Fabrication Characterization and Simulations of Nanoscale Memristors for Adoptable Future Technologies

Abstract

The first part of the thesis emphasizes the expediency of controlled defect incorporation in transition-metal-oxide (like TiO2) for obtaining excellent memristive features. In this regard, industry-friendly pulsed electron beam deposition and ion-beam implantation techniques are employed to fabricate the TiO2 switching layers with desired structural, electronic, and transport properties. Furthermore, using density functional perturbation theory-based simulations, a novel methodology is devised to achieve exotic electronic and transport characteristics of TiO2 by tuning the electron-phonon coupling under external agitation (strain and temperature). The second part of the thesis is focused on the exploration of the efficacy of transition-metal-dichalcogenide (such as MoS2) nanocrystals for obtaining highly stable and low-voltage resistive switching devices. The fabricated devices are extensively characterized by electrical measurements to extract optimum memory output (hysteresis loop, endurance, etc.) at very low-operation voltages ranging from ± 0.5V to ± 2.5V. In addition, various synaptic activities, i.e., short-term and long-term plasticity, Hebbian learning algorithm, and gradual conductance change are also emulated to establish their cognitive behavior, desirable for neuromorphic computing. Further, the active layers are investigated by different microscopic and spectroscopic techniques, as well as the density functional theory-based first-principles simulations for revealing the atomistic origin of memristive behavior. Overall, this thesis provides unique pathways for designing energy-efficient, high-performance, and scalable memristors for future NVM technology and neuromorphic computing.