Xiaomin Yang,Mengjie Zhang,Anastasios Koidis,Xiaodong Liu,Chuangzhong Guo,Zhenlin Xu,Xiaoqun Wei,Hongtao Lei
Currently, the identification and quantification of complex adulteration of high-value vegetable oils are still challenging. In this study, the extreme vertex design method was adopted to design representative multivariate adulterated camellia oil samples. Thereafter, 11 characteristic lipid species were identified by considering the statistically significant difference and categorical contribution. A discriminant method and linear regression model were established based on 11 key lipids. The accuracy rate of the model was 100 %, which could correctly discriminate the camellia oil with an adulteration ratio as low as 2.5 %. The root mean square error (RMSE) of the regression equation was close to 0, and the coefficient of determination (R 2 P ) was 0.9054. This method has been successfully applied to commercially available samples for validation purposes, detecting 27.3 % of camellia oil adulteration. Overall, the results indicate that the discriminating method and content prediction model can provide a potential strategy for authentication of multivariate vegetable oils.
Dietary supplements, with nutritional and health benefits, are widely used worldwide. Ensuring the quality of dietary supplements has become particularly important as demand increases. Current traditional methods of quality assessment tend to concentrate solely on the content of certain active ingredients; nevertheless, the bioactivity of a product, which serves as a more direct indicator of its quality, is frequently overlooked. Thus, a comprehensive "Content-Extractives-Activity" quality evaluation system was developed for dietary supplements using a portable near-infrared spectrometer combined with entropy weight method in this study. Taking Eucommia ulmoides granules as the object of study, near-infrared spectroscopy (NIRS) was used to predict the pinoresinol diglucoside content, ethanol-soluble extractives content, bioactivity indicators related to antioxidant damage such as cell viability value and intracellular malondialdehyde content, as well as comprehensive index. The best models all obtained satisfactory prediction results with correlation coefficient of prediction in the range of 0.89~0.97 and residual predictive deviation in the range of 1.81~4.19, which demonstrated the feasibility of the method for the rapid evaluation of the integrated quality of dietary supplements by NIRS.
危险品化学品经营许可证（带存储）