Optically Tunable Memristors Based On Organic-Inorganic Composite Materials

Jaafar, Ayoub H.

Abstract

Memristors or resistive memory devices (also called RRAMs) are one of the most promising emerging memory technologies for the next generation of memory devices with potential applications in computer memory, logic operations and neuromorphic computing. Memristor devices are unique because they have non-volatile memory operation in combination with fast nano-second switching, low power operation and high-density integration using nanoscale crossbar-array architectures. Fundamentally, memristors are two terminal devices; however, in certain applications, there are advantages in having an additional degree of freedom available to control their switching and electronic properties. This thesis work examines how this can be achieved using light.
Modulation of the electronic properties of memristors by optical means offers a new level of functional control, enabling the development of new types of optoelectronic devices and circuits, such as photonic integrated circuits with memory elements controllable by light. Memristors also have important applications in neuromorphic computing and in that context the dynamic and spatial patterning by light opens the route to new optically configurable and tunable synaptic circuits. In this thesis, two novel types of optically tunable switching memristors based on hybrid organic-inorganic composites are developed. The first type of device consists of vertically aligned zinc oxide nanorods (ZnO NRs) integrated with an optically active polymer, poly(disperse red 1 acrylate) (PDR1A), which expands and contracts upon irradiation by wavelength and polarization specific light. The second device type is instead based on gold nanoparticles (Au NPs) distributed in discrete conducting pathways throughout the same photoactive polymer. These break or re-connect upon exposure to light, and because there are only a few conducting pathways in a single device, much greater changes (3-4 orders of magnitude) occur in the electrical conductivity. Both device types exhibit optically dependent switching properties and have too the advantage of fabrication by very simple, low cost, solution-based methods.
The electronic properties of both optical RRAMs show bipolar switching behaviour with either sharp digital-like switching (Au NPs device) or a complex transition with space charge current limited like behaviour (ZnO NRs device). Both device types exhibit optical control of the devices’ electronic properties and in particular, we demonstrate for the first time, latched optical SET and RESET switching processes.
Lastly, this thesis examines how optical memristors can be used to control learning in neuromorphic applications. We demonstrate optical control of short-term and long-term memory and tunable learning through spike timing dependent (synaptic) plasticity. We believe this has important applications in the dynamic patterning of memristor networks, whereby both spatial and temporal patterning via light allows the development of new optically reconfigurable neural networks, adaptive electronic circuits and hierarchical control of artificial intelligent systems.

Citation

Jaafar, A. H. (2018). Optically Tunable Memristors Based On Organic-Inorganic Composite Materials. (Thesis). Univrsity of Hull. https://hull-repository.worktribe.com/output/4961915

• Thesis Type: Thesis
• Deposit Date: Dec 9, 2024
• Publicly Available Date: Dec 9, 2024
• Keywords: Mathematics; Physical sciences
• Public URL: https://hull-repository.worktribe.com/output/4961915
• Additional Information: School of Mathematics and Physical Sciences; University of Hull
• Award Date: Nov 16, 2018