A diffusion model of binary systems controlled by chemical potential gradient

Abstract

The paper presents a model of diffusion in a single phase with chemical potential gradient as the driving force of the process. Fick's laws are strictly empirical and the assumption that the concentration gradients are the driving forces of diffusion is far from precise. Instead, the gradient of chemical potential $\mu_{i}$ of component i is the real driving force. The matter of governing equations of models that incorporate this approach will be raised and discussed in this article. One of more important features is the ability to acquire results where diffusion against the concentration gradient may occur. The presented model uses the Finite Difference Method (FDM) and employs the CALPHAD method to obtain chemical potentials. The calculations of chemical potential are carried out for instant conditions - temperature and composition - in the entire task domain by Thermo-Calc via a TQ-Interface. Then the heterogeneity of chemical potentials is translated into mass transfer for each individual element. Calculations of two modelling tasks for one-dimension diffusion field were carried out. First: isothermal conditions with linear initial composition distribution and second: constant temperature gradient with uniform chemical composition in the specimen. Results for two binary solid solutions: Fe-C and Fe-Si, in the FCC phase for the given tasks will be presented. Modelling allows us to estimate the time needed to reach a desired state in a particular equilibrium or quasi-equilibrium state. It also shows the path of the composition change during the process. This can be used to determine whether the system at some point is getting close to the formation of another phase due to significant deviation from its initial conditions.