ZnO photoconductive synaptic devices for neuromorphic computing

Qiuchen Shang,Wenbo Peng,Tuo Song,Zeyang Li,Fangpei Li,Yongning He

DOI: https://doi.org/10.1016/j.mssp.2023.107489

IF: 4.1

2023-04-15

Materials Science in Semiconductor Processing

Abstract:In the field of artificial intelligence, neuromorphic computing offers the allure of human brain simulation. The capacity of optoelectronic synapses to intelligently handle optoelectronic input signals makes them a key component of neuromorphic computing. In this work, we fabricated ZnO-based optoelectronic synapses and investigated their electronic and photonic synaptic plasticity. The device fabrication process flow is as follows: photolithography is performed on ITO glass, followed by hydrochloric acid etching and degumming to obtain interdigital electrodes on the glass substrate, and a ZnO film is grown on the interdigitated electrodes by magnetron sputtering. A good Ohmic contact is formed between the ZnO and ITO electrodes, and the device exhibits a linear relationship between current and voltage under dark field conditions, and exhibits excellent synaptic properties and sensitivity in UV switching response experiments. Additional processes were subsequently introduced through growing ZnO nanowires and annealing, and the synaptic properties of ZnO device were improved. The shift from short-term to long-term memory, short-term plasticity, long-term plasticity, paired-pulse facilitation, and learning to experience behavior are only a few of the synaptic activities of the nervous system that have been successfully described. The manufactured array device also includes a special memory recall feature that may retrieve previous optoelectronic data. This work will advance the field of artificial intelligence and stimulate additional study on optoelectronic synapses.

engineering, electrical & electronic,materials science, multidisciplinary,physics, condensed matter, applied