🔬 materials science

Generalized Helicity-Dependent Magnetization Switching via Substrate Phonons

This study demonstrates that helicity-dependent magnetization switching via substrate phonons is a general phenomenon across diverse materials, revealing that its efficiency is governed by the heterostructure's full optical response rather than substrate absorption alone, as evidenced by systematic spectral shifts and interface-localized heating effects.

Original authors: F. G. N. Fennema, H. Damas, H. Yoshikawa, A. Tsukamoto, C. S. Davies, A. Kirilyuk

Published 2026-09-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: F. G. N. Fennema, H. Damas, H. Yoshikawa, A. Tsukamoto, C. S. Davies, A. Kirilyuk

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 microscopic world of magnets, scientists have long known that heat can scramble the orderly alignment of atomic spins, turning a magnet into a non-magnetic lump. But a more recent discovery has revealed a subtler, faster way to control magnetism: using the vibration of the material itself. Imagine the atoms in a solid not as static bricks, but as a crowd of people constantly jiggling. When these vibrations are forced to move in a specific, circular pattern, they can generate a tiny, effective magnetic field. This phenomenon, involving circularly polarized vibrations, offers a potential shortcut to flipping magnetic bits without using traditional electrical currents, a capability that could revolutionize how we store data. The question researchers have been asking is whether this effect is a rare trick that only works on specific, exotic materials, or if it is a universal rule that applies to almost any solid crystal.

A team of physicists has now answered this by testing the effect across a wide variety of common materials. They built a simple sandwich structure: a thin layer of a magnetic alloy called GdFeCo, placed on top of a solid crystal substrate. They then shone pulses of mid-infrared light onto the sample. This light was tuned to match the natural vibration frequency of the atoms in the substrate, causing the substrate to vibrate in a circular motion. The researchers used a special technique where the light's polarization was constantly switched, creating a pattern of alternating magnetic directions on the surface. If the circular vibrations were strong enough, they would force the magnetic layer to flip its direction in sync with the light's rotation.

The team tested seven different substrates, including titanium dioxide, quartz, and sapphire, which represent a diverse range of crystal structures and atomic arrangements. The results were clear and consistent: the magnetic layer flipped its direction on every single substrate they tried. This finding proves that the ability to switch magnetism using these substrate vibrations is not limited to a special few materials but is a broad, general phenomenon that works across many different types of crystals.

However, the story does not end with a simple "it works." The researchers discovered that the efficiency of this flipping process does not follow the most obvious rule. One might expect that the flipping would be strongest exactly where the substrate absorbs the most light energy, which happens when the light frequency perfectly matches the natural vibration of the atoms. Instead, the team found that the strongest flipping occurred at a slightly higher frequency, just before the peak absorption. Furthermore, when the light frequency did hit the peak absorption, the flipping actually became weaker, sometimes failing completely.

To understand why this happens, the researchers used computer simulations to track where the light energy went inside the material. They found that at the peak absorption frequency, the light energy gets trapped very close to the surface of the substrate, right next to the magnetic layer. This intense, localized energy deposition likely heats up the magnetic layer too much, which interferes with the flipping process. It is as if the energy meant to drive the switch is instead overheating the mechanism, causing it to jam. Away from this peak, the energy penetrates deeper and spreads out, avoiding this overheating problem and allowing the magnetic flip to happen cleanly.

The study also revealed that the strength of the effect depends on more than just how much light the material absorbs. It appears to be linked to how strongly the atoms in the crystal are charged and how they interact with each other during vibration. Some materials with similar absorption levels showed very different flipping efficiencies, suggesting that the internal "personality" of the crystal's vibrations matters just as much as the amount of light it catches.

Ultimately, this work shows that using the vibrations of a substrate to control magnetism is a robust and widely applicable method. It is not a fragile trick restricted to specific crystals, but a general physical principle. Yet, it also highlights that the process is delicate. The efficiency is not determined by a single factor like light absorption, but by a complex balance between exciting the vibrations and managing the heat they generate. By understanding this balance, scientists can better design future magnetic devices that use light to switch data at incredibly high speeds, turning the very ground beneath the magnetic layer into an active partner in the process.

