Study On Fourier Transform Infrared Spectra Of Rice And Millet

Rice is one of the most important foods in China. Millet won the favor of peoplebecause it has many therapeutic health effects such as strengthening one’s spleen,regulating stomach and so on. There are many kinds of rice and millet, and it is noteasy to judge them according to the appearance. In this dissertation, mid-infraredspectroscopy combined with statistic analysis was used to study6kinds of rice and7kinds of millet samples.Fourier transform infrared spectroscopy (FTIR) combined with principalcomponents analysis (PCA) and hierarchical cluster analysis (HCA) was used tostudy of77rice samples. The results showed the spectra were very similar and mainlycomposed of the absorption of starch and protein. The second derivative spectra ofthe samples exhibited obvious differences in the starch region. The second derivativespectra in the region of1800~500cm-1were used to perform correlation analysis,hierarchical cluster analysis, and principal component analysis. Results showed77rice samples were divided into six groups, the classification accuracy of principalcomponents analysis and cluster analysis are100%. It is proved that Fouriertransform infrared spectroscopy combined with statistical analysis might be used toclassify rice varieties.Fourier transform infrared spectroscopy combined with principal componentsanalysis and hierarchical cluster analysis was used to study white millet, yellow millet,glutinous millet, green millet, black millet, stale yellow millet and proso millet. Theresults showed that their original spectra were similar, but the second derivativespectra in1800~1400cm-1range were significant different. The second derivativespectral data were selected to perform multivariate analysis. Results showed that theclassification accuracy of principal component analysis and hierarchical clusteranalysis were about84.6%and92.3%respectively.This study demonstrated that FTIR spectroscopy combined with chemometricscould be applied to distinguish different kinds of rice and millet. Fourier transforminfrared(FTIR) spectroscopy combine with chemometrics analysis in theidentification of rice and millets are fast, accurate, and will be developed into a powerful method for detection of different varieties of cereal samples.