Ab initio modeling was used to predict diffusion relevant thermodynamic and kinetic information for dilute Ni–Cr and Ni–Fe alloys. The modeling results were then used to determine the phenomenological coefficient matrices and the tracer diffusion coefficients for both vacancy and interstitial mediated diffusion. In addition to predicting diffusion coefficients, this ab initio-based approach provides information typically inaccessible to experiments, including the different contributions to diffusion (e.g., electronic excitation effects), the species dependence of interstitial diffusion, and the deviations from Arrhenius-type relations, which were often used to describe and extrapolate experimental diffusion data. It was found that: (1) Cr was the fastest diffusing species in Ni by both vacancy and interstitial diffusion, followed by Fe and then Ni. The enhanced diffusivity of Cr was primarily due to differences in migration barriers and binding energies, not pre-exponential factors. (2) Fe and Cr solutes in Ni have weak interactions with vacancies but Cr solutes bind strongly to interstitial defects. (3) Cr exhibited non-Arrhenius behavior in both vacancy and interstitial mediated diffusion. (4) Temperature dependent electronic contributions have a significant impact on the diffusion in some cases. (5) The vacancy diffusion mechanism in Ni–Cr changed as a function of temperature resulting in vacancy–solute drag below 460 K.

Ab initio-Based Diffusion Theory and Tracer Diffusion in Ni–Cr and Ni–Fe Alloys. J.D.Tucker, R.Najafabadi, T.R.Allen, D.Morgan: Journal of Nuclear Materials, 2010, 405[3], 216-34