Abstract
An interpretation of empirical rules proposed by Miedema and collaborators to predict alloying tendencies amongst metals and heats of formation is elaborated. In these rules the alloy heat of formation is described by two parameters, φ*, the empirically adjusted electron workfunction andssthe interstitial electron density of the pure metal. The parameter n establishes a scale of attractiveness for impurities. However, since φ* and n are highly correlated, it is better to focus attention on φ d , that part of φ* which is independent of n. For non-transition metals the parameter φ d is related to non-locality or angular momentum dependence of the pseudo-potential which is important because the impurity screening cloud has an enhanced s and reduced p and d character; this is illustrated by soft X-ray emission data for AlMg alloys. The optimized form of the model potential has been used to reproduce φ d quantitatively for eight non-transition metals.