| Abstract Scope |
In this talk, we will review our most recent atomistic simulation studies to construct grain boundary (GB) “phase” (complexion) diagrams. Specifically, we have conducted hybrid Monte Carlo and molecular dynamics (hybrid MC/MD) simulations in semi-grand canonical ensembles to derive GB diagrams for several binaries, such as Ni-doped Mo [Yang et al. PRL 2018], Au-doped Si [Hu & Luo, Scripta 2019], and Ag-doped Cu systems. We find that the occurrence of first order GB transformations can be observed in Ni-doped Mo and Au-doped Si systems. For the latter case, our simulation shows that 1st order GB adsorption from Si “clean” GB to Au bilayer occurs at low temperature, but it becomes continuous at high temperature. Furthermore, the hexagonal Au segregation patterns identified by MC/MD simulations are verified by first-principles calculations. Finally, differential charge density maps show that strong charge transfer can prompt Au segregation at Si GBs. Most recent work systematically constructs GB diagrams for Ag-doped Cu via data-driven machine learning methods. |