Study On Determination Of Adulteration In Matcha Based On The Near-infrared Spectroscopy

Matcha is one of the products of tea deep processing,which show good economic due to the good quality characteristics..However,in the current market,the matcha products in good and bad are intermingled.The non-tea component are added to matcha,or sell seconds at best quality prices.This not only undermines the legitimate rights and interests of consumers,but also undermines the market order of the matcha,for which the establishment of fast and reliable method of camouflage detection method to maintain the market order and quality of matcha is of great significance.This study mainly aims at the market common matcha adulteration(white sugar,maltodextrin,mulberry leaf powder,barley powder)as research object.Near infrared spectral analysis technology,combined with chemometrics methods(principal component analysis and linear discriminant analysis,K nearest neighbor method and partial least squares),respectively established the model of quantitative analysis and qualitative pure matcha and matcha adulteration,and 4 kinds of adulteration of matcha.(1)Effect of pretreatment of matcha sample on quantitative model.In the quantitative model of the pressed and not pressed cake,the addition of white sugar and maltodextrin was used as the research object.The results showed that the not pressed cake model was better than that of the pressed cake.It can be used as a reference for the near infrared spectroscopy analysis of ultra-fine powder.(2)Near infrared spectroscopy(FTIR)combined with principal component analysis and linear discriminant analysis(PCA-LDA)and K nearest neighbor method were used to build models.A total of 42 samples of pure matcha and 150 samples of pseudo-opium matcha were collected to establish qualitative discriminant analysis to determine whether the matcha was adulterated and the adulterated type,and the results of PCA-LDA were superior to K nearest neighbor method.The recognition rate of the calibration set of pure matcha and matcha adulterated white sugar,pure matcha and matcha adulterated maltodextrin,pure matcha and matcha adulterated mulberry leaf powder,pure match and matcha ad ulterated barley flour,four kinds of fake matcha,were 98.3%,100%,91.7%,100%,100%,100%,respectively,and theprediction set recognition rate were respectively 96.5%,100%,87.5%,95.8%,90.3 %,95.3%.It can be seen that the models had very good precision and stabilization and they could quickly and accurately discriminate whether matha was adulterated.(3)Near infrared spectroscopy combined with partial least squares method,using four different spectral pretreatment methods of spectral pretreatment,establish quantitative model and several adulteration content in matcha.For the addition of white sugar,mulberry leaf powder,0~50% and 50~100% barley flour powder of the method of spectral pretreatment to smooth the results of the best,the results of the maximum and minimum normalization pretreatment of maltodextrin were the best.The correlation coefficients(Rc)of the calibration set were 0.991,0.994,0.997,0.997 and 0.997,respectively,and the RMSECV were 0.391,1.6,1.11,2.44 and 0.66;The correlation coefficient(Rv)of the prediction set were 0.9992,0.9984,0.9976,0.9925,0.9977,RMSEP were 0.365,1.99,1.13,1.93,0.761,respectively.It can be seen that the established quantitative analysis model can quantitatively analyze the content of the counterfeit objects in the matcha,and the accuracy and stability of the model can meet the general testing requirements.