A simply designed quasi-ratiometric fluorescence probe for the visual and on-site detection of levofloxacin (LVF) residues in milk and fish sample.

An europium metal organic framework (Eu-DBPA-Phen) was synthesized using 2,5-dibromoterephthalic acid (H₂DBPA) and 1-10-phenanthroline (Phen) as ligands. A straightforwardc quasi-ratiometric fluorescence probe was then developed for the detection of levofloxacin (LVF) by the simplistic combination of red-emitting Eu-DBPA-Phen and the inherent blue auto-fluorescence of the target. The probe exhibits the advantages of wide linear range (0.01-50 and 50-175 μM), good selectivity, low detection limit (4.53 nM) and fast response time. In addition, a smartphone-assisted fluorescent test strip analysis platform was established for the visual and on-site detection of LVF in milk and fish samples, achieving satisfactory recovery rates ranging from 101.7 % to 103.4 % and low standard deviations (RSD ≤2.890, n = 3). Furthermore, a hierarchical clustering algorithm integrating machine learning and smartphone-test strip platform was devised to streamline the detection process. The proposed intelligent detection platform introduces a novel approach for LVF detection, thereby enhancing food safety and human health.