English

Symmetry-Guided Computational Screening of Two-Dimensional Altermagnets with ab initio Hubbard Corrections

Materials Science 2026-08-03 v1

Abstract

Altermagnets combine compensated antiferromagnetic order with momentum-dependent spin splitting, offering a promising platform for spintronic applications without macroscopic magnetization or stray magnetic fields. Although a wide range of three-dimensional (3D) materials have been identified as altermagnets, two-dimensional (2D) altermagnets remain comparatively limited. In this work, we perform a high-throughput computational search for altermagnetism across 2710 materials in the Materials Cloud 2D Crystals (MC2D) database. Our approach combines symmetry-based screening with first-principles density functional theory calculations, including self-consistent Hubbard-UU corrections, to reliably capture magnetic ground states. Through a systematic exploration of magnetic configurations and their energetic stability, we identify 42 materials exhibiting altermagnetic ground states for at least one value of UU, of which 24 remain robust upon determination of the Hubbard-UU parameters from first principles--including 4 materials previously reported in the literature and 20 newly predicted candidates. These comprise promising monolayers such as metallic Fe2_2Si2_2SbO9_9, and insulating CoBrO, with spin splittings about 294 meV and 330 meV, respectively. Our results significantly expand the pool of potential 2D altermagnet candidates with favorable exfoliation energetics and provide valuable guidance for experimental efforts. In addition, this work establishes a high-throughput computational framework for reproducible discovery and characterization of altermagnetic materials.

Keywords

Cite

@article{arxiv.2608.02925,
  title  = {Symmetry-Guided Computational Screening of Two-Dimensional Altermagnets with ab initio Hubbard Corrections},
  author = {Anumita Bose and Nataliia Manko and Marco Gibertini and Antimo Marrazzo},
  journal= {arXiv preprint arXiv:2608.02925},
  year   = {2026}
}

Comments

12 pages, 5 figures