Deciphering grain-boundary composition-structure-mechanical property relationships via interfacial property diagrams

Grain boundaries (GBs) are planar crystal defects that significantly influence many mechanical properties and deformation processes of polycrystalline materials. Understanding the interfacial structures of GBs and their relationships to mechanical properties is a central topic in materials science. However, characterizing GB structures and their properties often requires highly sophisticated and time-consuming experimental procedures, making it a grand challenge in understanding the GB composition-structure-property relationships. Using copper-silver (Cu-Ag) as a modeling system, we adopted high-throughput atomistic simulations combined with machine learning techniques to uncover the relationships between GB compositions, structures, and mechanical properties by developing GB property diagrams for four types of GBs with different symmetry. These interfacial diagrams reveal that GB local structural features, such as disordering and free volume, plays more important roles in controlling the GB mechanical properties, while Ag segregation plays a minor role. Dislocation analysis shows that nonsymmetric GBs exhibit direction-dependent deformation behaviors, where stacking faults preferentially emitted into the grains with high-index planes accompanied by dislocation nucleation. This study not only expands the existing family of GB diagrams of mechanical properties to include various GB types, but also enhances our fundamental understanding of GB deformation mechanisms.