The relationship between component self-diffusion coefficients and interdiffusion coefficient for pseudobinary systems

Abstract

A general relationship between the interdiffusion ( D) and component self-diffusion coefficients ( D i ∗) for pseudobinary systems is derived using the basic definitions of the diffusion coefficients and the constraints appropriate for pseudobinary systems. The Hg 1 − x ,Cd x Te (MCT) system is used as an example of such a system. The relationship between D and the intrinsic chemical diffusion coefficient ( D i,) is the same as for a binary system and is described by the Darken equation. In pseudobinary systems there are relevant diffusion driving forces in addition to neutral chemical potential gradients and the relationships between D i and D i ∗ and D and D i ∗ are complicated by the diffusion of a common non-metal species. The physical meaning of the resulting equation is explained by the limiting behavior of the general equation. A Darken-type equation is obtained when there is only a neutral chemical potential gradient for metal diffusion species such as the case of a common non-metal species moving much faster than the metal species. However, when the common non-metal species diffuses more slowly than the metal components, the resulting interdiffusion coefficient is related to the D i ∗ by a Nernst-Planck-type equation, since the electroneutrality and/or the conservation of lattice sites requires the flux of metals to be coupled. The slow movement of the common non-metal species plays the same role as lack of attainment of vacancy equilibrium in binary systems. Our experimental results of the D and D i ∗ values in the MCT system support our theoretical analysis.