Systematic Study On The Identification Methods Of Mineral Traditional Chinese Medicine Calamine

Mineral Traditional Chinese Medicine?Mineral medicine for short?refers to the traditional Chinese medicine,including the natural minerals,the processing products of mineral materials and the fossils of ancient animal or animal bones.Calamine,recorded in the 2015 Edition Volume I of China Pharmacopoeia?ChP 2015??,is smithsonite belonging to carbonate minerals in the calcite group and mainly contains Zinc Carbonate?ZnCO₃?.As a drug for external application,it is widely used for treating dermatitis,prickly heat,eczema,urticarial and other skin diseases in traditional Chinese Medicine.Investigation of commercially available Calamine samples showed that there were a lot of fake and inferior Calamine goods in the market.It is worth noting that the samples in the same batch of mineral medicines are not always uniform and homogeneous,therefore the identification for every single sample in a batch is needed.To ensure the safe use and to meet pharmaceutical production need,it is necessary to establish an analytical method for rapid identification of large scale of Calamine samples.This paper makes a systematic study on the identification methods ofCalamine.Based on the accurate identification of samples by the macroscopic identification,microscopic identification,physicochemical identification,and XRD which accurately identifying the source of the samples,this paper focuses on NIR modeling.We select the correlation coefficient method,partial least squares algorithm to establish the qualitative and quantitative model of Calamine,analysis the rate of correct identification,to achieve rapid qualitative and quantitative analysis of Calamine.Further,to improve the accuracy of identification we have explored the application of intelligent modeling algorithm in modeling incuding SVM,BP-ANN and GA,CARS and other.Finally,all kinds of Calamine identification information and data are collected and converted into standardized text,graphics and map.Based on the identification information,the database is established and the information management platform of Calamine is constructed,to implement the query of Calamine samples information and identification of unknown samples.The contents and results of this study are as follows:1.To verify and evaluate the identification method for the Calamine samples?1?Macroscopic identificationEight characters of Calamine are extracted,including the shape,color,powdery,glossiness,surface roughness,porosity,texture and smell,which are used to identify 21 batches of commercially available Calamine samples.Based on the identification results,we analysis the feasibility of macroscopic identification.Compare the 6 corresponding characters of the sample with the ChP 2015?,and the results show,10 batches of samples are in accordance with the requirements of the Pharmacopoeia,7 batches of samples are not and 4batches of undetermined samples.Combined with subsequent identification results,the judgment results of 17 batches as genuine and counterfeit were completely correct,and the accuracy rate was 81%.?2?Physicochemical identification and Content determinationAccording to the ChP 2015?,identify 28 batches of Calamine samples by the physicochemical identification and content determination methods.Take1g powder with dilute hydrochloric acid dissolution,filtration,and the filtrate is added with potassium ferrocyanide solution for chemical reaction,observe the sediment color.The content of ZnO in Calamine samples was determined by EDTA titration.The results show,16 batches of samples that meet the requirements of the the ChP 2015?,which can produce white precipitation,or have a small amount of blue precipitate.Of which only 12 batches of samples were determined as genuine by content determination,false positive rate reached25%.The content of ZnO meet with the content determination has 15 batches,accounted for 53.6%and these samples by XRD method validation are mainly sourced from hydrozincite and smithsonite.2.Analysis of the Calamine samples of XRD phase compositionAccording to the analysis results of XRD and the content determination,15 batches of samples,which are sourced from hydrozincite and smithsonite and the Zn O content was more than 40%,can accurately identify as genuine Calamine.3 batches of calcined products and 1 batches of products processed by water-grinding,which phase composition contains ZnO and ZnO content was more than 56%,can accurately identif as processed Calamine genuine.The other samples are calcite,and they are all counterfeits.3.Investigated the origin of CalamineCalamine,first recorded in‘Materia Medica of external alchemy‘,which is named Calamine in English and Calamina in Latin.In the 2015 Edition Volume I of China Pharmacopoeia?ChP 2015??,it is smithsonite belonging to carbonate minerals in the calcite group and mainly contains Zinc Carbonate?ZnCO3?.According to the previous results,14 of 15 batches of crude Calamine samples are sourced from hydrozincite