Detection Of The Catechin Monomer In Green Tea Based On Spectroscopy

The content of tea polyphenols affects the merits of tea,which plays a decisive role in the quality of tea.The content of tea polyphenols was significantly different in different stages of tea processing,but there was a serious lack of rapid on-line determination method in the current production process.In this paper,spectroscopy was used to study the changes of tea polyphenols from fresh leaves to different stages of processing,and a rapid determination method of tea polyphenols in green tea was established.The main results of this research are as follows:(1)High-performance liquid chromatography(HPLC)was used to detect the contents of the main tea polyphenols in the different leaf positions(leaf 1 to 6)of three different varieties of tea(Juhuachun,Yingshuang,Zhenong 25).The results showed that the content of four kinds of catechins in tea leaves were:EGCG>ECG>EGC>EC.The contents of tea polyphenols in three varieties were as follows:Juhuachun>Zhenong 25>Yingshuang.And the content of tea polyphenols in each fresh tea leaves decreased with the increase of the number of leaf positions,that is,tea polyphenols content decreased gradually from the first leaves to the sixth leaf.(2)Quantitative determination of tea polyphenols in different leaf positions of different tea leaves was carried out with the use visible-near infrared spectroscopy.The models of the four kinds of tea polyphenols(EGCG,ECG,EGC,EC)based on the full spectral range all obtained good results with RP2 higher than 0.9.At the same time,the successive projection algorithm(SPA)was used to select 16,20,18 and 13 characteristic wavelengths for EGCG,ECG,EGC and EC respectively.Compared to the original 1050 wavelength,nearly 98%of the variables were eliminated.Although the performance of the models was degraded somewhat by the selection of the characteristic wavelength,the stability and accuracy of the model were still preserved.The Rp2 value of the four kinds of tea polyphenols was still above 0.9.To enhance the accuracy of quantitative determination of tea polyphenols,the LS-SVM algorithm was used to establish the nonlinear quantitative model based on the characteristic wavelength.The effect of the model was improved remarkably,the Rp2 values were 0.999,0.989,0.998 and 0.9878 for EGC,EGCG,EC and ECG,respectively.It can be seen that the robustness and accuracy of these models have reached an excellent state,and the working efficiency was improved due to the reduction of the wavelengths.Here,rapid and nondestructive determination of tea polyphenols has been truly realized.(3)Based on the hyperspectral imaging technique,the distribution and variation of various monomer contents in different leaf positions of different tea leaves were further compared.It can be seen from the hyperspectral imaging:?Analysis from the leaf shape,the size of leaf increased first and then decreased from the first to the sixth,among which the fourth leaf position leaves were the largest.?Analysis from the catechin content,the content of catechins in three kinds of tea leaves was as follows:Juhuachun>Zhenong 25>Yingshuang;the contents of four kinds of catechins were:EGCG>ECG>EGC>EC;the content of catechins in the six leaf positions decreased gradually,from the first leaf position to the sixth.And the content of catechins between the sixth leaf and the first leaf was significantly different.?Analysis from the internal structure,the content of catechins monomer in the main veins is relatively low.In addition to the main vein parenchyma cells,the other organization of the leaves almost have the accumulation of catechins and the vascular bundles and fence tissue accumulated more.Analysis from the subcellular level,catechins may mainly accumulate in the chloroplast and vessel wall.The hyperspectral image reveals the distribution and the synthesis mechanism of tea polyphenols in the tea leaves well.(4)Based on the visible-near-infrared spectroscopy,the content of tea polyphenols in three different varieties of green tea was quantitatively detected in each processing stage(plucking fresh tea leaves?spreading ?green removing?rolling?drying).The results showed that the content of tea polyphenols varied significantly in the process of processing,with the depth of each processing stage,the content of tea polyphenols reduced successively and the content of EGCG monomer decreased most obviously.At the same time,the RP2 of the model of each monomer content based on the full spectral range reached up to 0.9,and the characteristic wavelength is selected by both biPLS algorithm and SPA algorithm.By comparison,SPA algorithm is more suitable for this experiment.And 18,20,16 and 17 characteristic wavelengths were selected by SPA for EGCG,ECG,EGC and EC,respectively,which were less than that selected by biPLS algorithm.In addition,although the effect of the model based on the characteristic wavelength selected by SPA was slightly inferior to that based on the full spectral range,the Rp2 all reached more than 0.9 and the accuracy and stability of the model were still high.At the same time,compared with the model based on the full spectral range,the model reduces the number of modeling variables by nearly 98%,which undoubtedly provides the basis and means for the real-time online monitoring of tea quality in the process of green tea.