Soares Patricia Kaori, Scarminio Ieda Spacino
Departamento de Química, Universidade Estadual de Londrina, Londrina, 86051-990 Paraná, Brazil.
Phytochem Anal. 2008 Jan-Feb;19(1):78-85. doi: 10.1002/pca.1020.
Multivariate analysis and statistical mixture designs were used for chromatographic fingerprint preparation and authentication of the plant material of three species of the genus Bauhinia. The extracts were analysed by reversed-phase high-performance liquid chromatography. Mixture design gave an optimum solvent composition for extracting components from the plants of 36% dichloromethane, 17% ethanol and 47% ethyl acetate (by volume), while an optimum mobile phase for chromatographic analyses was found to be 27% methanol, 27% acetonitrile and 46% of water (by volume). Results from principal component analysis, hierarchical analysis and soft independent modelling by class analogy showed that Bauhinia candicans cannot be synonymous with B. forficata Link. It was also possible to trace the metabolic profile without identifying its chemical constituents and to determine a chromatographic discriminating region. The characteristics responsible for discrimination between B. candicans and B. forficata were more polar substances that presented peaks with retention times around 1.65 and 1.81 min.
采用多变量分析和统计混合设计对三种羊蹄甲属植物的植物材料进行色谱指纹图谱制备和鉴定。提取物通过反相高效液相色谱进行分析。混合设计得出从植物中提取成分的最佳溶剂组成是36%二氯甲烷、17%乙醇和47%乙酸乙酯（体积比），而色谱分析的最佳流动相为27%甲醇、27%乙腈和46%水（体积比）。主成分分析、层次分析和类相关软独立建模的结果表明，白花羊蹄甲与紫羊蹄甲并非同义词。在不鉴定其化学成分的情况下追踪代谢谱并确定色谱鉴别区域也是可行的。白花羊蹄甲和紫羊蹄甲之间鉴别所依据的特征是极性更强的物质，其峰的保留时间约为1.65和1.81分钟。
