A systematic calculation yields several substantial modifications to the Heine–Abarenkov model potential in a metal. Most important are an inhomogeneity correction to the valence–valence exchange and a large change in the energy at which the free ion model potential should be evaluated for an ion in a metal. The agreement with pseudopotentials fitted from Fermi surface data is excellent for Na, Mg, and Al. A discrepancy of about 0.01 Ry exists for K. Because the free ion model potential must be extrapolated quite far below the valence orbital energies, the use of a linear extrapolation is unlikely to be valid for any ion having a d state as its highest core level.
No takes yet. Share an insight, caveat, or question.
Ballentine et al. (1971) studied this question.