Geographical origin identification and quantitative calibration analysis of Polygonum perfoliatum L. by near infrared spectroscopy

Rapid near infrared (NIR) spectroscopic analysis models were developed to distinguish the geographical origin and determine the content of Polygonum perfoliatum L. The spectra of 94 samples of differing origin were recorded using two NIR instruments (Antaris II and Nicolet 6700). The spectral pretreatment methods used were the first-derivative (FD) and the standard normal variate correction (SNV). Distance match (DM) pattern was applied to classify the geographical origins of the samples, which had better predictive ability (>81%). For the quantitative calibration, the polyphenol and flavone models were established with a partial least squares (PLS) algorithm. The established models were evaluated with root mean square error of prediction (RMSEP) and correlation coefficients. The correlation coefficients of both calibration models were above 0.910, indicating satisfactory predictive abilities. This study demonstrated that NIR diffuse reflection spectroscopy can be used for rapid discrimination of geographical origin and content determination of P. perfoliatum L., and is a beneficial quality control method for this raw material.