Anal. Chem. 2008, 80, 8807–8813
Technical Notes Hydrogen Sensor Based on Au and YSZ/HgO/Hg Electrode for in Situ Measurement of Dissolved H2 in High-Temperature and -Pressure Fluids R. H. Zhang,* S. M. Hu, X. T. Zhang, and Y. Wang Open Research Laboratory of Geochemical Kinetics, Institute of Mineral Resources, Chinese Academy of Geological Sciences, Baiwanzhuang Road 26, Beijing 100037, P. R. China Gold as a hydrogen-sensing electrode for in situ measurement of dissolved H2 in aqueous solutions under extreme conditions is reported. The dissolved H2 sensor, constructed with a Au-based sensing element and coupled with a YSZ/HgO/Hg electrode, is well suited for determining dissolved H2 concentrations of aqueous fluids at elevated temperatures and pressures. The Au electrode is made of Au wire mounted in a quartz bar, which can be pressurized and heated in the high-pressure and -temperature conditions. The Au-YSZ sensor has been tested for its potential response to the concentrations of dissolved H2 in fluids by using a flow-through reactor at high temperatures up to 400 °C and pressures to 38 MPa. Good sensitivity and linear response between the hydrogen concentrations in the fluids and the H2 sensor potentials are reported for hydrogen gas in the concentration range of 0.1-0.001 M H2 in aqueous fluids at temperatures up to 340 °C and 30 MPa. Nernstian response of the cell potential to dissolved H2 in fluids was determined at 340 °C and 30 MPa, described as follows: ∆E ) 0.9444 + 0. 0603 log mH2 The experimental results indicate that the Au-YSZ/HgO/Hg cell can be used to measure the solubility of H2 in aqueous fluid at temperatures and pressures near to the critical state of water. Thus, this type of Au hydrogen sensor could be easily used for in situ measurement of H2 in hydrothermal fluids in a high-pressure vessel, or at midocean ridge, due to its structure of compression resistance. There is an ongoing need to accurately detect molecular hydrogen (H2) for a variety of applications in industrial process control and experimental and natural observations. Such applications include chemical, metallurgical, food and petroleum refining, rocket fuels, and electric power production.1 Particularly, the H2 sensor has been used to directly in situ measure H2, pH, and other chemical parameters of hydrothermal fluids in a wide temperature * To whom correspondence should be addressed. Tel: +8610 68329535. Fax: +8610 68327063. E-mail:
[email protected]. (1) Grimes, C. A.; Ong, L. G.; Varghese, O. K.; Yang, X.; Mor, G.; Paulose, M.; Dickey, E. C.; Ruan, C.; Pishko, M. V.; Kendig, J. W.; Mason, A. J. Sensors 2003, 3, 69–82. 10.1021/ac800948x CCC: $40.75 2008 American Chemical Society Published on Web 10/15/2008
range (2-400 °C) and at midocean ridge2-5 and H2-bearing fluids in chemical processes, such as Fischer-Tropsch type reactions.6 Scientists have been working on development of novel sensors for in situ measurement of H2 and other chemical parameters in liquids at extreme conditions.5-9 Recent advances in materials synthesis make great progress in the development of sensor technology. Yttria-stabilized zirconia (YSZ) ceramic-based chemical sensors have been used to in situ measure chemical parameters (pH, H2, H2S) of hydrothermal fluids in a high-pressure vessel and in the vents at midocean ridges.5,10-17 In situ technique for measuring/monitoring dissolved H2 in supercritical fluids was reported earlier. Its approach relied on (2) Mandelis, A.; Christofides, C. Physics, Chemistry and Technology of Solid State Gas Sensor Devices; Wiley: New York, 1993. (3) Madon, M. J.; Morrison, S. R. Chemical sensing with solid state Devices; Academic Press: New York, 1989. (4) Macdonald, D. D.; Mckubre, M. C. H.; Scott, A. C.; Wentrcek, P. R. Ind. Eng. Chem. Fundam. 