首页
>
期刊首页
>
中国化学工程学报(英文版)
>
2026年8期
>
Combining DFT calculations and machine learning for rational design of hydrogen electrocatalysts
Combining DFT calculations and machine learning for rational design of hydrogen electrocatalysts
摘要
The development of electrocatalysis,a core technology for the hydrogen economy,faces a dual challenge:overcoming the inherent difficulty in balancing catalyst activity,selectivity,and stability,and being plagued by the high costs and protracted timelines of conventional trial-and-error syn-thesis.This review integrates advances in density functional theory(DFT)calculations and machine learning(ML)for the rational design of hydrogen electrocatalysts.Firstly,it provides an overview of the basic reaction pathways in hydrogen electrocatalysis such as HER,HOR,OER,and ORR,and dis-cusses the types of data from DFT calculations and the normalization processing procedures.The discussed data cover elemental information,structural and energy data,and how these computational indicators can be mapped to macroscopic performance.Secondly,it introduces the data types con-tained in existing high-quality computational databases.Finally,it highlights ML applications in electrocatalysis,categorized into three areas:descriptor construction and screening methods based on expert experience,descriptor construction based on ML models,and descriptor-free rapid prediction and acceleration of large-scale screening using ML models.The potential of natural language pro-cessing for literature and data mining is also discussed,offering guidance and future perspectives for hydrogen electrocatalysis research.
年,卷(期):
2026, 96(8)
机标分类号:
G43
R338.1
TP393.09
在线出版日期:
2026年11月3日