Improved Sulfur Tolerance of SOFCs through Surface Modification of

Apr 3, 2018 - The AFL layer was prepared using the same organics (Menhaden fish oil, PVB, BBP, and ... Data analysis was done using CASAXPS software...
1 downloads 3 Views 7MB Size
Download PDF
Article www.acsaem.org

Cite This: ACS Appl. Energy Mater. 2018, 1, 1559−1566

Improved Sulfur Tolerance of SOFCs through Surface Modification of Anodes Thomas Hays,‡ A. Mohammed Hussain,‡ Yi-Lin Huang,† Dennis W. McOwen,‡ and Eric D. Wachsman*,‡ †

Maryland Energy Innovation Institute and ‡Department of Materials Science & Engineering, University of Maryland, College Park, Maryland 20742, United States ABSTRACT: SOFCs are a promising technology for high efficiency power generation with fuel flexibility; however, sulfur, a common contaminant in most hydrocarbon fuels, can cause severe anode degradation. Here we demonstrated that surface modification of Ni-GDC-based SOFC through nanoparticle infiltration drastically reduces the sulfur poisoning effect. Infiltrated SOFCs showed stable performance with sulfurcontaminated fuel for over 290 h, while unmodified SOFCs became inoperative after 60 h. We proposed that the nanosized GDC coating promotes sulfur removal through SO2 formation by increasing the density of reaction sites and providing a ready supply of oxygen to those sites. The significant benefit provided by this electrode treatment is promising for the integration of SOFCs with established fuel compositions. KEYWORDS: LT-SOFC, Ni-GDC anodes, coking, deactivation, nano ceria, sulfur tolerance, anticoking, stability

■

INTRODUCTION Solid oxide fuel cells (SOFCs) are efficient energy conversion devices that operate on versatile fuels, including hydrocarbons (e.g., natural gas). These fuels often have a certain amount of impurities (e.g., H2S), depending on the source and production methods.1 Fuel contamination is a potentially serious hurdle to the widespread adoption of SOFCs as an energy generation technology.2 For the past few decades, the leading anode materials system for SOFCs has been Ni-YSZ (yttrium-stabilized zirconia). NiYSZ-based anode-supported SOFCs operate in the ∼800 °C range and use nickel as the catalyst and current collector in the anode.3−5 These operating conditions make sulfur a particularly damaging fuel contaminant due to the high affinity for sulfur possessed by nickel and the fact that nickel sulfide formation is favorable.2,6−9 Both of these realities mean that even trace amounts of sulfur can destroy Ni-YSZ-based cells very quickly through both catalytic site occupation and cell cracking due to the volume change associated with nickel sulfide formation. Therefore, SOFC systems running on natural gas or other hydrocarbon fuels must use sacrificial scrubbers or an additional desulfurization system to remove sulfur. This adds material and maintenance costs to the system and increases the difficulty of deployment. More recently, low-to-intermediate temperature SOFCs based on ceria oxygen-ion conductors have become more common.10−13 These cells use a rare earth doped ceria, such as GDC, as the electrolyte and Ni-GDC cermet as the anode. These cells can experience similar performance loss issues to the Ni-YSZ cells when exposed to sulfur, but because of their lower operating temperature (550−650 °C), the mechanisms can differ. Physisorption of sulfur atoms to catalytic sites still occurs, but the favorable chemical reactions are different.12−14 © 2018 American Chemical Society

Sulfate compounds are more commonly formed, and both nickel and cerium can be affected. Sulfur occupying catalytic sites during cell operation leads to oxidation of the nickel instead of oxygen ions reacting with fuel species.2 The lower number of effective catalytic sites caused by the presence of sulfur will also enhance the damage from other chemical species which are present. This is due to a lower oxygen flux through the cell electrolyte that then limits the oxidation and removal of species such as carbon. Overall, the level of damage done to NiGDC-based IT-SOFCs by sulfur is less than for Ni-YSZ cells.13 One explanation for this difference is the variable oxidation state of the cerium atom. This variability means that the ceria lattice can readily give up lattice oxygen and prompt gaseous sulfur dioxide formation which lowers the local effective sulfur concentration, thus preventing chemical reactions with the anode structure.10,14 This phenomenon can be used to design SOFC anodes with greater tolerance to sulfur poisoning.15 By increasing the interface area between ceria and the nickel material in the anode, ceria can “shield” nickel from sulfur poisoning. However, the porosity of the anode should be preserved as much as possible to ensure no significant loss of catalytic sites. Therefore, any material added to the anode should have very small particle size. By use of a surface coating approach, the added ceria can have a disproportionately large effect relative to its total mass. In addition to shielding catalytic sites, surface coatings can improve the cell performance without needing to modify the original SOFC structure. Recently, infiltration of electrocatalysts has been identified as a promising technique to Received: December 29, 2017 Accepted: March 30, 2018 Published: April 3, 2018 1559

