Developing next-generation computing devices based on spintronics and magnonics requires understanding how spin textures can be tailored in patterned magnetic materials. Lithographically patterned micro/nanostructures can provide insights into the formation of magnetic domains consisting of flux closure states that are potential building blocks of magnetic memory. Complex oxides such as epitaxially grown La0.7Sr0.3MnO3 (LSMO) thin films exhibit strong correlations among spin, orbital, charge, and lattice degrees of freedom. As a soft ferromagnet, LSMO is a promising material for analyzing complex spin textures and controlling domain configurations by varying the micromagnet geometry. Using x-ray photoemission electron microscopy, thermally demagnetized LSMO donuts i.e., circular micromagnets with variable hole diameters, were imaged. The combination of quantitative image analysis and micromagnetic simulations allows probing and explaining the formation of spin textures consisting of either a single vortex or a pair of concentric vortices with opposite chirality in the donut structures. The results presented in this research are a step towards furthering our understanding of magnetic spin texture formation and vortex dynamics for developing spin-based devices.

AbstractGlobal computational power demands are sharply increasing and existing computing architectures are reaching the limits imposed by fundamental physics. Neuromorphic computing is a philosophy of computer design based on the human brain, which is much more energyefficient than manufactured computers. However, current materials are not optimized for emerging neuromorphic devices that require the ability to manipulate multiple electronic or magnetic states. Complex oxide materials exhibit a rich variety of interesting phenomena such as high temperature superconductivity, metal-to-insulator phase transitions, and colossal magnetoresistance. Perovskites are one of the members of the complex oxide family that have been studied for nearly a century and have shown promise in solving materials science problems in applications including data storage and computing, photovoltaic cells, solid oxide fuel cells, and light emitting diodes. Modern fabrication techniques enable synthesis of a wide variety of nanomaterials such as thin films, heterostructures namely bilayers or multilayers, and nanostructures such as quantum dots. Furthermore, there are many elements that can stably form the perovskite structure. Nanostructuring and the presence of interfaces in these systems allows for the functional properties to be tailored for a specific need and sometimes give rise to emergent phenomena, thus opening the door to many possible avenues of research.

Past studies have shown that perovskite oxide thin films readily undergo topotactic phasetransformations where either anion or cation vacancies form and rearrange into an ordered sublattice forming a new structure with distinct properties. Moreover, alternating the stacking order of heterostructures is a technique that has been proven to elicit differing outcomes in systems containing the same constituent layers. This thesis explores annealing-induced topotactic phase transformations of perovskite thin films and bilayers which cause structural and magnetic changes in the materials. There is limited research on the topotactic phase transformations in La0.7Sr0.3MnO3 thin films, and to date, no study has looked specifically at the effect of stacking order of LaCoO3-La0.7Sr0.3CoO3 bilayers regarding the phases produced by topotaxy. In this thesis, manganite thin films and cobaltite bilayers with alternating stacking order were synthesized and annealed under different conditions. The annealing process initiated topotactic phase transformations which were studied using X-ray diffraction, X-ray photoelectron spectroscopy and soft X-ray magnetic spectroscopy. Beginning from the perovskite phase, manganite thin films and cobaltite bilayers underwent structural and magnetic changes into perovskite-related phases including oxygen-deficient perovskite, Grenier, brownmillerite, and Ruddlesden-Popper phases. In the bilayer systems, the stacking order of the constituent layers influenced which phases were formed under different annealing conditions. The bilayers also behaved differently from their corresponding single layer thin films that were annealed under the same conditions, giving rise to phases that were not present in the single layer configurations. Along with the structural phase transformation, the magnetic properties of the material were also modified. Control of magnetic properties is important for application in spintronics—a growing field of study that makes use of the modulation and detection of spin states in a material for data storage and processing. Furthermore, access and control of numerous unique phases in the same material can pave the path for the next-generation computing devices.

