This article is one of five papers to be presented exclusively on the web as part of the October 2000 JOM-e the electronic supplement to JOM.
	The following article appears as part of JOM-e, 52 (10) (2000), http://www.tms.org/pubs/journals/JOM/0010/Fergus/Fergus-0010.html JOM is a publication of The Minerals, Metals & Materials Society

Sensors for Multifunctional Applications: Overview

Using Chemical Sensors to Control Molten Metal Processing

TABLE OF CONTENTS
INTRODUCTION ELECTROCHEMICAL SENSORS DISSOLVED GASES ALLOY CONTROL CONCLUSIONS References

Chemical sensors can provide valuable information about changes in the composition of a molten alloy during processing. Real-time information on these compositional changes can be used to optimize the process for efficiency or product quality. There are applications in which chemical sensors can be used to improve control in the processing of various molten metals, including steel, aluminum, and zinc. In some cases, sensors are commercially available and widely used. For other applications, sensors are under development or are too costly. In this article, the current status of sensor development for some of these applications will be discussed.

INTRODUCTION

Optimization of industrial processes using computer control algorithms requires real-time information about the various process parameters. The degree to which a process can be optimized depends on the quantity and quality of information about the process, so improved process control can be enhanced through the development of advanced sensors. In the processing of molten metals, one important parameter is the chemical composition of the molten metal. Interactions between a molten metal and the atmosphere can change the composition of the metal. In some cases, undesired elements, such as oxygen or hydrogen, can be incorporated into the molten metal. In other cases, the preferential reaction of certain elements with the atmosphere can result in unintended changes in the alloy composition. Counteractive measures are available to correct for these undesired changes, but the effectiveness of these corrective measures requires real-time information on the chemical composition of the alloy and the surrounding atmosphere. Chemical sensors have been developed to monitor the chemical composition during the processing of molten metals.^1-3

ELECTROCHEMICAL SENSORS

Sensors based on solid electrolytes have several advantages in the processing of molten metals. The conductivities of solid electrolytes increase with increasing temperature, so the high operating temperature required during the processing of molten metals is well-suited to solid electrolyte based sensors. The output of solid electrolyte based sensors is determined by the thermodynamic properties of the molten metal and reference electrode, so the sensor does not require calibration. The supporting electronics are relatively simple, since the output of an electrochemical sensor is a d.c. voltage. In addition, solid electrolytes are generally stable compounds, which can withstand the harsh chemical environment in molten metals.

Ionic conduction in a given solid electrolyte will generally occur through the transport of a specific ionic point defect. The voltage generated across the electrolyte is proportional to the logarithm of the concentration of the mobile species, according to the Nernst equation. For example, oxide-ion vacancies are mobile in zirconia, so zirconia-based electrolytes can be used for oxygen sensors. Fortunately, electrochemical sensors are not limited to detecting the species that is mobile in the electrolyte. The equilibrium between an immobile species and mobile species establishes a concentration of the mobile species, which can generate a measurable voltage that is related to the concentration of the immobile species. Furthermore, an additional phase (referred to as an auxiliary electrode) can be added to provide sensitivity to a species that is not present in the electrolyte. Thus, electrochemical sensors can be designed for detecting a wide variety of species through judicious selection of the electrolyte and electrode materials.

There are a number of applications where chemical sensors can improve efficiency and product quality during the processing of molten metals. In some of these applications, commercial sensors are available and widely used, while in others the needed sensors are either not currently available or their application is not cost effective.

DISSOLVED GASES

Molten metals are generally processed in air and thus can react with the present gases. Although air consists mostly of nitrogen, because oxides are generally more stable than nitrides, the two most important gases in the processing of molten metals are oxygen and water vapor.

Oxygen in Steel

Oxygen from reaction with the atmosphere is removed from molten steel by adding aluminum or silicon alloys, which react with the oxygen to form oxides. Determining the optimal amounts of these alloys to add requires knowledge of the amount of oxygen in the steel, which is provided by an oxygen sensor. The most successful and widely used sensor in molten metals is the steelmaking oxygen sensor.^4,5 This oxygen sensor is based on a stabilized zirconia electrolyte. The reference electrode is a metal/metal-oxide mixture (most commonly Cr/Cr₂O₃), the equilibrium of which establishes a reference oxygen partial pressure. Although oxygen sensors have been used for many years, there are areas for improvement.

Figure 1. Schematic oxygen sensors for use in molten steel: (a) commercial disposable sensor,5 (b) extended-life sensor made by pressing electrolyte around the reference electrode,6 (c) non-isothermal sensor in which the reference electrode is at lower temperature than working electrode.8

One area for improvement is lifetime. Current sensors are used for one measurement and then discarded. Replacing these disposable sensors with extended-life sensors would both improve the quality of data obtained (i.e., continuous measurements could be made) and reduce the costs. One approach for extending the lifetime of current oxygen sensors has been alternative fabrication techniques, which improve the seal between the reference electrode and the molten steel.⁶ Figure 1b schematically shows an oxygen sensor in which the zirconia electrolyte is isostatically pressed around the reference electrode. This design provides an improved seal for the reference electrode compared to the disposable sensor, shown schematically in Figure 1a. Another approach has been in the design of a non-isothermal sensor, in which the reference electrode is outside the molten steel (Figure 1c).^7-9 This reduced temperature lessens the requirements on the reference electrode seal, but introduces an additional voltage due to the temperature difference between the two electrodes. However, this additional voltage can be compensated for through calculations using the Seebeck coefficient of the electrolyte or through judicious selection of the reference electrode. Another approach to extending the life of oxygen sensors is to use an applied voltage to electrochemically reverse the degradation of the reference. ¹⁰

Another area for improvement is in extending the oxygen partial pressure range over which the sensor can be used. Steels with very low oxygen contents can establish an oxygen partial pressure that is low enough for significant electronic conductivity to occur in zirconia. Significant electronic conduction in the electrolyte can lead to an erroneous sensor output. One approach to extending the low oxygen content limit is replacing zirconia with an alternative electrolyte, such as a stable perovskite compound,⁹ which remains a pure ionic conductor to lower oxygen partial pressures. Another approach is to use a double-layer tube, which prevents electronic conduction while maintaining satisfactory thermal shock resistance.¹¹

Hydrogen in Aluminum

During molten aluminum processing, the most important dissolved gas is hydrogen, which is produced when aluminum reacts with moisture to form aluminum oxide and hydrogen.^12,13 The solubility of hydrogen in the liquid aluminum is much higher than that in solid aluminum, so dissolved hydrogen can lead to porosity during solidification. To produce sound castings, molten aluminum must often be degassed to remove hydrogen. The efficiency of these degassing processes can be improved using real-time measurement of the hydrogen content during the process.

Systems for measuring the hydrogen content in molten aluminum are currently commercially available.^13,14 In the most common systems, an inert gas (usually nitrogen) flows through a probe and over the molten metal, such that chemical equilibrium between the hydrogen partial pressure in the gas and the concentration of hydrogen in the molten aluminum can be established (i.e., according to Sievert's law). The sample gas is then transported to an analyzer where its hydrogen content is measured and then related to the amount of hydrogen dissolved in the alloy. Although these nitrogen carrier gas systems are used in industry, the cost of the system is too high for some applications. In response to the need for lower cost hydrogen sensors, research has been performed on developing low-cost solid-state electrochemical sensors.

A direct electrochemical sensor for hydrogen should conduct hydrogen ions (i.e., protons). Researchers have modified sodium-ion conducting materials, so that they conduct protons to form proton-conducting electrolytes.^15-17 However, some of these electrolytes cannot withstand the operation temperature in molten aluminum. In addition, sodium is often present in aluminum alloys and may interfere with the sensor output. Gee and Fray¹⁸ used a hydride electrolyte, CaH₂, with calcium metal for the electrolyte in a hydrogen sensor. Although the sensor successfully measured hydrogen content, difficulties with materials stability limited further development.¹³ Other researchers have used oxygen-ion conductors for hydrogen sensors.^19-22 However, the sensor output depends on a mixed potential between hydrogen, oxygen, and water vapor, which may cause difficulties in interpretation and lead to interference from the atmosphere.

A new class of proton-conducting oxides has been developed that offers potential materials for use in hydrogen sensors. The most widely studied of these proton-conducting oxides are based on strontium and barium cerate.²³ Zhuiykov^24,25 recently reviewed potential proton-conducting materials for high-temperature applications and recommended barium cerate, calcium zirconate and La₂M₂O₇ as materials that could be used above 500°C. Strontium and barium cerate have been used in hydrogen sensors.^26-29 However, because of its superior stability in molten aluminum, calcium zirconate doped with indium has received the most attention for hydrogen sensors for use in molten metals.^30,31 In addition to being used in hydrogen sensors for molten aluminum,^32-34 indium-doped calcium zirconate has been used in hydrogen sensors for other metals, including copper,^35,36 copper-zinc,³⁶ and silver.³⁶

Although most of the hydrogen sensors reported use hydrogen gas as the reference electrode, Zheng et al. used a condensed phase (Ca/CaH₂) reference electrode.^28,29 The hydrogen gas reference electrode provides a better-defined reference potential, but requires the transport of the reference gas to the sensor. A condensed reference electrode eliminates this requirement and can simplify the sensor design, which can reduce the cost of fabrication and improve reliability. However, oxidation of the Ca/CaH₂ reference electrode was observed even in controlled laboratory testing, so additional development is required to produce a sensor that is sufficiently robust for industrial application.

Figure 2. Schematic proton-conducting solid electrolyte based hydrogen sensor for use in molten aluminum.32-34

The solid electrolyte sensor is similar to the commercial nitrogen carrier gas sensor in that rather than directly measuring the hydrogen concentration dissolved in the molten metal, the hydrogen partial pressure in a gas, which is equilibrated with the metal, is measured. However, as shown in Figure 2, in the solid electrolyte based system (with the hydrogen gas reference electrode), the solid electrolyte forms a sample chamber, which is directly in contact with the molten metal. In the nitrogen carrier gas system, the sample gas must be transported from the melt to the analyzer. Elimination of this need for transport of the sample gas in the solid electrolyte based sensor simplifies the system and eliminates potential errors associated with the sampling process. Although this could lead to improved reliability and reduced cost, the solid electrolyte based sensor has not yet matured sufficiently to compete commercially with the nitrogen carrier gas system.

