Discovery of new type of magnetic materials

Origin. In 1964, Emmanuel Rashba and Solomon Pekar conjectured that a magnetic mechanism Soviet physics JETP (1964) — analogous to spin–orbit coupling but distinct from it — could produce spin effects useful for electronic and memory devices. The conjecture sat unconfirmed for half a century, and which material systems would realise the effect was unknown. Through 2018 conversations with Prof. Rashba (retired from Harvard), I set out to identify them.

Discovery (2019–2020). With Alex Zunger and Emmanuel Rashba, I derived that a class of antiferromagnets — including the textbook material MnF₂ — exhibits a non-relativistic, momentum-dependent spin splitting that is significantly larger than the conventional Rashba effect and does not depend on spin–orbit coupling (PRB 102, 014422 (2020)). Between 2019 and 2021 several other groups independently identified related effects in different materials; the class was subsequently named altermagnets.

Symmetry framework. I built the full symmetry and band-theory framework that classifies every crystal into seven spin-splitting types (SST-1 through SST-7), where non-relativistic spin-split antiferromagnetic order belongs to SST-4 (PR Materials 5, 014409 (2021)). Of the 1,651 magnetic space groups, 422 permit a non-relativistic spin polarisation. Inverse-design screening against MAGNDATA identified 203 candidate compounds, several of which I validated by first-principles calculation. I also introduced the concept of magnetic structural building units as a chemical design rule (Adv. Mater. 2023).

A second class. I subsequently identified a distinct family of compensated magnets (subset of SST-4) that is not altermagnet but hosts non-relativistic spin splitting at the Brillouin-zone centre — accessible in half-metals and insulators, and realisable through alloy design (PRL 2024).

What it could enable. These materials are a platform for terahertz spintronics, all-antiferromagnetic tunnel junctions (memory beyond Moore’s law), and unconventional superconducting pairing on a magnetically active substrate.