Thermal Contribution in the Electrical Switching Experiments with Heavy Metal / Antiferromagnet Structures

TL;DR

This study demonstrates that thermal effects, specifically localized Joule heating, significantly influence Hall resistance signals in antiferromagnetic spin-orbit torque experiments, highlighting the need to account for thermal artifacts in experimental design.

Contribution

It provides evidence that thermal contributions dominate Hall signals in these experiments and introduces a method to semi-quantitatively evaluate thermal effects.

Findings

Hall signals are primarily due to localized Joule heating.

Opposite TCR of Ta and Pt causes opposite Hall signal changes.

Quadratic relationship confirms thermal origin of signals.

Abstract

We examine the thermal origin of the detected "saw-tooth" shaped Hall resistance (Rxy) signals in the spin-orbit torque switching experiment for antiferromagnetic MnN. Compared with the results of the heavy metal / antiferromagnet bilayers (MnN/Ta), the qualitatively same "saw-tooth" shaped signals also appear in the samples with the heavy metal layer alone (either Ta or Pt) without MnN layer. In addition, The Rxy signal changes oppositely in the devices with Ta and Pt, due to the opposite temperature coefficient of resistivity (TCR) of the two materials. All those results are consistent with the "localized Joule heating" mechanism in devices with Hall crosses geometry. Moreover, by utilizing a structure with separated writing current paths and Hall cross area, the quadratic relationship between delta-Rxy and the writing current's amplitude is observed, which provides quantitative evidence of the thermal contribution. These results reveal the dominant thermal artifact in the widely used Hall crosses geometry for Neel vector probing, and also provide a strategy to semi-quantitatively evaluate the thermal effect, which can shed light on a more conclusive experiment design.