Current Induced Spin Dynamics in Antiferromagnets

Abstract

This thesis focuses on experimental investigations of detection and control of Néel order in insulating antiferromagnetic thin films, specifically 𝛼-Fe2O3 and NiO. Antiferromagnetic materials have been getting increasing attention from the spintronics community because of their interesting properties and for potential applications in information processing. Some of these properties are zero net magnetization, low magnetic susceptibility, and very fast intrinsic spin dynamics with frequencies in the THz regime. Detecting and manipulating antiferromagnetic order electrically and reliably is a crucial milestone for realizing antiferromagnetic devices, but also an experimentally challenging one. Magnetic switching reported in antiferromagnetic thin filmsthusfaris overwhelmingly based on electronic transport methods, which provide only spatially averaged information on antiferromagnetic states. In addition, the resulting magnetoresistance from switching is usually less than 1% and can be masked by other effects due to the large current densities (e.g. electromigration, heating, thermal stresses, etc.). So, it is not always possible to identify the source of the observed magnetoresistance signal.

To tackle this problem, we first took a direct approach of imaging the magnetic domains using soft X-rays, with a technique known as X-ray Magnetic Linear Dichroism - Photoemission Electron Microscopy (XMLD-PEEM). This technique takes advantage of the dependence of the absorption of incoming light on its polarization and the underlying anisotropies of the material – in this case magnetic ones. The main advantage of using a microscopy technique isthat it can provide local information on antiferromagnetic domains which can help to reveal the principal switching mechanism. In our experiments, we observed the repeatable current-induced switching of antiferromagnetic moments consistent with the electrical measurements. Our analysis of individual light polarizations as a function of sample rotation revealed the position of the Néel vector in 3D space. Furthermore, we identified two types of switching based on the location and repeatability: reversible and irreversible switching. For regions that switch once and are outside of current path, magneto-elastic effects are the most likely dominant switching mechanism. But for reversible and along the current path regions, both spin-orbit-torques and magnetoelastic could be the contributors.

Our second approach for understanding current-induced antiferromagnetic dynamics was to design and conduct an experiment that measures the torque on the Néel vector. This is done with harmonic Hall measurements in antiferromagnet/heavy metal bilayer structures. Harmonic measurements on ferro- and ferrimagnetic systems have been used before to characterize spin-orbit torques (SOT), which are one of the most effective ways to manipulate magnetic order. However, their effectiveness is less well-explored and quantified for antiferromagnets. SOT are a result of the spin-Hall effect in the heavy metal, in which a charge current leads to a spin accumulation at the interface. They can modify the orientation of the AFM Néel vector and this change can be detected electrically because of spin Hall magnetoresistance. In our work, we demonstrated a way to extract SOTs by fitting harmonic responses to a simple model. Surprisingly, we found that the field-like torques are two orders of magnitude larger than damping-like torques in our 𝛼-Fe2O3 /Pt heterostructures, implying that the spin-mixing conductance of the interface has the unusual property of having a large imaginary component. This also points to magneto-elastic effects likely being the dominant mechanism of current-induced switching studies. We also used harmonic measurements to detect spin transport in ferrimagnet/ antiferromagnet/ heavy metal trilayer heterostructures. A spin current generated by a thermal gradient – also known asthe spin Seebeck effect – in a ferrimagnet can travel across antiferromagnetic NiO layer and be detected by inverse spin-hall effect in the heavy metal layer. Our results showed that not only is spin transport possible through thin NiO layers but there is also a sharp increase in the spin Seebeck effect magnitude at intermediate temperatures. These results suggest that this technique can be used to characterize the response of thin antiferromagnetic materials and optimize their spin-transport characteristics.

In summary, we first show control of antiferromagnetic order using electrical pulses while detecting the resulting changes using x-rays. Then, we used harmonic measurements to quantify spin-orbit torques present in the same system. We also extended the harmonic studies to trilayer structure to explore spin transport across antiferromagnetic layer. Overall, our studies open up a promising path for future studies similar AFM/HM heterostructures, as well as a means that can be used in optimizing SOT on AFM for applications.

Download Paper