Lithiumferratebased Cathodes For Molten Carbonate Fuel Cells

M. T. Lanagan, I. Bloom,t T. D. Kaun.t J. Wolfenstine,1 and M. Krumpeltt

Energy Technology Division, Chemical Technology Divisiont Argonne National Laboratory Argonne, IL 60439-4838


Argonne National Laboratory is developing advanced cathodes for pressurized operation of the molten carbonate fuel cell (MCFC) at ~650°C [2-5]. To be economically viable for stationary power generation, molten carbonate fuel cells must have lifetimes of more than 25,000 h while exhibiting superior cell performance [6]. In the present technology, Iithiated NiO is used as the cathode. Over the lifetime of the cell, however, Ni + ions tend to transport to the anode, where they are reduced to metallic Ni [7]. With increased C02 partial pressure, the transport of Ni increases because of the increased solubility of NiO in the carbonate electrolyte. Although this process is slow in MCFCs operated at 1 atm and a low C02 partial pressure (about 0.1 atm), transport of nickel to the anode may be excessive at a higher pressure (e.g., 3 atm) and a high C02 partial pressure (e.g., about 0.3 atm). This transport is expected to lead eventually to poor MCFC performance and/or short circuiting.

Several alternative cathode compositions have been explored to reduce cathode solubility in the molten salt electrolyte [2-5,8]. For example, LiCo02 has been studied extensively [8,9,10] as a potential cathode material. The LiCo02 cathode has a low resistivity, about 1 Ci-cm, and can be used as a direct substitute for NiO. However, the high material cost may prevent large-scale implementation.

Argonne is developing advanced cathodes based on lithium ferrate (LiFeOj), which is attractive because of its very low solubility in the molten (Li.K^COs electrolyte [11], Because of its high resistivity (about 300 £2-cm), however, LiFe02 cannot be used as a direct substitute for NiO. Cation substitution is, therefore, necessary to decrease resistivity.

We determined the effect of cation substitution on the resistivity and deformation of LiFe02. The substituents were chosen because their respective oxides as well as LiFeC^ crystallize with the rock-salt structure.


Materials Synthesis and Characterization. Stoichiometric amounts of Li2C03, Fe203, CoO, NiO and/or MgO were mixed and heat treated at 825°C for 4 h. The powders were pressed into disks and sintered at 1000°C for 4 h. Upon cooling, the phase distribution of the sintered disks was characterized by X-ray diffraction. The relative densities of the pellets were in the range of 60-90%; LiCo02 had the highest (93% of theoretical) and pure LiFe02, the lowest (81%). All materials except LiCo02 had equiaxed grains (0.96-1.43 ¿um). LiCo02 had acicular grain morphology (5 x 1.5 ¡xm).

Resistivity and Deformation Rate Measurements. Resistivity was determined using the four-wire, dc, van der Pauw method at 650°C in air. Parallelepipeds (2x2x4 mm) were cut from the sintered pellets. The samples were deformed by compression at a constant deformation rate in air at 1000°C.

Cell Tests. Promising materials were tested in a 5 x 5-cm cell with a pressed (Li,K)2C03 electrolyte tile. The test cell was operated at 650 °C using an 02-C02 gas mixture which simulated pressurized operation at 3-5 atm.

Results and Discussion

Materials. We found that the range of homogeneity for Co-substituted LiFeOj is limited; several phases are formed during synthesis. All of the diffraction peaks for Mg-substituted LiFe02 were indexed to a rock-salt structure. Further analysis shows that the lattice parameter increases monotonically with MgO concentration, indicating a complete solid solution (Fig. 1). Fayard found that both CoO and MgO formed complete solid solutions with LiFe02 [12]. Apparently, Co solubility is affected by differences in synthesis conditions.

Resistivity Measurements. Resistivity results for MgO-substituted LiFe02 samples show that the minimum resistivity is approximately 100 ii-cm (at 650°C in air) and appears in the MgO concentration range of 6-12 mol% (Fig. 2). Clearly, the resistivity of Mg-substituted LiFeO^ is too high; lower resistivity values (-1 £2-cm) are necessary for MCFC cathodes. By using different cation substituents, we lowered the resistivity of a LiFeOrbased material (designated Material 1) to 1.3 £i-cm at 650°C.

Deformation Rate Studies. Figure 3 shows a log-log plot of the deformation rate vs stress at 1000°C for five materials: (a) pure LiFe02, (b) 12.5 m/o Ni0-LiFe02, (c) 9 m/o Mg0-LiFe02, (d) pure LiCo02, and (e) 5 m/o Li20-Ni0. Here, the deformation rate was normalized with respect to grain size and stress to the material density, and it was assumed that the deformation rate is controlled by lattice diffusion. The slope of all curves is close to unity, implying that they all exhibit the same deformation mechanism (e.g., diffusional creep and/or grain boundary sliding). Materials (a) and (d) have higher deformation rates than (e), lithiated NiO; Material (d) has the highest deformation rate. Doping LiFe02 with NiO or MgO decreases the deformation rate as compared to that of pure LiFe02. Materials (a), (b), (c), and (d) will compact more than lithiated NiO at 650°C.

