Discrimination of powdered herbal teas by Vis/NIR spectral reflectance and chemometrics

Abstract

Abstract The herbal tea market is projected to grow at an annual rate of 4.8 %, with the discrimination of these products appearing as an issue of food quality and safety. In this study the Vis/NIR spectroscopy combined with chemometrics was applied for discriminating five popular herbal teas (chamomile, boldo, lemon grass, carqueja, fennel) by using powdered samples. Dynamic sampling was applied for measuring the spectral signatures and different spectral pre-treatments were evaluated aiming at improving the discrimination accuracy. The Partial Least Squares Discriminant Analysis (PLS-DA) achieved high prediction accuracies (77.8–100 %), specificities (89.4–100 %) and sensitivities (66.1–100 %), with detrending and object-wise standardization pre-treatments correctly discriminating 100 % of the samples during the external validation. The Vis/NIR spectroscopy combined with chemometric analysis has great potential to discriminate powdered herbal teas, providing a non-destructive, fast, safe and chemical-free solution for automated quality control procedures in industries of tea processing.