Chiral spin textures and dynamics in multi-sublattice magnetic materials

Abstract

Spintronics is a research field that aims to understand and control magnetic spins on the nanoscale and should enable next-generation data storage and logic. A promising approach is to encode bits of information using nanoscale spin textures, such as chiral domain walls or skyrmions that can be translated by currents across racetrack-like wire devices. One technological and scientific challenge is to stabilize small spin textures and to move them efficiently with high velocities, which is critical for dense, fast memory. For the past decade, work has focused on using ferromagnetic heterostructures to host chiral spin textures. However, ferromagnets have fundamental limitations that inhibit further progress: large stray fields limit bit sizes and precessional dynamics limit operating speeds. In this thesis, we examine a broader class of multi-sublattice materials: ferrimagnets.

We show that by using ferrimagnets, the fundamental limits of ferromagnets can be overcome, realizing order-of-magnitude improvements in both size and speed. Using metallic, ferrimagnetic Pt/Gd₄₄Co₅₆/TaOx films with a sizeable Dzyaloshinskii-Moriya interaction (DMI), we realize a current-driven domain wall motion of 1.3 km s⁻¹ near the angular momentum compensation temperature and room-temperature-stable skyrmions with diameters close to 10 nm near the magnetic compensation temperature. For the first time, we show that the DMI is present in ferrimagnetic insulator garnet films and that the DMI necessitates a rare-earth ion in the magnetic insulator. Thickness dependent studies and interface engineering show that the DMI manifests at the ferrimagnetic insulator - substrate oxide interface.

We use a large spin-orbit torque from a Pt overlayer and the DMI to exploit ferrimagnetic dynamics, driving domain walls in low-damping and low-pinning GGG/TmIG/Pt heterostructures at velocities as high as 2.1 km s⁻¹. Moreover, by utilizing the ultra-low damping nature of Bi-YIG and an in-plane field, we can drive domain walls in GSGG/Bi-YIG/Pt at near relativistic velocities exceeding 4.0 km s-1, where the domain wall velocity is no longer limited by a velocity plateau defined by the in-plane field, but the magnon group velocity in Bi-YIG. These results show that multi-sublattice ferrimagnetic films are a promising materials system for next-generation data storage, paving a path forward for the field of spintronics.