近红外光谱技术快速测定杞菊地黄丸的水分含量

目的应用近红外光谱技术建立一种杞菊地黄丸(浓缩丸)中水分含量的快速测定方法。方法以甲苯法测定的样品中水分的含量为真实值,运用近红外漫反射光谱技术采集96份杞菊地黄丸(浓缩丸)样品的近红外漫反射光谱,结合偏最小二乘法(PLS)建立水分含量的定量分析模型。结果所建水分校正模型的相关系数(R2)和内部交叉验证均方差(RMSECV)分别为0.988 09和0.0587;经外部验证,模型的预测相关系数(r2)和预测均方差(RMSEP)分别为0.9969和0.075 2。结论该方法操作简便,无污染,结果准确可靠,可用于杞菊地黄丸(浓缩丸)中水分含量的快速测定。

Rapid Determination of Moisture in Qiju Dihuang Pills by Near-infrared Spectroscopy Technique

OBJECTIVE To establish a rapid analytical method for moisture in Qiju Dihuang Pills (condensed) by near-infrared diffuse reflectance technique. METHODS First, toluene method was used to determine the content of water in samples. At the same time, NIR diffuse reflectance spectra of 96 samples were collected. Then, Partial Least-Squares (PLS) method was used to create a quantitative model of calibration. RESULTS The correlation coefficients of the PLS calibration models of NIR spectroscopy for moisture was 0.988 09, the root-mean-square error of cross-validation (RMSECV) was 0.058 7; the correlation between the reference value of the testing sample set and the near infrared predictive value were 0.996 9; the root-mean-square error of prediction (RMSEP) was 0.075 2. CONCLUSION The method is rapid, nondestructive and accurate, which can be applied to the fast determination of moisture in Qiju Dihuang Pills (condensed).