and only 1 batch from smithsonite.The results show,?1?the method of the Chp can not distinguish between smithsonite and hydrozincite;?2?the smithsonite as Calamine resource is lack in the medicine market;?3?the crude Calamine samples are mostly hydrozincite.Therefore,it is suggested that the origin source minerals of Calamine should increase hydrozincite.4.Established the qualitative and quantitative model of NIR?1?Established the MRCC qualitative model:collected 62 batches of samples?including 32 batches of commercially available Calamine and 30batches of self-made Calamine?near infrared spectral data,the main characteristics of spectrum were 7500 cm^-14 000 cm^-1 region,spectral preprocessing by means of first derivative+9 point smoothing,using correlation coefficient method multi-reference,the qualitative model was established.The model prediction accuracy reaches 85%,and can also distinguish the crude Calamine,counterfeits and products processed.?2?Established the PLS quantitative model:using the partial least square method to establish the quantitative model of 66 batches of Calamine samples?including 36 crude Calamine samples and 30 positive samples?.The main characteristics of spectrum were 7500 cm^-14 000 cm^-1 region,spectral preprocessing by means of first derivative+13 point smoothing.The external verification proved that the root mean square?RMSEP?was 3.66.The abnormal spectrum was removed and the RMSEP and R² values were recalculated.The R² was 93.56%and the RMSEP value was 2.6.The prediction effect was improved,which indicated that the prediction ability of the model was better.5.Explored the application of intelligent algorithm in NIR identification?1?Established the SVM qualitative modelThe modeling of the algorithm is implemented in MATLAB.Collected the near infrared spectral data 62 batches of samples,spectral preprocessing by means of first derivative+9 point smoothing,and PCA method is used to reduce dimension.The SVM classification method provided by MATLAB libsvm toolkit is used to classify three categories samples of the crude Calamine,counterfeits and products processed.The SVM kernel function selects the linear kernel function,and the adjustment parameter is 100.Randomly selected 2/3 data as training set,1/3 data as validation set,and repeated tests 100 times,statistics the correctness of each validation set,and finally the average classification accuracy is counted.The prediction accuracy of the SVM model is 94.24%.?2?Established the BP-ANN qualitative modelBased on MRCC methods,the 3-layer structure of the network is established by MATLAB ANN toolbox.The learning rate of neural network was 0.1,the momentum factor was 0.9,and the number of learning was 50times.The verification result of the model shows that BP-ANN model can achieve accuracy rate at 95%.Compared with the previous MRCC model,the accuracy of the prediction is greatly improved.?3?Established the GA,CARS+PLS quantitative modelFirst select the multiple scatter correction?MSC?+second derivative+Norris?5,3?smoothing method to preprocess the NIR original spectral,establish NIR correction model based on PLS with 200 optimal characteristic wavelengths chosen by GA and CARS algorithm.The results show,the NIR quantitative model based on CARS algorithm has best effect.6.Construct the information management platform for identification of CalamineBased on the identification characteristics of mineral medicine and the data obtained by the project group,this research explored the construction method of the identification information database and realize the qualitative identification of unknown samples according to the data in the database.The system is based on the B/S structure and uses the JAVA EE+MySQL to realize the design of the front function interface and the backstage system management.It mainly includes data entry,all kinds of information retrieval,display,spectral comparison,data export,data addition and verification,user management and other functions.In summary,this paper mainly completed the following research.?1?Verify and evaluate the traditional identification method for the Calamine samples.?2?Investigated the origin of Calamine.Iit is suggested that the origin source minerals of Calamine should increase hydrozincite.?3?Based on content and XRD,established the MRCC qualitative model.The model prediction accuracy reaches 85%.?4?Explored the application of intelligent algorithm in NIR identification.Established the SVM qualitative model and the prediction accuracy of the SVM model is 94.24%.Established the BP-ANN qualitative model and BP-ANN model can achieve accuracy rate at 95%.The two models can also distinguish the crude Calamine,counterfeits and products processed.Established the CARS/GA+PLS quantitative model and the value of RMSEP is 1.4774,that shows the prediction effect of this model is better.?5?Explored the construction method of the information management platform for identification of Calamine,which provides the basis for identification,production and management of Calamine.