Physicists found spin-splitting evidence for a third kind of magnetism

Cobalt atoms wedged between layers of tantalum diselenide produce an electronic signature that physicists at the University of Central Florida read as altermagnetism, the magnetic order now ranked beside ferromagnetism and antiferromagnetism. The compound, Co_1/4TaSe₂, orders magnetically at 178 kelvins, far below room temperature, so it is a laboratory test case and not a device material.

The team led by UCF professor Madhab Neupane, with first author Milo Sprague, published the measurements in Nature Communications on August 20, 2026. UCF followed with its own announcement on September 18.

Co_1/4TaSe₂ and the ARPES measurements

The paper, titled “Observation of Altermagnetic Spin-Splitting in an Intercalated Transition Metal Dichalcogenide,” lists ten authors, among them theorist Igor Mazin and Nirmal Ghimire alongside Neupane and Sprague. Its abstract reports “clear signatures of altermagnetic spin splitting at the Fermi surface,” drawn from angle-resolved photoemission spectroscopy, which maps the energy and momentum of electrons knocked out of a crystal, and from density functional theory calculations. Magnetic measurements put the antiferromagnetic ordering temperature at 178 K.

UCF’s release adds that the group used spin-resolved measurements to tell spin-up from spin-down electrons. Neupane said that seeing the independent pieces of evidence converge gave the group confidence that it had identified a genuine layered altermagnet. An earlier preprint version of the work reports data taken at beamline 5-2 of the Stanford Synchrotron Radiation Lightsource and a momentum splitting of about 0.09 inverse angstroms between opposite-spin Fermi surfaces, visible in one slice of the crystal’s momentum space and absent in another.

Mainz and the Europhysics Prize behind the “third kind” label

The Nature Communications paper calls altermagnetism a novel magnetic phase that combines features of antiferromagnetism and ferromagnetic ordering. UCF’s release is titled “new type of magnetism” and its text says altermagnets combine desirable traits of both older kinds. Neither uses the words “third kind.”

The ranking comes from the theorists who classified the order, and the three physicists named below hold the prize for it. In July 2026, Johannes Gutenberg University Mainz announced that the 2026 Europhysics Prize went to Jairo Sinova of Mainz, Libor Šmejkal of the Max Planck institutes and the Czech Academy of Sciences, and Tomas Jungwirth of the Czech Academy and the University of Nottingham. The award, according to the release, recognizes that altermagnetism is “a third elementary form of collinear magnetism in addition to ferromagnetism and antiferromagnetism.” Sinova said that finding a new magnetic phase hidden for more than a hundred years shows that mature fields can still hold surprises.

The wording matters for the claim of a third kind. The UCF work did not discover the third class; it supplies one more material in which the predicted spin pattern shows up. The Mainz release says the symmetry classification arrived in 2022, that hundreds of candidate materials have been proposed, and that experimental confirmations followed worldwide.

Layered crystal, stray fields and the distance to devices

A ferromagnet has net magnetization, which is why it can store bits but also spills stray magnetic fields onto its neighbors. An ordinary antiferromagnet has no net magnetization, so it leaks no field, but its electrons are hard to read out. Altermagnets cancel net magnetization while still splitting electron bands by spin, so they could, in principle, carry spin currents without the stray fields. Neupane put it in UCF’s release as the ability to generate and detect spin currents without the negative effect of producing stray fields.

Two measurements anchor the case for this particular compound. One is the six-fold alternation of spin polarization across momentum space, which The Brighter Side of News reports swinging from roughly negative 13 percent to positive 13 percent, matching the pattern theory predicts for the crystal’s symmetry. The other is the temperature dependence in the paper’s abstract, where the band structure reconstructs on heating above the ordering temperature, consistent with the magnetic order being switched off. Each is a check that the splitting follows the magnetic order and not an unrelated feature of the sample, though the preprint’s note that the surface-versus-bulk question is unaddressed still applies.

The layered structure is the compound’s selling point. According to Sci.News, earlier altermagnet candidates such as manganese telluride and chromium antimonide are not van der Waals crystals, whereas Co_1/4TaSe₂ can in principle be stacked with other exotic materials.

The preprint carries its own caveat: spin splitting seen in photoemission does not by itself establish altermagnetism, and the interpretation also relies on the magnetic structure fitting the symmetry criteria. At 178 K, no memory chip is close. What the measurements add is a layered material in which the splitting can be studied, with the next steps being cleaner spin-resolved data and tests at device-relevant temperatures.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.