Standard ceramic process methods were used to fabricate NOx sensors containing 4.7 mol% Y2O3 – ZrO2 (PSZ, Advanced Ceramics) and 8 mol% Y2O3 – ZrO2 (FSZ, Tosoh Corp.). Polyvinyl butyral (B-76 Butvar) and ethanol were added to the electrolyte samples to form a mixture that was ball milled into a uniform slurry of which a portion was air dried to form a dry powder. The powder was uniaxially pressed under a load of 200 MPa to form pellets. Au wires were positioned over the electrolyte pellets and the remaining electrolyte slurry was coated over the electrodes. The PSZ and FSZ based sensors were fired at 1050˚C for 1 hour. Impedance studies were conducted using a Gamry Reference 600 for sensor operation over a temperature range of 600-700°C where NO and NO2 concentrations varied from 0 to 100 ppm for dry and humidified (3, 5, and 10% H2O) gas environments. Based on analysis of the impedance data the sensitivity, selectivity, oxygen dependence, activation energy, and response rate of the sensors was determined. The microstructure and morphology of the electrolytes were studied using SEM, BET, and MIP.
The impedance response indicated the PSZ based sensors demonstrated less cross-sensitivity to H2O in comparison to the FSZ based sensors. This was observed from the impedance data where the interfacial resistance decreased by 12% for FSZ based sensors when 10% water was added to the gas stream, whereas, a slight increase in interfacial resistance of approximately 2% occurred for PSZ based sensors under the same conditions. The addition of water likely resulted in hydroxyl groups that effected the interfacial conductivity of the sensors. It is possible that reactions involving hydroxyl groups were greater for the FSZ based sensors such that an increase in interfacial conductivity reduced the resistance at the Au/FSZ interface. The formation of hydroxyl groups can result from oxygen ions traveling through the electrolyte. As the oxygen ion conductivity is lower for PSZ in comparison to FSZ, it is possible that the difference in the water behavior for the PSZ versus the FSZ based sensors was due to a lower rate of hydroxyl formation at the PSZ/Au interface. The impedance data also indicated the NOx sensitivity was greater for FSZ versus PSZ based sensors for both dry and humidified gas conditions; and the FSZ based sensors had a more rapid response.
The activation energy of the sensors was approximately 1.0 and 1.2 eV for the PSZ and FSZ based sensors, respectively. The activation energy decreased slightly when NOx was present in the gas stream, and when water was added. Oxygen partial pressure measurements indicated dissociative adsorption was the dominant rate limiting for both dry and humidified gas environments for both PSZ and FSZ based sensors.
The effect of electrolyte particle size on water behavior was considered by studying the behavior of sensors composed of PSZ electrolytes containing 1 micron and 2 micron sized particles. The magnitude of the impedance response was larger for the sensors containing the 2 micron sized particles, however, the water behavior was similar for by types of PSZ based sensors. A slight increase in the impedance was observed when water was added to the gas stream for sensors containing PSZ irrespective of the electrolyte particle size.
Overall, the morphology and microstructure of the sensors did not appear to influence the behavior of the sensors in the presence of water. Rather, it is likely that the formation of hydroxyl groups and related reactions contributed to cross-sensitivity that was primarily observed for the FSZ based sensors. It is possible that the limitation of hydroxyl group reactions at PSZ based sensors restricted water cross-sensitivity, thereby improving the selectivity of the sensors to NOx.