Influence of wine components on the chemical and sensory quality of wines

Charles Diako

Wine is an alcoholic beverage containing numerous compounds that contribute to its overall quality. The overall objective of this dissertation was to examine the influence of matrix interactions in commercial red wines on the sensory and chemical properties of wines, and explore advanced methods of mathematical analyses of these data. In the first study, the influence of these matrix components and their interactions on wine quality was examined. Commercial Merlot wines (n=61) were evaluated for wine chemistry parameters, with the matrix components of interest identified as alcohol, tannin and mannoproteins. Sensory evaluation results showed complex interactions among tannins, alcohol and mannoproteins on the perception of most aromas, flavors, tastes and mouthfeel attributes (p<0.05). Since human subjects vary in their sensory perceptions, panel reproducibility and precision was addressed. Panelists’ variation in attribute evaluations was conceptualized as a linear operator called a bias matrix for feedback calibration and correction of panelists’ evaluations. Results showed that the bias matrix corrected the panelist ratings of the samples, leading to higher reproducibility and precision. No significant differences (p>0.05) were found between the original and corrected means. Predictive filtering showed that the panelists’ corrected means for the attributes were closer to the predicted panel mean compared to their unfiltered means. The application of the electronic tongue for the assessment of wine quality was further explored. Strong correlations (r2>0.930) were reported between the electronic tongue and the sensory perceptions of sweet, sour, bitter, burning, astringent and metallic. Further research on the application of the electronic tongue to discriminate among wines and building predictive models was performed. Non-linear methods showed high discrimination among the commercial wines (90.1%), with high prediction accuracy of the electronic tongue output using the chemical parameters (90.0%). These results showed the dependence of the intensity of the electronic tongue signal on the chemical components of the wines. This dissertation demonstrated how wine matrix components influenced perception through suppression and enhancement of various sensory attributes. In addition, advanced methods of chemical and sensory data analysis were developed and validated. Results from this study will be useful for winemakers for wine quality optimization.

Influence of wine components on the chemical and sensory quality of wines

Files and links (1)

Abstract

Metrics

Details