ALLOY CONTROL

Chemical sensors can also be used to monitor and control alloying additions. Some examples of applications where chemical sensors for monitoring the alloy composition can be used to improve process efficiency or product quality are described in the following sections.

Reclamation of Aluminum Scrap

One step in recycling aluminum is removal of the magnesium by a process referred to as "demagging."³⁷ The magnesium is removed from the molten aluminum by injecting chlorine gas, which reacts with the molten aluminum to form AlCl₃ gas. If magnesium is present in the molten alloy, MgCl₂ forms by an exchange reaction with the AlCl₃ gas and then floats to the surface of the alloy, reducing the concentration of magnesium in the liquid aluminum. Once the magnesium concentration is reduced below a certain level, MgCl₂ no longer forms, so the gaseous AlCl₃ leaves the molten alloy. In addition to being an undesired emission, the formation of AlCl₃ results in the loss of aluminum and excessive use of time and energy, all of which reduce process efficiency. Measurement of the magnesium concentration in the alloy during this process could be used to determine exactly when the magnesium concentration is sufficiently low for the process to be terminated.

Figure 3. Schematic magnesium sensors for use in molten aluminum: (a) molten chloride electrolyte contained with porous ceramic plug,40 (b) molten chloride electrode impregnated into porous ceramic,41 (c) solid electrolyte based.48

The first magnesium sensors were developed using molten chloride electrolytes,^38-41 which had been used to measure the chemical activity of magnesium in molten aluminum.^42-45 In a recent plant test, 200 magnesium sensors based on a molten chloride electrolyte performed well in both aluminum-magnesium binary alloys as well as magnesium-containing commercial alloys.⁴⁶ Although sensors using molten electrolytes have been successfully developed, there are potential improvements by using solid electrolytes. Solid electrolytes do not require a crucible to contain the electrolyte or frits to separate the electrodes from the electrolyte, as is required for a molten electrolyte. For example, the sensor shown in Figure 3a uses a porous ceramic plug to contain the molten electrolyte. The design can be simplified, as reported by Zhang et al.,⁴¹ by impregnating the porous ceramic with the molten electrolyte (Figure 3b). However, the system is further simplified by using a solid electrolyte, which would not require impregnation (Figure 3c). Thus, a solid electrolyte based sensor has fewer components and a simpler design, which can reduce the cost of fabrication.

Solid electrolyte sensors have been reported using either a b²-alumina⁴⁷ or MgF₂⁴⁸ electrolyte. The b²-alumina based sensor uses the equilibrium between magnesium (dissolved in the molten aluminum alloy), magnesium oxide and sodium oxide (both oxides are dissolved in b²-alumina) to generate a voltage, which corresponds to the magnesium activity in the molten alloy. The inclusion of sodium oxide in the electrode equilibrium results in the sensor output, under some conditions, being affected by sodium impurities in the melt.

Although MgF₂ is a fluoride-ion conductor, no additional phases are needed for the magnesium sensor, since the equilibrium between magnesium dissolved in the alloy and the MgF₂ electrolyte establishes the fluorine partial pressure that generates the cell voltage. The results of this sensor are in excellent agreement with results for the same two electrodes using liquid electrolytes and the sensor responds rapidly (1 minute or less) to changes in the magnesium content in the alloy.⁴⁸ The use of small electrolyte tubes (3 mm diameter) produces sensors that are resistant to thermal shock upon insertion directly into molten aluminum and generate the theoretical voltage within about ten minutes of insertion.⁴⁹ The MgF₂-based sensor has been tested in silicon-containing alloys to evaluate possible interference of this common alloying element with sensor performance.⁵⁰ The sensor output was not affected by small silicon additions in the Al-Mg-Si ternary alloys.

Microstructure Control

Another application for chemical sensors in processing molten metals is monitoring the concentration of alloying elements. This is particularly important in cases where a small amount of a reactive or volatile alloying element is added. In such cases, the alloying element may be preferentially lost, which can cause significant changes to a small initial concentration. An important example of this is the eutectic modification of aluminum alloys. One method for controlling the eutectic microstructure formed during casting of aluminum is to add small amounts of sodium or strontium (a given foundry will generally use either strontium or sodium).^51-55 Both of these elements are reactive and can preferentially oxidize or vaporize during processing. In addition, excess amounts of alkali metals can be detrimental by causing edge cracking or hot shortness during rolling.^56,57 Since the concentration of these two elements is critical for controlling the cast microstructure and the concentrations may change during processing, strontium and sodium sensors have been developed.

Sodium

Since b²-alumina is one of the most common solid electrolytes, electrochemical sodium sensors based on a b²-alumina electrolyte have been developed.^58-63 Although the simplest electrochemical cell would use sodium metal as the reference electrode, the high volatility and reactivity of sodium metal at the temperatures used for processing molten aluminum make this difficult. Therefore, alternative reference electrodes, such as NaCl-Na₂CO₃⁶⁰ and Na_0.75CoO₂,⁶⁰ have been used.

The sodium sensor provides valuable information on the actual sodium content in the alloy, which can be used to compensate for sodium lost due to preferential vaporization or oxidation. Further improvements to this process can be achieved by using the b²-alumina electrolyte cells to electrochemically add sodium to a molten aluminum alloy.⁶³ This is done by applying a voltage to the sensor, such that sodium ions are transported from the reference electrode into the aluminum alloy. Since b²-alumina is a pure ionic conductor, the current through the cell provides a direct measurement of the amount of sodium entering the alloy.

