基于HPLC指纹图谱和化学计量学评价白及饮片质量

目的：按照《中国药典》对市售36批次的白及饮片进行常规检验，并借助HPLC指纹图谱、聚类分析和多元统计分析等方法综合评价白及的质量现状。方法：检测白及饮片的水分、总灰分、二氧化硫残留量，通过HPLC方法建立指纹图谱，采用Thermo Hypersil Gold Aq色谱柱（4.6 mm×250 mm，5 μm），以乙腈为流动相A，0.05%磷酸水溶液为流动相B，梯度洗脱，检测波长为280 nm，将标准化后的共有峰峰面积数据分别导入SPSS和Simca-P软件进行聚类分析、偏最小二乘判别分析（PLS-DA）。结果：36批次白及饮片按标准检验全部合格，经指纹图谱确定6个共有峰，将聚类分析为两大类的样品共有峰信息进行PLS-DA分析，结果发现在得分图第一主成分上样品能够明显分为两类R2X（cum）=0.63，R2Y（cum）=0.71，Q2=0.59]，验证了聚类分析的结果，相应的载荷图显示5号峰对2组的区分贡献较大。结论：市售白及存在化学成分的差异，该聚类分析和多元统计分析相结合的方法可为白及饮片的质量控制提供参考依据。

Quality control of Bletilla striata based on chemometrics method and HPLC fingerprint

Objective: To make an routine inspection of Bletilla striata pieces decoction according to the Chinese Pharmacopoeia and to evaluate its quality by means of HPLC fingerprint, cluster analysis and multivariate statistical analysis. Methods: Chemical properties, including moisture, total ash and residues of sulfur dioxide, of the Bletilla striata pieces decoction were investigated. HPLC was used to determine the fingerprint chromatogram: Thermo Hypersil Gold Aq (4.6 nm×250 mm, 5 μm) column with gradient elution of acetonitrile and 0.05% phosphate acid, the detection wavelength was set at 340 nm, cluster analysis and partial least squares discriminant analysis (PLS-DA) were performed using SPSS and Simca-P software. Results: Six common peaks were found in the HPLC fingerprints of 36 sample batches from different habitats. Cluster analysis and PLS-DA model R2X (cum)=0.63, R2Y (cum)=0.71, Q2=0.59] showed clearly discriminated samples, with the NO.5 chromatographic peak being the major contributor to the differential profiles. Conclusion: Difference in chemical composition of Bletilla striata exists on the market and the study results can be the reference for quality control of decoction pieces of Bletilla striata.