🔬 materials science

Twist-tuned exchange and hysteresis in a bilayer van der Waals magnet

This study demonstrates that twisting bilayer CrSBr by approximately 3° modulates interlayer exchange interactions to induce robust magnetic hysteresis and a coherent antiferromagnetic switching response, establishing twist engineering as a viable route for programmable two-dimensional magnetic memories.

Original authors: Priyanka Mondal, Sonu Verma, Wenze Lan, Lukas Krelle, Ryan Tan, Regine von Klitzing, Kenji Watanabe, Takashi Taniguchi, Kseniia Mosina, Zdenek Sofer, Akashdeep Kamra, Bernhard Urbaszek

Published 2026-07-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Priyanka Mondal, Sonu Verma, Wenze Lan, Lukas Krelle, Ryan Tan, Regine von Klitzing, Kenji Watanabe, Takashi Taniguchi, Kseniia Mosina, Zdenek Sofer, Akashdeep Kamra, Bernhard Urbaszek

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 two sheets of a magical, magnetic material called Chromium Sulfur Bromide (CrSBr). Normally, if you stack these two sheets perfectly on top of each other, they act like a well-behaved team: the top sheet wants its tiny magnetic arrows pointing one way, and the bottom sheet wants them pointing the exact opposite way. They hold hands tightly in this "anti-parallel" dance, and if you try to force them to point the same way using a magnet, they switch instantly and smoothly. There's no hesitation, no memory of where they were before. It's like a light switch that clicks on and off with zero delay.

But here is where the scientists, Priyanka Mondal and her team, got creative. They took two of these sheets and twisted them slightly against each other—just by about 3°.

Think of it like placing two identical honeycombs on top of each other but turning one slightly. This creates a giant, wavy pattern called a "moiré superlattice." In this twisted setup, the way the two sheets talk to each other changes depending on exactly where you look. Some spots are tight, some are loose. The team found that this twist acts like a volume knob for the magnetic connection between the layers.

The Big Discovery: Magnetic Memory
When they applied a magnetic field to this twisted pair, something amazing happened that didn't happen in the perfect, untwisted version. The material started to show hysteresis.

In everyday terms, hysteresis is like a stubborn memory. Imagine you are trying to push a heavy door. You have to push hard to get it to open, but once it's open, it stays open even if you stop pushing. To close it, you have to push back hard in the other direction. The door doesn't just snap back the moment you let go; it "remembers" being open.

In the twisted CrSBr, the magnetic arrows stayed stubbornly pointing in the same direction (parallel) even after the scientists turned down the magnetic field to nearly zero. They had to reverse the field completely to force them back to their original opposite directions (antiparallel). This created a robust "loop" in the data, a signature of a material that can remember its state.

How They Knew
The team didn't just guess this was happening; they watched it happen using light. CrSBr is special because its "excitons" (tiny particles made of an electron and a hole) change their energy color based on how the magnetic layers are arranged. By shining a laser and measuring the light coming back, they could see the exciton energy shift.

The Perfect Stack: When they tested the untwisted sample, the exciton energy shifted smoothly from 1.332 eV to 1.320 eV as they increased the magnetic field. It switched back and forth instantly with almost no lag (a tiny hysteresis of just 0.18 T).
The Twisted Stack: In the twisted sample, the story changed. As they swept the magnetic field, the exciton energy shifted from 1.336 eV to 1.320 eV, but it got stuck! It stayed in the "parallel" state even when the field dropped to 3 mT (that's 0.003 Tesla). It didn't switch back until the field went all the way to 0 mT in the reverse direction. This created a wide, sturdy loop.

The "Monodomain" Surprise
Usually, when scientists twist magnetic materials, they expect to see a messy patchwork of different magnetic domains (like a quilt of different patterns) or complex swirling textures. The paper explicitly argues against this. Instead, the twisted CrSBr behaved like a single, giant, coherent unit. The whole sheet switched together, like a single monolith, rather than breaking into fragments. The twist didn't create chaos; it created a new, stable, and tunable "monodomain" state.

Not Just a Theory
The team built a mathematical model to explain this. They treated the two layers as two giant magnets (a "two-sublattice model"). Their math showed that the twist effectively lowered the strength of the connection between the layers, just enough to let both the "opposite" and "same" directions be stable for a while. This matched their experimental numbers perfectly, except for one tiny detail: the model predicted the switch would happen at a slightly different field strength than what they actually measured. The authors admit they don't fully understand that one small mismatch yet, but the rest of the picture is clear.

Why It Matters
This isn't just a cool trick with magnets. The team suggests that by twisting these 2D magnets, we can program their magnetic memory. You could potentially design a computer memory or a logic gate that works by simply twisting the layers to the right angle, rather than using electricity to write data. It's a new way to engineer materials, proving that a tiny twist of 3° can turn a forgetful magnet into a stubborn, memory-holding one.

They tested this at a chilly 4.7 K (that's very cold, just a few degrees above absolute zero) and found that this memory effect only happens when the magnetic field is applied along a specific direction (the b-axis). If they tried to push the magnet from the side (along the a-axis or c-axis), the memory vanished, and the material acted normally again.