1981, 20, 290–297. (5) Ding, K.; Seyfried, W. E., Jr J. Solution Chem. 1996, 25 (5), 421–433. (6) Workman, J., Jr.; Veltkamp, D. J.; Doherty, S.; Anderson, B. B.; Creasy, K. E.; Koch, M.; Tatera, J. F.; Robinson, A. L.; Bond, L.; Burgess, L. W.; Bokerman, G. N.; Ullman, A. H.; Darsey, G. P.; Mozayeni, F.; Bamberger, J. A.; Greenwood, M. S. Anal. Chem. 1999, 71, 121R–180R. (7) Walker, C. E.; Xia, Z.; Foster, Z. S.; Lutz, B. J.; Fan, Z. H. Electroanalysis 2008, 20 (6), 663–670. (8) Lutz, B.; Fan, Z. H.; Burgdorf, T.; Friedrich, B. Anal. Chem. 2005, 77, 4969–4975. (9) Sakthivel, M.; Weppner, W. Sensors 2006, 6, 284–297. (10) Macdonald, D. D.; Hettiarachchi, S.; Lenhart, S. J. J. Solution Chem. 1988, 17 (8), 719–732. (11) Macdonald, D. D.; Hettiarachchi, S.; Song, H.; Makela, K.; Emerson, R.; Ben-Haim, M. J. Solution Chem. 1992, 21, 849–881. (12) Macdonald, D. D.; Liu, J.; Lee, D. J. Appl. Electrochem. 2004, 34, 577– 582. (13) Kriksunov, L. B.; Macdonald, D. D. Sens. Actuators, B: Chem. 1994, 22, 201–204. (14) Zhang, R. H.; Zhang, X. T.; Hu, S. M. Mater. Lett. 2006, 60, 3170–3174. (15) Zhang, X. T.Study of Zr/ZrO2 film material for high temperature and pressure chemical sensor: manufacture and nature characteristics. Thesis, China University of Geosciences, 2005; in Chinese, with English abstract. (16) Zhang, X. T.; Zhang, R. H.; Hu, S. M. High Technol. Lett. 2004, (10), 360– 363. (17) Zhang, X. T.; Zhang, R. H.; Hu, S. M. J. Mater. Sci. 2007, 42 (14), 5632– 5640.
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
8807
the potential-metric cell Pt-YSZ/HgO/Hg.18-20 Virtually, all Pt sensor methods have limitations. When brought into contact with sulfide-bearing fluids, platinoid metals will be poisoned, and thus deactivated.21 And also, a reaction product of a Pt-sulfide compound could form on the Pt surface. In addition, at high temperatures, platinum could not be suited for making hydrogen electrode due to its strongly increased H2 permeability with increasing temperature.22 Recently, H2 oxidation at the surface of a gold electrode at high temperatures was demonstrated.23 The gold electrode can be used to measure H2 of aqueous fluids at high temperatures up to 400 °C, which would alternate the Pt electrode.5 This study is to develop a new construction method to fabricate a Au electrode in order to directly in situ measure H2 of hydrothermal fluids at high temperatures up to 400 °C. EXPERIMENTAL SECTION Design of High-Temperature and -Pressure Chemical Sensors. Gold is used as the H2-sensing electrode. By coupling the H2 electrode (Au) with the YSZ/HgO/Hg electrode, as depicted in cell eq 1, the overall potential of the cell is only a function of dissolved H2. Thus, the electrochemical cell is described as follows. Au|H2, H+, H2O|YSZ|HgO|Hg
(1)
The YSZ membrane electrode used here was modified after Macdonald et al.11 The YSZ tube (9% Y2O3) (from Luoyang Institute of Ceramics, Henan, China) is 6.35 cm in diameter, 0.082 in wall thickness, and 14.7 cm in length. The 1:1 mixture (volume ratio) of native Hg and HgO (99.9%, pure, red) is put into the bottom of the YSZ tube, which is ∼2.5 cm in height. A Pt wire is put into the mixture. Then, ceramics powder is mixed with water on the top of the mixture of Hg/HgO. The Pt wire is sealed into the YSZ tube, from which part of the Pt wire was exposed for connection to the voltmeter. The YSZ/HgO/Hg structure is shown in Figure 1. Preparing of H2 Sensor. A H2 sensor is fabricated by mounting a gold wire in quartz tubing at one end of a quartz tubing and the gold wire connected to an alloy wire of Mo-Ni-W inside of the tubing (wall is 3 mm inside, inner diameter is 0.5 mm). The alloy wire comes out the tubing as a lead connected to a potential meter. Melting each end of the quartz tubing, the quartz tubing becomes a gland, which looks like a bar. This Au electrode can be used to measure hydrogen gas in fluids at high temperatures up to 400 °C or higher. Au electrode structure is shown as Figure 2. (18) Hettiarachchi, S.; Makela, K.; Song, H.; Macdonald, D. D. J. Electrochem. Soc. 1992, 139, L3-L4. (19) Hammer, B.; Norskov, J. K. Nature 1995, 376, 238–240. (20) Ding, K.; Seyfried, W. E., Jr. Geochim Cosmochim Acta 1995, 59, 4769– 4773. (21) Clemens, J. D.; McKibben, M. A. In Hydrothermal Experimental Techniques; Ulmer, G. C., Barnes, H. L. Eds.; Wiley-Interscience: New York, 1987; p 138. (22) Gunter, W. D.; Myers, J.; Girsperger, S. In Hydrothermal Experimental Techniques; Ulmer, G. C., Barnes, H. L., Wiley-Interscience: New York, 1987; p 100. (23) Weewer, R.; Hemmes, K.; de Wit, J. H. W. J. Electrochem. Soc. 1995, 142, 389–397.
