OTHER TYPES OF FUEL CELLS
9.4 DIRECT CARBON FUEL CELLS
When toward the end of the nineteenth century Wilhelm Ostwald formulated his idea of using an electrochemical mechanism for direct conversion of the chemical energy of natural fuels to electrical energy, coal was the chief fuel in human hands. Even today the widespread use of petroleum products and the development of nuclear power notwithstanding, coal remains a very important component of world energy supply. Its share of all known natural fuel reserves worldwide is about 60%. In China today about 80% of the electrical energy is produced by coal-fired power stations, these being responsible for 70% of CO2
emissions and 90% of SO2emissions in that country (Cao et al., 2007).
Electrochemical utilization of coal’s energy would provide huge gains not only and not so much because of higher conversion efficiencies but also for other reasons. The thermal power stations emit CO2mixed with air and other gases as well as with uncombusted coal particles into the atmosphere. In a fuel cell, coal would be oxidized anodically in separate compartments closed off from the air.
From the gases evolved in these compartments, by-product hydrocarbons could readily be separated and then utilized, while particulates could be filtered off. The remainder, almost pure CO2, could be sent to underground storage. This would make a huge contribution to solving the problem of global warming caused by CO2emissions. The power plants with direct carbon fuel cells (DCFCs) could in principle be set up directly in coal mines to eliminate the numerous economic and ecological problems associated with long-distance coal transport.
There are two possibilities for electrochemical utilization of the chemical energy of coal: (1) via prior coal gasification and use of the hydrogen and/or carbon monoxide produced in this process in various fuel cells, and (2) by direct electrochemical oxidation within the fuel cell. Various methods of coal gasification are discussed in Chapter 11. Some attempts to realize the second approach, which is basically a much simpler one-step process, are discussed in the present section. [Note thatcoalandcarbonare somewhat interchangeable terms when discussing them as fuels, coal being a natural carbon material, carbon being a derived material (not necessarily from coal).]
Antoine Ce´sar Becquerel in France in 1855 and Pavel Yablochkov in Russia in 1877 built electrochemical devices using coal anodes in a molten KNO3
electrolyte (see Howard, 1945, and Liebhafsky and Cairns, 1968). In 1896, William Jacques obtained a U.S. patent for his invention of a ‘‘coal battery’’
with a coal anode and an iron cathode immersed in molten alkali NaOH (this battery was mentioned in Chapter 2). Despite the great doubts raised as to the nature of the processes taking place in it, the electrical performance of his 100- cell battery operating at 400 to 5001C was rather impressive: total power of 1.5 kW and current densities up to 100 mA/cm2. Much later, the prototype of a carbon fuel cell with a solid electrolyte working at a temperature of 10001C was built by Baur and Preis in Switzerland, 1938).
170 OTHER TYPES OF FUEL CELLS
9.4.2 Reactions and Thermodynamic Parameters
The reactions ideally occurring in DCFCs with an acidic electrolyte solution and their thermodynamic parameters at 251C are as follows:
Anode: Cỵ2H2O!CO2ỵ4Hỵỵ4e E0ẳ0:21 V ð9:14ị Cathode: O2ỵ4Hỵỵ4e!2H2O E0 ẳ1:23 V ð9:15ị Overall: CỵO2!CO2 E0ẳ1:02 V ð9:16ị DG0ẳ394:3 kJ=molẳ1:02 eV DH0 ẳ393:5 kJ=molẳ1:02 eV It will be seen below that these reactions actually take place not in aqueous solutions but at high temperatures and in other electrolytes (e.g., in molten carbonates or alkalies). Thus, in molten carbonates at 6001C, the equations can be formulated as follows:
Anode: Cỵ2CO23 !3CO2ỵ4e E0ẳ 0:02V ð9:17ị Cathode: O2ỵ2CO2ỵ4e!2CO23 E0ẳ1:10V ð9:18ị Equation (9.16) for the overall reaction remains unchanged:
DG0ẳ395:4 kJ=molẳ1:02 eV DH0ẳ394:0 kJ=molẳ1:02 eV It can be seen that the thermodynamic parameters of this reaction are practically independent of temperature. It is a remarkable feature of the reaction that its Gibbs free energy and its enthalpy have practically identical values, which implies that entropic losses are absent. This means that a 100%
conversion of the chemical energy of carbon to electrical energy is theoretically possible.
