Classification of complex gas mixtures from automotive leather using an electronic nose

A semiconductor gas sensor array combined with a routine for pattern recognition - a so-called electronic nose - for the detection of gas emissions from the leather used in car compartments is described. The gas sensors are 10 metal oxide semiconductor field effect transistors (MOSFETs) with gates of thin, catalytic metals, and five semiconducting metal oxide sensors. The sensor array data are processed by multivariate means using principal component analysis (PCA) and are shown to give similar and add additional information compared to gas chromatography-mass spectrometry (GC-MS) and a human sensory panel. The total volatile organic compound concentration as measured by GC did not differ between good and bad samples and could therefore not be used as a quality control tool, whilst the electronic nose together with pattern recognition could readily discover the deviating samples with unusual emitting gases. This set-up could be useful in on-line quality monitoring systems to detect anomalies in incoming car interior trim materials. (C) 2000 Elsevier Science B.V. All rights reserved.