Predictive gas sensor based on thermal fingerprints from Pt-SnO2 nanowires

The detection of volatile compounds is important for a broad variety of applications. Metal oxide gas nanosensors are tiny, inexpensive devices that can be integrated into any application, but they lack selectivity. On the other hand, electronic noses consisting of sensors arrays comprised of different active materials are complex as well as expensive to fabricate and use. This paper presents a novel approach using Pt-decorated tin oxide (SnO2) nanowires at different working temperatures to produce a virtual sensor array exploiting the thermal fingerprints of the different gases. With only one nanostructured material (Pt-SnO2) and 5 temperature values, the system could qualitatively and quantitatively discriminate all the gasestested (all reducing gases). The sensor could detect selectively which gas is present (with an accuracy of 100%) at what concentration (with an overall average error of approximately 14%, down to 3.7% for benzene). The results showed that single metal oxide resistive nanosensors could achieve a good level of real selectivity exploiting the thermal fingerprints from a temperature gradient.