At temperatures above 7501C, the Boudouard equilibrium can become established:
CỵCO2é2CO ð9:19ị
which during carbon oxidation would lead to the formation of carbon monoxide (CO), entailing some energy loss (only two electrons are liberated instead of four when a carbon atom is merely oxidized to CO).
* These values of the electrode potentials are referred to the equilibrium potential of the hydrogen electrode under the same conditions (carbonate melt at 6001C).
9.4 DIRECT CARBON FUEL CELLS 171
9.4.3 Work on Carbon Fuel Cells Since 1960
Despite the numerous problems encountered in earlier research, attempts to build versions of carbon fuel cells were continued when the fuel cell ‘‘boom’’
began in the 1960s.
Two characteristics are of importence when carbon (or coal) is used in fuel cells: the nature of the carbon material to be used and the physical state of this material. The notion of ‘‘carbon’’ is highly indefinite; it covers a large number of carbon materials, both natural and human-made. It includes various types of graphite, coke, and carbon black, differing very strongly in their structure and in the content of additional components, both volatile (hydrogen, oxygen, nitrogen, sulfur, organic compounds, etc.) and nonvolatile (mineral salts).
Carbon materials of natural origin could be used as such, or after having been subjected to pretreatments to eliminate undesired components.
The physical state of a carbon material in a fuel cell can be of two kinds: (1) relatively massive electrodes in the shape of plates or rods which may be cut out of graphite or pressed from powdered carbon materials (usually with different binders) and may serve as the current collector and as a consumable electrode;
or (2) highly disperse carbon powders, present as a slurry in a liquid electrolyte such as carbonate melt, which are in constant contact with a metal electrode serving as the current collector when they take part in the electrochemical reactions and electrons must be transferred.
In the work on DCFCs, different directions were followed with respect to the electrolytes: aqueous solutions (at temperatures below 1001C), high-tempera- ture melts of carbonates or sodium hydroxide, and high-temperature solid electrolytes.
Aqueous Solution
In 1979 Coughlin and Farooque advanced the idea of electrochemical gasifica- tion of coal by electrolysis of a suspension of carbon materials in a solution of sulfuric acid with platinum electrodes. Here, the cathodic reaction of hydrogen evolution,
2Hỵỵ2e!H2 E0ẳ0 V ð9:20ị occurs together with the anodic reaction (9.14), so that the overall reaction;
Cỵ2H2O!CO2ỵ2H2 E0ẳ0:21V ð9:21ị leads to electrolytic hydrogen evolution at a voltage of 0.21 V (instead of the 1.23 V needed in ordinary water electrolyzers). The device suggested is not a fuel cell but makes use of anodic carbon oxidation in aqueous solution. Details about this reaction were not communicated by the authors. It was shown later by Okada et al. (1981) and Dhooge et al. (1982) that at a temperature of 801C, 172 OTHER TYPES OF FUEL CELLS
such a reaction is actually possible. It proceeds since carbon samples contain iron as an impurity dissolving in the acid, thus producing the Fe3+/Fe2+redox system in the solution. This served as a mediator for coal oxidation. When iron was removed, the rate of carbon dropped to values too low for the practical generation of electrical energy in a fuel cell.
In 2006, Patil et al. continued research into this reaction, using carbon suspensions in a sulfuric acid solution to which 100 mM each of Fe3+ and Fe2+ions had been added on purpose. It could be shown that with a voltage of 0.6 to 1.0 V applied across a cell with platinum or platinum-alloy electrodes, even at 401C the reaction occurs with an acceptable current density of 30 mA/cm2. This is a rather promising result, since in ordinary electrolyzers a voltage of at least 1.7 V is needed for hydrogen evolution. The authors pointed out, though, that a number of aspects requiring further study are not clear with respect to the reaction.
