Classification of food, beverages and perfumes by WO3 thin-film sensors array and pattern recognition techniques

We have designed and fabricated a sensors array using WO3 thin films operat ing at the temperature of 180 degreesC for food and perfume analysis purpos es. The sensing surface of the rf sputtered WO3 thin films has been differe ntly activated by evaporated Pd. Au, Bi, Sb catalysts. Pattern recognition (PARC) techniques as principal components analysis (PCA) and cluster analys is (CA) have been used in order to classify and identify different classes of flavours samples such as olive oils and seed oils, fruit juices, tomato sauces, perfumes. Good separation among flavours samples belonging to homog eneous and comparable classes has been obtained by PARC methods. The headsp ace sampling has been referenced off-line by qualitative chromatograms of t he volatile compounds constituent the odour under test. The preliminary res ults show the feasibility of the electronic nose system to discriminate var ious flavours for food analysis and perfumes quality control. (C) 2001 Else vier Science B.V. All rights reserved.