Researchers at Louisiana State University and the University of Stuttgart have proposed a mechanism for routing spin information through altermagnets, a recently identified class of materials whose alternating magnetic moments cancel out any net magnetization while still letting electrons of opposite spin behave differently as they move through the crystal.
Illustration of a circular domain wall in an altermagnet. The orange and blue arrows show alternating magnetic moments gradually changing direction across the shaded ring, creating a texture that can steer electrons differently depending on their spin. Image credit: LSU
The study addresses a practical problem for spintronics, which encodes information in electron spin rather than charge: altermagnets can already treat spin-up and spin-down electrons differently without producing a stray magnetic field, but researchers have lacked a way to control where each spin state actually goes. The team's calculations show that gradual rotations in an altermagnet's internal magnetic order, such as those found at a domain wall separating two differently oriented regions, reshape the effective paths available to electrons depending on their spin - with the magnetic texture acting like a landscape of hills and valleys that bends spin-up and spin-down electrons differently, similar to how a lens bends light.
The effect is rooted in quantum geometry: near a domain wall, the spatial variation of an electron's spin state can be described by a quantity called the quantum metric, which combines with the altermagnet's spin-dependent electronic structure to give each spin its own effective geometry. One consequence the authors predict is electronic lensing, in which a domain wall focuses, defocuses, or bends electron trajectories depending on spin and on the texture's orientation relative to the crystal. "If we can engineer the texture, we engineer the landscape," Louisiana State University's Constantin Schrade said. "Then we can determine where spin-up goes and where spin-down goes."
The theory also offers a possible experimental signature for altermagnetism itself, which remains difficult to confirm directly. Around a circular domain wall, the authors predict a quadrupolar, four-lobed pattern of spin polarization for d-wave altermagnetic order and an octupolar, eight-lobed pattern for g-wave order - meaning the shape of the pattern could be used to identify not just altermagnetism itself but which symmetry class of altermagnetic order a material has. To detect it, the authors propose a four-terminal device that drives current through the material along two orthogonal directions while a scanning SQUID maps the resulting magnetic flux; combining the two flux measurements isolates the predicted spin-polarization signal from other contributions. The same texture effects also give rise to an emergent spin-orbit coupling that the authors say arises purely from the magnetic texture rather than relativistic effects.
The work is theoretical, and the authors describe it as a set of concrete predictions for experimentalists to test rather than a demonstrated device, but it points toward using magnetic textures to route spin information without the net magnetization that conventional magnets require.