Nitrospinics as a platform from orbital-torque memory to artificial intelligence

Abstract

The exploration of energy-efficient and functional spintronics has attracted considerable attention. Orbital transport has opened new pathways for current-induced torque generation beyond the conventional spin transport based on the spin Hall effect. In addition, artificial intelligence computing has been demonstrated using spintronic devices. Further progress of these devices can be anticipated through the development of unique and functional materials beyond the existing heavy metals and topological systems with strong spin-orbit coupling. The nitride materials exhibit unique chemical, magnetic, and structural versatility, including antiferromagnetism, high thermal stability, and compatibility with diverse device architectures. Here, we propose Nitrospinics as a conceptual and functional framework that exploits nitride materials for applications ranging from orbital-torque-based spintronic devices to artificial intelligence hardware. Using Cr2N, a two-dimensional nitride MXene with an atomic layered structure, as a prototype system, we discuss how nitrogen contributes to the structural stability, the orbital torque generation, and the interfacial orbital and spin conversion. We further outline key challenges and opportunities toward establishing nitride-based materials for next-generation computing technologies.