Artificial synapses based on electric stress induced conductance variation in vertical MoReS3 nanosheets

Abstract

As promising nonvolatile memory devices, memristors based on transition metal dichalcogenides, such as MoS2 and MoTe2, have received much attention for their polymorphism and distinct electronic characteristics. However, state-of-the-art memristors with horizontal placement of sandwiching 2D material layers suffer from poor performance in switching variability and endurance for the preparation and transfer process. Herein, we report a memristor based on vertical MoReS3 nanosheets spread on the surface of carbon fiber cloth. The atomic-scale thickness, combined with the structural transition from the T′ phase to the T″ phase in Janus MoReS3 nanostructures, lowers the spatial (device-to-device) and temporal (cycle-to-cycle) variation and prolongs the lifespan of the device during resistive switching processes. Furthermore, our memristor achieves a high device yield and accurate analogue programming and, thus, exhibits the synapse-like learning behavior such as short-term memory and long-term memory. These results demonstrate the potential of direct modulation of 2D materials with electric stress and motivate further research in implementation of artificial synapses.