Xylovolatile fingerprint of wines aged in barrels or with oak chips

Aim: In this research xylovolatile aromatic compounds were determined to highlight any compositional differences between wines aged in barrels and those produced using oak chips. Methods and results: Approximately 200 wines aged using oak chips or wood barrels were analysed, among which about 50 were aged at the experimental winery of the Research Centre of Viticulture and Enology in Asti (Piedmont, Italy). 17 different types of commercial oak chips were used in order to obtain a subset of reference samples. Several factors were considered, including degree of oak wood toasting, wood geographical origin and the type of wine (both red and white). Sample preparation for GC-MS analysis was performed by singlestep solid phase extraction, using polymeric SPE cartridges. More than 60 volatile molecules from oak were identified. Analytical results were explained using multivariate statistical analysis. A preliminarily step was carried out using principal component analysis (PCA), which showed interesting compositional differences between barrel and chip-aged wines with regards to methylvanillate, ethylvanillate, as well as furan derivative compounds. Further statistical analysis highlighted the clear impact of the type of wine (red or white) on the extraction of xylovolatile compounds from oak chips. Subsequently, in order to test if and how the selected chemical explanatory variables allow wine treatments to be discriminated, and to predict which group a new observation will belong to, a discriminant analysis (DA) was carried out on an independent dataset. More than 96 % of samples were correctly classified. Conclusions: Significant differences were highlighted between barrel- and chip-aged wines with regards to xylovolatile compounds. The analysis showed that several factors may influence the amount of aromatic compounds extracted from chips and barrels, especially the matrix composition. It was possible to deduce the aging process that the wines had undergone by selecting key molecules using multivariate methods. Significance and impact of the study: Via a suitable GC-MS method and a chemometric approach for the identification of discriminant xylovolatiles in wood aged wines, this study offers promising perspectives and useful tools for routine fraud inspection aiming at classifying wines according to their aging process.