论文部分内容阅读
目的:应用拉曼光谱技术对4种含碳酸钙(CaCO_3)的矿物类中药粉末进行快速鉴别。方法:收集不同来源的鱼脑石、鹅管石、花蕊石、南寒水石药材,共计共64批样品,粉碎为粉末,随机采集每种药材中3批粉末的X射线衍射(XRD)图谱和全部样品的粉末拉曼图谱。在用XRD图谱解析4种药材的物相组成的基础上,解析各药材的拉曼光谱特征。选取谱段200-300、350-400、670-700cm-1建立聚类分析模型(模型1);然后对主要含文石的矿物药建立偏最小二乘分析模型(模型2),以区分鱼脑石与鹅管石粉末。结果:模型1能将文石类与方解石类药材粉末准确区分,并且能区分方解石类的花蕊石和南寒水石,正确率为95%。模型2区分鹅管石和鱼脑石的正确率为90%。模型组整体预测正确率为90%,预测效果良好。结论:XRD技术可将含CaCO_3的具有不同晶型的矿物药粉末区分开,但难以区分同种晶型的不同药材粉末。而运用拉曼光谱技术建立的定性鉴别模型组可将4种含CaCO_3的矿物类中药粉末互相区分开。
OBJECTIVE: To rapidly identify four kinds of Chinese medicine powder with calcium carbonate (CaCO_3) by Raman spectroscopy. Methods: A total of 64 batches of fish brain stone, goose pipe stone, flower stone stone and southern cassia stone were collected from different sources, crushed into powder, X-ray diffraction (XRD) patterns of three batches of powder were collected randomly Powder Raman spectra of all samples. On the basis of analyzing the phase composition of four kinds of medicinal materials by XRD, the Raman spectra of each medicinal material were analyzed. Select the spectral bands 200-300,350-400,670-700cm-1 to establish a cluster analysis model (Model 1); then build a partial least-squares analysis model (model 2) Brain stone and goose pipe powder. Results: Model 1 can distinguish aragonite from calcite powder accurately, and can distinguish the flower stones of apatite from those of south, with a correct rate of 95%. The correct ratio of model 2 to goose pipe and fish brain stone is 90%. The overall prediction accuracy of the model group was 90%, and the prediction effect was good. Conclusion: The XRD technique can separate the powder with different crystal forms of CaCO_3, but it is hard to distinguish the different powder with the same crystal form. The use of Raman spectroscopy established qualitative identification model group can be four kinds of CaCO3-containing mineral Chinese medicine powder separated from each other.