High-Confidence AptapipetteIontronic Sensor for Analysisof Environmental Polystyrene Nanoplastics with Machine Learning-AssistedIonic Current Rectification

The pervasive accumulation of nanoplastics in ecosystems poses significant threats due to their bioaccumulation potential and ecotoxicity. Conventional detection methods suffer from complex pretreatment and limited on-site applicability. Here, we develop a DNA aptamer-functionalized borosilicate nanopipette, defined as aptapipette, for label-free detection of polystyrene (PS) nanoplastics by leveraging synergistic electrostatic interactions and steric hindrance effects. Stepwise modifications, including silicon nanowires/amination/aptamer, enable specific binding to PS nanoplastics, amplifying ionic current rectification (ICR) through enhanced surface charge density. The iontronic sensor achieves an ultralow detection limit down to 3.3 μg/L for aged PS nanoplastics with high selectivity and robustness. Impressively, a support vector machine-assisted method is integrated to decode ICR signals, establishing a quantitative contamination assessment model. This approach achieves 96.7% confidence in distinguishing pollution levels via principal component analysis and receiver operating characteristic curve (area under the curve = 0.998), transforming raw data into actionable environmental risk insights. This work integrates aptamer-based specific recognition, nanopipette-enabled iontronic sensing, and machine learning-assisted signal decoding, providing a promising tool for environmental nanoplastics monitoring and pollution stratification.