The application of discrimination technique to alcohols and tobaccos using tin-oxide sensors

Abstract After a brief introduction to the mechanisms involved in semiconductor gas sensors, questions of linearity and superposition are considered for mixtures of gases. It is established that Taguchi gas sensors (TGS) can provide a good approximation to a linear model and that superposition principles apply under certain conditions. This greatly facilitates the analysis of data and the identification of constituent components in a mixture of gases. Innovative techniques have been applied to alcohols and tobaccos with conspicuous success. These relate to pattern recognition and cluster separation, using simple mathematical procedures to enhance the identification of closely similar batches of test samples. It is confidently expected that many other applications for these methods will arise.