Spectroscopy Of Several Kinds Of Natural Chinese Traditional Medicine

The qualities of several natural traditional Chinese medicines were detected by infrared and near infrared spectroscopy analysis in this paper. Infrared spectroscopy has the advantages of non-destructive determination, no pollution, on-line determination and convenient operation. It is widely used in agriculture, industry, medicine and health, and other fields due to providing more information about molecular structure. The near infrared spectra consist of overtones and combination bands of the fundamental molecular absorptions found in the mid infrared region, and the generally overlapping vibrational bands may appear non-specific and badly resolved.The information extraction technology of near infrared spectra is the near infrared spectroscopy's key link. This paper focuses on the information extraction and processing during the process of detecting near infrared spectroscopy analysis. The chemical metrological methods, such as wavelet, partial least squares (PLS) and neural networks were applied to process infrared spectral data of several natural medicines.1. Based on the application of the Fourier transform infrared spectroscopy analysis, we know that ginseng presents the characteristic adsorption at 577, 1026, 1242 and 1417 cm-1, providing a convenient basis for our future research work. Based on I value calculation of ginseng adsorption at 3380, 1635 cm-1, we know that I value, age and producing area of ginseng have great relations, which provides another basis for the identification of ginseng. Ginseng from the different producing areas can be intuitively and clearly identified using two dimensional infrared spectroscopy technologies, which offers a new method for qualitative analysis of Chinese medicines.2. Effective compression of the spectral variables was performed by wavelet transform infrared spectroscopy with the spectral characteristics maintained and effectiveness in reducing spectral variables. The model for the classification of Jilin ginseng regions was established in conjunction with the artificial neural network. It is proved that the model forecasts good effect and the recognition rate reaches 92.5 percent. The model was constructed using support vector machine (SVM) method suitable for the small samples. Simultaneously, Utilizing polynomial and RBF kernel, the classification accuracy of two different parameters were compared. The results show that using RBF kernel while the value ofσadmission is 0.6; the classifier has the highest classification accuracy of 97.5%. Meanwhile, the use of SVM technology combined with wavelet transform infrared spectroscopy can correctly distinguish the producing area of Chinese herbal medicines and the recognition rate is much higher than that of the ANN network model. It provides a certain theoretical basis and technological support for further analysis and research of infrared spectrum of Chinese herbs so as to the circumstances of the small samples. 3. Infrared spectroscopy combined with artificial neural networks as well as wavelet transform techniques were used to identify 42 Epimedium samples from three different regions of Jilin Province, which achieved satisfactory results of the classification. Effective compression of the spectral variables was performed by wavelet transform infrared spectroscopy. It maintains a spectral characteristic, and is effective in reducing spectral variables. The model for the classification of Chinese herbal medicine Epimedium was established in conjunction with the artificial neural network. It is shown that the model forecasts good effect and the recognition rate reaches 95.2 percent. This presents that the use of infrared spectroscopy can quickly and accurately identify traditional Chinese medicines, while avoiding the separation and extraction of traditional spectrum analysis, so as to provide a more powerful and reliable basis for the scientific quality control and modern management of traditional Chinese medicines. Using cluster analysis on the samples, qualitative identification of the producing area of epimedium provides new ideas for producing area analysis of traditional Chinese medicines.4. The infrared spectra of coptis were measured under the conditions of preparation using Fourier transform infrared spectrometer, among the IR spectral which range from 1300 to 1600 cm-1 are the characteristic peaks of active ingredients berberine in coptis. With the first derivative spectral pretreatment, and the first derivative IR spectra of the samples were found to be similar, which can be used to determine the berberine content in comparison with the intensity of the original spectral peaks. By using normalization treatment method, the berberine content varies from top to bottom according to the sequence of the original spectral peaks intensities descendingly. Therefore, the basic relations between berberine content and preparation conditions were obtained, which provides suitable basis for the drugs preparation of the optimal content.