Abstract
Background: This work describes a fast, simple, sensitive, and low-cost method for the identification of resveratrol in different brands and varieties of red wines.
Methods: It was developed based on a comparison of the UV-VIS spectra of the samples and samples enriched with different concentrations of the trans-resveratrol standard. The spectra were analyzed by chemometric principal component analysis (PCA) and multivariate calibration.
Results: The PCA data indicated that only 4 main components made possible group samples based on the grape variety characteristics and/or production region.
Conclusion: From the construction of partial least squares (PLS) and multiple linear regression (RLM) models, it was possible to predict the sample trans-resveratrol content with that sample showing similarities between the groups observed in the PCA and the samples used in the model constructions. The predicted trans-resveratrol present in these samples ranged from 0.29 to 23.3 mg L-1. This multivariate method suggested a good predictive capacity of determination of resveratrol concentrations in commercial red wines.
Keywords: Multivariate methods, PCA, PLS, red wine, trans-resveratrol, UV-Vis.
Graphical Abstract
14