- Email: [email protected]
Y0.08Sr0.88TiO3–CeO2 composite as a diffusion barrier layer for stainless-steel supported solid oxide fuel cell
Journal of Power Sources 307 (2016) 385e390
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Y0.08Sr0.88TiO3eCeO2 composite as a diffusion barrier layer for stainless-steel supported solid oxide fuel cell Kun Joong Kim, Sun Jae Kim, Gyeong Man Choi* Fuel Cell Research Center/Department of Materials Science and Engineering, Pohang University of Science and Technology (POSTECH), Pohang, 790-784, Republic of Korea
h i g h l i g h t s STS-supported SOFC with YSZ and Ni-YSZ is tape-casted and co-fired at 1350 C. Y0.08Sr0.88TiO3eCeO2 composite is tested as a new diffusion barrier layer (DBL). The cell with DBL shows peak power density ~220 mW cm2 and maintains it at 700 C.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 April 2015 Received in revised form 24 December 2015 Accepted 27 December 2015 Available online xxx
A new diffusion barrier layer (DBL) is proposed for solid oxide fuel cells (SOFCs) supported on stainlesssteel where DBL prevents inter-diffusion of atoms between anode and stainless steel (STS) support during fabrication and operation of STS-supported SOFCs. Half cells consisting of dense yttria-stabilized zirconia (YSZ) electrolyte, porous Ni-YSZ anode layer, and ferritic STS support, with or without Y0.08Sr0.88TiO3eCeO2 (YST-CeO2) composite DBL, are prepared by tape casting and co-firing at 1250 and 1350 C, respectively, in reducing (H2) atmosphere. The porous YST-CeO2 layer (t ~ 60 mm) blocks interdiffusion of Fe and Ni, and captures the evaporated Cr during cell fabrication (1350 C). The cell with DBL and La0.6Sr0.4Co0.2Fe0.8O3d (LSCF) cathode achieved a maximum power density of ~220 mW cm2 which is stable at 700 C. In order to further improve the power performance, Ni coarsening in anode during cofiring must be prevented or alternative anode which is resistive to coarsening is suggested. This study demonstrates that the new YST-CeO2 layer is a promising as a DBL for stainless-steel-supported SOFCs fabricated with co-firing process. © 2016 Elsevier B.V. All rights reserved.
Keywords: Inter-diffusion Co-firing Reducing atmosphere Degradation
1. Introduction Solid oxide fuel cells (SOFCs) have possible applications as energy conversion devices due to their high energy conversion efficiency and the diversity of fuels [1]. Recent developments in the electrodes and electrolytes have decreased the operating temperature to ~600 C, so metal (e.g., Ni)-ceramic supports can be replaced by metal supports such as stainless steel (STS) [2]. The metal can provide high mechanical strength and thermal-shock resistance for the brittle ceramics and thus strengthen the cell [3]. Thus, STS-supported SOFCs (STS-SOFCs) may possibly be used for mobile application that requires both low temperature
* Corresponding author. E-mail addresses: [email protected] (S.J. Kim), [email protected] (G.M. Choi).
(K.J.
http://dx.doi.org/10.1016/j.jpowsour.2015.12.130 0378-7753/© 2016 Elsevier B.V. All rights reserved.
Kim),
[email protected]
operation and fast cycling speed. Appropriate selection of a supporting metal is essential to achieve high power density and durability. In addition to mechanical strength and electrical conductivity, other requirements such as reactivity with other components and matching of thermal expansion coefficients (TECs) also must be considered because these are directly related to power density and durability of the cell. STS is a popular supporting metal for metal-supported SOFCs (MS-SOFCs); it shows good TECcompatibility (10e12 ppm K1) with electrolyte materials and ferritic STS without Ni is cheaper than many special metals. STSsupported SOFCs are often fabricated by co-firing of layers at high-temperature in a reducing atmosphere to avoid metal oxidation. However, inter-diffusion between Fe, Cr in ferritic-STS support and Ni from the anode can occur during fabrication and during cell operation [4e6]. When Ni from the anode diffuses into the STS support, it changes STS from ferrite to austenite, and this
386
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
transformation causes undesirable TEC change [7]. In addition, diffusion of Cr into the anode may cause oxide scales to form on the Ni particles; this passivated part would increase anodic polarization, leading to the poor power density and fast degradation of the cell [8]. Several attempts have been made to overcome these problems. Fabrication of cells at relatively low temperature has been studied using methods such as electrophoretic deposition (EPD) [9], pulsed laser deposition (PLD) [10] and plasma spray. These cells have shown relatively high power density without Ni contamination. However, fabrication at low temperature requires high-cost equipment and entails a complicated procedure. Furthermore, cells with a large area and various shapes cannot be easily fabricated by these methods. Other approach to address the problems is the infiltration techniques to prepare the anode [11]. However, Ni coarsening during operation degrades power density [11]. Lastly, insertion of a diffusion barrier layer (DBL) has been explored extensively as a solution to avoid reaction of Ni and Fe/Cr [12e16]. The barrier layer must prevent inter-diffusion while allowing electron and gas transport. The layer should also have similar TECs with the other cell components, and be stable and compatible with the relevant operating and processing conditions. Applying a DBL is one of the most promising methods in terms of stability and practical use. Compositions based on LaCrO3 and LaMnO3 [8,13,14]; Cr2O3/Cr2MnO4, CeO2, and Ce0.8Gd0.2O2 (GDC) [15] have been evaluated as diffusion barriers. LaMnO3-based composition is expected to be decomposed in the fuel atmosphere [16], and Cr2O3 based has shown Cr diffusion into the Ni-based anode [15]. CeO2 and GDC form an effective diffusion barrier. However, the cation of CeO2 or GDC is partially reduced from Ce4þ to Ce3þ in low oxygen partial-pressure (Po2) leading to lattice expansion that may result in the mechanical failure [17], particularly during firing at high temperature. To compensate for the mechanical stability of CeO2, an additional composition should be considered. Donor (Y or La)-doped SrTiO3, evaluated as an alternative anode material, could be a good additive material to remedy the shortcomings of CeO2. (Y, Sr) TiO3 (YST) or (La, Sr)TiO3 (LST) are chemically stable in a reducing atmosphere. The electrical conductivity of donor-doped SrTiO3 sintered in reducing atmosphere is 1e2 orders of magnitude higher than that sintered in air or CeO2 [18,19]. In addition, the donordoped SrTiO3 based materials have TEC (12 ppm K1) similar to those of the typical electrolyte materials. Thus, the addition of donor-doped SrTiO3 to CeO2 could yield a stable diffusion barrier. In this study, 8 mol% Y-doped SrTiO3 (Y0.08Sr0.88TiO3) was selected due to its highest conductivity among YSTs [20] to composite with 50 wt.% CeO2 and the Y0.08Sr0.88TiO3eCeO2 composite was tested as a DBL. After co-firing of STS-supported SOFCs with and without DBL at elevated temperature (1250e1350 C) in the reducing atmosphere, compositional changes across Ni-YSZ anode, DBL, and STS layers were examined. The effect of DBL on the electrochemical performance and stability was monitored at 700 C. 2. Experiment Button-type SOFCs with and without Y0.08Sr0.88TiO3eCeO2 (YSTCeO2) composite as a DBL were fabricated by the tape-casting method (Fig. 1). The cell without DBL, (a) Pt/YSZ/Ni-YSZ/STS, was fired at 1250 C and the cell with DBL, (b) La0.6Sr0.4Co0.2Fe0.8O3d (LSCF)/YSZ/Ni-YSZ/YST-CeO2/STS, was fired at 1350 C. The firing temperatures were selected for the proper densification of YSZ layer. For the DBL, YST was prepared from Y2O3, SrCO3 and TiO2 (99.9%, High Purity Chemicals, Japan). The mixed powders were milled with zirconia balls in ethanol (99.9%, SAMCHEN Chemicals, Korea) for 12 h then the ethanol was removed by evaporation. The