Oxide electrolytes are susceptible to interference by oxygen and water vapor pressure, so other electrolytes have been investigated. One example is a sensor using a fluoride electrolyte (LaF₃),⁶⁴ which has also been used to measure the activity of lanthanum in molten aluminum.⁶⁵ Since LaF₃ does not contain sodium, a mixed-fluoride (NaLaF₄) auxiliary electrode is required to provide the sensitivity to sodium.⁶⁴ Although development of this fluoride-based sensor is not nearly as advanced as that of b²-alumina based sodium sensors, sensors based on fluoride, or other non-oxide electrolytes, can potentially be more resistant to interference from water vapor compared to oxide electrolytes-based sensors.

Strontium

Strontium sensors for use in molten aluminum have been reported using both oxide (strontium b-alumina)⁶⁶ and fluoride (SrF₂)⁶⁷ electrolytes. In the case of the strontium b-alumina based sensors, sodium is exchanged with strontium so that the electrolyte is a strontium-ion conductor. The SrF₂-based sensor uses the equilibrium between strontium dissolved in the alloy and the SrF₂ electrolyte to establish the measurable fluorine partial pressure (i.e., analogous to the MgF₂-based magnesium sensor previously described). Although the simplest reference electrode would be pure strontium, its reactivity and high melting point (relative to aluminum and magnesium) may be problematic for sensor operation and stability. Thus, a magnesium/MgF₂ mixture has been used as the reference electrode in the SrF₂-based strontium sensor.

Both of these strontium sensors have analogous magnesium sensors (i.e., b-alumina and fluoride based), so interference between the two alloying elements is possible. Magnesium can be added to aluminum alloys in relatively large amounts (several percent), so magnesium could possibly react with the strontium oxide (in the b-alumina) or the SrF₂. If a magnesium-containing phase (MgO or MgF₂) were to form on the working electrode, the sensor could respond to magnesium, rather than strontium, concentration. Thermodynamic analysis of both systems has shown that both electrolyte phases should be stable even for high levels of magnesium in the alloy.⁶⁸

Hot-Dip Galvanization of Steel

Hot-dip galvanization is a process in which a coating of zinc or a zinc alloy is deposited on steel sheet or wire by passing the sheet or wire through a molten zinc alloy bath. Although the primary constituent of the coating is zinc, small alloying additions are used to control the properties and microstructure of the coating.

Aluminum

The most important alloying addition used in hot-dip galvanization is aluminum. Aluminum additions are used to control coating thickness/roughness,^69-71 improve coating adherence/ductility,^72-74 and reduce dross formation, which can degrade coating quality.^75,76 Because of its strong effects on coating formation, the aluminum content must be precisely controlled.^76,77 Therefore, aluminum sensors have been developed.

Figure 4. Schematic aluminum sensors for use in molten zinc: (a) short-life molten chloride based sensor,79 (b) long-life molten chloride based sensors,79 (c) solid electrolyte based sensor.81-86

The most fully developed aluminum sensors are based on a molten chloride electrolyte, which contains AlCl₃ to provide the sensitivity to aluminum.^78-80 Disposable and durable (1- week life) molten-chloride-based sensors are commercially available, but are not cost-effective for some applications. During operation, AlCl₃ can react with water vapor and evaporate, which can limit the life of the sensor. To minimize loss of AlCl₃ and extend the sensor life, the long-life sensors include a b-alumina tube (in addition to the chloride electrolyte). The short- and long-life aluminum sensors are shown in Figure 4a and Figure 4b. Both sensors have a notched ceramic tube, which is broken off when the sensor is inserted into the melt and the lower density of the chloride is used to contain the electrolyte.

Researchers have developed aluminum sensors based on solid electrolytes, because such sensors have the potential for a simpler design (as shown in Figure 4c) and longer life, both of which can improve the cost effectiveness of the sensor. Sensors using zirconia-based electrolytes with an Al₂O₃ auxiliary electrode have been reported.^81-83 One problem with zirconia-based aluminum sensor is the formation of a continuous Al₂O₃ layer on the surface of the electrolyte, which eliminates the three-phase contact (solid electrolyte, auxiliary electrode, melt) needed to maintain the reference potential. In addition, the low oxygen partial pressure in molten aluminum may lead to electronic conductivity in the zirconia electrolyte.

Aluminum sensors have also been developed using fluoride electrolytes (SrF₂-LaF₃,⁸⁴ CaF₂,⁸⁵ and MgF₂^86,87), which remain pure ionic conductors in more reducing conditions as compared to zirconia. All these fluoride-based sensors use AlF₃ as the auxiliary electrode and all have been shown to respond to aluminum concentration. In general, the response time and reproducibility of the solid electrolyte based sensors is inferior to those of the molten chloride based electrolytes. However, the potential for lower fabrication cost and longer life may make solid electrolyte based sensors a more cost-effective alternative in the future.

Antimony

Another alloying element used to control the microstructure of hot-dip galvanized coatings is antimony.^69,88,89 Specifically, control of the antimony concentration is used to control the degree to which large grains (referred to as spangles) form on the galvanized coating. Spangle formation can affect both the appearance and properties, such as corrosion resistance and paintability, of the coating.

