Int. J. Appl. Ceram. Technol., 3 [4] 302–311 (2006)
Ceramic Product Development and Commercialization
High-Temperature Ceramic Gas Sensors: A Review Sheikh Akbar,* Prabir Dutta, and Chonghoon Lee Center for Industrial Sensors and Measurements (CISM), Ohio State University, Columbus, Ohio 43210
Identifying chemical species and their quantification have become important in many industrial applications involving high temperatures and chemical contaminants. Center for Industrial Sensors and Measurements has developed TiO2 semiconducting sensors, zirconia and lithium phosphate-based electrochemical sensors, and a sensor array for high-temperature emission control. The underlying theme in our sensor development has been the use of materials science and chemistry to promote high-temperature performance with selectivity. This article presents key results of previous studies on CO, NOx, CO2, and O2 sensors, and scope for future development.
Introduction Intelligent systems based on sensors and controls can be used for health and safety (e.g., medical diagnostics, air quality monitoring, and detection of toxic, flammable, and explosive gases), energy efficiency, and emission control in combustion processes, and industrial process control for improved productivity and product quality. There is a continuing need for the development of fast, sensitive, rugged, reliable, and low-cost sensors for applications in harsh industrial environments found in heat treating, metal processing and casting, glass, ceramic, pulp and paper, automotive, aerospace, utility and power, chemical and petrochemical, and food-processing industries. Table I, for example, summarizes some of the key industrial applications of
This work was supported by CISM through the National Science Foundation contract no. EEC-9523358, the Department of Energy contract no. DE-FC26-03NT41615, and NASA-GMI contract no. NNC04AA48A. *
[email protected] r 2006 The American Ceramic Society
an oxygen gas sensor.1–3 Similar data can be compiled for other sensors. The monitoring and control of combustion-related emissions are a top priority in many industries. The availability of reliable sensors along with predictive emission modeling tools would provide a better control of combustion, leading to reduction of toxic emissions and subsequent energy savings. According to a U.S. DOE report,4 harsh environment sensors are predicted to save 0.25 quadrillion BTU/year of energy across all energy-consuming industries, identified as Industries of the Future (IOF). Sensors capable of providing threedimensional maps of emission profiles will allow for feedback control systems of combustion processes, resulting in lower emissions and efficient use of fuels. Emissions-monitoring sensors for these applications include those for CO, NOx, O2, CO2, hydrocarbons (HCs), and volatile organic compounds (VOCs). The control of combustion and emissions from glass tanks with the use of high-temperature sensors is a top priority of the glass industry. Emissions are the byproduct of combustion, batching, fining, and melting
High-Temperature Ceramic Gas Sensors
www.ceramics.org/ACT
Table I. Industry Petrochemical industry
Oil and gas
Steelmaking
Power generation
Cement
Refinery
303
Oxygen Gas Detection Ranges in Various Industrial Processes Process
Styrene vent gas Trans-alkylation regeneration gas Stacked gas Polystyrene Ethylene feed Reactor inlet and reactor outlet Oxidizer vent gas Heavy ends fired boiler Absorber vent gas Stripper vent gas VCM vent gas Product storage/N2 blanket ACH vent blower safety Steam condensate/leaving plant Methanator and NH3 converter outlet CO2 feed to urea plant Inlet of recovered VCM compressor Outlet to N2header/purity Wet gas flare header Thermal incinerator stack Dry gas production lines Liquid extraction facility Landfill gas feed to incinerator boiler Pig iron production Coke oven for benzene emissions control Hydrogen/N2 blanketing for colored oxide prevention Furnace (at 14001C) O2 detection to prevent embrittlement Dust collection system/baghouses–airleak detection Prevent slag oxidation of steel Excess O2 for combustion efficiency Pollution abatement SF6 filling Nuclear power plant (PWR/BWR) for monitoring Gaseous waste treatment Vacuum drag system Rotary kiln string Pre-calciner string Electrostatic precipitator Portable combustion efficiency Catalyst regeneration Desulfide separator vent gas Sulfur recovery unit flue gas to SRU stack incinerator Process gas/‘‘Burn off only’’ Partial combustion – furnace process
Concentration ranges 0–2000 ppm in H2/N2/CO2 0–15% in O2/N2/CO/CO2 0–5% in flue gas 0–5% in N2/styrene vapors 0–5% in CO2/AR/CH4 0–25% in CO2/AR/CH4 0–25% in N2, cumene, H2O 0–5% in flue gas 0–5% in CO2/H20/Cl2/HCl 0–5% in CO2/H20/Cl2/HCl 0–10/0–100 ppm 0–10% 0–5% in N21HCN1acetone 0–150 ppm in 90% CO 0–0.5% in N2/H2/CO2 0–1% in CO2, H2, N2, Ar 0–5% in VCM (96%), H2Ov 0–10 ppm in N2 0–2000 ppm in C3 hydrocarbons 15% O2 in N2, CO, SO2, CO2 0–10 ppm in LNG 0–100 ppm in LNG 0–5% in CH4, CO2, O2 0–2, 0–5% in N2, CO/CO2, H2Ov 0–2, 0–10% incoke oven gas 0–10 ppm or less in H2/N2 0–5, 0–10% in N2, H2Ov, CO2 0–5% in air1dust from BOF 0–10, 0–100 ppm in N2/Ar blanketing 0–5% in flue gas 0–10 ppm in SFG 0–25% in air 0–5, 0–10% in N2 0–5% in N2/H2 0–5%, 0–10% in flue gas/dusty 0–5%, 0–10% in flue gas/dusty 0–5, 0–10% in flue gas 0–10% in flue gas 0–1, 0–5, 0–25% in flue gas 0–25% in process gas 0–5% in flue gas 0–5% in N2/CO2 0–5% in C.B1H.C1CO
304
International Journal of Applied Ceramic Technology—Akbar, Dutta, and Lee
Table I. Industry Food and beverage
Pulp and paper and glass industry
Continued
Process
Concentration ranges
