星点设计-效应面法优化龙须藤的大孔树脂纯化工艺

目的：优选D-101型大孔树脂纯化龙须藤的工艺条件。方法：采用HPLC测定3',4',5',5,7-五甲氧基黄酮（黄酮1）和3',4'-亚甲二氧基-5',5,6,7-四甲氧基黄酮（黄酮2）含量,流动相乙腈（A）-水（B）梯度洗脱（0~20 min,30%~40%A;20~36 min,40%~48%A）,检测波长324 nm。以黄酮1和黄酮2洗脱率和纯度的"总评归一值"为因变量,上样液质量浓度、乙醇体积分数和用量为自变量,通过星点试验对自变量各水平进行多元线性回归和二项式拟合,利用效应面法优选纯化工艺并进行预测分析。结果：二次多项式拟合的相关系数明显优于多元线性回归方程,最佳纯化工艺参数为上样液质量浓度0.11 g·mL^-1,洗脱剂乙醇体积分数84.41%,乙醇用量13.94 BV;黄酮1和黄酮2平均洗脱率分别为78.45%（RSD 0.32%）,78.55%（RSD 0.33%）,纯度依次为1.14%（RSD 0.88%）,0.67%（RSD 1.49%）。结论：通过星点设计-效应面法优选的龙须藤大孔树脂纯化工艺稳定可行,预测性好。

Optimization of Macroporous Resin Purification Process for Bauhinia championii by Central Composite Design-Response Surface Methodology

Objective: To optimize purification technology of Bauhinia championii with D-101 macroporous resin. Method: HPLC was adopted to determine contents of 3', 4', 5', 5, 7-pentamethoxy flavone (flavone 1) and 3', 4'-methylenedioxy-5', 5, 6, 7-tetramethoxy flavone (flavone 2) with mobile phase of acetonitrile (A) -water (B) gradient elution (0-20 min, 30%-40%A;20-36 min, 40%-48%A) and detection wavelength of 324 nm.Taking overall desirability of desorption rates and purities of flavone 1 and flavone 2 as dependent variable, while the concentration of sample solution, ethanol concentration and volume as independent variables, multiple linear regression and binomial fitting of each level of independent variables was handled by central composite design, response surface methodology was adopted to optimize technological parameters. Result: Quadratic equation described relationship between dependent and independent variables was obviously better than multiple linear regression equation.Optimum technological parameters were as following:sample concentration 0.11 g·mL^-1, ethanol concentration 84.41%, eluent volume 13.94 BV;Average elution rates of flavone 1 and flavone 2 were 78.45% (RSD 0.32%) and 79.55% (RSD 0.33%), purities were 1.14% (RSD 0.88%) and 0.67% (RSD 1.49%). Conclusion: Optimized purification technology was feasible and stable with good predictability.