氢脆会导致金属材料发生突发性失效，这在核能设施和氢能储存系统中是一个严峻挑战。

Hydrogen embrittlement causes unpredictable material failure, posing a significant challenge for nuclear facilities and hydrogen storage systems.

研究方法 Methods

该研究结合了第一性原理计算、基于ab-initio数据训练的机器学习势的分子动力学模拟，以及用于热力学平衡分析的Ising模型。

This study combines Density Functional Theory (DFT) calculations, Molecular Dynamics (MD) simulations using a machine learning (ML) potential trained on ab-initio data, and Ising models for thermodynamic equilibrium analysis.

研究亮点 Highlights

• 表征了氢修饰的“易核”和“硬核”两种截然不同的螺位错核结构。

• Characterized two distinct hydrogen-decorated core structures: the "easy core" and the "hard core" configurations.

• 热力学分析证明，在广泛的温度和氢浓度范围内，氢偏聚的“硬核”结构更为稳定。

• Thermodynamic analysis reveals that the hydrogen-decorated "hard core" is favored across a wide range of temperatures and concentrations.

• 定量证实氢偏聚显著提升了位错运动能垒，导致螺位错被“锁住”，解释了高浓度下的氢致硬化现象。

• Quantitatively confirmed that hydrogen segregation significantly increases energy barriers, leading to the "locking" of screw dislocations and H-induced hardening.

研究意义 Significance

This work clarifies the transition from the dilute regime (enhanced mobility) to the segregation regime (dislocation locking), providing foundation for understanding H-induced plasticity.

图文概览 Graphical Overview

图 1 用于DFT计算的135原子和180原子位错超胞模型。

Fig. 1 Dislocation supercell models containing 135 and 180 atoms used for DFT calculations.

图 2 氢原子在易核、原子列中心及硬核位错配置下的弛豫结构。

Fig. 2 Relaxed structures of hydrogen atoms in easy, column-centered, and hard core dislocation configurations.

图 3 易核与硬核位错偏聚模型的Ising哈密顿量参数定义。

图 4 DFT和机器学习势计算的易核与硬核结构间的自由能差

Fig. 4 Free energy difference between easy and hard core structures calculated via DFT and ML potential.

图 5 分子动力学模拟与Ising模型预测的硬核配置氢偏聚剖面对比。

作者 Authors

Thomas Leveau, Lisa Ventelon,

Mihai-Cosmin Marinica, Emmanuel Clouet

第一作者及通讯作者

First & Corresponding Author

Thomas Leveau

研究单位 Affiliations

法国原子能与可替代能源委员会 (CEA)，巴黎-萨克雷大学 (Université Paris-Saclay)，法国国家科学研究中心 (CNRS)。