Multiscale micromagnetic-atomistic modelling of the effects of MgO/CoFeB interfacial anisotropy, Dzyaloshinskii-Moriya interactions, and spin transfer torque in HDD TMR reader

Abstract

Finite difference multiscale micromagnetic – atomistic HDD reader model has been developed, allowing to simulate realistic device designs, while benefiting from atomic scale resolution in the most critical parts, such as tunnelling barrier or interface between ferromagnetic and antiferromagnetic layers. The model features LLG and LLB solvers for dynamics calculations, with consistent time steps for the atomistic and micromagnetic regions; constrained Monte-Carlo solvers for modelling of equilibrium properties, such as temperature dependence of magnetization and magnetocrystalline anisotropy. In this presentation we will demonstrate some of the model’s capabilities by considering effects of perpendicular anisotropy at the reference layer CoFeB / tunnel barrier MgO / free layer CoFe interfaces, as well as Dzyaloshinskii-Moriya interactions (DMI) arising at the same interfaces. Furthermore, effects of spin transfer torque (STT), introduced in the model via Slonczewski formalism are considered. First we compare fully atomistic, fully micromagnetic, and multiscale atomistic-micromagnetic models of a limited size trilayer, consisting of a CoFe pinned layer, antiferromagnetically coupled to a CoFeB reference layer via RKKY exchange coupling through a thin Ru layer (forming a synthetic antiferromagnet structure), and a free layer, separated from RL by an MgO tunnelling barrier. In-plane and out-of-plane field sweeps are used to characterize magnetic behaviour and to obtain the spin valve’s transfer curves, and FMR spectra are used to characterize layers’ anisotropy and magnetic noise, induced by thermal fluctuations. Then we present results of a full multiscale reader model, where the benefits of having an atomistic region around the tunnelling barrier are used to realistically simulate temperature dependent effects of interfacial anisotropy, DMI, and STT on the reader’s performance.