Packaging quality testing of inert atmos (CO2 or N2) of packaged foods; bottled or canned beverages CO2 fermentation gas recovery/scrubbing for headspace bottling process Black liquor recovery (BLR) Combustion efficiency percent O2 monitored at several points to obtain complete profile of furnace
processes.5–9 Critical air pollutants include oxides of nitrogen (NOx), oxides of carbon (CO/CO2), oxides of sulfur (SOx), and particulates. In 1994, the domestic glass industry consumed approximately 200 trillion BTUs and the melting process is only about 50% efficient. The glass industry in its document ‘‘Glass: A Clear Vision for a Bright Future’’ sets as a goal for 2020 to reduce air/water emissions by 20%.10 Across IOF, 60 tons of NOx are generated per year, with estimates that the glass industry may produce 40,000 ton/year. The reduction of NOx from glass tanks in compliance with emission regulations is one of the stated goals of the glass industry. Precise detection of CO2 is important for air quality monitoring, fire detection, and engine exhausts. Air quality monitoring is important in an aircraft/spacecraft compartment or as part of a building ventilation system. Fire detection can be improved beyond the standard smoke detection system by concurrently monitoring the chemical signature of a fire.1,11 One possible chemical signature to determine fire condition is the CO/CO2 ratio. Exhaust gas sensors can reduce emissions and lower operating costs of solid oxide fuel cells (SOFC). There are several locations in the SOFC generator where gas sensors can be utilized to achieve various objectives.12 While in one region monitoring CO/H2 ratio and CH4 would determine the efficiency of the reforming process, CO2 sensing in a second region can provide feedback for the electrochemical conversion of CO. In a third region, monitoring CO, H2, O2, and CO2 would determine the efficiency of the total combustion process and provide feedback to the controls for optimum fuel utilization. Commercial gas detection systems using infrared spectroscopy, gas chromatography/mass spectrometry (GC/MS), and chemiluminescent analysis are available
Vol. 3, No. 4, 2006
0–100 ppm, 0–3% in N2 or CO2 0–100 ppm in CO2 0–5% in extremely dirty application 0–5% in N2, O2, HC, H2Ov
with good detection limits and fast response times.13 However, these instruments are bulky, expensive, need maintenance, incompatible with high-temperature environments, and are typically located outside the combustion environment with additional gas sampling systems. Miniaturized gas sensors are necessary for in situ high-temperature monitoring and feedback control for combustion optimization. SiC is used for gas sensing, but silicide formation limits its use for high-temperature use.14 High-temperature-stable piezoelectric, langasite-based surface acoustic wave (SAW) sensors can be used at 9351C.15 However, using physical properties such as mass, dielectric constant, temperature, and surface stress changes produced by sorption of gas molecules is too complex for harsh environments. Recently developed ionization sensors16 using carbon nano-tubes appear promising, but high-temperature use is not yet anticipated. On the other hand, solid-state chemical gas sensors, based on either oxide semiconductors or bulk electrolytic properties of ion-conducting ceramics for in situ high-temperature combustion environments, hold promise and can be readily miniaturized if reference gases are not required. Our research group at the NSF Center for Industrial Sensors and Measurements (CISM) has developed a series of high-temperature sensors for CO, O2, NOx, CO2, and hydrocarbon detection.13,17–33 The underlying theme in our sensor development has been the use of materials science and chemistry to promote high-temperature performance with selectivity. For NOx and hydrocarbon sensors, we have exploited microporous aluminosilicate chemical filters, for CO sensing the use of dopants that alter electrical and chemical characteristics of titania, for O2 sensing the use of metal–metal oxide references, and for CO2 sensing, a novel electrode and electrolyte system. In addition, we have used
High-Temperature Ceramic Gas Sensors
www.ceramics.org/ACT
305
nonlinear regression techniques for gas concentration predictions for two- and three-sensor array systems. This article presents state of the art in solid-state gas sensors including latest developments in CISM. Specific Gas Sensors Electrochemical NOx Sensor
Solid-state gas sensors using oxygen ion-conducting yttria-stabilized zirconia (YSZ) are suitable for mixedpotential NOx sensors at temperatures higher than 5001C.34,35 There has been particular focus on discovering optimal metal–oxide electrodes.36,37 Sensors operating in the chronoamperometric mode by polarizing the working electrode and recording the current–time relation are also under study.38–40 For detecting total NOx (NO1NO2) in fluctuating oxygen, a multichamber design is the most well-known approach to construct electrochemical sensors.40–42 In this design, a gas mixture passes through a diffusion channel into one or two chambers constructed of laminated YSZ sheets. The first chamber is for oxygen pumping; noble metal electrodes electrochemically convert the NOx mixture into NO or NO2 exclusively, which is then detected by either potentiometric or amperometric methods at the last stage. Electrode impedance change at a fixed AC frequency has also been reported for NOx detection.43 Further improvements needed to meet the requirements of realworld applications are a quick response, high selectivity, temperature compatibility (400–8001C), long-term stability, simple design, and low cost.
