Electronic energy bands corresponding to the 3d and 4s states of nonmagnetic, bcc chromium were calculated using the augmented-plane-wave method. A charge density corresponding to the occupied states plus the atomic core was used to construct a new crystal potential. This procedure was repeated until the charge density was self-consistent. The self-consistent charge density is compared with the atomic density and with the initially assumed crystal density. A direct application of Slater's superlattice model of an antiferromagnetic band structure was made for chromium assuming spin-up localization on the cube corners and spin-down localization in the body-centered position. The same procedure for self-consistency is yielding a large net spin per site of greater than three Bohr magnetons of d-like character and less than 0.05 magnetons of the same sign for the conduction electrons. The spin densities for this situation are given along with a discussion of the approximations leading to this physically incorrect result.
No takes yet. Share an insight, caveat, or question.
A. C. Switendick (1966) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: