引用本文:
【打印本页】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 2764次   下载 1637 本文二维码信息
码上扫一扫!
分享到: 微信 更多
酸枣仁总黄酮提取工艺及其预测模型研究
李 游1, 靳爱仙2, 梁宗锁1
1.西北农林科技大学 生命科学学院;2.国家林业局 调查规划设计院
摘要:
【目的】 研究优化酸枣仁总黄酮的提取工艺。【方法】 采用星点设计 效应面法,对酸枣仁总黄酮的超声提取工艺进行优化。以乙醇体积分数、液(体积,mL)料(质量,g)比、提取时间为自变量,以总黄酮提取量为因变量建立预测模型,并进行了验证。【结果】 总黄酮提取量与3个影响因素不能用线性关系进行描述,采用二次多项式的拟合效果较好,复相关系数 R2=0.887 9,具有较高的可信度。优选的最佳工艺为:乙醇体积分数84%,液料比40 mL/g,提取时间35 min;最佳工艺验证结果与模型预测值相差2.30%,总黄酮提取率达到95.05%。【结论】 星点设计 效应面法可以快速、简便地优化酸枣仁总黄酮的超声提取工艺,所建模型预测性良好,工艺稳定可行。
关键词:  酸枣仁  总黄酮  提取工艺  星点设计-效应面法
DOI:
分类号:
基金项目:中国科学院院地合作项目“酸枣规范化生产及综合开发”(2007-1)
Optimization of the extraction technics of total flavones from Semen-ziziphi spinosae and its mathematical model
Abstract:
【Objective】 The extraction technics of total flavones in Semen ziziphi spinosae was optimized. 【Method】 Ultrasonic extraction process of the total flavones in Semen ziziphi spinosae was optimized with central composite design-response surface methodology.Linear or nonlinear mathematic models were used to evaluate the relationship between the independent variable (concentration of ethanol,liquid-solid ratio,and extraction time) and the dependent variable (extraction quantity).Optimum process conditions were predicted with response surface methodology,and then validated.【Result】 Second-order quadratic model was more suitable than linear model for evaluation of the relationship between independent variable and dependent variable,and it's regression coefficient was as high as 0.887 9.The optimum process conditions were obtained as follows:84% ethanol,liquid solid ratio at 40 mL/g and extracting for 35 min.The deviation between observed and predicted values of extraction quantity was 2.30%,the extraction rate up to 95.05%.【Conclusion】 Central composite design-response surface methodology can be used to optimize the extraction process of the total flavones in Semen ziziphi spinosae and optimized process is reliable.
Key words:  Semen ziziphi spinosae  total flavone  extraction technique  central composite design response surface methodology