Neuromorphic computing, which utilises human brain-like computing architecture for energy-efficient operation, is gaining interest due to the increasing demand for artificial intelligence. A key component of this computing system is the use of oscillator-based spiking neural networks (SNNs) or oscillatory neural networks (ONNs). These networks require compact and low-power solid-state oscillators to function effectively. Metal-oxide-metal (MOM) devices, which can exhibit volatile threshold switching or negative differential resistance (NDR), are a promising candidate for fabricating such oscillators. This NDR is due to the super-linear temperature dependence of the oxide conductivity, and oxides that exhibit an insulator-metal transition (IMT) provide a basis for low-power operation due to their large conductivity change over a narrow temperature range. Very few of the oxides that exhibit reliable NDR have IMT temperatures above room temperature.

V3O5 is a stable phase of vanadium-oxide systems that exhibit IMT at ~420 K. In my thesis, I have explored the structural, electrical, and thermal properties of V3O5 thin films and their application as a functional oxide in MOM relaxation oscillators. Planar V3O5 devices show electroforming-free volatile threshold switching and negative differential resistance (NDR) with stable (less than 3% variation) cycle-to-cycle operation. The physical mechanisms underpinning these characteristics are investigated using a combination of electrical measurements, in situ thermal imaging, and device modelling. This shows that conduction is confined to a narrow filamentary path due to the self-confinement of the current distribution and that the NDR response is initiated at temperatures well below the IMT temperature, where it is dominated by the temperature-dependent conductivity of the insulating phase. Further, I investigate the dynamics of individual and coupled V3O5-based relaxation oscillators and show that capacitively coupled devices exhibit rich non-linear dynamics, including frequency and phase synchronisation. Finally, a leaky integrate-and-fire (LIF) neuron is demonstrated using a V3O5-based threshold switching device, which shows excitatory and inhibition spiking behaviour like biological spiking afferent nerve. These results establish V3O5 as a promising material for volatile threshold switching and advance the development of robust solid-state neurons for neuromorphic computing.