Therefore, appropriate number of layers deserves comprehensive considerations. An electronic nose containing 16 sensors withfive to twenty layers of porous In2O3microtubes was prepared (see SI,Fig. S4). The Ni–Cr heating wires with the same electronic resistance were selected as the heaters in ceramic tubes for each individual sensor to ensure the same working temperature under same voltage. The working temperature of all individual sensor units in the electronic nose is about 340°C. The performance of the electronic nose, equipped with gradient layers of porous In2O3 microtubes, is shown in Fig. 3 for ethanol as a test gas at concentrations of 100 ppm, 300 ppm, 1000 ppm, 2000 ppm, and 3300 ppm. As presented inFig. 3a, each of In2O3SMO sensors shows a different but reliable response (Ra/Rg) for ethanol. All the experiments were
repeated in triplicate, and a correlation of the logarithm of the
average sensor response versus the logarithm of the ethanol concentration was obtained (Fig. 3b). The linearfitting was quite good after the exposure to the wide concentration ranges of ethanol. Although the difference amongfitting curves is small, none of them overlapped or paralleled (see SI,Table S1), proving that each of the sixteen sensor units was independent. More importantly, the linear relationship between the logarithms of the average sensor response versus the logarithm of the ethanol concentration provided a wide measurement range for the quantitative analysis of ethanol.