Giant orbital-magnon conversion driven perpendicular magnetization switching

Technical Summary: Giant Orbital-Magnon Conversion Driven Perpendicular Magnetization Switching

Problem Statement
The development of beyond-Moore information technologies relies on efficient interconversion among electron charge, spin, orbital, and magnon degrees of freedom. While charge-to-spin (C-S) conversion has enabled spin-transfer torque (STT) and spin-orbit torque (SOT) devices, and spin-to-magnon (S-M) conversion allows for heat-free spin transport, a direct coupling between orbital angular momentum (L) and magnons (M) has remained elusive. Specifically, magnetization switching via orbital-to-magnon (L-M) conversion has not been achieved. A significant bottleneck in orbitronics is the inherently low L-S conversion coefficient in common ferromagnets (e.g., CoFeB, NiFe), which limits the strength of orbital torque and hinders perpendicular magnetization switching in orbital material-ferromagnet heterostructures.

Methodology
The authors fabricated bilayer and trilayer devices consisting of an orbital metal (Ti, W, or Ta) and an antiferromagnetic insulator (NiO) interfaced with a ferromagnet (CoFeB or Ni).

• Device Structures: Key structures included Ti (5 nm)/NiO ($t_{NiO}$)/CoFeB (6 nm) for spin-torque ferromagnetic resonance (ST-FMR) measurements, and Hall-bar devices with perpendicular magnetic anisotropy (PMA) such as Ti/NiO/Ti/CoFeB/MgO for switching and second-harmonic Hall voltage measurements.
• Characterization Techniques:
  • ST-FMR: Used to measure the damping-like torque efficiency ($\theta_{DL}$) and extract effective Hall conductivities at room temperature.
  • Second-Harmonic Hall Voltage: Employed to characterize torque efficiency and verify the magnon origin of the torque, specifically analyzing temperature dependence near the Néel temperature of NiO.
  • Current-Induced Switching: Pulsed currents (50 $\mu$s width) were applied under in-plane magnetic fields to induce perpendicular magnetization switching, verified by anomalous Hall resistance and magneto-optical Kerr effect (MOKE) microscopy.
• Modeling: A theoretical model was developed to separate the total effective Hall conductivity into orbital ($\sigma_{O}^{eff}$) and magnon ($\sigma_{M}^{eff}$) components, allowing for the quantitative extraction of the L-M conversion coefficient ($C_{MI}$) and diffusion lengths.

Key Results

1. Observation of L-M Conversion: In Ti/NiO/CoFeB devices, the damping-like torque efficiency ($\theta_{DL}$) increased with NiO thickness up to 3 nm before decreasing, a behavior distinct from direct orbital torque. This indicated the emergence of a magnon torque mediated by the insulating NiO layer.
2. Temperature Dependence: Second-harmonic measurements revealed that $\theta_{DL}$ peaked at characteristic temperatures (~60 K for 3 nm NiO, ~80 K for 5 nm NiO), coinciding with the magnetic ordering temperatures of the NiO layers. This confirms the torque is magnon-originated, as magnon currents are boosted by enhanced spin fluctuations near the Néel temperature.
3. Quantification of Conversion Coefficient: By modeling the thickness dependence of orbital and magnon conductivities, the authors determined the orbital diffusion length in NiO to be ~3.18 nm and the magnon diffusion length to be ~25.4 nm. The L-M conversion coefficient ($C_{MI}$) in NiO was quantitatively determined to be approximately 0.033, rising to ~0.06 when interface transparency is considered. This efficiency is over an order of magnitude higher than in traditional orbital systems lacking the L-M process.
4. Perpendicular Magnetization Switching: The study demonstrated robust, room-temperature perpendicular magnetization switching in CoFeB layers driven by the orbital-induced magnon torque. The switching polarity reversed with the direction of the in-plane magnetic field, confirming a current-induced torque mechanism.
5. Superior Efficiency: Devices utilizing the Ti/NiO bilayer exhibited a switching efficiency ($\eta$) nearly three times higher than control devices relying solely on spin torque (e.g., W/Ti/CoFeB) and more than 20 times higher than Ti/CoFeB bilayers lacking the NiO layer. In W-based devices, the dominance of the positive orbital-induced magnon torque over the negative spin-induced torque was observed when NiO thickness exceeded 2 nm, reversing the switching polarity.

Significance and Claims
The paper claims to report the first experimental demonstration of L-M conversion in an orbital metal/antiferromagnetic insulator bilayer at room temperature. The primary significance lies in establishing a direct link between orbitronics and magnonics. By utilizing the L-M conversion mechanism, the authors overcame the limitation of weak orbital torque in standard ferromagnets, achieving efficient magnetization control without the need for heavy metals with strong spin-orbit coupling. The findings provide a new platform for developing advanced nano-devices based on orbital-driven magnonic phenomena, offering a pathway toward energy-efficient data processing and storage with perpendicular magnetic anisotropy. The work suggests that magnetic insulators can convert orbital angular momentum into magnons with remarkable efficiency, significantly expanding the material palette for future orbitronic research.