Selective gas sensor based on one single SnO2 nanowire

Nowadays the analysis of ambient atmosphere in order to monitor the presence of dangerous gases and volatile compounds is more and more important. For this reason, a network of tiny sensors capable to discriminate the presence of pollutants and distinguish them is crucial. We present here a resistive sensor based on a single tin oxide nanowire (60 nm in diameter and 3.5 mu m long) that can detect the presence of different gases and estimate their concentration in the range of 1-50 ppm. The SnO2 nanowire (NW) is grown by chemical vapor deposition and then used to bridge to metal electrodes. Under a temperature gradient, 5 signals can be extracted, forming the thermal fingerprint of each specific gas that can be present in the measuring chamber. Applying machine learning algorithms to these thermal fingerprints, the system can recognize which gas is present in the chamber (with an 94.3% accuracy) and estimate the concentration of the gas (with an average error of 24.5%). The limit of detection has been found to be under 1 part per million for all the gases tested.