Abstract The failure of Moore’s Law has hindered traditional silicon-based devices from meeting future demands for large-scale information storage and data processing. Recently, synaptic mimicry, inspired by biological nervous systems, has gained attention. Memristors, with their intrinsic properties, can regulate conductance through charge or magnetic flux, mimicking synaptic mechanisms, and can serve as synaptic mimicry elements. These devices switch between high and low resistance states when subjected to an electric field. Resistive switching phenomena have been observed in materials like transition metal oxides, perovskites, and solid-state electrolytes. Graphene oxide, known for its conductivity, ductility, adjustable band gap, and ease of room-temperature processing, has garnered research interest. Graphene oxide memristors present advantages such as low power consumption, high density, and uniformity, ideal for high-performance components. This study investigates graphene oxide-based memristors with an Ag/GO/Ag structure. Devices show large switching ratios and excellent electrical performance. Synaptic mimicry was explored through short-term and long-term memory tests by adjusting input pulses. Results demonstrate that a single memristor can simulate key synaptic plasticity functions, paving the way for AI applications in neuromorphology. These findings highlight the potential of graphene oxide memristors for future information processing and data storage.
Sui et al. (Tue,) studied this question.