Diffusion in Ni-Zr Melts: Insights from Statistical Mechanics and Atomistic Modeling
A Kromik and EV Levchenko and C Massobrio and AV Evteev, ADVANCED THEORY AND SIMULATIONS, 1, 1800109 (2018).
DOI: 10.1002/adts.201800109
An accurate database of diffusion properties of Ni-Zr melts is generated within the framework of the molecular-dynamics method in conjunction with a semi-empirical many-body interatomic potential. The reliability of the model description of Ni-Zr melts is confirmed via comparison of the simulation results with the existing experimental data on diffusion properties of Ni-Zr melts. A statistical mechanical formalism is employed to understand the behavior of the cross-correlation between the interdiffusion flux and the force caused by the difference in the average random accelerations of atoms of different species in the short time limit t -> 0. This theoretical description is exploited to analyze the simulation data on the diffusion properties of Ni-Zr melts. On this basis, it is found that in the composition range 0.25 less than or similar to c(NI) less than or similar to 0.5 both single-particle and collective diffusion dynamics slow down homogeneously upon undercooling of Ni-Zr melts. Furthermore, it is inferred that such homogeneous dynamical slowdown is related to the enhanced stability of undercooled melt against crystallization. As a consequence, Ni-Zr alloys within this composition range are identified as viable glass formers.
Return to Publications page