🔬 materials science

Magnon band splitting without altermagnetism in CuF2

This combined computational and experimental study demonstrates that CuF $_2$ , despite possessing altermagnetic symmetries, does not exhibit altermagnetism because its specific monoclinic structure induces interlayer couplings and strong Dzyaloshinskii-Moriya interactions that suppress band splitting and pin the magnetic order, thereby highlighting the critical role of relativistic effects in determining the actual magnetic state of 3d materials.

Original authors: Ioannis Rousochatzakis, Oleg Janson, Alexander A. Tsirlin

Published 2026-10-09

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Ioannis Rousochatzakis, Oleg Janson, Alexander A. Tsirlin

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 solid-state physics, scientists are constantly searching for new ways to control how electricity and magnetism move through materials. One of the most promising recent discoveries is a state of matter called altermagnetism. To understand it, imagine a material where the magnetic spins of its atoms are perfectly balanced, canceling each other out so the material has no overall magnetic pull, yet the internal arrangement of these spins creates a hidden structure. In a normal magnet, this balance usually means the material behaves the same way regardless of direction. But in an altermagnet, this specific internal symmetry causes the energy levels of electrons or magnetic waves to split apart, creating distinct paths for them to travel. This splitting is highly desirable because it could lead to faster, more efficient electronic devices that do not require a net magnetic field to operate.

For a material to be an altermagnet, it must possess a very specific crystal shape that allows this symmetry to exist. Researchers have long suspected that a common mineral called copper fluoride, or CuF2, fits this description. Its atoms are arranged in a structure similar to rutile, a pattern known to support these special magnetic properties. If CuF2 were truly an altermagnet, it would display a unique signature: its magnetic waves, known as magnons, would split into two separate bands that behave differently, much like left-handed and right-handed versions of the same wave. This would be a major confirmation of the theory and a potential breakthrough for future technology.

However, a team of researchers from the United Kingdom and Germany has now shown that CuF2 is not an altermagnet, at least not in the way it naturally occurs. By combining precise computer simulations with careful laboratory measurements, they discovered that the material's actual structure is more distorted than the ideal models predicted. This distortion changes the magnetic interactions between the atoms so fundamentally that the expected splitting of magnetic waves driven by altermagnetic symmetry does not happen. Instead, the researchers found that the material exhibits a different kind of magnetic behavior driven by a subtle relativistic effect, which creates a splitting that looks similar to the altermagnetic kind but has a completely different origin.

The journey to this conclusion began with a close look at the physical shape of the copper fluoride crystals. While the ideal version of this material should have a symmetrical, four-sided structure, the real crystals the team studied were stretched and twisted into a different shape. Using powerful computers to calculate the energy of the atoms in both the ideal and the real shapes, the researchers found that the twisted, distorted version is significantly more stable. Even if they tried to squeeze the material with high pressure to force it into the ideal shape, it would remain distorted. They also tested whether stretching the material on a flat surface could stabilize the ideal shape, but this approach only made the material magnetic in a way that would cancel out the desired effects. Essentially, the natural state of the material prevents it from ever becoming the altermagnet that theory had hoped for.

With the structure confirmed, the team turned their attention to how the atoms talk to each other magnetically. They built a detailed model of the magnetic forces, treating the copper atoms as tiny magnets connected by invisible springs. Their calculations revealed that the strongest connection is between atoms in the same flat layer, forming a grid-like pattern. However, there are also weaker connections linking these layers together. Crucially, these connections between layers are not uniform; they are tilted and angled in a specific way that forces the magnetic order to repeat itself over a larger distance. This larger repeating pattern effectively doubles the size of the magnetic unit cell, a change that destroys the delicate symmetry required for altermagnetism. Because the symmetry is broken, the magnetic waves cannot split in the unique way that defines an altermagnet.