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials

Figure 1. I−V measurement of (a) reference cell and (b) GDC-infiltrated cell. (c) Long-term stability of maximum power density for GDCinfiltrated cell with rapid failure of reference cell. (d) Constant OCV of GDC-infiltrated cell with decaying reference cell. Fuel: 20 ppm of H2S, 50:50 H2:CH4, T = 650 °C.

enhance stability and performance of conventional SOFC electrodes.16 It has been shown that polarization resistance of SOFC anodes can be improved using infiltrated coatings containing GDC at relatively low wt % loadings. For example, Nb-doped SrTiO3 infiltrated with GDC (along with a chosen catalytic metals such as Pd, Ru, Pt, and Ni) has greatly improved the electrochemical properties of anodes at low temperature ranges (400−650 °C).17−19 Specifically, Kurokawa et al. investigated the effect of H2S (40 ppm of H2S in H2) on a ceria-coated Ni-YSZ anode; the resulting cell showed good stability over a period of 500 h at 700 °C under a current load of 0.4 A/cm2.20 Further, Mo0.1Ce0.9O2+δ-modified Ni-YSZ using the infiltration technique has shown very good sulfur tolerance in the presence of 50 ppm of H2S in H2 investigated at 750 °C.21 Also, infiltration of nickel into a porous YSZ anode scaffold showed improved tolerance to both coking and sulfur poisoning.22 The increased density of catalytic sites caused by the infiltration treatments was found to significantly improve anode performances; such positive effects are attributed mainly to increased surface area and decreased particle sizes of the infiltrated nanoparticles.22−24 In this study, Ni-GDC-based SOFC button cells are evaluated for resistance to sulfur poisoning by testing electrochemical performance under constant current operating conditions. Cells constructed using modified anodes with increased ceria loading were compared to unmodified cells and were found to have far more stable polarization resistance and total area specific resistance (ASR). The electrochemical performance of each cell was tested, followed by a suite of post-test characterization including scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, and

energy-dispersive X-ray spectroscopy (EDS). The purpose of this was to search for any microstructural or compositional changes in the anodes that could be correlated to the cell performance changes observed during testing. Additionally, the amount of sulfur present after testing would help illuminate the mechanism of interaction with the catalyst and ion conducting materials.

■

EXPERIMENTAL SECTION

Porous NiO-GDC anodes were prepared using tape casting. The tapecasting recipe contained ethanol as solvent, Menhaden fish oil as dispersant, polyvinyl butyral (PVB) as binder, benzyl butyl phthalate (BBP) as a plasticizer, and 10 μm poly(methyl methacrylate) (PMMA) as pore-former (3 wt %). To obtain the desired SOFC structure, a 30 μm thick as-prepared GDC tape was laminated with a finer microstructured NiO-GDC anode functional layer (AFL) tape and prelaminated NiO-GDC anode support layers (ASL) using a hot press (180 °F). The AFL layer was prepared using the same organics (Menhaden fish oil, PVB, BBP, and PMMA) but with smaller particles of NiO and GDC (1−2 μm). The finer microstructure of this layer enhances the anode reaction kinetics by providing more available reaction sites. The laminated tapes were stepwise heat-treated to burn out the PMMA pore-formers and any organics, followed by sintering at 1450 °C for 4 h. The resulting half-cell consists of a 550 μm thick porous NiO-GDC ASL scaffold and an ∼20 μm thick dense GDC electrolyte. The LSCF-GDC cathode was then deposited on the prepared half-cell by screen printing and dried in an oven at 100 °C for 2 h. The cathode layer was then sintered at 1100 °C for 2 h. The prepared anode-supported full cells were infiltrated with GDC precursor (cerium nitrate hexahydrate, gadolinium nitrate hexahydrate) on the anode side. The Ce0.9Gd0.1O2−δ precursor (1 M concentration) was prepared by dissolving nitrates of cerium and gadolinium (Alfa Aesar) in H2O. A few drops of the infiltrate solution were deposited in the porous anodes. Samples were then placed under vacuum for 10 min. Between each successive infiltration step, the 1560

