Photoelectric Synaptic Device Based on Bilayerd O<sub>R</sub>/O<sub>P</sub>-InGaZnO for Neuromorphic Computing
Jieru Song, Jialin Meng, Lu Chen, Tianyu Wang, Hao Zhu, Qingqing Sun, David Wei Zhang, Lin Chen
Abstract
Photoelectric synaptic devices have exhibited remarkable potential in the field of neuromorphic computing (NC), facilitating high-speed and energy-efficient neuromorphic operations. Here, we explored the application of Indium-Gallium-Zinc-Oxide (IGZO) in two-terminal artificial synaptic devices for NC. The dielectric layer of the device, comprising an oxygen-rich layer and an oxygen-poor layer, effectively harnessed the migration of oxygen vacancies at their interface to exhibit memristive properties. The device emulated synaptic characteristics successfully under both electrical and optical stimuli, including excitatory post-synaptic current (EPSC), paired-pulse facilitation (PPF), long-term potentiation (LTP), and long-term depression (LTD). Furthermore, the recognition of handwritten digits was realized based on the synaptic weight resulting from light pulse stimulation, and the recognition rate was up to 92.6%. The results illustrated the advanced potential of IGZO-based memristors in NC.