Despite ruling out altermagnetism, the researchers uncovered something equally fascinating: the material is dominated by a surprisingly strong interaction called the Dzyaloshinskii-Moriya interaction. This is a force that arises from the motion of electrons and the way they interact with the atomic nuclei, a phenomenon rooted in the theory of relativity. In most magnetic materials, this force is very weak, but in copper fluoride, it is unusually strong, about thirty percent as powerful as the main magnetic connection between atoms. This strong force causes the magnetic spins to tilt slightly, rather than pointing in perfectly opposite directions. This tilting, combined with the way the layers interact, locks the direction of the magnetic order to a specific axis within the crystal.

The presence of this strong force has a direct impact on the magnetic waves traveling through the material. The researchers calculated that these waves do indeed split into separate bands, a feature that initially looked like the signature of altermagnetism. However, they demonstrated that this splitting is caused entirely by the strong relativistic force tilting the spins, not by the special symmetry of an altermagnet. Furthermore, unlike the waves in a true altermagnet, which would have a distinct "handedness" or chirality, the waves in this material do not possess this property. The splitting is real, but it is a different phenomenon entirely.

To confirm these theoretical findings, the team performed experiments measuring how the material responds to magnetic fields and how it scatters neutrons. The measurements showed no evidence of a weak magnetic pull that would have indicated a different type of ordering. Instead, the data matched their model of a material where the spins are slightly tilted and locked into a specific direction. The experiments also confirmed that the magnetic waves split, but the pattern of this splitting matched the predictions for a system driven by the strong relativistic force, not the altermagnetic symmetry.

The study serves as a crucial reminder that identifying new states of matter requires more than just looking at the shape of a crystal. While symmetry arguments can suggest that a material might be an altermagnet, the actual magnetic interactions and the precise arrangement of atoms can override these predictions. In the case of copper fluoride, the material's natural distortion and its strong relativistic forces conspire to prevent the altermagnetic state from forming. The researchers' work highlights the importance of combining computer simulations with real-world measurements to understand the true nature of magnetic materials. It shows that even when a material does not turn out to be the exotic state scientists were hoping for, the detailed investigation can reveal rich and complex physics that deepens our understanding of how magnetism works in the solid world.

Problem Statement
The identification of altermagnetism—a state characterized by the lifting of Kramers degeneracy in fully compensated magnetic structures due to specific structural symmetries—has largely relied on symmetry arguments. While the rutile structure is theoretically predicted to host altermagnetism (e.g., in $MnF_2$ and $FeF_2$ ), experimental confirmation of the resulting chiral magnon band splitting has been ambiguous, often attributed to minute dipolar or magnon-phonon effects rather than altermagnetic exchange. $CuF_2$ , a rutile-like $3d$ difluoride, appears on initial lists of altermagnetic candidates. However, experimental neutron diffraction indicates a magnetic ordering vector of $k=(1/2, 0, 0)$ , implying a doubling of the magnetic unit cell. Theoretical consensus suggests that such a $k \neq 0$ order should restore double degeneracy, effectively suppressing altermagnetic band splitting, yet the material's specific magnetic interactions and the origin of its ground state remained unquantified. The central problem addressed is whether $CuF_2$ exhibits altermagnetism in its real, distorted monoclinic structure, or if its magnetic properties are governed by other mechanisms that mimic altermagnetic signatures.

Methodology
The authors employed a combined computational and experimental approach:

