Identification of typical wine aromas by means of an electronic nose

In the field of electronic noses it is not very usual to find many applications in wine detection. Most of them are related to discrimination of wines in order to prevent their illegal adulteration and detection of off-odours but their objective is not the identification of wine aromas. In this paper, an e-nose using headspace as an extraction technique is used for the identification of typical aromatic compounds present in white and red wines. The descriptors of these compounds are: fruity, floral, herbaceous, vegetative, spicy, smoky, and microbiological and they are responsible for the usual aromas in wines; concentrations differ from 2 to 10 times the threshold concentration humans can smell. Some of the measured aromas are: pear, apple, peach, coconut, rose, geranium, cut green grass, mint, vanilla, clove, almond, toast, wood and butter. Principal component analysis (PCA) shows datasets of this group of compounds are clearly separated and radial basis neural networks (RB-NN) show a 98% rate of success in classification.