UC San Diego

The development of novel materials and electronic devices with distinctive physical

characteristics is revolutionizing the path toward the next major advancement in neuromorphic

computing. Vanadium oxides (VOx) are particularly notable among functional materials as

leading contenders for the future of oxide electronics. This distinction arises from their metal-

insulator transitions (MITs) at or near room temperature, which can be induced by heat, electric

fields, or other external factors. These properties enable the implementation of neuronal or synaptic functionalities, making them essential components for constructing neuromorphic

systems. For technological applications, one of the major challenges in complex oxides is to

obtain material phases with reliable electronic performance in thin films. However, for materials

like vanadium oxides, due to existence of large number of oxidation states, synthesis conditions

for one phase are often detrimental to the stability of the other. Synthesizing thin films that

incorporate multiple vanadium oxide phases is particularly challenging, especially if different

phases must be organized into well-defined spatial patterns for applications. Despite decades of

research, a universal methodology for obtaining high-quality thin films of a specific VOx phase

is still lacking. In Chapter 2 and 3 of dissertation, we proposed and established an efficient solid-

state reaction laser annealing (SRLA) approach to directly write regions of different local

chemical compositions. Using this method, we achieved the controlled local recrystallization of a

uniform V2O3 thin film into VO2, V3O5, and V4O7 regions exhibiting sharp 1st and 2nd order

metal-insulator transitions over a wide range of critical temperatures. We utilized this approach

to pattern spiking oscillators with distinct electrical behavior directly in the vanadium oxide thin

film without employing elaborate lithography fabrication. Our method opens a pathway to

synthesizing a wide range of artificially micropatterned composites, with precision and control

unattainable in the conventional material fabrication methods. In Chapter 4 of dissertation, we

extended our study to superconductivity, and present a rapid thermal hydrogenation approach to

manipulate the critical transition temperature (Tc) of Nb thin films. In analog to the MIT,

superconductor-based neuromorphic computing is another promising aspect of this technological

field. Our works offer fundamental understanding of hydrogen incorporation into superconductor

materials.