Fig. 1. Schematic representation of the equilibration of NOx by Pt loaded inside zeolite Y.
We have successfully fabricated potentiometric sensors for total NOx detection at high temperatures. By using Pt-loaded zeolite Y (PtY) as a filter, NOx species in the gas stream are brought to an equilibrium mixture of NO and NO2 determined solely by the background oxygen concentration and filter temperature, as demonstrated schematically in Fig. 1. The equilibrated NOx gas is then measured with a planar potentiometric sensor comprised of YSZ electrolyte and metal–oxide electrodes.26 Figure 2 shows data with the zeolite filter at 6001C and the sensor at 5001C, and demonstrates that the sensor response to either NO or NO2 (100–000 ppm) is essentially identical, thus leading to a total NOx sensor.
–25.0 3%O –30.0 –35.0
EMF/mV
3%O
3%O 100ppm NO
100ppm NO
–40.0 400ppm NO
–45.0
400ppm NO
600ppm NO 1000ppm NO
–50.0
600ppm NO
1000ppm NO
–55.0 0
50
100
150
200
Time (min)
Fig. 2.
Potentiometric sensor response to NO and NO2 after gases have passed through a Pt-zeolite Y filter (background gas 3% O2/N2).
306
International Journal of Applied Ceramic Technology—Akbar, Dutta, and Lee
Vol. 3, No. 4, 2006
By increasing the temperature difference between the filter and the sensor, the signal magnitude can be increased. The advantages of high sensitivity, no CO interference (because of filter) and minimal O2 interference, quick response/recovery time, good long-term stability, and total NOx-measuring capability make the sensor attractive for real-life applications. Semiconducting CO Sensor
The change in electrical conductance of a metal oxide in the presence of a gas is a relatively simple method of detection. ZnO, SnO2, TiO2, and Ga2O3 have been extensively studied for gas-sensing applications.44 SnO2-based CO sensors for domestic applications operate at temperatures between 1001C and 3001C. Addition of dopants to modify sensitivity has been extensively studied. Platinum-activated SnO2 thin films were found to be more sensitive, operable at lower temperatures, and with improved response time (2–3 min) as compared with a pure SnO2 film.45 SnO2 thin films prepared from pyrolysis of an arosol showed about 10 times higher sensitivity to CO after doping with Pt (1–12 wt%), with long response times (20 min).46 A platinum-impregnated TiO2 thick film was found to be more sensitive to CH4 than CO at 6001C.47 A Ga2O3 thin film-based CO sensor developed by the Siemens AG group in Germany48 shows promise for high-temperature applications, although it has cross-sensitivity to changes in oxygen, hydrogen, and hydrocarbons. The main drawbacks of the metal oxide sensors for reducing gases at high temperatures (400–8001C) are (i) high cross-sensitivity (interference from other gases), (ii) drift of the sensor response, and (iii) poor recovery. To achieve better stability and reproducibility for CO sensing at temperatures above 6001C, we have adopted the strategy of doping TiO2 (anatase) with lanthanum oxide, which provides microstructural, crystallographic, and electrical stability during long-term operation at high temperatures, thus minimizing drift. To the La-stabilized anantase, we then add CuO (labeled ALC) for improving sensitivity.19 Figure 3 shows the results of changes in the electrical resistance of an ALC sample upon exposure to CH4 and CO at 6001C. Sensitivity is defined as the steady-state resistance (R) of the sensor at a given concentration of the target gas divided by the resistance in the absence of the target gas (Ro).
Fig. 3. Change in resistance of an La-stabilized anantase, we then add CuO (ALC) sample at 6001C upon exposure to CH4 and CO.
The sensor response to CH4 is practically none, making it selective to CO.
Electrochemical CO2 Sensor
CO2 gas sensors with NASICON electrolyte and Na2CO3-sensing electrode have been well studied.49,50 However, these materials are sensitive to humidity, and sensor performance deteriorates with time.51 Moreover, these electrolytes are not compatible with microelectronic fabrication technology for integration of several sensors into electronic chips and mass production for commercialization. On the other hand, lithium is known to be more resistive to humidity than other alkali metals. Li3PO4 electrolyte has been fabricated as a LIPON glass thin film in battery applications.52–54 Therefore, it is a potential electrolyte candidate for a thin film-based CO2 gas sensor. CO2 sensors operating at high temperatures (400–8001C) with minimal interference from humidity and lacking the need for reference are required. CISM has developed a CO2 sensor28,29 based on a Li-conducting electrolyte that shows excellent sensing performance in the lab as well as in automobile exhaust tests.30 This sensor is composed of Li2CO3 and a Li2TiO31TiO2 mixture as the sensing and the reference electrodes, respectively, and Li3PO4 as an electrolyte. It does not need a reference gas because reference electrode reactions are independent of CO2 gas. The
High-Temperature Ceramic Gas Sensors
www.ceramics.org/ACT
307
600
EMF (mV)
500 400 300 200 100 0 2.5
600C 550C 500C 600C Theoretical 550C Theoretical 500C Theoretical
3
3.5
4 4.5 5 log (PCO /[ppm])
5.5
6
Fig. 4. Comparison of the measured EMF and the theoretically calculated EMF at 5001C, 5501C and 6001C.
stoichiometry of Li3PO4 at temperatures above 5001C.30
measured EMF is determined by EMF ¼
ðmoLi2 CO3 þ moTiO2 moLi2 TiO3 moCO2 Þ 2F RT ln PCO2 2F
Fig. 5. Electronic conduction parameter calculated with various concentrations of CO2 at 4001C, 5001C, and 6001C.
