Analysis of tomato taste using two types of electronic tongues

Abstract In this study the potential of two types of electronic tongues as rapid techniques to analyze taste is evaluated. The first electronic tongue was developed at the University of Saint-Petersburg and comprises of 18 potentiometric sensors. The second electronic tongue was the ASTREE electronic tongue developed by Alpha M.O.S. (Toulouse, France) which consists of a set of seven sensors which are commercially available. Six Belgian tomato cultivars were classified according to similarity in taste profile using both multisensor systems. The tomato cultivars were selected based on their difference in sweetness and sourness as perceived by trained sensory panels. The concentration of sugars (glucose and fructose), organic acids (citric acid, malic acid and glutamic acid) and minerals (Na and K) were also determined with reference techniques. Multivariate statistical data analysis techniques as principal components analysis (PCA), canonical discriminant analysis (CDA) and partial least squares regression (PLS) were used to classify tomato cultivars according to similarity in taste profile and to quantitatively relate the taste compounds to the sensory panel scores. Both electronic tongues were very well suited to classify tomato cultivars based on their taste profile. To quantify individual sugars, acids and minerals in a complex mixture the system which was developed at the University of Saint-Petersburg was highly appropriate, but this system could not predict general sweetness and umami taste as evaluated by the sensory panel. The ASTREE electronic tongue on the other hand was suitable to quantify glutamic acid and Na, but the sensor readings were poorly correlated to the sweetness, sourness, saltiness and umami in tomato as tasted by the sensory panel.