Thermodynamic measurements of the activity of zinc in zinc-antimony alloys have been made using a molten chloride electrolyte.⁹⁰ However, there are no reports of similar measurements of the activity of antimony in zinc-antimony alloys, which is more difficult because zinc is more active than antimony and will thus react with potential electrode materials. For example, if antimony oxide were used as the auxiliary electrode with an oxide-ion conducting electrolyte in a zinc-antimony alloy, the antimony oxide would be reduced by zinc to form zinc oxide, and the sensor would then respond to zinc rather than antimony.

An antimony sensor using a b-alumina electrolyte has been reported.^91,92 The sensor uses a NaSbO₃ auxiliary electrode to provide the sensitivity to antimony and has been shown to respond to antimony concentrations between 0.02 wt.% and 1 wt.%.

Another antimony sensor has been reported using a zirconia electrolyte.⁹³ As previously mentioned, antimony oxide is not stable in zinc. Therefore, an intermetallic compound, ZrSb₂, was used as the auxiliary electrode. The cell output generated a Nernstian response, which agreed with that predicted by calculations from thermodynamic measurements of the activity of zinc in zinc-antimony alloys. However, the lifetime of the sensor was limited due to reaction of the ZrSb2 auxiliary electrode with zinc to form zinc-zirconium intermetallic compounds.

CONCLUSIONS

Chemical sensors are valuable tools for improving the efficiency and quality control in the processing of molten metals. Oxygen sensors in molten steel are the most successful and widely used example, and their use will expand in the future as the sensors are further improved. Other sensors, such as hydrogen sensors for molten aluminum and aluminum sensors for molten zinc, are commercially available, but are cost prohibitive for some applications. In addition, other sensors, such as a magnesium sensor for molten aluminum, have shown promise in laboratory testing, but need additional development for commercialization. The use of chemical sensors in the metallurgical processing will continue to expand as the performance, reliability, and cost effectiveness of current sensors are improved and as new sensors are developed.

