A Nanostructured Architecture for Reduced‐Temperature Solid Oxide Fuel Cells

Yucun Zhou,Hao Wu,Ting Luo,Jianqiang Wang,Yixiang Shi,Changrong Xia,Shaorong Wang,Zhongliang Zhan

DOI: https://doi.org/10.1002/aenm.201500375

IF: 27.8

2015-01-01

Advanced Energy Materials

Abstract:A nanostructured architecture is proposed for reduced-temperature solid oxide fuel cells, featuring thin scandium-stabilized zirconia electrolytes sandwiched between two porous 430L metallic layers. Impregnating nanoscale catalysts into the porous 430L scaffolds—Ce0.8Sm0.2O1.9/Ni at the anodes and SmBa0.5Sr0.5Co2O5 at the cathodes—yields encouraging power densities of 1.02 W cm−2 at 600 °C and 0.6 W cm−2 at 550 °C. Solid oxide fuel cells (SOFCs) represent one of the most efficient means for direct conversion of fuels into electricity with minimal environmental impact.1 Traditionally, SOFCs are all-ceramic with yttria-stabilized zirconia (YSZ) electrolytes, Ni-YSZ anodes, and lanthanum strontium manganite cathodes.2 The typical 750–1000 °C operating temperature provides important advantages including high efficiencies and high power densities; it also introduces difficult challenges related to seals, interconnects, startup time, and long-term durability. Lowering the operating temperature to the regime of 500–600 °C,3 as enabled by reducing the electrolyte thickness,4 utilizing more conductive electrolytes,5 or optimizing the electrode materials/microstructures,6-8 can at least partly mitigate the above problems. In particular, nanoscale engineering by chemical solution impregnation has produced highly active and advanced electrodes at temperatures ≤600 °C, with high versatility in selecting the infiltrated catalytic nanoparticles and the supporting oxide scaffolds.9 Notwithstanding these impressive progresses, the inherent brittleness of the high-porosity oxide scaffolds, as required to facilitate the catalyst infiltration, remains one of the main sources of SOFC durability limitations. Due to their attractive features such as plasticity, weldability, high electrical and thermal conductivities, porous metal materials have been widely used for filtration and separation, flow distribution, and catalyst supports.10 Nonetheless, porous metal materials have not seriously been considered as primary components in functioning electrodes due to lack of catalytic activity and structural stability under the traditional SOFC environments. A distinctive advantage with reduced-temperature (≤600 °C) SOFCs is their feasibility of using inexpensive alloys to support the electrochemically active layers—anode, electrolyte, and cathode—while simultaneously collecting the electrical current.11-13 Table 1 summarizes some material sets developed for metal-supported SOFCs together with power densities measured in H2.14-18 Impregnating nanoporous Ce0.8Sm0.2O1.9 (SDC) and/or Ni catalysts into macroporous 430L stainless steels yielded hierarchical composites (e.g., nano-SDC/Ni@430L) with high catalytic activities toward hydrogen electro-oxidation, obviating the need for porous and fragile ceramic scaffolds.16, 17 In the present work, porous 430L stainless steels were for the first time used to support nanoscale SmBa0.5Sr0.5Co2O5 (SBSCO) catalysts. Such nano-SBSCO@430L composites efficiently catalyzed oxygen reduction reactions at 500–600 °C, thereby enabling a novel SOFC architecture with minimized ceramic components, enhanced reliability, and reduced materials costs. In particular, thin scandium-stabilized zirconia (SSZ) electrolyte SOFCs with nano-SDC/Ni@430L anodes and nano-SBSCO@430L cathodes produced encouraging power densities at reduced temperatures, e.g., 1.02 W cm−2 at 600 °C, when operating on humidified hydrogen fuels and air oxidants. Figure 1a shows a typical scanning electron microscopy (SEM) image of tri-layer structures of “porous 430L | dense SSZ | porous 430L”, as fabricated by laminating and firing tape-casted green sheets at 1325 °C in diluted hydrogen (Supporting Information). The SSZ electrolytes were fully dense and typically 15 μm thick. Inexpensive 430L stainless steels were used as the metallic scaffold material due to their good corrosion resistances under atmospheric conditions and compatible coefficients of thermal expansion with the SOFC electrolyte materials (11.4 × 10−6 K−1 for 430L vs 10.8 × 10−6 K−1 for SSZ). Use of ammonium oxalate as the fugitive material in the tape casting formulation yielded a porous microstructure with a mean pore size of ≈15 μm and a porosity of ≈45% (Figure S1, Supporting Information). High oxidation resistances of the porous 430L scaffolds were confirmed by negligible weight gains measured in air at 600 °C (Figure S2, Supporting Information). On the fuel electrode side, the 430L pore surfaces had a nanoporous and well intraconnected coating of SDC/Ni (Figure 1b). The air electrode side was coated with nanoporous SBSCO oxides (Figure 1c). SBSCO oxides were selected as the active cathode catalyst due to their excellent mixed oxide ionic and electronic conductivities that made the entire SBSCO surfaces highly active for oxygen reduction reactions (Figure S3, Supporting Information).19 Furthermore, well intraconnected SBSCO coatings provided effective pathways for transport of oxide ions to the adjacent SSZ electrolytes (Figure S3, Supporting Information). On the anodes, oxide ions diffused through the bulk SDC coatings to the triple-phase boundaries, where SDC, Ni, and H2 came into contact to facilitate hydrogen oxidation reactions (Figure S3, Supporting Information). Both SDC/Ni and SBSCO coatings were deposited by impregnating the mixed nitrate solutions in appropriate ratios, followed by calcinations in air at 650 °C (Supporting Information). Calcinations at higher temperatures were not pursued so as to suppress the formation of oxide scales on the 430L scaffolds and the growth of the deposited catalysts. Formation of SDC/Ni and SBSCO in the coatings was independently confirmed by X-ray diffraction patterns of powders, synthesized by calcinating the infiltrate solutions at 650 °C (Figure S4a,b, Supporting Information). Note that some minor impurities of BaCoO3 were also observed in the cathode catalysts (Figure S4b, Supporting Information), even though citric acid was added as a chelating agent to promote crystallization of perovskite structures. Examination of as-synthesized catalysts by transmission electron microscopy (TEM), with the results illustrated in Figure 2a for SDC/Ni and in Figure 2b for SBSCO, revealed a typical porous aggregate morphology that consisted of 10–40 nm particles with intercrystalline porosities. This observation is consistent with the XRD crystalline sizes calculated by the Debye–Scherrer method—18 nm for SDC, 35 nm for Ni, and 19 nm for SBSCO. The selected area electron diffraction (SAED) patterns for SDC (Figure 2c) and SBSCO (Figure 2e) show the presence of diffuse rings along with some Bragg spots, indicative of a highly defective or polycrystalline nature for these two catalysts. Differently, the SAED patterns for Ni (Figure 2d) suggest that these nanoparticles are single crystals. The associated characteristic speckle patterns in Figure 2d should arise from the adjacent smaller single crystals of Ni metals. Electrochemical measurements were performed on as-fabricated fuel cells, i.e., nano-SDC/Ni@430L | SSZ | nano-SBSCO@430L, with 3% humidified hydrogen fuels in the anodes and dry air oxidants in the cathodes at 500–600 °C. The catalyst loadings were 10 wt% for SDC/Ni and 16 wt% for SBSCO. Figure 3 shows typical polarization curves of cell voltages and power densities versus current densities. The open circuit voltage values increased from 1.104 to 1.122 V with decreasing temperature, and were within 20 mV of the calculated Nernst potentials. These results suggest that the SSZ electrolytes co-fired with the porous 430L scaffolds were gas impermeable, consistent with the SEM observation in Figure 1a. The maximum power densities measured were 1.02, 0.86, 0.63, 0.41, and 0.27 W cm−2 at 600, 575, 550, 525, and 500 °C, respectively. These values are very competitive with the common all-ceramic SOFCs,4, 7 or prior metal-supported SOFCs with zirconia-based electrolytes,12, 17, 20 which can be ascribed to good catalytic activities of nanoscale catalysts and high oxide ionic conductivities of SSZ electrolytes. For instances, metal-supported SOFCs with the FeCr alloy supports, Ce0.9Gd0.1O2-δ-Ni impregnated cermet anodes, Sc2O3 and Y2O3 co-doped ZrO2 electrolytes and in situ sintered La0.6Sr0.4CoO3-δ cathodes produced peak power densities of 1.14 W cm−2 at a higher temperature of 650 °C.13 Power densities reported here were also 20–30% higher than those obtained at 550–600 °C for our prior metal-supported fuel cells with similar nano-SDC/Ni@430L anodes and SSZ electrolytes, but with in situ sintered ESB-Ag cathodes (e.g., 0.77 W cm−2 at 600 °C).18 Although there are prior reports of excellent all-ceramic fuel cells that produced power densities of ≈1 W cm−2 at 500–550 °C, they were obtained with much more conductive electrolytes such as Gd3+-doped ceria21 or strontium- and magnesium-doped lanthanum gallate.8, 19 Electrochemical impedance spectroscopy (EIS) was used to better understand the electrode reactions and quantitatively determine their contributions to the internal resistances. Since it is not so straightforward to deconvolve EIS data of functioning fuel cells and relate them to the anode or cathode contributions, symmetrical electrode fuel cells were fabricated from tri-layer structures of “porous 430L | dense SSZ | porous 430L” with the same catalysts impregnated into both 430L scaffolds. Impedance data were collected in a uniform atmosphere—97% H2–3% H2O for nano-SDC/Ni@430L anodes and dry air for nano-SBSCO@430L cathodes (Figure S5, Supporting Information). Such symmetrical configurations allow for the polarization resistance to be explicitly determined for each electrode under the typical SOFC operating conditions. Representative Nyquist plots of the impedance data at 600 °C are shown in Figure 4a for nano-SDC/Ni@430L anodes and in Figure 4b