Final Report: Mixed Potential-Based Miniature Sensors for Real-Time On-Vehicle NOx Monitoring From Mobile SourcesEPA Contract Number: 68D02076
Title: Mixed Potential-Based Miniature Sensors for Real-Time On-Vehicle NOx Monitoring From Mobile Sources
Investigators: Nair, Balakrishnan
Small Business: Ceramatec Inc.
EPA Contact: Manager, SBIR Program
Project Period: October 1, 2002 through July 31, 2003
Project Amount: $99,932
RFA: Small Business Innovation Research (SBIR) - Phase I (2002) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , SBIR - Air Pollution , Small Business Innovation Research (SBIR)
The objective of this Phase I research project was to demonstrate the feasibility of using a mixed potential-based sensor for monitoring NOx gases in an exhaust stream in a temperature range of 450-550°C. To demonstrate feasibility, Ceramatec, Inc.'s objective was to design and fabricate a sensor that would eliminate many of the problems that have previously limited the effective application of mixed potential sensors. These limitations include cross-sensitivity between nitrogen monoxide (NO) and nitrogen dioxide (NO2); cross-sensitivity with other components typical of engine exhaust such as carbon monoxide (CO), carbon dioxide (CO2), sulfur dioxide (SO2), and steam (H2O); a slow response time; and the demonstration of meaningful sensor signal in gases with varying oxygen content. The Phase I work included:
· Sensor fabrication.
· Electrode characterization.
· Electrode adhesion optimization.
· Electrochemical testing, including sensor response characterization as a function of NO and NO2 concentration.
· Incorporation of appropriate modifications in the sensor system to overcome the limitations described above.
· Evaluation of the effectiveness of these modifications in enhancing sensor performance.
· Detailed cross-sensitivity testing of the sensor system in the presence of other gases with and without modifications to the sensor design.
· Response time characterization of the sensors.
As a first step in the project, electrolyte tubes were fabricated through conventional ceramic processing techniques using isostatic pressing and sintering. The conductivity of the electrolyte materials was accurately measured through four-point conductivity measurements on bar specimens at the target use temperatures. Because the adhesion of electrode materials was known to be poor prior to the Phase I project, effort was directed at improving the electrode adhesion for practical use of the sensor. A proprietary method was developed that substantially improved the adhesion of the electrode materials, which performed very well in a tape peel test.
After completing the design and fabrication of the electrode/electrolyte system, a series of detailed electrochemical tests were performed to characterize the sensor performance with varying concentrations of NO and NO2 in various nitrogen/oxygen mixtures. As was reported in the literature, the signs of the sensor responses to NO and NO2 were opposite, and the sensor response was not interpretable under varying NO/NO2 ratios. During Phase I, a proprietary technique was developed to overcome this limitation and after appropriate modifications to the sensor system, it was demonstrated that the sensor was completely insensitive to the NO/NO2 ratio, and that a total NOx reading could be obtained irrespective of the starting concentrations of NO and/or NO2. This represented the successful completion of one of the major goals for Phase I.
Another important goal was demonstrating the feasibility of operation under various oxygen concentrations. As expected, the sensor response was a function of the oxygen concentration. However, it was demonstrated that if the oxygen concentration could be measured, perhaps by incorporation of a second oxygen-sensing electrode on the same sensor, the NOx concentration could be calculated from the measured voltage and the oxygen concentration. This was the second major technical accomplishment in Phase I.
Ceramatec, Inc.'s subsequent efforts were focused on addressing the issue of cross-sensitivity of the sensor to other gases in the exhaust stream. Through a careful set of experiments, it was determined that the electrode was intrinsically insensitive to some constituents such as H2O and CO2. However, with no modifications, the electrode showed some cross-sensitivity to CO and SO2. Through a modification of the sensor system, the cross-sensitivity to CO was nearly completely eliminated. Further, through yet another modification in the sensor system, the tolerance to SO2 was demonstrated to be sufficient for the sensor to perform in systems that use cleaner fuels, and little cross-sensitivity was seen up to 25 ppm of SO2.
