Chemical crosstalk between heated gas microsensor elements operating in close proximity

Gas microsensor arrays often have closely-spaced elements, typically separa ted by hundreds of microns. For such devices, crosstalk between elements op erated within a gaseous environment is a concern because sensing materials held at elevated temperatures have an increased probability for disrupting gas flows and activating gas-film interactions that can consume analytes or evolve reaction products. To explore such effects in a microarray, microho tplate array platforms were used to sense carbon monoxide. Carbon monoxide sensing was chosen as a model system because the oscillatory kinetics of CO oxidation on Pt films are known to exhibit sensitive gas-phase coupling. U nder proper conditions, a Pt/SnO2 microsensor was observed to oscillate bet ween two stable CO/O-2 coverage ratios. The high CO coverage state results in higher film conductance. and the oscillation frequency is extremely sens itive to gas-phase CO concentration. Crosstalk was observed between adjacen t microsensors with Pt particles supported on SnO2 films, as evidenced by s ynchronization of sensor response and mixed-mode oscillations (similar to t hose observed in CO oxidation on Pt films). The range of crosstalk effects was studied by operating a single device in an oscillatory CO sensing mode and heating neighboring elements in an array. The operation of nearby senso rs is believed to produce reactions that effectively lower the CO partial p ressure at the monitored device, thereby reducing the period of oscillation . The magnitude of the effect was calculated from the frequency change usin g a CO concentration calibration, and the effect is greater than 10% when o perating 15 neighboring sensors. The effects of heating neighboring microse nsors have also been studied for hydrogen and methanol sensing on both Pt/S nO2 and bare SnO2 microsensor arrays. While crosstalk effects are observed for these gases, the effects we observe on Pt/SnO2 sensors are less pronoun ced than in the case of CO sensing. A less than 1% effect occurs for methan ol sensing and the largest effect for H-2 sensing was an apparent concentra tion decrease of approximate to4% when heating 15 neighboring devices in 10 mu mol/mol (10 ppm) of the analyte. On the bare SnO2 devices, we observe a n apparent increase in concentration of methanol and H-2 of approximately 5 and 15%, respectively. While crosstalk is only measured for the case of co nductometric sensing in this work, similar phenomena are likely to occur fo r sensors that utilize other detection principles. Published by Elsevier Sc ience B.V.