In this paper, a new, lighter, version of the highly accurate Moleko, Allnatt and Allnatt formalism for describing both tracer (self) and collective diffusion kinetics in multicomponent random alloys is presented. Verification of the resulting expressions is performed by means of kinetic Monte Carlo simulation. The accuracy of the new formalism is much higher than that of the combined Manning and Holdsworth and Elliott formalism discussed recently. Using this formalism the possible range of the tracer diffusion ratio of the highest to the lowest atomic component is examined for equiatomic (or near equiatomic) binary, ternary, quaternary and quinary alloys. It is shown that in the random alloy model, the correlation effect is the highest with a reduction of the fastest tracer diffusion by 40-55%, when moving from two pure metals to their equiatomic binary alloy. By adding the third component (with an intermediate mobility) this effect can be further increased with a possible total reduction of the fastest tracer diffusion by up to 70% (depending on the combinations of mobilities), while adding the fourth component brings this reduction up to 80% and with a possible maximum of up to 85% reduction for the 5component alloy (again depending on the combinations of mobilities). But the slowest diffusing components are not affected by this. This suggests that kinetics arguments alone are not enough for explaining the sluggish diffusion observed of all atomic components in (equiatomic) high entropy alloys.
The self-or tracer diffusivity of one component in a binary alloy is often required when there is knowledge of the other component’s self-or tracer diffusivity and the interdiffusivity (and the thermodynamic factor). In the present paper, this problem is addressed for the random alloy model by applying three possible approximations having different levels of accuracy: Darken (low level of accuracy), Manning (medium level of accuracy) and Moleko, Allnatt and Allnatt (MAA) (high level of accuracy). There are unexpectedly large differences between the results of these approximations that sometimes are reflected in the high sensitivity of the vacancy-wind factor to the level of approximation. Generally, for the application of Manning and the MAA approximations, it is found that there is a difference in the number of self-diffusivity roots depending on whether the tracer diffusivity is available for the faster diffuser or for the slower diffuser and depending on how close the composition is to the forbidden (according to Manning’s description) region. Provided that the interdiffusion coefficient (divided by the thermodynamic factor) is greater than the available self-diffusion coefficient multiplied by its complementary composition, the application of the Darken approximation always results in one self-diffusivity root.
In this paper, the diffusion isotope effect and diffusion mechanism are investigated by means of molecular dynamics simulations in two liquid alloys, Ni-Ag and Ni-Cu. The values for the diffusion isotope effect parameter allow for the estimate of the number of atoms which are moving cooperatively in a basic diffusion event as experienced by a given atomic species. It is shown that the composition dependence of ND is typically very small. However, the temperature dependence of this parameter is much more pronounced. In addition, it is shown that, on average, in these alloys and temperatures considered, ND is limited to the range: 5<ND<17. This is consistent with results of molecular dynamics simulations on the average coordination number calculations. This would suggest that, together with a given atom, depending on temperature, the neighbouring atoms are all involved in the basic diffusion event.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.