Abstract
The authors present an extension of the blip-function expansion of Weeks, Chandler and Andersen (WCA) (1971) to liquid binary alloys. It turns out that the fact that realistic interionic pair potentials tend to be non-additive for most alloys poses a serious problem. They propose to resolve this problem by modifying the separation of the pair potentials in repulsive short-range and long-range terms in such a way that the repulsive parts are always additive, the non-additivity being entirely absorbed in the long-range terms. They show that this results in a first important improvement of the structure of liquid alloys over a simple hard-sphere picture. However, we emphasise that this should only be considered as a first step towards a systematic application of thermodynamic perturbation theory to liquid alloys.