Positive enthalpy of mixing and short-range order topology in atomic transport: Cu–Ta mutual diffusion coefficients via Maxwell–Stefan diffusion theory

Immiscible systems offer unique insights into atomic interactions and structural evolution, especially in amorphous materials. In this study, we investigate amorphous Cu/Ta alloys using GPU-assisted molecular dynamics simulations for robust sampling and reliable data acquisition, covering a temperature range from 373 to 523 K. We compute mutual diffusion coefficients at the atomic scale and analyze their temperature and composition dependence, supported by Maxwell–Stefan theory. The results demonstrate a complex interplay between diffusion and structure, with diffusion coefficients systematically varying with composition and temperature. Advanced analysis tools – including Voronoi topology, Kirkwood–Buff integrals, and short- and medium-range order statistics – reveal that atomic-scale heterogeneity and positive enthalpy of mixing drive the observed diffusion behavior. Our findings highlight that structural disorder in amorphous alloys produces significantly higher diffusion coefficients compared to crystalline counterparts, emphasizing the importance of incorporating topological and thermodynamic factors into models of atomic transport in immiscible systems.