8808
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
Figure 1. Au electrode: 1, metal lead wire; 2, heat-shrinkable PTFE tube; 3, insulation coating and mixed Teflon and graphite seal; 4, steel seal; 5, mixed Teflon and graphite seal; 6, steel connect; 7, pressure fitting; 8, Au wire; 9, quartz tube; 10, connection of Au wire and alloy of mixed Mo, Ni, and W metals.
Experimental Approach. A high-temperature and -pressure experimental system was set up to measure the response of the Au-YSZ/HgO/Hg sensor potentials to dissolved H2 in fluids. The electrochemical experiments were performed in a computercontrolled Ti flow reactor, which is especially resistant to corrosion by acid NaCl-bearing fluids. That is a flow-through reactor system with a pressure vessel. For testing H2 sensor, the Au-YSZ/HgO/ Hg electrodes were then installed into the high-pressure flow reactor. The experimental flow reactor used to determine the sensitivity of the gold electrode to H2 in fluids at high temperatures and pressures is shown schematically in Figure 3a and b. A backpressure regulator was used to control the pressure of the flowing fluid in the measurement system. An additional pressure gauge provided pressure readings for the flowing system of the H2bearing fluid (H2 1% and N2 99%). Also, two pressure sensors were connected to both the flowing system of aqueous fluid and the flowing system of H2 + N2, which provide accurate pressure readings with ±0.01 MPa (0.1 bar). The high-pressure liquid pump offers a liquid flow velocity range from 0.1 to 9.9 mL/min. When we operated the measurement system of flowing fluids and maintained a constant high pressure of the fluids, e.g., at 30-32 MPa, the pressure of the flowing fluids varied within ±0.05 MPa (0.5 bar). The pressure of the measurement system was maintained to be stable through the back-pressure regulator. The heating system was used to control the cell temperature including a thermocouple, a heating furnace, and a proportional temperature controller. Temperature reading is accurate within ±0.1 °C, and stability within ±0.1 °C.
aqueous solution, which comes out from the outlet of the experimental system. For the third step, the gas cylinder is closed and the pressure of the fluid flowing though system is increased immediately. And also, the influent fluid continuously passes through the vessel; thus the process of decreasing H2 in the fluid occurs in the experimental system. The voltmeter records the response of the Au-YSZ/HgO/Hg cell with the concentration of H2 dissolved in the aqueous solution, during the decrease of H2 in the solution. In the whole process, the output H2-bearing solution samples are continuously collected. The H2-bearing fluid samples were collected continuously from the outlet by using a glass bottle with a plastic cover. One bottle (25 mL, SGE Co) of aqueous sample is taken in 20 min (more or less, depending on flow rate). A pinhead is mounted at the outlet tube. As the pinhead is inserted into the plastic cover of the sample bottle, the H2-bearing fluids pass through the pinhead into the bottle. This sampling method is to avoid the room air pollution. Dissolved H2 concentrations in the fluid were determined by gas chromatography (GC-9A, Shimadzu; thermoconductivity detector, C-R2A Shimadzu; Soom injector, 25-mL glass bottle, SGE).
Figure 2. YSZ/HgO/Hg electrode structure: 1, Pt wire; 2, Teflon tube; 3, mixed Teflon and graphite seal; 4, steel seal; 5, graphic seal; 6, steel connection; 7, graphic seal; 8, HgO/Hg; 9, YSZ tube; 10, ceramic fillings.
