| Abstract Scope |
The nature of solid–liquid interfaces is essential for processes such as metal-catalyzed nanostructure growth. In particular, during self-catalyzed GaN nanowire growth by molecular beam epitaxy, the presence of molten metal, such as liquid Ga on solid GaN, is crucial for the successful incorporation of N atoms into the crystal. A deeper understanding of these interfaces is vital for elucidating the growth mechanisms of semiconductor crystal and the related nanostructures. In this study, molecular dynamics simulations with machine-learning interatomic potentials reveal interface-induced ordering and charge state variations in liquid Ga in contact with solid GaN. Advanced free energy sampling calculations quantify the energy barriers encountered by N adatoms during adsorption and migration across different interfacial facets, as well as at step edges. These free energy landscapes enable quantitative predictions using kinetic Monte Carlo simulations to assess the influence of nitrogen flux conditions, revealing diffusion-controlled growth kinetics for GaN. |