🔬 mesoscale physics

Two Strategies to Measure Spin-Orbit-Torque Efficiency Acting on the Insulating Magnet Li $_{0.5}$ Al $_{0.7}$ Fe $_{1.8}$ O $_4$

This paper evaluates two strategies for measuring spin-orbit-torque efficiency in the insulating magnet Li $_{0.5}$ Al $_{0.7}$ Fe $_{1.8}$ O $_4$ , finding that lateral + longitudinal spin-torque ferromagnetic resonance yields a reliable efficiency of ~0.07 while Sagnac interferometry, though sensitive, produces conflicting results likely due to non-magnetic contributions.

Original authors: Orion Smedley, Thow Min Jerald Cham, Daisy O'Mahoney, Sanyum Channa, Xin Yu Zheng, Anna Janni, Lauren J. Riddiford, Yuhan Liang, Bozo Vareskic, Zbigniew Galazka, Yunqiu Kelly Luo, Yuri Suzuki, Daniel

Published 2026-09-15

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Orion Smedley, Thow Min Jerald Cham, Daisy O'Mahoney, Sanyum Channa, Xin Yu Zheng, Anna Janni, Lauren J. Riddiford, Yuhan Liang, Bozo Vareskic, Zbigniew Galazka, Yunqiu Kelly Luo, Yuri Suzuki, Daniel C. Ralph

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of modern electronics, scientists are constantly searching for ways to move information without moving physical matter. One promising avenue involves using the "spin" of electrons—a fundamental property that makes them act like tiny magnets—to carry data. This field, known as spintronics, relies heavily on materials that can generate and manipulate these magnetic signals. A particularly exciting class of materials for this purpose is magnetic insulators. Unlike the metal wires we use every day, these materials do not conduct electricity, which means they do not waste energy as heat when signals pass through them. They also possess a unique ability to transmit magnetic waves over long distances with very little loss, making them ideal candidates for future, ultra-efficient computing devices. However, because they do not conduct electricity, measuring how effectively they respond to magnetic forces is notoriously difficult. Standard electrical tools often fail to distinguish the true signal from background noise, and traditional optical methods frequently lack the sensitivity to see the subtle changes in these thin, non-conductive films.

To solve this puzzle, researchers at Cornell and Stanford universities turned their attention to a specific magnetic insulator called lithium aluminum ferrite. They wanted to measure how efficiently a layer of platinum, a heavy metal, could twist the magnetism of this insulator when an electric current flowed through the platinum. This twisting force, known as spin-orbit torque, is the engine that would drive future magnetic devices. The team employed two distinct strategies to capture this effect. The first method involved a technique called spin-torque ferromagnetic resonance, which uses microwaves to shake the magnetic material and measures how it responds. By carefully analyzing the electrical signals generated during this process, they were able to separate the true magnetic twisting force from other confusing electrical artifacts that often plague measurements on insulating materials. This approach revealed a clear, measurable efficiency: the platinum layer was able to exert a torque on the insulator that was comparable to what is seen when platinum acts on standard metal magnets.

The second strategy attempted to measure the same effect using light. The researchers used a highly sensitive optical device called a Sagnac interferometer, which is designed to detect tiny changes in the orientation of a magnetic field by observing how light reflects off the surface. While this method is powerful enough to detect signals from very thin films, the results it produced were puzzling. The optical measurements suggested that the twisting force was far weaker than what the microwave method had found—so weak, in fact, that it seemed physically inconsistent with other known behaviors of the material. The researchers concluded that the optical technique was likely being misled by a hidden factor. They suspect that the light was not just detecting the tilt of the magnetic insulator itself, but was also picking up a signal from the electrons in the platinum layer that were temporarily misaligned. Because the magnetic insulator interacts so weakly with light, this secondary signal from the metal layer overwhelmed the true signal from the insulator, leading to an inaccurate reading.

Ultimately, the study demonstrates that while optical methods are powerful, they can be deceptive when applied to insulating magnets with very weak optical signatures. The microwave-based resonance method proved to be the more reliable tool in this specific context, providing a trustworthy measurement of how effectively spin-orbit torque can act on these promising materials. The findings confirm that lithium aluminum ferrite is a viable candidate for next-generation spintronic devices, capable of being manipulated efficiently by standard heavy metals. By successfully distinguishing the true magnetic signal from the noise, the researchers have cleared a path for more accurate measurements in the field, ensuring that future developments in energy-efficient computing are built on solid, verified data rather than misleading artifacts.