Experimental Characterization: Magnetization measurements were performed on commercial $CuF_2$ powder using a SQUID magnetometer to determine magnetic susceptibility and remnant magnetization. X-ray diffraction (XRD) confirmed the monoclinic crystal structure.
Ab Initio Calculations: Density Functional Theory (DFT) calculations were conducted using the Generalized Gradient Approximation (GGA) and GGA+U methods (FPLO and VASP codes). These were used to analyze the electronic band structure, calculate total energies for hypothetical tetragonal versus experimental monoclinic phases under pressure and strain, and derive exchange coupling parameters.
Wannier Function Analysis: Maximally projected Wannier functions were constructed to map the electronic structure onto an effective tight-binding model, identifying dominant hopping integrals ( $t_1, t'_2, t^+_b, t^+_{bc}$ ).
Spin Model Construction: The hopping integrals were mapped to a spin- $1/2$ Heisenberg model including nearest-neighbor ( $J_1$ ), second-neighbor ( $J'_2$ ), and interlayer couplings ( $J^+_b, J^+_{bc}$ ). The Dzyaloshinskii-Moriya (DM) interaction was extracted from the full bilinear exchange matrix.
Thermodynamic and Dynamic Simulations:
- Quantum Monte Carlo (QMC): Used to simulate magnetic susceptibility for the square-lattice model and 3D extensions.
- Exact Diagonalization (ED): Applied to small lattices ( $4\times4$ ) to account for DM anisotropy in the susceptibility fit.
- Mean-Field Theory: Used to incorporate all exchange terms and DM interactions for a comprehensive fit to experimental susceptibility.
- Linear Spin-Wave Theory (LSWT): Employed to calculate magnon dispersions, dynamical spin structure factors ( $S_\perp$ ), and chiral structure factors ( $C$ ) for both single-layer and 3D models.

Key Contributions and Results

Structural Stability: The study confirms that $CuF_2$ adopts a monoclinic structure at ambient conditions due to the Jahn-Teller distortion of $Cu^{2+}$ ions. Ab initio calculations show that hydrostatic pressure does not invert this distortion to a tetragonal phase. While epitaxial strain on a tetragonal substrate could stabilize the tetragonal phase, it simultaneously induces ferromagnetism via the $d_{3z^2-r^2}$ orbital, precluding a spin-compensated altermagnetic state under practical conditions.
Magnetic Model: The material is identified as a quasi-2D square-lattice antiferromagnet with weakly coupled layers. The dominant in-plane exchange is $J_1 \approx 115$ K. Two synergistic antiferromagnetic interlayer couplings ( $J^+_b$ and $J^+_{bc}$ ) amount to 4% and 8% of $J_1$ , respectively. These couplings are oblique to the square planes, leading to a magnetic unit cell doubling ( $k=(1/2, 0, 0)$ ).
Role of DM Interactions: The study identifies unusually strong Dzyaloshinskii-Moriya interactions, with a ratio $|D|/J_1 \approx 0.3$ . This is significantly larger than in parent cuprates like $La_2CuO_4$ . The DM interaction causes spin canting ( $\approx 14^\circ$ ) and, combined with the "order-by-disorder" effect, pins the Néel vector to the crystallographic $b$ -axis.
Absence of Altermagnetism: The doubling of the magnetic unit cell restores the double degeneracy of the magnon bands, effectively suppressing any altermagnetic splitting. The magnetic ground state is fully compensated with no weak ferromagnetism detected (remnant magnetization $< 1 \times 10^{-5} \mu_B$ /f.u.).
Magnon Band Splitting: Despite the absence of altermagnetism, the calculated spin-wave spectra show a splitting of magnon bands. However, the authors demonstrate that this splitting is driven by the relativistic DM anisotropy, not by altermagnetic exchange. Crucially, unlike the chiral magnons predicted for altermagnets, the magnon bands in $CuF_2$ are non-chiral. The chiral structure factor $C(q, \omega)$ vanishes in the 3D system due to the cancellation of chirality between antiferromagnetically coupled layers.

Significance
The paper concludes that conclusive identification of altermagnetism requires going beyond symmetry arguments to analyze specific magnetic interactions and ground states. In $CuF_2$ , the experimentally observed magnetic order ( $k \neq 0$ ) precludes altermagnetism, even though the parent crystal structure possesses the necessary symmetry. The observed magnon band splitting is a relativistic effect arising from strong DM interactions, which can mimic altermagnetic signatures but lack the chiral nature characteristic of true altermagnets. This work highlights that in $3d$ magnets, even weak spin-orbit coupling can produce significant effects (such as large DM anisotropy and spin canting) that must be distinguished from altermagnetic phenomena. The findings suggest that $CuF_2$ serves as a platform for studying the impact of strong DM anisotropy on quantum spin dynamics in a 2D antiferromagnet, rather than as a candidate for altermagnetic applications.

More like this