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials

Figure 2. Nyquist plots over time for (a) reference cell and (b) GDC-infiltrated cell. (c) Stable ohmic ASR for sulfur-exposed cells. (d) Growth in non-ohmic ASR for reference cell with stable GDC-infiltrated cell. Fuel: 20 ppm of H2S, 50:50 H2:CH4, T = 650 °C. exposed to the fuel stream for 24 h at 650 °C and then cooled in hydrogen (25 sccm flow rate). TEM images were taken using a JEM 2100 FE-TEM.

samples were heat-treated at 400 °C for an hour to decompose the nitrates. Cell weight was measured prior to infiltration and after each firing. The infiltration cycles were repeated to get approximate loading of ∼3.62 wt % of added GDC in all samples. This process resulted in a fine coating of GDC throughout the interior pore volume of the anode layer. The electrochemical performance of these infiltrated cells was compared to unmodified Ni-GDC based SOFCs. SOFCs were tested using a standardized procedure composed of cycles lasting slightly longer than 1 h. In each test cycle, an impedance spectrum measurement, a 0.033A/s galvanodynamic scan, and a 1 h 0.1 A/cm2 galvanostatic measurement were taken. A Solartron CellTest system composed of a 1470E potentiostat, and a 1400 frequency response analyzer was used for electrochemical measurements. Once the cell reached the operating temperature of 650 °C, a baseline test cycle was completed on humidified hydrogen. Following this cycle, a fuel composition of 50:50 H2:CH4 (Airgas, 99.99%) was used. The chosen concentration of 20 ppm of H2S was then added to the fuel stream through a mass flow controller using a source tank of 2500 ppm of H2S balanced in CO2 (Airgas, ±2% H2S content). Following the long-term testing of these cells, the anode surface and cross section were examined using scanning electron microscopy (SEM) (Hitachi SU-70) with a field emission gun equipped with a Bruker XFlash silicon drift EDS detector. XPS measurements were performed in a Kratos Axis 165 X-ray photoelectron spectrometer, with a monochromatic aluminum X-ray source operating at 1400 keV. Data analysis was done using CASAXPS software. Raman spectroscopy was performed using a Horiba Jobin Yvon LabRam ARAMIS Raman microscope with a 532 nm laser. The instrument was calibrated with a Si wafer at 520.7 cm−1. All spectra were normalized to the 1030−1280 cm−1 region, corresponding to the Ni−O band. In addition to full cell testing, a number of powder samples were prepared for TEM analysis. In order to mimic the SOFC anode, sections of fired anode functional layer (already prepared for SOFC fabrication) composed of fine nickel and GDC grains were calcined at 900 °C for 2 h. The heat-treated composites were crushed using a mortar and pestle to create a powder. Four powder samples were exposed to sulfurized fuels. Two samples were coated with GDC solution, and two served as unmodified controls. One test-control pair was exposed to humidified H2 and the other to a humidified 50:50 H2:CH4 mixture (Airgas, 99.99%). In each case, the gas flow rate was 125 sccm. Both streams contained 40 ppm of H2S. Each sample was

■

RESULTS AND DISCUSSION Figures 1a and 1b show I−V curves of the reference and the infiltrated cells, respectively, as a function of sulfur exposure time. There was a significant difference in the performance of the cells containing the infiltrated GDC as compared to the reference cell once exposed to sulfur. Figure 1c shows the trends in maximum power density of both cells over time, with points corresponding to the peaks in Figures 1a and 1b. Both cells displayed a significant drop in power output within 5 h of exposure, caused by reversible sulfur adsorption on the anode triple phase boundary (TPB) sites.14,25−27 However, while the power loss continued until failure in the reference cell, the infiltrated cell stabilized and experienced no additional power loss for over 290 h. The open circuit voltage (OCV) of both cells is shown in Figure 1d. The reference cell shows a constant decrease in OCV with a slope of 0.1/100 h. In contrast, the OCV of the infiltrated cell remained very stable throughout the test. The time-dependent changes of impedance spectra of the reference and the infiltrated cell are shown in Figures 2a and 2b. A significant increase in the total ASR was observed in all SOFCs within the first 3 h of sulfur exposure. The trends in the ohmic and non-ohmic portions of the cell impedance are summarized in Figures 2c and 2d. The ohmic portion of ASR was relatively unchanged in both cells for the duration of the testing, as shown in Figure 2c. The growth of the non-ohmic portion of the cell impedance, shown in Figure 2d, was largely responsible for the performance loss of the reference cell, which is in agreement with results in the literature2,6,10,13,28 and is dominated by the polarization resistance of the anode. The sharp initial increase in the nonohmic component of the ASR in all cells is likely the result of sulfur occupying catalytic sites and reducing effective TPB 1561

