Multiscale approach for modeling magnetization properties of inhomogeneous ultrathin magnetic layers

TL;DR

This paper introduces a multiscale modeling approach combining atomistic and macrospin methods to analyze the magnetic properties of inhomogeneous ultrathin iron layers, accounting for surface roughness and nanostructure effects.

Contribution

The study presents a novel multiscale framework that links atomic-scale roughness to macrospin dynamics, enabling more realistic modeling of ultrathin magnetic layers.

Findings

Magnetization dynamics are significantly affected by surface morphology.

Resonant frequency can decrease by up to an order of magnitude due to inhomogeneities.

Transition temperatures are modifiable by nanostructure shapes and roughness.

Abstract

We report on spin atomistic calculations used to model static and dynamic magnetic properties of inhomogeneous ultrathin iron films. Active magnetic layers in next-generation spintronic devices are becoming so thin that they exhibit some variable degree of roughness at the low-scale making them magnetically inhomogeneous. We propose a multiscale approach to progressively shift from a rough atomic-scale system to an ensemble of macrospins. By studying nanoscale islands of atoms in contact with each other, we demonstrate that ultrathin rough layers can be described by a set of macrospins coupled by a Heisenberg-like exchange interaction driven by the existence and shape of nanoconstrictions linking the islands. We show that the magnetization dynamics at 0 K is strongly impacted by this surface morphology since the resonant frequency of a typical ultrathin iron layer can drop by up to an order of magnitude due to inhomogeneities. Additionally, we used Monte Carlo simulations to determine the ferromagnetic-paramagnetic and spin reorientation transition temperatures for various morphology parameters and we show how nanoconstrictions and shapes of the atomic clusters can modify these transition temperatures. Our results demonstrate the possibility to account for the morphology of ultrathin structures with significant roughness. We believe that our approach makes it possible to model complete devices as close as possible to experimental reality.