@article{HUANG2026, 
author = {Xinyu HUANG and Jihong WANG and Lin WANG and Hao SHI and Wenting JIAO and Tangjie MU and Yang GAO and Menglai ZHANG and Lei ZHANG and Junchao REN and Kun YIN and Hui YU and Yong WANG},
title = {Scalable TCNQ-Doped Graphene Oxide Sensor for Neuromorphic Machine Vision},
year = {2026},
journal = {Photonic Sensors},
volume = {16},
number = {2},
pages = {9560013},
keywords = {Neuromorphic machine vision, graphene oxide, programmable sensor arrays, convolutional neural networks},
url = {https://www.sciopen.com/article/10.26599/PhoS.2026.9560013},
doi = {10.26599/PhoS.2026.9560013},
abstract = {Neuromorphic machine vision systems have garnered substantial interest due to their potential for achieving autonomous control through real-time visual perception and processing. Two-dimensional materials offer a highly promising option for neuromorphic photodetectors due to their tunable electrical and optical properties, as well as their compatibility with heterogeneous integration. However, the fabrication of such devices often involves inefficient or expensive processes, limiting their widespread commercial adoption. To address these challenges, we develop a scalable 7,7,8,8-tetracyanoquinodimethane (TCNQ)-doped graphene oxide (p-GO) sensor by utilizing easily prepared graphene oxide as a substrate and employing a surface charge transfer doping strategy to modulate its charge state. This device exhibits a high and linearly tunable responsivity. By varying the applied bias voltage, the responsivity can be adjusted from 1.4 mA/W to 25 mA/W. With precise control over the photoelectric response at 76 levels, we establish a 3×3 array of p-GO sensors as programmable kernels for optical image edge processing and convolutional neural networks, achieving the impressive accuracy of up to 97.7% in letter recognition tasks. We anticipate that this work will significantly enhance the widespread adoption and commercial utilization of neuromorphic detectors.}
}