dried powder was calcined at 1200 C for 5 h in air. X-ray diffraction (XRD) measurement confirmed that the resultant powder is a single-phase YST. The synthesized YST and an equal weight of CeO2 (99.9%, High Purity Chemicals, Japan) with 15 wt.% of corn starch (D.C. Chemicals, Korea) as a pore former consist of DBL powder. To prepare the slurry for tape casting, powders were mixed with binder solution composed of toluene and ethanol as solvents, polyvinyl butyral (PVB, B-76) as a binder (10% of raw powders) and dioctyl phthalate (DOP) as a plasticizer (50% of binder contents). Then, the mixture of powders and binder solution was milled with zirconia balls (diameter; 10 mm and 5 mm) for 72 h. Green sheets of DBL were cast to a thickness of 40e50 mm after drying. Commercial stainless-steel powder (STS-434L, 400e500 mesh, Daekwang Industry, Korea) was chosen as a material for metal support. Planetary-milled (Pulverisette 6, Netzsch, Germany) NiO (1 mm, 99.97%, Kojundo chemical, Japan) powder and 8 mol% Y2O3-stabilized ZrO2 (YSZ, 0.3 mm, TZ-8YS, Tosoh, Japan) powder were mixed in 60:40 wt.% for anode slurries. YSZ slurries for electrolyte were prepared in a similar method to form green sheets. As-cast green sheets of support, DBL, anode, and electrolyte were used for the cells with or without DBL. Both green cells with and without DBL layer were laminated at 30 MPa and 60 C for 20 min to achieve the desired thickness, then punched out to yield a circular green-cell of 24-mm diameter. The green cells were heated to burn binder at or below 400 C for 12 h and co-fired at 1350 C for the cell with DBL and 1250 C for the cell without DBL, respectively, for 3 h in dry hydrogen atmosphere. To avoid deformation of the sample during the firing process, a porous ZrO2 plate was used to apply a vertical load (1.5 g cm2) to the cell. The shrinkage curves of YSZ, STS434L, Ni-YSZ and YST-CeO2 tapes (50 5 mm2 size) were determined simultaneously by positioning and firing all the tapes along the temperature gradient of a tube furnace (754e1277 C) in dry H2. The temperature of each position was measured using an R-type thermocouple. The heating rate was 2 C/min and the cooling rate was 5 C/min. For the electrochemical measurement of cells, ceramic bond (Model 571, Aremco, USA) was used to mount the sintered cell on an alumina tube. Platinum (#6082, Heraeus, Germany) and La0.6Sr0.4Co0.2Fe0.8O3d (LSCF, AGC Seimi Chemical Co., Japan) pastes, respectively, were screen-printed as a cathode (0.502 cm2) for the cell with and without DBL, respectively. Although Pt cathode is stable and easy to prepare, it is expected to show high polarization resistance. A mixed ionic and electronic conductor (MIEC) cathode (LSCF) was selected for the cell with DBL to obtain a better performance. For firing of the LSCF cathode, the cell was pre-fired at 900 C for 2 h while maintaining the reducing atmosphere in the anode and open air in the cathode. Pt was used as a cathode for the cell without DBL and was not pre-fired. After pre-firing of the cell with LSCF cathode, the cell was cooled to room temperature, then Pt paste and mesh were attached on the cathode as a current collector. The electrochemical performances of both cells were evaluated using wet H2 gas as a fuel gas and open air as an oxidant gas. The expected (or calculated) Po2 in the anode gas was ~1022 atm at 800 C. The microstructure of cross-section of the cell was examined using a field emission scanning electron microscope (FE-SEM, Model XL30S FEG, Philips Electron Optics B.V., Netherlands). Atomic (Ni, Fe, Cr) distributions across anode, DBL, and STS support layers were examined using an energy dispersive spectroscopy (EDS) line profile to examine the possible reaction between the layers. The current-voltage-power (IeVeP) curve and the impedance were measured using an AC impedance analyzer (VSP, Bio Logic Science instruments, France) using wet H2 (97% H2 þ 3% H2O) as a fuel gas and open air as an oxidant gas. Current-voltage-power measurements were conducted at 700 C for 40 h.
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
387
Fig. 1. Schematics of two STS-supported SOFCs; (a) Pt/YSZ/Ni-YSZ/STS cell (without DBL, sintered at 1250 C) (b) LSCF/YSZ/Ni-YSZ/YST-CeO2/STS cell (with DBL, sintered at 1350 C).
3. Results and discussion We used different firing temperatures for two cells. To obtain pore-free YSZ electrolyte during co-firing, the required minimum temperatures were 1250 C for the cell without DBL and 1350 C for the cell with DBL. This 100 C difference in firing temperature may be due to the poor sinterability of DBL. Linear shrinkage curves for STS support, YST-CeO2 DBL, Ni-YSZ anode, and YSZ electrolyte were obtained and compared at temperatures between 754 and 1277 C (Fig. 2). The initial shrinkage values (1e4%) at ~740 C differ due to NiO reduction in Ni-YSZ, burn-out of pore former (15 wt.%) in YSTCeO2 layer, in addition to the burning of binder. At the highest firing temperature (1277 C), the shrinkage of DBL is approximately onehalf of YSZ. This means that DBL in the cell may impede overall shrinkage of the cell, especially the densification of electrolyte. The poor sintering or shrinkage is the nature of inert composite where YST or CeO2 has limited solubility each other and thus impede the shrinkage of counterpart particles. Ni-YSZ composite and STS show the similar shrinkage values that are intermediate between those of YSZ and DBL. We also fabricated the cell without DBL at 1350 C, however, Ni-YSZ anode after firing in dry H2 gas showed very dense microstructure which is not suitable for use in an electrochemical cell and was excluded in the discussion. The diffusion profiles of both of as-sintered cells were analyzed to determine the effect of YST-CeO2 DBL on the diffusion during fabrication process. In the cell without DBL (Fig. 3a), both Fe and Cr
Shrinkage [%]
0 5 10 Fig. 3. EDS line profile of fractured cross-section of the cells (a) without DBL (co-fired at 1250 C) and (b) with DBL (co-fired at 1350 C).
15 20 25 700
YSZ Ni-YSZ YST-CeO2 STS
800
900
1000
1100
1200
1300
Temperature [°C] Fig. 2. Shrinkage curves for YSZ, STS434L, Ni-YSZ and YST-CeO2. Green tapes of 50 5 mm2 size for each composition were sintered simultaneously using tube furnace with temperature gradient. Gas: dry hydrogen, temperature range: 754e1277 C. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
of STS diffused into the Ni-YSZ anode during co-firing at 1250 C as previously reported [8,21]. In the cell with DBL, diffusion of Fe and Cr was prevented during co-firing at 1350 C (Fig. 3b). Although Cr was found in the DBL, little Cr was found in the anode layer. During high-temperature fabrication, diffusion of Fe was completely prevented from STS to DBL since no Fe was found in DBL layer. In contrast, Cr may have diffused or evaporated and may have been captured by the DBL layer. Thus YST-CeO2 seemed to function well as a DBL during co-firing. The time-dependent changes of impedance spectra of both cells
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
6
0h 40 h
(a) w/o DBL cell
4 2 0 0
15
20
(b) w/ DBL cell
42 0h 40 h
0.6
28 0.3
14
700 °C Air | 97 % H + 3 % H O 0
40
80
120
160
-2
2
10
56
0.9
0.0
1.5
0 200
1.0
(b) w/ DBL cell
0.0 -0.5
240
1.2
0.5
0
1
2
3
4
2
Zreal [Ω cm ] Fig. 4. Impedance spectra of cells (a) without and (b) with DBL during operation for 40 h at 700 C. 97% H2 þ 3% H2O mixture was used as fuel and air as oxidant gases. Total resistance of cell without DBL increased severely with time.
0.9
180
0.6
120
0.3
60
-2
Power density [mW cm ]
-Zimag [Ω cm ]
5
70
1.2
700 °C Air | 97 % H2 + 3 % H2O
Cell voltage [V]
(a) w/o DBL cell
Cell voltage [V]
2
8
Power density [mW cm ]
-Zimag [Ω cm ]
388
were obtained to see the effect of DBL during operation at 700 C. The impedance spectra of the cell without DBL increased severely in size with time (Fig. 4a), but those of the cell with DBL showed little change during operation for 40 h (Fig. 4b). The cell with DBL shows much lower Ohmic and polarization area-specific resistance (ASR) values than the cell without DBL, both initially and after 40 h operation (Table 1). At initial operation, the higher Ohmic ASR of the cell without DBL (1.1 U cm2) than the cell with DBL (0.19 U cm2) may be caused by a reaction layer such as a resistive NieCr oxide phase formed between the Ni-YSZ anode and the STS. The polarization ASR difference (3.5 versus 6.4 U cm2 at 0 h) mainly originated from the difference in cathode materials (Pt for the cell without DBL vs. LSCF for the cell with DBL), inter-diffusion, and anode microstructure. Although the origin of strong increase of polarization ASR over time for the cell without DBL is not clear, Fe/ Cr contamination (especially Cr evaporation) at the anode is strongly suspected because the microstructure of Pt cathode and Ni-YSZ anode might be relatively stable at 700 C. To support this, we have fabricated an electrolyte (YSZ)esupported cell with symmetrical Pt cathode and observed impedance change with time at 700 C in air; the total polarization ASR increased from 1.2 to 1.6 U cm2 for 40 h. Meanwhile, the polarization ASR for the present cell without DBL increased from 6.4 to 18.4 U cm2 for the same time. Thus degradation due to Pt cathode is not significant and it is mostly due to anode. The stable polarization ASR of the cell with DBL suggests that Cr captured by the YST-CeO2 layer degrades little
0.0
0
200
400
600
0 800
-2
Current density [mA cm ] Fig. 5. Current-voltage and currentepower curves of cells (a) without DBL and (b) with DBL tested at 700 C for 40 h. 97% H2 þ 3% H2O mixture was used as fuel and air as oxidant gases.
the cell performance at 700 C for 40 h. Therefore, we conclude YSTCeO2 works as a diffusion barrier layer for at least 40 h of operation after high-temperature firing. We also measured currentevoltage (IeV) curves of each cell type at 700 C, and observed that both cells exhibited OCV value of z1 V (Table 1 and Fig. 5). However, the power density differed significantly between two cells. The cell without the DBL achieved a maximum power density (MPD) of only ~62 mW cm2 (Table 1; Fig. 5a) and the MPD degraded rapidly as expected. Meanwhile, the cell with DBL achieved a MPD of ~220 mW cm2 (Table 1; Fig. 5b). The reason for the enhanced performance and durability for the cell with DBL is explained by lower ASRs compared to the cell without DBL as discussed in impedance analysis: lower Ohmic ASR due to little resistive phase, lower polarization ASR due to MIEC cathode and the anode free from Cr contamination. Although the total ASR of the cell with DBL is nearly constant for 40 h in OCV condition, a slight decrease in MPD may be due to the degradation of cathodic performance [22] under current in addition to the slight decrease in OCV from 1.09 to 1.08 V (Table 1). Although applying DBL enhanced
Table 1 ASRs and power density of the cells with and without DBL. Types
Time [h]
Ohmic ASR [U cm2]
Polarization ASR [U cm2]
OCV [V]
Max power density [mW cm2]
Cell without DBL
0 40 0 40
1.1 0.61 0.19 0.21
6.4 18.4 3.5 3.5
1.12 1.12 1.09 1.08
62 35 220 203
Cell with DBL
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
389
Fig. 6. S.E.M. images of the cell with DBL: (a) YSZ electrolyte/Ni-YSZ anode, (b) YST-CeO2 DBL, (c) LSCF cathode/YSZ electrolyte. Ni-YSZ anode was co-fired at 1350 C in reducing atmosphere.
both the performance and stability of the cell, the total ASRs for the cell are still limited by the large polarization resistance (Table 1; Fig. 4b); the polarization ASR to total ASR ratio is ~94% and an arc at low frequency is much larger than that at high frequency. The origin of large ASR is not clear at present. However, a large part of it may come from cathode since oxygen reduction reaction (ORR) is often becomes slow with decreasing temperature. The contribution of anodic ASR also cannot be entirely neglected since the polarization is strongly dependent upon the change of H2 content in anode gas (not shown here) [23]. To confirm the performance of electrodes, microstructures of both anode and cathode were examined after test for the cell with DBL (Fig. 6); the microstructure of (a) Ni-YSZ anode was less porous than that of either (b) YST-CeO2 DBL or (c) LSCF cathode. Ni particles of Ni-YSZ anode may have coarsened during firing in reducing atmosphere at high temperature after reduction from NiO and the coarsening may have contributed to the large polarization ASR shown in Fig. 4b [24]. Thus, for further improvement of anodic performance, the anode microstructure must be controlled. Use of a 65-mm-thick YST-CeO2 layer have provided stable electrochemical performance, which means that the layer successfully prevents inter-diffusion of Fe/Cr and Ni between STS support and Ni-based anode during cell processing and operation. Compared to the reported DBL compositions such as Cr2O3-based composition [15] and Cu-YSZ cermet [25], the YST-CeO2 shows better chemical stability in a reducing atmosphere and better diffusion-barrier ability. Cr2O3-based composition has shown Cr diffusion into Ni-based anode after high-temperature co-firing in reducing atmosphere. A Cu-YSZ cermet DBL was found to block Ni diffusion into the support, but not able to block Fe and Cr diffusion into the anode. The cell with CeO2 DBL showed limited lifetime (~165 h) due to the start of breakaway oxidation of the metallic substrate [15]. Meanwhile cell with LaCrO3-based and LaMnO3based [8,13,14] DBL showed good stability for long operation time (degradation rate ~1% 1000 h1 at 800 C), however, the stability of these cells is still questionable when they are co-fired at high temperature. Our results demonstrated the stability of cell during
co-firing at 1350 C as well as during operation at 700 C for 40 h. Long-term durability (>500 h) will be further tested in the future for the reliability of the STS-supported SOFCs with YST-CeO2 layer. In terms of the cell performance, the dense microstructure in anode becomes the limiting factor. The use of alternative anode which is resistive to agglomeration is suggested.
4. Conclusions Y0.08Sr0.88TiO3eCeO2 (YST-CeO2) composite was used as diffusion barrier layer (DBL) to prevent mutual diffusion of Fe and Cr in stainless-steel (STS) and Ni in anode during co-firing and operation of STS-supported solid oxide fuel cells (SOFCs). SOFCs with and without DBL were prepared using tape-casting, screen-printing, and subsequent co-firing at elevated temperature in reducing atmosphere. After co-firing of both cells, atomic distribution across Ni-YSZ anode, YST-CeO2 DBL, and STS support was examined using an energy dispersive spectroscopy (EDS) line profile to confirm the effect of DBL on the inter-diffusion. The power density and impedance spectra for both cells at the initial stage (0 h) were compared as well as after 40 h of operation. A porous YST-CeO2 layer effectively blocked inter-diffusion of Fe and Ni, and captured the evaporated Cr during cell fabrication (1350 C) and operation. To improve further the electrochemical performance of the cell, Ni coarsening in anode during co-firing must be prevented or alternative anode which is resistive to coarsening is suggested. This study demonstrates that the new YST-CeO2 layer could be a promising as DBL for stainless-steel-supported SOFCs.
Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (grant no. 2011-0023389).
390
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
References [1] S.C. Singhal, Solid State Ionics 135 (2000) 305e313. [2] N. Christiansen, J.B. Hansen, H. Holm-Larsen, S. Linderoth, P.H. Larsen, P.V. Hendriksen, M. Mogensen, in: 7th European SOFC Forum, 2006, p. B034. [3] Y. Matus, L. De Jonghe, C. Jacobson, S. Visco, Solid State Ionics 176 (2005) 443e449. [4] S. Visco, C. Jacobson, L. De Jonghe, in: SECA Workshop, 2003. [5] M. Brandner, in: Solid Oxide Fuel Cells IX (SOFC-IX), 2005, p. 1235. [6] D. Hathiramani, in: e-Proceedings ICACC, Cocoa Beach, Florida, USA, 2006. [7] Z. Yang, D. Paxton, K. Weil, J. Stevenson, P. Singh, Materials Properties Database for Selection of High-temperature Alloys and Concepts of Alloy Design for SOFC Applications, PNNL-14116. Prepared for the U.S. Department of Energy under Contract DE-AC06-76RL01830, 2002. [8] T. Franco, K. Schibinger, Z. Ilhan, G. Schiller, A. Venskutonis, ECS Trans. 7 (1) (2007) 771e780. [9] P. Banace, N.P. Brandon, B. Girvan, P. Holbeche, S. O'Dea, B.C.H. Steele, J. Power Sources 131 (2004) 86e90. [10] S. Hui, D. Yang, Z. Wang, S. Yick, C. Deces-Petit, W. Qu, A. Tuck, R. Maric, D. Ghosh, J. Power Sources 167 (2007) 336e339. [11] M.C. Tucker, G.Y. Lau, C.P. Jacobson, L.C. DeJonghe, S.J. Visco, J. Power Sources 175 (2008) 447e451. [12] I. Villarreal, M. Rivas, L.M. Rodriguez-Martinez, L. Otaegi, A. Zabala, N. Gomez, M.A. Alvarez, I. Antepara, N. Arizmendiarrieta, J. Manzanedo, M. Olave, A. Urriolabeitia, N. Burgos, F. Castro, A. Laresgoiti, ECS Trans. 25 (2) (2009)
689e694. [13] T. Franco, M. Brandner, M. Ruttinger, G. Kunschert, A. Venskutonis, L.S. Sigl, ECS Trans. 25 (2) (2009) 681e688. [14] P. Szabo, J. Arnold, T. Franco, M. Gindrat, A. Refke, A. Zagst, A. Ansar, ECS Trans. 25 (2) (2009) 175e185. [15] M. Brandner, M. Bram, J. Froitzheim, H.P. Buchkremer, D. Stover, Solid State Ionics 179 (2008) 1501e1504. [16] N. Minh, T. Takahashi, Science and Technology of Ceramic Fuel Cells, Elsevier, Netherlands, 1995. [17] A. Atkinson, Solid State Ionics 95 (1997) 249e258. [18] O.A. Marina, N.L. Canfield, J.W. Stevenson, Solid State Ionics 149 (2002) 21e28. [19] M.R. Pillai, I.W. Kim, D.M. Bierschenk, S.A. Barnett, J. Power Sources 185 (2008) 1086e1093. [20] X. Huang, H. Zhao, W. Shen, W. Qiu, W. Wu, J. Phys. Chem. Solids 67 (2006) 2609e2613. [21] I. Villarreal, C. Jacobson, A. Leming, Y. Matus, S. Visco, L. De Jonghe, Electrochem. Solid State Lett. 6 (9) (2003) A178eA179. [22] A. Hagen, R. Barfod, P.V. Hendriksen, Y.-L. Liu, S. Ramousse, J. Electrochem. Soc. 153 (6) (2006) A1165eA1171. [23] M.C. Tucker, G.Y. Lau, C. Jacobson, L. De Jonghe, S.J. Visco, J. Power Sources 171 (2007) 477e482. [24] H.J. Cho, G.M. Choi, Solid State Ionics 180 (2009) 792e795. [25] M.C. Tucker, G.Y. Lau, C. Jacobson, S.J. Visco, L. De Jonghe, J. Power Sources 195 (10) (2010) 3119e3123.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Y0.08Sr0.88TiO3eCeO2 composite as a diffusion barrier layer for stainless-steel supported solid oxide fuel cell Kun Joong Kim, Sun Jae Kim, Gyeong Man Choi* Fuel Cell Research Center/Department of Materials Science and Engineering, Pohang University of Science and Technology (POSTECH), Pohang, 790-784, Republic of Korea
h i g h l i g h t s STS-supported SOFC with YSZ and Ni-YSZ is tape-casted and co-fired at 1350 C. Y0.08Sr0.88TiO3eCeO2 composite is tested as a new diffusion barrier layer (DBL). The cell with DBL shows peak power density ~220 mW cm2 and maintains it at 700 C.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 April 2015 Received in revised form 24 December 2015 Accepted 27 December 2015 Available online xxx
A new diffusion barrier layer (DBL) is proposed for solid oxide fuel cells (SOFCs) supported on stainlesssteel where DBL prevents inter-diffusion of atoms between anode and stainless steel (STS) support during fabrication and operation of STS-supported SOFCs. Half cells consisting of dense yttria-stabilized zirconia (YSZ) electrolyte, porous Ni-YSZ anode layer, and ferritic STS support, with or without Y0.08Sr0.88TiO3eCeO2 (YST-CeO2) composite DBL, are prepared by tape casting and co-firing at 1250 and 1350 C, respectively, in reducing (H2) atmosphere. The porous YST-CeO2 layer (t ~ 60 mm) blocks interdiffusion of Fe and Ni, and captures the evaporated Cr during cell fabrication (1350 C). The cell with DBL and La0.6Sr0.4Co0.2Fe0.8O3d (LSCF) cathode achieved a maximum power density of ~220 mW cm2 which is stable at 700 C. In order to further improve the power performance, Ni coarsening in anode during cofiring must be prevented or alternative anode which is resistive to coarsening is suggested. This study demonstrates that the new YST-CeO2 layer is a promising as a DBL for stainless-steel-supported SOFCs fabricated with co-firing process. © 2016 Elsevier B.V. All rights reserved.
Keywords: Inter-diffusion Co-firing Reducing atmosphere Degradation
1. Introduction Solid oxide fuel cells (SOFCs) have possible applications as energy conversion devices due to their high energy conversion efficiency and the diversity of fuels [1]. Recent developments in the electrodes and electrolytes have decreased the operating temperature to ~600 C, so metal (e.g., Ni)-ceramic supports can be replaced by metal supports such as stainless steel (STS) [2]. The metal can provide high mechanical strength and thermal-shock resistance for the brittle ceramics and thus strengthen the cell [3]. Thus, STS-supported SOFCs (STS-SOFCs) may possibly be used for mobile application that requires both low temperature
* Corresponding author. E-mail addresses: [email protected] (S.J. Kim), [email protected] (G.M. Choi).
(K.J.
http://dx.doi.org/10.1016/j.jpowsour.2015.12.130 0378-7753/© 2016 Elsevier B.V. All rights reserved.
Kim),
[email protected]
operation and fast cycling speed. Appropriate selection of a supporting metal is essential to achieve high power density and durability. In addition to mechanical strength and electrical conductivity, other requirements such as reactivity with other components and matching of thermal expansion coefficients (TECs) also must be considered because these are directly related to power density and durability of the cell. STS is a popular supporting metal for metal-supported SOFCs (MS-SOFCs); it shows good TECcompatibility (10e12 ppm K1) with electrolyte materials and ferritic STS without Ni is cheaper than many special metals. STSsupported SOFCs are often fabricated by co-firing of layers at high-temperature in a reducing atmosphere to avoid metal oxidation. However, inter-diffusion between Fe, Cr in ferritic-STS support and Ni from the anode can occur during fabrication and during cell operation [4e6]. When Ni from the anode diffuses into the STS support, it changes STS from ferrite to austenite, and this
386
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
transformation causes undesirable TEC change [7]. In addition, diffusion of Cr into the anode may cause oxide scales to form on the Ni particles; this passivated part would increase anodic polarization, leading to the poor power density and fast degradation of the cell [8]. Several attempts have been made to overcome these problems. Fabrication of cells at relatively low temperature has been studied using methods such as electrophoretic deposition (EPD) [9], pulsed laser deposition (PLD) [10] and plasma spray. These cells have shown relatively high power density without Ni contamination. However, fabrication at low temperature requires high-cost equipment and entails a complicated procedure. Furthermore, cells with a large area and various shapes cannot be easily fabricated by these methods. Other approach to address the problems is the infiltration techniques to prepare the anode [11]. However, Ni coarsening during operation degrades power density [11]. Lastly, insertion of a diffusion barrier layer (DBL) has been explored extensively as a solution to avoid reaction of Ni and Fe/Cr [12e16]. The barrier layer must prevent inter-diffusion while allowing electron and gas transport. The layer should also have similar TECs with the other cell components, and be stable and compatible with the relevant operating and processing conditions. Applying a DBL is one of the most promising methods in terms of stability and practical use. Compositions based on LaCrO3 and LaMnO3 [8,13,14]; Cr2O3/Cr2MnO4, CeO2, and Ce0.8Gd0.2O2 (GDC) [15] have been evaluated as diffusion barriers. LaMnO3-based composition is expected to be decomposed in the fuel atmosphere [16], and Cr2O3 based has shown Cr diffusion into the Ni-based anode [15]. CeO2 and GDC form an effective diffusion barrier. However, the cation of CeO2 or GDC is partially reduced from Ce4þ to Ce3þ in low oxygen partial-pressure (Po2) leading to lattice expansion that may result in the mechanical failure [17], particularly during firing at high temperature. To compensate for the mechanical stability of CeO2, an additional composition should be considered. Donor (Y or La)-doped SrTiO3, evaluated as an alternative anode material, could be a good additive material to remedy the shortcomings of CeO2. (Y, Sr) TiO3 (YST) or (La, Sr)TiO3 (LST) are chemically stable in a reducing atmosphere. The electrical conductivity of donor-doped SrTiO3 sintered in reducing atmosphere is 1e2 orders of magnitude higher than that sintered in air or CeO2 [18,19]. In addition, the donordoped SrTiO3 based materials have TEC (12 ppm K1) similar to those of the typical electrolyte materials. Thus, the addition of donor-doped SrTiO3 to CeO2 could yield a stable diffusion barrier. In this study, 8 mol% Y-doped SrTiO3 (Y0.08Sr0.88TiO3) was selected due to its highest conductivity among YSTs [20] to composite with 50 wt.% CeO2 and the Y0.08Sr0.88TiO3eCeO2 composite was tested as a DBL. After co-firing of STS-supported SOFCs with and without DBL at elevated temperature (1250e1350 C) in the reducing atmosphere, compositional changes across Ni-YSZ anode, DBL, and STS layers were examined. The effect of DBL on the electrochemical performance and stability was monitored at 700 C. 2. Experiment Button-type SOFCs with and without Y0.08Sr0.88TiO3eCeO2 (YSTCeO2) composite as a DBL were fabricated by the tape-casting method (Fig. 1). The cell without DBL, (a) Pt/YSZ/Ni-YSZ/STS, was fired at 1250 C and the cell with DBL, (b) La0.6Sr0.4Co0.2Fe0.8O3d (LSCF)/YSZ/Ni-YSZ/YST-CeO2/STS, was fired at 1350 C. The firing temperatures were selected for the proper densification of YSZ layer. For the DBL, YST was prepared from Y2O3, SrCO3 and TiO2 (99.9%, High Purity Chemicals, Japan). The mixed powders were milled with zirconia balls in ethanol (99.9%, SAMCHEN Chemicals, Korea) for 12 h then the ethanol was removed by evaporation. The
dried powder was calcined at 1200 C for 5 h in air. X-ray diffraction (XRD) measurement confirmed that the resultant powder is a single-phase YST. The synthesized YST and an equal weight of CeO2 (99.9%, High Purity Chemicals, Japan) with 15 wt.% of corn starch (D.C. Chemicals, Korea) as a pore former consist of DBL powder. To prepare the slurry for tape casting, powders were mixed with binder solution composed of toluene and ethanol as solvents, polyvinyl butyral (PVB, B-76) as a binder (10% of raw powders) and dioctyl phthalate (DOP) as a plasticizer (50% of binder contents). Then, the mixture of powders and binder solution was milled with zirconia balls (diameter; 10 mm and 5 mm) for 72 h. Green sheets of DBL were cast to a thickness of 40e50 mm after drying. Commercial stainless-steel powder (STS-434L, 400e500 mesh, Daekwang Industry, Korea) was chosen as a material for metal support. Planetary-milled (Pulverisette 6, Netzsch, Germany) NiO (1 mm, 99.97%, Kojundo chemical, Japan) powder and 8 mol% Y2O3-stabilized ZrO2 (YSZ, 0.3 mm, TZ-8YS, Tosoh, Japan) powder were mixed in 60:40 wt.% for anode slurries. YSZ slurries for electrolyte were prepared in a similar method to form green sheets. As-cast green sheets of support, DBL, anode, and electrolyte were used for the cells with or without DBL. Both green cells with and without DBL layer were laminated at 30 MPa and 60 C for 20 min to achieve the desired thickness, then punched out to yield a circular green-cell of 24-mm diameter. The green cells were heated to burn binder at or below 400 C for 12 h and co-fired at 1350 C for the cell with DBL and 1250 C for the cell without DBL, respectively, for 3 h in dry hydrogen atmosphere. To avoid deformation of the sample during the firing process, a porous ZrO2 plate was used to apply a vertical load (1.5 g cm2) to the cell. The shrinkage curves of YSZ, STS434L, Ni-YSZ and YST-CeO2 tapes (50 5 mm2 size) were determined simultaneously by positioning and firing all the tapes along the temperature gradient of a tube furnace (754e1277 C) in dry H2. The temperature of each position was measured using an R-type thermocouple. The heating rate was 2 C/min and the cooling rate was 5 C/min. For the electrochemical measurement of cells, ceramic bond (Model 571, Aremco, USA) was used to mount the sintered cell on an alumina tube. Platinum (#6082, Heraeus, Germany) and La0.6Sr0.4Co0.2Fe0.8O3d (LSCF, AGC Seimi Chemical Co., Japan) pastes, respectively, were screen-printed as a cathode (0.502 cm2) for the cell with and without DBL, respectively. Although Pt cathode is stable and easy to prepare, it is expected to show high polarization resistance. A mixed ionic and electronic conductor (MIEC) cathode (LSCF) was selected for the cell with DBL to obtain a better performance. For firing of the LSCF cathode, the cell was pre-fired at 900 C for 2 h while maintaining the reducing atmosphere in the anode and open air in the cathode. Pt was used as a cathode for the cell without DBL and was not pre-fired. After pre-firing of the cell with LSCF cathode, the cell was cooled to room temperature, then Pt paste and mesh were attached on the cathode as a current collector. The electrochemical performances of both cells were evaluated using wet H2 gas as a fuel gas and open air as an oxidant gas. The expected (or calculated) Po2 in the anode gas was ~1022 atm at 800 C. The microstructure of cross-section of the cell was examined using a field emission scanning electron microscope (FE-SEM, Model XL30S FEG, Philips Electron Optics B.V., Netherlands). Atomic (Ni, Fe, Cr) distributions across anode, DBL, and STS support layers were examined using an energy dispersive spectroscopy (EDS) line profile to examine the possible reaction between the layers. The current-voltage-power (IeVeP) curve and the impedance were measured using an AC impedance analyzer (VSP, Bio Logic Science instruments, France) using wet H2 (97% H2 þ 3% H2O) as a fuel gas and open air as an oxidant gas. Current-voltage-power measurements were conducted at 700 C for 40 h.
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
387
Fig. 1. Schematics of two STS-supported SOFCs; (a) Pt/YSZ/Ni-YSZ/STS cell (without DBL, sintered at 1250 C) (b) LSCF/YSZ/Ni-YSZ/YST-CeO2/STS cell (with DBL, sintered at 1350 C).
3. Results and discussion We used different firing temperatures for two cells. To obtain pore-free YSZ electrolyte during co-firing, the required minimum temperatures were 1250 C for the cell without DBL and 1350 C for the cell with DBL. This 100 C difference in firing temperature may be due to the poor sinterability of DBL. Linear shrinkage curves for STS support, YST-CeO2 DBL, Ni-YSZ anode, and YSZ electrolyte were obtained and compared at temperatures between 754 and 1277 C (Fig. 2). The initial shrinkage values (1e4%) at ~740 C differ due to NiO reduction in Ni-YSZ, burn-out of pore former (15 wt.%) in YSTCeO2 layer, in addition to the burning of binder. At the highest firing temperature (1277 C), the shrinkage of DBL is approximately onehalf of YSZ. This means that DBL in the cell may impede overall shrinkage of the cell, especially the densification of electrolyte. The poor sintering or shrinkage is the nature of inert composite where YST or CeO2 has limited solubility each other and thus impede the shrinkage of counterpart particles. Ni-YSZ composite and STS show the similar shrinkage values that are intermediate between those of YSZ and DBL. We also fabricated the cell without DBL at 1350 C, however, Ni-YSZ anode after firing in dry H2 gas showed very dense microstructure which is not suitable for use in an electrochemical cell and was excluded in the discussion. The diffusion profiles of both of as-sintered cells were analyzed to determine the effect of YST-CeO2 DBL on the diffusion during fabrication process. In the cell without DBL (Fig. 3a), both Fe and Cr
Shrinkage [%]
0 5 10 Fig. 3. EDS line profile of fractured cross-section of the cells (a) without DBL (co-fired at 1250 C) and (b) with DBL (co-fired at 1350 C).
15 20 25 700
YSZ Ni-YSZ YST-CeO2 STS
800
900
1000
1100
1200
1300
Temperature [°C] Fig. 2. Shrinkage curves for YSZ, STS434L, Ni-YSZ and YST-CeO2. Green tapes of 50 5 mm2 size for each composition were sintered simultaneously using tube furnace with temperature gradient. Gas: dry hydrogen, temperature range: 754e1277 C. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
of STS diffused into the Ni-YSZ anode during co-firing at 1250 C as previously reported [8,21]. In the cell with DBL, diffusion of Fe and Cr was prevented during co-firing at 1350 C (Fig. 3b). Although Cr was found in the DBL, little Cr was found in the anode layer. During high-temperature fabrication, diffusion of Fe was completely prevented from STS to DBL since no Fe was found in DBL layer. In contrast, Cr may have diffused or evaporated and may have been captured by the DBL layer. Thus YST-CeO2 seemed to function well as a DBL during co-firing. The time-dependent changes of impedance spectra of both cells
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
6
0h 40 h
(a) w/o DBL cell
4 2 0 0
15
20
(b) w/ DBL cell
42 0h 40 h
0.6
28 0.3
14
700 °C Air | 97 % H + 3 % H O 0
40
80
120
160
-2
2
10
56
0.9
0.0
1.5
0 200
1.0
(b) w/ DBL cell
0.0 -0.5
240
1.2
0.5
0
1
2
3
4
2
Zreal [Ω cm ] Fig. 4. Impedance spectra of cells (a) without and (b) with DBL during operation for 40 h at 700 C. 97% H2 þ 3% H2O mixture was used as fuel and air as oxidant gases. Total resistance of cell without DBL increased severely with time.
0.9
180
0.6
120
0.3
60
-2
Power density [mW cm ]
-Zimag [Ω cm ]
5
70
1.2
700 °C Air | 97 % H2 + 3 % H2O
Cell voltage [V]
(a) w/o DBL cell
Cell voltage [V]
2
8
Power density [mW cm ]
-Zimag [Ω cm ]
388
were obtained to see the effect of DBL during operation at 700 C. The impedance spectra of the cell without DBL increased severely in size with time (Fig. 4a), but those of the cell with DBL showed little change during operation for 40 h (Fig. 4b). The cell with DBL shows much lower Ohmic and polarization area-specific resistance (ASR) values than the cell without DBL, both initially and after 40 h operation (Table 1). At initial operation, the higher Ohmic ASR of the cell without DBL (1.1 U cm2) than the cell with DBL (0.19 U cm2) may be caused by a reaction layer such as a resistive NieCr oxide phase formed between the Ni-YSZ anode and the STS. The polarization ASR difference (3.5 versus 6.4 U cm2 at 0 h) mainly originated from the difference in cathode materials (Pt for the cell without DBL vs. LSCF for the cell with DBL), inter-diffusion, and anode microstructure. Although the origin of strong increase of polarization ASR over time for the cell without DBL is not clear, Fe/ Cr contamination (especially Cr evaporation) at the anode is strongly suspected because the microstructure of Pt cathode and Ni-YSZ anode might be relatively stable at 700 C. To support this, we have fabricated an electrolyte (YSZ)esupported cell with symmetrical Pt cathode and observed impedance change with time at 700 C in air; the total polarization ASR increased from 1.2 to 1.6 U cm2 for 40 h. Meanwhile, the polarization ASR for the present cell without DBL increased from 6.4 to 18.4 U cm2 for the same time. Thus degradation due to Pt cathode is not significant and it is mostly due to anode. The stable polarization ASR of the cell with DBL suggests that Cr captured by the YST-CeO2 layer degrades little
0.0
0
200
400
600
0 800
-2
Current density [mA cm ] Fig. 5. Current-voltage and currentepower curves of cells (a) without DBL and (b) with DBL tested at 700 C for 40 h. 97% H2 þ 3% H2O mixture was used as fuel and air as oxidant gases.
the cell performance at 700 C for 40 h. Therefore, we conclude YSTCeO2 works as a diffusion barrier layer for at least 40 h of operation after high-temperature firing. We also measured currentevoltage (IeV) curves of each cell type at 700 C, and observed that both cells exhibited OCV value of z1 V (Table 1 and Fig. 5). However, the power density differed significantly between two cells. The cell without the DBL achieved a maximum power density (MPD) of only ~62 mW cm2 (Table 1; Fig. 5a) and the MPD degraded rapidly as expected. Meanwhile, the cell with DBL achieved a MPD of ~220 mW cm2 (Table 1; Fig. 5b). The reason for the enhanced performance and durability for the cell with DBL is explained by lower ASRs compared to the cell without DBL as discussed in impedance analysis: lower Ohmic ASR due to little resistive phase, lower polarization ASR due to MIEC cathode and the anode free from Cr contamination. Although the total ASR of the cell with DBL is nearly constant for 40 h in OCV condition, a slight decrease in MPD may be due to the degradation of cathodic performance [22] under current in addition to the slight decrease in OCV from 1.09 to 1.08 V (Table 1). Although applying DBL enhanced
Table 1 ASRs and power density of the cells with and without DBL. Types
Time [h]
Ohmic ASR [U cm2]
Polarization ASR [U cm2]
OCV [V]
Max power density [mW cm2]
Cell without DBL
0 40 0 40
1.1 0.61 0.19 0.21
6.4 18.4 3.5 3.5
1.12 1.12 1.09 1.08
62 35 220 203
Cell with DBL
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
389
Fig. 6. S.E.M. images of the cell with DBL: (a) YSZ electrolyte/Ni-YSZ anode, (b) YST-CeO2 DBL, (c) LSCF cathode/YSZ electrolyte. Ni-YSZ anode was co-fired at 1350 C in reducing atmosphere.
both the performance and stability of the cell, the total ASRs for the cell are still limited by the large polarization resistance (Table 1; Fig. 4b); the polarization ASR to total ASR ratio is ~94% and an arc at low frequency is much larger than that at high frequency. The origin of large ASR is not clear at present. However, a large part of it may come from cathode since oxygen reduction reaction (ORR) is often becomes slow with decreasing temperature. The contribution of anodic ASR also cannot be entirely neglected since the polarization is strongly dependent upon the change of H2 content in anode gas (not shown here) [23]. To confirm the performance of electrodes, microstructures of both anode and cathode were examined after test for the cell with DBL (Fig. 6); the microstructure of (a) Ni-YSZ anode was less porous than that of either (b) YST-CeO2 DBL or (c) LSCF cathode. Ni particles of Ni-YSZ anode may have coarsened during firing in reducing atmosphere at high temperature after reduction from NiO and the coarsening may have contributed to the large polarization ASR shown in Fig. 4b [24]. Thus, for further improvement of anodic performance, the anode microstructure must be controlled. Use of a 65-mm-thick YST-CeO2 layer have provided stable electrochemical performance, which means that the layer successfully prevents inter-diffusion of Fe/Cr and Ni between STS support and Ni-based anode during cell processing and operation. Compared to the reported DBL compositions such as Cr2O3-based composition [15] and Cu-YSZ cermet [25], the YST-CeO2 shows better chemical stability in a reducing atmosphere and better diffusion-barrier ability. Cr2O3-based composition has shown Cr diffusion into Ni-based anode after high-temperature co-firing in reducing atmosphere. A Cu-YSZ cermet DBL was found to block Ni diffusion into the support, but not able to block Fe and Cr diffusion into the anode. The cell with CeO2 DBL showed limited lifetime (~165 h) due to the start of breakaway oxidation of the metallic substrate [15]. Meanwhile cell with LaCrO3-based and LaMnO3based [8,13,14] DBL showed good stability for long operation time (degradation rate ~1% 1000 h1 at 800 C), however, the stability of these cells is still questionable when they are co-fired at high temperature. Our results demonstrated the stability of cell during
co-firing at 1350 C as well as during operation at 700 C for 40 h. Long-term durability (>500 h) will be further tested in the future for the reliability of the STS-supported SOFCs with YST-CeO2 layer. In terms of the cell performance, the dense microstructure in anode becomes the limiting factor. The use of alternative anode which is resistive to agglomeration is suggested.
4. Conclusions Y0.08Sr0.88TiO3eCeO2 (YST-CeO2) composite was used as diffusion barrier layer (DBL) to prevent mutual diffusion of Fe and Cr in stainless-steel (STS) and Ni in anode during co-firing and operation of STS-supported solid oxide fuel cells (SOFCs). SOFCs with and without DBL were prepared using tape-casting, screen-printing, and subsequent co-firing at elevated temperature in reducing atmosphere. After co-firing of both cells, atomic distribution across Ni-YSZ anode, YST-CeO2 DBL, and STS support was examined using an energy dispersive spectroscopy (EDS) line profile to confirm the effect of DBL on the inter-diffusion. The power density and impedance spectra for both cells at the initial stage (0 h) were compared as well as after 40 h of operation. A porous YST-CeO2 layer effectively blocked inter-diffusion of Fe and Ni, and captured the evaporated Cr during cell fabrication (1350 C) and operation. To improve further the electrochemical performance of the cell, Ni coarsening in anode during co-firing must be prevented or alternative anode which is resistive to coarsening is suggested. This study demonstrates that the new YST-CeO2 layer could be a promising as DBL for stainless-steel-supported SOFCs.
Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (grant no. 2011-0023389).
390
K.J. Kim et al. / Journal of Power Sources 307 (2016) 385e390
References [1] S.C. Singhal, Solid State Ionics 135 (2000) 305e313. [2] N. Christiansen, J.B. Hansen, H. Holm-Larsen, S. Linderoth, P.H. Larsen, P.V. Hendriksen, M. Mogensen, in: 7th European SOFC Forum, 2006, p. B034. [3] Y. Matus, L. De Jonghe, C. Jacobson, S. Visco, Solid State Ionics 176 (2005) 443e449. [4] S. Visco, C. Jacobson, L. De Jonghe, in: SECA Workshop, 2003. [5] M. Brandner, in: Solid Oxide Fuel Cells IX (SOFC-IX), 2005, p. 1235. [6] D. Hathiramani, in: e-Proceedings ICACC, Cocoa Beach, Florida, USA, 2006. [7] Z. Yang, D. Paxton, K. Weil, J. Stevenson, P. Singh, Materials Properties Database for Selection of High-temperature Alloys and Concepts of Alloy Design for SOFC Applications, PNNL-14116. Prepared for the U.S. Department of Energy under Contract DE-AC06-76RL01830, 2002. [8] T. Franco, K. Schibinger, Z. Ilhan, G. Schiller, A. Venskutonis, ECS Trans. 7 (1) (2007) 771e780. [9] P. Banace, N.P. Brandon, B. Girvan, P. Holbeche, S. O'Dea, B.C.H. Steele, J. Power Sources 131 (2004) 86e90. [10] S. Hui, D. Yang, Z. Wang, S. Yick, C. Deces-Petit, W. Qu, A. Tuck, R. Maric, D. Ghosh, J. Power Sources 167 (2007) 336e339. [11] M.C. Tucker, G.Y. Lau, C.P. Jacobson, L.C. DeJonghe, S.J. Visco, J. Power Sources 175 (2008) 447e451. [12] I. Villarreal, M. Rivas, L.M. Rodriguez-Martinez, L. Otaegi, A. Zabala, N. Gomez, M.A. Alvarez, I. Antepara, N. Arizmendiarrieta, J. Manzanedo, M. Olave, A. Urriolabeitia, N. Burgos, F. Castro, A. Laresgoiti, ECS Trans. 25 (2) (2009)
689e694. [13] T. Franco, M. Brandner, M. Ruttinger, G. Kunschert, A. Venskutonis, L.S. Sigl, ECS Trans. 25 (2) (2009) 681e688. [14] P. Szabo, J. Arnold, T. Franco, M. Gindrat, A. Refke, A. Zagst, A. Ansar, ECS Trans. 25 (2) (2009) 175e185. [15] M. Brandner, M. Bram, J. Froitzheim, H.P. Buchkremer, D. Stover, Solid State Ionics 179 (2008) 1501e1504. [16] N. Minh, T. Takahashi, Science and Technology of Ceramic Fuel Cells, Elsevier, Netherlands, 1995. [17] A. Atkinson, Solid State Ionics 95 (1997) 249e258. [18] O.A. Marina, N.L. Canfield, J.W. Stevenson, Solid State Ionics 149 (2002) 21e28. [19] M.R. Pillai, I.W. Kim, D.M. Bierschenk, S.A. Barnett, J. Power Sources 185 (2008) 1086e1093. [20] X. Huang, H. Zhao, W. Shen, W. Qiu, W. Wu, J. Phys. Chem. Solids 67 (2006) 2609e2613. [21] I. Villarreal, C. Jacobson, A. Leming, Y. Matus, S. Visco, L. De Jonghe, Electrochem. Solid State Lett. 6 (9) (2003) A178eA179. [22] A. Hagen, R. Barfod, P.V. Hendriksen, Y.-L. Liu, S. Ramousse, J. Electrochem. Soc. 153 (6) (2006) A1165eA1171. [23] M.C. Tucker, G.Y. Lau, C. Jacobson, L. De Jonghe, S.J. Visco, J. Power Sources 171 (2007) 477e482. [24] H.J. Cho, G.M. Choi, Solid State Ionics 180 (2009) 792e795. [25] M.C. Tucker, G.Y. Lau, C. Jacobson, S.J. Visco, L. De Jonghe, J. Power Sources 195 (10) (2010) 3119e3123.