It was recalled that, in the absence of kinetic cross interactions between diffusing components, intrinsic diffusion could be described by a simple atomic mobility model in which intrinsic diffusion was driven only by the chemical potential gradient of each component. In systems where the diffusional interactions between components could not be ignored, the interactions could be related to a vacancy wind effect in which the net vacancy flux made an additional contribution to the total intrinsic flux of a component. New relationships were derived in order to describe the interdiffusion fluxes in terms of atomic mobilities, plus a vacancy wind parameter. These relationships were applied at selected points on the diffusion paths of γ-phase U-Pu-Zr diffusion couples at 750C. The calculated atomic mobilities and vacancy wind parameter were used to assess the contribution which the vacancy wind effect made to the intrinsic diffusion of the individual components. The results showed that the vacancy wind contribution to intrinsic diffusion could be greater than the contribution that arose from chemical potential gradients. When these 2 contributions acted in opposite directions, the net intrinsic diffusion of a component could be contrary to its own chemical potential gradient. This was the first reported experimental evidence for intrinsic U diffusion up its own chemical potential gradient.

Vacancy-Wind Contributions to Intrinsic Diffusion M.C.Petri, M.A.Dayananda: Philosophical Magazine A, 1997, 76[6], 1169-85