Quality Control Of Traditional Chinese Medicine Using Near Infrared Spectroscopy

Near infrared(NIR)spectroscopy is currently the most widely used process analytical technology(PAT)in the pharmaceutical industry.Its application in the quality control of traditional Chinese medicine has also attracted much attention in recent years.Traditional Chinese medicine is often very complex in composition,and the production processes are distinctive from that for pharmaceutical chemicals.As a result,the application of NIR spectroscopy in the traditional Chinese medicine domain faces major challenges,in particular in chemometric model development.The research conducted in this paper was aimed to address the challenges.The research is summarized below:Firstly,the effects of fiber optic cable and temperature on the quantitative model of polysaccharides concentration were investigated.1)The effects of different fiber optic cable lengths and pretreatment methods on the performance of partial least squares(PLS)models were compared.These models with different fiber optic cable lengths showed good fitness but bad prediction ability after processing the spectra using 1st derivative + detrend.Furthermore,the variables of Polysaccharides were selected by ganatic algorithm(GA).It was found that the PLS model with shorter fiber optic cable length presented better prediction ability.It was illustrated that fiber optic cable length had an effect on the NIR analysis result of polysaccharides determination,and appropriate length could guarantee the detection accurancy.2)The effect of temperature on the performance of PLS models was compared.Principal component analysis(PCA)was employed for processing the spectra of the samples at the two temperatures after 1st derivative pretreatment,it was found that the spectra were grouped to two clusters.The quantitative models of polysaccharides concentration built at both temperatures had similar performance and prediction ability,but the versatility between models was not good.A mixed model was thus built using spectra with different temperatures,which showed good prediction ability at both temperatures.When collecting near infrared spectra,the temperature should be controlled to guarantee the detection accurancy.Secondly,a prepared solution of traditional Chinese medicine was considered.Quantitative models of its index components(polysaccharides concentration,soluble solids concentration,pH value)during prepared stage were built using NIR.There was no characteristic absorption peak of the polysaccharides in the near infrared region,so the traditional modeling methods were not effective.In this section,the effects of pretreatment methods and variable selection methods on model performance were considered to select characteristic absorption peak of the polysaccharides.The optimal pretreatment method for the quantitative model of polysaccharides concentration was found to be 1st derivative + detrend.410 variables were selected by GA,the Rc2 was 0.969,and the relative error was less than 10%,which is the minumum prediction error of polysaccharides detection in the literature.The results indicated that the model of NIR spectroscopy could satisfy the industry's requirement.At last it showed how to build a monitoring interface and detect the outlier by T2-Q,which could provide a reference for application of NIR spectroscopy in traditional Chinese medicine.Thirdly,the concentrated solution of traditional Chinese medicine was investigated.Quantitative models of its index components(polysaccharides concentration,solids concentration)were built during the concentrated stage using GA-PLS.These models showed good prediction ability and could be used to control the quality of traditional Chinese medicine concentrate on line.