Pseudopotential Theory of Metallic Lithium

Publisher Website
Google Scholar
Abstract
A detailed calculation of the electronic structure of a screened lithium ion is made, with a view to testing some of the assumptions and approximations which have been used in pseudopotential work. Within the confines of a linearly screened Austin pseudopotential we conclude that for lithium, (i) the Slater approximation in the treatment of core-valence exchange is qualitatively incorrect, (ii) the phase shifts considerably overestimate the screening charge, and (iii) the core states are little different from those of the free ion. It is emphasized that when calculating the form factor, it is better to use the Born approximation, which automatically scales by describing the forward scattering correctly, than to use approximate phase shifts which violate significantly the Friedel sum rule.