Noise spectroscopy data analysis-based gas identification with a single MOX sensor

Air quality monitoring and analysis have become a major issue in recent years. Metal oxide (MOX) gas sensors are very sensitive, due to high variability of their resistivity in presence of gas. However, they are not selective, i.e. it is not possible to determine the nature and concentration of the gas using only variations in the resistance of the sensor. Noise spectroscopy is one of the solutions to improve selectivity. In this paper, we evaluate recent noise spectroscopy-based gases identifications methods, to distinguish the nature of different gases using a single MOX sensor. From noise measurements performed on MOX gas sensors with tungsten trioxide (WO3) sensing layer, under several nitrogen dioxide (NO2), ozone (O-3) and carbon monoxide (CO) concentrations in dry air, we have created a database. This database is increased by extracting the spectral attributes of noise responses, and then reduced by using principal component analysis to extract only useful information. The results obtained in this study demonstrated that it is possible to distinguish three air quality gases with only one single WO3 sensor.

Automated summary

Metal oxide (MOX) gas sensors are sensitive but not selective. Noise spectroscopy is a solution to improve selectivity. This study successfully distinguished the nature and concentration of three different gases using noise spectroscopy-based gas identification methods.