In short, by twisting two sheets of CrSBr, the scientists found a way to make the magnetic layers "hold their breath" in a new state, creating a robust, tunable magnetic memory that didn't exist before.

Problem Statement
Twist engineering has proven effective in reconstructing electronic band structures in 2D quantum materials like graphene and WSe₂, leading to correlated states such as superconductivity and insulating phases. However, extending this paradigm to van der Waals (vdW) magnets to control magnetic order remains an open challenge. While twisting bilayer magnets can create spatially varying interlayer exchange interactions that stabilize emergent spin textures or coexisting domains, there is a need to understand how twist angles influence magnetic hysteresis and whether they can enable robust, programmable magnetic switching in a monodomain regime. Specifically, the behavior of bilayer Chromium Sulfur Bromide (CrSBr)—a material with strong intralayer ferromagnetic exchange and antiferromagnetic (AFM) interlayer coupling—under twist engineering requires investigation.

Methodology
The study employs a combination of experimental characterization and theoretical modeling:

Sample Fabrication: High-quality bulk CrSBr crystals were grown via chemical vapor transport. Monolayer flakes were mechanically exfoliated and assembled into twisted bilayers using a dry transfer technique with a polydimethylsiloxane (PDMS) stamp. A precise rotational misalignment of approximately 3° was introduced between two monolayer CrSBr flakes, sandwiched between hexagonal boron nitride (hBN) layers.
Experimental Characterization: Magneto-optical properties were probed using low-temperature (4.7 K) polarization-resolved photoluminescence (PL) spectroscopy in a confocal microscope. Magnetic fields were applied along the three principal crystallographic axes (a, b, and c) using a vector magnet. The exciton energy shifts were tracked as a function of the magnetic field to infer magnetic configurations. Spatially resolved measurements were conducted at different locations across the twisted sample to map local variations.
Theoretical Modeling: A two-sublattice model was developed to describe the system. Given the strong intralayer ferromagnetic exchange relative to interlayer exchange, the complex position-dependent exchange arising from the moiré superlattice was approximated as a spatially averaged effective interlayer exchange ( $J^*_{inter}$ ). The magnetic free energy was minimized to determine ground states and switching fields for different field orientations.

Key Results

Emergence of Hysteresis: Pristine (untwisted) bilayer CrSBr exhibits reversible, symmetric magnetic switching with negligible hysteresis. In contrast, the twisted bilayer (∼3°) displays robust magnetic hysteresis when the magnetic field is applied along the magnetic easy axis (b-axis).
Exciton Energy as a Probe: The hysteresis is observed as a hysteretic evolution of the A-exciton energy. In the twisted sample, the exciton energy redshifts from the AFM state (1.336 eV) to the ferromagnetic (FM) state (1.320 eV) with a switching field of just 3 mT during the forward sweep, while the reverse sweep switches at 0 mT, creating a wide hysteresis loop.
Directional Dependence: Hysteresis is specific to the magnetic easy axis (b-axis). Fields applied along the intermediate (a-axis) or hard (c-axis) directions result in switching without hysteresis in both pristine and twisted samples.
Spatial Variability: Spatially resolved PL measurements reveal that the width of the hysteresis loop and the critical switching fields vary significantly across the sample (e.g., switching fields ranging from ∼3 mT to ∼0.16 T at different positions), consistent with local modulations of the average interlayer exchange due to the moiré pattern.
Monodomain Behavior: Unlike other twisted magnetic systems that form fragmented spin textures or domains, the twisted CrSBr bilayer exhibits coherent averaging across the moiré supercell, resulting in an effective monodomain response where the system switches between global parallel and antiparallel states.

Theoretical Insights
The two-sublattice model successfully captures the experimental behavior by attributing the hysteresis to a twist-induced reduction in the effective interlayer exchange ( $J^*_{inter}$ ). This reduction stabilizes both parallel and antiparallel spin configurations over a broad range of magnetic fields along the easy axis. The model predicts switching fields ( $h_{AP+}$ and $h_{P-}$ ) that align with experimental data, allowing for the quantitative extraction of effective exchange strength and anisotropy parameters. However, the authors note a discrepancy where the theoretical model predicts a canted state in certain ranges that is not fully consistent with all experimental switching behaviors, particularly regarding the transition from the antiparallel state at high fields.

Significance and Claims
The paper establishes twist engineering as a powerful route to achieve programmable magnetic hysteresis in 2D magnets without increasing fabrication complexity (e.g., via doping or strain). The authors claim that this work demonstrates a new regime for twisted-bilayer magnets characterized by coherence over the moiré length scale, distinct from the formation of spin textures. By harnessing the robustness of antiferromagnetic order and the tunability of interlayer exchange, the findings suggest potential applications in spin filters, reconfigurable logic gates, and memory devices based on antiferromagnetic states. The study provides a framework for designing moiré magnets with tailored magnetic responses through static configuration studies.

More like this