基于拉曼光谱的不同产地黄芪鉴别研究

摘要

Objective:To discriminate Huangqi ( Radix Astragali seu Hedysari) from different producing areas using Raman spectrosco-py.Methods:The Raman spectroscopy of 100 batches of Huangqi from four producing areas of Inner Mongoli, Shanxi, Heilongjiang, and Gansu were measured and analyzed by principal component analysis ( PCA) method to discriminate between samples from different producing areas.The discriminant model of producing areas based on partial least squares analysis ( PLS) was establish and verified. Results:The differences between Raman spectra of Huangqi samples from five producing areas mainly displayed in the peak intensity of the spectrum, indicating that the biochemical substances contained in Huangqi from different areas have a significant difference in the composition ratio.The PCA could distinguish samples from different producing areas.In the 3-D scatter plot with PC1, PC2 and PC3 score as the three axes, five kinds of representative symbols for four different producing areas were effectively classified, without any o-verlap.Moreover, PLS discriminant model can correctly identify the producing areas of the Huangqi samples from Inner Mongoli, Hei-longjiang, and Gansu.Conclusion:The results demonstrate that the combination of Raman spectroscopy technology, PCA and PLS can successfully identify origins of Huangqi, showing great potential of this method in the origin identification of other medicinal herbs.%目的：利用拉曼光谱技术建立黄芪饮片产地鉴别方法。方法：检测来自内蒙古、山西、黑龙江、甘肃四个地区共100批次黄芪饮片的拉曼光谱，通过主成分分析（ PCA）对不同产地样本进行判别区分，利用偏最小二乘判别法（ PLS）建立产地判别模型并予以验证。结果：四个产地黄芪饮片拉曼谱线差异主要体现在谱峰强度方面，表明不同产地的黄芪饮片所含生化物质在组成比例上有较大差异。 PCA分析可对不同产地的黄芪饮片进行区分，主成分得分PC1、PC2、PC3构成的三维散点图中，代表不同产地的数据散点各自聚集，没有重叠。 PLS产地判别模型可正确鉴别内蒙古、山西、黑龙江、甘肃等四产地黄芪饮片。结论：本研究表明拉曼光谱技术结合PCA及PLS方法，可实现黄芪饮片产地的快速鉴别。同时该方法还可拓展应用于其他饮片产地鉴别，具有良好应用前景。