Extractable and bound tannin in Merlot fruit and its partitioning during fermentation

Tannin extraction during winemaking has been a concern in the wine industry for quite some time due to its influence on astringency, aging, and color stability in wine. This concern usually focuses on the inability to extract enough tannin during fermentation and the winemaker's understanding of how to optimize or normalize tannin extraction. The problem of tannin extraction hinders a winemaker's ability to predict tannin levels in finished wine based on the tannin levels at harvest, and also affects winemaking decisions during fermentation. It has been found that not all of the tannin in the fruit can be accounted for in the finished wine. The unaccountable tannin that is missing is thought to be bound to the cell wall matrix of the fruit during fermentation, and therefore not extracted with further winemaking practices. Recent studies that accounted for the tannin that is bound to the cell wall matrix (the bound tannin) found a recovery range of 70% to 108% for the missing tannin from the fruit to the wine, on the commercial scale.;A small experimental fermentation was performed in triplicate on Merlot fruit from Oakville, CA at the UC Davis pilot winery. Samples of fruit, pomace, lees, and finished wine were sampled and volumes and weights throughout fermentation were recorded to account for all of the tannin from the fruit to the finished wine. Fruit was sampled prior to primary fermentation, pomace samples were collected after pressing, and lees and wine samples were collected after malolactic fermentation finished and the wine was racked. Fruit samples (separated into skins and seeds), pomace along with pomace seeds separated, and lees samples were ground separately in liquid nitrogen, and extracted with 70% acetone to account for "extractable" tannin analysis by the Adams-Harbertson Tannin Assay. Wine samples were also analyzed by this method. The insoluble solids filtered after acetone extraction were used to account for "bound" tannin by a common acid-butanol method for the analysis of proanthocyanidins. With both extractable and bound tannin measurements, 98.5% of the tannin in the fruit was accounted for in the fermentation products. Therefore, accounting for both extractable and bound tannin on a small scale level allows for all of the tannin from the fruit to be accounted for throughout fermentation and in the finished product.;Carbon-13 Cross Polarization Magic Angle Spinning Nuclear Magnetic Resonance (CPMAS NMR) was used to confirm the identity of the bound tannin and characterize its linkage to the cell wall material. It was found that the tannin remaining in the samples after acid-butanol treatment was condensed tannin by the peak at 155ppm. Mesocarp of the fruit was found to have no bound tannin and could thus serve as a reliable control for analysis of grapes. It was noted that the peak at 155ppm became broader and shifted upfield slightly after the acid-butanol treatment. The linkage of condensed tannin to the cell wall matrix is not yet known but it is suggested that this may be an ether linkage.