An artificial neural network model based on CDs@MOF/COF composites for the ultra-sensitive fluorescence detection of glyphosate
Glyphosate (Gly) overuse threatens public health and ecosystems, necessitating highly sensitive detection. This work developed a novel ratiometric fluorescent probe, CDs@Cu-MOF/TMTA-COF, by integrating carbon dots (CDs), metal organic frameworks (MOFs), and covalent organic frameworks (COFs). We encapsulated CDs within Cu-MOF in-situ grown onto sp 2 ‑carbon conjugated TMTA-COF. Gly binding to Cu 2+ causes MOF structural collapse, releasing CDs and enhancing its fluorescence at 425 nm. Simultaneously, Gly inhibits competitive absorption and acceptor-photoinduced electron transfer process between TMTA-COF and Cu-MOF, increasing COF fluorescence at 550 nm. This dual-signal change enables ratiometric Gly detection from 0.2 to 40 μM, with a detection limit of 0.07 μM. This probe was successfully used to measure Gly in soil, Yellow River water and soybean (recoveries: 95.3–104.1 %). Additionally, the back propagation neural network further enables fast detection of Gly. This work demonstrates the potential of MOF/COF hybrids for fluorescence sensing for pesticide analysis and paves the way for intelligent sensing platforms.
危险品化学品经营许可证（带存储）