Investigating the Inadvertent Transfer of Vitis Labrusca Associated Odors to Vitis Vinifera Wines

Open Access
Smith, Jared C
Graduate Program:
Food Science
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 14, 2014
Committee Members:
  • Ryan John Elias, Thesis Advisor
  • John E Hayes, Thesis Advisor
  • Gregory Ray Ziegler, Thesis Advisor
  • Denise Michelle Gardner, Thesis Advisor
  • Aroma Scalping
  • Odorant Scalping
  • Polymers
  • Wine
  • Methyl Anthranilate
  • 2-Aaminoacetophenone
Methyl anthranilate (MA) and 2-aminoacetophenone (2AAP) are volatile compounds present in certain Vitis labrusca grapes, such as Concord and Niagara. Reportedly, MA is responsible for the distinct grapey odor of Concord wines and 2AAP is responsible for the “foxy” character of Niagara wines. For the vast majority of Vitis vinifera wines, these odors are not expected to be present at any detectable level; however, in recent years, there have been anecdotal reports in northeastern U.S. wine regions of these odors appearing in wines prepared solely from V. vinifera varieties. MA and 2AAP also have low sensory detection thresholds (300 µg/L and 0.5-2 µg/L, respectively) therefore even trace levels of these compounds in V. vinifera wines are perceptible by human assessors. The sources and modes of carryover of foxy/native odors into V. vinifera wines is currently unknown, and we hypothesizes that the major mechanism by which V. vinifera wines accumulate V. labrusca-associated character is through scalping by winery equipment during juice and wine processing. As it is common for V. labrusca and V. vinifera varieties to be processed in the same facilities in the northeastern U.S., certain materials used during winemaking may be capable of transferring these volatile compounds between wines and juices. In order to test this, I developed procedures to access the ability of various polymeric materials widely used during wine production to scalp MA and 2AAP from spiked model solutions (100mg/L). Following scalping, I subjected those resins to un-spiked model solutions, measuring the desorption of the odorants over a 144h period. I then tested the polymers that transferred MA and 2AAP in model solutions in real wine and juice, using spiked (5mg/L) Concord wine and juice in the scalping study, and V. vinifera wine (Chablis) and juice (Muscat) to investigate desorption. I found MA and 2AAP were scalped from both model and real systems at various rates by polymeric materials relevant to winemaking. I also found that scalped MA and 2AAP desorbed from polymers upon introduction to new solutions. I found that the levels of scalping and desorption was largely dependent on the structural characteristics of the polymers. PVC and rubber allowed for significantly (p<0.05) larger amounts of MA and 2AAP to be transferred between solutions in comparison to HDPE, LDPE and polysulfone. This likely results from the higher amorphous content of PVC and rubber which allows for greater sorption and desorption of the odorants. I also found that polymers containing a higher degree of crystallinity, like LDPE and HDPE, led to transfer of the odorants to a lesser extent, while polysulfone failed to transfer any MA or 2AAP between solutions. I observed no definitive trends based on the solution (wine or juice) MA and 2AAP were being scalped from or desorbed into. This is likely due to the wines weakly ethanolic (12%) content which does not drastically affecting the polarity of the solutions and varying effects of components of the systems acting as either copermeants or competitors during sorption and desorption of 2AAP. Copermeants are molecules that increase the levels of sorption of other molecules, while competitors do the exact opposite as their presence minimizes the amount of sorption of the analyte of interest. I observed the same trends for all of the tested polymers between model and real systems. These results suggests that the model study was a relatively successful means for screening the majority of the polymers in terms of their ability to transfer MA and 2AAP between real wines and juices. In addition, polymeric materials may act as a vehicle for the inadvertent introduction of native odorants into wines produced solely from V. vinifera varieties. Further investigation should be carried out to confirm the validity of these findings in an actual wine production environment. I also tested the ability for various cleaning solutions affiliated with commercial wine production to remove MA and 2AAP from polymers as potential means to prevent this inadvertent transfer in wineries that use the same equipment for V. labrusca and V. vinifera processing. As I expected, increasing both ethanol content and temperature of the cleaning solutions lead to greater desorption of MA and 2AAP from the polymers. Even at the highest ethanol content (80%) and higher temperature (75°C) treatment only up to 40% of the odors were desorbed after 60 minutes of exposure to the cleaning solution. This suggests that while these odorants can be desorbed, the exact protocol for removing the compounds needs to be optimized in order to ensure that they are removed at levels which render the polymer suitable for processing V. vinifera varietals. Lastly, I performed human sensory evaluation in order to determine the rejection threshold for 2AAP in wines. I found that even at the highest concentration tested, which was almost 50 times above its detection threshold (.5 µg/L), the spiked 2AAP wines were not significantly (p>0.05) preferred over the un-spiked control. This suggests that 2AAP may not be offensive to certain common V. vinifera wine consumers in certain regions of the country. However, other factors need to also be considered when interpreting this result as the presence of 2AAP in V. vinifera wines may affect preference in other populations.