Claude Opus 5.5 Agents Identify Two Room-Temperature Magnetic Semiconductor Candidates

A research team using Claude Opus 5.5 agents has identified two candidate materials for next-generation spintronic memory: a newly designed compound, YBaMnFeO₅, and an existing material, KV[Cr(CN)₆], first synthesized in 1999. Both are predicted to be Luttinger-compensated (LC) semiconductors, a class of materials that combines the zero net magnetism of antiferromagnets with the spin-sorting capabilities of ferromagnets, enabling high-density, low-interference data storage.

Candidate 1: YBaMnFeO₅

The agents designed YBaMnFeO₅, a new compound composed of yttrium, barium, manganese, iron, and oxygen. Simulations using density functional theory (PBE+U and HSE06) predict the following properties:

• Band Gap: 2.35 eV, with spin sorting occurring at both sides of the gap.
• Spin Windows: 1.0 eV for holes and 1.4 eV for electrons, significantly larger than the 26 meV thermal fluctuation at room temperature.
• Magnetic Ordering: Predicted to retain magnetism up to 420 K (raw simulation) or 490 K (calibrated).

However, the design requires manganese and iron atoms to arrange in a perfect checkerboard pattern. Simulations indicate this order collapses into a random mix at 950 K, while standard synthesis occurs at 900–1300 °C, making the useful ordered form difficult to produce.

Candidate 2: KV[Cr(CN)₆]

The agents identified KV[Cr(CN)₆], a Prussian blue-family material first made in 1999, as a viable LC semiconductor. While its zero net magnetism was known, its potential as a spin-sorted semiconductor had not been previously recognized or quantified.

• Band Gap: Predicted at 2.1 eV.
• Spin Windows: 2.6 eV for holes and 1.6 eV for electrons.
• Magnetic Ordering: The 1999 sample maintained magnetic order up to 376 K (103 °C), above room temperature.
• Structure: The crystal structure naturally locks chromium and vanadium into distinct sites, avoiding the synthesis challenges faced by YBaMnFeO₅.

Current limitations for KV[Cr(CN)₆] include the fact that existing samples are hydrated powders with a small residual magnetic moment (0.125 Bohr magnetons per formula unit), whereas the perfect crystal is predicted to have zero. Neither the band gap nor the spin sorting has been experimentally measured yet.

Implications

The findings suggest that room-temperature Luttinger-compensated semiconductors may already exist in known chemical libraries. The team has published all calculations, code, and caveats on GitHub to facilitate independent verification and further research into practical spin-based technologies.