As the final goal of the Phase I project, Ceramatec, Inc., set out to demonstrate that, through appropriate modifications, a fast 90 percent response time on the order of 1 second could be attained. Testing with various sensor configurations clearly showed that response times on the order of 1-3 seconds consistently could be obtained. Although target response times for control sensors have been noted to be on the order of 250 milliseconds, some end users informed the company that many sensor developers refer to the time constant (i.e., the 63 percent response time) as the "response time," which is approximately three times less than the 90 percent response time due to the logarithmic relationship between output voltage and concentration. By this criterion, Ceramatec, Inc.'s response times routinely are in the range of 300 milliseconds to 1 second. Therefore, the company essentially has achieved its objective with further definition of response time to be clarified during the Phase II effort.
Having successfully accomplished all the goals outlined in the Phase I proposal, attention was directed at some new and unexpected problems that were identified during the Phase I project. The first was the problem of sensor drift. A set of detailed experiments during a period of weeks with the same sensor identified a few key parameters that influence sensor drift. Future design modifications will be incorporated to minimize any effects from sensor drift.
The second new issue that was identified, based on a discussion with a potential end user of the technology, was the issue of "oxygen response time" (i.e., the time taken by the sensor to respond to a change in oxygen concentration rather than NOx concentration). Preliminary experiments were performed to evaluate this issue, and initial results were very encouraging and will be pursued further in the Phase II project.
Finally, because the sensor is designed to be part of an emissions control system that would reduce the NOx emissions to a significantly lower level than current levels, several end users suggested that it would be important to demonstrate sensitivity to very low NOx levels on the order of 1 ppm. Therefore, Ceramatec, Inc., performed sensor testing at very low NOx levels of 1-20 ppm and found that, as expected due to the logarithmic nature of the response, the sensitivity to low levels of NOx was very high and that NOx levels as low as 1 ppm could be accurately detected.
Previously, mixed potential-type NOx sensor technologies have been limited by problems of cross-sensitivity with other gas constituents commonly found in diesel exhaust, as well as an inability to provide a meaningful signal in varying NO/NO2 mixtures. A novel proprietary sensor system developed by Ceramatec, Inc., has overcome many of the problems previously associated with NOx sensors. Sensors have been fabricated and tested that have the following characteristics: (1) operation temperature of 500-600°C; (2) excellent sensitivity in NOx levels of 1-1,200 ppm; (3) 90 percent response times as fast as 1-3 seconds; (4) insensitivity to various NO/NO2 ratios in the exhaust stream; (5) very low cross-sensitivity to CO, CO2, H2O, and low levels of SO2; and (6) ability to operate in oxygen-containing environments with no requirement for a pumping cell.
The new sensor technology, which is a potentiometric mixed potential sensor, has a number of significant advantages over other NOx sensor systems, and therefore represents an exciting technology for diesel engine manufacturers to meet the demanding requirements of emissions control sensors. Other competing high-temperature sensors are amperometric sensors that depend on the use of a very small orifice that acts as a gas-limiting element that controls the amount of gas that can enter the pumping chamber. To make these amperometric sensors reproducible from one to the next, the size of this orifice has to be very accurately controlled. In addition, the orifice can become plugged with time or simply change geometrical dimension through long-term annealing or partial clogging of particles on the edges. Other advantages of a potentiometric mixed potential-type sensor over an amperometric sensor include very high sensitivity to low concentrations of NOx due to the logarithmic response curve, much lower power requirement for operation, and substantially lower design complexity.
During the Phase I research project, Ceramatec, Inc., stayed in close contact with potential end users of the technology such as diesel engine and automobile manufacturers, as well as system integrators. A commercialization report on the sensor technology, prepared by Foresight Science and Technology, indicated that Ceramatec Inc.'s technology was very competitive with other sensor technologies being developed. During Phase I and in the proposed Phase II project, a major U.S. diesel engine manufacturer committed to an in-kind contribution to the program in the form of sensor testing in simulated engine exhaust. Other targets identified by Foresight Science and Technology were contacted to discuss the potential of the technology, and the response was very encouraging. The technology also captured the interest of a major U.S. advanced turbine manufacturer that was interested in substantial cost sharing, through in-house research and development, in a future program for the development of higher end sensors for advanced turbine emissions monitoring at very low concentrations below 10 ppm.