Time variation of concentration-dependent interdiffusion coefficient obtained by numerical simulation analysis

A newly developed numerical diffusion model that conserves solute is coupled with a recently reported forward simulation technique to address the key flaws in conventional methods, including the Boltzmann-Matano, Sauer-Freise and Hall procedures, which are used to study the dependence of interdiffusion coefficient on concentration. The new model involves a combination of fully explicit finite difference analysis with the Leapfrog / Dufort-Frankel scheme to effectively eliminate non-trivial simplification assumptions that degrade accuracy. Numerical analyses of experimental data obtained in the present work and those reported by other authors in the literature show that despite not being generally acknowledged, the concentration dependency of interdiffusion coefficient can significantly vary with time isothermally, even in binary systems. In these situations, the use of a time-independent function, through which interdiffusion coefficient changes with concentration, as commonly derived by the conventional methods, to compute diffusion-controlled kinetics or solute distribution, can be grossly unreliable. Furthermore, unawareness or disregard of this vital concept can lead to significant misinterpretation of experimental data, such as misidentification of the underlying mechanism of microstructural changes by phase transformation reactions.