References

1. D.J. Fray, "The Use of Solid Electrolytes as Sensors for Applications in Molten Metals," Solid State Ionics, 86-88 (1996), pp. 1045-1054.
2. S. Seetharaman and D. Sichen, "Development and Application of Electrochemical Sensors for Molten Metals Processing," Emerging Separation Technologies for Metals II, ed. R.G. Bautista (Warrendale, PA: TMS, 1996), pp. 317-340.
3. D.J. Fray, "Potentiometric Gas Sensors for Use at High Temperatures," Mater. Sci. Tech., 16 (2000), pp. 237-242.
4. E.T. Turkdogan and R.J. Fruehan, "Review of Oxygen Sensors for Use in Steelmaking and of Deoxidation Equilibria," Can. Metall. Quart., 11 (2) (1972), pp. 371-384.
5. M. Iwase and Y. Waseda, "Recent Developments in Electrochemical Oxygen Sensors Used for Iron and Steelmaking," High Temp. Mater. Proc. 7 (2-3) (1986), pp. 123-131.
6. W.L. Worrell and Q. Liu, "Development of an Extended-Life Oxygen Sensor for Iron and Steel Melts," Solid State Ionics, 40-41 (1990), pp. 761-763.
7. T.H. Etsell and C.B. Alcock, "Non-isothermal Probe for Continuous Measurement of Oxygen in Steel," Solid State Ionics, 3/4 (1981), pp. 621-626.
8. K.T. Jacob and S.K. Ramasesha, "Design of Temperature-Compensated Reference Electrodes for Non-Isothermal Galvanic Sensors," Solid State Ionics, 34 (1989), pp. 161-166.
9. C.B. Alcock et al., "New Electrochemical Sensors for Oxygen Determination," Solid State Ionics, 53-56 (1992), pp. 39-43.
10. F. Li, Z. Zhu, and L. Li, "A New Way Extending Working-Life of Oxygen Sensor in Melt," Solid State Ionics, 70/71 (1994), pp. 555-558.
11. Q. Liu, "The Development of High Temperature Electrochemical Sensors for Metallurgical Processes," Solid State Ionics, 86-88 (1996), pp. 1037-1043.
12. S. Shivkumar, L. Wang, and D. Apelian, "Molten Metal Processing of Advanced Cast Aluminum Alloys," JOM, 43 (1) (1991), pp. 26-32.
13. M.M. Makhlouf, L. Wang, and D. Apelian, Measurement and Removal of Hydrogen in Aluminum Alloys (Des Plaines, IL: AFS, 1998), pp. 29-38.
14. X.-G. Chen et al., "Comparing Hydrogen Testing Methods for Wrought Aluminum," JOM, 46 (8) (1994), pp. 34-38.
15. R. Palombari and M. Casciola, "Proton-Metal Ion Conduction in Monoalkalai Salt Forms of a-Zirconium Phosphate," Solid State Ionics, 47 (1991), pp. 155-159.
16. S.F. Chehab et al., "Hydrogen Sensor Based on Bonded Hydronium," Solid State Ionics, 45 (3-4) (1991), pp. 299-310.
17. J. Gulens et al., "Hydrogen Electrolysis Using a NASICON Solid Protonic Conductor," Solid State Ionics, 28-30 (1988), pp. 622-626.
18. R. Gee and D.J. Fray, "Instantaneous Determination of Hydrogen Content in Molten Aluminum and Its Alloys," Metall. Trans. B, 9B (1978), pp. 427-430.
19. G. Lu, N. Miura, and H. Yamazoe, "High-Temperature Hydrogen Sensor Based on Stabilized Zirconia and a Metal Oxide Electrode," Sensors and Actuators B, 35-36 (1996), pp. 130-135.
20. G. Lu, N. Miura, and N. Yamazoe, "Mixed Potential Hydrogen Sensor Combining Oxide Ion Conductor with Oxide Electrode," J. Electrochem. Soc., 143 (7) (1996), pp. L154-L155.
21. N. Hara and D.D. MacDonald, "Development of Dissolved Hydrogen Sensor Based on Yttria-Stabilized Zirconia with Noble Metal Electrolytes," J. Electrochem. Soc., 144 (12) (year), pp. 4152-4157.
22. Y. Tan and T.C. Tan, "Characteristics and Modeling of a Solid State Hydrogen Sensor," J. Electrochem. Soc., 141 (2) (1994), pp. 461-467.
23. T. Norby, "Proton Conduction in Oxides," Solid State Ionics, 40/41 (1990), pp. 857-862.
24. S. Zhuiykov, "Development of High-Temperature Hydrogen Sensor Based on Pyrochlore Type of Proton-Conductive Solid Electrolyte," Ceram. Eng. Sci. Proc., 17 (3) (1996), pp. 179-186.
25. S. Zhuiykov, "Hydrogen Sensor Based on a New Type of Proton Conductive," Ceramic. Int. J. Hydrogen Energy, 21 (9) (1996), pp. 749-759.
26. H. Iwahara et al., "Nernstian Hydrogen Sensor Using BaCeO₃-Based, Proton-Conducting Ceramics Operative at 200-900°C," J. Electrochem. Soc., 138 (1) (1991), pp. 295-299.
27. M. Zheng and X. Chen, "Preparation and Electrochemical Characterization of SrCeO₃-Based Proton Conductor," Solid State Ionics, 70/71 (1994), pp. 595-600.
28. M. Zheng and X. Zhen, "SrCeO₃-Based Solid Electrolyte Probe Sensing Hydrogen Content in Molten Aluminum," Solid State Ionics, 59 (1993), pp. 167-169.
29. M. Zheng and X. Zhen, "Hydrogen Probe Equipped with SrCeO₃-Based Proton Conductor and Ca/CaH₂ Reference Electrode," Metall. Mater. Trans. B, 24B (1993), pp. 789-794.
30. T. Yajima et al., "Proton Conduction in Sintered Oxides Based on CaZrO₃," Solid State Ionics, 47 (1991), pp. 271-275.
31. N. Kurita et al., "Proton Conduction Domain of Indium-Doped Calcium Zirconate," J. Electrochem. Soc., 142 (5) (1995), pp. 1552-1559.
32. T. Yajima et al., "Measurement of Hydrogen Content in Molten Aluminum Using Proton Conducting Ceramic Sensor," Keikinzoku, 42 (5) (1992), pp. 263-267.
