Artificial\nSynapse Based on van der Waals Heterostructures with Tunable Synaptic\nFunctions for Neuromorphic Computing

Abstract

Two-dimensional\n(2D) materials and van der Waals heterostructures have attracted tremendous\nattention because of their appealing electronic, mechanical, and optoelectronic\nproperties, which offer the possibility to extend the range of functionalities\nfor diverse potential applications. Here, we fabricate a novel multiterminal\ndevice with dual-gate based on 2D material van der Waals heterostructures.\nSuch a multiterminal device exhibited excellent nonvolatile multilevel\nresistance switching performance controlled by the source–drain\nvoltage and back-gate voltage. Based on these features, heterosynaptic\nplasticity, in which the synaptic weight can be tuned by another modulatory\ninterneuron, has been mimicked. A tunable analogue weight update (both\non/off ratio and update nonlinearity) of synapse with high speed (50\nns) and low energy (∼7.3 fJ) programming has been achieved.\nThese results demonstrate the great potential of the artificial synapse\nbased on van der Waals heterostructures for neuromorphic computing.