ZnS Optoelectronic Synapses with Tunable Volatile and Nonvolatile Resistive Switching for Neuromorphic Learning

Artificial optoelectric synapses that integrate dual functions of memristor and a selector hold significant practical potential for developing next-generation high-density neuromorphic computing systems. Herein, an optoelectronic synaptic memristor based on ZnS is proposed, enabling reversible switching between long-time stable (>10⁴ s) nonvolatile bipolar resistive switching characteristics (on/off ratio >10⁵) and a tunable volatile self-rectifying mode (rectification ratio ranging in 45-10⁴), thus functioning as integrated selector and memristor. The device realizes multilevel memory and effectively emulates diverse biological synaptic plasticity behaviors under electrical or near-infrared (808 and 980 nm) light stimulation, while also reproducing the brain-like Ebbinghaus "learning-forgetting-consolidation" rules. This work holds scientific significance for advancing the potential of devices with self-suppressing sneak current functionality in brain-inspired neuromorphic computing systems.