Perovskite oxides, with the ABO3 structure where the A cations are typically alkaline earth or rare earth elements and B cations are transition metals, have attracted significant attention due to their potential applications in spintronics, magnetic devices, and neuromorphic computing. These materials exhibit a wide range of functional properties, including ferromagnetism, metal to insulator transitions, and controllable phase transformations, which arise from the intricateinterplay of spin, charge, lattice, and orbital degrees of freedom. Importantly, perovskite oxides demonstrate a remarkable sensitivity to external stimuli such as external magnetic fields, lattice strain, and ion migrations, allowing for precise control over their emerging magnetic and electronic properties. This characteristic distinguishes them from metallic systems, offering unique path for the design and manipulation of functional materials with tailored properties. As a result, perovskite oxides hold great promise for the development of advanced electronic and magnetic devices with enhanced performance and functionality.

Controlling interfacial magnetic phenomena between hard and soft ferromagnetic (FM) layers in heterostructures plays a key role in next generation spintronic and magnetic memory devices. In this dissertation, the tuning of interfacial properties through control of the thin film strain and layer thickness in La2/3Sr1/3CoO3 (LSCO)/La2/3Sr1/3MnO3 (LSMO) bilayers are discussed. Soft x-ray magnetic spectroscopy was performed to study the valence states and bonding configurations of the transition metal ions as well as the magnetic exchange coupling of the bilayers. Magnetocrystalline anisotropy and exchange bias were further explored LSCO/LSMO bilayers grown on NdGaO3 substrates, where soft x-ray linear dichroism spectra highlighted the differing electron occupancy along the two in plane directions of the LSMO layer with increasing LSCO layer thickness.

Investigating the potential of ion migration induced alteration on functional properties has become a prominent research focus in the rapidly evolving field of neuromorphic computing. Perovskite oxides, including cobaltites (LSCO), manganites (LSMO), and ferrites (La0.7Sr0.3FeO3, LSFO), have emerged as prime candidates for their remarkable characteristics, such as high oxygen vacancy conductivity, relatively low oxygen vacancy formation energy, and the strong interplay between magnetic and electronic properties with oxygen stoichiometry. This dissertation delves into the evolution of physical properties exhibited by LSCO, LSMO, and LSFO thin films when subjected to highly reducing environments, specifically H2 and vacuum conditions. In the cobaltite systems, the introduction of H2 leads to the formation of oxygendeficient phases at lower temperatures compared to vacuum annealing but remain in the perovskite phase without phase transitions at higher hydrogenation temperatures. The manganites, on the other hand, require more reduced conditions under vacuum annealing to achieve a mixed phase of brownmillerite and Ruddlesden Popper phases. Notably, the ferrite films maintain their oxygen deficient perovskite phase even when subjected to the most reducing conditions but shows significant differences in their electronic properties. The strong correlation between crystal structure and the magnetic/electronic properties underscores the potential of leveraging ion migration as a foundation for emerging applications such as neuromorphic computing.

Perovskite oxides ABO3, where A is usually an alkaline-earth or rare-earth cation and B a transition metal element, have received a great deal of interest because of their wide range of functional properties such as ferroelectricity, superconductivity, and ferromagnetism, with a unique sensitivity to external stimuli including chemical doping, lattice strain, and electrical and magnetic fields. This sensitivity makes perovskite oxides highly tunable materials, and further investigation into their fundamental physical properties are critical for future device development. Unlike metallic systems that have been studied for decades for a variety of applications such as hard disk drives and logic/memory devices, thin films of perovskite oxides are a relatively new field and they show diverse functional properties compared to their metal counterparts, creating novel pathways for designing specific device applications. The work in this dissertation was focused on extending our current knowledge on the magnetic perovskite oxides, and further exploring the ability to precisely tune and control their functional properties.

Resonant x-ray reflectivity, soft x-ray photoemission electron microscopy (X-PEEM), and magnetometry measurements were employed to study the physical properties of perovskite oxide Nd0.5Sr0.5MnO3 (NSMO) thin films grown on (110)-oriented SrTiO3 substrates, where it displays coupled magnetic and electronic transitions from paramagnetic/insulator to ferromagnetic (FM)/metal and then to antiferromagnetic (AFM)/charge-ordered insulator with decreasing temperature, due to the anisotropic strain state induced by the underlying substrate. The FM and AFM properties of the NSMO film were probed as a function of temperature using soft x-ray magnetic spectroscopy, and the coexistence of lateral FM and AFM domains was demonstrated using X-PEEM, showing both vertical and lateral magnetic phase separation in the film. These characterization techniques demonstrate the multiple magnetic and electronic phase transitions in the NSMO films, and thus the possibility and diversity of functional properties to be applied in next generation devices.

