A Noise Spectroscopy-Based Selective Gas Sensing with MOX Gas Sensors

Abstract : Communicated by Zoltan Gingl We propose a new method for obtaining a °uctuation-enhanced sensing (FES) signature of a gas using a single metal oxide (MOX) gas micro sensor. Starting from our model of adsorption-desorption (A-D) noise previously developed, we show theoretically that the product of frequency by the power spectrum density (PSD) of the gas sensing layer resistance °uctuations often has a maximum which is characteristic of the gas. This property was experimentally con¯rmed in the case of the detection of NO 2 and O 3 using a WO 3 sensing layer. This method could be useful for classifying gases. Furthermore, our noise measurements con¯rm our previous model showing that PSD of the A-Dnoise in MOX gas sensor is a combination of Lorentzians having a low frequency magnitude and a cut-o® frequency which depends on the nature of the detected gas.