33. T. Yajima et al., "A New Hydrogen Sensor for Molten Aluminum," Sensors and Actuators B, 13-14 (1993), pp. 697-699.
34. T. Yajima et al., "Application of Hydrogen Sensor Using Proton Conductive Ceramics as a Solid Electrolyte to Aluminum Casting Industries," Solid State Ionics, 79 (1995), pp. 333-357.
35. N. Kurita et al., "The Measurement of Hydrogen Activities in Molten Copper Using Oxide Protonic Conductor," Metall. Mater. Trans. B, 27B (1996), pp. 929-935.
36. N. Fukatsu et al., "Hydrogen Sensor for Molten Metals Usable up to 1500 K," Solid State Ionics, 113-115 (1998), pp. 219-227.
37. B.L. Tiwari, "Demagging Processes for Aluminum Alloy Scrap," JOM, 34 (7) (1982), pp. 54-58.
38. D.H. DeYoung, J.B. Moreland, and R. Mutharasan, "Closed Loop Melt Composition Control by In-line Computer-Aided Alloying," Light Metals 1995, ed. J. Evans (Warrendale, PA: TMS, 1995), pp. 840-850.
39. B.L. Tiwari and B.J. Howie, "Electrochemical Probe for Measuring Magnesium Concentration in Molten Aluminum," U.S. patent 4,601,810 (22 July 1986).
40. B.J. Howie and B.L. Tiwari, "Determination of Magnesium in Molten Aluminum Alloy Using an Electrochemical Sensor," Light Metals 1989, ed. P.G. Campbell (Warrendale, PA: TMS, 1989), pp. 895-902.
41. L Zhang et al., "Electrochemical Sensor for Measuring Magnesium Content in Molten Aluminum," J. Appl. Electrochem., 26 (3) (1996), pp. 269-275.
42. B.L. Tiwari, "Thermodynamic Properties of Liquid Al-Mg Alloys Measured by the EMF Method," Metall. Trans. A, 18A (1987), pp. 1645-1651.
43. M.M. Tsyplakova and Kh.L. Strelets, "Study of the Thermodynamic Properties of the Magnesium-Aluminum System by the EMF Method," J. Applied Chem. USSR, 42 (11) (1969), pp. 2354-2359.
44. E.E. Lukachenko and A.M. Pogodayev, "The Thermodynamic Functions of Liquid Mg-Al Alloys," Russian Metallurgy (Metally), 5 (1971), pp. 69-72.
45. G.R. Belton and Y.K Rao, "A Galvanic Cell Study of Activities in Mg-Al Liquid Alloys," Trans. Metall. Soc. AIME, 245 (1969), pp. 2189-2193.
46. J. Vangrunderbeek et al., "Continuous In-line Monitoring of Magnesium in Aluminum," Light Metals 1999, ed. C.E. Eckert (Warrendale, PA: TMS, 1999), pp. 1005-1009.
47. S. Larose, A. Dubreuil, and A.D. Pelton, "Solid Electrolyte Probes for Magnesium, Calcium and Strontium in Molten Aluminum," Solid State Ionics, 47 (1991), pp. 287-295.
48. J.W. Fergus and S. Hui, "Solid Electrolyte Sensor for Measuring Magnesium in Molten Aluminum," Metall. Mater. Trans. B, 26B (1995), pp. 1289-1291.
49. J.W. Fergus, "Chemical Sensors for Use in Processing Molten Metals," AFS Transactions, 98-22 (1998), pp. 125-130.
50. J.W. Fergus et al., "Anomalous Output of Magnesium Sensor," Mater. Sci. Tech., 13 (1997), pp. 533-536.
51. S.-Z. Lu and A. Hellawell, "Modification of Al-Si Alloys: Microstructure, Thermal Analysis, and Mechanisms," JOM, 47 (11) (1995), pp. 38-40.
52. B. Closset et al., "Microstructures and Properties of Strontium Treated Aluminium Electrical Conductor Alloys," Light Metals 1996, ed. W. Hale (Warrendale, PA: TMS, 1996), pp. 737-744.
53. D. Emadi et al., "The Effects of Sr-Modification on the Melt Hydrogen Content and the Hydrogen Solubility in the Solid and Liquid Al-Si Alloys," in Ref. 52, pp. 721-728.
54. F. Paray et al., "Metallurgical Effects of Strontium on Wrought 6061 Alloys," in Ref. 52, pp. 717-712.
55. P.C. Van Wiggen, "The Flexible AlSr 10/15 Rod," in Ref. 52, pp. 755-759.
56. D.J. Fray, "Possible Uses of Sensors in the Aluminium Industry," Mater. Sci. Tech., 3 (1987), pp. 61-65.
57. C.J. Simensen and M. Nilmani, "A Computer Model for Alkali Removal from Molten Aluminium," in Ref. 52, pp. 995-1000.
58. D.J. Fray, "Solid Electrolytes and the Analysis of Molten Metals," Chem. Ind., (1992), pp. 445-448.
59. D.J. Fray and R.J. Brisley, "Determination of the Sodium Activity in Aluminum and Aluminum Silicon Alloys Using Sodium Beta Alumina," Metall. Trans. B, 14B (1983), pp. 435-440.
60. L. Zhang et al., "Reference Electrode of Simple Galvanic Cells for Developing Sodium Sensors for Use in Molten Aluminum," Metall. Mater. Trans. B, 27B (1996), pp. 794-800.
61. P.C. Yao and D.J. Fray, "Sodium Activity Determinations in Molten 99.5% Aluminium Using Solid Electrolytes," J. Appl. Electrochem., 15 (1985), pp. 379-386.
62. J.C. Dekeyser et al., "An Electrochemical Sensor for Aluminum Melts," Sensors and Actuators B, 24-25 (1995), pp. 273-275.
63. G. Doughty et al., "b-Alumina for Controlling the Rate of Sodium Addition to Aluminum Alloys," Solid State Ionics, 86-88 (1996), pp. 193-196.
64. Q. Zhang, "Fluoride-Based Sodium Sensor for Use in Molten Aluminum," M.S. Thesis, Auburn University, 1998.
65. X. Lisheng, S. Zhitong, and W. Changzhen, "Activity of Dissolved La in Liquid Al," Scand. J. Metall., 24 (1995), pp. 86-90.
