Quantitative Analysis of Panax ginseng by FT-NIR Spectroscopy

Abstract

Near-infrared spectroscopy (NIRS), a rapid and efficient tool, was used to determine the total amount of nine ginsenosides in Panax ginseng. In the study, the regression models were established using multivariate regression methods with the results from conventional chemical analytical methods as reference values. The multivariate regression methods, partial least squares regression (PLSR) and principal component regression (PCR), were discussed and the PLSR was more suitable. Multiplicative scatter correction (MSC), second derivative, and Savitzky-Golay smoothing were utilized together for the spectral preprocessing. When evaluating the final model, factors such as correlation coefficient (R (2)) and the root mean square error of prediction (RMSEP) were considered. The final optimal results of PLSR model showed that root mean square error of prediction (RMSEP) and correlation coefficients (R (2)) in the calibration set were 0.159 and 0.963, respectively. The results demonstrated that the NIRS as a new method can be applied to the quality control of Ginseng Radix et Rhizoma.

Figure 1

Chemical structures of the nine ginsenosides.

Figure 2

NIR spectra of ginseng samples obtained from original data (a) and MSC processing (b).

Figure 3

NIR spectra for ginseng samples processed by 1st derivative + SG + MSC (a) and MSC + SG + 2nd derivative (b).

Figure 4

Correlation diagram between the NIR model calculated values and the actual values of the total amount of nine ginsenosides.