气相色谱-质谱指纹图谱结合化学计量学方法分析不同产地艾叶挥发油的差异

Analysis of Differences of Volatile Oil in Folium Artemisiae Argyi from Different Origins by GC-MS Fingerprint Chromatograms Combined with Chemometrics

将来自10个不同产地的艾叶样品粉碎、过筛,参照《中华人民共和国药典》(2020年版)进行提取。取40 g样品,加入700 mL水和100 mL饱和氯化钠溶液,于100℃加热回流5 h。放置10 h,分取0.1 mL,用正己烷稀释至10 mL,加入1.0 g无水硫酸钠,过0.22μm滤膜,滤液按照优化的仪器工作条件测定。用气相色谱-质谱法(GC-MS)建立10批样品的指纹图谱,共鉴定出103个成分,共有成分21个。将10批艾叶挥发油的指纹图谱导入指纹图谱相似度评价系统,以来自河南省南阳市样品的指纹图谱为参照图谱,以桉油精色谱峰为参照峰进行多点校正,所得相似度结果显示,10批样品可分为3类,和后续通过对21个共有成分色谱峰进行主成分分析、偏最小二乘法-判别分析和系统聚类分析所得分类结果一致。其中,偏最小二乘法-判别分析模型拟合效果好,聚类效果较优。

Folium Artemisiae argyi samples from 10 different origins were crushed, screened and extracted according to the Pharmacopoeia of the People′s Republic of China(2020 Edition). After an aliquot(40 g) of the sample was taken, 700 mL of water and 100 mL of saturated sodium chloride solution were added. The mixture was heated at 100 ℃ for 5 h under reflux, and placed for 10 h. An aliquot(0.1 mL) was taken, diluted to 10 mL with n-hexane, and mixed with 1.0 g of anhydrous sodium sulfate. The above mixture was passed through a 0.22 μm filter membrane, and the filtrate was determined in accordance with the optimized operating conditions of the instrument. The GC-MS fingerprint chromatograms of the volatile oil in 10 batches of Folium Artemisiae argyi were established, and a total of 103 components with 21 common components were identified. The fingerprint chromatograms of the volatile oil in 10 batches of Folium Artemisiae argyi were imported into the similarity evaluation system of fingerprint chromatogram, and multi-point correction was made with the fingerprint of sample from Nanyang city in Henan province as the reference chromatogram and the chromatographic peak of eucalyptol as the reference peak. It was found by the results of similarity that the 10 batches of samples could be divided into three categories, in accordance with the subsequent classification results given by principal component analysis, partial least squares-discriminant analysis and hierarchical-cluster analysis of chromatographic peaks from 21 common components. Amoung which, the model of partial least squares-discriminant analysis had good fitting effect and the clustering effect.