Low-temperature oxidation (below 2001C) of carbon has received attention from research workers not only because of its potential utility in DCFCs but also because of the harm it does in many types of fuel cells and other electrochemical devices where platinum catalysts on carbon supports are used. It was in this connection that Choo et al. (2007) studied the anodic oxidation mechanism of graphite in sulfuric acid.
Melts
Studies of electrochemical carbon oxidation in carbonate melts at 7001C were performed by Weaver et al. (1981) at the Stanford Research Institute (SRI) in Menlo Park, California. They used rods of different carbon materials as the electrodes. The electrode potentials were measured relative to a gold reference electrode in an atmosphere of CO2+O2at the same temperature. The electrodes proved to be more active the lower the degree of crystallinity of the original carbon powder used to press the rods. The electrodes had open-circuit potentials around 1.1 V. At a current density of 100 mA/cm2, the potential of the most active sample was 0.8 V (and as high as 0.9 V when the temperature was raised to 9001C).
At the Lawrence Livermoore National Laboratory, a fuel cell using a semisolid suspension of carbon powder in a molten carbonate was developed.
Porous nickel was used as the cathode material. At a temperature of 8001C and a voltage of 0.8 V, current densities of 50 to 125 mA/cm2were obtained. In one test, a current density of 27 mA/cm2was drawn for 30 hours (Cherepy et al., 2005).
Hackett et al. (2007) reported experiments with DCFCs where rods prepared from different carbon materials were used as the anodes. The cathodes were made from iron–titanium alloy. Melts of NaOH at temperatures in the range 600 to 7001C were used as the electrolyte. The most stable operation with high performance figures was obtained with graphite anodes. The OCV value was 0.788 V. At the optimum temperature of 6751C, the voltage was 0.45 V at a current density of 140 mA/cm2. Anodes made of other carbon materials had higher OCVs (up to 1.044 V), but exhibited inferior and less stable performance
9.4 DIRECT CARBON FUEL CELLS 173
when current was drawn. The equation for the anodic reaction in NaOH melt can be written as
Cỵ4OH!CO2ỵ2H2Oỵ4e ð9:22ị An undesirable side reaction is carbonation of the alkali by the CO2evolved, which involves the equilibrium
CO2ỵ2OHéCO23ỵH2O ð9:23ị When steam is added to the oxygen flow, the equilibrium (9.22) can be shifted somewhat to the left.
Solid-Oxide Electrolytes
A difficulty of principle arises when using a solid electrolyte in DCFCs. In fact, in cells with a liquid electrolyte (solution or melt), the entire surface area of the carbon material is in contact with the electrolyte (is wetted by the electrolyte).
In cells with a solid electrolyte, to the contrary, the contact between the solid carbon material and the solid electrolyte is a mere point contact, and the working surface area is much smaller.
Two ways to overcome this difficulty have been suggested. The first (Pointon et al., 2006) makes combined use of a solid-oxide electrolyte and liquid (molten) carbonate electrolyte. The oxygen electrode was separated by the solid electro- lyte from the carbonate melt that contained the suspended carbon particles.
The second (Gu¨r and Huggins, 1992) used a cell consisting of two halves separated by a solid electrolyte in the shape of a tube. The inside and outside surface of the tube were coated with a layer of platinum. The inside of the tube was in contact with ambient air. The outside was in contact with a closed compartment holding air and the carbon samples. The solid electrolyte was maintained at a temperature of 9321C, while the compartment holding the carbon was maintained (in one experiment) at a level of 9551C. Because of the reaction between oxygen and carbon, a reduced oxygen equilibrium pressure was set up in the compartment. This led to a considerable oxygen pressure gradient between the two sides of the solid electrolyte, and hence to a potential difference of about 1.05 V. When a current of 10 mA/cm2 was drawn, this difference (the discharge voltage) dropped to a value of about 0.4 V.
REFERENCES Section 9.1
Codina G., J. R. Perez, M. Lopez-Atalaya, et al.,J. Power Sources,48, 293 (1994).
Joerissen L., J. Garche, Ch. Fabian, G. Yomazic,J. Power Sources,127, 98 (2004).
Johnson D. A., M. A. Reid,J. Electrochem. Soc.,132, 1058 (1985).
