Chemical crosstalk between heated gas microsensor elements operating in close proximity

Abstract

Listen

Translate article

Gas microsensor arrays often have closely-spaced elements, typically separated by hundreds of microns. For such devices, crosstalk between elements operated 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, microhotplate 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. Under proper conditions, a Pt/SnO 2 microsensor was observed to oscillate between two stable CO/O 2 coverage ratios. The high CO coverage state results in higher film conductance, and the oscillation frequency is extremely sensitive to gas-phase CO concentration. Crosstalk was observed between adjacent microsensors with Pt particles supported on SnO 2 films, as evidenced by synchronization of sensor response and mixed-mode oscillations (similar to those 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 sensors is believed to produce reactions that effectively lower the CO partial pressure at the monitored device, thereby reducing the period of oscillation. The magnitude of the effect was calculated from the frequency change using a CO concentration calibration, and the effect is greater than 10% when operating 15 neighboring sensors. The effects of heating neighboring microsensors have also been studied for hydrogen and methanol sensing on both Pt/SnO 2 and bare SnO 2 microsensor arrays. While crosstalk effects are observed for these gases, the effects we observe on Pt/SnO 2 sensors are less pronounced than in the case of CO sensing. A less than 1% effect occurs for methanol sensing and the largest effect for H 2 sensing was an apparent concentration decrease of ≈4% when heating 15 neighboring devices in 10 μmol/mol (10 ppm) of the analyte. On the bare SnO 2 devices, we observe an apparent increase in concentration of methanol and H 2 of approximately 5 and 15%, respectively. While crosstalk is only measured for the case of conductometric sensing in this work, similar phenomena are likely to occur for sensors that utilize other detection principles.