Our lab asks how electron spin — not just charge — can be used to store, move, and process information. That question sits behind two technologies we're actively working toward: spin-orbit-torque MRAM, a faster and more energy-efficient alternative to conventional memory, and probabilistic bits for neuromorphic (brain-inspired) computing. We get there by growing our own magnetic thin films, building nanoscale devices from them, and reading out their spin and charge transport with precision low-noise measurements.
Our research spans three major thrust areas at the intersection of magnetism, spintronics, and advanced materials.
Conventional memory is either fast or non-volatile, rarely both. Spin-orbit torque devices promise to change that — and we study the physics that determines how well they can. A central theme in spintronics is the inter-conversion of charge and spin currents. Recently, a focus has been on magnetic insulators where spin transport occurs through spin-wave propagation and spin currents can be generated by either spin injection or thermal gradients. These phenomena can be studied in simple bilayer films consisting of a ferrimagnetic (FIM) or Antiferromagnetic (AFM) insulator, and a heavy metal (HM) with large spin-orbit coupling such as Pt,W and so on. Spin to charge current conversion in such bilayers occurs by the inverse spin-Hall and Rashba-Edelstein effects.
Spin to charge conversion enables determination of the spin Seebeck effect (SSE). A thermal gradient across the FIM film produces a spin current into a neighboring heavy metal film, resulting in a transverse charge current or a voltage across the heavy metal film in an open circuit situation. Recent studies have demonstrated significant torques associated with thermally generated spin currents in magnetic tunnel junctions. Thermal gradients, thus, provide a convenient route to characterize the spin transport as well as a means to study the inverse effects, such as the spin torque on the AFM/ FIM magnetization in response to spin currents associated with charge current flow in the HM.
Schematic diagram of SOC and SOT phenomenon: (a) A three-dimensional representation of SOT, (b) Schematic diagram of spin Hall effect, (c) Rashba–Edelstein effect with an internal electric field ER perpendicular to the film plane, (d) SOT effect with field like and damping like torque.
In SMDL lab, we design spin–orbit torque (SOT)-driven devices to enable reconfigurable memory and neuromorphic logic. By combining SOT with magnetic fields, we achieve multistate memristive behavior, advancing beyond binary switching. Our work also surveys the potential of 2D van der Waals materials to enhance SOT effects for ultralow-power spintronic applications. Recent work from the group includes a self-induced spin-orbit torque discovered in nominally symmetric MgO/NiFe/MgO trilayers, traced to an interfacial dead layer (Physical Review Applied, in press), a phase-locked optical modulation technique that actively suppresses planar Nernst artefacts in harmonic Hall measurements (ACS Applied Electronic Materials, in press), and current-driven skyrmion dynamics in defect-engineered racetracks aimed at multistate and spin-logic memory.
Emergence of considerable thermoelectric effect due to the addition of an underlayer in Pt/Co/Pt stack
Pt/Co/NiFe/Pt stack with the experimental geometries and SOT directions.
Initialization-free multistate memristor: Synergy of spin–orbit torque and magnetic fields
Berry curvature and topology give 2D materials torque behavior ordinary magnets don't show — behavior we want to turn into practical switching mechanisms. We are actively investigating unconventional torque mechanisms arising from Berry curvature, topological effects, and nonlinear Hall phenomena in 2D materials and heterostructures. This work aims to unlock new pathways for achieving field-free switching, probabilistic bit behavior, and next-generation neuromorphic architectures. This includes engineering a large, tunable Berry curvature dipole in sputtered polycrystalline magnetic multilayers to realize a pronounced nonlinear Hall effect, and extending these ideas to exfoliated two-dimensional crystals such as MoTe2 and Ta3NiSe6 as part of a broader multi-year program moving toward twisted and magic-angle 2D systems.
Typical 2D device.
Many magnetic configurations stay hidden to standard measurements — our characterization techniques are built to reveal them. We investigate the magnetic states and switching behaviors in nanoscale systems like nanotubes and multilayer films, using advanced techniques such as remanent ferromagnetic resonance and planar Hall measurements. These studies reveal hidden magnetic configurations and enable new device functionalities. We are also developing noise spectroscopy — including 1/f and thermal Hall noise measurements on Heusler and van der Waals magnets — as a new diagnostic for topological band structure and domain dynamics, alongside microwave instrumentation such as loop-gap cavity resonators and compact coplanar-waveguide geometries for spin-torque ferromagnetic resonance.
Comparison of experimentally obtained M-H measurement and simulated M-H for nanotubes.
Variation of the FMR absorption with frequency at remanent field (RFMR)