Ion migration-induced modification of physical properties is another emerging research direction in the search for tunable materials that can revolutionize the growing field of neuromorphic computing. Among the candidate materials, perovskite oxides including cobaltites (La0.7Sr0.3CoO3, LSCO and LaCoO3, LCO) and ferrites (La0.7Sr0.3FeO3, LSFO) are ideal candidates due to their high oxygen vacancy conductivity, relatively low oxygen vacancy formation energy, and strong coupling of the magnetic and electronic properties to the oxygen stoichiometry. The evolution of the physical properties of LSCO, LCO and LSFO thin films upon exposure to highly reducing environments was studied in this dissertation. In the cobaltite systems, the rarely-reported crystalline Ruddlesden-Popper (RP) phase was observed, which involved the loss of both oxygen anions and cobalt cations upon annealing, where the cobalt is found as isolated Co ions or Co nanoparticles. First principles calculations confirm that the concurrent loss of oxygen and cobalt ions is thermodynamically possible. In the ferrite system, however, the films remained in the oxygen-deficient perovskite phase even when annealed at the most reducing conditions explored in this dissertation. The strong correlation of the magnetic and electronic properties to the crystal structure highlights the potential of utilizing ion migration as a basis for emerging applications such as neuromorphic computing.

Artificial spin ices (ASIs) are lithographically-patterned arrays of interacting magnetic nanoislands which have traditionally been designed such that each nanoisland possesses an Ising-like magnetic configuration (i.e., behaves as a bar magnet). These systems were originally introduced as an engineered analog of spin-ice crystals whose atomic-level magnetic ordering could be emulated and visualized with conventional magnetic microscopy techniques. Since this initial study, ASIs have served as a platform for designing and investigating phenomena including phase transitions between different global magnetic ordering states, magnetically-reconfigurable spin wave dynamics, and computation in nanomagnet-based devices.

Many ASI studies have used Ni80Fe20-based nanoislands whose material and geometric properties enforce the formation of uniform, Ising-like magnetizations in each nanoisland. While these systems have facilitated the investigation of magnetic ordering phenomena in various array geometries, ASIs fabricated from ferromagnetic complex oxides offer opportunities to explore how the nanoisland material properties and magnetostatic interactions can influence the behavior of the individual nanoislands and ASI as a whole. Understanding how these magnetic parameters can tailor the nanoisland functionalities will be essential in the development of nanomagnet-based computing devices.

This dissertation focuses on ASIs fabricated from the ferromagnetic perovskite oxide La0.7Sr0.3MnO3 (LSMO) to understand how the interaction between magnetostatically-coupled nanostructures can influence the formation of spin textures. Soft x-ray photoemission electron microscopy (X-PEEM) was used to perform magnetic domain imaging of LSMO-based ASIs. Through these imaging studies, these systems were observed to allow the formation of both Ising states, observed in traditional ASI systems, as well as complex spin textures (CSTs), comprised of single and double vortices. The formation of these different spin texture states were found to be sensitive to the ASI geometry as well as the magnetization of the neighboring nanoislands. Through the use of micromagnetic simulations, an energetic analysis of the system was performed which revealed how spin texture formation arises from a competition between effects intrinsic to each nanoisland as well as the magnetostatic interaction between adjacent nanoislands. This competition is facilitated by the unique combination of magnetic material properties of LSMO.

The unusual domains stabilized in LSMO-based ASIs also raises the question of the relaxation kinetics of CST-bearing ASI systems. Insights into the thermal relaxation behavior of LSMO-based ASIs was provided using X-PEEM and an in-situ pulsed annealing protocol which allows for “snapshots” of the time-dependent relaxation to be acquired. By tracking the domain changes in each nanoisland as a function of the annealing time, several types of apparent domain conversion were observed whose relative transformation rates evolved over time. These results suggest that the domain formation kinetics of each individual nanoisland is sensitive to the evolving magnetizations of the neighboring nanoislands.

These static and dynamic studies of LSMO-based ASIs demonstrate how magnetostatic interactions can be used to tailor spin texture formations in these systems. Moreover, this work highlights the potential for material-driven ASI studies to provide deeper insights into ASI and domain formation physics.