Towards electronic smelling of ketones and alcohols at sub- and low ppms by pinky-sized on-chip sensor array with SnO2 mesoporous layer gradually engineered by near IR-laser

We consider an on-chip sensor array based on a mesoporous layer of SnO2 nanoparticles to be screen printed on the multielectrode-supplied Si/SiO2 substrate as a chemiresistive building platform for portable and personalized in situ instruments. To differentiate the local oxide layer properties we apply Nd:YAG laser whose scanning etched various layer areas at varied power driven by working current in 24.8 A–26.7 A range. As a result, the SnO2 layer has dual-grad modified properties as, (i) a spatial modification of thickness down to nm-range, and (ii) the change of oxidation state with appearance of traces of SnO, which both result in a great varying of gas-sensing properties of local sensor elements over the array. To test the functionality of the chip, we could detect vapors of four ketones (acetone, cyclopentanone, cyclohexanone, 2-octanone) and four alcohols (methanol, ethanol, isopropanol, butanol), at sub-, down to ca. 100 ppb, and low, up to 10, ppm concentrations with their selective recognition via processing the array's vector signal by linear discriminant algorithm. Primarily, we show differences in the interaction of ketones and alcohols with SnO2 surface by first-principle calculations in frames of density functional theory to serve as fundamental receptor pre-requisites for the analyte's selective discrimination by the oxide layer under the multisensor concept to employ here. We consistently show that two modes of the sensor operation could be rather equally applied to the array as, (i) UV LED, 366 nm wavelength, irradiation at room temperature, and (ii) heating up to approx. 583 K. While the heating provides faster and higher chemiresistive responses, the UV-excited mode provides more selective vector signals, lower energy consumption, and a higher signal-to-noise ratio.