Abstract. The feasibility of using Fourier transform infrared (FT-IR) spectroscopy combined with chemometrics to determine the ß-carotene and lutein contents in green tea was investigated in this study. The relationship between pigment contents and spectral responses was explored by partial least squares (PLS), least squares support vector machine (LS-SVM), and extreme learning machine (ELM) methods. Next, 30 and 29 effective wavenumbers (EWs) for ß-carotene and lutein, respectively, were selected according to the weighted regression coefficients of the PLS regression models, and simplified determinant models were built on the extracted EWs. The ELM models based on the EWs obtained the best results, with correlation coefficients of calibration (rc) and prediction (rp), and residual prediction deviation (RPD) of 0.977, 0.946, and 2.84, respectively, for ß-carotene and 0.975, 0.937, and 2.88, respectively, for lutein. The overall results indicate that FT-IR spectroscopy combined with chemometrics could be a rapid and accurate alternative method for determining carotenoid pigments in green tea. Keywords: ß-carotene, Chemometrics, Fourier transform infrared spectroscopy, Green tea, Lutein.
Analysis of ‘legal high’ substances and common adulterants using handheld spectroscopic techniques