A Low-Cost Flexible Optoelectronic Synapse Based on ZnO Nanowires for Neuromorphic Computing

Sensors (Basel). 2024 Dec 5;24(23):7788. doi: 10.3390/s24237788.

Abstract

Neuromorphic computing, inspired by the brain, holds significant promise for advancing artificial intelligence. Artificial optoelectronic synapses, which can convert optical signals into electrical signals, play a crucial role in neuromorphic computing. In this study, we successfully fabricated a flexible artificial optoelectronic synapse device based on the ZnO/PDMS structure by utilizing the magnetron sputtering technique to deposit the ZnO film on a flexible substrate. Under UV light illumination, the device exhibits excellent synaptic plasticity, including excitatory postsynaptic current (EPSC), short-term potentiation (STP), and paired-pulse facilitation (PPF). By growing ZnO nanowires, we improved the fabrication processes and further enhanced the synaptic properties of the device, demonstrating long-term potentiation (LTP) and the transition from short-term memory (STM) to long-term memory (LTM). Additionally, the device exhibits outstanding flexibility, maintaining stable synaptic plasticity under bending conditions. This device shows broad application potential in mimicking visual systems and is expected to contribute significantly to the development of neuromorphic computing.

Keywords: ZnO; flexible optoelectronic synapse; photoconductivity; synaptic plasticity.