On the Theory of Metallic Lithium

Abstract

The pseudopotential formalism has been extensively applied to the interpretation of different physical properties of simple metals.^1–3 However there are some well-known drawbacks to this approach, (i) Its use requires an accurate knowledge of both the core-state eigenfunctions and the core-state eigenvalues. Unfortunately these are not known very accurately. In fact the core-state eigenvalue shifts in the metal as compared to the free ion.⁴ For this reason model potentials have been introduced.^4–7 However, Heine and Abarenkov,⁵ Ashcroft⁶ and Shaw⁷ make no attempt to base the shape of their model potentials on any physical principle. This and the missing information about the core shift show the importance of first-principle pseudopotentials. (ii) In reality, the electron gas screens the ionic pseudopotentials. This effect requires a knowledge of the dielectric function of the electron gas. However, there is still some uncertainty about the best approximation to include the effects of exchange and correlation.^8–12