UV-Responsive GaO x /ZnMgO QDs Heterojunction Optoelectronic Synapse for Self-Denoising Image Processing

作者：Zilong Guo, Z.W. Li, Q. W. Liu, ChunWei Zhang, J. H. Zhang, Hao Kan, Yang Li · 发表于：IEEE Transactions on Electron Devices · 年份：2026 · DOI：10.1109/ted.2025.3649620 · 被引用次数：1 · 研究领域：Advanced Memory and Neural Computing、Neural Networks and Reservoir Computing、Ferroelectric and Negative Capacitance Devices

In the fields of industrial manufacturing, biomedical diagnostics, and environmental monitoring, computing systems with von Neumann architecture suffer from inefficiencies due to the separation of UV sensing, memory, and processing, leading to increased complexity and hardware overhead. In-sensor computing utilizing UV-responsive optoelectronic neuromorphic devices offers a promising alternative by bypassing these limitations for efficient, low-power UV information processing. Here, we present a UV-responsive optoelectronic synapse based on a GaOx/ZnMgO quantum dots (QDs) heterojunction. This device selectively detects UV light and intrinsically suppresses noise, enabling integrated clear perception and fast computation. Various biological synaptic behaviors are demonstrated on the device under 300 nm UV illumination, including short-term memory (STM) to long-term memory (LTM) transition, learning–forgetting–relearning cycles, and dark adaptation behaviors. Leveraging synaptic features, a UV-driven neuromorphic vision system built from the device achieved 97% recognition accuracy on handwritten digits even in the presence of RGB noise. These results highlight the potential of GaOx/ZnMgO QDs heterojunction synapses for robust, noise-resilient in-sensor computing and high-fidelity UV image processing.