There are the gas cylinder (H2 1% and N2 99%, 10 MPa pressure), gas flow meter, gas flow controller, and gas pressure gauge, connected to the experimental system, shown in Figure 3a. There are several holes made on the top and bottom of the pressure vessel, in which the fluid inlet, fluid outlet, and thermocouple can be put. And also there are two holes on the side face of the vessel, where two different electrodes can be put, shown in Figure 3b. Thus, the thermocouple, tubing, Au electrode, and YSZ/HgO/Hg electrode were sealed into the pressure vessel. The experimental system is connected to a computer, which records all of data of cell potentials, temperature, and pressure and monitors all of data simultaneously. It can obtain 6-channel data in 1 s. Experimental Method. Experiments commenced with the continuous flow (2-4 mL/min) of 3.5 wt % NaCl solution. For the test solutions, we prepare deionized and degassed water by adding NaCl (3.5 wt % NaCl). As the fluids flowed onto the reactor, cell potential was measured with a voltmeter with an input impedance of 1013 Ω. To operate the experiment for measurement of H2 in the fluid, we pump the NaCl-H2O solution to pass through the pressure vessel and heat the vessel from 20 to 340 °C at a pressure of 7.5 MPa. The pressure was maintained to be stable during heating. For the second step, the open gas cylinder allows the gas (H2 + N2) to flow through the vessel (1 mL/s) at a pressure between 7.5 and 8 MPa by adjusting the flow controller and back-pressure regulator. The gas flowmeter shows how many volumes of gas input to the vessel. Thus, the pressure vessel is full of H2-bearing
RESULTS AND DISCUSSION Cell Potential Measurement. As H2-bearing aqueous fluids pass through the vessel and Au-YSZ/HgO/Hg cell, the potential response will be measured, monitored on the computer screen, and digitally recorded in the computer. The computer monitor system can record and show eight variables, such as potential from Au-YSZ/HgO/Hg, potential from Au-ground, potential from YSZ/HgO/Hg-ground, temperature, pressure of the fluid system, pressure of the gas flow, and extra signals (or we could put more sensors in the system). Figure 4 shows the entire process of testing the H2 sensor of Au-YSZ/HgO/Hg. In the first step of operating the experimental system, aqueous fluid is pumped to flow through pressure vessel, heating the vessel to 340 °C at 7.5 MPa; in the second step, H2 + N2 gas is injected at 7.5-8.0 MPa, which is adjusted to a little higher than the pressure of the liquid pressure in the vessel. The sensor reveals a rapid response to increase of the concentration of dissolved H2 in the effluent aqueous solutions, shown as the A-B part of the potential curve of the Au electrode in Figure 4. Figure 4 indicates the measured potential of the Au-YSZ/HgO/ Hg cell increases with time, as we began to inject gas of H2 + N2. In the third step, injecting of the gas is stopped, we continue to pump aqueous solution (without H2) to the vessel and increase the pressure of the input aqueous solution up to 30 MPa. Thus, the concentration of H2 in the solution becomes less and less, and the Au-YSZ/HgO/Hg cell potential decreases slowly (from ∼-760 to -870 mV), which is the response with decreasing concentration of H2 in aqueous solution, shown as the C-D part of the potential curve of the Au-YSZ/HgO/Hg cell in Figure 4. We first inject H2 + N2 gas to the aqueous solution in the vessel, and then we input aqueous solution without H2 to the vessel, in which changes of H2 concentrations in the solution are also depicted in Figure 5. Figure 5 displays the cell potential response to the changes of the concentration of H2 dissolved in aqueous solution. Also, Figure 5 indicates that the cell potential decreases from -750 to -880 mV with elapsed time, because H2 dissolved in aqueous solution in the vessel decreased. Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
8809
Figure 3. High-temperature and high-pressure hydrogen measurement system for calibrating the Au-YSZ/HgO/Hg sensor. (a) 1, furnace; 2, pressure vessel; 3, electrode; 4, temperature controller; 5, voltammeter; 6, computer and monitor; 7, liquid pump; 8, liquid reservoir; 9, connection; 10, gas cylinder; 11, pressure gauge; 12, back-pressure regulator; 13, fluid outlet for temporary sampling; 14, fluid outlet; 15, flow fluid meter; 16, one-direction valve; 17, valve; 18, pressure gauge; 19, tube for high-pressure fluid sample; 20, adjust valve; A, gas controlling system; B, temporary sampling of high-pressure fluid. (b) High-pressure vessel: 1, Au electrode; 2, YSZ/HgO/Hg electrode; 3, fluid inlet and fluid outlet; 4, hole for thermocouple; 5, upper part of vessel; 6, lower part of vessel.
