近红外光谱测定清热解毒口服液中绿原酸与黄芩苷的波段优选

目的]运用近红外光谱测定清热解毒口服液中绿原酸与黄芩苷含量的波段并建立数学模型。方法]结合移动窗口片最小二乘(MWPLS)法对波段进行了优选,分别建立测定清热解毒口服液中绿原酸与黄芩苷含量的定量分析模型,比较了不同的光谱预处理方法的效果。结果]采用MWPLS方法对6个厂家30个样品建模选出绿原酸与黄芩苷的最优波段分别为1 254～1 306 nm和1 176～1 266nm,使用其他2个厂家10个批次样品进行检验,绿原酸和黄芩苷实测值与预测值检验均方根误差分别为1.637和0.857,相关系数分别为0.993 7和0.987 7。结论]定量模型预测精度高,稳定性较好,可用于清热解毒口服液生产过程中的质量控制。

Determination of Chlorogenic Acid and Baicalin in Qingrejiedu Oral Liquid by Near Infrared Spectroscopy and Waveband selection

Objective ] Near infrared （NIR） spectroscopy combined with waveband selection was applied for determination of chlorogenic acid and baicalin in Qingrejiedu oral liquid. Method] Combined with MWPLS, the wave band was optimized. The quantitative analysis model was established for determining chlorogenic acid and baicalin content in Qingrejiedu oral liquid. The effects of different spectral pretreatment meth- ods were compared. Resuh ] According to Qingrejiedu oral liquid from 30 of 6 factory, the optimized waveband of chlorogenic acid and baica- lin were 1 254 to 1 306 nm and 1176 to 1 266 nm, 10 batches of other 2 factories were selected as validation set, the validation root mean square error of prediction （V-SEP） and validation correlation coefficients of prediction （V-RP） for chlorogenic acid and baicalin were 1. 637, 0.993 7 and 0. 857, 0. 987 7. F Conclusion] Prediction values of the samples are accurate and reliable, and it is fit for on-line detection.