Domain Wall Motion and Interfacial Dzyaloshinskii-Moriya Interactions in Pt/Co/Ir$(t_\mathrm{Ir})$/Ta Multilayers

TL;DR

This study investigates how adding Ir layers to Pt/Co/Ta multilayers affects interfacial Dzyaloshinskii-Moriya interactions, revealing increased DMI strength, changes in domain wall behavior, and the importance of advanced modeling for accurate DMI measurement.

Contribution

It introduces a comprehensive analysis of Ir thickness effects on DMI in multilayers and proposes a micromagnetic model to better interpret experimental velocity data.

Findings

Ir increases magnetic moment by suppressing dead layers

DMI strength correlates with Ir thickness and is similar at Ir/Co and Pt/Co interfaces

Velocity minima are influenced by in-plane field and depinning field variations

Abstract

The interfacial Dzyaloshinskii-Moriya interaction (DMI) is important for chiral domain walls (DWs) and for stabilizing magnetic skyrmions. We study the effects of introducing increasing thicknesses of Ir, from zero to 2 nm, into a Pt/Co/Ta multilayer between the Co and Ta. We observe a marked increase in magnetic moment, due to the suppression of the dead layer at the interface with Ta, but the perpendicular anisotropy is hardly affected. All samples show a universal scaling of the field-driven domain wall velocity across the creep and depinning regimes. Asymmetric bubble expansion shows that DWs in all of the samples have the left-handed N\'{e}el form. The value of in-plane field at which the creep velocity shows a minimum drops markedly on the introduction of Ir, as does the frequency shift of the Stokes and anti-Stokes peaks in Brillouin light scattering measurements. Despite this qualitative similarity, there are quantitative differences in the DMI strength given by the two measurements, with BLS often returning higher values. Many features in bubble expansion velocity curves do not fit simple models commonly used to date, namely a lack of symmetry about the velocity minimum and no difference in velocities at high in-plane field. These features are explained by the use of a model in which the depinning field is allowed to vary with in-plane field in a way determined from micromagnetic simulations. This theory shows that velocity minimum underestimates the DMI field, consistent with BLS returning higher values. Our results suggest that the DMI at an Ir/Co interface has the same sign as the DMI at a Pt/Co interface.