Genetic algorithm was applied to search the energy minimization configuration of vacancy clusters in α-Fe. Molecular statics calculations and dynamics annealing relaxation were employed to calculate the formation and binding energies of vacancies and 3D vacancy or vacancy-hydrogen (H) clusters, as well as 2D vacancy or vacancy-H clusters on (111), (011) and (211) planes. Our calculations show that vacancies prefer to form 3D clusters and vacancy dislocation loops are difficult to form, while vacancy-H clusters prefer to shape into 2D clusters, especially on (211) planes. Since H prefers the directional bonding, a vacancy cluster with H atoms tend to form a vacancy dislocation loop with its slip direction along the 〈100〉 direction and on the (211) habit plane. Our results are consistent with the experimental observations, and provide a possible mechanism for the formation of vacancy dislocation loops in α-Fe. Furthermore, we have also explored how dislocation loops trap self-interstitial atoms, vacancies and H atoms. It is of interest to note that H atoms strongly bound to a 〈100〉 vacancy dislocation loop, and are able to enhance the ability for the dislocation loop to further trap vacancies and reduce its ability to absorb self-interstitial atoms, thus promoting the growth of the 〈100〉 vacancy dislocation loops.

应用遗传算法搜索α-Fe中空位簇的能量最小化构型。利用分子静力学计算和动力学退火弛豫来计算空位和3D空位或空位氢（H）簇以及（111），（011）和（2D）上的2D空位或空位H簇的形成和结合能。 211）飞机。我们的计算表明，空位倾向于形成3D簇，并且空位错位环很难形成，而空位H簇则倾向于形成2D簇，尤其是在（211）平面上。由于H偏向于定向键，具有H原子的空位簇趋于形成空位错环，其滑移方向沿<100>方向且位于（211）惯性平面上。我们的结果与实验观察结果一致，并为α-Fe中空位错环的形成提供了可能的机制。此外，我们还探索了位错环如何俘获自填隙原子，空位和H原子。值得注意的是，H原子与一个 <100>空位错环，并能够增强位错环进一步捕获空位的能力并降低其吸收自填隙原子的能力，从而促进了<100>空位错环的生长。