Magnon polarons unveiled in a kagome antiferromagnet

Magnons and phonons – quanta of spin waves and lattice vibrations – are fundamental excitations in solids. Magnons arise when the ordered alignment of atomic magnetic moments is disturbed, producing waves of spin fluctuations. Meanwhile, phonons describe the collective vibrations of the atoms themselves. In most materials, these two types of excitations exist separately, but when they interact strongly they merge into new hybrid states called magnon polarons. These coupled modes are not just a curiosity – they can give rise to new physical effects such as unusual heat transport, spin currents carried by atomic vibrations, or even exotic chiral phonons. As such, they are of great interest for future spintronic and magnonic devices, where information might be carried and processed using spin waves instead of electric charge.

Despite this promise, it has been difficult to find a clean model system in which spin–lattice coupling can be unambiguously identified and studied in detail. In many magnets, the crystal lattice is rather complex, with numerous atoms per unit cell producing a dense spectrum of phonon modes. This makes it hard to disentangle which vibrations are actually interacting with magnetic excitations and to what extent. What has long been missing is a simple, well-understood material where the coupling is both strong and clearly observable.

In this study, such a model system has been identified in the kagome antiferromagnet Mn₃Ge (Figure 1). The kagome lattice is a two-dimensional network of corner-sharing triangles, and it is known to trap spins in a frustrated geometry in which they cannot all align simultaneously. This frustration leads to non-collinear magnetic order, which in turn enhances the possibility of coupling between spin and lattice degrees of freedom. The key feature of Mn₃Ge is its relatively simple structure: the lattice is dominated by magnetic manganese atoms, with only minor contributions from germanium. This clarity makes it possible to isolate and interpret the hybridised modes without ambiguity.

 

Click figure to enlarge

Fig. 1: The crystal and magnetic structures of Mn₃Ge. The unit cell contains two layers of Mn atoms that form adjacent kagome planes. The nearest in-plane (J₁, left) and out-of-plane (J₂, right) magnetic exchange interactions are shown as bonds between the atoms. 

Inelastic X-ray scattering (IXS) at ID28 was performed on Mn₃Ge with high energy and momentum resolution as well as high intensity, which is critical for resolving complex spectra. Its sensitivity to only the lattice part of the magnon polarons becomes its main advantage in comparison with neutron scattering, the only alternative technique. Indeed, as the latter probes both the lattice and the spin, the hybrid excitations are difficult to distinguish from a simple superposition of uncoupled phonon and magnon signals.

The ID28 experiments revealed a striking difference between high and low temperatures. At 500 K, well above the material’s Néel temperature of 370 K, the IXS spectra displayed only the expected bare phonon branches, in excellent agreement with ab initio lattice-dynamics calculations in the non-magnetic state. However, when cooled well below its magnetic ordering temperature, the spectra changed drastically. Instead of a simple overlap in momentum–energy space, certain magnon and phonon modes avoided one another, creating a distinct hybridisation gap of about 2 meV. This “avoided crossing” is the hallmark of strong coupling and provides direct evidence that magnons and phonons in Mn₃Ge are no longer independent but form mixed excitations.

To understand the microscopic origin of this effect, a minimal theoretical model was developed. This showed that the coupling is driven by interlayer exchange interactions (J₂) – strong internal magnetic fields binding neighbouring kagome planes – that are particularly sensitive to lattice distortions of E_2g symmetry. These vibrations, involving collective displacements of manganese atoms within the kagome planes, modulate the exchange interactions and resonate with spin excitations. The outcome is a set of hybrid modes with both magnetic and vibrational character, which was confirmed in excellent agreement between experiment and simulation.

The IXS spectrum at 500 K reveals two weakly dispersive optical phonons in full agreement with ab initio calculations of lattice dynamics without magnetism. At a low temperature of 80 K, the spectrum consists of three magnon–polaron branches that are perfectly captured within the theoretical model (Figure 2).
 

Click figure to enlarge

Fig. 2: A comparison between theory (left column) and the experimental data (right column). The pure lattice-dynamics spectrum above the magnetic ordering temperature (500 K, top row) and the hybrid magnon-phonon excitations below (80 K, bottom row).
 

The significance of this discovery is twofold. From a fundamental standpoint, it provides the long-sought example of a “clean” system where magnon-phonon hybridisation can be scrutinised. This bridges the gap between theoretical predictions and experiment, offering a platform to explore new phenomena. From a technological perspective, it demonstrates that strong magnetoelastic interactions can be controlled in real materials, in which spin and lattice dynamics can be deliberately tuned. Such control could enable new forms of spintronic devices that exploit vibrations to manipulate information. Additionally, the physical mechanisms of Mn₃Ge are likely to apply to a broader class of the kagome magnets, which attract a lot of attention due to many other intriguing phenomena. 

Principal publication
Strong Magnon-Phonon Coupling in the Kagome Antiferromagnets, A.S. Sukhanov et al., Phys. Rev. Lett. 135, 086703 (2025); https://doi.org/10.1103/gymx-jk1g