Comment on ‘‘Magnetism and local order:Ab initiotight-binding theory’’

In a recent paper entitled ``Magnetism and local order: Ab initio tight-binding theory,'' Feng Liu et al. [Phys. Rev. B 39, 6914 (1989)] presented a model for the determination of the electronic structure of finite systems with reduced symmetry. The various matrix elements appearing in the tight-binding formulation are determined from the self-consistent treatment of the dimer. A simple method based upon the real space enables us to show that in the case of iron the intrasite matrix elements of the tight-binding calculation are not transferable to bulk and slabs without caution. Actually, the non-self-consistent calculation leads to a charge transfer that is at least an order of magnitude greater than that found in the present self-consistent one.