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials

Figure 3. SEM micrographs of SOFC anodes before testing: (a) reference cell; (b) GDC-infiltrated cell with GDC nanograins on nickel particle. After testing: (c) carbon-fouled anode of reference cell after sulfurized methane−hydrogen exposure for 70 h; (d) pristine anode of GDC-infiltrated cell after sulfurized methane−hydrogen exposure for 290 h.

Figure 4. EDS mapping of reference cell and GDC-infiltrated cell anodes following testing in sulfurized methane−hydrogen showing significant carbon deposition in the reference cell anode. Spectra for S, C, Ce, Gd, and Ni are shown.

shown in Figure 3. In the infiltrated anodes, deposited material was observed on the surface of exposed grains seen in Figure 3b. These features are GDC deposits from the infiltration and were not present in the reference cell shown in Figure 3a. Following testing in the sulfur-containing fuel, the anode of the reference cell (Figure 3c) contained a large amount of deposited material. EDS was employed to determine the composition. EDS measurements performed on the cross section of the cells after testing, shown in Figure 4, revealed a large contrast in the amount of carbon present and a small difference in sulfur between the reference and infiltrated cells. This finding indicates that sulfur poisoning of the anode will encourage carbon deposition, whereas the infiltrated anode had very little carbon buildup. Carbon growth on anode surfaces will occur when hydrocarbons are present, and there is insufficient oxygen available to oxidize both the hydrogen and carbon present in

length. The longer term degradation observed in the reference cell would then be the result of fouling and chemical degradation of the anode structure from sulfur and carbon. The degradation effect was seen after the first 5 h in the reference cell and caused an increase of 6.2%/h in the nonohmic ASR. The degradation phenomenon was not present in the infiltrated cells. The infiltrated cells also displayed an increase in the non-ohmic ASR initially but then remained virtually unchanged for the duration of the test following this initial change, suggesting that the nanoparticle infiltration strongly inhibits structural damage to the anode from sulfur. This observation supports the hypothesis of increased ceria− nickel contact area helping to remove surface sulfur through SO2 formation rapidly enough to prevent significant reactions with the anode materials and the resulting ASR increase. SEM analysis of the anode cross sections showed some notable differences in the microstructure of the various cells, as 1562

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials

Figure 5. Raman spectra of reference, infiltrated, and clean reference cell anodes. (a) Shifted ceria peak in infiltrated cell. (b) Significant carbon peak growth on unmodified anode.

Figure 6. Sulfur, cerium, carbon, and nickel XPS spectra: (a) reference cell; (b) infiltrated cell; (c) clean reference cell.

the fuel. Any available oxygen supplied by the cell will preferentially react with hydrogen, which leads to solid carbon being deposited within the anode structure. The resilience of the infiltrated anode to carbon deposition is a promising result for IT-SOFCs that are able to operate on commercially available fuels. Further investigation into the nature of the contaminant species in the cell anodes was performed using Raman spectroscopy and XPS. These surface analysis techniques were used to determine the form taken by carbon and sulfur as well as the state of the anode materials after fuel exposure. Figure 5 shows Raman spectra of reference, infiltrated, and clean (unsulfurized) reference cells. Figure 5a shows the cerium

oxide peak of each cell. The slight peak shift and broadening seen in the infiltrated cerium oxide spectra is attributed to the ceria nanoparticles from infiltration.29 The infiltrated and reference cells showed a stark contrast in the 1250−1750 cm−1 wavenumber range, shown in Figure 5b. Compared to a clean, unmodified cell exposed only to wet hydrogen, the infiltrated cell and reference cell exposed to sulfurized hydrocarbon fuel gained peaks corresponding to the carbon D-band at 1350 cm−1, and the reference cell exposed to sulfurized fuel gained a large peak corresponding to the carbon G-band at 1580 cm−1.30 The carbon G-band peak indicates carbon is present in the form of graphite on the anode surface.31,32 The carbon D-band seen in the spectra of both the infiltrated and reference cells 1563

