An Interfacial Nanospike‐Structured Cathode for Low Temperature Solid Oxide Fuel Cells

A nanospike structured cathode is ­fabricated by combining a discharge treatment with a low-temperature sintering process. The fuel cell with a nanospike-structured cathode exhibits superior electrochemical performance at low operating temperatures (350–550 °C) and demonstrates certain thermal-cycling stability. Lowering the operating temperature of solid oxide fuel cells (SOFCs) has been widely recognized as a feasible strategy for widening material selections, enhancing durability, and reducing cost of SOFC systems.1-4 The polarization loss of cathode electrode is a major factor limiting the overall SOFC efficiency due to the sluggish oxygen reduction reactions (ORRs) particularly in low temperature regime of 350–650 °C.5-7 There have been significant efforts in developing cathode materials that can decrease polarization loss at reduced temperatures, such as ABO3 type perovskites8-11 and recently developed layered perovskites.12, 13 Alternatively high performance SOFCs have been developed through nanoscale engineering. The widely used method is to deposit nanoparticles into a pre-fired electrode scaffold through the precipitation and decomposition of a metal salt solution. The generated surface nanoparticles could significantly increase active sites for ORRs, and consequently reduce electrode polarization resistance.14, 15 By impregnating La0.6Sr0.4Co0.2Fe0.8O3-δ into Gd-doped ceria,16 Barnett et al. obtained a cathode polarization resistance of 0.24 Ω cm2 at 600 °C, fourteen times lower than ­non-infiltrated one. Recently a micro-nano porous Sm0.5Sr0.5CoO3-δ/La0.9Sr0.1Ga0.8Mg0.2O3-δ hybrid cathode, prepared by ion-impregnation, exhibited a polarization resistance of 0.021 Ω cm2, and the corresponding maximum power density of 2.02 W cm−2 was achieved at 650 °C.17 Essentially the infiltration is an additive approach to add the catalytically active materials onto internal surface of electrodes to enhance the ORRs.18 Here we report the fabrication and evaluation of a novel interfacial nanospike-structured cathode. The interfacial nanospikes grew directly from the internal surface of cathode framework toward void space within the porous cathode. The growth of interfacial nanospikes was enabled by a discharging voltage applied on the button cell under low temperature operating conditions. The fabricated interfacial nanospike-structured cathode is unique in open literature and may significantly improve the electrochemical performance of SOFCs at low temperatures of 350–550 °C. The anode-support/electrolyte assembly of NiO-Sm0.2Ce0.8O1.9 (SDC)/SDC was first pre-sintered at 1350 °C. The slurry of layered perovskite PrBaCo2O5+δ (PBCO) cathode was painted onto electrolyte and then fired at low temperature of 450 °C in a test furnace. During the low temperature firing process, the humidified hydrogen was supplied to the anode so that the NiO in the anode was fully reduced to Ni. Afterwards a discharge voltage of 0.1 V was applied to the cell while the current density response was monitored. The history of cell response (Figure S1, Supporting Information) showed a transient process, increasing from zero to 2.77 A cm−2 then to 5.1 A cm−2 in about 30 minutes and gradually stabilized at 5.8 A cm−2. The polarization performance (V-I curve) was then obtained under different low temperatures. After electrochemical test, the microstructure of the button cell was characterized. As shown in Figure 1a and 1b, the needle-like nano-spikes were uniformly distributed on the pore walls of the cathode. The transmission electron microscope (TEM) images showed that the length of the nano-spikes varies from a few ten nanometer to a few hundred nanometer with diameter about 20 nm, and each of the spikes was composed of a row of nano-beads connected with one another (Figure 1c and 1d). These nanospikes spread all over the cathode interface up to 10 μm (Figure S2, Supporting Information). The lattice-resolved high resolution TEM (HRTEM) image of a grain at the tip of a nanospike shows the presence of highly crystalline nature (Figure 1e), which corresponds to the (102) and (100) crystal planes of the double perovskite structure with lattice inter-planar spacing of d102 = 0.283 nm and d100 = 0.405 nm. These parameters are in agreement with the XRD data of PBCO (Figure S3, Supporting Information), where d102 = 0.274 nm and d100 = 0.391 nm for the synthesized PBCO powders respectively, indicating that the lattice structure of nanospikes is the same as that of the cathode