Potentiometric O2 Sensor
ð1Þ
Figure 4 shows the sensor response to CO2 gas concentrations in the temperature range of 500–6001C, along with the theoretically predicted Nernstian response. As can be seen from Fig.4, the sensor response to CO2 shows a systematic deviation from the prediction of the Nernst equation particularly at low PCO2 . The electronic conduction of Li3PO4 was observed to increase with decreasing PCO2 based on the EMF measurement. The conduction domain constructed by EMF measurement demonstrates that change in the Li activity in the sensing side of the cell due to PCO2 variation drives the Li3PO4 electrolyte to a mixed (n-type electronic and ionic) conduction region at low PCO2 as can be seen in Fig.5. Hebb–Wagner DC polarization measurements also proved that n-type electronic conduction in Li3PO4 becomes significant when Li3PO4 is in contact with a mixture of Li2CO3 and gold as a reversible electrode.30 The transference numbers obtained from both the EMF measurement and the Hebb–Wagner polarization measurement indicated that the origin of the non-Nernstian behavior of the CO2 sensor is due to the lithium mass transport from the Li2CO3 sensing electrode to the Li3PO4 electrolyte, resulting in non-
YSZ-based oxygen sensors are widely used in automobile engine combustion management and other combustion processes. A variety of practical designs of oxygen sensors ranging from tubular YSZ-based potentiometric type to planar amperometric type with extended electronic circuits exist for different applications. The most abundantly used potentiometric-type sensor uses air as the reference gas electrode. In many applications, an external reference gas needs associated plumbing and the sensor package is bulky and not amenable to positioning near the combustion source. The driving force in oxygen sensor miniaturization has been to remove the need for an external gas reference. There are few publications on reference-free oxygen sensors. O2 gas can be generated electrolytically at the electrolyte/electrode junction, generating a reference PCO2 :55 This sensor was designed for use as an on–off l sensor to measure rich versus lean engine exhaust. Another on–off sensor uses solid CeO2 (with 25 mol% ZrO2) as the material stores oxygen at high temperatures and can act as the reference O2 source.56 The use of metal/metal oxide mixture as a reference has been known since 1960 as a Rhine’s pack in a galvanic cell arrangement (sensing side) Pt electrode |YSZ electrolyte| M/MO solid reference | Pt electrode.23,57 The M/MO mixture provides a source–sink oxygen
International Journal of Applied Ceramic Technology—Akbar, Dutta, and Lee
308
reservoir with a fixed activity provided by the equilibrium y xM þ O2 $ Mx Oy ð2Þ 2 Equilibrium is achieved when the oxygen gas has thoroughly permeated both the metal and the oxide. In any EMF measurement, it is necessary to establish local thermodynamic equilibrium at the relevant phase boundaries. It is therefore necessary that the galvanic cell behaves reversibly, that is, the voltage should return to the equilibrium value even if small currents are passed through the cell in either direction. Local thermodynamic equilibrium can be disturbed by electrode–electrolyte reactions, or due to an excessive oxygen flux through the cell. Maskell and Steele58 indicate that when the reference oxygen partial pressure is of the order of the ppm level (106 atm range) or below, the exchange current density at the electrode is very low. In this case, the metal/metal–oxide redox couple can only be effective if
Vol. 3, No. 4, 2006
there is some presence of either CO2, or H2O or both above the metal/metal–oxide reference, to act as intermediaries, effectively carrying oxygen gas between the metal/metal–oxide couple and the electrode and/or the metal/metal–oxide system is in direct contact with the electrode, so that the metal/metal–oxide systems participate via a direct solid-state reaction. Based on Rhine’s pack idea, we developed a tubular design sensor using Ni/NiO23 that showed satisfactory performance. To demonstrate the feasibility of a miniaturized planar design, a probe was fabricated (Fig. 6a), where sealing was achieved using a gold paste and firing at 10001C for 1 h. The fabricated sensor was tested under different oxygen partial pressures at 5001C and 6001C after a heat treatment at 5001C in a nitrogen atmosphere for 12 h. Figure 6b shows the semi-logarithmic plot of the sensor output (mV) against the partial pressure of oxygen (PCO2 ) from 2% to 21% measured at 6001C on different days. The sensor was found to exhibit reproducible sensitivity with a fast response.
Electrodes
Sensor Arrays and Algorithms Gold seal
YSZ Electrolyte Ni/NiO powder
5 mm
Alumina base cup 10 mm (a) 90 Sensor output (mV)
600 °C
70
50
–1.8
–1.5
– 1.2
– 0.9
– 0.6
Log [pO (atm.)] (b)
Fig. 6. (a) Schematic design of the planar sensor and (b) sensitivity plot of the planar oxygen sensor at 6001C measured on different days.
