Evolution of sugars in cider brandy aged in oak barrels: a contribution to its characterization

A chemometric study was carried out to typify cider brandies according to the type of wood employed in the maturation process and their aging time. Monosaccharides, previously derivatized with p-aminobenzoic ethyl ester, were analyzed using a reversed-phase high-performance liquid chromatographic method. Univariate data treatment was not sufficient to enable differentiation of the classes of cider brandies on the basis of wood type and maturation time. Two linear combinations of original variables, ascertained by principal components analysis, provided an adequate data structurization. A mathematical decision rule was established to classify cider brandies with prediction capacities of 92 and 97% using an LDA method and Bayesian analysis, respectively. The use of the PLS algorithm allowed the authors to differentiate cider brandies according to the age and type of oak used in the aging process.