Abstract
As artificial intelligence (AI) systems continue to scale, they will demand storage architectures that combine high capacity, high speed, and energy efficiency, as well as new computing paradigms that can overcome the bottlenecks of conventional computing architectures. This has motivated the search for new non-volatile memory technologies and alternative computing approaches such as neuromorphic computing. Quantum materials, including magnetic and multiferroic materials, are particularly promising in this context because the presence of multiple order parameters and associated ultrafast dynamics offer rich functionalities for high-density information storage and ultrafast signal generation.
In this talk, I will first discuss a proof-of-principle demonstration of out-of-plane damping-like torque generation using canted ferromagnets. This approach produces an efficient out-of-plane component of the spin-orbit torque, enables current-induced switching of perpendicular magnetic devices at zero applied magnetic field and suggests a viable route toward thermally stable, high-density, and energy-efficient magnetic memory devices. I will then present a new scheme for electrically controlling spin transport in magnetic media over long distances by using ferroelectric polarization to modulate antiferromagnetic magnons in a multiferroic material. The resulting magnon flow can exert spin-orbit torques on an adjacent ferromagnet, establishing a pathway toward non-volatile and electrically tunable spin transport. Together, these studies highlight the potential of magnetic and multiferroic quantum materials as a versatile platform for next-generation memory, computing, and signal-generation technologies.