All experiments demonstrated that the cell potential changes linearly in response to the concentration of H2 dissolved in aqueous fluids at high temperatures and pressures. Calibration. The overall electrochemical cell is described as eq 1; the electrochemical cell reaction occurs at Au electrode as follows: 2H++2 e- S H2(g)
(2)
The other reaction is present at the YSZ/HgO/Hg electrode as HgO + 2H++2e- S Hg + H2O
(3)
The potential ∆E (V)P,T for the Au-YSZ/HgO/Hg cell as a function of the fugacity of H2 dissolved in aqueous solution, fH2 can be described as 8810
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
[
∆E(V) ) E°Hg⁄HgO,T +
( )]
KHγH2 2.303RT log 2F aH2O
+
2.303RT log(mH2)(4) 2F where E°Hg/HgO,T refers to the standard potential of the HgO/Hg reference electrode at the appropriate temperature T, which is determined by reaction 2; R is gas constant; F refers to Faraday constant; KH refers to Henry’s law constant; γH2 stands for activity coefficient for dissolved H2; aH2O is the activity of H2O; mH2 is the molar concentration of H2 dissolved in aqueous solution. In the equation, T is K. The fugacity of H2, fH2 can be determined from measured cell voltage ∆E (V)H2,T,P as follows: ∆E(V)H2,T,P ) ∆Eo-
( )
fH2 2.303RT log 2F aH2O
(5)
Figure 4. Measured voltage response of the Au-YSZ/HgO/Hg cell to the concentration of dissolved H2 in aqueous fluid as demonstrated by flow-through experiment at 340 °C and 30 MPa. The recorded cell potential and cell temperature are automatically recorded and plotted against time. (a) A-B part shows injecting the hydrogen gas into the fluid inside vessel, at 300 °C and at 7.5 MPa, with simultaneous input of aqueous solution into the fluid inside vessel; (b) C-D part shows that decreasing concentration of H2 in fluid as temperature reaches 340 °C and 30 MPa. The different color lines show the potentials between Au-YSZ/HgO/Hg electrodes, Au electrode-ground, and YSZ/HgO/Hg electrode-ground.
where ∆E °cell is the cell potential at standard state, which can be calculated from the standard state potential of E°(YSZ/HgO/Hg) and E°(Au), and fH2 is the fugacity of H2. The sensor potential, ∆E(V)H2,T,P is solely a function of fH2 at a given dissolved H2 concentration and at a fixed T-P condition. Fugacity of H2 fH2 can be described by the Henry’s law constant KH and H2 concentration in the aqueous solution as follows
Figure 5. Measured voltage response of the Au-YSZ/HgO/Hg cell to the concentration of dissolved H2 in aqueous fluid as demonstrated by flow-through experiment at 340 °C and 30 MPa. The recorded cell potential and cell temperature are plotted against time, which shows changes of concentration of dissolved H2 in the fluid, i.e., decreasing concentration of H2 in the fluids with elapsed time as the temperature reaches 340 °C and 30 MPa. The different color lines show the potentials between Au-YSZ/HgO/Hg electrodes, Au electrodeground, and YSZ/HgO/Hg electrode-ground.
varies linearly with H2 concentration in the fluids, which follows the linear equation Y ) -944.4- 60.2X, R2 ) -0.93. The measured slope of the regression line is 60.2(0.06023), which is consistent with the theoretical prediction. Theoretical value of Nernstian slope (2.303RT)/(2F) is 60.8(0.06076) at 340 °C. See eq 4. Based on the experimental results of the potential response of hydrogen sensor of the Au-YSZ/HgO/Hg to the concentrations of H2 in solution at 340 °C and 30 MPa, the cell potential can be described as follows: ∆EAu-YSZ⁄HgO⁄Hg(V) ) 0.9444 + 0.06023 log mH2
mH2 )
f H2 KHγH2
(6)
Contrarily, (∆E(V)H2,T,P and mH2), KH2, and γH2 can be calculated based on the experimental results..24-30 A series of measurement experiments of the Au-YSZ/HgO/Hg sensor were performed in NaCl fluids of dissolved H2 in the temperature range from 20 to 350 °C. The results revealed spectacular cell potential response (Figure 6a). Thus, a plot of the Au-YSZ/HgO/ Hg sensor potential against the measured concentration of hydrogen in fluids at 340 °C is shown in Figure 6a. Figure 6b shows the potential response of the Au-YSZ/HgO/Hg cell to dissolved H2 at 340 °C and 30 MPa. The measured cell potential (24) Naumov, G. B.; Ryzhenko, B. N.; Khodakovsky, I. L. Handbook of Thermodynamic Data; USGS-WRD-74-001, 1974. (25) Haar, L.; Gallagher, J.; Kell, G. NBSINRC Steam Tables; Hemisphere: Bristol, PA, 1984. (26) Johnson, J. W.; Oelkers, E. H.; Helgeson, H. C. Comput. Geosci. 1992, 18, 899–947. (27) Robie, R. A.; Hemingway, B. S.; Fisher, J. R. U. S. Geol. Surv. Bull. 1978, 1452. (28) Kishima, N.; Sakai, H. Geochem. J. 1984, 18, 19–29. (29) Ding, K.; Seyfried, W. E. Jr. Eos: Trans. Am. Geophys. Union 1990, 71, 1680. (30) Shock, E. L.; Helgeson, H. C.; Sverjensky, D. A. Geochim. Cosmochim. Acta 1989, 53, 2157–2183.
