Chen Enze, Heo Tae Wook, Wood Brandon C, Asta Mark, Frolov Timofey
Department of Materials Science and Engineering, University of California, Berkeley, CA, USA.
Materials Science Division, Lawrence Livermore National Laboratory, Livermore, CA, USA.
Nat Commun. 2024 Aug 15;15(1):7049. doi: 10.1038/s41467-024-51330-9.
Grain boundaries (GBs) profoundly influence the properties and performance of materials, emphasizing the importance of understanding the GB structure and phase behavior. As recent computational studies have demonstrated the existence of multiple GB phases associated with varying the atomic density at the interface, we introduce a validated, open-source GRand canonical Interface Predictor (GRIP) tool that automates high-throughput, grand canonical optimization of GB structures. While previous studies of GB phases have almost exclusively focused on cubic systems, we demonstrate the utility of GRIP in an application to hexagonal close-packed titanium. We perform a systematic high-throughput exploration of tilt GBs in titanium and discover previously unreported structures and phase transitions. In low-angle boundaries, we demonstrate a coupling between point defect absorption and the change in the GB dislocation network topology due to GB phase transformations, which has important implications for the accommodation of radiation-induced defects.
晶界（GBs）对材料的性能和表现有着深远影响，这凸显了理解晶界结构和相行为的重要性。近期的计算研究表明，随着界面处原子密度的变化，会存在多种晶界相，因此我们引入了一个经过验证的开源巨正则界面预测器（GRIP）工具，该工具可自动进行晶界结构的高通量巨正则优化。虽然此前对晶界相的研究几乎都只聚焦于立方体系，但我们展示了GRIP在应用于六方密堆积钛时的效用。我们对钛中的倾斜晶界进行了系统的高通量探索，并发现了此前未报道的结构和相变。在小角度晶界中，我们证明了点缺陷吸收与由于晶界相变导致的晶界位错网络拓扑变化之间的耦合，这对辐射诱导缺陷的容纳具有重要意义。
