Abstract
The theoretical equations proposed for calculating D 0 of the diffusion‐constant‐temperature relation D=D 0 e −E/RT are critically discussed in the light of presently available data. It is concluded that none of the relations gives satisfactory agreement with experiment. Empirical analysis of experimental data covering a wide range of D 0 and E values showed that the quantity E/Tm , where Tm is the melting point in °K, is the main factor in determining the value of D 0, and that D 0 depends approximately exponentially on E/Tm . In addition to many other diffusion constants, it is shown that the anisotropicdiffusion of zinc and the widely differing diffusion constants for self‐diffusion in α‐ and γ‐iron are included in the proposed empirical correlation. A possible theoretical basis for the empirical correlation is described based on the idea of local melting or disordering. The quantity E/Tm is identified with the entropy of activation.

This publication has 3 references indexed in Scilit: