- Email: [email protected]
Silver–praseodymium oxy-sulfate cermet: A new composite cathode for intermediate temperature solid oxide fuel cells
Journal of Power Sources 306 (2016) 611e616
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Silverepraseodymium oxy-sulfate cermet: A new composite cathode for intermediate temperature solid oxide fuel cells Tao Yang*, Aliaksandr L. Shaula, Sergey M. Mikhalev, Devaraj Ramasamy, Duncan P. Fagg TEMA, Mechanical Engineering Department, University of Aveiro, 3810-193 Aveiro, Portugal
h i g h l i g h t s A novel AgePr2O2SO4 composite cathode was developed. High oxygen storage capacity facilitated gaseous oxygen to be adsorbed. A highly active property towards oxygen reduction reactions was achieved. The highest power density reached 1.5 W/cm2 at 800 C. High stability of the cell at 800 C under load was maintained for 24 days.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 September 2015 Received in revised form 17 December 2015 Accepted 18 December 2015 Available online xxx
AgePr2O2SO4 is identified as a promising new composite material to enhance the cathodic oxygen reduction reaction in solid oxide fuel cells. AgePr2O2SO4 was studied in terms of synthesis, stability of Pr2O2SO4 in CO2, and electrochemical behavior as a cathode. The performance of the composite cathode was assessed as a function of temperature by A.C. impedance using a symmetrical cell arrangement in oxygen. The global performance of an anode-supported fuel cell AgePr2O2SO4/CGO/NiO-CGO was also assessed, the highest power density being 1.5 W cm2 at 800 C. A longevity test of this cell performed at 800 C under load for 24 days demonstrated high stability with AgePr2O2SO4 cathode. © 2015 Elsevier B.V. All rights reserved.
Keywords: Ag-Pr2O2SO4 Novel cathode Solid oxide fuel cells Oxygen reduction reaction High power density
1. Introduction Solid oxide fuel cells (SOFCs) are devices of commercial and environmental importance that effectively convert the chemical energy of fuels directly to electricity. They combine advantages of environment-friendly electricity generation with fuel and oxidant flexibility, opening up an extensive range of potential applications, such as stationary power for isolated areas, public transportation power, and even portable power. Nonetheless, the marketability of these applications depends on the achievement of stable fuel cell operation at minimal cost. In order to attain this requirement a reduction in the SOFC working temperature, to at least 600e800 C, has been suggested as desirable [1e5]. The benefits of lowering the operating temperature include fast start-up of the system, an
* Corresponding author. E-mail address: [email protected] (T. Yang). http://dx.doi.org/10.1016/j.jpowsour.2015.12.098 0378-7753/© 2015 Elsevier B.V. All rights reserved.
increase in materials stability and a reduction in cost. However, with the decrease in the working temperature, the polarization resistances of the electrode and of the electrode/electrolyte interface can increase dramatically [4]. This is because the catalytic oxidation/reduction reactions on the surface of the electrode and the ionic transport within the electrode slow down and take major responsibility for impaired power density of the whole fuel cell [4]. In order to deal with this challenge, a variety of perovskite-type ABO3 and Aurivillius-type phases have been developed and improved as potential intermediate temperature SOFC cathodes [2,6]. With respect to the typical ABO3 perovskite structure, strontium-doped lanthanum manganite (La1-xSrxMnO3-d, LSM) mixed with yttria-stabilized zirconia (ZrO2)0.92(Y2O3)0.08, a classical ionic conductor, has traditionally been the composite cathode of choice. Here (ZrO2)0.92(Y2O3)0.08 provides oxygen ionic conductivity while the LSM phase predominantly works as an electronic conductor [7]. Single-phase mixed ionic and electronic conductors
612
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
(MIECs) such as La0.6Sr0.4Co0.8Fe0.2O3-d (LSCF) and Sm0.5Sr0.5Co0.5O3-d (SSC) have also been suggested to reduce the cathodic interfacial polarization resistance at lower temperatures [8,9]. Shao and Haile [10] replaced the rare-earth elements (Ln) with alkaline-earth (Ba) and presented Ba0.5Sr0.5Co0.8Fe0.2O3d (BSCF) as a new cathode material for reduced-temperature SOFC operation. The power densities of cells using this cathode reached values as high as 1010 mW cm2 at 600 C when operated with humidified hydrogen as the fuel and air as the oxidant [10]. For the Aurivillius-type structures, it was found that the Bi2Cu0.1V0.9O5.35 (known as BICUVOX10) showed very good electrochemical performance at 600 C [11e14]. However, its disadvantage of low electronic transference number requires the addition of an electronically superior conducting phase to form a composite. For example, Liu and Xia [6,15] mixed silver powder with BICUVOX10, to provide the required electronic conductivity offering a peak power density of 443 mW cm2 at 550 C. Tao et al. prepared a coreeshell Au-BICUVOX10 cathode fabricated on top of a finger-like CGO electrolyte to increase the triple phase boundary (TPB) length, and achieved a power density of 470 mW cm2 at 550 C by this modification [16,17]. Such previous work has, therefore, shown that a suitable cathode for an intermediate temperature fuel cell should offer sufficient electronic and ionic conductivities, either as a single-phase MIEC or as a composite of two phases to provide the MIEC nature. In an unconventional departure from previous materials for this goal, the current study offers a completely new composite cathode for SOFCs based on combination of silver particles and porous praseodymium oxy-sulfate, Pr2O2SO4. Such oxy-sulfate materials have traditionally been suggested as oxygen storage materials due to their large oxygen storage capacity [18]. Although there are various mechanisms suggested to be involved in the oxygen storage in Pr2O2SO4, for example that the distortion of the oxygen sub-lattice generates some mobile oxide-ions and that the change in the oxidation states of Pr3þ/Pr4þ can catalyze oxygen exchange, it is the redox ability of sulfur that is mainly attributed for the high levels of oxygen storage capacity [19]. Due to the ability of the Ln oxy-sulfates to store oxygen [20e22], they have been applied as catalysts for the high-temperature wateregas shift reaction [23], and for anaerobic catalytic CO oxidation [24]. In the present work, a novel composite cathode of AgePr2O2SO4 is prepared and carefully characterized in terms of stability, microstructure and electrochemical behavior. In this composite it is envisaged that silver will function as the electronic component, while the Pr2O2SO4 phase may provide mobile oxygen species and fast oxygen exchange [25]The performance of an anode-supported fuel cell of composition AgePr2O2SO4/CGO/NiOCGO was assessed, including longevity testing of an optimized cell operated at 800 C for a period of 24 days under load. 2. Experimental
counting time of 2 s per step. Thermogravimetric (TG) analysis and differential scanning calorimetry (DSC) were carried out using the Netzsch Jupiter instrument in dry air atmosphere with heating and cooling rates of 5 C min1. Before sintering, oxy-sulfate and composite powders were pressed uniaxially under 50 MPa into bars of diameter 10 mm and length 40 mm, and then pressed isostatically under 600 MPa during 30 min using Hiperbaric 55 hydrostatic press. The bars were sintered at 900 C for 24 h. The level of porosity was determined by a “slice and view” approach using a FEI Helios 450S focused ion beam scanning electron microscope. Dilatometric characterization was performed by a MI-900 Michelson Laser Interferometer. The total conductivity of the bulk composites was measured by a 4-probe DC technique in air. 2.2. Preparation and charaterisation of symmetrical assemblies and fuel cells The Pr2O2SO4 oxy-sulfate and silver powders (60:40 wt.%) were mixed with terpineol in a NOAH NQM-2 planetary ball-milling machine at 400 rpm for 2 h followed by ultrasonication for 1 h, to break possible agglomerates. Ce0.9Gd0.1O2-d (CGO) powder (Alfa Aesar) was isostatically pressed at 200 MPa during 1 min into disks of 1.5 mm thickness and 10 mm diameter, followed by sintering at 1600 C for 5 h in air (heating and cooling rates of 1.5 C min1). Thin AgePr2O2SO4 films (0.264 cm2 of effective area) were deposited over the CGO disk using a WS-650-23 Laurell spin coater at 2000 rpm for 30 s, to provide a cathode electrode thickness of around 20 mm. For formation of symmetrical cathode-electrolytecathode assemblies, this process was repeated on the reverse side of the pellet after drying in air. The symmetrical assemblies were then annealed at 900 C for 5 h in air, with heating and cooling rates of 1.5 C min1. For the complete fuel cell, CGO electrolyte powder was spin-coated onto partially-sintered (900 C, 5 h) NiO/CGO (70 wt.% of NiO, Alfa Aesar) disks with thickness of 0.9 mm and diameter of 15 mm, followed by sintering at 1300 C for 10 h in air. The AgePr2O2SO4 (40 wt.% of silver) composite was then spincoated onto the center of electrolyte of the prepared anode/electrolyte assembly, as described for the symmetric cell. The reference electrode was prepared by attaching and firing platinum paste onto the electrolyte around the cathode (AgePr2O2SO4) as a ring. The cell morphology was observed using a Jeol JSM-6500F scanning electron microscope (SEM). The impedance spectra of the symmetric assembly were taken from 550 to 800 C in oxygen using Princeton 273A potentiostat/galvanostat in the frequency range of 0.01 Hze1 MHz using a signal amplitude of 50 mV. The power output of the fuel cell was monitored with the Arbin SOFC Testing System from 550 to 800 C with wet hydrogen (3% of H2O) as fuel and oxygen as oxidant with the same feeding rates of 80 ml min1. Long-term stability testing was performed at 800 C with a load of 0.2U, a constant load corresponding to the maximum power density, with gas feeding rates of 50 ml min1. The respective current was recorded automatically every 10 min, to a total of 24 days.
2.1. Preparation and characterization of composite cathode 3. Results and discussion Pr2O2SO4 oxy-sulfate powder was prepared using a solid-state reaction method using praseodymium (III) sulfate octahydrate Pr2(SO4)3$8H2O (Alfa Aesar). Before heating, this precursor was ground in an agate mortar and pestle. Then the powder was annealed at 850 C for 5 h in air (heating and cooling rates of 10 C min1). After pyrolysis, the formed Pr2O2SO4 oxy-sulfate powder was intimately mixed with silver powder (particle size of ~1.3 mm, Alfa Aesar) in order to prepare AgePr2O2SO4 mixtures (20, 40, 50, 60 and 80 wt.% of silver). X-ray diffraction (XRD) patterns were collected using a Bruker D-8 Advance diffractometer with a CuKa (l ¼ 0.15405 nm) radiation, at increments of 0.02 2q and
3.1. Structural, dilatometric and conductivity study Fig. 1 shows the TG and DSC profiles of the precursor Pr2(SO4)3$8H2O upon heating in the temperature range 25e1350 C. Three main weight loss events can be observed. Before 550 C, a 18% weight loss can be correlated to the dehydration of Pr2(SO4)3$8H2O to Pr2(SO4)3, as confirmed by complementary XRD studies (Fig. 2a). From 550 to 850 C, Pr2(SO4)3 remained stable, whereas after 850 C another 20% weight loss occurred identified by XRD to be that of transformation of Pr2(SO4)3 into Pr2O2SO4. At
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
613
1150 C, further weight loss coupled to a strongly exothermic event indicates the onset of decomposition of Pr2O2SO4 into oxides, as confirmed by XRD (Fig. 2a). The combined TG and XRD results of Figs. 1 and 2(a), respectively, confirm that Pr2O2SO4 can be easily synthesized by the solid state method and offers stability in air up to 1100 C. To investigate the stability of the Pr2O2SO4 in CO2, reaction with pure CO2 was attempted for 20 h at 800 C. The XRD pattern by peak intensity shows that less than ~5 wt.% of Pr2O2SO4 powder reacted with carbon dioxide to produce praseodymium carbonate (Fig. 2b). Fig. 3 summarizes information on the coefficient of thermal expansion (CTE) and the total conductivity of the AgePr2O2SO4 ceramics as a function of Ag content. The CTE of pure Pr2O2SO4 is shown to be 11.8$106 K1. With increase in Ag content, the CTE is noted to ascend linearly in agreement with theoretical prediction (Fig. 3). The conductivity as a function of Ag content measured at 700 C (Fig. 3) exhibits typical percolation behavior with the limit of percolation shown to be at approximately 40 wt.% Ag. Taking into account the limit of percolation and minimization of thermal expansion mismatch of the cathode with the electrolyte, the optimal content of 40 wt.% of Ag was selected for further electrochemical testing. 3.2. Microstructural studies The micrograph in Fig. 4a and corresponding EDS elemental mapping, Fig. 4b, illustrate the size of the Pr2O2SO4 grains to be 1e5 mm and to be well dispersed with the Ag particles. The silver is noted to form a continuous network inside the cathode, corresponding well with the percolation in electrical behavior observed in Fig. 3. Fig. 4c and d presents the cross-sectional view and EDS elemental mapping of a fractured surface of the prepared fuel cell, respectively, showing the deposited cathode film with ~5 mm thickness being well bonded with the CGO electrolyte (~10 mm thick), which in turn is supported by the NiO-CGO anode. The elemental distribution in Fig. 4d testifies an intimate and even distribution of the Ag and Pr2O2SO4 and NiO-CGO phases of the cathode and anode composites, respectively, and a clear compositional separation between each fuel cell layer. 3.3. Impedance spectroscopy characterization of the symmetrical assemblies The impedance spectra obtained for symmetrical cathode/
Fig. 1. TG and DSC curves of Pr2(SO4)3$8H2O measured in a flow of air (heating rate of 5 C min1).
Fig. 2. The XRD study of Pr2(SO4)3$8H2O transformation (a) and XRD profile of Pr2O2SO4 after attempted reaction with pure CO2 (b).
electrolyte/cathode assembly measured in flowing oxygen, without the application of DC bias, correspond to two partially overlapped arcs and an offset along the real Z'axis at high frequency. The high
Fig. 3. Thermal expansion coefficient and conductivity (700 composites.
C)
of AgePr2O2SO4
614
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
frequency offset, R1, represents the ohmic resistance of the measuring cell and the electrolyte, and was subtracted from Fig. 5 for clarity. The impedance responses were, thus, described by a resistor, R1, in series with a set of two parallel RjjCPE elements, where R is a resistance and CPE is a constant phase element defined as ZCPE ¼ Q1(ju)n, where the resultant fitting parameters can be related to the true capacitance by the equation C ¼ R(1 - n)/nQ1/n. The total electrode polarization resistance, Rp, expressed as Rp ¼ R2 þ R3, is shown to decrease with increasing temperature, Fig. 6, with an activation energy (Ea) of 1.32 eV in the temperature range 550e800 C, a value coincident to that reported in the literature for pure Ag cathodes, ~1.4 eV [26]. The Arrhenius plot (Fig. 6) further indicates that the higher-frequency arc, corresponding to resistance R2, constitutes the main contribution to Rp under the measured conditions. The respective capacitance values of this response are in the order of 106 F cm2 concurring with typical values expected for interfacial polarization resistance [27e29]. Conversely, the much larger capacitance of the lower frequency, R3, response, 103 F cm2, is typical of that expected for the chemical capacitance of solid-state diffusion and chemical exchange of oxygen species [27e29]. This polarization resistance is observed to contribute significantly to the total electrode resistance only at lower temperatures (Fig. 6). By following the oxygen dependencies of polarization resistances, the impact of Pr2O2SO4 additions on the electrochemical performance of a common SOFC cathode material La0.6Sr0.4Co0.2Fe0.8O3-d (LSCF) has been recently documented in the literature [25]. The presence of the Pr2O2SO4 in the LSCF cathode was noted to beneficially impact the interfacial polarization resistance and the dominant polarization resistance associated with solid-state diffusion and chemical exchange of oxygen species. The reported capacitance values for these contributions correspond well with those recorded for the two main electrode processes in the current
Fig. 5. Impedance spectra of AgePr2O2SO4jCGOjAgePr2O2SO4 symmetrical assembly at 550e800 C in oxygen, OCV. The high frequency offset along the real Z'axis has been subtracted for clarity and is not shown.
work, thus, reinforcing their similar attribution [25]. Nonetheless, in contrast to LSCF-Pr2O2SO4, Fig. 6 shows the current AgePr2O2SO4 cathode to be limited by the higher frequency response related to the resistance of interfacial polarization. Moreover, a significant change in the activation energy of the lower frequency polarization term, R3, is revealed with increasing temperature. The activation energy of this term at lower temperatures <700 C, 1.15 eV, compares well with that documented for absorption and diffusion of atomic oxygen inside a pure silver electrode, 1.05 eV [26], whereas the significant increase in
Fig. 4. SEM micrographs and corresponding EDS mapping of the AgePr2O2SO4 cathode (a and b) and the fuel cell (c and d). The red color refers to Pr-La, green - to Ag-La, blue - to SKa, cyan - to O-Ka or Ni-Ka signal, and yellow - to Ce-La. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
615
both phases (silver and oxy-sulfate) maintained their original structures. Nonetheless, small traces of Ag2SO4 were also observed (Fig. 8), suggesting that some chemical interaction had occurred on long term operation, albeit without producing a notable impact on the electrochemical cell performance. 4. Conclusions
Fig. 6. Arrhenius plot of the total (Rp), higher-frequency (R2) and lower-frequency (R3) electrode resistances of AgePr2O2SO4 (40e60% wt.), compared with total electrode resistances (Rp) of La0.6Sr0.4Co0.2Fe0.8O3-d (LSCF) [25], LSCF- Pr2O2SO4 (50-50% vol.) [25] and AgeSm1.8Ce0.2CuO4 (10e90% wt.) [30].
activation energy recorded at higher temperatures may reflect an increasing role of Pr2O2SO4 in these processes, comparable to the high levels of activation energy recorded for this term for the Pr2O2SO4-LSCF composite, ~2 eV [25]. To put the performance of the AgePr2O2SO4 cathode material into context, Fig. 6 further compares the current data to total polarization resistances recorded for LSCF and Pr2O2SO4-LSCF cathodes, prepared in a similar way by the present authors [25], and also that of a silver containing composite, AgeSm1.8Ce0.2CuO4 [30] measured under similar oxidizing conditions. Of these materials, the current AgePr2O2SO4 material is observed to offer the peak performance. As noted in the introduction, it has been previously suggested in the literature that the high oxygen storage capacity of Pr2O2SO4, and the available presence of the Pr-redox couple, may facilitate oxygen chemical exchange. In the current case this may assist the well-known catalytic properties of silver for the oxygen reduction reaction [6,31], leading to overall polarization resistances that are lower than those for similar composite materials in the literature [25].
A novel AgePr2O2SO4 composite has been synthesized and proved to be an efficient and relatively stable cathode for the CGO/ CGO-NiO electrolyte-anode system. This cell successfully generated power densities as high as 1.140 and 1.500 W cm2 at 750 and 800 C, respectively, and the power output maintained its original value for the 24 days of operation under real load mode at 800 C. Taking into account the redox behavior of Pr3þ/Pr4þ in AgePr2O2SO4 composite, we can assume that the ready presence of oxide-ions and reported high oxygen storage capacity in the Pr2O2SO4 phase facilitated gaseous oxygen to be adsorbed as negatively charged species onto the cathode surface, in agreement with other recent studies of composite cathodes with Pr2O2SO4 leading to a low polarization resistance for solid-state diffusion and chemical exchange of oxygen species [30].The results of the power density measurements of the single fuel cell showed the AgePr2O2SO4 composite to be a potential cathode for use in intermediate-temperature SOFCs.
3.4. Power density and long term stability test of the fuel cell Fig. 7a shows the currentevoltage and corresponding power density curves of the single fuel cell, AgePr2O2SO4/CGO/NiO-CGO, measured in O2/H2 (3 vol.% H2O) gradient. At each applied load, the corresponding IeV data points were recorded after a stabilization time of 1 h. Maximum power density values at 550, 600, 650, 700, 750, and 800 C were 0.075, 0.283, 0.492, 0.805, 1.140 and 1.500 W cm2, respectively. Variations between different samples were less than 10%. This fuel cell, offering a relatively high power output, was then tested for long-term stability at 800 C. The external resistance load of the fuel cell was kept at 0.2 Ohm, a value corresponding to the maximum power density obtained in the initial measurement. The initial fuel cell output voltage, corresponding to this maximum in power density, was 0.55 V. The current was recorded automatically every 10 min, to a total of 24 days (Fig. 7b). The power density at the end of the test maintained 1.45 W cm2; corresponding to a 3.3% drop in power output per 580 h. The cathode XRD analysis after the stability test showed that
Fig. 7. Current density dependencies of voltage and power density at 550e800 C (a) and long-term stability test at 800 C (b).
616
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
Fig. 8. The XRD pattern of the cathode after stability test.
Acknowledgments The authors gratefully acknowledge funding from the FCT, QREN, FEDER, COMPETE, POPH, FCT Investigator Programme, projects SFRH/BPD/86336/2012, IF/00280/2012 and PEst-C/EME/ UI0481/2013, Portugal and the European Social Fund, European Union. References [1] F. Abraham, M.F. Debreuille-Gresse, G. Nowogrochi, Solid State Ion. 28e30 (1988) 529e532.
[2] B.C.H. Steele, Solid State Ion. 129 (2000) 95e110. [3] J.B. Goodenough, A. Manthiram, M. Paranthaman, Y.S. Zhen, Mater. Sci. Eng. B12 (1992) 357e360. [4] C.R. Xia, M.L. Liu, Solid State Ion. 144 (2001) 249e255. [5] E.P. Murray, S.A. Barnett, Solid State Ion. 143 (2001) 265e273. [6] C.R. Xia, M.L. Liu, Adv. Mater. 14 (2002) 521e523. [7] C. Sun, R. Hui, J. Roller, J. Solid State Electrochem 14 (2009) 1125e1144. [8] E.P. Murray, M.J. Sever, S.A. Barnett, Solid State Ion. 148 (2002) 27e34. [9] T. Hibino, A. Hashimoto, T. Inoue, J. Tokuno, S. Yoshida, M. Sano, Science 288 (2000) 2031e2033. [10] Z.P. Shao, S.M. Haile, Nature 431 (9) (2004) 170e173. [11] A.V. Chadwick, C. Colli, C. Maltese, G. Morrison, I. Abrahams, A. Bush, Solid State Ion. 119 (1999) 79e84. [12] M. Alaga, A. Ammar, R. Essalim, B. Tanouti, F. Mauvy, R. Decourt, Solid State Sci. 7 (2005) 1173e1188. [13] C.K. Lee, A.R. West, Solid State Ion. 86e88 (1996) 235e239. [14] J.C. Boivin, G. Mairesse, Chem Mater. 10 (1998) 2870e2888. [15] C.R. Xia, F.L. Chen, M.L. Liu, Electrochem. Solid State Lett. 4 (5) (2001) A52eA54. [16] T. Yang, F. Li, D.G. Xia, J. Power Sources 195 (2010) 2514e2519. [17] T. Yang, C.M. Almeida, D. Ramasamy, F.A. Loureiro, J. Power Sources 269 (2014) 46e53. [18] M. Machida, Nagasaki Symposium on Nano-Dynamics, 2008. NSND2008. [19] J. Lian, X. Sun, Z. Liu, J. Yu, X. Li, Mater. Res. Bull. 44 (2009) 1822e1827. [20] M. Machida, T. Kawano, M. Eto, D. Zhang, K. Ikeue, Chem. Mater. 19 (2007) 954e960. [21] M. Machida, K. Kawamura, K. Ito, K. Ikeue, Chem. Mater. 17 (2005) 1487e1492. [22] J. Liang, R. Ma, F. Geng, Y. Ebina, T. Sasaki, Chem. Mater. 22 (2010) 6001e6007. [23] I. Valsamakis, M.F. Stephanopoulos, Appl. Catal. B 106 (2011), 25e263. [24] K. Ikeue, M. Eto, D.J. Zhang, T. Kawano, M. Machida, J. Catal. 248 (2007) 46e52. [25] F.A. Loureiro, T. Yang, D.G. Stroppa, D.P. Fagg, J. Mater. Chem. A 3 (2015) 12636e12641. [26] M. Kleitz, L. Dessemond, T. Kloidt, M.C. Steil, Solid Oxide Fuel Cells IV (1995) 527e536. [27] Y.M. Kim, S.I. Pyun, J.S. Kim, G.J. Lee, J. Electrochem. Soc. 154 (2007) B802eB809. [28] S.B. Adler, Chem. Rev. 104 (2004) 4791e4843. [29] V.C. Kourkoutis, F. Tietz, S. Bebelis, Fuel Cells 9 (2009) 852e860. [30] L.P. Sun, H. Zh, Q. Li, L.H. Huo, J.P. Viricelle, C. Pijolat, Int. J. Hydrogen Energy 38 (2013) 14060e14066. [31] J.S. Guo, A. Hsu, D. Chu, R.R. Chen, J. Phys. Chem. C 114 (10) (2010) 4324e4330.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Silverepraseodymium oxy-sulfate cermet: A new composite cathode for intermediate temperature solid oxide fuel cells Tao Yang*, Aliaksandr L. Shaula, Sergey M. Mikhalev, Devaraj Ramasamy, Duncan P. Fagg TEMA, Mechanical Engineering Department, University of Aveiro, 3810-193 Aveiro, Portugal
h i g h l i g h t s A novel AgePr2O2SO4 composite cathode was developed. High oxygen storage capacity facilitated gaseous oxygen to be adsorbed. A highly active property towards oxygen reduction reactions was achieved. The highest power density reached 1.5 W/cm2 at 800 C. High stability of the cell at 800 C under load was maintained for 24 days.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 September 2015 Received in revised form 17 December 2015 Accepted 18 December 2015 Available online xxx
AgePr2O2SO4 is identified as a promising new composite material to enhance the cathodic oxygen reduction reaction in solid oxide fuel cells. AgePr2O2SO4 was studied in terms of synthesis, stability of Pr2O2SO4 in CO2, and electrochemical behavior as a cathode. The performance of the composite cathode was assessed as a function of temperature by A.C. impedance using a symmetrical cell arrangement in oxygen. The global performance of an anode-supported fuel cell AgePr2O2SO4/CGO/NiO-CGO was also assessed, the highest power density being 1.5 W cm2 at 800 C. A longevity test of this cell performed at 800 C under load for 24 days demonstrated high stability with AgePr2O2SO4 cathode. © 2015 Elsevier B.V. All rights reserved.
