神经形态工程学
光电流
光电子学
材料科学
光探测
钻石
紫外线
光电探测器
暗电流
基质(水族馆)
可见光谱
光学
计算机科学
物理
人工智能
人工神经网络
复合材料
地质学
海洋学
作者
Zitong Liu,Yang Bing,Peitao Liu,Longhai Shen,Ouxiang Zhou,Hanning Xu,Xiaokun Guo,Xinglai Zhang,Junchao Wang,Lijie Deng,Siyu Feng,Lizhi Feng,Baodan Liu,Dongming Sun,Xing‐Qiu Chen,Xin Jiang,Bing Leng
出处
期刊:ACS Nano
[American Chemical Society]
日期:2025-06-02
卷期号:19 (22): 21130-21142
被引量:4
标识
DOI:10.1021/acsnano.5c05737
摘要
Diamond, an important electronic and optoelectronic material, exhibits excellent performance in deep ultraviolet photodetection. However, its lack of visible light response and nonvolatile photocurrent hinders its application in neuromorphic sensors that mimic the human eye. In this work, we report a diamond-based neuromorphic retina perception system (NRPS) that demonstrates broad-spectrum photoresponse beyond the human visual range, nonvolatile photocurrent, and wavelength sensitivity. This outcome is achieved by introducing a large number of defects and doping states into the diamond film through heteroepitaxial growth on the SiO2 substrate and Si doping. By adjusting the input light and electrical stimulation, human eye's pupil and image memory behavior are successfully simulated. More interestingly, the diamond NRPS exhibits a significant difference in the decay characteristic of photocurrent under different wavelengths of illumination. Based on this characteristic, the NRPS achieves recognition of unknown colors and trajectory tracking of dynamic light dots. Furthermore, the image recognition accuracy is improved from 67.5% to 81.7% by filtering the visible light information and extracting useful ultraviolet light information with the device. This study offers an effective strategy to develop a diamond-based neuromorphic sensor.
科研通智能强力驱动
Strongly Powered by AbleSci AI