Investigations on two aspects of two important classes of nanomaterials: resistive switching with synaptic applications of tio2 based memory devices and nonlinear optical properties of perovskite rbpbi3 using sspm technique

Abstract

In the human brain, distributed parallel processing happens due to a vast interconnected network of neurons. For this reason, the human brain can do high-speed data processing tasks and self-learning, language processing, and prediction tasks. And also, the human brain consumes very low power for doing these tasks. So, the human brain is a perfect example in which Von Neumann's bottlenecks are not seen. Researchers do their research in this domain to perfectly mimic the human brain for advanced neuromorphic computing and memory device applications. Titanium dioxide (TiO2) nanostructures represent a unique combination of shape and functionality where the dimensionality of the nanostructures directly influences the material's properties. Recent research is mainly focused on the cost-effectiveness and efficiency of the device. For these reasons, TiO2 thin films have been considered the primary material for memory device applications. First part of this thesis entitled "Study of Resistive Switching and Synaptic Properties on TiO2-Based Memory Devices and Its Application as Artificial Vision Sensor with Machine Learning Approach", aimed to investigate the resistive switching (RS) behaviours with synaptic functionalities of TiO2-based thin film devices. The TiO2 films are synthesized using the hydrothermal technique. Another major aim is to create a machine learning classifier model that collaborates with the RS device to artificially mimic the human vision system. After hydrothermal synthesis of TiO2 thin films, to obtain the best performing device, the growth time has varied from 30 min to 5 h. Following basic characterizations were done for all samples for deep scientific understanding. Fundamental memristor properties are evaluated by current-voltage (I-V) characteristics, followed by retention test, endurance test, and on/off current ratios measurement for all the devices. One hour (abbreviated asT1) device was selected as the best device by comparing all the above-said test results. This device exhibited up to 104 seconds retention time with 599 continuous complete endurance cycles and 8x104 on/off current ratio. In the next phase, all synaptic functions like STDP, SRDP, LTP, LTD, PPF, and PPD are emulated by the T1 device successfully for both electrical and optical stimuli. After that, we used the T1 device as an artificial retina for image sensing purposes in terms of light intensity. After a successful image is sensed, we put that image in the Support Vector Machine (SVM) Linear Classifier model (MATLAB program) for recognition purposes. T1 device successfully gave the best result for each input light. This T1 sample can be used in a robotic vision system as an artificial vision sensor (artificial retina) with an experimental linear Classifier model. In the second work, titled "Determination of Nonlinear Refractive Index (n2) and Third Order Susceptibility ((3)) for RbPbI3 Perovskite Nanorods in Different Solvents Using SSPM Method," we have done a nonlinear optical study of RbPbI3 nanorods. From the above study, the nonlinear optical parameters such as nonlinear refractive index and third-order susceptibility have been calculated using the spatial self-phase modulation (SSPM) technique. This study was done for two solvents to investigate the nonlinear refractive index's dependency on the solvents' viscosity. RbPbI3 halide perovskite nanorods are synthesized using a low-cost, simple chemical route. Basic characterizations like XRD, FESEM, UV-Vis absorbance, and EDS were done to understand this perovskite sample deeply. This type of study has been done for the first time for this material. As a result, RbPbI3 emerges to be a promising material for nonlinear optical applications. The calculated nonlinear refractive indexes were 2.165x10-5 and 2.088x10-5, and third-order susceptibility was 9.67x10-9 and 8.031x10-9 for NMP and 2-propanol solvents, respectively.