scaffold. The selected area electron diffraction (SAED) pattern of the boxed area confirmed the long-range order crystal structure. The element contents in the nanospikes determined by an energy dispersive x-ray spectroscope (EDX) showed the appearance of metal Pr, Ba, Co and Cu, where Cu was due to the substrate of the sample (Figure S4). All of the above results have cross-verified that the generated interfacial nanospikes are PBCO. The microstructure of the anode was also examined using SEM technique. However, interfacial nano-spikes were not found on the internal surface of porous anodes with/without discharging treatment (Figure S5a,b, Supporting Information). The Brunauer-Emmett-Teller (BET) test result showed that the specific surface areas of the anodes were 0.69 m2 g–1 and 0.66 m2 g–1 with/without discharging treatment respectively (Figure S5c, Supporting Information), further confirming that the effect of discharging treatment was negligible on the micro-morphology of Ni-SDC porous anode. Recently, Ivers-Tiffée et al. have found that a nanometer-scaled and nanoporous Ni/YSZ interlayer between the electrolyte and anode functional layer was formed after applying a short-time reverse current treatment across the anode. The corresponding anode performance was significantly improved.19-21 It was recognized that the reduction of the electrolyte due to extremely low oxygen partial pressure and a reverse high current density together lead to the formation of nanoporous interlayer. In our current experiment, an external voltage was applied on the entire cell in steady of merely anode and the obvious nanospikes occurred at the internal surface of porous cathode where oxygen partial pressure was high. More than likely, the in situ growth mechanism of interfacial ­nanospikes is different from that of above in-situ generated nanoporous interlayer, which, we believe, needs significant effort of further studies. At this stage, it is speculated that the applied cell voltage plays an important role on the growth of interfacial nanospikes during low temperature sintering process. When the operating voltage of 0.3 V was used at 450 °C, the nanospike was less and shorter (Figure 2a); with the lower applied voltages of 0.2 V and 0.1 V, the interfacial nanospikes become denser and longer (Figure 2b and 2c). Without applying a voltage, the interfacial nanospikes could not be found on the pore walls of the cathode (Figure S6, Supporting ­Information). Under the same discharge voltage (0.1 V), the growth of interfacial nanospikes was also affected by sintering temperature. With the sintering temperature of 550 °C, the interfacial nano-beads could be found on the pore walls of the cathode (Figure S7a, Supporting Information). With the lower sintering temperatures of 500 °C and 450 °C, the interfacial nanospikes became longer with length ranging from 100 nm to 500 nm (Figure S7b and c, Supporting Information). These observations suggest that the applied voltage favors the growth of interfacial nanospikes while the temperature tends to smooth them out. The pore wall with interfacial-nanospikes is a unique microstructure for the cathode of SOFCs, which was never reported in open literature. The BET test results indicated that the specific surface area of the cathode without discharging voltage treatment was 3.82 m2 g–1; after discharging treatment under 0.1V, the specific surface area of the cathode was increased to 12.51 m2 g–1 (Figure S8, Supporting Information). This result confirmed that the interfacial nanospikes significantly enhanced the effective surface area and reaction sites for ORRs in the cathode. It is expected that the increased cathode surface area will improve the cell performance. The electrochemical performance of the anode-supported cell Ni-SDC/SDC/PBCO with thickness of 760 μm/8 μm/20 μm respectively was further evaluated, in which the nanospike-structured PBCO cathode was fabricated at the sintering temperature of 450 °C and discharging voltage of 0.1 V. As shown in Figure 3a, the cell exhibited maximum power densities (Pmax) of 1.453, 1.044, 0.657, and 0.316 W cm−2 at 550, 500, 450, and 400 °C, respectively. The Pmax of 0.135 W cm−2 could be achieved at 350 °C, the lowest testing temperature of SOFCs in open literature. As a comparison, the identical cell was fabricated, in which the cathode was sintered at 450 °C without any discharging voltage treatment. The SEM image did not show any presence of nanospikes in the cathode (Figure S6, Supporting Information). The maximum power density of the corresponding cell