Zirconia-based electrochemical sensors for measuring the oxygen concentration in the automotive exhaust have been used on vehicles for nearly 40 years. The oxygen sensor is an electrochemical cell based on oxygen-ion conducting zirconia. The sensor consists of a single electrochemical cell which has one electrode exposed to a reference atmosphere (typically air) and the other electrode exposed to the measurement (exhaust) gas and operates as a simple Nernst cell. The open circuit voltage of this zirconia electrochemical cell changes with variations in the oxygen partial pressures existing adjacent to its two electrodes. The emf generated by this Nernst cell gives a measurement of the concentration of oxygen in the exhaust gas. This measurement can be used by the engine control unit to adjust the air-to-fuel ratio so as to maintain it close to stoichiometry, thereby minimizing the unwanted emissions. All gasoline (and some diesel) vehicles produced today utilize an oxygen sensor to control their emission systems.

Although successfully used for stoichiometric A/F control for many years, Nernst-type oxygen sensors are not useful for applications away from stoichiometry because of their low sensitivity (i.e. weak logarithmic dependence of the emf on oxygen partial pressure). However, double-zirconia-cell sensors have been developed [1,2] which have much higher sensitivity and thus a wider range of operation and are therefore applicable to A/F measurement and control from very rich to very lean A/F mixtures. These sensors are based on oxygen pumping, which involves the transfer of oxygen from one side of a zirconia cell to the other by passing an electric current through the device. Since these sensors operate over a wide range, they are referred to as “wide-range oxygen” or “universal exhaust gas oxygen” (UEGO) sensors. Most gasoline vehicles, as well as diesel powertrains which operate lean, employ UEGO sensors to meet the emission requirements.

A double-cell oxygen-pumping device (UEGO sensor) also can be used to measu­re the concentration of other oxygen-containing molecules in the exhaust gas, e.g. H₂O, CO₂, SO₂, and NO_x. In order to mea­sure the H₂O concentration, a voltage is applied across the pump-cell that is greater than the dissociation potential for H₂O. A measurement of the pumping current through the cell provides a measure of the oxygen generated from the decomposition of water and thus a measure of the H₂O concentration in the gas. However, in the exhaust gas this measurement is confounded by the fact that molecular oxygen and other molecules also are dissociated. If the concentrations of these other molecules were constant, then the concentration of H₂O could still be determined uniquely by subtracting the contribution of these other molecules. However, the concentrations of the different oxygen-containing molecules in the exhaust gas vary with engine operating conditions. This problem can be overcome by varying the voltage across the pump cell and measuring the pumping currents at various dissociation potentials. At a lower potential the concentration of oxygen can be measured. A second voltage larger than the dissociation potential of the measurement gas, H₂O e.g., is applied to measure the concentration of oxygen plus water. The water concentration can be determined by subtracting the pumping current due to oxygen from the pumping current due to the combination. Other gases, CO₂ e.g., can also be measured in a similar fashion. In principle, these measurements can be achieved without any modification to the existing sensor hardware. A simple modification to the electronic control circuitry is all that is required to perform this type of measurement.

Figure 1 shows the sensitivity of a commercially available UEGO sensor for the dissociation of H₂O as a function of dissociation potential. A typical operating value for the sensing voltage Vs for oxygen measurement is 450 mV. The plot below was generated from data measured in the lab at various water concentrations in a carrier gas of dry nitrogen. From the plot it is evident that water dissociation is occurring at potentials less than 800 mV. Additionally, the output begin to saturate at a potential approaching 1100 mV, suggesting that full dissociation of the water has been achieved. The pumping current of the sensor is linear as a function of water concentration for Vs = 1080 mV. The application of this technology to measure ambient humidity on-board a vehicle during fuel cut will be discussed.