Artificial synaptic plasticity and associative learning of the nanostructured MoS₂/Cu₂S heterojunction based memristor

Based on the isomorphism between memristor and biosynapse, memristor can be applied to simulate the artificial synaptic behavior, thus the memristor can achieve brain-like neuromorphic computing. In this work, MoS₂ film was deposited on a Cu₂S/Ti substrate to successfully fabricating an Ag/MoS₂/Cu₂S/Ti heterostructure based memristor. Further, it was found the current-voltage (I-V) curve of the device exhibits typical non-volatile resistance switching characteristics, and maintained stable cycling performance and high switching ratio. By analyzing the formation and rupture of Ag⁺ ions and sulfur vacancy conductive filaments, the charge transfer mechanism of the Ag/MoS₂/Cu₂S/Ti memristor was investigated. By applying specific pulse sequences to the device, various biosynaptic functions, such as short-term potentiation (PPF), long-term potentiation/depression (LTP/D), and spike-timing dependent plasticity (STDP), were successfully simulated. Finally, in the associative learning-Pavlov's dog experiment simulation, the device requires only a few training iterations to re-establish the conditioned response, demonstrating excellent learning capabilities.