Extended spin models and quantum statistics in atomistic simulations

Abstract: An extended atomistic spin model allowing for studies of the finite temperature magnetic properties of alloys is proposed. The model is obtained by extending the Heisenberg Hamiltonian via a parameterization from a first principles basis, interpolating from both the low temperature ferromagnetic and the high temperature paramagnetic reference states. This allows us to treat magnetic systems with varying degree of itinerant character within the model. Satisfactory agreement with both previous theoretical studies and experiments are obtained in terms of Curie temperatures and paramagnetic properties. The proposed model is not restricted to elements but is also applied to binary alloys, such as the technologically important material Permalloy, where significant differences in the finite magnetic properties of Fe and Ni magnetic moments are found. The proposed model strives to find the right compromise between accuracy and computational feasibility for accurate modeling, even for complex magnetic alloys and compounds.
Traditionally, finite temperature atomistic simulations use classical (Boltzmann) statistics. We have incorporated quantum (Bose-Einstein) statistics instead and this corrects the low temperature properties in particular which is reflected in both magnetization and the magnetic specific heat, the latter allowing for improved modeling of the magnetic contribution to free energies. A central property of the method is the magnon density of states at finite temperatures, and we have compared several different implementations for obtaining it. The method has no restrictions regarding chemical and magnetic order of the considered materials. This is demonstrated by applying the method to elemental ferromagnetic systems, including Fe and Ni, as well as Fe-Co random alloys and the ferrimagnetic system GdFe3.