Cell Tests. A fibrous cathode of Material 1 was used in a cell test. The cell potential at 160 mA/cm2 as a function of time is shown in Fig. 4. After an initial break-in period, a constant potential of 850 mV was observed for 1700 h. Polarization experiments showed a small degradation at 1706 h (see Fig. 5). These results show that LiFe02-based cathodes have good performance and durability at high C02 partial pressures.


Several LiFe02-based materials with cation substitutions were synthesized and characterized. With the proper choice of substituents, the resistivity of LiFe02 was lowered to about 1 fl-cm.

The LiCo02- and LiFe02-based cathode materials have deformation rates higher than that of lithiated NiO at 1000°C. They will compact more at 650°C. Doping the alternative cathodes decreases their deformation rates.

The LiFe02-based cathodes exhibit cell potentials which are stable for long periods of . time under simulated pressurized operation. LiFeCVbased materials continue to show promise as cathodes for the MCFC.


The authors thank H. Wilson Abernathy III, Bryant Polzin, and Ryan Thayer for their assistance in sample preparation. This research was supported under a contract with M-C Power



1. Faculty Research Participant. Present address: Dept. of Chemical-Biochemical Engineering, University of California, Irvine, CA.

2. A. P. Brown, G. H. Kucera, M. F. Roche, D. D. Chu, and E. Indacochea, "Performance of Alternative Cathodes for Molten Carbonate Fuel Cells," Fuel Cell Seminar Program and Abstracts, November 29-December2,1992, Tucson, AZ, p. 125.

3. G. Kucera, A. Brown, D. Chu, and E. Indacochea, "Development of Alternative Components for Molten Carbonate Fuel Cells," presented at EPRI/GRI Fuel Cell Workshop on Fuel Cell Technology Research and Development, April 29-30,1992, San Francisco, CA.

4. J. L. Smith, G. H. Kucera, and A. P. Brown, "Development of Cathode Materials and Structures for the Molten Carbonate Fuel Cell," Proc. 2"d Symp. on Molten Carbonate Fuel Technology, J. Selman, D. A. Shores, H. C. Maru and I. Uchida, eds., Vol. 90-16 (Oct. 1990), The Electrochemical Soc., p. 226-ff, and references cited therein.

5. G. H. Kucera and J. L. Smith, "Electronically Conductive Ceramics for High Temperature Oxidizing Environments," US Pat. No. 4,564,567 (January 1986).

6. C. Yuh, M. Faooque, and H. Maru, "Selection of Materials for Carbonate Fuel Cell Use," Electrochemical Society Extended Abstracts, Vol. 94-2 (Oct. 1994).

7. N.Q. Minh, 'Technological Status of Nickel Oxide Cathodes in Molten Carbonate Fuel Cells," J. Power Sources, Vol. 24, pp. 1-19 (1988).

8. L. Plomp, E. F. Sitters, J. B. J. Veldhuis, R. C. Makkus, and S. B. van der Molen, "Performance and Dissolution-Precipitation Behavior of Molten Carbonate Fuel Cell Cathodes," Proc. Electrochem. Soc., Vol. 93-3, pp. 171-185 (1993) and L. Plomp, R. C. Makkus, E. F. Sitters, and G. Rietveld, "Endurance Issues and Materials Development in Molten Carbonate Fuel Cells," Fuel Cell Seminar Program and Abstracts, November 28-December 1,1994, San Diego, CA, p. 164.

9. B. Bergman, E. Fontes, C. Lagergren, G. Lindbergh, A. Lundblad, S. Schwartz, D. Simonsson, and

C. Sylwan. "Swedish Research on MCFC," Fuel Cell Seminar Program and Abstracts, November 29-December 2,1992, Tucson, AZ, p. 73.

10. M. Carewska, S. Scaccia, G. Qhidelli, L. Giorgi and E. Simonetti, "Synthesis and Electronic Conductivity of LiCo02 Materials for MCFC Alternative Cathodes," Fuel Cell Seminar Program and Abstracts, November 28-December 1,1994, San Diego, CA, p. 176 and references cited therein.

11. D. Shores and Y. Qu, "Dissolution of Oxides in Molten Carbonates," Proc. 3rd International Symp. on Carbonate Fuel Cell Technology, Electrochem. Soc., Vol. 93-3, pp. 356-367 (1993).

12. M. Fayard, "Sur les Phénomènes Ordre-Désordre dans l'Oxyde Mixte LiFe02 et sur les Propriétés de ses Solutions Solides avec Quelques Oxydes du Type NaCl," Annales de Chimie, pp. 1313-1335 (1961).

0 0

Post a comment