Improving the performance of gas sensors for electronic noses using zeolites as selectivity modifiers
The demanding problem of lack of selectivity in semiconducting oxide gas sensors was addressed by the combination of different technologies: shape and size selective catalysts (zeolites), chromium-titanium oxide sensing material and multi-electrode sensor arrays. Sensor devices were fabricated with additional shape and size selective catalysts (zeolites) of three different types (ZSM-5, (3 and Y) either printed over the top of the sensing material or admixed with it. The shape and size selectivity of the zeolite sensors has been demonstrated in their ability to discriminate a range of volatile organic, flavour and fragrance compounds of different molecular size and shape. To promote the selective catalytic activity of the sensors the zeolites were catalytically modified (ion exchanged) by the controlled addition of chromium catalysts. The modification increased their catalytic activity in a controlled way allowing them to selectively crack longer species into smaller ones without resulting in complete combustion of the analytes. Arrays of these sensors have been used in an electronic nose with small number of sensors and a sufficiently large variance in the response for reliable and repeatable discrimination of gases. The discrimination was as good as, and in some cases better than, that achieved using much larger sensor arrays, while also allowing additional discrimination of gases on the basis of their reactivity. Investigation of different parameters such as the chromium loading of the zeolites, the thickness of the zeolite layer, and the sensor operating temperature, allowed further improvements in the discrimination between the different tested compounds. Computational simulations of the interactions of a range of volatile organic molecules with different zeolites were performed using a commercially available software to select promising zeolite catalyst materials for the construction of new sensors and to compare computational predictions with experimental results.