Characterisation of food freshness with sensor arrays

The detection of volatile compounds with chemical gas sensors is of increasing interest in the non-destructive determination of food quality. We have investigated the problem in relation to fish freshness. The relevant compounds which may be monitored in the gas phase and which correlate with the status of freshness are alcohols, carbonyls, amines and mercaptanes with a chain length of up to 9 carbon atoms. These compounds show typical concentration changes over a time period of four weeks under defined storage conditions. In earlier studies of fish freshness mainly gas chromatography (GC) was used to detect low concentrations of these compounds. We now introduce a cheap portable sensor system consisting of an array of amperometric sensors, a heated catalyst and a subsequent 'intelligent' data evaluation. The electrochemically inactive compounds released by the fish were decomposed at different temperatures by a platinum catalyst in front of the sensor array. The chemometric techniques principal component analysis (PCA) and principal components regression (PCR) were used for the identification and quantification of these compounds. For the calibration of the sensor system, test gases were monitored at different concentrations and different catalyst temperatures. Based upon this calibration set, a model was developed and applied to identify gas mixtures and to then characterise the freshness of a trout monitored over several days.