Sun Xiaodong, Zhang Min, Wang Pengjiao, Chen Junhua, Yang Shengjun, Luo Peng, Gao Xiuli
State Key Laboratory of Functions and Applications of Medicinal Plants, School of Pharmacy, Guizhou Medical University, Guiyang 550025, China.
Microbiology and Biochemical Pharmaceutical Engineering Research Center of Guizhou Provincial Department of Education, Guizhou Medical University, Guiyang 550004, China.
Foods. 2022 Aug 8;11(15):2376. doi: 10.3390/foods11152376.
Paprika is a widely consumed spice in the world and its authentication has gained interest considering the increase in adulteration cases in recent years. In this study, second-order fingerprints acquired by liquid chromatography with fluorescence detection (HPLC-FLD) were first used to detect and quantify adulteration levels of Chinese paprika samples. Six different adulteration cases, involving paprika production region, cultivar, or both, were investigated by pairs. Two strategies were employed to reduce the data matrices: (1) chromatographic fingerprints collected at specific wavelengths and (2) fusion of the mean data profiles in both spectral and time dimensions. Afterward, the fingerprint data with different data orders were analyzed using partial least squares (PLS) and n-way partial least squares (N-PLS) regression models, respectively. For most adulteration cases, N-PLS based on second-order fingerprints provided the overall best quantitation results with cross-validation and prediction errors lower than 2.27% and 20.28%, respectively, for external validation sets with 15-85% adulteration levels. To conclude, second-order HPLC-FLD fingerprints coupled with chemometrics can be a promising screening technique to assess paprika quality and authenticity in the control and prevention of food frauds.
辣椒粉是全球广泛食用的一种香料，鉴于近年来掺假案例的增加，其鉴定受到了关注。在本研究中，首次使用液相色谱 - 荧光检测（HPLC - FLD）获得的二阶指纹图谱来检测和量化中国辣椒粉样品的掺假水平。研究了六种不同的掺假情况，涉及辣椒粉的产地、品种或两者，通过两两组合进行研究。采用了两种策略来减少数据矩阵：（1）在特定波长下收集的色谱指纹图谱；（2）光谱和时间维度上平均数据轮廓的融合。之后，分别使用偏最小二乘法（PLS）和n - 路偏最小二乘法（N - PLS）回归模型对不同数据阶次的指纹数据进行分析。对于大多数掺假情况，基于二阶指纹图谱的N - PLS提供了总体最佳的定量结果，对于掺假水平为15 - 85%的外部验证集，交叉验证和预测误差分别低于2.27%和20.28%。总之，二阶HPLC - FLD指纹图谱结合化学计量学可以成为一种有前景的筛选技术，用于在控制和预防食品欺诈中评估辣椒粉的质量和真实性。
