Analysis of VOCs with a tin oxide sensor array

Abstract

Using a sensor array of 15 thin film tin oxide sensors, both the single-component classification and the multicomponent analysis of volatile organic compounds (VOCs) have been carried out. The classification has been accomplished through the techniques of principal component analysis (PCA) and artificial neural networks (ANNs). The multicomponent analysis has been carried out in two stages: first, linearization of the responses, secondly, multivariate linear regression. Four multivariate (MVA) regression methods have been used: classical least squares (CLS), inverse least squares (ILS), principal component regression (PCR) and partial least squares (PLS). The PCA classification permitted to distinguish three families of VOCs: aliphatic and aromatic, chlorinated and oxygenated compounds. ANNs classification discriminated six VOCs gases with a success rate of 71%. The best results from the multicomponent analysis were obtained for the ILS and PCR methods.