Huang, L. ; Chen, T. B. ; Liu, M. H. ; Rao, G. F. ; Yang, H.
RISK-ASSESSMENT; PLANT MATERIALS; LIBS; PELLETS
In seeking a novel method with the ability of green analysis in monitoring toxic heavy metals residue in fresh leafy vegetables, laser-induced breakdownspectroscopy (LIBS) was applied to prove its capability in performing this work. The spectra of fresh vegetable samples polluted in the lab were collected by optimized LIBS experimental setup, and the reference concentrations of cadmium (Cd) from samples were obtained by conventional atomic absorption spectroscopy after wet digestion. The direct calibration employing intensity of single Cd line and Cd concentration exposed the weakness of this calibration method. Furthermore, the accuracy of linear calibration can be improved a little by triple Cd lines as characteristic variables, especially after the spectra were pretreated. However, it is not enough in predicting Cd in samples. Therefore, partial least-squares regression (PLSR) was utilized to enhance the robustness of quantitative analysis. The results of the PLSR model showed that the prediction accuracy of the Cd target can meet the requirement of determination in food safety. This investigation presented that LIBS is a promising and emerging method in analyzing toxic compositions in agricultural products, especially combined with suitable chemometrics. (C) 2017 Optical Society of America