Answers are grounded in this paper and cite their sources.
From ab initio pseudopotential calculations of the solid the authors extract atomic-like electron densities which, when overlapped in the crystal, reproduce the self-consistent density except for those components for which the structure factor is zero. They demonstrate the universality of the optimized densities for different crystal structures at varying volumes. These densities are the optimized choice for the Harris functional, and they compare their result with the Finnis contraction of the free atom density and with the effective medium theory ansatz of embedding an atom in a homogeneous electron gas.
No takes yet. Share an insight, caveat, or question.
Chetty et al. (1991) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: