Interpreting signals from an array of non-specific piezoelectric chemical sensors

Abstract

Pattern recognition methods were tested using a gas sensor array consisting of eight interchangeable quartz crystals coated with different sorbent layers. The system is designed in such a way as to allow headspace sampling from a jar or vial. Using a two-sample t-test the two sampling methods were found to be significantly different at the 95% confidence level. It was found that the application of principal component analysis, multivariate analysis of variance and discriminant function analysis to the gradient of the initial responses and over-all response magnitudes allowed hexane, o-xylene, toluene, dodecane and tetradecane to be distinguished.