All sensors, no matter how well designed, tend to exhibit some cross-sensitivity when working at high temperatures, necessitating the development of sensor arrays, which in turn requires miniaturized sensors. Most of the micromachined sensor arrays to date are based on well-developed integrated circuit fabrication technology using silicon substrates for applications under 4001C.59,60 Chemical and physical sensors have been successfully integrated with all the microelectronics required for data processing and transaction on a single chip as a monolithic gas-sensing system,61,62 but operate at ambient temperatures. High-temperature miniaturized gas sensor arrays do not yet exist. Methods to extract information from sensor arrays can be divided into two categories: those for qualitative information acquisition and those for quantitative extraction.63–66 The most well known representatives of the first category are pattern recognition-based approaches, while multicomponent analysis or regression techniques fall into the second category. There exists a large variety of pattern recognition methods,67–69 including back-propagation artificial neural networks (BP-ANN), feed-forward neural networks (FNN), principal component analysis (PCA),
www.ceramics.org/ACT
High-Temperature Ceramic Gas Sensors
probabilistic neural network (PNN), learning vector quantization (LVQ) neural networks, soft independent modeling of class analogy (SIMCA), linear discriminant analysis (LDA), Bayesian linear discriminant analysis (BLDA), mahalanobis linear discriminant analysis (MLDA), and the nearest-neighbor (NN) pattern recognition algorithm. Multicomponent analysis or regression methods are commonly regarded as traditional means in chemometrics for sensor applications. The most popular regression methods include partial least square (PLS), principal component regression (PCR), both of which are treated as linear regression, and the nonlinear extensions of PLS. PLS and PCR are quite similar in principle.68 They tend to be overly optimistic when the data tend to be characterized by more measured variables than observations. In addition, as both PLS and PCR are linear regression techniques, they are most successful when sensor responses are known to be linear. The nonlinearity problem can be alleviated by using different versions of PLS, for example poly-PLS, which uses a polynomial fit for the PLS inner relation or some novel nonlinear extension of linear PLS.67,68 But these methods require high computational costs and the results of these calculations are often not transparent. Algorithms that use the physics of sensor response and have predictive capabilities will be an advance in the field. We have reported results on an anatase-based twosensor array for detecting CO and O2.33 In this work, a multivariate nonlinear regression approach, called kernel ridge regression (KRR), was developed for modeling sensor array behavior and to identify gas composition quantitatively from raw sensor array data.33 Originally derived from the concepts of the support vector machine theory,70 KRR has many interesting properties when applied in gas sensor array application. As most sensor behavior demonstrates nonlinearities, KRR can handle this type of sensor response readily due to its nonlinear nature. Secondly, the computational expense increases as the number of basis functions used in regression increases. By using a novel approach to achieve close-form solutions, KRR circumvents the explosive increase of computational cost while obtaining good accuracy. More importantly, the selection of basis functions is based on the sensor response. Furthermore, unlike many other techniques that depend heavily on individual sensor selectivity, KRR decouples the colinearity between sensors in an array and introduces the
309
concept of inter-sensor orthogonality to characterize the group properties of a sensor array system. In addition, KRR requires a relatively small amount of training data, a much shorter period of training time, and no redundant sensor or training data, as is mentioned in.71 Finally, KRR has the capability of extrapolation, which is often a good starting point in predicting sensor array behavior in unknown ranges.
Conclusions Current chemical sensor technologies based on electrochemistry, calorimetry, chromatography, and spectroscopic techniques are either too expensive or cumbersome for in situ measurements in industrial environments involving high temperature and chemical contaminants. Ceramic gas sensors are attractive because of their low cost, simple structure, ease of fabrication, and compatibility with electronic systems. However, for commercial success, major advances in these sensors are required in terms of selectivity, longterm stability, and miniaturization. Although MEMS processing has enjoyed a great deal of success in the development of physical sensors and actuators, its application in chemical sensors is very limited. One key technical challenge is that the new materials developed for chemical sensing have not been implemented in thin-film applications.
