星点设计-响应面法优化山茱萸总苷的提取及纯化工艺

Application of response surface methodology with central composite design in optimizing the extraction and purification of cornel total iridoid glycosides

目的:采用星点设计-响应面法优化山茱萸总苷的提取及纯化工艺。方法:以乙醇浓度、液料比和提取时间为自变量,以莫诺苷、马钱苷和獐牙菜苷的收率以及提取物得率为因变量,通过对自变量的总评归一值进行多元线性回归和二项式拟合,以响应面优选最佳的超声提取工艺。以乙醇浓度、洗脱体积、上样量为自变量,以莫诺苷、马钱苷和獐牙菜苷的收率以及洗脱物得率为因变量,通过对自变量的总评归一值进行多元线性回归和二项式拟合,以响应面优选最佳的大孔树脂纯化工艺。结果:确定提取工艺条件为乙醇浓度22%～25%,液料比12～15倍,提取时间30～40 min;纯化工艺条件为乙醇浓度35%～45%,洗脱体积7～9 BV,上样量40～50 mg.g-1。2个步骤工艺验证的总评归一预测値与实测值的偏差分别为-1.8%和2.7%。结论:采用星点设计-响应面法优选山茱萸总苷提取及纯化工艺,方法简便,预测性良好。

Objective：To optimize extraction and purification techniques for cornel total iridoid glycosides response surface methodology with central composite design. Methods：Ethanol concentration, extraction time and ratio of sol-vent to raw material were selected as independent variables, the overall desirability of the yields of morroniside, loga-nin and sweroside and the extract yield were set as dependent variables. Multiple linear regression （MLR） and bi-nomial fitting were used to estimate the relationship between independent and dependent variables, response surface methodology was applied to optimize the process of ultrasound extraction. Ethanol concentration, elution volume and sampling amount were selected as independent variables, the overall desirability of the yields of morroniside,loganin and sweroside and the elution yield were set as dependent variables. MLR and binomial fitting were used to estimate the relationship between independent and dependent variables, response surface methodology was used to optimize the process of macroporous resin purification. Results：The optimal conditions of extraction process were determined as follows ：The raw material was ultrasound extracted for 30-40 min with 12-15 fold of 22%-25% aqueous eth- anol;the optimal conditions of macroporous resin purification process were determined as follows：after sampling 40-50 mg·g^-1 of macroporous resin, 35%-45% aqueous ethanol eluted the column with 7-9 BV of column volume. The deviation of the above two processes between observed and predicted values was - 1.8% and 2. 7% , respectively. Conclusion：Application of response surface methodology with central composite design in optimizing the extraction and purification of cornel total iridoid glycosides is simple and highly predictive.