Altermagnetism: a third kind of magnetic order

What it is

Altermagnetism is a form of magnetic order in which the tiny magnetic moments on neighbouring atoms point in opposite directions and cancel, so the material has no overall magnetisation. Even so, electrons moving through it behave differently depending on their spin, a property usually associated with ordinary magnets.

The explanation lies in the geometry of the crystal. In an altermagnet, the atoms whose moments point one way are surrounded by a version of their environment that is rotated or mirrored compared with that of the atoms pointing the other way, rather than an identical copy shifted along the lattice. Because the two sets of sites cannot be swapped by a simple shift or by inversion, electrons travelling in some directions favour one spin, while those travelling in other directions favour the opposite spin. Physicists describe the resulting pattern of spin splitting, which alternates in sign around momentum space, by analogy with the shapes of atomic orbitals, calling it d-wave, g-wave or i-wave. Crucially, this splitting arises from the magnetic order and the crystal structure themselves rather than from spin-orbit coupling, a weaker relativistic effect, so in principle it can be sizeable.

The same symmetry also shapes spin waves, the collective ripples that travel through magnetic order. A peer-reviewed study using polarised X-ray scattering reports spin waves with a definite handedness in manganese telluride, a candidate altermagnet, with a spin polarisation that varies with momentum in line with the material's symmetry[1].

How it differs from ferromagnets and antiferromagnets

In a ferromagnet such as iron, the moments line up, which gives a net magnetisation and electronic bands split by spin. As a result, ferromagnets are easy to sense and steer with magnetic fields, and they underpin hard drives and magnetic memory. Their drawbacks are stray fields, which let neighbouring bits interfere with one another, and switching dynamics that are comparatively slow.

In a conventional antiferromagnet, the moments alternate and cancel, and the sites with opposite spin are related by a shift or by inversion. Together with time-reversal symmetry, that relationship forces the energy bands of the two spins to coincide, so there is no spin splitting. Antiferromagnets produce no stray fields and their magnetic dynamics are much faster, but they are hard to read out electrically.

Altermagnets combine features of both: zero net magnetisation like an antiferromagnet, and spin-split bands like a ferromagnet. Several materials long catalogued as antiferromagnets turn out to fit this description. Using a framework called spin-group symmetry, theorists argue that altermagnetism is a distinct third class of collinear magnetism rather than a subtype. That combination explains the appeal of altermagnetic spintronics, which would carry information in electron spin without relying on a net magnet.

Where it is stuck

Identifying genuine altermagnets is harder than it sounds. Confirming the predicted spin splitting requires demanding measurements such as spin-resolved photoemission. Some early candidates have been questioned after closer study suggested that their magnetic order was weaker than assumed, or absent altogether. Separating altermagnetic signals from those caused by spin-orbit coupling or impurities also takes care.

Control is another hurdle. With no net magnetisation, external magnetic fields gain little purchase, yet the material can still form domains in which the pattern of spin splitting is reversed. Reliably setting, switching and reading those domains, ideally with electric currents at room temperature, remains an open problem. So does finding materials that combine robust magnetic order at room temperature, suitable conductivity and compatibility with chip fabrication.

Finally, it is not yet clear how large the useful effects will be in working devices, or whether the theoretical advantages over ferromagnets and antiferromagnets will hold up against real-world disorder and manufacturing. The field is young, and many of its predictions still await experimental tests.