The mutual interaction of host atoms and interstitial atoms was expected to induce phase transitions and order–disorder transitions. Thus, such correlated site occupations had to be taken into account in the computational design of materials such as steels and membranes for hydrogen purification. The cluster variation-based simple cube approximation was used here to predict phase transformations in a hypothetical system and the effect of alloying elements upon the distribution of atoms and vacancies in face-centered cubic structured transition-metal hosts. The cube cluster was composed of both metal and interstitial sub-lattice sites whose mutual interaction was expressed in terms of pair potentials.

Computational Design of Interstitial Alloys - Effect of Metal-Interstitial Interactions on Atom Distributions. M.I.Pekelharing, A.J.Böttger: Computational Materials Science, 2009, 45[2], 561-7