人工神经网络优化厚朴提取工艺及其“发汗”前后的含量测定

Optimization of Extraction Technology of Magnolia officinalis by Artificial Neural Network and Determination of Its Content before And after Sweating

目的:优化厚朴提取工艺,从厚朴提取物的化学成分变化,阐释厚朴"发汗"的必要性。方法:采用正交设计结合人工神经网络模型的方法,以提取物干膏得率、厚朴酚、和厚朴酚三个指标的综合评分为评价指标,对厚朴提取时的溶剂浓度、料液比、提取时间和提取次数进行优化,确定最佳提取工艺;采用HPLC法对"发汗"及"未发汗"厚朴提取物进行含量测定,对比其"发汗"前后的厚朴酚与和厚朴酚的含量变化。结果:筛选的最佳提取工艺为加入厚朴样品100 g,加入70%乙醇提取,料液比为1∶8,提取时间为90 min,提取次数为2次;"发汗"后厚朴的厚朴酚提取率提高45.04%,和厚朴酚含量提高32.27%。结论:正交设计结合BP人工神经网络模型的方法稳定可行,具有良好重复性;"发汗"能增加厚朴中厚朴酚及和厚朴酚的提取率,从化学成分角度阐释了厚朴"发汗"的科学性与必要性。

Objective:To optimize the extraction process of Magnolia officinalis and explain the necessity of sweating from the changes of chemical constituents in the extracts of Magnolia officinalis.Methods:Orthogonal design combined with artificial neural network model was used to optimize the solvent concentration,material-liquid ratio,extraction time and extraction times of Magnolia officinalis with the comprehendscore of dry extract yield,magnolol and honokiol as evaluation indexes.HPLC method was used to determine the extractcontent of"sweating"and"non sweating"Magnolia officinalis.The changes of magnolol and honokiol before and after"sweating"were compared.Results:The optimum extraction process was to add 100 g of Magnolia officinalis and 70%ethanol for extraction.The ratio of material to liquid was 1:8,the extraction time was 90 minutes,and the extraction times were 2 times.Conclusion:The method of orthogonal design combined with BP artificial neural network model is stable and feasible,and has good repeatability."Sweating"can increase the extraction rate of magnolol and honokiol in Magnolia officinalis.The scientific and necessary of"sweating"of Magnolia officinalis was explained from the chemical constituents of the extracts.