Published online by Cambridge University Press: 07 December 2007
Introduction. Aroma is often used as a fruit quality marker. Several researchers have used the analysis of volatile compounds to characterize dates (Phoenix dactylifera). Nevertheless, it is difficult to analyze this subtle aroma in spite of the many techniques available, such as liquid-liquid extractions and headspace analysis. The global approach of the e-nose with a minimum handling of the sample is a fast and efficient alternative to these techniques. This rather new tool, simulating the behavior of the human nose, thus aroused our interest for application to dates. Materials and methods. Dates were harvested, in 2005, at the Tamar stage from nine date varieties, in Zagora, Morocco. They were frozen after harvest and sent by air to the laboratory of CIRAD (Montpellier, France). All data were collected at the same time. Five hundred g of dates were taken at random, pitted and cut into small pieces. After homogenization, an aliquot of 3.0 g was put into a 10-mL headspace flask and immediately sealed. Six replicates per sample were prepared for later electronic-nose analysis. An electronic nose (e-nose) from Alpha MOS, equipped with Metal Oxide Sensors (MOS), was used. Data were analyzed using Principal Component Analysis (PCA). The Discrimination Index (DI), created by Alpha MOS, was used to describe the discriminatory power of the groups’ separation. Alpha MOS considers that good discrimination and reliability are achieved for DI over 80%. Results. Discrimination indexes > 88% and means of the coefficients of variation obtained for each sensor < 4% demonstrate the high performance and good reproducibility of the e-nose. Conclusion. The e-nose is sensitive to date volatile emissions. All the date varieties studied were well separated. This fast screening technique, creating a specific olfactory fingerprint as a genuine identity card, could be used to follow changes in maturity, assure the identity of lots of dates and detect adulteration.