Full vision adaptation in mixed-light conditions enabled by dynamic water adsorption/desorption

Article https://doi.org/10.1038/s41467-026-73217-7 Full vision adaptation in mixed-light conditions enabled by dynamic water adsorption/desorption Received: 13 January 2026 Check for updates 1234567890():,; 1234567890():,; Accepted: 5 May 2026 Jia Zhu1,2,3 , Wantao Liu1, Wanxin Huang4,5, Xiangjie Chen1, Xuewei Feng 6, Xin Luo 7,8, Kai Xu 9, Min Gao 1, Haifeng Ling 5, Chaoyun Song 9 , Huanyu Cheng 10 & Yuan Lin 1,11 Mimicking the human eye’s ability to autonomously adapt to diverse and mixed illumination conditions remains a fundamental challenge in artificial vision systems. Although substantial progress has been made in materials and device engineering, current adaptive vision architectures still depend heavily on complex circuitry or algorithms and are typically restricted to uniform illumination owing to the strong intensity-dependence of photosensitivity. Here, this work presents a highly adaptive TiO₂/PEDOT:PSS photomemristor that leverages the tunable conductivity of PEDOT:PSS together with the optoelectronic response of TiO₂. The photothermal effect dynamically modulates the water absorption/desorption equilibrium in PEDOT:PSS, enabling reversible suppression or enhancement of photosensitivity under bright or dim illumination, respectively. By combining with artificial neural networks (ANNs), the artificial vision system based on TiO₂/PEDOT:PSS photomemristor arrays achieves a high accuracy of 91.3% in image recognition under mixedlight conditions—without the need for complex circuitry or algorithms. This work may establish a new approach for designing autonomous, efficient, and high-performance neuromorphic vision systems to advance the development of autonomous driving and humanoid robots. Visual perception is one of the most important features of biological systems, including humans. Studies have revealed that over 80% of external environmental information is acquired through the visual system1,2. Designing artificial vision systems with similar features to biological systems is highly demanding for cutting-edge applications, including intelligent manufacturing, autonomous driving, and medical diagnostics. Earlier research on photodetectors has primarily focused on enhancing static performance metrics such as sensitivity and operational wavelengths, attempting to approach or even exceed those of human vision3,4. Owing to its superior sensitivity and dynamic 1 School of Materials and Energy, University of Electronic Science and Technology of China, Chengdu, China. 2Yangtze Delta Region Institute (Quzhou), University of Electronics Science and Technology of China, Quzhou, China. 3 Energy and Information Materials Key Laboratory of Sichuan Province, University of Electronic Science and Technology of China, Chengdu, China. 4School of Automation and Engineering, University of Electronic Science and Technology of China, Chengdu, China. 5Nanjing University of Posts & Telecommunications (NJUPT), Nanjing, China. 6School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai, China. 7Guangdong Provincial Key Laboratory of Magnetoelectric Physics and Devices, School of Physics, Sun Yat-sen University, Guangzhou, China. 8State Key Laboratory of Optoelectronic Materials and Technologies, School of Physics, Sun Yat-sen University, Guangzhou, China. 9 Department of Engineering, King’s College London, London, UK. 10Department of Engineering Science and Mechanics, The Pennsylvania State University, University Park, Pennsylvania, USA. 11Medico-Engineering Cooperation on Applied Medicine Research Center, University of Electronics Science and Teche-mail: ; ; ; nology of China, Chengdu, China. Nature Communications | (2026)17:4965 1 Article adaptability, the human eye can process visual information under diverse illumination conditions, exhibiting a photodetection dynamic range exceeding 160 dB5,6. This capability allows humans to effectively respond to various illumination scenarios, such as watching a movie in a dark theater or driving in the glare of oncoming headlights7. However, conventional electronic photodetectors largely lack this fundamental capability of biological vision—the dynamic self-adaptation to complex and varying environments8–10. Until very recently, adaptation has been recognized as one of the key aspects in the design of artificial vision systems11. However, conventional photodetectors in previous literature reports have fixed photosensitivity and lack the capacity of adapting to different illumination conditions, resulting in image distortion and degraded recognition accuracy under fluctuating lighting conditions12. Neuromorphic behavior has been widely recognized as a key route to achieving light or visual adaptation13. Tremendous progress has been made in the design and fabrication of neuromorphic devices, including two-terminal diodes or memristors and three-terminal fieldeffect transistors (FETs) or organic electrochemical transistors (OECTs)14–18, which primarily exploit the hysteresis behavior of semiconductors14. For instance, a MoS₂-based phototransistor has demonstrated tunable photosensitivity by introducing charge trap states to regulate the photocurrent. However, the adaptation of the MoS₂-based phototransistor requires external bias modulation, fundamentally restricting its feasibility in practical implementations. Even though adaptation in a P3HT/MAPbI₃ heterostructure has been achieved by leveraging the antagonistic interplay between a PN and Schottky junction without the need for external bias modulation, its performance remains highly sensitive to the thickness of P3HT with very low tolerance16. Furthermore, its limited tunable range of photosensitivity necessitates extra preprocessing and algorithms to achieve high recognition accuracy. Given these limitations, the development of neuromorphic vision devices with embodied adaptation to varying or even mixed illumination still remains a crucial challenge. This work presents a highly adaptive TiO₂/PEDOT:PSS-based optoelectronic synapse, manifested by a rapid adaptation time and a significant suppression of photoresponse in bright light, to enable accurate vision recognition in complex illumination environments. This superior dynamic performance is enabled by the unique inorganic/organic heterojunction architecture. The TiO2/PEDOT:PSS interface provides an ideal platform for regulating the water adsorption/desorption dynamics on PEDOT:PSS, which in turn precisely modulates the resistance and postsynaptic current of the photomemristor. Specifically, under bright light, the device exhibits swift adaptation characterized by a rapidly decaying postsynaptic current, which even decreases below the baseline dark current. This embodied adaptation, quantified by its speed and efficiency, surpasses that of previously reported vision-adaptive synaptic devices. It enables the dynamic amplification of subtle low-intensity light signals while simultaneously suppressing high-intensity backgr (...truncated)