Variation of some fermentative sulfur compounds in Italian "millesime" classic sparkling wines during aging and storage on lees.

Fermentative sulfur compounds are recognized as strongly affecting wine aroma, mainly as off-flavors, but recently also as possible positive contributors to wine quality and typicity in still wines. Nevertheless, no evidence has been provided for the influence of these molecules on sparkling wine aroma, except for peculiar volatile thiols found in French Champagne. According to the traditional method, the second fermentation, occurring in sealed bottles, is the essence of the procedure. After this fermentation, sparkling wine is aged on yeast lees for a period ranging from a few months to several years so that yeast autolysis can take place. So far, no evidence is provided for the effect of yeast contact duration on the level of sulfur compounds. Following a HS-SPME/GC-MS method, 13 sulfur compounds, that is, ethylmercaptan, dimethyl sulfide, diethyl sulfide, dimethyl disulfide, diethyl disulfide, methyl thioacetate, ethyl thioacetate, 2-mercaptoethanol, 2-(methylthio)-1-ethanol, 3-(methylthio)-1-propanol, 4-(methylthio)-1-butanol, benzothiazole, and 5-(2-hydroxyethyl)-4-methylthiazole, were quantified in several Italian sparkling wines, produced according to the traditional method in two wineries from Trentino-South Tyrol, region. Additionally, the influence of winemaking technology differences and vintage effects on the evolution of the quoted sulfur compounds was considered. This investigation was carried out by coupling the HS-SPME data with those obtained by SPE method and relevant to other volatile compounds, which are considered as winemaking markers. This work furnished the first evidence of the effect of aging and lees contact at different storage temperatures on the levels of these analytes in sparkling wines. Significant increments were observed for dimethyl sulfide, diethyl sulfide, 2-(methylthio)-1-ethanol, 3-(methylthio)-1-propanol, and 4-(methylthio)-1-butanol during aging with a different variation slope possibly due to the remarkably different storage temperatures. No clear influence of lees contact duration was found for the majority of the sulfur compounds considered.