The gas sensors array approach to monitoring and control of air quality

Indoor environments are composed by exogenous and endogenous compounds. In absence of proper filtering and conditioning of air, the quality of indoor air is necessarily worse than the external one. In addition to outdoor sources, a number of volatile compounds have an indoor origin. They are exhaled from furniture, wall tapestries, masonry materials, eventual animals, moulds and fungi. Last but not least, the presence of humans is a further source of volatile compounds contributing to the overall composition of the air. Some of these volatiles are known to be harmful, for instance nitric acid (HNO3), ozone (O3), nitrogen oxides (NOx), hydrogen chloride (HCl), sulphur dioxide (SO2), ammonia (NH3), formaldehyde (H2CO), acetic acid (C2H4O2). Working places are characterized by additional compounds, often harmful, related to the specific chemical and physical processes. In this context, it is important to control these compounds indoor, to protect the workers, and outdoor, to protect the nearby citizens. It is also important to consider that many compounds, which may be present as traces, may be dangerous either as consequence of acute or long-time exposures. The standard approach to the measure and monitoring of air quality consists in the application of a number of analytical techniques aimed at decomposing the air in its basic components and to determine quantitatively the amount of each (analysis in Greek means decomposition). Rapid and distributed detection methods should be based on low-cost, highly integrated devices such as electronic sensors. However, the selective detection of known harmful compounds should require the development of specific devices. Specificity requires the development of dedicated chemical receptors. Such a development requires vast efforts for each compound, and given the large kinds of molecules of interest it can hardly result in distributed and low-cost devices. Furthermore, selectivity is hardly combined with other fundamental requirements for instance the reversibility. Being most of the interactions based on thermodynamics equilibrium, specificity automatically means strong, and then not reversible, bindings. In the last two decades an alternative methodology to approach the analysis of chemically characterised samples was introduced. This method is based on the nonselective character of many solid-state chemical sensors. Surprisingly, this feature, detrimental from the analytical point view, defines an analogy between a set of nonselective sensors and the receptors of the animal olfaction [1]. This analogy is based on the fact that both artificial sensors and natural receptors are non-selective in the sense that each receptor is sensitive to more odorant compounds and each compound is sensed by more receptors [2]. Thus, the identification of an odour is not determined by the signals of a single receptor, but from the combination of signals of all the receptors. This is called combinatorial selectivity. Once the link was established a growing number of researchers worked to develop this concept towards artificial systems able to mimic some functions of the human olfaction [3]. Efficient gas sensor arrays are made of broadly selective sensors but with different sensitivity parameters. It is important to remind that modern sensors are electronic devices that create signals (analogue or digital) that carry information about the sensed compounds. Such devices are logically made up of two main components: the sensing material (or receptor) and the transducer. Gas sensor arrays are typically based on the same transducers but carrying different receptors. In this paper a different approach is presented. It is based on the integration of different sensors technologies gathered together in order to expand t