All-Solution-Processed IGZO Optoelectronic Synaptic Transistor with Dual-Mode Operation toward Artificial Vision...

Realizing the dual-mode electric and optical synaptic plasticity within one neuromorphic device is impressive for the construction of a compact artificial visual system. Here, we proposed indium gallium zinc oxide (IGZO) photoelectric synaptic transistors utilizing all-solid-state electrolytes (Li-doped ZrO₂) as gate dielectric layers. The device was fabricated by using a facile and cost-effective all-solution method. The synaptic transistor exhibited dual-mode electric and optical synaptic plasticity. Meanwhile, the tunable conductance is achieved through electric potentiation and depression processes, demonstrating the potential for realizing neuromorphic computing. Based on this, a simulated convolutional neural network was designed to realize handwriting digit recognition, achieving an accuracy of 96.8%. Additionally, sophisticated neuromorphic applications such as logic operations, Pavlov's classical experiment, and pupillary reflex simulation were successfully realized. Therefore, the designed transistor demonstrates significant potential for future applications in artificial vision.