Study On Quality Control Of Schisandra Based On Low Field Nuclear Magnetic Resonance

The quality of Chinese herbal medicine determines its efficacy,and the source of origin is the main factor affecting its quality.In addition,the moisture content,storage conditions,processing conditions and other medicinal materials have an impact on their efficacy and medicinal safety.Therefore,simple methods for identification of traditional Chinese medicines from different places are important to quality control.In this paper,we propose a simple and effective analytical method for identifying the origin of Chinese medicinal materials based on low field nuclear magnetic resonance（LF-NMR）.Firstly,Carr-Purcell-Meiboom-Gill（CPMG）data of five species of Schisandra from from LF-NMR were processed with principal component analysis（PCA）and cluster analysis（CA）.Five species of Schisandra were successfully distinguished and the total cumulative contribution rate of the model was 99.52%.Secondly,Water content and volatile oil content were determined according to the methods prescribed by the Chinese Pharmacopoeia（2015 edition）.Partial least squares regression（PLSR）models of water content and volatile oil content were established based on CPMG data with partial least squares regression method,a completely cross-validated method was used to determine the residual factor of the model for moisture and volatile oil content of Schisandra.In the moisture prediction model,the residual factor number is 2,correction set correlation coefficient R²_cal and cross validation correlation coefficient R²_cv were 0.9939 and0.9485,the root mean square error of calibration（RMSEC）is 2.998%;the root mean square error of cross-validation（RMSECV）is 9.947%.In the prediction model of volatile oil content,the residual factor number is 3,R²_cal and R_cv² are 0.9993and 0.9944,RMSEC and RMSECV are0.2300%and 0.7517%,respectively.The prediction model was verified by the unknown Schisandra.Relative error of water content prediction was 4.115%²0.08%,and relative error of volatile oil content prediction was 0.8929%¹9.55%.These results indicate that the models have higher predictive power.Our research have shown that it is feasible to identify origin places of same kind Chinese medicine by the application of LF-NMR combined with chemometrics.This study proposes a method based on low field nuclear magnetic resonance（LF-NMR）and chemometrics to identify the source of Chinese medicinal materials and identify their water and volatile oil content.