Sensory and nephelometric analysis of tannin fractions obtained by ultrafiltration of red wines
Abstract
The assessment of red wine mouthfeel relies primarily on the sensory description of its tannic properties. This evaluation could be improved by gaining a better understanding of the physicochemical properties of these tannins. Hence, the objectives of the present study were threefold: (1) to gain an insight into the sensory properties of subpopulations of proanthocyanidic tannins of different molecular sizes obtained through several ultrafiltration steps, (2) to quantify the kinetics of haze formation of these proanthocyanidic tannins in a dynamic polyvinylpyrrolidone (PVP) precipitation test, (3) to determine whether a correlation exists between the sensory and the precipitation data. Two wines from different grape varieties, gamay and merlot from Switzerland, were ultrafiltered to provide four tannic fractions differing by the range of their polymerization degrees. Then, these fractions were added (individually or in combination) into their native wine matrix according to a specific experimental design. The reconstituted wines were then characterized by PVP precipitation tests using a dynamic nephelometer and by sensory assessment following two methods, Quantitative Descriptive Analysis, and Temporal Dominance of Sensations (TDS). For the sensory test, a panel of twenty trained judges generated a list of seven mouth-feel descriptors. Wines containing the fraction with the largest size tannins (range of highest polymerization degrees) were perceived as being more astringent and mouth-drying after spitting. Wines containing with the smallest size tannins added (range of lowest polymerization degrees) were perceived as more voluminous. With regards to the TDS, wines containing the largest tannins fraction were the most intense and mouth-drying, and the astringency and mouth-drying perceptions were the most persistent attributes. Wines containing the fraction with the smallest tannins were judged to be the most voluminous, bitter and sour. Finally, a positive correlation, RV=0.58, p=0.001, between the results of the precipitation test and the Quantitative Descriptive Analysis could be established.
Issue: Macrowine 2016
Type: Poster
Authors
*HES-SO