction

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月30日

查看全文
  • Wiljar Lobjakas,
  • Christian M. Clausen,
  • Jaak Nerut,
  • Jan Rossmeisl

摘要

High-entropy alloys are increasingly recognised as promising catalyst materials due to their highly tunable compositions. In particular, the Cantor alloy, CrCoFeMnNi, has attracted significant interest as a low-cost, earth-abundant alternative to Pt-group catalysts. However, the Cantor alloy is susceptible to degradation in acidic environments. To address this, Pt has been proposed as an addition to the Cantor alloy to leverage the high catalytic activity and surface stability of Pt. In this study, we identify and elucidate the catalytic activity trend for the oxygen reduction reaction (ORR) by screening CrCoFeMnNiPt alloys containing 50–100 at.% Pt and up to 50 at.% of each 3d element at 10 at.% resolution using a purely computational, machine learning-accelerated density-functional-theory workflow. Assuming the formation of a Pt skin through surface reconfiguration, a machine learning-accelerated simulation protocol infers the strain-corrected adsorption strengths of key reaction intermediates and uses them as descriptors for kinetic modelling. Our results show that Pt-modification of the Cantor alloy is a potentially viable strategy towards active and durable ORR catalysts. However, the simulations also predict binary Pt-rich alloys with approximately 80 at.% Pt to exhibit the highest ORR activity due to synergistic ligand and strain effects. As this model identifies known alloys with high ORR performance, we suggest that this framework contributes to a clear understanding of how bulk composition affects the reaction kinetics of complex surface alloys.