66. A.J. Kirchnerova and A.D. Pelton, "Solid Electrolyte Probe for Strontium Employing a SrCl₂-AgCl/Ag Reference and a Thermodynamic Evaluation of the SrCl₂-AgCl System," Solid State Ionics, 93 (1996), pp. 165-170.
67. D. Hardy, "Fluoride-Based Strontium Sensor for Use in Molten Aluminum," M.S. Thesis, Auburn University, 1998.
68. J.W. Fergus, "Sensors for Use in Systems Containing Multiple Reactive Metals," Light Metals 1999, ed. C.E. Eckert (Warrendale, PA: TMS, 1999), pp. 1131-1134.
69. F.A. Fasoyinu and F. Weinberg, "Spangle Formation in Galvanized Sheet Steel Coatings," Metall. Trans. B, 21B (1990), pp. 549-558.
70. K.L. Lin et al., "Growth Behavior and Corrosion Resistance of 5% Al-Zn Coating," Corrosion, 49 (9) (1993), pp. 759-762.
71. R. Gutenberg, J. Lait, and F. Weinberg, "Changing Al and Pb Bath Concentrations in Galvanized Sheet Steel," Can. Metall. Quart., 29 (4) (1990), pp. 307-312.
72. L.A. Rocha and M.A. Barbosa, "Microstructure, Growth Kinetics, and Corrosion Resistance of Hot-Dip Galvanized Zn-5% Al Coatings," Corrosion, 47 (7) (1991), pp. 536-541.
73. Y. Yoshitaka, M. Arai, and T. Nakamori, "Effect of Al in Molten Zinc on Adhesion Strength in Galvannealed Steel," Tetsu-to Nagane, 80 (8) (1994), pp. 67-72.
74. S.J. Makimattila et al., "The Effect of Intermetallic Layer on the Adherence of a Hot-Dip Galvanized Coating," Scripta Metall., 19 (2) (1985), pp. 211-214.
75. H. Yamaguchi and Y. Hisamatsu, "Reaction of Dross Formation in Continuous Galvanizing," Tetsu-to Hagane, 60 (1) (1974), pp. 96-103.
76. V. Jagannathan, "Emerging Technologies in the Hot-Dip Coating of Automotive Sheet Steel," JOM, 45 (8) (1993), pp. 48-51.
77. N.-Y. Tang, "Refined 450°C Isotherm of Zn-Fe-Al Phase Diagram," Mater. Sci. Tech., 11 (1995), pp. 870-873.
78. S. Yamaguchi et al., "Development of Aluminum Sensor for Molten Zinc Bath Using Composite Salt Electrolyte," CAMP-ISIJ, 4 (1991), p. 669.
79. S. Yamaguchi, N. Fukatsu, and H. Kimura, "Development of Al sensor in Zn Bath for Continuous Galvanizing Processes," International Galvatech '95 Conf. Proc. (Warrendale, PA: ISS, 1995), pp. 647-655.
80. N. Qiang, N.Y. Tang, and G.R. Adams, "Applications of Al Sensors in Continuous Galvanizing," Sensors and Modeling in Materials Processing, ed. S. Viswanathan, R.G. Reddy, and J.C. Malas (Warrendale, PA: TMS, 1997), pp. 397-408.
81. S. Matsubara et al., "Determination of Aluminum Content in Molten Zinc by the E.M.F. Method Using Zirconia Solid Electrolyte," ISIJ International, 35 (5) (1995), pp. 512-518.
82. T.C. Wilder, "Method for Determining the Concentration of a Metal in an Alloy Melt," U.S. patent 3,816,269 (11 June 1974).
83. S. Matsubara et al., "Determination of Aluminum Sensor for Molten Zinc Bath Using Zirconia Solid Electrolyte," Tetsu-to Hagane, 79 (2) (1993), pp. 180-186.
84. C.B. Alcock and B. Li, "Electrochemical Sensor for Determining the Level of a Certain Metal in Metals and Alloys," U.S. patent 5,256,272 (26 October 1993).
85. S. Matsubara et al., "Determination of Aluminum Content in Molten Zinc by the E.M.F. Method Using Calcium Fluoride Solid Electrolyte," Materials Transactions JIM, 36 (10) (1995), pp. 1255-1262.
86. J.W. Fergus and S. Hui, "Solid-State Aluminum Sensor for Use in Molten Zinc," Sensors and Modeling in Materials Processing, ed. S. Viswanathan, R.G. Reddy, and J.C. Malas (Warrendale, PA: TMS, 1997), pp. 929-935.
87. J.W. Fergus, "The Status of Chemical Sensors for Hot-Dip Galvanization," JOM, 48 (9) (1996), pp. 38-41.
88. F.A. Fasoyinu and F. Weinberg, "The Surface-Topography of Sheet Steel Galvanized Coatings," Can. Metall. Quart., 32 (2) (1993), pp. 185-192.
89. S. Chang and J.C. Shin, "Effect of Antimony Addition on Hot Dip Galvanized Coating," Corrosion, 36 (1994), pp. 1425-1436.
90. I.B. Rubin, K.L. Komarek, and E. Miller, "Thermodynamic Properties and Compound Cluster Formation in Liquid Zinc-Antimony Alloys," Z. Metallkde., 65 (1974), pp. 191-199.
91. G.M. Kale, A.J. Davidson, and D.J. Fray, "Solid-State Sensor for Measuring Antimony in Non-ferrous Metals," Solid State Ionics, 86-88 (1996), pp. 1101-1105.
92. D.J. Fray and R.V. Kumar, "Method for Measuring a Minor Element in a Molten Metal," U.S. patent 5,192,404 (09 March 1993).
93. J.W. Fergus and S. Hui, "Solid Electrolyte Based Galvanic Cell for Measuring the Antimony Concentration in Molten Zinc," J. Electrochem. Soc., 143 (1996), pp. 2498-2502.

Jeffrey W. Fergus is with the Materials Research and Education Center at Auburn University.

For more information, contact J.W. Fergus, Auburn University, Materials Research and Education Center, 201 Ross Hall, Auburn University, AL 36849; (334) 844-3405; fax (334) 844-3400; e-mail jwfergus@eng.auburn.edu