Total: 1
The behavior of surface triple junctions (STJ) at emergent grain boundaries on free surfaces are critical contributors to material properties, particularly at the nanoscale. Interest has been stimulated by recent experimental observations of local grain rotation, the restructuring of emergent grain boundaries and the resultant levels of stress generation, and computationally by the introduction of the core-shift method for building wedge disclination structures at STJ and the establishment of relationships between atomic calculation and the continuum elastic theory. Despite this progress, the lowest-energy structure for the surface triple junction is still unknown. Here, we carried out the experimental observation of the local surface deformation at STJ, the local stress analysis and ergodic searches for metastable structures. We identified the zipped Y-shaped notch, rather than the zipped V-shaped notch or wedge disclination, as the most stable structures, which was subsequently verified with machine learning methods for a wide range of boundaries. Local stress analysis explains this energetic preference. Our findings provide fundamental insight into STJ energetics and offer a possible framework for understanding kinetically diffusive deformation and the pathway for engineering related material properties, in particular thin-film properties.