First-principles calculation of the parameters used by atomistic magnetic simulations

TL;DR

This paper reviews methods for calculating parameters of atomistic magnetic models from first principles, emphasizing the importance of relativistic effects and the KKR Green function method for accurate finite-temperature and dynamical magnetic simulations.

Contribution

It provides an overview of computational techniques to derive magnetic Hamiltonian parameters from first principles, highlighting the role of spin-orbit coupling and relativistic calculations.

Findings

KKR Green function method effectively calculates magnetic parameters.

Relativistic SDFT is crucial for accurate magnetic anisotropy and DMI.

New computational scheme for spin-lattice interactions.

Abstract

While the ground state of magnetic materials is in general well described on the basis of spin density functional theory (SDFT), the theoretical description of finite-temperature and non-equilibrium properties require an extension beyond the standard SDFT. Time-dependent SDFT (TD-SDFT), which give for example access to dynamical properties are computationally very demanding and can currently be hardly applied to complex solids. Here we focus on the alternative approach based on the combination of a parameterized phenomenological spin Hamiltonian and SDFT-based electronic structure calculations, giving access to the dynamical and finite-temperature properties for example via spin-dynamics simulations using the Landau-Lifshitz-Gilbert (LLG) equation or Monte Carlo simulations. We present an overview on the various methods to calculate the parameters of the various phenomenological Hamiltonians with an emphasis on the KKR Green function method as one of the most flexible band structure methods giving access to practically all relevant parameters. Concerning these, it is crucial to account for the spin-orbit coupling (SOC) by performing relativistic SDFT-based calculations as it plays a key role for magnetic anisotropy and chiral exchange interactions represented by the DMI parameters in the spin Hamiltonian. This concerns also the Gilbert damping parameters characterizing magnetization dissipation in the LLG equation, chiral multispin interaction parameters of the extended Heisenberg Hamiltonian, as well as spin-lattice interaction parameters describing the interplay of spin and lattice dynamics processes, for which an efficient computational scheme has been developed recently by the present authors.