Negative Photoconductivity in Nanowires/QDs Heterojunction Network for Neuromorphic Visual Perception

Neuromorphic visual systems, integrating image information acquisition, processing, and storage, hold significant potential for applications in smart security, healthcare, and other fields. Currently, commercial human visual prosthetic systems primarily utilize Si-based CMOS logic circuits to simulate synaptic functions. Compared to traditional Si-based materials, metal oxide nanowires offer unique advantages, such as a larger specific surface area and sustained photoconductivity. However, artificial synaptic devices based on metal oxide nanowires can only achieve unidirectional photoresponse enhancement, posing challenges in mimicking hyperpolarization behavior of cone and rod cells and further developing image classification and recognition functions. Herein, an artificial photo-synaptic device with negative photoconductivity based on metal oxide nanowires/perovskite quantum dots heterojunction fabricated via low-cost electrospinning is proposed. The carrier dynamics process is investigated using advanced characterization techniques. It is revealed that the negative photoconductivity is induced by heterojunction-associated conduction carrier consumption. Based on the device array, the optical encryption function is achieved. The array with adjustable conductivity is used to develop an artificial vision system, reducing redundant data and noise while improving recognition accuracy from 51% to 99%. This study highlights the promising potential of optoelectronic synaptic devices based on metal oxide nanowires and perovskite heterojunctions for artificial visual systems.