Analysis Of Rutin And Protein In Buckwheat By Near-Infrared Spectroscopy (NIRS)

Buckwheat is Dicotyledon Plant, which is one of the most important crops to be developed as health care and therapeutical food. Analyzing buckwheat various nutritions content by near-infrared spectroscopy would realize quicky, accurate, lossless, environmental protection test. The paper takes 100 varieties of common buckwheat and 80 varieties of tartary buckwheat as samples, and gets contents of rutin, protein, starch, amylase and amylopectin in buckwheat. Based on the examination, scan the buckwheat flour by MPA from German Bruker Corporation, and get the spectrums of buckwheat. Then establish near-infrared spectroscopy calibration models of each components, and improve them. Finally determine the optimum modeling parameters. The results as follow.The common buckwheat rutin content amplitude is 0.180%-0.688%, the mean value is 0.396%±0.108%. The tartary buckwheat rutin content amplitude is 0.590%-1.840%, and the mean value is 1.250%±0.220%. The common buckwheat protein content amplitude is 4.220%-9.103%, and the mean value is 5.911%±0.876%.The tartary buckwheat protein content amplitude is 6.190%-10.870%, and the mean value is 8.114%±1.074%. The indexes of tartary buckwheat are higher than common buckwheat.The paper establishes near-infrared calibration models of buckwheat various components, and determined the optimal parameters of models. The optimum conditions of common buckwheat rutin calibration model are that near-infrared spectroscopy regions are 7501.9-6098 cm^-1 and 5453.8-4597.6 cm^-1, the spectrum pretreatment method is COE, the number principal component is 12. The parameters of model are that R2 value achieves 87.31%, the RMSECV is 0.0355, the RMSEP achieves 0.046.The optimum conditions of tartary buckwheat rutin calibration model are that near-infrared spectroscopy regions are 7501.9-6098 cm^-1 and 5453.8-4597.6 cm^-1, the spectrum pretreatment method is MSC, the number of principal component is 7. The parameters of model are that R2 value achieves 91.57%, the RMSECV is 0.0571, the RMSEP achieves 0.161.The optimum conditions of total buckwheat rutin calibration model are that near-infrared spectroscopy regions are 9013.9-7498 cm^-1,5453.8-4597.6 cm^-1, the spectrum pretreatment method is First Derivative +MSC, the number of principal components is 8. The parameters of model are that R2 value achieves 92.79%, the RMSECV is 0.093, the RMSEP achieves 0.046.The optimum conditions of common buckwheat protein near-infrared calibration model are that the spectrum region is 9000-4000 cm-1; the spectrum pretreatment method is MSC, the number of principal components is 11. The parameters of model are the R2 value achieves 99.41%, the RMSECV is 0.067, the RMSEP achieves 0.212.The optimum conditions of tartary buckwheat protein near-infrared calibration model are that the spectrum region is 9000-4000cm^-1 , the spectrum pretreatment method is first derivative + vector normalization, the number of principal components is 10. The parameters of model are the R2 value achieves 97.59%, the RMSECV is 0.163, the RMSEP achieves 0.090.The optimum conditions of total buckwheat protein calibration model are that near-infrared spectroscopy regions is 9000-4000 cm^-1 , the spectrum pretreatment method is rst derivative + vector normalization, the number principal components is 8. The parameters of model are the R² value achieve 98.88%, the RMSECV is 0.154, the RMSEP achieves 0.145.Indicated from the experiment that it is completely feasible to establish the near-infrared calibration models of buckwheat rutin and protein . The models have outstanding predictive and analysis effects.