A Weight-Tunable Oxide Synaptic Transistor

Artificial neural systems demonstrate outstanding performance by processing sensory data and realtime contexts in parallel, far surpassing conventional von Neumann computers in terms of energy efficiency. This property has sparked widespread interest within the field of artificial intelligence. In recent years, research on electric-double-layer synaptic transistors has attracted much attention due to their similarity in ionic motion modulation to that of biological synapses, thereby demonstrating a wealth of potential application scenarios. In this paper, we demonstrate a solid-state electrolyte-gated transistor that uses TaO x with a unique ionic composition as the insulating layer and transparent indium tin oxide (ITO) as the semiconductor layer to prepare artificial synaptic thin-film transistors (TFTs) with signal transmission and self-learning properties. The devices exhibit significant memory holding, memory bank voltage exceeds 6V at operating voltages below 10V. The shift of oxygen Void Sites in the insulating layer in synaptic transistor, entrained by electrical forces created from input signals, plays a crucial role in simulating synaptic behavior. In addition, the device successfully simulates the enhancement and inhibition of synaptic weights, such as excitatory postsynaptic current (EPSC), inhibitory response (IPSC), paired pulse facilitation (PPF), and long-term potentiation (LTP).