Near‐Infrared Optoelectronic Memristor with All‐Optical Modulation for Microscopic Biological Motion Recognition
Jiaqi Han, Hongchen Mao, Ya Lin, Ke Peng, L. Meng, Yadan Ding, Xuanyu Shan, Ye Tao, Jiajuan Shi, Zhongqiang Wang, Hancheng Zhu, Xiaoning Zhao, Haiyang Xu, Yichun Liu
Abstract
Near-infrared (NIR) synaptic devices are characterized by unique optical sensitivity and synaptic plasticity, possessing high potential to significantly broaden the visual spectrum accessible to humans. Moreover, these devices seamlessly integrate neuro-morphological computing with infrared optical communication, attracting significant attention for future low-power, high-efficiency neuromorphic visual systems. However, the irreversible optical response of most optical synapses under short-wavelength infrared light severely restricts the practical application of these devices. To overcome this limitation, this work proposes a NIR optoelectronic memristor with all-optical modulation based on ZnO-upconversion nanoparticles (ZnO-UCNPs) nanocomposite film. The proposed device exhibits bidirectional light response characteristics under 350 and 980 nm light stimulations. The memristive mechanism can be attributed to the photothermal effect of upconversion nanoparticles in the nanocomposite film in conjunction with optical excitation in ZnO. Due to the all-optical plasticity, a series of logical functions with fault-tolerance capability, including are demonstrated in the proposed synaptic device. In addition, by leveraging the tissue-penetrating capability of NIR light, the proposed device can be used in microscopic biological motion detection tasks and edge detection functions. This work introduces a promising method for developing fully optically controlled memristors and advancing the field of neuromorphic vision.