only reached 0.764 W cm−2 at 700 °C, 0.515 W cm−2 at 650 °C, 0.305 W cm−2 at 600 °C, and 0.162 W cm−2 at 550 °C, respectively (Figure S9a, Supporting Information). Furthermore, a traditional anode-supported Ni-SDC/SDC/PBCO was also evaluated, in which the PBCO cathode was sintered at 1000 °C. As shown in Figure S9b, the Pmax of this cell only reached 0.355 W cm−2 at 550 °C, which was four times less than that of the novel cell with interfacial nanospike-structured cathode. When the temperature increased to 700 °C, the Pmax reached 1.044 W cm−2, but still less than that of the novel cell operated at 550 °C. The impedance spectra at open circuit condition showed that electrode polarization resistance (Rp) was as low as 0.048 Ω cm2 at 550 °C and 1.32 Ω cm2 at 350 °C respectively (Figure 3c). By contrast, the Rp of Ni-SDC/SDC/PBCO cell with traditional cathode reached 0.291 Ω cm2 at 550 °C (Figure S10, Supporting Information). The SEM image of the cathode sintered at 1000 °C was shown in Figure S11, Supporting Information. The interfacial nanospike-structured cathode also showed comparable performance with the infiltrated cathode. For example, the cell with a micro-nano porous Sm0.5Sr0.5CoO3-δ/La0.9Sr0.1Ga0.8Mg0.2O3-δ hybrid cathode prepared by impregnation process exhibited the Pmax of 0.91 W cm−2 at 550 °C.17 The thermal-cycling experiment in the temperature range of 200–550 °C was carried out to examine the stability of nanospike-structured PBCO cathode. As shown in Figure 3b, the Pmax exhibited 4.4% degradation while the overall Rp ­demonstrated 10.3% increase after 40 thermal cycles. The ­corresponding traditional cell without nanospikes (with thermal-cycling 350–700 °C) showed 23.6% decrease of power density and 52.1% increase of Rp. These results indicated a certain stability of the entire cathode layer against thermal-cycling operation, which was consistent with nanospike morphology in post-tested cells. The growth of interfacial nanospikes is dependent on the discharge voltage and low sintering temperature, which in turn affects the electrochemical performance of the cell. A set of nanospike-structured PBCO cathodes was fabricated under low sintering temperatures of 450, 500 and 550 °C respectively. At each of sintering temperatures, a discharge voltage of 0.1, 0.2 and 0.3 V was used. As shown in Figure 4, the Pmax increases with increasing the operating temperature from 400 to 550 °C. For a given sintering temperature, lowering the ­discharge voltage from 0.3 V to 0.1 V during the cathode ­fabrication process could improve the Pmax. According to the character of SOFC polarization curve, the lower voltage would lead to higher current density. The higher current density in turn would improve the growth of interfacial nanospikes and thus cell performance, e.g., for sintering temperature of 450 °C and discharge voltage of 0.3 V and 0.1 V, the Pmax reached 1.19 W cm−2 and 1.45 W cm−2 at 550 °C respectively. It also can be seen that for a given discharge voltage lowering the sintering temperature from 550 to 450 °C, the corresponding cell performance was also improved, e.g., at discharge voltage of 0.3 V and the sintering temperature of 550 and 450 °C, the corresponding peak power density was 0.75 W cm−2 and 1.19 W cm−2 respectively at operating temperature of 550 °C. Therefore a suitable low sintering temperature favors the growth of nanospikes and thus the corresponding cell performance. In summary, an interfacial nanospike-structured PBCO cathode has been successfully developed by low-temperature sintering (450–550 °C) in combination with electrochemical voltage discharge treatment. Although the fundamental mechanisms are not fully understood yet and need further study, the experimental results well demonstrated that the SOFCs with interfacial nanospike-structured cathodes could obtain very good performance at low temperatures. This novel approach will potentially bring a new paradigm towards nanoscale ­engineering of porous cathode for low-temperature high performance SOFCs, which otherwise would be very difficult for conventional high temperature sintering process (>1000 °C). Electrolyte and Electrode Powder Preparation and Basic Characterizations: PrBaCo2O5+δ (PBCO) cathode powders were synthesized using modified Pechini process, where citrate and ethylenediaminetetraacetic acid (EDTA) were employed as parallel complexing agents. Pr6O11 was first dissolved in nitric acid; the calculated amount of Ba(NO3)2·9H2O and Co(NO3)2·6H2O was dissolved in EDTA-NH3 aqueous solution under heating and stirring conditions. An appropriate amount of citric acid was then added in the solution. The molar ratio of EDTA : citric acid : total of metal cations was controlled around 1:1.5:1. After converted into viscous gel under heating and stirring conditions, the solution was ignited to flame and result in ash. The resulting ash-like material was afterwards calcined in air at 950 °C for 3 h. The fine PBCO powders were then mixed thoroughly with a 6 wt% ethylcellulose-terpineol binder (Heraeus, V006) to obtain the cathode slurry. The glycine-nitrate process (GNP) was employed to synthesize super loose Sm0.2Ce0.8O1.9 (SDC) electrolyte powders with the filled density as low as 0.032 g cm−3, only 0.46% theoretical density (6.958 g cm−3) of the bulk SDC. During the process of acid-base neutralization reactions between nitric acid and ammonia, ammonium nitrate was produced. The firing of ammonium nitrate released significant amount of heat in a very short period of time. Therefore, the volume-to-weight ratio of synthesized SDC powder can be adjusted by different amount of nitric acid. The phase identification of prepared cathode and electrolyte powders was studied with X-ray powder diffraction by Cu-Kα radiation (D/Max-gA, Japan). A scanning rate of 5° min−1 was used in the range of 20° < 2θ < 80°. Fuel Cell Fabrication, Cathode Pre-treatment and Electrochemical Test: The anode-supported SDC bi-layer (Φ = 15 mm) was prepared using a dry-pressing/co-firing method. NiO + SDC + starch mixture (60%:40%:20% in weight) was first pre-pressed at 200 MPa as an anode substrate. Then the loose SDC powder synthesized above was uniformly distributed onto the anode substrate, co-pressed at 250 MPa and sintered subsequently at 1350 °C for 5 h to densify the SDC membrane (≈8 μm). The cathode slurry was then painted on SDC electrolyte film and sintered at 450, 500, and 550 °C for 0.5 h in air to form single cells of NiO-SDC/SDC/PBCO (≈20 μm). The silver paste (Heraeus, C8728) painted onto the cathode surface was used as a current collector. The silver wires were connected to both the anode substrate and cathode respectively for electrochemical measurements. The button cell was assembly onto an alumina tube using silver paste. The circumference of the anode substrate was covered with ceramic slurry (Aremco, Ceramabond 552). The humidified hydrogen was supplied to the anode while the cathode was exposed to the ambient air. The temperature of the cell assembly was controlled by a test furnace. The furnace temperature was increased from room temperature to an elevated temperature, (e.g., 450 °C, 500 °C, and 550 °C respectively) with step size of 3 °C min–1. The furnace temperature was dwelled at the elevated temperature for 10 hours to complete anode reduction. Subsequently, the button cells were discharged at constant loads of 0.1, 0.2 and 0.3 V at the elevated temperature, (e.g., 450 °C, 500 °C, and 550 °C respectively). The flow rate of fuel was 40 ml min−1 controlled by a mass flow meter while the current response history was recorded by a Scribner 890ZV. Afterwards voltage-current curves were measured at a 30 mV s−1 scan rate at temperature range of 350–550 °C. The electrochemical impedance spectra (EIS) of the cells were obtained by a Solatron 1260 frequency response analyzer in combination with a Solartron 1287 potentiostat under open-circuit conditions. The sweeping frequency range was from 0.1 to 105 Hz. Microstructure Observation and Characterization of the Nanospikes at Cathode Interface in Post-Test Cells: The microstructure and morphology of the grown nanospikes on the surface of cathode particles in the vicinity of electrolyte/cathode interface were observed by field emission scanning electron microscope (FESEM, Zeiss). The cathode particles were completely dispersed into ethanol with ultrasonic vibration and observed by transmission electron microscopy (TEM, Hitachi HF2000) operating at 200 kV equipped with energy-dispersive X-ray spectroscopy (EDX) and high resolution TEM (JEOL 2010F) to obtain the details about crystal lattice, element distribution and selected area electron diffraction pattern. The specific surface areas of the cathode and anode electrodes were determined by nitrogen adsorption/desorption measurements at 77 K using a BET analyzer (Quantachrome, NOVA 2000). This work was supported by National Science Foundation, grant no. CMMI-1100085, the US Department of Energy, Basic Energy Sciences, grant no. DE-SC0001061, and the University of South Carolina, Office of Research and Graduate Education. 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.