Altermagnets show no net magnetization, yet they generate spin current as efficiently as ferromagnets, as demonstrated by a new study published in Nature by a team led by Helena Reichlová at the Institute of Physics (FZU) of the Czech Academy of Sciences in collaboration with Fèlix Casanova of CIC nanoGUNE in Spain. The result opens the way to smaller and more robust spintronic devices.
At the FZU, Helena Reichlová leads the Max Planck Dioscuri Centre research group, which, among other things, studies a compound of manganese and silicon. The material has proven itself in the lab because it can be measured at readily accessible temperatures, is non-toxic and forms high-quality thin films. Reichlová has been studying it since 2019, and together with her colleagues she has already used it to demonstrate several important phenomena.
What was still missing was a result with direct relevance to spintronics. “For a long time, I wanted to demonstrate a spin valve on this material as well, a typical spintronic device we know from ferromagnets,” says Reichlová.
A spin valve is a device in which one magnetic element emits a spin current, and another detects it. It usually consists of layers stacked on top of each other. This is not possible with the manganese-silicon compound, however, because the spin current cannot be sent perpendicular to the thin-film interface into the next layer. Reichlová and her colleagues therefore chose lateral spin valves, in which the spin current travels within the plane of the thin film. Another advantage of these lateral devices is that the spin current can be separated from the charge current.
Fabricating these devices is complex, so in 2022 Reichlová’s team began collaborating with Fèlix Casanova of the Spanish research centre CIC nanoGUNE, whose group has long worked with them. Four years later, this collaboration led to a publication in Nature .
“Lateral spin valves are complex devices, and I didn’t want to start developing them on my own, especially with completely new materials. It made more sense to team up with someone who has experience with them,” Reichlová explains.
Demanding production
Fine-tuning the fabrication of the devices alone took about a year. The key part of the work was done by the study’s first author, doctoral student Jone Mencos, who initially struggled with the material losing its magnetic properties in nanostructures. She eventually brought the production to a working state, after which the team spent another year figuring out what the measured signals looked like and what exactly they revealed. Theorists then joined the research and suggested which measurements still needed to be added. “For roughly two years, we weren’t sure whether the research would lead to a result,” says Reichlová.
Success was far from certain, as no one had carried out a similar measurement on this type of material, and the experiment relied on an unconventional geometry. While in ferromagnets spin is injected across the entire contact area, in the altermagnet this happens only over a small section and along specific crystal directions. It was therefore unclear whether the resulting weak signal could be detected at all. “I was surprised that it works at all. But research rarely brings a sudden moment of discovery. The measured signal is only the beginning, and only careful analysis shows what it really means,” Reichlová emphasizes.
Most of the laboratory work was done by students, with the main share falling to Jone Mencos, who also spent three months at the FZU. The project drew on the knowledge of the material gained by FZU student Antonín Baďura, who also contributed to sample characterisation. Mencos, Reichlová, and Casanova discussed the data regularly, mostly online. According to Reichlová, it proved to be an advantage that each of them has slightly different knowledge and experience and can bring their own perspective to interpreting the results.
Two types of magnets in one
Besides its electric charge, an electron also carries spin, which makes it behave like a tiny magnet. Conventional electronics uses only charge, whereas spintronics also works with spin and promises faster and more energy-efficient devices. To do so, it needs a spin current, a flow of electrons in which one spin orientation prevails. Today, spin currents are generated by ferromagnets, but their magnetic field affects their surroundings. As devices shrink and are packed densely together, they begin to interfere with one another. Antiferromagnets do not have this problem because their magnetic moments cancel each other out, but they cannot generate a spin current.
Altermagnets, described only in recent years, combine the advantages of both groups. They have zero net magnetization, yet thanks to the symmetry of their crystal, electrons in them split by spin almost as strongly as in ferromagnets. Until now, however, no experiment had clearly shown that a spin current can actually be extracted from them.
In their Nature study, the team led by Reichlová and Casanova describes how they became the first to electrically inject a pure spin current from a magnet with zero net magnetization into a non-magnetic material. They also provided unambiguous proof of the so-called spin-splitter effect, a mechanism characteristic of altermagnets in which a spin current arises perpendicular to the flow of charge.
The next stage of research
According to Reichlová, practical applications of the result are still difficult to predict. They will depend on materials development, engineering solutions, and the willingness of technology companies to invest in new architectures. “Our work is fundamental research. Above all, we try to understand how these materials work, and their specific use is the next step,” she says.
The collaboration with Casanova’s group continues. The team plans to study other materials, but the manganese-silicon compound itself still raises many open questions, for instance, how the signal depends on direction or whether the devices can be made even smaller. For now, a ferromagnet serves as the spin current detector in the devices, and the researchers would like to find out whether they can do without it entirely.
Nature
Experimental study
Not applicable
Time-reversal-symmetry-breaking spin injection from a compensated magnet
28-Oct-2026
There is no COI related to this release.