Phytochemical analysis and antioxidant activity of eight cultivars of tea (Camellia sinensis) and rapid discrimination with FTIR spectroscopy and pattern recognition techniques

Abstract

Background: Tea (Camellia sinensis L.O. Kuntze) is one of the most commonly consumed beverages globally, with several beneficial health effects. The composition of leaves is affected by different factors such as climate and tea cultivar. Methods: In this study, the total phenolic, flavonoid, and tannin contents, and antioxidant activities of eight cultivars of tea growing in Iran were determined. The epigallocatechin gallate (EGCG) contents were measured by high-performance liquid chromatography (HPLC) analysis. The Fourier-transform infrared (FT-IR) spectra were used to construct a supervised pattern recognition model using a genetic algorithm-support vector machine (GA-SVM) for the classification of cultivars. Results: The results indicated a wide variation of total phenolic content (178.04 to 316.09 mg gallic acid equivalents (GAE)/g extract). Also, the flavonoid contents ranged from 25.54 to 41.1 mg quercetin equivalents )QE(/g extract. All the samples had close tannin amounts (ranging from 40.32 to 45.90 mg as tannic acid equivalent (TAE)/g extract). There was a significant linear relationship among total phenolic and flavonoid contents and antioxidant properties. The cultivars DN, PK2, and C.Y.9 had the highest content of phenolic and flavonoid content as well as the best antioxidant activity. The EGCG contents were from 2.66 to 4.12%. The highest amount of EGCG was found in cultivars 282, PK2, and C.Y.9. The discriminative region of FT-IR spectra (1350-1650 cm-1) were selected using a GA-SVM. This model showed 100% sensitivity and specificity for training and test sets. Conclusion: The leaves phytochemical compositions and antioxidant effect are deeply dependent on the type of tea cultivars. The cultivars PK2 and C.Y.9 can be considered richer sources of polyphenols, especially EGCG. The eight different cultivars can be classified based on chemical components using the recorded FT-IR spectra.