New Interpretation of Spin-Wave Behavior in Nickel

Abstract

Results from recent calculations of the neutron-scattering cross section for energy transfer up to 1 eV have led to a new picture of spin waves in nickel, one in which spin waves exist out to the zone boundary. Our prediction of the dispersion curve and peak widths is in general agreement with data from subsequent neutron-scattering experiments. This provides additional evidence that the "average" spin-splitting energy for nickel is in the 300-to 400-meV range.

Wannier-function approach to spin excitations in solids

“…84 The authors calculated the magnetic susceptibility within the RPA for rigidly spin-split bands, but the agreement with the available experimental data was not good. Cooke et al [23][24][25] showed that it is necessary to take the k dependence of the exchange splitting into account for a quantitative comparison between theory and experiment. They used a tight-binding description of the electronic energy bands, and the ferromagnetism was driven by an empirical on-site Coulomb interaction between the 3d electrons with two adjustable parameters.…”

“…In previous treatments of spin waves in a many-body context the Coulomb interaction was chosen either as a simple Hubbard-type U parameter or as a model potential with an adjustable range parameter. 24,25,33 Recent ab initio studies of the bare and the screened Coulomb interaction in 3d transition metals focused only on the NM state. 48,56,[61][62][63][64][65][66] Here we present a detailed study of the matrix elements in the MLWF basis for the proper FM state of the 3d transition metals Fe, Co, and Ni.…”

“…These authors studied the spin-wave dispersion of 3d ferromagnets and obtained reasonable agreement with experiments; in particular, the optical branch in the spin-wave dispersion of fcc Ni was correctly described. 25 Using a similar approach, Mills and co-workers [27][28][29][30] carried out extensive calculations to explore the spin dynamics in ultrathin ferromagnetic ͑FM͒ films on nonmagnetic ͑NM͒ substrates. Recently more realistic treatments of the spin-wave spectra in 3d ferromagnets were reported by Savrasov 31 and Buczek et al 32 within time-dependent DFT ͑TDDFT͒ and by Karlsson and Aryasetiawan 33 within MBPT.…”

“…23,24 First-principles calculations of the magnetic response function using realistic energy bands and wave functions are very rare. Initial attempts by Cooke et al [23][24][25][26] within the framework of many-body perturbation theory ͑MBPT͒ were based on a tight-binding description of the electronic energy bands. These authors studied the spin-wave dispersion of 3d ferromagnets and obtained reasonable agreement with experiments; in particular, the optical branch in the spin-wave dispersion of fcc Ni was correctly described.…”

Theoretical probing of inelastic spin-excitations in adatoms on surfaces

“…Since these types of calculations for magnetic materials are computationally expensive, very few DFT-based calculations, even for bulk systems, were performed [19,20], and it is only recently that a few groups have undertaken the task of calculating dynamical transverse magnetic response functions for bulk materials [21][22][23], thin films [24] and adatoms or nanostructures on surfaces [28,29]. Several studies based on empirical tight binding (ETB) theory and MBPT were published [25][26][27] which, although relying on the accuracy of fitting parameters, have advanced the understanding of many effects accompanying spin-excitations. Our first-principles scheme uses TD-DFT (see Refs.…”