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials exposed to sulfurized hydrocarbon fuel is caused by defects and impurities in carbon structures.31 In addition to these peaks, a small but sharp peak was present in all three samples at 1550 cm−1, though at varying intensity. This peak is caused by amorphous carbon and is likely partially caused by contamination from handling of the samples. These findings support the results from EDS measurements showing a significant difference in the carbon present on the anodes of the infiltrated and reference cells. In particular, the high-intensity carbon G-band peak (1550 cm−1) in the sulfurexposed reference cell further confirms that significant solid carbon formation for the non-ceria infiltrated anodes. XPS analysis of the same sample set used for Raman spectroscopy provided further insights into the state of the surface of each cell’s anode. Figures 6a, 6b, and 6c show the XPS spectra of S 2p, Ce 3d, C 1s, and Ni 2p on the reference, infiltrated, and clean cell, respectively. The clean reference cell contained no sulfur signal as expected and confirms that any sulfur present in the other cells is not due to sources external to the experiment. In contrast, both sulfur exposed cells with or without surface modification show the sulfur signals in the form of sulfate.33,34 Quantitative analysis of sulfur concentration on each cell shows that the sulfur concentration in the infiltrated cell was approximately half of the concentration in the reference. One of the XPS analysis findings for the anodes was the variation in cerium oxidation state between samples. Cerium 3+ and 4+ have distinct 3d spectra, and the prevailing oxidation state of cerium in the sample can be estimated based on the shape of the measured spectra. The cerium in the infiltrated cell was primarily Ce3+, identical to cerium in a cell that had never been exposed to sulfur, as indicated by the peak at binding energy ∼886 eV. In contrast, the reference cell that was exposed to sulfur contained more Ce4+, indicated by the strong peak at binding energy ∼884 eV.35 These results suggest that the addition of the ceria coating is preventing the lasting oxidation of the cerium present in the anode surface and is related to the increase of anode sulfur tolerance. The Ni 2p spectra show a clear difference in the relative surface ratio of metallic nickel (dark yellow, ∼853 eV)36 between the clean reference cell and the two cells exposed to sulfur. According to the spectra of both sulfur-exposed cells, multiple nickel species are copresent on the surface and the signals are overlapping such that we cannot accurately quantify the possible nickel−sulfur surface species. Another key difference between the XPS measurements of the reference and infiltrated cells was seen in the C 1s spectra. All three samples showed a peak at ∼285 eV, but only the reference cell possessed a second peak at ∼290 eV. The absence of the carbon 1s peak at binding energy ∼290 eV on the surface of the infiltrated cell reinforces the Raman spectroscopy and EDS findings that without the GDC nanoparticle coating sulfur poisoning makes the cell vulnerable to carbon deposition. TEM performed on untreated anode powder exposed to sulfurized methane showed a significant carbon presence on nickel grains, shown in Figure 7a. This further supports the findings from XPS and Raman spectroscopy that carbon formation is a major contributor to cell degradation under these conditions. Anode powder treated with the ceria solution was seen to be carbon free after an identical gas exposure, as shown in Figure 7b. The pristine nature of this surface after 24 h of sulfurized methane exposure suggests that the degradation mechanism that leads to coking does not occur when the

Figure 7. TEM images of (a) carbon growths on nickel grain in untreated anode powder following 24 h exposure to sulfurized methane-hydrogen and (b) pristine surface of GDC-infiltrated anode powder following 24 h exposure to sulfurized methane−hydrogen. Fuel: 40 ppm of H2S, 50:50 H2:CH4, T = 650 °C.

infiltration treatment has been applied. This suggests a passive mechanism of protection offered by the GDC coating in which the nano-GDC donates lattice oxygen to contaminant species that arrive at the surface. Figure 8 shows the conceptual schemes for the sulfur protection mechanism. Based on the experimental evidence on

