基于近红外光谱技术对硫磺熏蒸白芷的定性鉴别和定量研究

白芷中的香豆素类成分在硫磺熏蒸后显著下降，因此需要对硫磺熏蒸的白芷建立快速定性鉴别，并建立欧前胡素含量定量模型，实现药材质量控制的目的。该文使用近红外（NIR）漫反射光谱技术采集硫磺熏蒸和非硫磺熏蒸的白芷的NIR光谱，以一阶导数和矢量归一化的方法进行预处理，采用OPUS/INDENT定性分析软件Ward’s Algorithm法进行聚类分析，在8 806.0~3 811.0 cm^-1的聚类分析结果可以简单，快速地鉴别白芷是否被硫磺熏蒸。同时，以高效液相色谱法测定的欧前胡素分析值为参照，结合OPUS/QUANT定量分析软件偏最小二乘法，建立欧前胡素含量的定量模型。该模型的相关系数（R²），内部交叉验证均方差（RMSECV）和预测均方差（RMSEP）分别为0.982 8，0.006 8和0.011 8。建立的欧前胡素NIR定量模型可准确地用于快速测定白芷中欧前胡素的含量。

Qualitative and quantitative research on sulfur fumigation of Angelicae Dahuricae Radix (Baizhi) by near-infrared spectroscopy

The contents of coumarins in the sulfur fumigated Angelicae Dahuricae Radix (Baizhi, ADR) were reduced significantly. To achieve the quality control of ADR, the qualitative identification of sulfur fumigated ADR and quantitative model of imperatorin content should be established. The near-infrared (NIR) spectrograms of non-sulfur and sulfur fumigated ADR were collected by NIR diffuse reflectance spectroscopy technology and pretreated by the method of first derivative derivation and vector normalization. The Ward's Algorithm method was used for the cluster analysis. The non-sulfur and sulfur fumigated ADR can be quickly identified in the range of 8 806.0-3 811.0 cm^-1 based on the cluster analysis. The NIR quantitative model of imperatorin was established by the contents of imperatorin determined by HPLC in combination with partial least squares regression analysis. According to the calibration model established in this study, correlation coefficients (R²), the root-mean-square error of cross-validation (RMSECV), and the root-mean-square error of prediction (RMSEP) for imperatorin were 0.982 8, 0.006 8, 0.011 8, respectively. The quantitative model of imperatorin can be applied to determine the content of imperatorin in ADR accurately.