A comparative study of three chemometrics methods combined with excitation-emission matrix fluorescence for quantification of the bioactive compounds aesculin and aesculetin in Cortex Fraxini

FRONTIERS IN CHEMISTRY(2022)

引用 2|浏览1
暂无评分
摘要
Cortex Fraxini is an important traditional Chinese herbal medicine with various medical functions. Aesculin and aesculetin are the main effective components of Cortex Fraxini. The fluorescence signals of the two compounds have a high degree of overlap with each other, making quantitative analysis difficult with conventional analytical methods. In the present study, different chemometrics methods, including lasso regression (LAR), interval partial least squares (iPLS), and multidimensional partial least squares (N-PLS) methods, were employed and combined with excitation-emission matrix (EEM) fluorescence for the purpose of accurate quantification of aesculin and aesculetin in Cortex Fraxini samples. The most satisfactory results were obtained by using the N-PLS method based on the EEM spectra without scatterings, with correlation coefficient of calibration and prediction values higher than 0.9972 and 0.9962, respectively, root mean squared errors for calibration and prediction values lower than 0.0304 and 0.1165, respectively, and recovery values in the range of 83.32%-104.62%. The obtained credible models indicated that the N-PLS method combined with EEM spectra has the advantages of being green, low cost, and accurate and it is a good strategy for the determination of active compounds in complex samples. To further confirm the accuracy of the obtained results, the same samples were analyzed by the recognized ultra-performance liquid chromatography method.
更多
查看译文
关键词
Cortex Fraxini, excitation-emission matrix fluorescence, second-order calibration, machine learning, quantitative analysis, UPLC
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要