Prediction of Temperature Dependence of Impurity Diffusion Coefficients in Liquid Metal Based on a Hard-Sphere Model from Measurements Using Shear Cell Technique and Stable Density Layering

This study aims to establish predictive formulas for the temperature dependence of impurity diffusion coefficient based on a hard-sphere (HS) model and the measurement results. The impurity diffusion coefficients of Sb, Bi, and In in liquid Sn were measured using the shear cell technique and stable density layering at 773 K and 973 K (500 °C and 700 °C, respectively) with suppression of natural convection. The temperature dependence of the impurity diffusion coefficient can be predicted by multiplying the ratio of the solvent to the solute of the following three factors by the self-diffusion coefficient of the solvent as the slope: (i) the square of mean atomic diameter, (ii) the first peak of the pair distribution function calculated by the HS model, and (iii) the square root of the converted atomic weight. If the ratio of the atomic diameter is close to one, the temperature dependence of the impurity diffusion coefficient can also be predicted with an accuracy similar to the abovementioned relationship by multiplying the following two factors by the self-diffusion coefficient of the solvent as the slope: (i) the atomic diameter ratio of the solvent to the solute and (ii) the thermodynamic factor. The predictive formulas based on the HS model showed an accuracy of approximately ±10 pct for the experimental values from 573 K to 973 K (300 °C to 700 °C).