Technical Summary: Two Strategies to Measure Spin-Orbit-Torque Efficiency Acting on the Insulating Magnet Li0.5Al0.7Fe1.8O4

Problem Statement
Insulating magnetic materials offer significant potential for spintronics due to low Gilbert damping constants, which facilitate efficient magnetic manipulation and long-distance magnon transmission. However, characterizing the spin-orbit torque (SOT) efficiency in these systems presents distinct challenges. Electrical measurement techniques, such as second-harmonic Hall detection, often suffer from poor signal-to-noise ratios and are susceptible to artifacts like current-induced magnon creation and annihilation, which can lead to order-of-magnitude errors in efficiency calculations. Conversely, conventional optical methods based on magneto-optical Kerr effect (MOKE) imaging generally lack the sensitivity required to measure SOT efficiency in thin films (approx. 10 nm) of magnetic insulators, which typically exhibit very weak magneto-optical coupling coefficients compared to metallic ferromagnets.

Methodology
The authors investigated two distinct strategies to measure the damping-like spin-orbit torque efficiency ( $\xi_{DL}$ ) acting on thin films (14.4 nm) of the low-damping magnetic insulator lithium aluminum ferrite (Li $_{0.5}$ Al $_{0.7}$ Fe $_{1.8}$ O $_4$ , or LAFO) in a bilayer with Platinum (Pt, 2 nm).

Lateral + Longitudinal Spin-Torque Ferromagnetic Resonance (ST-FMR): The authors employed a lateral + longitudinal ST-FMR geometry. This approach utilizes both longitudinal ( $V_{XX}$ ) and transverse/Hall ( $V_{XY}$ ) DC mixing voltages generated under microwave excitation. By analyzing the angular dependence of these voltages relative to the in-plane magnetic field, the method allows for the mathematical separation of the true spin-torque signal from artifacts caused by spin-pumping voltages and resonant heating, which typically confound conventional longitudinal ST-FMR measurements in insulating systems.
Sagnac Interferometry: As an alternative to conventional MOKE, the authors utilized Sagnac interferometry to measure current-induced magnetic deflection. This technique measures the phase difference between time-reversed circularly polarized light beams, offering high sensitivity (50–100x better than unmodulated MOKE) and inherent rejection of thermal artifacts. The procedure involved calibrating the magneto-optic coupling constant ( $\kappa$ ) via out-of-plane field sweeps and then measuring the modulation of the Kerr rotation ( $\Delta\theta_k$ ) induced by an AC current in the presence of an in-plane magnetic field.

Key Results

ST-FMR Measurements: Using the lateral + longitudinal ST-FMR technique, the authors determined a damping-like spin-orbit torque efficiency of $\xi_{DL}^{ST-FMR} \approx 0.07$ (specifically 0.07(1) from AMR-based extraction and 0.08(2) from PHE-based extraction). This value is comparable to, though slightly lower than, efficiencies observed for Pt acting on metallic ferromagnets (typically 0.09–0.12). The study notes that neglecting the spin-pumping correction in a conventional analysis would have yielded a significantly underestimated value of $\approx 0.029$ .
Sagnac Interferometry Measurements: The Sagnac measurements successfully detected signals from the 14.4 nm LAFO films, demonstrating the technique's sensitivity. However, the inferred torque efficiency was significantly lower, yielding $|\xi_{DL}^{Sagnac}| \approx 0.0009$ to $0.0018$. This result is more than an order of magnitude smaller than the ST-FMR result and is inconsistent with other observations, such as the ability of Pt/LAFO systems to drive steady-state nano-oscillations at similar current densities.
Discrepancy Analysis: The authors found that the Sagnac signal did not scale as expected with the magnetic field strength, deviating from the theoretical $(B + \mu_0 M_{eff})^{-1}$ dependence.

Significance and Claims
The paper concludes that the lateral + longitudinal ST-FMR method provides a more reliable and trustworthy measurement of SOT efficiency for insulating magnets like LAFO, yielding a value ( $\xi_{DL} \approx 0.07$ ) that aligns with physical expectations for Pt interfaces.

Regarding the Sagnac results, the authors speculate that the discrepancy arises because the Sagnac signal in insulating magnets with very weak magneto-optical coupling is not solely determined by the tilting of the magnetization. They propose that a net out-of-plane canting of the spins of itinerant electrons in the Pt layer—induced by the exchange field of the LAFO and spin-orbit scattering—may generate a Kerr rotation signal that opposes the signal from the magnetization tilting, thereby suppressing the net measured signal.

The study establishes that a torque efficiency of $\xi_{DL} \approx 0.07$ is sufficient to support the integration of low-damping magnetic insulators with heavy metals for investigating magnon generation, transmission, and control. The work highlights the necessity of separating artifact contributions in electrical measurements and cautions against relying solely on optical deflection measurements for insulators with weak magneto-optic coefficients without accounting for potential contributions from itinerant electron spins.

More like this