ELECTRONIC STRUCTURE AND ORDERING
- 1 December 1977
- journal article
- Published by EDP Sciences in Le Journal de Physique Colloques
- Vol. 38 (C7) , C7-272
- https://doi.org/10.1051/jphyscol:1977753
Abstract
In most cases ordering processes in alloys are described by using simple models (in general the Ising model) in which phenomenological pairwise interactions are introduced. The problem is therefore to enquire whether such interactions can actually be defined from the electronic structure of the alloy (elastic effects will not be considered here). This is possible in the case of normal metals (alkaline metals, ...) by starting from the pseudopotential theory [l]. Though problems remain for treating non isoelectronic alloys, appreciable progresses have been made in this field these last years [2] and it seems that careful pseudopotential calculations can lead to realistic results [3]. The case of transition metals is a priori more complex since the total energy of a pure metal cannot be written as the sum of pair interactions. Rather accurate calculations of the cohesive energy of these metals have nevertheless been performed [4]. In the other hand a theory is now available (the coherent potential approximation : CPA [5]) for describing concentrated disordered alloys ; thus energies of formation can be estimated [6]. The CPA can be extended in order to deal with partial ordering ; in this way ordering energies can be obtained [7]. It turns out that these ordering energies vary as the square of the order parameter, exactly as in the phenomenological theories [8]. It is then possible to define a generalized perturbation method and therefore effective pair interactions [9]. Some calculations have been made with very simple models ; they are in qualitative agreement with the experimental results, at least for the ClCs-type compounds of the series TiFe, VMn, . . . These pair interactions have some particular properties : - they depend on the concentration ; this was also true in the case of normal metals, but this effect can be larger here. For example the sign of some interactions can change as a function of the concentration [10] ; - they are short ranged, except when the atomic potentials are not very different from the average atomic potential ; - they are specific to ordering processes and should not be used for calculating other physical propertiesKeywords
This publication has 0 references indexed in Scilit: