Degradation study of yttria-doped barium cerate (BCY) electrolyte in protonic ceramic fuel cells under various operating conditions

Transcription

1 Journal of the Ceramic Society of Japan 123 [4] Paper Degradation study of yttria-doped barium cerate (BCY) electrolyte in protonic ceramic fuel cells under various operating conditions Mi Young PARK, Young Jin KIM and Hyung-Tae LIM ³ School of Materials Science and Engineering, Changwon National University, 20 Changwondaehak-ro, Changwon, Gyeongnam , Republic of Korea Yttria doped barium cerate (BCY) electrolyte, Ni+BCY anode supported cells were fabricated, and their stability and the mechanism of their degradation were investigated through constant current tests under various operating conditions, especially negative cell voltage operation (with respect to degradation phenomenon due to cell imbalance in a series connected stack). The results of electrochemical tests (I V characteristics and impedance spectra) indicate that the degradation rate was significant when the cell was operated under higher current densities (regardless of the sign of cell voltage) and only the ambient air was used for the cathode. Post-material analyses revealed that microstructural and compositional changes were obvious in the BCY electrolyte and the BCY of the cathode functional layer because of BCY decomposition in the wet atmosphere at the cathode. Thus, the present work concludes that the degradation rate of BCY electrolyte-based cell depends on operating conditions, i.e., the amount of current density (the water vapor production rate at the cathode) and the air flow rate (flushing water vapor at the cathode) The Ceramic Society of Japan. All rights reserved. Key-words : Proton conductor, Protonic ceramic fuel cells, Degradation [Received January 30, 2015; Accepted February 18, 2015] 1. Introduction Typical solid-oxide fuel cells (SOFCs) are operated at temperatures above 800 C to obtain sufficient oxygen ion conduction of the yttria stabilized zirconia (YSZ) electrolyte, and this high operating temperature leads to advantages, such as high efficiency and fuel flexibility. 1),2) On the other hand, the lifetime of a high-temperature SOFC stack is limited by thermal degradation, such as the coarsening of particles in the electrodes and chemical interactions between the electrolyte and electrodes, due to high-temperature operation. 1) Additionally, when an oxygen ion conducting electrolyte is employed in a fuel cell, water vapor is formed at the anode side, resulting in the dilution of the fuel. For these reasons, numerous studies have been conducted to reduce the operating temperature of SOFCs, i.e., to increase long-term stability; therefore, SOFCs based on proton conducting electrolytes (protonic ceramic fuel cells, PCFCs) have attracted considerable attention in recent years. 3) 8) Among many proton conducting solids, such as perovskite-type oxides, yttria-doped barium cerates (BCY) and barium zirconate (BZY), have been extensively studied due to their high ion (proton) conductivity at lower temperatures (<600 C). 6) 8) These materials are known to exhibit mixed proton, oxygen ion and electron/hole conduction depending upon the operating temperature and atmosphere. 6) Barium cerates are known to be the highest proton conductor, but their poor chemical stability in atmosphere containing water vapor and carbon dioxide limits the application of these materials ³ Corresponding author: H.-T. Lim; htaelim@changwon. ac.kr Preface for this article: DOI The Ceramic Society of Japan DOI for electrolyte in PCFCs. 9),10) The corresponding reactions with water vapor (H 2 O) and carbon dioxide (CO 2 ) can be written as BaCeO 3 þ CO 2! BaCO 3 þ CeO 2 ; ð1þ BaCeO 3 þ H 2 O! BaðOHÞ 2 þ CeO 2 : ð2þ Many research groups have carried out various material analyses using X-ray diffraction (XRD), SEM (scanning electron microscopy), TEM (transmission electron microscopy), etc. on doped BaCeO 3 to detect possible changes in phase, composition, and microstructure after aging in H 2 O/CO 2 -rich atmosphere as a function of dopants and their composition. 11) 17) It is noted that most of this research has been conducted on BCY powder materials or specimens themselves, not on the BCY electrolyte of fuel cells under actual fuel cell operating conditions. Obvious correlation between real fuel cell testing conditions and accelerating degradation processes has not been determined. 9) Thus, BCY electrolyte demands further chemical stability studies at the level of practical application. Meanwhile, efforts have been made to improve the chemical stability of doped BaCeO 3 as an electrolyte. Studies have been made in the following areas: 1) synthesis of yttria-doped BaCeO 3 BaZrO 3 solid solution (BCZY) to achieve both the advantages of the high ion conductivity of BCY and the high chemical stability of BZY, 17) 19) 2) design of dual and multi-layers consisting of a BZY layer to protect the BCY layer from H 2 O and CO 2, 20) 24) and 3) co-doping of BaCeO3 with M(M= Ta, Ti, Nb, Sn and In) and Y. 25) 27) Guo et al. and Bae et al. proved the good stability of BCY-BCZY dual-layer cells with a (Ba, Sr)(Co, Fe) oxide cathode and a BCY/BZY/BCY multi-layer cell with a Pt cathode under fuel cell operating conditions over 10 h. 21),23) Recently, it was reported that a pure BCY single-layer cell showed no degradation for 1,000 h of long-term operation with a very low current density ( A/cm 2 ) and 257

2 Park et al.: Degradation study of yttria-doped barium cerate (BCY) electrolyte in protonic ceramic fuel cells under various operating conditions an unconventional cathode, Nd2NiO4+d. at 600 C.9) On the other hand, Zhao et al. observed significant degradation of a BCY electrolyte-based cell under a constant voltage output of 0.5 V at 700 C,25) which implies that the stability of BCY electrolytebased cells depends on various factors related to cell components and operating conditions (such as current density and temperature). If water vapour formation at the cathode side is one of the critical factors contributing to BCY electrolyte degradation, the stability of BCY electrolyte-based cells should be strongly dependent on the kind of electrolyte material used (especially close to the cathode side) and the air flow rate depressing the water vapour pressure. In the present study, BCY electrolytebased cells were fabricated, and their stability and the mechanism of their degradation were investigated, especially under severe operating conditions, such as a negative voltage and low air flow rate. One of the critical SOFC stack degradation phenomena is abnormal operation under a negative voltage due to cell imbalance (cell-to-cell variation) in a series-connected stack, causing rapid degradation due to anode/ysz electrolyte interface delamination.28) In previous studies, the degradation mechanism of YSZ electrolyte-based cells was studied under a negative voltage condition.29) 32) In the present work, constant current (CC) tests were conducted on BCY electrolyte-based cells from a positive to a negative voltage as a function of air flow rate. Before and after the CC tests, the I V curves and impedance spectra were measured and compared to trace any increase in ohmic and/or non-ohmic resistance of the cells. Post-material analyses were conducted using X-ray diffraction (XRD) and scanning electron microscopy (SEM) combined with energy dispersive spectroscopy (EDS) to identify any change in microstructure and phase after operation under severe conditions. 2. Experimental Procedure anode functional and support layers was reduced using a gas mixture containing 10% H2 with the balance nitrogen for 2 h at 600 C. The anode gas was then changed to 100% H2 while air was supplied to the cathode side, both at the flow rate of 200 ml/min. After the cell voltage reached ³1 V, the initial I V characteristics and impedance spectra were measured, and then constant current (CC) tests were conducted successively from a positive to a negative cell voltage using a Bio-Logic SP 240 potentiostat/galvanostat. Impedance measurements were taken within the frequency range of 200 khz 0.1 Hz with a signal amplitude of 10 mv under an open circuit. The duration of each CC test was about 2 h. Between CC tests, the cell was returned to the open-circuit condition, and then the I V characteristics and impedance spectra were measured at the hydrogen/air flow rate of 200 ml/min.to trace any change in electrochemical performance during the CC tests. This test procedure was repeated on the new ones without air being supplied from a gas tank, which means that only the ambient air was used for the cathode, to investigate the effect of the air flow rate on cell stability. Before and after the tests, material analyses were conducted using XRD and SEM with EDS on the electrolyte and the cathode functional layers to trace any change in the phase and microstructure of the cells. 3. Resutls and Discussion Figures 2(a) and 2(b) show SEM micrographs of the cross section of the BCY electrolyte and the cathode functional layer (BCY+LSCF), respectively, before the tests. The electrolyte appears to be quite dense with negligible porosity. EDS analyses were also conducted, and the atomic compositions of the electrolyte and the cathode functional layers were obtained. The theoretical atomic ratio of Ce to Ba in BCY 15 is The measured value, the ratio of Ce to Ba, in the BCY single layer is A detailed description of the fabrication of BCY (BaCe0.85Y0.15O2.925, 8.5 m2/g, Kceracell)-electrolyte, Ni + BCY anodesupported cells can be found elsewhere.33) The cathode functional layer and the cathode current collector consist of BCY + LSCF (La0.6Sr0.4Co0.2Fe0.8O3-d, FCM, 4 8 m2/g) composite and LSFC, respectively. A schematic of anode-supported cell (Ni-BCY/ BCY/LSCF-BCY) is shown in Fig. 1. The cell diameter and the active area (the cathode area) is ³2.3 cm and ³1.3 cm2, respectively. Electrochemical tests were carried out as follows. Before hydrogen fuel was supplied to the anode side, NiO in the SEM images of the cross section of the BCY electrolyte (a) and the cathode functional layer (BCY+LSCF) (b) before the tests, indicating the atomic ratio of Ce to Ba. Fig. 2. Fig. 1. BCY). 258 A schematic of anode-supported cell (Ni-BCY/BCY/LSCF-

3 Journal of the Ceramic Society of Japan 123 [4] (a) (c) (b) Fig. 3. (a) Voltage plot vs. time at 0.56 A/cm 2 (positive voltage condition) with hydrogen/air flow rate of 200 ml/min; (b) Voltage plot vs. time at 0.91 A/cm 2 (negative voltage condition) with hydrogen/air flow rate of 200 ml/min.; (c) Voltage current density and power density measured before and after the CC tests; (d) Impedance spectra measured before and after the CC tests. (d) about the same as the theoretical value for the cathode functional layer and the electrolyte. Figures 3(a) and 3(b) show the cell voltage plot vs. time during the first CC test at 0.56 A/cm 2 (³0.7 V) and the second CC test at 0.91 A/cm 2 (³¹0.2 V), respectively, at the hydrogen/ air flow rate of 200 ml/min. A significant voltage drop was observed only during the second CC test. Between CC tests (after the reduction and after the first and the second CC test for 2 h), power tests were carried out as shown in Fig. 3(c). The value of the initial maximum power density (³0.22 W/cm 2 ) of the anodesupported BCY cell at 600 C is acceptable in comparison with other research groups reports. 22),34) After the CC test at ³0.70 V, the power density at the higher current region increased, and this improved performance may be attributed to the further reduction of NiO in the anode, judging from the slope change in the I V curve in the higher current region. On the other hand, the power density was significantly reduced after negative voltage operation at ³0.91 A/cm 2. Moreover, the open-circuit voltage (OCV) decreased to ³0.9 V. Note that any gas crossover evidences due to cell cracking and/or sealant damages were not be detected after the test. Thus, it can be said that this OCV drop is attributed to some chemical changes occurred in the cell during the severe operation. Impedance spectra corresponding to before the test, after the CC test at a positive voltage and after the CC test at a negative voltage are shown in Fig. 2(d). The semicircle at the low frequencies dominated by gas diffusion (agreed with the power test result) decreased after the low-current test at a positive voltage, while the high-frequency intercept of the semicircle on the real axis shifted, and the semicircle size increased after the high-current test at a negative voltage, indicating increment in both ohmic and non-ohmic resistance. A similar CC test protocol was repeated on a BCY single-layer cell without air supply. As observed with an air flow rate of 200 ml/min, the cell voltage slightly increased during the lower current density (0.21 A/cm 2 ) CC test at a positive voltage (³0.7 V), while it dramatically decreased during the high current density (0.60 A/cm 2 ) CC test at a negative voltage (³¹0.2 V), as shown in Figs. 4(a) and 4(b), respectively. The improvement and degradation in electrochemical performance were reflected in the I V curve and impedance spectra obtained after each CC test for 2 h, as shown in Figs. 4(c) and 4(d), respectively. As compared with the air flow rate of ³200 ml/min, the cell degradation was more serious, indicating the dependency of the BCY cell stability on the air supply rate (for flushing water vapor) as well as the amount of current density. Jung et al. investigated the transport properties of BCY electrolyte in an anode-supported cell using embedded reference electrodes. It was found that BCY electrolyte close to electrolyte/ electrode interfaces exhibits non-negligible electronic conduction while the electronic resistance is very high in the middle region. 33) It was reported that some electronic conduction in solid electrolyte is beneficial to cell stability with regard to abnormal operation under a negative voltage in a series-connected fuel cell stack. 30) 32) In the case of pure ionic conducting materials, such as YSZ, the cell showed significant degradation after negative voltage operation with the observation of anode delamination due to high internal partial pressure in the electrolyte close to the anode side. 29) From this point of view, BCY cells were expected to be resistant to degradation caused by negative voltage operation (in fact, any delamination was observed after the tests in the present work). However, their performance was significantly degraded after abnormal operation (negative cell voltage conditions). Thus, a different degradation mechanism should be taken into account for BCY cells. Figures 5(a) and 5(b) show SEM micrographs including EDS results from the BCY electrolyte and the cathode functional layer, respectively, after the CC tests with positive and negative cell voltages with the air flow rate of ³200 ml/min. Compared with those before the test, the atomic ratio of Ce to Ba increased to in the electrolyte region, especially the upper part close to the 259

4 Park et al.: Degradation study of yttria-doped barium cerate (BCY) electrolyte in protonic ceramic fuel cells under various operating conditions (a) (c) (b) (d) (a) Voltage plot vs. time at 0.21 A/cm2 (positive voltage condition) with hydrogen flow rate of 200 ml/min and without air supply; (b) Voltage plot vs. time at 0.60 A/cm2 (negative voltage condition) with hydrogen flow rate of 200 ml/min and without air supply; (c) Voltage current density and power density measured before and after the CC tests; (d) Impedance spectra measured before and after the CC tests. Fig. 4. Fig. 5. SEM images of the cross section of the BCY electrolyte (a) and the cathode functional layer (BCY+LSCF) (b) after the test with air supply (Fig. 3), indicating the atomic ratio of Ce to Ba. 260 cathode side producing water vapor, while that in the cathode functional layer decreased to Note that the reported oxygen ion transport number of BCY 15 electrolyte in a fuel cell is ³0.1 at 600 C,35) and this value is much lower than proton transport number ³0.9, implying that the influence of water vapor on the cell is dominant at the cathode side, not the anode side. Similar microstructural and compositional changes were observed without air supply as shown in Figs. 6(a) and 6(b). In Figs. 7(a) 7(c), XRD patterns of the cathode functional layer are compared among before (as fired at 1000 C) after the CC test with the air flow rate of ³200 ml/min and after the CC test without air supply, obtained from the cathode functional layer surface of each case. It is noted that some of BCY cells were fabricated and tested without a cathode current collector (LSCF) for the purpose of this analysis. The initial XRD patterns simply correspond to the combination of BCY and LSCF XRD patterns [Fig. 7(a)]. However, additional peaks (unknown phases) were detected in the cases of after the test [Figs. 7(b) and 7(c)]. Especially after the operation without air supply, the intensities of the BCY and LSCF peaks were reduced, and some of the BCY peaks disappeared. As mentioned in the introduction, BCY can be decomposed to CeO2 and Ba(OH)2 in a water-containing atmosphere. As the current density increased, water vapor pressure increased, leading to BCY decomposition in the electrolyte as well as the cathode region. Thus, based on the EDS and XRD results, it can be said that the Ba ions of Ba(OH)2 from the decomposed BCY diffuse and react with LSCF, resulting in the formation of unknown phases (oxides) including Ba, Fe, Sr, etc. Note that LSCF is known to be chemically unstable and decomposed, under certain operating conditions (such as water vapor vol % and working temperature).36) Thus, the BCY+LSCF

5 Journal of the Ceramic Society of Japan 123 [4] composite cathode degradation process should accelerate more as the current density increases, i.e., water vapor pressure increases. CC tests were conducted on the BCY cells at relatively higher current density, but under a positive voltage condition, with hydrogen and air both at the flow rate of 200 ml/min. In the first case, the cell voltage vs. time plot at 0.75 and 0.60 A/cm2 is shown in Fig. 8(a), and the power test and impedance results (obtained before and after the CC test) are shown in Figs. 8(b) and 8(c), respectively. Because the voltage rapidly dropped under 0.75 A/cm2, the current was reduced to 0.60 A/cm2. The I V curves and impedance spectra indicate that the degradation aspects are similar to those in the previous cases. Based on these SEM images of the cross section of the BCY electrolyte (a) and the cathode functional layer (BCY+LSCF) (b) after the test without air supply (Fig. 4), indicating the atomic ratio of Ce to Ba. Fig. 6. results, we can say that the degradation of BCY cell is related to the amount of current density, not the sign of cell voltage. After the tests, the electrolyte and the cathode functional layer were analyzed by SEM and EDS, and the corresponding microstructure and atomic ratio (Ce/Ba) are shown in Figs. 9(a) and 9(b). As Fig. 8. (a) Voltage plot vs. time at 0.75 and 0.60 A/cm2 (positive voltage condition) with hydrogen/air flow rate of 200 ml/min; (b) Voltage current density and power density measured before and after the CC tests; (d) Impedance spectra measured before and after the CC tests. XRD patterns of the cathode functional layer surface: (a) before the tests, (b) after the test with air supply (Fig. 3) and (c) after the test without air supply (Fig. 4). Fig

6 Park et al.: Degradation study of yttria-doped barium cerate (BCY) electrolyte in protonic ceramic fuel cells under various operating conditions Fig. 9. SEM images of the cross section of the BCY electrolyte (a) and the cathode functional layer (BCY+LSCF) (b) after the test at a positive voltage (Fig. 8), indicating the atomic ratio of Ce to Ba. seen in the figures, Ba is deficient in the electrolyte, while it is rich in the cathode functional layer, as observed in the previous tests. Thus, to prevent the degradation of BCY electrolyte-based cells due to severe operation conditions, such as negative cell voltage, the use of materials that are durable against a wet atmosphere (for example, barium zirconate) is suggested as a protective layer between the cathode functional layer and the BCY electrolyte and also in the cathode functional layer. A degradation study on this kind of fuel cell constitution will be reported in a future paper. 4. Conclusion In the present work, anode-supported cells (Ni-BCY/BCY/ LSCF-BCY) were fabricated and CC tested from positive to negative cell voltage to investigate the relation between degradation rate and cell components/operating conditions. The BCY cells showed a significant degradation rate during CC tests with increasing current density and decreasing air flow rate at the cathode (regardless of the sign of cell voltage). Post material analyses revealed that Ba became deficient in the electrolyte while it became rich in the cathode functional layer with the formation of secondary phases after operation. Therefore, the present work concludes that the degradation rate of BCY-based cell depends on operating conditions, such as the amount of current density (the water vapor production rate at the cathode) and the air flow rate (flushing water vapor at the cathode). Acknowledgement This research is financially supported by Changwon National University in References 1) H. Yokokawa, H. Tu, B. Iwanschitz and A. Mai, J. Power Sources, 182, (2008). 2) S. C. Singhal and K. Kendall, High Temperature Solid Oxide Fuel Cell: fundamentals, Design and Application, Elsevier (2004). 3) H. G. Bohn and T. Schober, J. Am. Ceram. Soc., 83, (2000). 4) K. Katahira, Y. Kohchi, T. Shimura and H. Iwahara, Solid State Ionics, 138, (2000). 5) K. D. Kreuer, S. Adams, W. Munch, A. Fuchs, U. Klick and J. Maier, Solid State Ionics, 145, (2001). 6) H. Iwahara, Solid State Ionics, 77, (1995). 7) H. Iwahara, Solid State Ionics, 86 88, 9 15 (1996). 8) T. Norby, Solid State Ionics, 125, 1 11 (1999). 9) J. Dailly and M. Marrony, J. Power Sources, 240, (2013). 10) K. Xie, R. Yan, X. Chen, S. Wang, Y. Jiang, X. Liu and G. Meng, J. Alloys Compd., 473, (2009). 11) C. W. Tanner and A. V. Virkar, J. Electrochem. Soc., 143, (1996). 12) Z. Wu and M. Liu, J. Electrochem. Soc., 144, (1997). 13) S. V. Bhide and A. V. Virkar, J. Electrochem. Soc., 146, (1999). 14) N. Taniguchi, C. Nishimura and J. Kato, Solid State Ionics, 145, (2001). 15) L. Jingde, W. Ling, F. Lihua, L. Yuehua, D. Lei and G. Hongxia, J. Rare Earths, 26, (2008). 16) R. Gawel, K. Przybylski and M. Viviani, J. Therm. Anal. Calorim., 116, (2014). 17) P. Sawant, S. Varma, B. N. Wani and S. R. Bharadwaj, Int. J. Hydrogen Energy, 37, (2012). 18) E. Fabbri, A. D Epifanio, E. D. Bartolomeo, S. Licoccia and E. Traversa, Solid State Ionics, 179, (2008). 19) Y. Yoo and N. Lim, J. Power Sources, 229, (2013). 20) E. Fabbri, D. Pergolesi, A. D Epifanio, E. D. Bartolomeo, G. Balestrino, S. Licoccia and E. Traversa, Energy Environ. Sci., 1, (2008). 21) Y. Guo, R. Ran and Z. Shao, Int. J. Hydrogen Energy, 35, (2010). 22) J. Qian, W. Sun, Q. Zhang, G. Jiang and W. Liu, J. Power Sources, 249, (2014). 23) K. H. Bae, D. Y. Jang, H. J. Jung, J. W. Kim, J.-W. Son and J. H. Shim, J. Power Sources, 248, (2014). 24) S. M. Choi, J.-H. Lee, H. S. An, J. S. Hong, H. C. Kim, K. J. Yoon, J.-W. Son, B.-K. Kim, H.-W. Lee and J.-H. Lee, Int. J. Hydrogen Energy, 39, (2014). 25) F. Zhao and F. Chen, Int. J. Hydrogen Energy, 35, (2010). 26) L. Bi, S. Q. Zhang, S. M. Fang, Z. T. Tao, R. R. Peng and W. Liu, Electrochem. Commun., 10, (2008). 27) K. Xie, R. Q. Yan and X. Q. Liu, J. Alloys Compd., 479, L40 L42 (2009). 28) A. V. Virakr, J. Power Sources, 172, (2007). 29) H.-T. Lim and A. V. Virkar, J. Power Sources, 185, (2008). 30) H.-T. Lim and A. V. Virkar, ECS Trans., 25, (2009). 31) M. Y. Park, Y.-G. Jung and H.-T. Lim, Solid State Ionics, 262, (2014). 32) M. Y. Park, H. Bae and H.-T. Lim, J. Kor. Ceram. Soc., 51, (2014). 33) M. G. Jung, Y. J. Kim, Y.-G. Jung and H.-T. Lim, Int. J. Hydrogen Energy, 39, (2014). 34) M. Zunic, L. Chevallier, E. Di Bartolomeo, A. D Epifanio, S. Licoccia and E. Traversa, Fuel Cells (Weinheim, Ger.), 11, (2011). 35) D.-K. Lim, M.-B. Choi, K.-T. Lee, H.-S. Yoon, E. D. Wachsman and S.-J. Song, Int. J. Hydrogen Energy, 36, (2011). 36) R. R. Liu, S. H. Kim, S. Taniguchi, T. Oshima, Y. Shiratori, K. Ito and K. Sasaki, J. Power Sources, 196, (2011). 262