Solution-Processed Small Molecule Memristors: From Nanowire Arrays to Thin-Films

Conventional computing architectures based on the Von Neumann model are nearing their physical and operational limitations, driven by the breakdown of Moore’s law, memory bottlenecks, and challenges in heat dissipation. As traditional approaches become less viable, alternative computing paradigms such as neuromorphic computing are gaining attention. Neuromorphic systems emulate the neural structures of the brain, enabling efficient integration of processing and memory. Among various implementations, organic electronic memristors have emerged as promising candidates for next-generation neuromorphic devices. This thesis investigates memristor devices fabricated from two squaraine derivatives, demonstrating multistate memory behaviour and strong retention capabilities. Squaraine nanowire-based memristors were developed and shown to exhibit key neuromorphic functionalities, including synaptic plasticity features such as long-term potentiation and depression over 1000 cycles, paired-pulse facilitation on the millisecond scale, and spike-timing dependent plasticity in alignment with Hebbian learning principles. The operational mechanisms of these devices were analysed under varying environmental and thermal conditions using a suite of electronic and electrochemical characterisation techniques. While space-charge limited current and injection/extraction processes remain inconclusive, ionic contributions are also ruled out. The results point towards a combined influence of light, moisture, and oxygen in modulating the memristive behaviour. In the final chapter, the work advances to scalable thin-film architectures, demonstrating reduced energy consumption per synaptic event, achieved through reduced voltage input, along with multistate conductance. These thin film memristors were able to demonstrate synaptic plasticity through pulsed paired facilitation and spike-time dependent plasticity. These findings underscore the potential of squaraine based memristors as energy-efficient, scalable components for future neuromorphic computing platforms. By bridging organic materials science and neuromorphic engineering, this research contributes to the development of next generation computing technologies that transcend the limitations of current architectures, offering a pathway toward more adaptive, efficient, and biologically inspired information processing systems.