Quantitative Analysis By NIR Spectrum Technic For The Main Components Of Savor In Green Tea

When optics , data processing technic of the computer, theoretics of chymistry metrology are improving ,and new instruments of NIR come forth, and with retread of software edition, Stabilization, practicability and veracity of Near infrared spectroscopy have being incessant improved. Being excellence of speediness, high efficient, convenient and capability of measured more components at one time, NIR technique has been acquainted by peoples, is used in many fields. And it is one of analytical technique which has made rapid progress at present.This project was to measurate lixivium, tea polyphenols(TPP), amino acids(AA) and theine of green tea soup by national normal measure.and then constitute calibration models for tea polyphenols(TPP), amino acids(AA) and theine of green tea soup (the percent of lixivium) by using different spectra pretreatment methods and regression methods for tea samples. Spectroscopy pretreatment methods were multiplication scatter correction (MSC), standard normalized variate (SNV), Svitzky Golay (SG), Normalization, first derivative and second derivative; regression methods were principal components regression (PCR) and partial least square (PLS). The results indicated that the PLS was the optimal method for calibration models of three components. It indicated that by internal cross validation : the tea polyphenols(TPP) model with none spectra pretreatment was the best spectra pretreatment, the standard error of calibration (SEE), correlation coefficient (r)of calibration set (C-set), standard error of cross- validation (SEP) and and validation set (V-set) were 1.1455,0.8792,1.0600,0.8623; the AA model with log spectra pretreatment was the best, its SEE, SEE-r,SEP, and SEP-r were 0.2518,0.9728,0.2597,0.9722; the theine model with fi none spectra pretreatment was the best, its SEE, SEE-r SEP, and SEP-r were 0.5553,0.9648,0.5315,0.9554. TPP calibration model had less effect than AA and theine calibration models. It indicated that by external cross validation, correlation coefficient (R)of true content of samples and forecasted content of samples: R of TPP is 0.8935, R of AA is 0.9541, R of theine is 0.9386.The percent of TPP ,AA and theine in soup lixivium is about 70%,tiptop is upwards 80%.So grade of savour correlate with three savour component of tea soup. This article constituted statistic equation by statistical analysis, expecting to obtain grade of savour apace and by rule and line.By analyzing relativity of the main quality components and organoleptic evaluation in tea soup of 42 tea samples, it can be found that TPP is prominent minus correlation with savor grade(P- value is 0.0107, less than 0.05), AA is prominent positive correlation with savor grade(P- value<.0001),but theine has not prominent correlation with savor grade.It is not all-sided that it analyzes for single component, By integrating analyse of every component, it can set up statistic equation:Y=84.251 -0.263X1+3.295X2-0.703X3 (Y: savor grade, X1: TPP, X2: AA, X3: theine) .This project makes NIR analysis for the main quality components(tea polyphenols(TPP), amino acids(AA) and theine) in tea soup of 42 tea samples, and has set up calibration models which have better effect, it can forecast content of sealed samples by scan sealed samples and import into calibration models, which is save trouble. Finally forecast content of the main quality components are imported into statistic equation, and can obtain savor grade of tea samples. This project obtain calibration models and statistic equation by NIR analysis and statistic analysis, which afford academic foundation for mensuration of the main quality components of tea soup and savor grade of instrumental mensuration.