Modeling strategies for hydrogen reduction of high-purity metals: From DFT to ReaxFF and machine learning

Zhimeng Shao, Bowen Gao, Zhifang Hu, Honglin Jiang, Qidong Zhang, Zhihe Dou, Yanxi Yin,
Modeling strategies for hydrogen reduction of high-purity metals: From DFT to ReaxFF and machine learning,
Materials Today Physics,
Volume 59,
2025,
101903,
ISSN 2542-5293,
https://doi.org/10.1016/j.mtphys.2025.101903.
(https://www.sciencedirect.com/science/article/pii/S2542529325002597)
Abstract: High-purity metals are regarded as indispensable in aerospace, nuclear energy, and semiconductor technologies due to their outstanding physical and chemical properties. Hydrogen reduction constitutes a fundamental step in their production, but the associated processes are highly complex. These processes include hydrogen adsorption and dissociation, electronic-state reorganization, and successive intermediate-phase transformations. To elucidate these microscopic mechanisms, computational modeling has emerged as an indispensable approach. In recent years, density functional theory (DFT), reactive force field molecular dynamics (ReaxFF-MD), and machine-learning interatomic potentials (MLIPs) have been increasingly employed, each providing complementary strengths in accuracy, efficiency, and transferability. This review summarizes the applications of DFT, ReaxFF, and MLIPs in elucidating the hydrogen reduction mechanisms of high-purity metals. The respective advantages and limitations of these approaches in describing electronic structures, capturing dynamical processes, and bridging atomistic scales are critically examined. In addition, perspectives on future methodological developments are outlined, with the aim of advancing mechanistic understanding and facilitating the construction of more reliable multiscale models for hydrogen reduction.
Keywords: High-purity metals; Hydrogen reduction; DFT; ReaxFF; Machine-learning interatomic potentials