Study On Identification Of Several Species Of Mineral Medicines Based On X-Ray Diffraction,Raman And Near-infrared Spectroscopy

Objective:To provide rapid and accurate method for the identification and quality control of mineral medicine,X-ray diffraction?XRD?,Raman spectroscopy and near-infrared?NIR?spectroscopy were applied to dealing with the major problems in the identification research of mineral medicine.Methods:?1?XRD Fourier patterns collected from sample powders were analyzed to determine the original mineral resources of Chloriti Lapis,Micae Lapis Aureus and Vermiculitum by comparing with characteristic traits.First derivative,Vector normalization and 21 point smoothing are used to pretreat the selected spectrum band from 0.68¹.77 nm.Then the data were analyzed by fuzzy cluster.?2?Raman spectra were collected from 20 baches of samples including Cinnabaris,Calomelas,Realgar,Arsenicum,Lithargyrum,Plumbum Rubrum and Sulfur.Spectral characteristics and their corresponding molecular vibration forms were analyzed,and the assignments of characteristic peaks were given.?3?XRD,NIR and Raman spectra were collected from samples including Alumen and calcined Alumen,and the spectral characteristics and differences were analyzed.?4?Samples of eleven easily-confused mineral medicines,including alumen,natrii sulfas,natrii sulfas exsiccatus,gypsum fibrosum,calcite,yellow croaker otolith,borax,talcum,quartz,sal ammoniac and sal Praeparatum,were collected from different sources.Firstly,morphological method and XRD method were used to identify their authenticity.Then Raman spectra of samples were collected.Raman spectral characteristics peaks and their corresponding molecular vibration forms were analyzed.Characteristic intensity data extraction method and principal component analysis?PCA?were performed to reduce the dimensionality of Raman spectral data.The Raman identification model was subsequently built by SVC algorithm.Finally,NIR spectra of samples were collected.The NIR qualitative model of eleven easily-confused mineral medicines was established based on distance discrimination method.Results:?1?It has been found experimentally that the original mineral resource of 7 sample powders of Chloriti Lapis is biotite schist belonging to metamorphic mineral.The original mineral resource of 3 sample powders of high-quality Micae Lapis Aureus is vermiculite biotite schist belonging to metamorphic mineral.The original mineral resources of 3 sample powders of Vermiculitum are phlogopite and vermiculite phlogopite.?2?The characteristic spectral bands were in the area below 800 cm^-1,the characteristic peaks were sharp and correspond to main composition.Raman spectra of the same kind of medicinal materials have good consistency,and there is distinct difference between different kinds of medicinal materials.?3?there are significant differences between Alumen and its counterfeit Ammonium alum.In XRD patterns,Ammonium alum has a peak at d=0.61,which Alumen does not.In NIR spectra,Ammonium alum has a characteristic peaks at 4 650 cm^-1,which Alumen does not.In Raman spectra,the strongest peak of Alumen split into 2 peaks 990 cm^-1 and 974 cm^-1.The above differences could be used for the rapid identification of Alumen and its counterfeit Ammonium alum.Besides,calcined Alumen appears 9characteristic peaks?d=0.79,0.41,0.36,0.29,0.27,0.24,0.23,0.18,0.17?in XRD patterns and 2 characteristic peaks(7 000 cm^-1 and 5 150 cm^-1)in NIR spectra.?4?Through an observation on the morphological characteristics as well as an XRD analysis of eleven mineral medicines,it is found that the morphological identification can only be used to distinguish medicinal materials with intact shape and evident features,which proves to be limited to some extent.Instead,XRD can discriminate the eleven mineral drugs with obvious effects and accurate results.In the research with Raman spectroscopy,apart from the Raman spectra of talcum and sal Praeparatum,the remaining spectra were seen with sharp peaks and clear characteristics.The 3-fold cross validation?3-CV?accuracy of the SVM model established based on extracting characteristic intensity data from spectra pretreated by first derivation was98.61%,and the prediction accuracies of training set and validation set were100%.As for the PCA-SVM model,when the spectra pretreated by vector normalization and the number of principal components?NPC?is 7,the 3-CV accuracy and prediction accuracies were all reached 100%.Both models have good performance and strong prediction capacity.In the research with NIR,NIR identification model was established based on two libraries.Library 1was built through standard algorithm?Euclidean distance method?with 4000¹0 000 cm^-1 as modeling spectra and vector normalization+first derivation+smoothing as the preprocessing method.Library 2 was built through factorization method with 4 000⁷ 000 cm^-1 as modeling spectra and second derivation as the preprocessing method.The validation report shows that the samples of training set can be uniquely identified,and the prediction accuracy of validation set was 96.6%,which mains the NIR model has a strong prediction ability.Conclusion:In contrast with traditional identification methods,XRD,Raman and NIR techniques turn out to have significant advantages regarding their application of dealing with the major problems in the identification research of mineral medicine.The details of ingredients contained in mineral medicines can be directly acquired by analyzing XRD phase.XRD is accurate and reliable when applied to differentiate mineral medicines which are difficult to identify with morphological characteristics.Raman spectra are featured with sharp peaks and clear characteristics.The chemical component information in mineral medicine can be gained by assigning peaks to molecular vibration so as to identify the species or authenticity of medicinal materials.It is also workable by building a mathematical model to achieve rapid identification for mineral medicine.NIR can achieve rapid identification and quantitative detection by establish qualitative and quantitative models with chemometrics methods.Raman and NIR technology are worthy to be popularized as rapid detection of mineral medicine.The research provides practical experience for XRD,NIR and Raman spectroscopy technologies regarding the rapid identification of mineral medicine.In application,based on the actual situation,one optimum analytical method or the combination of the three can be used in order to ensure the reliability of results.