Figure 8. Schematic of proposed poisoning and stability mechanisms: (a) Normal fuel oxidation reactions at TPB. (b) Adsorbed sulfur blocks TPB site, carbon deposition begins. (c) Fully deactivated anode with adsorbed sulfur and carbon growths. (d) High density of TPB sites and improved oxygen transport from nanoparticles prevents sulfur poisoning.

the aged cells, the presence of GDC nanoinfiltrants plays an important role of surface protection. The infiltrated GDC increases the density of reaction sites in the anode, mitigating the impact of sulfur adsorption. The high ionic conductivity of nanoparticle GDC allows a continuous pathway for oxygen to transport to the TPB sites, providing sufficient oxygen to oxidize the surface sulfur intermediates into gaseous SO2. The constant oxygen flux, pumped from cathode to anode while SOFC operating, is essential for preventing permanent anode degradation, caused by carbon growth and damage to the anode structure.

■

CONCLUSION Infiltrating a small wt % of GDC into porous Ni-GDC SOFC anodes significantly decreased the degradation caused by sulfur exposure. When exposed to a hydrogen/methane mixture containing 20 ppm of hydrogen sulfide, an unmodified SOFC 1564

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials

(9) Brightman, E.; Ivey, D. G.; Brett, D. J. L.; Brandon, N. P. The Effect of Current Density on H2S-Poisoning of Nickel-Based Solid Oxide Fuel Cell Anodes. J. Power Sources 2011, 196 (17), 7182−7187. (10) Riegraf, M.; Zekri, A.; Yurkiv, V.; Costa, R.; Schiller, G.; Friedrich, K. A. Assessment of Sulfur Poisoning of Ni/CGO-Based SOFC Anodes. ECS Trans. 2017, 77 (10), 149−156. (11) Kakaç, S.; Pramuanjaroenkij, A.; Zhou, X. Y. A Review of Numerical Modeling of Solid Oxide Fuel Cells. Int. J. Hydrogen Energy 2007, 32 (7), 761−786. (12) Zheng, L. L.; Wang, X.; Zhang, L.; Wang, J.-Y.; Jiang, S. P. Effect of Pd-Impregnation on Performance, Sulfur Poisoning and Tolerance of Ni/GDC Anode of Solid Oxide Fuel Cells. Int. J. Hydrogen Energy 2012, 37 (13), 10299−10310. (13) Zhang, L.; Jiang, S. P.; He, H. Q.; Chen, X.; Ma, J.; Song, X. C. A Comparative Study of H2S Poisoning on Electrode Behavior of Ni/ YSZ and Ni/GDC Anodes of Solid Oxide Fuel Cells. Int. J. Hydrogen Energy 2010, 35 (22), 12359−12368. (14) Gong, M.; Liu, X.; Trembly, J.; Johnson, C. Sulfur-Tolerant Anode Materials for Solid Oxide Fuel Cell Application. J. Power Sources 2007, 168 (2), 289−298. (15) Yun, J. W.; Yoon, S. P.; Han, J.; Park, S.; Kim, H. S.; Nam, S. W. Ceria Coatings Effect on H2S Poisoning of Ni/YSZ Anodes for Solid Oxide Fuel Cells. J. Electrochem. Soc. 2010, 157 (12), B1825−B1830. (16) Ding, D.; Li, X.; Lai, Y.; Liu, M. Enhancing SOFC Cathode Performance by Surface Modification through Infiltration. Energy Environ. Sci. 2014, 7, 552−575. (17) Hussain, A. M.; Høgh, J. V. T.; Jacobsen, T.; Bonanos, N. Nickel-Ceria Infiltrated Nb-Doped SrTiO3 for Low Temperature SOFC Anodes and Analysis on Gas Diffusion Impedance. Int. J. Hydrogen Energy 2012, 37 (5), 4309−4318. (18) Hussain, A. M.; Hogh, J. V. T.; Zhang, W.; Bonanos, N. Efficient Ceramic Anodes Infiltrated with Binary and Ternary Electrocatalysts for SOFCs Operating at Low Temperatures. J. Power Sources 2012, 216, 308−313. (19) Hussain, A. M.; Høgh, J. V. T.; Zhang, W.; Blennow, P.; Bonanos, N.; Boukamp, B. A. Effective Improvement of Interface Modified Strontium Titanate Based Solid Oxide Fuel Cell Anodes by Infiltration with Nano-Sized Palladium and Gadolinium-Doped Cerium Oxide. Electrochim. Acta 2013, 113, 635−643. (20) Kurokawa, H.; Sholklapper, T. Z.; Jacobson, C. P.; De Jonghe, L. C.; Visco, S. J. Ceria Nanocoating for Sulfur Tolerant Ni-Based Anodes of Solid Oxide Fuel Cells. Electrochem. Solid-State Lett. 2007, 10 (9), B135−B138. (21) Chen, Y.; Bunch, J.; Jin, C.; Yang, C.; Chen, F. Performance Enhancements of Ni-YSZ Electrode by Impregnation of Mo0.1Ce0.9O2+d. J. Power Sources 2012, 204, 40−45. (22) Blinn, K. S.; Liu, M. BaZr0.9Yb0.1O3_d-Modified Bi-Electrode Supported Solid Oxide Fuel Cells with Enhanced Coking and Sulfur Tolerance. J. Power Sources 2013, 243, 24−28. (23) Liu, Y.; Wang, F.; Chi, B.; Pu, J.; Jian, L.; Jiang, S. P. A Stability Study of Impregnated LSCF-GDC Composite Cathodes of Solid Oxide Fuel Cells. J. Alloys Compd. 2013, 578, 37−43. (24) Liu, Z.; Liu, M.; et al. LSM-Infiltrated LSCF Cathodes for Solid Oxide Fuel Cells. J. Energy Chem. 2013, 22 (4), 555−559. (25) Bartholomew, C. H. Mechanisms of Catalyst Deactivation. Appl. Catal., A 2001, 212 (1−2), 17−60. (26) Mirfakhraei, B.; Paulson, S.; Thangadurai, V.; Birss, V. Enhanced Hydrogen Oxidation Activity and H2S Tolerance of Ni-Infiltrated Ceria Solid Oxide Fuel Cell Anodes. J. Power Sources 2013, 243, 95− 101. (27) Yokokawa, H.; Tu, H.; Iwanschitz, B.; Mai, A. Fundamental Mechanisms Limiting Solid Oxide Fuel Cell Durability. J. Power Sources 2008, 182 (2), 400−412. (28) Huang, Q. A.; Hui, R.; Wang, B.; Zhang, J. A Review of AC Impedance Modeling and Validation in SOFC Diagnosis. Electrochim. Acta 2007, 52 (28), 8144−8164. (29) Choi, H. C.; Jung, Y. M.; Kim, S. B. Size Effects in the Raman Spectra of TiO2 Nanoparticles. Vib. Spectrosc. 2005, 37 (1), 33−38.

