Construction of NH₂-MIL-125@ZnIn₂S₄ Z-Scheme Heterojunctions for Deep-Learning Assisted Photoelectrochemical Sensing of Dopamine

Abstract

Dopamine (DA) plays a pivotal role in modulating various physiological systems. Therefore, the ultrasensitive detection of DA holds substantial importance for the diagnosis and treatment of neurological disorders. Herein, a deep-learning-assisted smart PEC biosensor is designed by synthesizing an NH₂-MIL-125@ZnIn₂S₄ Z-scheme (NH₂-MIL-125@ZIS) heterostructure as a photocathode. The heterojunction, integrating a titanium-based metal-organic framework (NH₂-MIL-125) with ZIS nanosheets, achieves optimized band alignment and interfacial charge transfer, which significantly improves the electron-hole separation and yields a 2.05-fold photocurrent than that of pristine NH₂-MIL-125. In the ultrasensitive detection of DA, the biosensor demonstrated a broad linear detection range (0.01 μM-1 mM) and an ultralow detection limit of 2.75 × 10^-9 M (S/N = 3). With the assistance of deep learning (DL), an artificial neural network was employed to achieve an intelligent analysis of DA with outstanding predictive ability (R² = 0.9851). This work transcends conventional PEC biosensing by integrating advanced materials engineering with artificial intelligence, establishing a new paradigm for high-performance, intelligent neurotransmitter monitoring.