Abstract: In the present study, the chemical composition of pure powders of Ziziphi spinosae Semen (ZSS) and Ziziphi mauritianae Semen (ZMS) and their mixtures in different proportions was compared by 1H nuclear magnetic resonance (1H NMR)-based metabolomic fingerprinting. Then the obtained spectral data were analyzed by multivariate analysis to develop a new method for rapid detection of ZSS adulteration. Principal component analysis (PCA) performed on the 1H NMR data revealed that only a few samples at low adulteration ratios were overlapped with ZSS. In addition, the data were also analyzed by partial least squares-discriminant analysis (PLS-DA). A clear separation was seen between ZSS and 40% adulterated samples. Then a PLS model was established based on the 1H NMR data and was applied to detect adulteration in ZSS falsified by intentional addition of different concentrations of ZMS, and a quantitative calibration model was established. The PLS regression approach provided a highly reliable and accurate prediction model. The results demonstrated that 1H NMR spectroscopy coupled with PLS is a fast and effective technique for checking adulteration in ZSS. Meanwhile, the method provides a potential tool for quality control of ZSS.

Key words: Ziziphi spinosae Semen, Ziziphi mauritianae Semen, adulterants, 1H nuclear magnetic resonance, metabolomics, partial least squares