Study Of Danshen And Baishao Herb By Infrared Spectroscopy

In this thesis, Danshen (Salvia miltiorrhiza) and Baishao(Paeonia lactiflora) herbs are studied by the multi-steps macro-fingerprint method which combines general infrared spectroscopy (IR), second derivative spectroscopy and two-dimensional correlation infrared spectroscopy (2D-IR), and Partial Least Squares (PLS) regression, SIMCA (Soft Independent Modeling of Class Analogy) chemometric method. The main conclusions are as follows.①Based on multi-steps macro-fingerprint method, Danshen herbs from 6 different areas are identified. In the 2D-IR spectra, Danshen herbs from Shandong, Shanxi, Henan and Sichuan provinces all appear the strongest characteristic absorption peaks at 1575cm-1, which belongs to stretching vibration of C=C. It means that C=C is stronger sensitive to thermal perturbation. But in the 2D-IR spectra of Danshen herbs from Zhejiang and Yunnan provinces, the position of characteristic absorption peaks are different from other areas during the range from 1500 to 1800cm-1. It means that the sort and relative content of aromatic compounds in Zhejiang and Yunnan Danshen herbs are different from other areas.②Based on IR spectra and SIMCA, Danshen herbs from 7 different areas can be classified and predicted. The prediction accuracy is more than 86.7% for every area.③Characteristic absorption peaks of Danshen effective constituent are found by analyzing the spectra of Danshen, tanshinone IIA and salvianolic acid B. It is the fingerprint characteristics of Danshen, and also used as basis of identification. The second derivative spectrum of Danshen herb appears characteristic absorption peaks at 1746, 1730, 1691, 1670 (νC=O of ketone and carboxylic acid), 1610, 1580, 1543, 1509, 1489, 1453 cm-1(skeletal vibration of aromatic compounds).④Characteristic absorption peaks of Baishao effective constituent are found by analyzing the spectra of Baishao and paeoniflorin. It is the fingerprint characteristics of Baishao, and also used as basis of identification. The IR spectrum of Baishao extract appears characteristic absorption peaks at 1041, 1102, 992 cm-1 (νC-O of indicant), 925 cm-1(ring stretching vibration of indicant). The second derivative spectrum of Baishao extract appears characteristic absorption peaks at 1714cm-1(νC=O of aromatic ester),1602, 1584, 1453 cm-1(skeletal vibration of aromatic compounds),1281, 1178 cm-(1antisymmetric and symmetricνC-O-C of ester),1102, 1077, 1048, 1026, 988cm-1(νC-O of indicant),942, 922, 895cm-1(ring stretching vibration of indicant).⑤IR spectra and content data of paeoniflorin of Baishao are used to build models by PLS algorithm. 31 samples are used to calibrate, and 6 samples are used to validate. The R 2 of calibration model is 0.9696, when the pretreatment is second derivative and MSC, and waveband is 2950~650 cm-1. The R 2 of calibration model is 0.9713, when the pretreatment is second derivative and SNV+De-trending, and waveband is 2900~650 cm-1. The relative deviation between predicted and actual value is less than 2.5% for every unknown sample. The quantitative analysis method using IR and PLS algorithm has advantages of simple and rapid. This method provides new idea and solution to control quality of main constituent in traditional Chinese medicine.