(7)
According to eq 4, theoretical prediction of the Au-YSZ/HgO/ Hg cell potential in H2-bearing aqueous solution at high temperatures can be carried out based on the thermodynamic data and previous study.20,24-26,31 Thus, on the standard hydrogen electrode scale, E°HgO/Hg, T at high temperature, and the Henry law’s constant for H2 can be evaluated. E°HgO/Hg, T is ∼0.8464 at 340 °C.32 In fact, Henry law’s constant for H2 at all various conditions would not be offered in the published literature, and particularly a few experiments are concentrated to KH for H2 at high temperatures close to the critical state of water.31,32 Some investigators suggested that the values of Hc at high temperature can be estimated with Ostwald’s expression KH = PH2/ RTCW. According to this method, we calculated Henry law’s constant for H2 at 340 °C, log KH)1.39.33 See Appendix part 1. Considering an aqueous solution is very dilute (NaCl-H2O) and low concentration of hydrogen dissolved in aqueous solution, aH2O and γH2 can be taken as a unit. And assuming that this theoretic prediction of KH can be used in our experiment at 340 (31) Macdonald, D. D.; Hettiarachchi, S.; Lenhart, S. J. J. Solution Chem. 1988, 17, 719–732. (32) Eklund, K.; Lvov, S. N.; Macdonald, D. D. J. Electroanal. Chem. 1997, 437, 99–110. (33) Spear, J. R.; Walker, F. J.; McCollom, T. M.; Pace, N. R. Proc. Natl. Acad. Sci. U. S. A. 2005, 102 (7), 2261–2672.
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
8811
from a subcritical to the critical state of water. Our experiments for calibration Au-YSZ/HgO/Hg cell and measurement of H2 in aqueous solutions provide new approach to determine the Henry constant of H2 in water and a dilute aqueous system at high temperatures and pressures. Equation 4 is reformed as follows:
(
Log
KHγH2 ) EAu-YSZ⁄HgO⁄Hg(V) - E°Hg⁄HgO, T aH2O
[
(4′) )] ⁄ 2.303RT 2F
2.303RT log(mH2 2F
Based on experimental results of solubility of H2 in NaCl-H2O at 340 °C and 30 MPa, and the statistical calculation of 31 groups data (as shown in Figure 6a), the average values of log KH2) 1.633 at 340 °C. Thus, theoretical expression for Nernstian response of the Au-YSZ/HgO/Hg cell to dissolved H2 at 340 °C and 30 MPa (eq 4) can be rewritten as,
Figure 6. Measurement potentials as a function of dissolved H2 concentration in fluids at 340 °C and 30 MPa. (a) The symbol [ represents results from the Au-YSZ/HgO/Hg cell. The dashed line is the theoretical calculation using the data of E°HgO/Hg, T from ref 33 and Henry constant calculated by using the method of ref 35. Experiment measurements indicate that a linear response of the Au-YSZ/HgO/Hg cell potential to the concentration of dissolved H2 in aqueous fluid at 340 °C and 30 MPa. The slope is 944.4 of the regression line (solid, Y ) 944.4 - 60.2X, R2 ) -0.93), which is coincident with theoretical line. (b) Measurement potentials as a function of dissolved H2 concentration in fluids at 340 °C and 30 MPa during 8 days. The symbol 0 represents results from the Au-YSZ/ HgO/Hg cell at 340 °C on the first day; ] refers to that on the fourth day; [stands for that on the fifth day; and 2 refers to that at 350 °C and 30.8 MPa on the eighth day. The Nernstian slop and linear equation can be figured out, based on the data collected at 340 °C, as shown in (a). But at 350 °C, Y ) -996.1-62.6X, R2 ) -0.8, which are closed to the theoretical values. Nernstian slop at 350 °C is 61.8 mV.
°C and 30 MPa.32 Thus, eq 4 can be described as follows: ∆EAu-YSZ⁄HgO⁄Hg(V) ) 0.9306 + 0.06076 log mH2
(8)
As applied, the experimental results about dilute hydrogen in water at elevated temperature and pressure,28,34 for instance, the Henry constant of hydrogen in water, log KH ) 2.1, the eq 4 could be shown as ∆EAu-YSZ⁄HgO⁄Hg(V) ) 0.974 + 0.06076 log mH2
(8′)
The difference between eq 8 and eq 8′ is derived from the difference in Henry constant of hydrogen in water at 340 °C. See Appendix 2. Our experimental measurement of ∆EAu-YSZ/HgO/Hg (V) at 340 °C and 30 MPa, shown as eq 7, apparently agrees with these predictions of ∆E cell. See eq 8 and Appendix part 2. It is important to assess H2 solubility in fluids and hydrogen behavior under extreme conditions, particularly within the region (34) Kishima, N.; Sakai, H. Fugacity-concentration relationship of dilute hydrogen in water at elevated temperature and pressure. ESPL 1984, 67, 79-86.
8812
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
∆EAu-YSZ⁄HgO⁄Hg(V) ) 0.9456 + 0.06076 log mH2
(9)
In addition, this type of Au hydrogen sensor could easily be used for in situ measurement of H2 in hydrothermal fluids in a high-pressure vessel, or at midocean ridge, particularly in longterm monitoring due to its structure of compression resistance. We have performed experiments in the long term of 10 days to test the stability of the H2 sensor. Figure 6b shows that the Aubased sensor was tested at high temperatures in the long-term experiments and displays the results in the first day, fourth, fifth, and eighth day. The experiments demonstrate that the potential response of Au-YSZ/HgO/Hg cell to H2 concentrations in fluids has long-term stability. And experimental data indicate that the measured cell potentials follow Nernstian law predictions for dissolved H2 concentrations in fluids. This paper is not going to discuss the electrochemical mechanism in detail. Many investigators reported that the generally accepted sensing mechanism at the YSZ electrodes, such as the Pt-YSZ H2 sensor and other related sensors, can be expressed as a combination of simultaneous, competing oxidation-reduction reactions: 2e- + 1/2O2 ) O2-(YSZ) and H2+ O2-(YSZ) ) H2O + 2e-. Concerning the Pt(Au)-YSZ H2 sensor, the electrochemical reaction and ion transportation could happen to the hydrogen halfcell, Au (Pt) | 1/2H2(g) | H+(aq) ||, in which hydrogen gas is allowed to bubble over a Au (or Pt) electrode, that is 1/2 H2(g) ) H+(aq) + e-. And simultaneously the reactions occurred on the YSZ electrode, H2O ) 2H+(aq) + O2- (YSZ), and 2e- + 1/2O2 ) O2-(YSZ). See Appendix part 3. CONCLUSION The novel structure of hydrogen sensor is constructed of the Au-YSZ /HgO/Hg cell, which includes the Au electrode combined with the YSZ /HgO/Hg reference electrode. Au electrode is made of Au wire mounted in a quartz bar, which can be pressurized and heated in high-pressure and -temperature conditions. The H2 sensor of Au-YSZ has been tested and calibrated at high temperatures up to 400 °C and pressures to 38 MPa. The experimental measurements of the Au sensor potential against the YSZ/HgO/Hg reference electrode indicate that the
sensor potential varies linearly with the concentration of dissolved H2 in NaCl-H2O at 340 °C and 30 MPa. Experiments demonstrated that Nernstian response of the cell potential to dissolved hydrogen H2 was determined at 340 °C and 30 MPa. ∆E ) 0.9444 + 0. 0603 log mH2. These experiments and calculations indicate that the novel structure the Au-YSZ/HgO/Hg cell can be used to in situ measure the solubility of H2 in aqueous fluids at temperatures and pressures near the critical state of water. In sum, these experiments and theoretical calculations indicate that this kind of Au electrode may be easily used for in situ measurement of the concentration of dissolved H2 in hydrothermal fluids at high pressures, in the laboratory, reaction engineering, and geothermal systems, particularly in long-term monitoring. The experimental equipment system is effective for