Detection and discrimination of noxious gases using metal oxide semiconductor gas sensor arrays with pattern recognition techniques

Gas sensing properties of micro-electronic MOS gas sensor arrays were characterized for the noxious gases, carbon monoxide (CO), nitrogen oxides (NO_X), ammonia (NH₃) and formaldehyde (HCHO), and the sensor responses were investigated for the discrimination of gas species by using pattern recognition technique. In three binary mixed gas systems with oxidizing and reducing gases, the gas sensing behaviors and sensor responses were examined. The gas sensing behaviors to the mixed gas systems suggested that specific adsorption and selective activation of the adsorption sites might offer the priority for the adsorption of specific gas in gas mixtures. Gas sensitivity data were extracted and their patterns were recognized with PCA techniques. The PCA results showed good separation among the mixed gas systems, and it suggested that the gas mixture tests could be well classified and discriminated in the noxious gases and their mixtures.