References 1. D. Kohl, J. Kelleter, and H. Petig, ‘‘Detection of Fires by Gas Sensors,’’ Sensors Update, 9 161–223 (2001). 2. N. Docquier and S. Candel, ‘‘Combustion Control and Sensors: A Review,’’ Prog. Energy Combustion Sci., 28 [2] 107–150 (2002). 3. H. Kubler, ‘‘Oxygen Sensor For Control of Wood Combustion,’’ Wood Fiber Sci., 24 [2] 141–146 (1992). 4. Glass Industry: Technology Road Map, DOE Report, April 2002. 5. T. W. Samadhi, J. C. Elliott, L. E. Jones, and A. G. Clare, ‘‘Sodium Sulfate Decomposition in Dry Atmospheres,’’ Glastech. Ber. Glass Sci. Technol., 74 [3] 47–56 (2001). 6. T. W. Samadhi, L. E. Jones, and A. G. Clare, ‘‘SOx Emissions from Glass Manufacturing: Decomposition of Sodium Sulfate as Influenced by Carbon,’’ Glastech. Ber. Glass Sci. Technol., 73, C2 361–368 (2000). 7. A. Kucuk, A. G. Clare, and L. E. Jones, ‘‘Differences Between Surface and Bulk Properties of Glass Melts I: Compositional Differences and Influence of Volatilization on Composition and Physical Properties,’’ J. Non-Cryst. Solids, 261 28–38 (2000). 8. A. Kucuk, A. G. Clare, and L. E. Jones, ‘‘Differences Between Surface and Bulk of Glass Melts II: Influence of Redox Ratio on Surface Properties of Silicate Melts,’’ Phys. Chem. Glasses, 41 75–80 (2000). 9. L. E. Jones and M. B. Ingram, ‘‘Measuring Hazardous Air Pollutants,’’ Am. Ceram. Soc. Bull., 74 [7] 66–67 (1995). 10. Glass IOF Technology Roadmap: www.eere.energy.gov/industry/glass/
310
International Journal of Applied Ceramic Technology—Akbar, Dutta, and Lee
11. W. L. Grosshandler, ‘‘Towards the Development of a Universal Fire Emulator-Detector Evaluator,’’ Fire Safety J., 29 113–127 (1997). 12. K. Hassmann, ‘‘SOFC Power Plants, the Siemens–Westinghouse Approach,’’ Fuel Cells, 1 [1] 78–84 (2001). 13. N. Szabo, C. Lee, J. Trimboli, O. Figueroa, R. Ramamoorthy, S. MidlamMohler, A. Soliman, H. Verweij, P. Dutta, and S. Akbar, ‘‘Ceramic-Based Chemical Sensors, Probes and Field-Tests in Automobile Engines,’’ J. Mater. Sci., 38 [21] 4239–4245 (2003). 14. A. L. Spetz, S. Nakagomi, and S. Savage, ‘‘High Temperature SiC-FET Chemical Gas Sensors,’’ Advances in Silicon Carbide Processing and Application. eds., S. E. Saddow and A. Agarwal. Artech House Publishers, Boston, 29–67, 2004. 15. J. A. Thiele and M. Pereira da Cunha, ‘‘High temperature SAW gas sensor on Langasite,’’ in Proceedings of IEEE Sensors, Vol. 2, IEEE International Conference on Sensors, 2nd, Toronto, ON, Canada, 2003, 769–772. 16. A. Modi, N. Koratkar, E. Lass, B. Wei, and P. M. Ajayan, ‘‘Miniaturized Gas Ionization Sensors Using Carbon Nanotubes,’’ Nature, 424 [6945] 171–174 (2003). 17. L. D. Birkefeld, A. M. Azad, and S. A. Akbar, ‘‘Carbon Monoxide and Hydrogen Detection by Anatase Modification of Titanium Dioxide,’’ J. Am. Ceram. Soc., 75 2694–2698 (1992). 18. P. K. Dutta, A. Ginwalla, B. Hogg, B. Patton, B. Chwieroth, Z. Liang, P. Gouma, M. Mills, and S. A. Akbar, ‘‘Interaction of Carbon Monoxide with Anatase Surfaces at High Temperatures: Optimization of a Carbon Monoxide Sensor,’’ J. Phys. Chem. B, 103 4412–4422 (1999). 19. N. Savage, S. A. Akbar, and P. K. Dutta, ‘‘Titanium Dioxide Based High Temperature Carbon Monoxide Selective Sensor,’’ Sens. Actuators B, 72 239– 248 (2001). 20. N. Savage, B. Chwieroth, A. Ginwalla, B. R. Patton, S. A. Akbar, and P. K. Dutta, ‘‘Composite n–p Semiconducting Titanium Oxides as Gas Sensors,’’ Sens. Actuators B, 79 17–27 (2001). 21. N. F. Szabo, H. Du, S. A. Akbar, A. A. Soliman, and P. K. Dutta, ‘‘Microporous Zeolite Modified Yttria Stabilized Zirconia (YSZ) Sensors for Nitric Oxide (NO) Determination in Harsh Environments,’’ Sens. Actuators B, 82 142–149 (2002). 22. A. Kohli, C. C. Wang, and S. A. Akbar, ‘‘Niobium Pentoxide as a LeanRange Oxygen Sensor,’’ Sens. Actuators B, 56 121–128 (1999). 23. A. K. M. S. Chowdhury, S. A. Akbar, S. Kapileshwar, and J. R. Schorr, ‘‘A Rugged Oxygen Gas Sensor with Solid Reference for High Temperature Applications,’’ J. Electrochem. Soc., 148 G91–G94 (2001). 24. B. Narayanan, S. A. Akbar, and P. K. Dutta, ‘‘A Phosphate-Based Proton Conducting Solid Electrolyte Hydrocarbon Gas Sensor,’’ Sens. Actuators B, 87 [3] 480–486 (2002). 25. C. Reddy, P. K. Dutta, and S. A. Akbar, ‘‘Detection of CO in a Reducing, Hydrous Environment Using CuBr as Electrolyte,’’ Sens. Actuators B, 92 351– 355 (2003). 26. N. F. Szabo and P. K. Dutta, ‘‘Strategies for Total NOx Measurement with Minimal CO Interference Utilizing a Microporous Zeolitic Catalytic Filter,’’ Sens. Actuators B, 88 [2] 168–177 (2003). 27. N. F. Szabo and P. K. Dutta, ‘‘Correlation of Sensing Behavior of Mixed Potential Sensors with Chemical and Electrochemical Properties of Electrodes,’’ Solid State Ionics, 171 183–190 (2004). 28. C. Lee, S. A. Akbar, and C. O. Park, ‘‘Potentiometric type CO2 Gas Sensor with Lithium Phosphrous Oxynitride Electrolyte,’’ Sens. Actuators B, 80 234– 242 (2001). 29. C. O. Park, C. Lee, S. A. Akbar, and J. Hwang, ‘‘The origin of Oxygen Dependence in a Potentiometric CO2 Sensor with Li-Ionconducting Electrolytes,’’ Sens. Actuators B, 88 53–59 (2003). 30. C. Lee, P. K. Dutta, R. Ramamoorthy, and S. A. Akbar, ‘‘Study of Mixed Ionic and Electronic Conduction in Li3PO4 Electrolyte for a CO2 Gas Sensor,’’ J. Electrochem. Soc., 153 H4–H1 (2006). 31. O. L. Figueroa, C. Lee, S. A. Akbar, N. F. Szabo, J. A. Trimboli, P. K. Dutta, N. Sawaki, A. A. Soliman, and H. Verweij, ‘‘Temperature-controlled CO, CO2 and NOx Sensing in a Diesel Engine Exhaust Stream,’’ Sens. Actuators B, 107 839–848 (2005). 32. R. Ramamoorthy, P. K. Dutta, and S. A. Akbar, ‘‘Oxygen Sensors: Materials, Methods, Designs and Applications,’’ J. Mater. Sci., 38 [21] 4271–4281 (2003). 33. M. L. Frank, M. D. Fulkerson, B. R. Patton, and P. K. Dutta, ‘‘TiO2-Based Sensor Arrays Modeled with Nonlinear Regression Analysis for Simultane-
34.
