🔬 materials science

X-ray imaging of antiferromagnetic octupole domains in Mn $_3$ Sn

This study demonstrates the successful real-space imaging of antiferromagnetic octupole domains in Mn $_3$ Sn using scanning transmission X-ray microscopy with X-ray magnetic circular dichroism, confirming the bulk nature of the contrast and establishing a powerful method for investigating time-reversal symmetry-breaking antiferromagnets.

Original authors: Max T. Birch, Sebastian Wintz, Yuhan Sun, Akiko Kikkawa, Markus Weigand, Takahisa Arima, Yoshinori Tokura

Published 2026-08-10

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Max T. Birch, Sebastian Wintz, Yuhan Sun, Akiko Kikkawa, Markus Weigand, Takahisa Arima, Yoshinori Tokura

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

Imagine the world of tiny magnets as a bustling city of invisible dancers. For decades, scientists have been fascinated by two main types of dancers: the "Ferromagnets," who all hold hands and spin in the exact same direction, creating a strong, unified magnetic field that we can easily feel and use in things like fridge magnets and hard drives. Then there are the "Antiferromagnets," a more chaotic group where neighbors spin in opposite directions, canceling each other out perfectly. To the naked eye (and even to most sensors), these dancers seem to stand still, producing no net magnetic field at all. This makes them incredibly hard to see or control, yet they are the stars of a new, exciting field called spintronics. Why? Because they are super-fast, immune to outside magnetic interference, and don't leave behind messy "stray fields" that mess up neighboring devices.

Recently, a special new class of these anti-dancers has emerged: the "TRS-breaking" antiferromagnets. These are unique because, even though their neighbors cancel out, their specific dance formation breaks a fundamental rule of time symmetry, allowing them to behave a bit like ferromagnets in certain ways. Think of them as a group of dancers who, while spinning in opposite directions, are arranged in a pattern that creates a hidden, powerful "octupole" force—a complex, multi-pole magnetic shape that is usually invisible to standard cameras. The big question for scientists has been: How do we actually see these hidden patterns in real space? If we can't see the dancers, we can't choreograph the next generation of ultra-fast, super-efficient computer memory. This is where the story of a special material called Mn3Sn comes in.

The Invisible Dance Floor: Seeing the Unseeable

In this study, a team of researchers decided to take a closer look at a material called Mn3Sn, which acts like a non-collinear antiferromagnet. Imagine the atoms in this material not as a simple line of dancers, but as a hexagonal dance floor where the spins form a 120-degree triangle pattern. This specific arrangement creates a "magnetic octupole"—a fancy way of saying the spins form a complex, three-dimensional shape that breaks time-reversal symmetry. The problem? These octupoles are notoriously difficult to photograph. Standard magnetic cameras just see "nothing" because the spins cancel out.

To solve this, the researchers built a tiny, custom-made device using a focused ion beam (FIB), essentially using a super-precise atomic scalpel to carve a Hall bar (a specific shape for measuring electricity) out of a single crystal of Mn3Sn. They made this slice incredibly thin—about 250 nanometers thick, which is roughly 400 times thinner than a human hair—and sandwiched it between gold and platinum contacts. They even gave it a protective coat of aluminum oxide so it wouldn't rust in the air.

The Magic Flashlight

To see the invisible dance, the team used a powerful tool called Scanning Transmission X-ray Microscopy (STXM) at a giant particle accelerator called BESSY II. They didn't use a regular flashlight; they used a beam of soft X-rays tuned to a very specific energy: 639.25 electron volts (eV). This energy is just below the peak where manganese atoms usually absorb light (the Mn L3 edge).

Here is the clever trick: The researchers used circularly polarized light, which is like a beam of light that spins as it travels, either clockwise or counter-clockwise. When this spinning light hits the Mn3Sn, it interacts differently with the magnetic octupoles depending on their orientation. The researchers measured how much light was absorbed when the beam was spinning one way versus the other. The difference between these two measurements is called X-ray Magnetic Circular Dichroism (XMCD).

Usually, this difference is zero for antiferromagnets. But because of the unique "octupole" shape in Mn3Sn, the researchers found a tiny, but measurable, difference. They detected a contrast of about 0.2%. To put that in perspective, if the total light intensity were a full glass of water, the difference they saw was less than two drops. Yet, their instrument was so sensitive that it could detect changes as small as 0.02%, allowing them to clearly distinguish between different magnetic domains.

Mapping the Switch

The team didn't just take a picture; they watched the dance change. They cooled the device down to 200 Kelvin (a very chilly -73°C) to reset the magnetic state, then warmed it up to 260 Kelvin (-13°C) to let the octupole order return. As they did this, they applied an external magnetic field to see if they could force the dancers to switch partners.

They found that as they swept the magnetic field back and forth, the XMCD contrast flipped. When the field went one way, the "bright" spots in their image became "dark," and vice versa. This proved that they were successfully switching the octupole domains. To confirm this wasn't just a visual trick, they measured the "Anomalous Hall Effect" (a voltage that appears when electricity flows through a magnetic material) in the exact same device. The voltage jumped and dropped in perfect sync with their X-ray images. This double-check confirmed that the contrast they saw was real and came from the bulk of the material, not just the surface.

What They Found (and What They Didn't)

The researchers successfully mapped the "octupole domains" in Mn3Sn, showing that they could switch these domains using a magnetic field. They observed that the switching wasn't 100% perfect; the magnetic field they could apply (up to ±250 mT) wasn't quite strong enough to force every single domain to flip completely, leaving some mixed zones where the dance was still confused. They also noted that at lower temperatures (220 K), the material entered a different "spiral" phase where this switching effect disappeared, leaving only a weak, linear response.

