A Three Electrode Mixed Potential Sensor for Gas Detection and Discrimination

Abstract

A major goal for sensors in the automotive exhaust emissions control and monitoring field is the development of a single sensor to monitor NOx, NH3, CO, and hydrocarbons. Mixed potential electrochemical sensors can be readily tuned to be selective towards different gas mixtures by changing the composition of their electrode or applying a voltage or current bias. We have developed a three-electrode sensor comprised of Pt, La0.8Sr0.2CrO3 (LSCO), and Au0.5Pd0.5 electrodes with a porous (Yttria-stabilized Zirconia) YSZ electrolyte which can detect single and binary gas mixtures of NO, C3H8, and CO. The sensitivity of the electrode pairs were investigated and it was found that AuPd+Pt was most sensitive to CO while LSCO+Pt is best at detecting C3H8. From the sensor responses, we have trained artificial neural networks to use the voltage responses to recover the gas concentrations and classify the gas mixtures.