35. 36.
37.
38. 39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
Vol. 3, No. 4, 2006
ously Determining CO and O2 Concentrations at High Temperatures,’’ Sens. Actuators B, 87 [3] 471–479 (2002). F. Me´nil, V. Coillard, and C. Lucat, ‘‘Critical Review of Nitrogen Monoxide Sensors for Exhaust Gases of Lean Burn Engines,’’ Sens. Actuators B, 67 [1–2] 1–23 (2000). N. Docquier and S. Candel, ‘‘Combustion Control and Sensors: A Review,’’ Prog. Energy Combust. Sci., 28 107–150 (2002). N. Miura, H. Kurosawa, M. Hasei, G. Lu, and N. Yamazoe, ‘‘Stabilized Zirconia-Based Sensor Using Oxide Electrode for Detection of NOx in HighTemperature Combustion-Exhausts,’’ Solid State Ionics, 86-8 1069–1073 (1996). G. Lu, N. Miura, and N. Yamazoe, ‘‘High-Temperature Sensors for NO and NO2 Based on Stabilized Zirconia and Spinel-Type Oxide Electrodes,’’ J. Mater. Chem., 7 1445–1449 (1997). G. Reinhardt, R. Mayer, and M. Rosch, ‘‘Sensing Small Molecules with Amperometric Sensors,’’ Solid State Ionics, 150 [1–2] 79–92 (2002). N. Miura, G. Lu, M. Ono, and N. Yamazoe, ‘‘Selective Detection of NO by Using an Amperometric Sensor Based on Stabilized Zirconia and Oxide Electrode,’’ Solid State Ionics, 117 [3, 4] 283–290 (1999). P. Schmidt-Zhang, K. P. Sandow, F. Adolf, W. Go¨pel, and U. Guth, ‘‘A Novel Thick Film Sensor for Simultaneous O2 and NO Monitoring in Exhaust Gases,’’ Sens. Actuators B, 70 [1–3] 25–29 (2000). T. Ono, M. Hasei, A. Kunimoto, T. Yamamoto, and A. Noda, ‘‘Performance of the NOx Sensor Based on Mixed Potential for Automobiles in Exhaust Gases,’’ JSAE Rev., 22 [1] 49–55 (2001). S. Zhuiykov, T. Nakano, A. Kunimoto, N. Yamazoe, and N. Miura, ‘‘Potentiometric NOx Sensor Based on Stabilized Zirconia and NiCr2O4 Sensing Electrode Operating at High Temperatures,’’ Electrochem. Comm., 3 [2] 97– 101 (2001). N. Miura, M. Nakatou, and S. Zhuiykov, ‘‘Impedancemetric Gas Sensor Based on Zirconia Solid Electrolyte and Oxide Sensing Electrode for Detecting Total NOx at High Temperature,’’ Sens. Actuators B, 93 [1–3] 221– 228 (2003). U. Lampe, M. Fleischer, N. Reitmeier, H. Meixner, J. B. McMonagle, and A. Marsh, ‘‘New Metal Oxide Sensors: Materials and Properties,’’ Sensors Update. eds., W. Go¨pel and J. Hesse. VCH, New York, 1–36, 1996. M. D. Giulio, G. Micocci, A. Serra, A. Tepore, R. Rella, and P. Siciliano, ‘‘Characteristics of Reactively Sputtered Pt-SnO2 Thin Films for CO Gas Sensors,’’ J. Vac. Sci. Technol. A, 14 [4] 2215–2219 (1996). A. V. Tadeev, G. Delabouglise, and M. Labeau, ‘‘Influence of Pd and Pt Additives on the Microstructural and Electrical Properties of SnO2-Based Sensors,’’ Mater. Sci. Eng. B, 57 [1] 76–83 (1998). A. Ruiz, J. Arbiol, A. Cirera, A. Cornet, and J. R. Morante, ‘‘Surface Activation by Pt-Nanoclusters on Titania for Gas Sensing Applications,’’ Mater. Sci. Eng. C, 19 [1–2] 105–109 (2002). T. Schwebel, M. Fleischer, H. Meixner, and C.-D. Kohl, ‘‘CO-Sensor for Domestic Use Based on High Temperature Stable Ga2O3 Thin Films,’’ Sens. Actuators B, 49 [1–2] 46–51 (1998). T. Maruyama, S. Sasaki, and Y. Saito, ‘‘Potentiometric Gas Sensor for Carbon Dioxide Using Solid Electrolytes,’’ Solid State Ionics, 23 [1-2] 107–12 (1987). N. Miura, S. Yao, Y. Shimizu, and N. Yamazoe, ‘‘Carbon Dioxide Sensor Using Sodium Ion Conductor and Binary Carbonate Auxiliary Electrode,’’ J. Electrochem. Soc., 139 [5] 1384–1388 (1992). Y. Sadaoka, Y. Sakai, M. Matsumoto, and T. Manabe, ‘‘Solid State Electrochemical Carbon Dioxide Gas Sensor Using Zircon-Based Sodium Ionic Conductors,’’ J. Mater. Sci., 28 [8] 2035–2039 (1993). J. B. Bates, N. J. Dudney, D. C. Lubben, G. R. Gruzalski, B. S. Kwak, Yu Xiaohua, and R. A. Zuhr, ‘‘Thin-Film Rechargeable Lithium Batteries,’’ J. Power Sources, 54 [1] 58–62 (1995). B. Wang, J. B. Bates, F. X. Hart, B. C. Sales, R. A. Zuhr, and J. D. Robertson, ‘‘Characterization of Thin-Film Rechargeable Lithium Batteries with Lithium Cobalt Oxide Cathodes,’’ J. Electrochem. Soc., 143 [10] 3203–3213 (1996). Yu Xiaohua, J. B. Bates, G. E. Jellison, and F. X. Hart, ‘‘A Stable Thin-film Lithium Electrolyte: Lithium Phosphorous Oxynitride,’’ J. Electrochem. Soc., 144 524–532 (1997). S. Kimura, S. Ishitani, and H. Takao, ‘‘Principles and Development of a Thick-Film Zirconium Oxide Oxygen Sensor,’’ ACS Symp. Ser., 309 101–20 (1986).