Lopez-Atalaya M., G. Codina, J. R. Perez, et al.,J. Power Sources,39, 147 (1992).
174 OTHER TYPES OF FUEL CELLS
Murthy A. S. N., T. Srivastava,J. Power Sources,27, 119 (1989).
Skyllas-Kazacos M., D. Kasherman, et al.,J. Power Sources,35, 399 (1991).
Wen Y. H., H. M. Zhang, et al.,Electrochim. Acta,51, 3775 (2006).
Section 9.2
Betso S. R., M. H. Klapper, L. B. Andreasson, in: Biological Aspects of Electrochem- istry, Birkha¨user, Basel and Stuttgart, Germany, 1971, p. 162.
Gao F., Y. Yan, L. Sub, et al.,Electrochem. Commun.,9, 989 (2007).
Prasad, D., T. K. Sivaram, S. Berchmans, V. Yegnaraman,J. Power Sources,160, 991 (2006).
Tarasevich M. R., Yu. G. Chirkov, V. A. Bogdanovskaya, A. V. Kapustin,Electrochim.
Acta,51, 418 (2005).
Topcagic S., S. D. Minteer,Electrochim. Acta,51, 2168 (2006).
Walker A. L., C. W. Walker,J. Power Sources,160, 123 (2006).
Wingard L. W., C. C. Liu, N. L. Nagda,Biotechnol. Bioeng.,13, 629 (1971).
Yaropolov A. I., S. D. Varfolomeev, I. V. Berezin, et al., FEBS Lett.,71, 306 (1976).
Zhang E., W. Xu, G. Diao, C. Shuang,J. Power Sources,161, 820 (2006).
Section 9.4
Baur E., H. Preis,Z. Elektrochem.,43, 727 (1937);44, 695 (1938).
Cao D., Y. Sun, G. Wang,J. Power Sources,167, 250 (2007).
Cherepy N. J., R. Krueger, K. J. Fiet, et al.,J. Electrochem. Soc.,152, A80 (2005).
Choo H.-S., T. Kinumoto, S.-K. Jeong, et al., J. Electrochem. Soc., 154, B1017 (2007).
Coughlin R.W., M. Farooque,Nature,279, 301 (1979).
Dhooge P. M., D. L. Stillwell, S. M. Park,J. Electrochem. Soc.,129, 1719 (1982);130, 1029 (1983).
Gu¨r N. M., R. A. Huggins,J. Electrochem. Soc.,139, L95 (1992).
Hackett G. A., J. W. Zondlo, R. Svensson,J. Power Sources,168, 111 (2007).
Howard H. C.,Direct Generation of Electricity from Coal and Gas(Fuel Cells), Wiley, New York, 1945.
Jacques W. W., U.S. patent 555,511 (1896);Z. Elektrochem.,4, 286 (1910).
Liebhafsky H. A., E. J. Cairns,Fuel Cells and Batteries, Wiley, New York, 1968.
Okada G., V. Guruswamy, J. O’M. Bockris,J. Electrochem. Soc.,128, 2097 (1981).
Patil P., Y. D. Abreu, G. G. Botte,J. Power Sources,158, 368 (2006).
Pointon K., J. Irvine, J. Bradley, S. Jain,J. Power Sources,162, 750 (2006).
Weaver R. D., S. C. Leach, L. Nanis, Proc. 16th Intersociety Energy Conversion Engineering Conference, ASME, New York, 1981, p. 717.
Reviews and Monographs
Bagotsky V. S., A. M. Skundin, Chemical Power Sources, Academic Press, London, 1980.
REFERENCES 175
Cao D., Y. Sun, G. Wang, Direct carbon fuel cells: Fundamentals and recent developments,J. Power Sources,167, 250 (2007).
Ponce de Leo´n C., A. Frı´as-Ferrer, J. Gonza´lez-Garcı´a, D. A. Sza´nto, F. C. Walsh, Redox flow cells for energy conversion,J. Power Sources,160, 716 (2006).
Vincent C. A., B. Scrosati,Modern Batteries: An Introduction to Electrochemical Power Sources, Edward Arnold, London, 1997.
176 OTHER TYPES OF FUEL CELLS
CHAPTER 10