Transformation of Analog to Digital Resistive Switching in Cu Implanted ITO/NiO/Ag Device for Neuromorphic Applications

Abstract

Both analog and digital resistive switching are essential components in the neuromorphic computing system. This work reports the influence of Cu ions for the transformation of analog to digital resistive switching in ITO/NiO/Ag memristor devices. The undoped and low-concentration Cu doping illustrates the analog switching, whereas higher doping demonstrates the digital characteristics. At higher bias voltage, the Schottky barrier is developed at both ITO/NiO and NiO/Ag interfaces. The increasing and decreasing of current conduction with the escalating number of cycles for both the polarity in undoped and low doped is elucidated by the electrode-dominated mechanism in terms of reduction and enhancement of Schottky barrier height at the interface, respectively. The digital switching characteristic due to the formation and rupturing of the vacancy filament at higher doped sample is induced due to the boosting of vacancies above the critical amount using ion implantation. The synergic effect of current conduction due to local Cu migration and oxygen vacancies can be utilized as a learning and forgetting process for neuromorphic applications.