for nano-SBSCO@430L cathodes, where the electrode polarization resistances were obtained by taking the differences between the higher- and lower-frequency real-axis intercepts. Figure 4a shows that the anode polarization resistance was RP,A = 0.116 Ω cm2 at 600 °C. In contrast, the state-of-the-art Ni-YSZ cermet anodes had to work at temperatures above 750 °C to obtain comparably low RP,A values.22 Good catalytic activities of the nano-SDC/Ni@430L composites for hydrogen electro-oxidation at reduced temperatures can be ascribed to superior redox capability of nanoscale SDC,15 which becomes preferentially facilitated within the vicinity of Ni particles due to spillover of highly reactive atomic hydrogen from nickel to doped ceria as observed by Crozier and co-workers.23 Note that trapping of Ni particles within the excessive SDC ceramic matrix (SDC/Ni = 80/20 wt) helped to stabilize the nanostructured catalysts and thus maintain the low RP,A values over the long run.16 Figure 4b shows that the cathode polarization resistance was RP,C = 0.093 Ω cm2 at 600 °C. Such low RP,C values make nano-SBSCO@430L composites competitive with the well-known SOFC cathodes such as La0.6Sr0.4Co0.2Fe0.8O3-δ, Ba0.5Sr0.5Co0.8Fe0.2O3-δ, and Sm0.5Sr0.5CoO3-δ,6, 8, 24 especially given that the SBSCO catalysts were coated onto the pore walls of the inert 430L scaffolds. The combined RP,A and RP,C value at 600 °C was in a reasonable agreement with the total interfacial polarization resistance measured for the functioning fuel cells (Figure S6, Supporting Information), especially given the difference in the measurement conditions. The pure ohmic losses (RO) were taken from the higher-frequency real-axis intercepts of the Nyquist plots (Figure 4a,b), and were largely the same for the symmetrical anode or cathode fuel cells within the measurement uncertainties, i.e., RO ≈ 0.123 cm2 at 600 °C. Based upon the oxide ionic conductivity of 0.014 S cm−1 for SSZ electrolytes as measured at 600 °C using the standard four-probe DC method, the calculated area specific resistance for the 15 μm thick electrolytes is REL = 0.11 Ω cm2. Comparable REL and RO values indicate that the 430L scaffolds can efficiently collect the electrical currents on both electrodes while supporting the nanoscale SDC/Ni and SBSCO catalysts. The RO, RP,A and RP,C values at varied temperatures are summarized in Figure 4c, showing that the ohmic losses, the anode, and cathode polarizations almost equally contributed to the internal resistances for the present metal-supported fuel cells. A higher activation energy for oxygen reduction than for hydrogen oxidation (1.13 vs 0.73 eV) suggests that optimizing the cathode composition and microstructure would be more effective in further enhancing the fuel cell performance at lower temperatures. The most distinctive feature for the present fuel cells is the use of porous 430L scaffolds to support nanoscale SBSCO particles as the active cathode catalysts. Such a nonrandom structure made the thermal expansion coefficients (CTE) of nano-SBSCO@430L composites largely dominated by the 430L scaffolds (11.4 × 10−6 K−1),25 even though SBSCO has a much larger CTE value (20.3 × 10−6 K−1).26 Operation at a reduced temperature regime of 500–600 °C, as enabled by the highly conductive SSZ electrolytes, minimized the formation of gaseous Cr-species from the supporting 430L scaffolds as well as the interaction between SBSCO and 430L, which may otherwise react with the supported SBSCO catalysts and thus result in a degradation in the cathode catalytic activity. This argument was supported by stable power densities measured for functioning fuel cells over a 60 h duration (Figure S7, Supporting Information) and almost unchanged RP,C values during a 250 h measurement of impedance data for symmetrical cathode fuel cells (Figure S8, Supporting Information). In summary, a new architecture has been proposed for reduced-temperature SOFCs with the distinctive feature of thin electrolyte films placed between two porous 430L alloys. Using highly conductive SSZ electrolytes allowed for small ohmic losses at reduced temperatures, and impregnating nanoscale catalysts into the porous 430L scaffolds—SDC/Ni at the anodes and SBSCO at the cathodes—yielded impressively low electrode polarization resistances. The resulting metal-supported SOFCs exhibited promising power densities in hydrogen fuels, e.g., 1.02 W cm−2 at 600 °C and 0.6 W cm−2 at 550 °C. Compared with the traditional SOFCs, the new architecture displays attractive advantages of higher robustness and lower material cost, as enabled by reducing the usage of fragile ceramic components and increasing the dominance of inexpensive stainless steel alloys. Y.C.Z., H.W., and T.L. contributed equally to this work. The authors gratefully acknowledge the financial support of National Basic Research Program of China (Grant No. 2012CB215401) and the National Natural Science Foundation of China (Grant Nos. 51172266, 51072219, 51106173, and 51071169). As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.