Universitätspublikationen
Refine
Year of publication
- 2018 (2)
Document Type
- Article (2)
Language
- English (2) (remove)
Has Fulltext
- yes (2) (remove)
Is part of the Bibliography
- no (2)
Keywords
- wine (2) (remove)
Institute
A full understanding of the origin, formation and degradation of volatile compounds that contribute to wine aroma is required before wine style can be effectively managed. Fractionation of grapes represents a convenient and robust method to simplify the grape matrix to enhance our understanding of the grape contribution to volatile compound production during yeast fermentation. In this study, acetone extracts of both Riesling and Cabernet Sauvignon grape berries were fractionated and model wines produced by spiking aliquots of these grape fractions into model grape juice must and fermented. Non-targeted SPME-GCMS analyses of the wines showed that several medium chain fatty acid ethyl esters were more abundant in wines made by fermenting model musts spiked with certain fractions. Further fractionation of the non-polar fractions and fermentation of model must after addition of these fractions led to the identification of a mixture of polyunsaturated triacylglycerides that, when added to fermenting model must, increase the concentration of medium chain fatty acid ethyl esters in wines. Dosage-response fermentation studies with commercially-available trilinolein revealed that the concentration of medium chain fatty acid ethyl esters can be increased by the addition of this triacylglyceride to model musts. This work suggests that grape triacylglycerides can enhance the production of fermentation-derived ethyl esters and show that this fractionation method is effective in segregating precursors or factors involved in altering the concentration of fermentation volatiles.
In this study, a portable electronic nose (E-nose) prototype is developed using metal oxide semiconductor (MOS) sensors to detect odors of different wines. Odor detection facilitates the distinction of wines with different properties, including areas of production, vintage years, fermentation processes, and varietals. Four popular machine learning algorithms—extreme gradient boosting (XGBoost), random forest (RF), support vector machine (SVM), and backpropagation neural network (BPNN)—were used to build identification models for different classification tasks. Experimental results show that BPNN achieved the best performance, with accuracies of 94% and 92.5% in identifying production areas and varietals, respectively; and SVM achieved the best performance in identifying vintages and fermentation processes, with accuracies of 67.3% and 60.5%, respectively. Results demonstrate the effectiveness of the developed E-nose, which could be used to distinguish different wines based on their properties following selection of an optimal algorithm.