Technical Summary: Generalized Helicity-Dependent Magnetization Switching via Substrate Phonons

Problem Statement
Recent experimental work has demonstrated that circularly polarized transverse-optical (TO) phonons in specific substrates (sapphire and glass-ceramic) can drive deterministic, helicity-dependent reversal of magnetization in adjacent magnetic layers. This phenomenon relies on the generation of effective magnetic fields by the rotational motion of circularly polarized ionic charges. However, the universality of this substrate-mediated mechanism remains unresolved. It is unclear whether this effect is confined to specific crystal symmetries or if it depends on specific optical properties of the substrate. Furthermore, it is unknown whether switching efficiency is governed solely by the strength of resonant phonon absorption or if other factors, such as the full optical response of the heterostructure, play a critical role.

Methodology
To address these questions, the authors investigated helicity-dependent magnetization switching in magnetic Gd $_{24}$ FeCo layers (20 nm) grown on seven different diamagnetic substrates: TiO $_2$ , $\alpha$ -SiO $_2$ , LiNbO $_3$ , KTaO $_3$ , ZnO, MgO, and Al $_2$ O $_3$ . These substrates span six different point groups and exhibit a wide variety of infrared-active phonon spectra.

The samples consisted of a magnetic layer sandwiched between Si $_3$ N $_4$ capping and interlayers. Excitation was performed using mid-infrared (mid-IR) laser pulses from the FELIX free-electron laser facility. The study employed a polarization-modulated transient grating excitation scheme, where two orthogonally polarized pulses create a spatially periodic pattern of alternating optical helicity. This allows for the writing of tracks of reversed magnetization that follow the local optical helicity.

Switching efficiency was quantified using Faraday microscopy to visualize the magnetic state. The efficiency was determined by analyzing the contrast of the magnetization reversal tracks relative to the background. To understand the underlying physics, the authors performed transfer-matrix calculations (using pyGTM) to simulate the electric field distribution and the time-averaged volumetric power loss density ( $P_{loss}$ ) within the heterostructures across the relevant wavelength range.

Key Results

Generalization of the Effect: Helicity-dependent switching was observed for all seven investigated substrates, regardless of their point group symmetry or specific phonon mode. This confirms that substrate-mediated phononic switching is a general phenomenon not restricted to specific crystal structures.
Spectral Mismatch: While the switching spectra broadly overlap with the Reststrahlen bands of the substrates, a systematic spectral mismatch was observed. The maximum switching efficiency consistently occurred on the high-frequency (shorter wavelength) side of the corresponding TO-phonon resonance, rather than at the resonance peak or the absorption maximum.
Decoupling from Absorption: The switching efficiency does not simply follow the substrate absorption spectrum. For instance, TiO $_2$ and $\alpha$ -SiO $_2$ exhibit comparable peak absorption coefficients, yet TiO $_2$ shows strong switching while $\alpha$ -SiO $_2$ shows only weak switching. Similarly, LiNbO $_3$ and KTaO $_3$ show different efficiencies despite comparable absorption peaks.
Role of Local Heating: Transfer-matrix simulations revealed that near the TO-phonon resonance, strong substrate absorption causes the optical energy deposition to become highly concentrated within the first 200 nm below the substrate interface. This localization suggests that as the resonance is approached, increased local heating of the adjacent magnetic layer may suppress the switching efficiency. The observed spectral optimum likely reflects a balance between efficient phonon excitation and the suppression of switching due to excessive interface-localized heating.
LO-TO Splitting Correlation: The study suggests that the LO-TO splitting (a measure of polar strength and Born effective charge) may serve as a qualitative indicator of a phonon mode's capacity to generate an orbital magnetic moment. Modes with larger LO-TO splittings generally correlated with higher maximum switching efficiencies, though this trend was not universal.

Significance and Claims
The paper establishes that substrate-mediated phononic switching is a broadly applicable mechanism for controlling magnetization in adjacent magnetic layers, extending beyond the previously studied sapphire and glass-ceramic systems.

Crucially, the authors demonstrate that the efficiency of this switching is not determined by substrate absorption alone. Instead, the response is governed by the full optical response of the heterostructure. The systematic shift of the switching maxima away from the TO-phonon resonances indicates that additional processes, such as local heating induced by the spatial confinement of energy deposition near the interface, significantly influence the outcome.

The results highlight the substrate as an active component of the heterostructure, where its phonon spectrum and optical response directly influence helicity-dependent magnetization reversal. While the study identifies local heating as a likely factor in the suppression of switching near resonance, the recovery of switching on the long-wavelength side of the resonance (where absorption decreases) remains an unresolved feature, suggesting that the spectral response cannot be described solely by direct resonant excitation of the identified TO mode.

More like this