Crucially, they ruled out the idea that this effect was caused by simple ferromagnetism. If the material were acting like a normal magnet, the X-ray signal would have been negative at this energy level. Instead, they saw a positive signal, which matches the theoretical prediction for an octupole order driven by a specific term called $T_z$ (an anisotropic spin density). This confirmed that they were indeed imaging the complex, hidden octupole structure, not just a simple magnetic field.

Why This Matters

This paper doesn't claim to have built a new computer yet, but it has handed scientists a powerful new camera lens. By proving that XMCD-STXM can see these hidden octupole domains with high sensitivity, the researchers have opened the door to studying how these materials switch, how they move, and how they behave in real-time. This is a vital step toward using these "invisible" magnets for the ultra-fast, ultra-secure spintronic devices of the future. The team showed that even when the magnets cancel each other out, if you look at them with the right kind of spinning light, you can see the secret patterns that make them so special.

Technical Summary: X-ray Imaging of Antiferromagnetic Octupole Domains in Mn3Sn

Problem Statement
Antiferromagnetic (AFM) systems offer significant advantages for spintronics, including robustness against external magnetic fields, the absence of stray fields, and ultrafast spin dynamics. However, their compensated spin structures make them notoriously difficult to image and control in real space. While recent research has identified a class of unconventional antiferromagnets that break time-reversal symmetry (TRS)—such as non-collinear Mn3Sn and collinear altermagnets—enabling physical responses typically associated with ferromagnets (e.g., anomalous Hall effect, magneto-optical Kerr effect, and X-ray magnetic circular dichroism), direct real-space imaging of their domain structures remains a challenge. Previous imaging of Mn3Sn domains has relied on surface-sensitive techniques like MOKE or nitrogen vacancy microscopy. There is a need for a bulk-sensitive method capable of resolving the specific multipolar order parameters, such as magnetic octupoles, that drive these TRS-breaking phenomena.

Methodology
The authors fabricated a Hall bar device from a single crystal of Mn3Sn (Type A, exhibiting a helical modulation below ~260 K) using a focused ion beam (FIB). The device was thinned to ~250 nm and capped with Al2O3 to prevent oxidation. Electrical contacts were established via in situ Pt deposition.

The study employed Scanning Transmission X-ray Microscopy (STXM) at the MAXYMUS beamline (BESSY II) to image the antiferromagnetic domains. Key experimental parameters included:

Energy Selection: Measurements were performed at the Mn $L_3$ pre-edge (nominal 639.25 eV), where the X-ray magnetic circular dichroism (XMCD) signal is dominated by the intra-atomic magnetic-dipole term ( $T_z$ ) rather than the spin moment ( $S_z$ ), which is negligible in these systems.
Imaging Protocol: The sample was cooled to 200 K (spiral AFM phase) and warmed to 260 K (non-collinear AFM phase) to initialize a multidomain state. Images were acquired using left- and right-circularly polarized light at 100 nm pixel resolution.
Data Processing: XMCD contrast was derived by subtracting the logarithms of normalized images acquired with opposite helicities.
Correlative Measurements: The same device was subjected to magnetic field sweeps (up to ±250 mT) while simultaneously measuring XMCD contrast and the anomalous Hall resistivity ( $\rho_{yx}$ ).

Key Results

Real-Space Imaging of Octupole Domains: The authors successfully resolved antiferromagnetic octupole domains in Mn3Sn. The images revealed distinct bright and dark contrast regions corresponding to spin clusters oriented parallel or antiparallel to the X-ray beam.
High Sensitivity and Contrast: The technique achieved a maximum XMCD contrast of approximately 0.2% at the ideal energy, with a noise floor below 0.02%. This high sensitivity allowed for the robust identification of weak octupole domain contrast.
Spectral Confirmation: Line scans across domain boundaries confirmed contrast reversals. The spectral dependence of the XMCD signal showed a positive sign at the $L_3$ edge, consistent with the theoretical prediction for octupole order (as opposed to the negative contrast typical of ferromagnetic Mn). The data aligned quantitatively with reference spectra from previous studies.
Hysteretic Switching: Under an applied magnetic field, the XMCD contrast exhibited a clear hysteresis loop at 260 K, corresponding to the switching of octupole domains. This switching behavior closely tracked the hysteretic response of the anomalous Hall effect measured in the same device.
Phase Dependence: At 220 K (within the antiferromagnetic spiral phase), the XMCD contrast and Hall effect showed only a weak linear response, confirming that the strong hysteretic signal is specific to the non-collinear octupole-ordered phase.

Significance and Claims
The paper establishes XMCD-based STXM as a powerful real-space imaging method for TRS-breaking antiferromagnets. The authors claim that this technique successfully overcomes the difficulty of imaging compensated magnetic structures by exploiting the finite XMCD contrast arising from the $T_z$ term in magnetic space groups lacking specific symmetries.

The study demonstrates that XMCD-STXM can directly probe bulk multipolar spin textures (specifically magnetic octupoles) and correlate them with macroscopic transport properties (anomalous Hall effect) within the same device. The authors position this method as superior to surface-limited optical techniques and electron microscopy for these materials, noting its element specificity, depth sensitivity, and compatibility with buried or encapsulated devices. Furthermore, the authors suggest that this approach enables future studies of the dynamics, switching, and symmetry-tunable phenomena in TRS-breaking antiferromagnets, including the emerging class of altermagnets.

More like this