The presence of excessive heavy metal content in food poses potential risks to human health.
This paper presents the fabrication of a novel type of Carbon Dots (CDs) using green tea residue as a carbon source, which can be applied for simultaneous detection of Hg2+, Pb2+, Fe3+, and Cu22+ through particle swarm optimization (PSO) based optimized variable-weighted least-squares support vector machine (VWLS-SVM) model and the partial least squares discriminant analysis (PLSDA) method.
The utilization of PSO-VWLS-SVM model discovered and verified two fluorescence enhancement markers of CDs, namely isoquercitrin and 5-methyl furfural in green tea residues. By employing PLSDA, simultaneous qualitative and quantitative determination of these four metal ions was achieved. These CDs are capable of detecting four types of metal ions at low concentrations even when there are high concentrations of other metal ions and amino acids. More importantly, the CDs were successfully applied for the detection of Hg2+, Pb2+, Fe3+, and Cu2+ in real food samples. The recovery rates of four metal ions spiked into five different matrices were found to be the range of 99.1–101.3%, while both intra-day and inter-day relative standard deviations remained below 0.5% for all samples. This study on chemometrics-assisted exploration into formation mechanisms of endogenous CDs provides theoretical guidance for enhancing their fluorescence properties and expanding their application in heavy metal detection in food.