Microsensor Arrays

A well-known problem of tin dioxide is its lack of selectivity (Chap. 2). This situation is usually dealt with by using an array of sensors in combination with multi-component or pattern recognition algorithms such as principal component regression (PCR), multi-way analysis or artificial neural networks (ANN) [142, 143]. Doping of the tin dioxide also changes its selectivity characteristics to different gases [68]. Another parameter that can be varied is the operation temperature. The use of an array of microhotplates with individually controlled temperatures, the hotplates of which are covered with different sensitive materials, increases the overall information that can be extracted from metal-oxide-based gas sensing systems.