became nonfunctional after 70 h while the infiltrated SOFC operated stably for over 290 h. The increase in polarization resistance associated with anode poisoning was far smaller and did not grow over time in the infiltrated cell. The negative impact of sulfur exposure manifested as two distinct degradation mechanisms. These were identified as adsorption of sulfur on triple-phase boundaries and carbon buildup on the anode surfaces. In this case, coking results in more damage to the anode due to the high amount of carbon in the fuel relative to the sulfur content. However, sulfur occupying a high number of anode reaction sites both reduces performance and allows coking to proceed and is therefore of greater concern for cell health. The infiltration prevented carbon buildup by providing improved oxygen ion transport to the surface which promoted the removal of sulfur and carbon via SO2 and CO2 formation so that the concentration of contaminants remained low.

■

AUTHOR INFORMATION

Corresponding Author

*(E.D.W.) E-mail [email protected]. ORCID

Yi-Lin Huang: 0000-0002-1886-3352 Eric D. Wachsman: 0000-0002-0667-1927 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding for this work was provided by the National Energy Technology Laboratory (contract #4319920) through subcontract to Redox Power Systems. We acknowledge the support of the Maryland NanoCenter and its AIMLab for the use of the SEM and TEM used in this study. We additionally acknowledge Dr. Karen Gaskell in the Surface Analysis Center of the UMD Department of Chemistry and Biochemistry for assistance in obtaining XPS and Raman spectroscopy data.