www.ceramics.org/ACT
High-Temperature Ceramic Gas Sensors
56. N. Rajabbeigi, B. Elyassi, A. Khodadadi, S. Mohajerzadeh, and M. Sahimi, ‘‘A Novel Miniaturized Oxygen Sensor with Solid-State Ceria–Zirconia Reference,’’ Sens. Actuators B, 100 [1–2] 139–142 (2004). 57. B. C. H. Steele, ‘‘High Temperature Thermodynamic Measurements Involving Solid Electrolyte Systems,’’ Electromotive Force Measurements High Temperature Systems. ed. C. B. Alcock. Elsevier, New York, 3–27, 1968. 58. W. C. Maskell and B. C. H. Steele, ‘‘Solid State Potentiometric Oxygen Gas Sensors,’’ J. Appl. Electrochem., 16 475–489 (1986). 59. A. Hierlemann, O. Brand, C. Hagleitner, and H. Baltes, ‘‘Microfabrication Techniques for Chemical/Biosensors,’’ Proc. IEEE, 91 [6] 839–863 (2003). 60. S. Semancik, R. E. Cavicchi, M. C. Wheeler, J. E. Tiffany, G. E. Poirier, R. M. Walton, J. S. Suehle, B. Panchapakesan, and D. L. DeVoe, ‘‘Microhotplate Platforms for Chemical Sensor Research,’’ Sens. Actuators B, 77 [1–2] 579–591 (2001). 61. C. Hagleitner, A. Hierlemann, D. Lange, A. Kummer, N. Kerness, O. Brand, and H. Baltes, ‘‘Smart Single-Chip Gas Sensor Microsystem,’’ Nature, 414 [6861] 293–296 (2001). 62. M. Graf, D. Barrettino, S. Taschini, C. Hagleitner, A. Hierlemann, and H. Baltes, ‘‘Metal Oxide-Based Monolithic Complementary Metal Oxide Semiconductor Gas Sensor Microsystem,’’ Anal. Chem., 76 [15] 4437–4445 (2004). 63. R. K. Sharma, P. Chan, Z. Tang, G. Yan, I. Hsing, and J. Sin, ‘‘Sensitive, Selective and Stable Tin Dioxide Thin-Films for Carbon Monoxide and Hydrogen Sensing in the Integrated Gas Sensor Array Applications,’’ Sens. Actuators B, 72 [2] 160–166 (2001).
311
64. B. Joo, N. Choi, Y. Lee, J. Lin, B. Kang, and D. Lee, ‘‘Pattern Recognition of Gas Sensor Array Using Characteristics of Impedance,’’ Sens. Actuators B, 77 [1–2] 209–214 (2001). 65. J. W. Gardner, E. L. Hines, and M. Wilkinson, ‘‘Application of Artificial Neural Networks to an Electronic Olfactory System,’’ Meas. Sci. Technol., 1 446–451 (1990). 66. J. Haugen and K. Kvaal, ‘‘Electronic Nose and Artificial Neural Network,’’ Meat Sci., 49 [Suppl. 1] S273–S286 (1998). 67. R. E. Shaffer, S. L. Rose-Pehersson, and R. A. McGill, ‘‘A Comparison Study of Chemical Sensor Array Pattern Recognition Algorithms,’’ Anal. Chim. Acta, 384 [3] 305–317 (1999). 68. P. C. Jurs, G. A. Bakken, and H. E. McClelland, ‘‘Computational Methods for the Analysis of Chemical Sensor Array Data from Volatile Analytes,’’ Chem. Rev., 100 [7] 2649–2678 (2000). 69. T. Sundic, S. Marco, A. Perera, A. Pardo, S. Hahn, N. Barsan, and U. Weimar, ‘‘Fuzzy Inference System for Sensor Array Calibration: Prediction of CO and CH4 Levels in Variable Humidity Conditions,’’ Chemometr. Intel. Lab. Syst., 64 103–122 (2003). 70. N. Cristianini and J. Shawe-Taylor, An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods, Cambridge University Press, Cambridge, U.K., 2000. 71. G. Niebling and R. Muller, ‘‘Non-Linear Signal Evaluation with Linear Regression Techniques for Redundant Signals,’’ Sens. Actuators B, 25 [1–3] 805–807 (1995).
