近红外光谱法监测甜菊糖苷生产过程液中乙醇浓度的应用研究

Application of near infrared spectroscopy in monitoring ethanol concentration in stevia glycoside production

建立近红外光谱法快速检测甜菊糖苷生产过程液中乙醇浓度的方法。首先采用配置的方式获得系列浓度的甜菊糖生产过程液乙醇浓度,再用近红外智能透射方式对不同浓度的甜菊糖苷生产过程液进行扫描。以理论乙醇浓度作为化学值,采用偏最小二乘法建立乙醇浓度模型,对比了不同的光谱预处理方法对模型的影响。建立的乙醇浓度模型交叉验证值(RMSECV)和残差分析误差(RPD)分别为0.413和2.6。方法检测绝对误差在1.2%以内,检测精密度在2%以内,检测用时5min以内。方法的建立解决了生产过程液中乙醇浓度无法准确检测的难题,为甜菊糖生产工艺控制提供技术手段,提高控制水平,对植物提取行业生产过程溶剂浓度的准确检测也极具借鉴意义。

A method of rapid determination of ethanol concentration in stevia glycoside production by near infrared spectroscopy was established. First,the ethanol concentration of stevioside production process was obtained by configuration,and then the liquid stevioside product was scanned by near infrared intelligent transmission method.Ethanol concentration model was established by partial least square method with theoretical ethanol concentration as chemical value,and effects of different pretreatment methods on the model were compared. The cross-validation value(RMSECV)and residual analysis error(RPD)of the established ethanol concentration model were 0.413 and 2.6,respectively. The method was simple,rapid and accurate. The method solved the problem that ethanol concentration in the production process could not be accurately detected. The method provided technical method for the control of stevia production process as well as the control level,and provided reference for the accurate detection of solvent concentration in the production process of plant extraction industry.