Automated Analysis to Reveal Grain Boundary Phase Microstructures

Abstract

We develop a method for analyzing grain boundary (GB) microstructures that identifies distinct interfacial phases and the dislocation line defects separating them. Similar to bulk materials, GBs can adopt multiple distinct interfacial phases and undergo first-order phase transitions. When these phases coexist, their spatial arrangement and phase junctions constitute a GB microstructure, characterized by variations in excess properties and line defects with associated dislocation content. Despite this intrinsic heterogeneity, our ability to quantitatively characterize GB microstructures remains limited, as it requires identification of individual GB phases, phase-resolved excess properties, and the Burgers content of phase junctions, capabilities not available in existing automated methods. Here, we present an automated tool that performs interfacial microstructure mapping for planar coincidence site lattice GBs. The method identifies the spatial distribution of GB phases, quantifies phase-specific excess properties, and estimates the Burgers content of GB phase junctions. We demonstrate the approach using three representative cases: (i) quantification of mass transport during diffusion-limited GB phase transformations; (ii) identification of structurally indistinguishable phases formed by vacancy and interstitial loops; and (iii) characterization of GB microstructures containing phase nuclei. More broadly, this framework enables quantitative studies of GB evolution processes, including spinodal decomposition and coarsening with direct implications for GB deformation, creep, and migration.