Cross-flow filtration is an increasingly common post-fermentation process in the wine industry. Because of the nature of the membranes used in this process, our hypothesis was that cross-flow filtration would not have a significant impact on the sensory or chemical properties of either white or red wines.

To investigate this, a California white wine blend and a California red wine blendwere filtered in three 570 L lots using a Bucher Vaslin cross-flow unit with a nominal 0.22 micron polyethersulfone membrane. The unfiltered control was sent directly to the bottling line without filtration.

Panelists evaluated the wi nes nine times during an eight-month descriptive analysis panel with replicated tasting at each time point. UV-VIS spectrophotometry and the Adams-Harbertson assay were used to determine color and phenolic content of the filtered and unfiltered wines respectively.

The effect of filtration was found to be significant for one sensory attribute out of 16 total measured in the white wine. Similar analysis for the red wines indicated that six sensory attributes were significantly different out of 16 total in the red wine.

Unfiltered red wines were found to be higher in earthy, grassy, oak, and smoke aromas compared to filtered wines and lower in mixed berry and stone fruit aromas compared to the filtered wines, but only after two months in bottle.

Cross-flow filtration was found to have a stabilizing effect on the sensory profile of both wines. For both red and white wines, there were significant changes in color and phenolic profile with filtration, but it is not clear, especially for red wines, that the changes in phenolic compound concentration were large enough to be detected sensorially.

(We recommend that you consult the full text of this article)