Memristors: A Missing Element is a Boon Toward the Development of Neuromorphic Computing and AI

The advancement and research in memristor technology are proving to be highly beneficial in various fields, particularly in the realm of Artificial Intelligence and in assisting patients with autism disorders. Memristors possess unique capabilities such as simultaneous programming and data storage, emulation of brain cell behavior, and power-independent stability, making them a promising component. The chapter aims to delve into the manufacturing process of memristors and their crossbar structure. It is likely to discuss the utilization of heterojunction layers in the development of nanowire (NW) HfO2 memristors and acknowledge the device’s remarkable endurance for nearly two decades. However, along with the growing research in this field, there are challenges that must be addressed, such as enhancing conductivity and storage capacity. The chapter is also expected to cover the fabrication methods employed to evaluate multi-layered memory stores and the investigation of the device’s surface morphology using the FEGSEM technique. Overall, the chapter will provide insights into the potential of memristors and their role in advancing various fields, particularly in the realm of neuromorphic computing. The chapter appears to discuss the properties and applications of two types of memristor devices: TiO2 nanoparticles and Gd-doped HfO2 nanoparticles. Through X-ray diffraction analysis, it has been observed that TiO2 nanoparticles possess a crystalline nature, and the presence of oxygen vacancies increases with heat application. Various studies have examined the Gd-doped HfO2 nanoparticles memristor device, including annealing, energy-dispersive X-ray, and photoluminescence examinations. The findings indicate device was subjected to an annealing process at a temperature of 600 °C degrees and demonstrates a notable enhancement in the terms of leakage current and interference state density factor. Additionally, the chapter delves into the potential applications of memristor devices in advanced neurocomputing, specifically as synaptic simulators connecting pre- and postsynaptic neurons. The chapter examines the utilization of memristor devices incorporating TiO2 nanoparticles with two different Ag and Au electrodes by highlighting their properties and applications. The nanoparticle morphology verified using atomic force microscopy (AFM). Memristor behavior does indeed resemble that of a synaptic emulator, as it can modify its resistance based on the electrical signal history passing through it. This characteristic positions memristors as a promising candidate for constructing artificial neural networks and other neuromorphic computing systems. Regarding the comparison between memristors with Ag and Au contacts, measurements of capacitor–voltage curves have shown that Ag-based memristors exhibit superior charge-storage capabilities. This is attributed to Ag’s lower work function compared to Au, enabling efficient electron exchange with other materials. As a result, Ag serves as an excellent electrode material for memristors, facilitating efficient charge storage and release. Memristor devices with Ag electrodes have been proposed as pre- and post-neuronal systems and have demonstrated comparable performance to Au electrode memristors. Additionally, these devices may offer significant advantages in the field of neuromorphic computing by potentially enhancing performance and energy efficiency. However, further research is necessary to fully comprehend the capabilities and limitations of Ag electrode memristors for neuromorphic computing applications.