Spin-polarized Electronic Band Structure Research Articles

Effects of band structure and spin-independent disorder on conductivity and giant magnetoresistance in Co/Cu and Fe/Cr multilayers.

We develop a tight-binding model for calculating conductivity in multiband spin-dependent systems within the Kubo-Greenwood formalism. The model includes spin-independent disorder in the on-site atomic energies, representing intrinsic defects in real systems, and realistic spin-polarized electronic band structure. The model is applied to calculating conductivity in elemental $3d$ metals and giant magnetoresistance (GMR) in magnetic Co/Cu and Fe/Cr multilayers. We find that for disorder, producing a realistic resistivity of the multilayers, the values of GMR are in quantitative agreement with those observed experimentally. We demonstrate how the conductivity and GMR depend on the features of the electronic band structure and degree of disorder in the system. In particular, we show that (i) the $d$ electrons make an important contribution to the current in magnetic $3d$ metals and multilayers, (ii) the $sp\ensuremath{-}d$ hybridization is crucial for GMR, (iii) increasing disorder causes a decrease of the spin asymmetry of conductivity in magnetic metals and a drop of GMR in multilayers, (iv) GMR for the current-perpendicular-to-plane geometry is typically a factor of 2 higher than that for the current-in-plane geometry, and (v) the semiclassical treatment of conductivity applied to magnetic multilayers results in overestimated values of GMR due to the neglect of interband transitions.

Magnetocrystalline anisotropy of iron within limitation of standard band approach

An attempt to get more insight into the ab initio calculations of magnetic anisotropy in ferromagnetic 3 d transition metals is presented. The calculations are performed for bulk crystalline iron. We use the tight-binding Slater-Koster (SK) interpolation scheme for the self-consistent spin-polarized electronic band structure of the LMTO-ASA method. Electronic structure effects produced by the spin-orbit (SO) interaction are studied within the perturbation theory up to fourth-order terms. There is only one adjustable parameter which describes SO coupling. All other parameters in this paper are calculated from first principles. The discrepancy between calculated and experimental values of anisotropy constants is significantly reduced by taking into account the fourth-order corrections alone. The widely discussed numerical precision of the total energy calculations and the local density approximation of the density functional formalism seem to be important. However these are not the only possible deficiencies of the general theoretical framework which are essential for handling the problem of magnetocrystalline anisotropy in itinerant magnets. In this context the possibility of an hitherto unexploited approach to the calculation of the magnetocrystalline anisotropy is indicated.