Study On Rapid Detection Method Of Tea Meal Based On Near Infrared Spectroscopy

Tea meal is rich in tea saponin,protein,polysaccharides,etc.According to the tea meal containing different components,it is widely used in pharmaceuticals,health products,pond cleaning,feed,military and other industries.In this study,chemical analysis method is used to detect the tea saponin,oil content,protein and moisture content in the collection of 164 batches of tea meal samples which were produced in Jiangxi province.Then based on these data,near-infrared(NIR)spectroscopy technology is combined with chemometrics method and used for quantitative analysis of the main components in these tea meal samples.At first,the moisture content,tea saponin,protein and oil content in the collection of164 batches of tea meal samples were determined through the heating and drying method,hot ethanol extraction gravimetric method,Kjeldahl method and Soxhlet extraction gravimetric method,respectively.The results showed that the moisture content is in the range of 6%-13%,and its content is mainly concentrated in the range of 8%-10% which is accounting for 95.1% of the total,and the time-consuming is 8-10 h.The content of tea saponin is in the range of 11%-31%,and its content is mainly concentrated in the range of14%-23% which is accounting for 93.9% of the total,and the time-consuming is 10-12 h.The protein content is in the range of 1%-19%,and its content is mainly concentrated in the range of 6%-13% which is accounting for 96.4% of the total,and the time-consuming is 6-8 h.The oil content is in the range of 0%-10%,and its content is mainly concentrated in the range of 1%-5% which is accounting for 84.7% of the total,and the time-consuming is 10-12 h.This paper mainly discussed the influencing factors of the establishment of NIR technology model from the aspects of experiment parameter selection,spectral acquisition,processing data method and model establishment,and validated the model.The influenceof the thickness and the number of scans on the NIR spectrum is studied,and the optimum thickness and the number of scans are 4 mm and 32 times respectively.With the established experimental parameters and conditions,the collected tea meal samples were conducted by spectral acquisition.The first-order derivative,second-order derivative,9-point smoothing and multiple scattering correction technique were used to the collected samples.The PLS and PCR quantitative calibration techniques were used to establish the NIR technology model.The data of tea saponin,oil content,protein and moisture content in tea meal were verified by the established model.The correlation coefficients of calibration set and validation set of tea saponin were 0.9593 and 0.9398 and the standard errors were 1.93% and 1.77%,respectively.The correlation coefficients of calibration set and verification set of protein were 0.9691% and 0.9488% and the standard errors were 1.83% and 1.87%,respectively.The correlation coefficients of calibration set and validation set of oil content were 0.9593 and 0.9398,and the standard errors were1.82% and 1.69%,respectively.The correlation coefficients of calibration set and verification set of moisture content were 0.9593 and 0.9398,and the standard errors were1.67 %and 1.63%,respectively.All the results show that the NIR spectroscopy technology is fast and efficient,and can be applied to the analysis of tea meal and other agricultural products.This study will provide some theoretical support for the NIR spectroscopy in analysis of food and pharmaceutical products,and lay a good foundation for the application of NIR technology...