Topological insulators have been demonstrated to have record-large spin-orbit torque efficiency by our group both at low temperature [1] and room temperature [3]. This ultrahigh efficiency is attributed to the unique spin-momentum locking of the topological surface states. Thanks to the insulating bulk, the spin-orbit torque can be effectively tuned by the gate voltage [3], promising future electric-field control of spin-orbit torque devices. However, the integration of room temperature magnetic thin film with perpendicular magnetic anisotropy has proven hard due to the two-dimensional nature of the topological insulators (Bi2
Se3
family). Recently, we utilize the Mo as a buffer layer and achieve room temperature perpendicular magnetic anisotropy in a topological insulator/Mo/CoFeB/MgO heterostructure [4], which is compatible with industry back-end-of-line process. We are working on the full realization of topological insulator-based spin-orbit torque magnetic memory.
Magnetic insulators are attractive to spintronic studies thanks to their low damping and thus long spin information propagation length [1]. Different from magnetic metals, the spin information in magnetic insulators propagates in the form of spin waves or magnons, which are free of mobile carrier scattering. However, there are two critical challenges for magnetic insulator-based spintronics/magnonics, write and read.
Write: Traditionally, magnetization of magnetic insulators can only been manipulated by external magnetic field, which is not energy efficient. While the spin transfer torque does not work due to their insulating nature, we have demonstrated that the spin-orbit torque from an adjacent heavy metal can efficiently switch the magnetization of magnetic insulator or excite magnetization resonance [2-3].
Read: The magnetization of magnetic insulator is difficult to readout since there is no electrical conduction. Recently, we have shown that the exchange coupling between magnetic insulator and heavy metal allows effective readout of information in magnetic insulators through anomalous Hall effect in the heavy metal layer [4].
Monolayer transition metal dichalcogenides, including MoS2
and WSe2
, have been shown to generate spin-orbit torque through Rashba-Edelstein effect [1]. To further improving spin-orbit torque efficiency, we are working on the spin-orbit torque generation from the 1T' phase WTe2, which has very strong spin-orbit coupling and nontrivial topology in the form of single layer.
Skyrmion is a whirlpool spin texture in magnetic materials, which could be the smallest spin texture in magnetic thin films in nature given the strong enough Dzyaloshinskii–Moriya interaction. The skyrmion in magnetic thin films at room temperature stabilized by the interfacial Dzyaloshinskii–Moriya interaction is promising [1] for practical applications due to its compatibility with the industry back-end-of-line process. A prototype skyrmion shift device for memory application has been demonstrated by our group [2]. Also, to eliminate the external field, we have employed the antiferromagnet to exchange bias the skyrmion and realized the zero-field skyrmion [3]. To reduce energy consumption, we are looking for skyrmions in room temperature magnetic insulators due to low damping. Signatures of skyrmions in magnetic insulators, topological Hall effect, have been observed [4].
Ferrimagnets provide great tunablity and are easy to manipulate compared with the antiferromagnets. Due to different g-factors in a ferrimagnet, the magnetization compensation temperature and the angular momentum compensation temperature can be different. At the magnetization compensation temperature, it behaves like a static antiferromagnet (from a energy point of view) and is immune to the external field. Meanwhile, it could support ultratiny spin texture (skyrmion) due to the absence of stray field. At the angular momentum compensation temperature, it behaves like a dynamic antiferromagnet and has very high characteristic resonance frequency. Meanwhile, the spin texture mobility shows a peak near this temperature.
Ferrimagnetic insulator has the advantage of ferrimagnetic metals and could potentially have much lower dissipation. We have explored the various ferrimagnetic garnet, like Tm3
Fe5
O12
and Tb3
Fe5
O12
from various perspectives. We show that their information, such as compensation temperature, can be readout through anomalous Hall effect [1]. Also, we show that the compensated ferrimagnetic insulator can be energy efficiently switched by the spin-orbit torque [2].
We are working on the switching of antiferromagnetic order using spin-orbit torques.
Here are a few keywords describing our research topics:
The techniques that we use frequently are the following:
Facilities in SQML:
Research platform: HKUST has state-of-the-art cleanroom (www.nff.ust.hk/) and material characterization facilities (http://www.mcpf.ust.hk/), which allow us to make nanometer-scale devices and observe atomic effects for many exciting and important applications. Based on this platform, SQML aims to be a device research lab with world-class thin film deposition, device fabrication, quantum transport, and microwave characterization tools.