Keywords: Ag-Pr2O2SO4 Novel cathode Solid oxide fuel cells Oxygen reduction reaction High power density
1. Introduction Solid oxide fuel cells (SOFCs) are devices of commercial and environmental importance that effectively convert the chemical energy of fuels directly to electricity. They combine advantages of environment-friendly electricity generation with fuel and oxidant flexibility, opening up an extensive range of potential applications, such as stationary power for isolated areas, public transportation power, and even portable power. Nonetheless, the marketability of these applications depends on the achievement of stable fuel cell operation at minimal cost. In order to attain this requirement a reduction in the SOFC working temperature, to at least 600e800 C, has been suggested as desirable [1e5]. The benefits of lowering the operating temperature include fast start-up of the system, an
* Corresponding author. E-mail address: [email protected] (T. Yang). http://dx.doi.org/10.1016/j.jpowsour.2015.12.098 0378-7753/© 2015 Elsevier B.V. All rights reserved.
increase in materials stability and a reduction in cost. However, with the decrease in the working temperature, the polarization resistances of the electrode and of the electrode/electrolyte interface can increase dramatically [4]. This is because the catalytic oxidation/reduction reactions on the surface of the electrode and the ionic transport within the electrode slow down and take major responsibility for impaired power density of the whole fuel cell [4]. In order to deal with this challenge, a variety of perovskite-type ABO3 and Aurivillius-type phases have been developed and improved as potential intermediate temperature SOFC cathodes [2,6]. With respect to the typical ABO3 perovskite structure, strontium-doped lanthanum manganite (La1-xSrxMnO3-d, LSM) mixed with yttria-stabilized zirconia (ZrO2)0.92(Y2O3)0.08, a classical ionic conductor, has traditionally been the composite cathode of choice. Here (ZrO2)0.92(Y2O3)0.08 provides oxygen ionic conductivity while the LSM phase predominantly works as an electronic conductor [7]. Single-phase mixed ionic and electronic conductors
612
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
(MIECs) such as La0.6Sr0.4Co0.8Fe0.2O3-d (LSCF) and Sm0.5Sr0.5Co0.5O3-d (SSC) have also been suggested to reduce the cathodic interfacial polarization resistance at lower temperatures [8,9]. Shao and Haile [10] replaced the rare-earth elements (Ln) with alkaline-earth (Ba) and presented Ba0.5Sr0.5Co0.8Fe0.2O3d (BSCF) as a new cathode material for reduced-temperature SOFC operation. The power densities of cells using this cathode reached values as high as 1010 mW cm2 at 600 C when operated with humidified hydrogen as the fuel and air as the oxidant [10]. For the Aurivillius-type structures, it was found that the Bi2Cu0.1V0.9O5.35 (known as BICUVOX10) showed very good electrochemical performance at 600 C [11e14]. However, its disadvantage of low electronic transference number requires the addition of an electronically superior conducting phase to form a composite. For example, Liu and Xia [6,15] mixed silver powder with BICUVOX10, to provide the required electronic conductivity offering a peak power density of 443 mW cm2 at 550 C. Tao et al. prepared a coreeshell Au-BICUVOX10 cathode fabricated on top of a finger-like CGO electrolyte to increase the triple phase boundary (TPB) length, and achieved a power density of 470 mW cm2 at 550 C by this modification [16,17]. Such previous work has, therefore, shown that a suitable cathode for an intermediate temperature fuel cell should offer sufficient electronic and ionic conductivities, either as a single-phase MIEC or as a composite of two phases to provide the MIEC nature. In an unconventional departure from previous materials for this goal, the current study offers a completely new composite cathode for SOFCs based on combination of silver particles and porous praseodymium oxy-sulfate, Pr2O2SO4. Such oxy-sulfate materials have traditionally been suggested as oxygen storage materials due to their large oxygen storage capacity [18]. Although there are various mechanisms suggested to be involved in the oxygen storage in Pr2O2SO4, for example that the distortion of the oxygen sub-lattice generates some mobile oxide-ions and that the change in the oxidation states of Pr3þ/Pr4þ can catalyze oxygen exchange, it is the redox ability of sulfur that is mainly attributed for the high levels of oxygen storage capacity [19]. Due to the ability of the Ln oxy-sulfates to store oxygen [20e22], they have been applied as catalysts for the high-temperature wateregas shift reaction [23], and for anaerobic catalytic CO oxidation [24]. In the present work, a novel composite cathode of AgePr2O2SO4 is prepared and carefully characterized in terms of stability, microstructure and electrochemical behavior. In this composite it is envisaged that silver will function as the electronic component, while the Pr2O2SO4 phase may provide mobile oxygen species and fast oxygen exchange [25]The performance of an anode-supported fuel cell of composition AgePr2O2SO4/CGO/NiOCGO was assessed, including longevity testing of an optimized cell operated at 800 C for a period of 24 days under load. 2. Experimental
counting time of 2 s per step. Thermogravimetric (TG) analysis and differential scanning calorimetry (DSC) were carried out using the Netzsch Jupiter instrument in dry air atmosphere with heating and cooling rates of 5 C min1. Before sintering, oxy-sulfate and composite powders were pressed uniaxially under 50 MPa into bars of diameter 10 mm and length 40 mm, and then pressed isostatically under 600 MPa during 30 min using Hiperbaric 55 hydrostatic press. The bars were sintered at 900 C for 24 h. The level of porosity was determined by a “slice and view” approach using a FEI Helios 450S focused ion beam scanning electron microscope. Dilatometric characterization was performed by a MI-900 Michelson Laser Interferometer. The total conductivity of the bulk composites was measured by a 4-probe DC technique in air. 2.2. Preparation and charaterisation of symmetrical assemblies and fuel cells The Pr2O2SO4 oxy-sulfate and silver powders (60:40 wt.%) were mixed with terpineol in a NOAH NQM-2 planetary ball-milling machine at 400 rpm for 2 h followed by ultrasonication for 1 h, to break possible agglomerates. Ce0.9Gd0.1O2-d (CGO) powder (Alfa Aesar) was isostatically pressed at 200 MPa during 1 min into disks of 1.5 mm thickness and 10 mm diameter, followed by sintering at 1600 C for 5 h in air (heating and cooling rates of 1.5 C min1). Thin AgePr2O2SO4 films (0.264 cm2 of effective area) were deposited over the CGO disk using a WS-650-23 Laurell spin coater at 2000 rpm for 30 s, to provide a cathode electrode thickness of around 20 mm. For formation of symmetrical cathode-electrolytecathode assemblies, this process was repeated on the reverse side of the pellet after drying in air. The symmetrical assemblies were then annealed at 900 C for 5 h in air, with heating and cooling rates of 1.5 C min1. For the complete fuel cell, CGO electrolyte powder was spin-coated onto partially-sintered (900 C, 5 h) NiO/CGO (70 wt.% of NiO, Alfa Aesar) disks with thickness of 0.9 mm and diameter of 15 mm, followed by sintering at 1300 C for 10 h in air. The AgePr2O2SO4 (40 wt.% of silver) composite was then spincoated onto the center of electrolyte of the prepared anode/electrolyte assembly, as described for the symmetric cell. The reference electrode was prepared by attaching and firing platinum paste onto the electrolyte around the cathode (AgePr2O2SO4) as a ring. The cell morphology was observed using a Jeol JSM-6500F scanning electron microscope (SEM). The impedance spectra of the symmetric assembly were taken from 550 to 800 C in oxygen using Princeton 273A potentiostat/galvanostat in the frequency range of 0.01 Hze1 MHz using a signal amplitude of 50 mV. The power output of the fuel cell was monitored with the Arbin SOFC Testing System from 550 to 800 C with wet hydrogen (3% of H2O) as fuel and oxygen as oxidant with the same feeding rates of 80 ml min1. Long-term stability testing was performed at 800 C with a load of 0.2U, a constant load corresponding to the maximum power density, with gas feeding rates of 50 ml min1. The respective current was recorded automatically every 10 min, to a total of 24 days.
2.1. Preparation and characterization of composite cathode 3. Results and discussion Pr2O2SO4 oxy-sulfate powder was prepared using a solid-state reaction method using praseodymium (III) sulfate octahydrate Pr2(SO4)3$8H2O (Alfa Aesar). Before heating, this precursor was ground in an agate mortar and pestle. Then the powder was annealed at 850 C for 5 h in air (heating and cooling rates of 10 C min1). After pyrolysis, the formed Pr2O2SO4 oxy-sulfate powder was intimately mixed with silver powder (particle size of ~1.3 mm, Alfa Aesar) in order to prepare AgePr2O2SO4 mixtures (20, 40, 50, 60 and 80 wt.% of silver). X-ray diffraction (XRD) patterns were collected using a Bruker D-8 Advance diffractometer with a CuKa (l ¼ 0.15405 nm) radiation, at increments of 0.02 2q and
3.1. Structural, dilatometric and conductivity study Fig. 1 shows the TG and DSC profiles of the precursor Pr2(SO4)3$8H2O upon heating in the temperature range 25e1350 C. Three main weight loss events can be observed. Before 550 C, a 18% weight loss can be correlated to the dehydration of Pr2(SO4)3$8H2O to Pr2(SO4)3, as confirmed by complementary XRD studies (Fig. 2a). From 550 to 850 C, Pr2(SO4)3 remained stable, whereas after 850 C another 20% weight loss occurred identified by XRD to be that of transformation of Pr2(SO4)3 into Pr2O2SO4. At
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
613
1150 C, further weight loss coupled to a strongly exothermic event indicates the onset of decomposition of Pr2O2SO4 into oxides, as confirmed by XRD (Fig. 2a). The combined TG and XRD results of Figs. 1 and 2(a), respectively, confirm that Pr2O2SO4 can be easily synthesized by the solid state method and offers stability in air up to 1100 C. To investigate the stability of the Pr2O2SO4 in CO2, reaction with pure CO2 was attempted for 20 h at 800 C. The XRD pattern by peak intensity shows that less than ~5 wt.% of Pr2O2SO4 powder reacted with carbon dioxide to produce praseodymium carbonate (Fig. 2b). Fig. 3 summarizes information on the coefficient of thermal expansion (CTE) and the total conductivity of the AgePr2O2SO4 ceramics as a function of Ag content. The CTE of pure Pr2O2SO4 is shown to be 11.8$106 K1. With increase in Ag content, the CTE is noted to ascend linearly in agreement with theoretical prediction (Fig. 3). The conductivity as a function of Ag content measured at 700 C (Fig. 3) exhibits typical percolation behavior with the limit of percolation shown to be at approximately 40 wt.% Ag. Taking into account the limit of percolation and minimization of thermal expansion mismatch of the cathode with the electrolyte, the optimal content of 40 wt.% of Ag was selected for further electrochemical testing. 3.2. Microstructural studies The micrograph in Fig. 4a and corresponding EDS elemental mapping, Fig. 4b, illustrate the size of the Pr2O2SO4 grains to be 1e5 mm and to be well dispersed with the Ag particles. The silver is noted to form a continuous network inside the cathode, corresponding well with the percolation in electrical behavior observed in Fig. 3. Fig. 4c and d presents the cross-sectional view and EDS elemental mapping of a fractured surface of the prepared fuel cell, respectively, showing the deposited cathode film with ~5 mm thickness being well bonded with the CGO electrolyte (~10 mm thick), which in turn is supported by the NiO-CGO anode. The elemental distribution in Fig. 4d testifies an intimate and even distribution of the Ag and Pr2O2SO4 and NiO-CGO phases of the cathode and anode composites, respectively, and a clear compositional separation between each fuel cell layer. 3.3. Impedance spectroscopy characterization of the symmetrical assemblies The impedance spectra obtained for symmetrical cathode/
Fig. 1. TG and DSC curves of Pr2(SO4)3$8H2O measured in a flow of air (heating rate of 5 C min1).
Fig. 2. The XRD study of Pr2(SO4)3$8H2O transformation (a) and XRD profile of Pr2O2SO4 after attempted reaction with pure CO2 (b).
electrolyte/cathode assembly measured in flowing oxygen, without the application of DC bias, correspond to two partially overlapped arcs and an offset along the real Z'axis at high frequency. The high
Fig. 3. Thermal expansion coefficient and conductivity (700 composites.
C)
of AgePr2O2SO4
614
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
frequency offset, R1, represents the ohmic resistance of the measuring cell and the electrolyte, and was subtracted from Fig. 5 for clarity. The impedance responses were, thus, described by a resistor, R1, in series with a set of two parallel RjjCPE elements, where R is a resistance and CPE is a constant phase element defined as ZCPE ¼ Q1(ju)n, where the resultant fitting parameters can be related to the true capacitance by the equation C ¼ R(1 - n)/nQ1/n. The total electrode polarization resistance, Rp, expressed as Rp ¼ R2 þ R3, is shown to decrease with increasing temperature, Fig. 6, with an activation energy (Ea) of 1.32 eV in the temperature range 550e800 C, a value coincident to that reported in the literature for pure Ag cathodes, ~1.4 eV [26]. The Arrhenius plot (Fig. 6) further indicates that the higher-frequency arc, corresponding to resistance R2, constitutes the main contribution to Rp under the measured conditions. The respective capacitance values of this response are in the order of 106 F cm2 concurring with typical values expected for interfacial polarization resistance [27e29]. Conversely, the much larger capacitance of the lower frequency, R3, response, 103 F cm2, is typical of that expected for the chemical capacitance of solid-state diffusion and chemical exchange of oxygen species [27e29]. This polarization resistance is observed to contribute significantly to the total electrode resistance only at lower temperatures (Fig. 6). By following the oxygen dependencies of polarization resistances, the impact of Pr2O2SO4 additions on the electrochemical performance of a common SOFC cathode material La0.6Sr0.4Co0.2Fe0.8O3-d (LSCF) has been recently documented in the literature [25]. The presence of the Pr2O2SO4 in the LSCF cathode was noted to beneficially impact the interfacial polarization resistance and the dominant polarization resistance associated with solid-state diffusion and chemical exchange of oxygen species. The reported capacitance values for these contributions correspond well with those recorded for the two main electrode processes in the current
Fig. 5. Impedance spectra of AgePr2O2SO4jCGOjAgePr2O2SO4 symmetrical assembly at 550e800 C in oxygen, OCV. The high frequency offset along the real Z'axis has been subtracted for clarity and is not shown.
work, thus, reinforcing their similar attribution [25]. Nonetheless, in contrast to LSCF-Pr2O2SO4, Fig. 6 shows the current AgePr2O2SO4 cathode to be limited by the higher frequency response related to the resistance of interfacial polarization. Moreover, a significant change in the activation energy of the lower frequency polarization term, R3, is revealed with increasing temperature. The activation energy of this term at lower temperatures <700 C, 1.15 eV, compares well with that documented for absorption and diffusion of atomic oxygen inside a pure silver electrode, 1.05 eV [26], whereas the significant increase in
Fig. 4. SEM micrographs and corresponding EDS mapping of the AgePr2O2SO4 cathode (a and b) and the fuel cell (c and d). The red color refers to Pr-La, green - to Ag-La, blue - to SKa, cyan - to O-Ka or Ni-Ka signal, and yellow - to Ce-La. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
615
both phases (silver and oxy-sulfate) maintained their original structures. Nonetheless, small traces of Ag2SO4 were also observed (Fig. 8), suggesting that some chemical interaction had occurred on long term operation, albeit without producing a notable impact on the electrochemical cell performance. 4. Conclusions
Fig. 6. Arrhenius plot of the total (Rp), higher-frequency (R2) and lower-frequency (R3) electrode resistances of AgePr2O2SO4 (40e60% wt.), compared with total electrode resistances (Rp) of La0.6Sr0.4Co0.2Fe0.8O3-d (LSCF) [25], LSCF- Pr2O2SO4 (50-50% vol.) [25] and AgeSm1.8Ce0.2CuO4 (10e90% wt.) [30].
activation energy recorded at higher temperatures may reflect an increasing role of Pr2O2SO4 in these processes, comparable to the high levels of activation energy recorded for this term for the Pr2O2SO4-LSCF composite, ~2 eV [25]. To put the performance of the AgePr2O2SO4 cathode material into context, Fig. 6 further compares the current data to total polarization resistances recorded for LSCF and Pr2O2SO4-LSCF cathodes, prepared in a similar way by the present authors [25], and also that of a silver containing composite, AgeSm1.8Ce0.2CuO4 [30] measured under similar oxidizing conditions. Of these materials, the current AgePr2O2SO4 material is observed to offer the peak performance. As noted in the introduction, it has been previously suggested in the literature that the high oxygen storage capacity of Pr2O2SO4, and the available presence of the Pr-redox couple, may facilitate oxygen chemical exchange. In the current case this may assist the well-known catalytic properties of silver for the oxygen reduction reaction [6,31], leading to overall polarization resistances that are lower than those for similar composite materials in the literature [25].
A novel AgePr2O2SO4 composite has been synthesized and proved to be an efficient and relatively stable cathode for the CGO/ CGO-NiO electrolyte-anode system. This cell successfully generated power densities as high as 1.140 and 1.500 W cm2 at 750 and 800 C, respectively, and the power output maintained its original value for the 24 days of operation under real load mode at 800 C. Taking into account the redox behavior of Pr3þ/Pr4þ in AgePr2O2SO4 composite, we can assume that the ready presence of oxide-ions and reported high oxygen storage capacity in the Pr2O2SO4 phase facilitated gaseous oxygen to be adsorbed as negatively charged species onto the cathode surface, in agreement with other recent studies of composite cathodes with Pr2O2SO4 leading to a low polarization resistance for solid-state diffusion and chemical exchange of oxygen species [30].The results of the power density measurements of the single fuel cell showed the AgePr2O2SO4 composite to be a potential cathode for use in intermediate-temperature SOFCs.
3.4. Power density and long term stability test of the fuel cell Fig. 7a shows the currentevoltage and corresponding power density curves of the single fuel cell, AgePr2O2SO4/CGO/NiO-CGO, measured in O2/H2 (3 vol.% H2O) gradient. At each applied load, the corresponding IeV data points were recorded after a stabilization time of 1 h. Maximum power density values at 550, 600, 650, 700, 750, and 800 C were 0.075, 0.283, 0.492, 0.805, 1.140 and 1.500 W cm2, respectively. Variations between different samples were less than 10%. This fuel cell, offering a relatively high power output, was then tested for long-term stability at 800 C. The external resistance load of the fuel cell was kept at 0.2 Ohm, a value corresponding to the maximum power density obtained in the initial measurement. The initial fuel cell output voltage, corresponding to this maximum in power density, was 0.55 V. The current was recorded automatically every 10 min, to a total of 24 days (Fig. 7b). The power density at the end of the test maintained 1.45 W cm2; corresponding to a 3.3% drop in power output per 580 h. The cathode XRD analysis after the stability test showed that
Fig. 7. Current density dependencies of voltage and power density at 550e800 C (a) and long-term stability test at 800 C (b).
616
T. Yang et al. / Journal of Power Sources 306 (2016) 611e616
Fig. 8. The XRD pattern of the cathode after stability test.
Acknowledgments The authors gratefully acknowledge funding from the FCT, QREN, FEDER, COMPETE, POPH, FCT Investigator Programme, projects SFRH/BPD/86336/2012, IF/00280/2012 and PEst-C/EME/ UI0481/2013, Portugal and the European Social Fund, European Union. References [1] F. Abraham, M.F. Debreuille-Gresse, G. Nowogrochi, Solid State Ion. 28e30 (1988) 529e532.
[2] B.C.H. Steele, Solid State Ion. 129 (2000) 95e110. [3] J.B. Goodenough, A. Manthiram, M. Paranthaman, Y.S. Zhen, Mater. Sci. Eng. B12 (1992) 357e360. [4] C.R. Xia, M.L. Liu, Solid State Ion. 144 (2001) 249e255. [5] E.P. Murray, S.A. Barnett, Solid State Ion. 143 (2001) 265e273. [6] C.R. Xia, M.L. Liu, Adv. Mater. 14 (2002) 521e523. [7] C. Sun, R. Hui, J. Roller, J. Solid State Electrochem 14 (2009) 1125e1144. [8] E.P. Murray, M.J. Sever, S.A. Barnett, Solid State Ion. 148 (2002) 27e34. [9] T. Hibino, A. Hashimoto, T. Inoue, J. Tokuno, S. Yoshida, M. Sano, Science 288 (2000) 2031e2033. [10] Z.P. Shao, S.M. Haile, Nature 431 (9) (2004) 170e173. [11] A.V. Chadwick, C. Colli, C. Maltese, G. Morrison, I. Abrahams, A. Bush, Solid State Ion. 119 (1999) 79e84. [12] M. Alaga, A. Ammar, R. Essalim, B. Tanouti, F. Mauvy, R. Decourt, Solid State Sci. 7 (2005) 1173e1188. [13] C.K. Lee, A.R. West, Solid State Ion. 86e88 (1996) 235e239. [14] J.C. Boivin, G. Mairesse, Chem Mater. 10 (1998) 2870e2888. [15] C.R. Xia, F.L. Chen, M.L. Liu, Electrochem. Solid State Lett. 4 (5) (2001) A52eA54. [16] T. Yang, F. Li, D.G. Xia, J. Power Sources 195 (2010) 2514e2519. [17] T. Yang, C.M. Almeida, D. Ramasamy, F.A. Loureiro, J. Power Sources 269 (2014) 46e53. [18] M. Machida, Nagasaki Symposium on Nano-Dynamics, 2008. NSND2008. [19] J. Lian, X. Sun, Z. Liu, J. Yu, X. Li, Mater. Res. Bull. 44 (2009) 1822e1827. [20] M. Machida, T. Kawano, M. Eto, D. Zhang, K. Ikeue, Chem. Mater. 19 (2007) 954e960. [21] M. Machida, K. Kawamura, K. Ito, K. Ikeue, Chem. Mater. 17 (2005) 1487e1492. [22] J. Liang, R. Ma, F. Geng, Y. Ebina, T. Sasaki, Chem. Mater. 22 (2010) 6001e6007. [23] I. Valsamakis, M.F. Stephanopoulos, Appl. Catal. B 106 (2011), 25e263. [24] K. Ikeue, M. Eto, D.J. Zhang, T. Kawano, M. Machida, J. Catal. 248 (2007) 46e52. [25] F.A. Loureiro, T. Yang, D.G. Stroppa, D.P. Fagg, J. Mater. Chem. A 3 (2015) 12636e12641. [26] M. Kleitz, L. Dessemond, T. Kloidt, M.C. Steil, Solid Oxide Fuel Cells IV (1995) 527e536. [27] Y.M. Kim, S.I. Pyun, J.S. Kim, G.J. Lee, J. Electrochem. Soc. 154 (2007) B802eB809. [28] S.B. Adler, Chem. Rev. 104 (2004) 4791e4843. [29] V.C. Kourkoutis, F. Tietz, S. Bebelis, Fuel Cells 9 (2009) 852e860. [30] L.P. Sun, H. Zh, Q. Li, L.H. Huo, J.P. Viricelle, C. Pijolat, Int. J. Hydrogen Energy 38 (2013) 14060e14066. [31] J.S. Guo, A. Hsu, D. Chu, R.R. Chen, J. Phys. Chem. C 114 (10) (2010) 4324e4330.