EEFM combined with N-PLS method for the quantitative determination of psoralen in cosmetics

A sample and effective approach for quantitative determination of the prohibited component named psoralen in cosmetics was proposed on the basis of emission-excitation fluorescence matrix (EEFM) coupled with N-PLS method. The standard samples were divided into the calibration set and prediction set; the N-PLS model was established on the calibration set and the predictive ability of the established model was validated by prediction set. In this regard, some statistical parameters and figures of merit, such as correlation coefficient of calibration (Rc2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{{\text{c}}}^{2}$$\end{document}) and prediction (Rp2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{{\text{p}}}^{2}$$\end{document}), root mean square error of calibration (RMSEC) and prediction (RMSEP), and recovery were investigated to evaluate the performance of proposed method. The N-PLS model developed on the EEFM showed Rc2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{c}^{2}$$\end{document} and Rp2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{{\text{p}}}^{2}$$\end{document} of 0.9951 and 0.9982, respectively, and showed RMSEC and RMSEP of 1.5636 and 2.7618, respectively, demonstrating the established model has good linearity and predictive ability. Moreover, the recovery rates in the range of 105.40–123.61% demonstrated that the proposed strategy has good accuracy. Finally, the established model applied to the real cosmetic samples showed that the psoralen was not detected, and the obtained results were in accordance with those of the validated gas chromatography method. The satisfactory results indicated that EEFM combined with the N-PLS method can be employed as a reliable tool for quality monitoring of cosmetic products.