Trends Towards An Electronic Nose-Electronic Tongue Fusion Applied to Discrimination of Wines

LANGENHEIM M.; LOVINO M.; MONGE M.E.; MIZRAHI D.; NEGRI M.R.

Praga

Workshop; I Czech-Argentine Biennale Workshop ?e-Golems?: Interdisciplinary aspects of human-machine co- existence and co-operation; 2005

Wines of different wineries and different varietals were discriminated by using a combined approach of a gas sensors array (electronic nose) and liquid phase sensors. Different commercial Argentine wines of two varietals, Malbec and Chardonnay, were analyzed. The gas phase of the wines was analyzed with an electronic nose developed in our laboratory. The liquid phase was analyzed by pH, electrical conductivity and PVC-based membrane potentiometric measurements. The whole data set was studied with different combinations of the sensors used such as e-nose, e-nose + conductivity, e-nose + conductivity + pH, etc. These data were examined with two multivariate methods: principal component analysis (PCA) and cluster analysis (CA). The best discrimination was obtained when the sensors signals corresponding to the measurements of both gas and liquid phases were jointly analyzed. PCA plots showed a clear qualitative discrimination of the wines, and the cluster analysis correctly classified thirty three over a total of thirty four samples. The complete analysis showed that only the simultaneous analysis of gas and liquid phases provided a relatively simple, effective and fast discrimination of wines of different wineries and different varietals, in contrast to the individual e-nose or e-tongue analysis.