The Study Of Applying Basis Of Quantitative Analysis In Near-infrared Fourier Transform Raman Spectroscopy

Fourier-transform Raman spectroscopy (NIR-FT-Raman) is a fast technique that can provide component information about intact samples. In this thesis, the applying basis of quantitative analysis using NIR-FT-Raman spectroscopy has been studied. Besides the traditional quantitative analysis method that based on the intensity of Raman peak, we have combined NIR-FT-Raman technique with Chemometrics to perform a quantitative determination of Ibuprofen tablet and Amino acid liquid compound sample. There are three multivariate calibration methods has been used, including linear Principle Component Regression (PCR), Partial Least Squares Regression (PLS) and nonlinear Artificial Neural Network (ANN).The results indicate that using PLS can establish comparatively good analysis model. The relate coefficient of Ibuprofen and Starch PLS model is 0.9974 and 0.9990, respectively. The relate coefficient of Lysine, Proline and Valine PLS model is 0.9974,0.9989 and 0.9990, respectively. Simutaneously, the predition result is influenced by the exciting power evidently.In addition, Dispersion Raman spectroscopy (DRS) and Fourier-transform Near-infrared spectroscopy (FT-NIRS) had been used to quantitative analysis solid and liquid samples, respectively. The quantitative analysis results of NIR-FT-Raman spectroscopy close to the results of FT-NIRS and indicate that NIR-FT-Raman spectroscopy more suit to quantitative analysis multivariate compounds compare to DRS.