This article reports a method for determining organic acids by CE in order to obtain a characteristic fingerprint with a view to classifying white wine samples with the aid of multivariate statistics. The proposed method was optimized for the determination of the most abundant organic acids in wine (viz. succinic, malic, tartaric, citric, acetic and lactic). The acids were separated in a running buffer consisting of 180 mM disodium hydrogen phosphate, 0.5 mM CTAB and 10% methanol, adjusted to pH 7.5. The applied voltage was 10 kV and the working temperature 251C. The precision of the method was assessed by replicate analysis of a wine sample; the resulting RSD was less than 5.2% in peak area and less than 1% in migration time for all organic acids. Calibration curves exhibited good linearity throughout the studied concentration range for all acids. Principal component analysis and linear discriminant analysis allowed accurate classification of wine samples with variable alcohol contents from their electrophoretic fingerprints. The discriminant potential of various electropherogram regions was examined. Based on the results, the proposed method is an effective choice for classifying wine samples.
IntroductionThe winemaking sector is one of the main motors of the economy in Spain. Spanish white wines are high-quality products much appreciated by consumers inside the country and also, increasingly, abroad [1]. Therefore, being able to confirm their authenticity is of paramount importance with a view to assuring their quality to an expanding international market. Multivariate analysis has been widely used for this purpose in recent years.Wines [11,12] or nuclear magnetic resonance spectroscopy [13]. In this context, conventional methods have been supplemented with electronic tongues, noses and other sensor array devices for wine-monitoring purposes [14][15][16][17]. These instrumental techniques have been used in combination with chemometric tools such as artificial neural networks in order to classify white wines [18]. Separation techniques such as CE have also been used to classify wines [19][20][21][22]. To our knowledge, electropherogram profiles, which are constructed from the CE response of organic acids, have never, to date, been used as fingerprints for white wines and their classification with chemometric models. In this study, we used electropherogram fingerprints obtained with CE to classify white wines for the first time.Organic acids present in grape juices and wines are important because they influence the sensory properties (flavour, colour and aroma), and the physical and microbiological stability, of the product. Acetic, lactic, succinic, malic, citric and tartaric are the most abundant organic acids in wines. Tartaric, malic and citric acid are obtained directly from the grape, whereas succinic, lactic and acetic acid form during the fermentation process. These organic acids are also very useful for detecting wine alteration and/ or disease [23].Malic acid is present in large amounts in green gra...