■

REFERENCES

(1) Fellah, M. F. Adsorption of Hydrogen Sulfide as Initial Step of H2S Removal: A DFT Study on Metal Exchanged ZSM-12 Clusters. Fuel Process. Technol. 2016, 144, 191−196. (2) Sasaki, K.; Susuki, K.; Iyoshi, A.; Uchimura, M.; Imamura, N.; Kusaba, H.; Teraoka, Y.; Fuchino, H.; Tsujimoto, K.; Uchida, Y.; Jingo, N. H2S Poisoning of Solid Oxide Fuel Cells. J. Electrochem. Soc. 2006, 153 (11), A2023−A2029. (3) Eguchi, K. Fuel Flexibility in Power Generation by Solid Oxide Fuel Cells. Solid State Ionics 2002, 152−153, 411−416. (4) de Souza, S. Reduced-Temperature Solid Oxide Fuel Cell Based on YSZ Thin-Film Electrolyte. J. Electrochem. Soc. 1997, 144 (3), L35− L37. (5) Takeguchi, T.; Kani, Y.; Yano, T.; Kikuchi, R.; Eguchi, K.; Tsujimoto, K.; Uchida, Y.; Ueno, A.; Omoshiki, K.; Aizawa, M. Study on Steam Reforming of CH4 and C2 Hydrocarbons and Carbon Deposition on Ni-YSZ Cermets. J. Power Sources 2002, 112 (2), 588− 595. (6) Weber, A.; Dierickx, S.; Russner, N.; Ivers-tiffée, E. Sulfur Poisoning of Ni-Based SOFC-Anodes − Short and Long Term Behavior. ECS Trans. 2017, 77 (10), 141−147. (7) Offer, G. J.; Mermelstein, J.; Brightman, E.; Brandon, N. P. Thermodynamics and Kinetics of the Interaction of Carbon and Sulfur with Solid Oxide Fuel Cell Anodes. J. Am. Ceram. Soc. 2009, 92 (4), 763−780. (8) Cheng, Z.; Zha, S.; Liu, M. Influence of Cell Voltage and Current on Sulfur Poisoning Behavior of Solid Oxide Fuel Cells. J. Power Sources 2007, 172 (2), 688−693. 1565

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566

Article

ACS Applied Energy Materials (30) Li, X.; Liu, M.; Lee, J.; Ding, D.; Bottomley, L. A.; Park, S.; Liu, M. An Operando Surface Enhanced Raman Spectroscopy (SERS) Study of Carbon Deposition on SOFC Anodes. Phys. Chem. Chem. Phys. 2015, 17 (33), 21112−21119. (31) Dupuis, A. Characterizing Carbon Materials with Raman Spectroscopy. Prog. Mater. Sci. 2005, 50, 929−961. (32) Bose, S. M.; Gayen, S.; Behera, S. N. Theory of the Tangential G-Band Feature in the Raman Spectra of Metallic Carbon Nanotubes. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72 (15), 153402. (33) Legrand, D. L.; Nesbitt, H. W.; Bancroft, G. M. X-Ray Photoelectron Spectroscopic Study of a Pristine Millerite (NiS) Surface and the Effect of Air and Water Oxidation. Am. Mineral. 1998, 83, 1256−1265. (34) Siriwardane, R. V.; Poston, J. A., Jr.; Fisher, E. P.; Shen, M. S.; Miltz, A. L. Decomposition of the Sulfates of Copper, Iron(II), Iron(III), Nickel, and Zinc: XPS, SEM, DRIFTS, XRD, and TGA Study. Appl. Surf. Sci. 1999, 152 (3-4), 219−236. (35) Truffault, L.; Ta, M. T.; Devers, T.; Konstantinov, K.; Harel, V.; Simmonard, C.; Andreazza, C.; Nevirkovets, I. P.; Pineau, A.; Veron, O.; Blondeau, J. P. Application of Nanostructured Ca Doped CeO2 for Ultraviolet Filtration. Mater. Res. Bull. 2010, 45 (5), 527−535. (36) Rodriguez, J. A.; Li, S. Y.; Hrbek, J.; Huang, H. H.; Xu, G.-Q. The Interaction of Ni and Fe with Sulfur and Molybdenum-Sulfide Surfaces: A TDS, XPS and Hydrogen-Chemisorption Study. Surf. Sci. 1997, 370 (1), 85−95.

1566

DOI: 10.1021/acsaem.7b00354 ACS Appl. Energy Mater. 2018, 1, 1559−1566