testing and calibrating a high-temperature and high-pressure chemical sensor. And experiments also provide a new method to assess the Henry constant of hydrogen in water and aqueous solutions at high temperatures and pressures. APPENDIX 1. Some investigators suggested that the values of Hc at high temperature can be estimated with Ostwald’s expression, KH = PH2/RTCW, who used these data: R is the gas constant (0.0821 L atm mol-1 K-1), T is the temperature (K), and CW is the concentration of the gas in the water, and it is assumed CW is 1 × 10-8 mol · L-1. PH2 is 1.22 × 10-5 atm (1 atm ) 101.3 kPa)] 32 2. The experiments on hydrogen dissolved in water at elevated temperatures and pressure28,34 were carried out by using Dicksontype hydrothermal apparatus. Hydrogen concentrations were measured in these experimental systems containing magnetitehematite, Ni-NiO, and fayalite-magnetite-quartz oxygen buffers. The solutions were put in the gold bag inside of the pressure vessel, when operating experiments. The problem is that gold metal can adsorb hydrogen gas, and the measured concentration of hydrogen could be a little low. Therefore, log KH values were possible higher than the actual values. Eklund et al.32 reported on the Hg|HgO|ZrO2(Y2O3)| NaOH(aq)|H2(Pt) cell at temperatures ranging from 298.15 (25 °C) to 723.15 K (450 °C) and at a pressure of 275 bar, to assess the viability of ceramic sensors for measuring Henry’s constant for hydrogen in high subcritical (473.15 K < T < 647.30 K) and supercritical (T > 647.30 K) aqueous systems. The Hg|HgO|ZrO2(Y2O3)|H+,H2O ceramic membrane electrode, when combined with a platinum electrode, provides a convenient means of evaluating hydrogen activity and Henry’s constant for hydrogen in high subcritical and supercritical aqueous systems. The calculated Henry’s constants for dissolved molecular hydrogen in pure water were compared with the available literature data28,34 and our data, and they
are basically in agreement, but KH at 350 °C is higher than the others, log KH ) 2.388. Why are they different? Maybe, it is due to our experiments being carried out in a NaCl-H2O system. 3. Mechanism of transportation of ion through the YSZ membrane in Pt(Au)-YSZ H2 sensor and other metal-YSZ cells were reported.31,35-37 Many investigators considered that as membrane of an oxygen ion-conducting electrolyte, yttria-stabilized zirconia, oxygen, which is fed to one side of the membrane, is reduced by the cathode to oxygen ions via the overall half-cell reaction 2e- + 1 ⁄ 2O2 ) O2-(YSZ)
(A1)
Oxygen ions thus created migrate selectively through the membrane to the anode, where they undergo a similar half-cell reaction with a H2 to produce H2O as follows: H2(g) + O2-(YSZ) ) H2O + 2e-
(A2)
In addition, for the Pt(Au)-YSZ H2 sensor, the electrochemical reaction, and ion transportation happening in the hydrogen halfcell, Au | 1/2H2(g) | H+(aq) ||, in which hydrogen gas is allowed to bubble over a Au (or Pt) electrode having a specially treated surface, which catalyzed the reaction 1 ⁄ 2H2(g) ) H+(aq) + e-
(A3)
And the potential determining process on both sides of the membrane was attributed to the equilibrium18,38 H2O ) 2H+(aq) + O2-(YSZ)
(A4)
and also combined with reaction (A1). ACKNOWLEDGMENT This study was supported by the Ministry of Science and Technology of China, Sea 863 project (2001AA612020-3, 2003AA612020-3), Basic Research Project (2001DEA30041, 2002DEA30084, 2003DEA 2C021), DY105-03-01, and NSFC (20373064, 50602042). Received for review May 8, 2008. Accepted September 6, 2008. AC800948X (35) Adler, S. B. Chem. Rev. 2004, 104, 4791–4843. (36) Martin, L. P.; Pham, A.-Q.; Glass, R. S. Solid State Ionics 2004, 175, 527– 530. (37) Bessler, W. G.; Warnatz, J.; Goodwin, D. G. Solid State Ionics 2007, 177, 3371–3383. (38) Lvov, S. N.; Zhou, X. Y.; Ulmer, G. C.; Barnes, H. L.; Macdonald, D. D.; Ulyanov, S. M.; Benning, L. G.; Grandstaff, D. E.; Manna, M.; Vicenzi, E. Chem. Geol. 2003, 198, 141–162.
Analytical Chemistry, Vol. 80, No. 22, November 15, 2008
8813
