Research progress in artificial synapses based on two-dimensional materials

Abstract Artificial synapse is one of the most important fundamental neuromorphic components in neuromorphic network, a novel computing architecture capable of integrating perception, memory and computation, therefore offering a promising pathway to overcoming the von Neumann bottleneck. Synapse plays a central role in mimicking learning and memory functions through synaptic weight update. A variety of materials have been explored for the fabrication of artificial synapses. Among them, two-dimensional (2D) materials stand out due to their excellent stability at atomic thickness and their rich physical properties such as the responsiveness to optical, electrical and mechanical stimulus, and the coupling between various physical fields. These advantages make 2D material-based artificial synapse a research hotspot. This article briefly reviews the recent progress of artificial synapses based on 2D materials. Both two-terminal memristors and three-terminal memtransistors are summarized based on their core working mechanisms, systematic classification and in-depth discussion are also carried out. Finally, the challenges of 2D-material-based synaptic devices are discussed, along with their broad application prospects.