Wu Qing-Qing, Li Zheng, Chen Miao-Hua, Yuan Cheng, Zhu Yuan-Cheng, Xu Jing-Juan, Zhao Wei-Wei
State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing, 210023, P. R. China.
Adv Mater. 2025 Jun 25:e2504338. doi: 10.1002/adma.202504338.
Close imitation of synaptic metaplasticity is an important objective in the neuromorphic domain. Progress has been made in solid-state electronics with high-voltage dynamics, which, nevertheless, marks a significant inconsistency with the biological systems in aqueous media. Here, the concept of reticular photoelectrochemical transistor (RPECT) is proposed and devised that can realize metaplasticity with biochemical modulation. Based on the ambipolar behavior of a metal-organic framework channel gated by a photosensitive hydrogen-bonded organic framework electrode, biochemically modulated positive/negative photoconductivity and metaplasticity with the typical features, e.g., the nonmonotonic enhanced depression effect region and the threshold sliding are achieved. Taking advantage of such unique properties, in-sensor preprocessing and in-memory computing are further implemented for efficient image recognition. This work realizes the aqueous metaplasticity by a new device of RPECT, which also introduces the biochemical modulation into image recognition, providing a perspective for future development of machine vision processing.
紧密模拟突触可塑性是神经形态领域的一个重要目标。固态电子学在高电压动力学方面取得了进展，然而，这与水性介质中的生物系统存在显著的不一致。在此，提出并设计了网状光电化学晶体管（RPECT）的概念，其能够通过生化调制实现可塑性。基于由光敏氢键有机框架电极门控的金属有机框架通道的双极性行为，实现了生化调制的正/负光电导率以及具有典型特征（例如非单调增强的抑制效应区域和阈值滑动）的可塑性。利用这些独特特性，进一步实现了传感器内预处理和内存内计算以进行高效图像识别。这项工作通过新型RPECT器件实现了水性可塑性，还将生化调制引入图像识别，为机器视觉处理的未来发展提供了一个视角。
