ABSTRACT
Artificial intelligence (AI)-integrated smartphone-based handheld determination platform, based on 3D printed accessory, Al3+-triggered aggregation-induced red-emssion enhanced carbon dots (CDs) test strip, and smartphone with self-developed YOLO v3 AI algorithm-based application, proves the feasibility for intelligent real-time on-site quantitation of F- through tracking a consecutive fluorescence (FL) colour change. CDs, manifesting dual emission of moderate green emission at 512 nm and weak red one at 620 nm under 365 nm excitation, were synthesized hydrothermally from alizarin carmine and citric acid. CDs@Al3+, with distinct aggregation-induced red-emssion enhancement and green-emssion quenchment, were prepared by adding Al3+ to the CDs solution. Inspiringly, due to intrinsic ratiometric FL variation (I620/I512), CDs@Al3+ engender a successive FL colour variation from red to green in response to different concentrations of F- with low limit of detection of 7.998 µM and wide linear range of 150-1200 µM based on excellent linearity correlation between R/G value and F- concentration. Furthermore, F- content in tap water, toothpaste and milk could be intelligently, speedily, and straightforwardly analyzed through the AI-integrated smartphone-based handheld detection platform. It is fervently desired that our study will motivate a brand-new perspective for the promotion of efficacious detection strategy and the extension of practical application promise.