Cobalt Alloying Effect on Improvement of Ni/YSZ Anode-Supported Solid Oxide Fuel Cell Operating with Dry Methane

Abstract

Although the cermet consisting of nickel and zirconia is useful as the solid oxide fuel cell anode, the conventional cermet consisting of nickel and yttria-stabilized zirconia (Ni/YSZ) is prone to be deactivated for a direct supply of hydrocarbon fuels. This paper describes a method to retard the deactivation of the Ni/YSZ-based cermet anode by being modified with appropriate incorporation of Co as the alloying element into Ni. The single-type anode-supported solid oxide fuel cell (SOFC) consisting of the Ni_0.75Co_0.25/YSZ cermet anode, YSZ film electrolyte, and La_0.8Sr_0.2MnO₃ cathode exhibited a prolonged stable SOFC performance at 750°C for dry methane (CH₄) feeding compared to the one using the Ni/YSZ anode. The microstructural modification of the cermet anode caused a carbon tolerance effect against CH₄ accompanied by a low anodic polarization resistance.

1. Introduction

The power generation by burning fossil fuels is well known for releasing a large amount of carbon dioxide into the atmosphere that causes global warming. The overall power-generation efficiency at a thermal power plant provided principally by the Carnot cycle, which converts heat energy into electric energy, will not be sufficient in our future sustainable society. Solid oxide fuel cells (SOFCs) directly converting fuel to electricity are attractive due to their high-power generation efficiencies, environmental friendliness, and flexibility for fuels that can use hydrogen or hydrocarbons.^1–4) Although the electric energy generated by a hydrogen fuel cell is highly attractive in the light of water emission as the exhaust with no environmental impact, the storage and supply systems for hydrogen have not been well established. Directly feeding gaseous hydrocarbon fuels to SOFCs without an external reforming process has been thus recently studied.^4–9) One of the most beneficial alternative fuels for SOFCs is methane (CH₄), the major component of natural gas, coal-bed gas, and biogas and its high overall fuel efficiency.^2,10–13) The high operating temperatures of SOFCs from about 700°C to about 1000°C provide a spontaneous thermal decomposition of CH₄, which can make a supplied CH₄ fuel favorable for the electrochemical oxidations by the oxide ions transported through the electrolyte; the typical reactions taking place at the anode according to eqs. (1)–(5):   

\begin{equation} \text{CH$_{4}$} \rightarrow \text{C} + \text{4H$_{\text{ad}}$} \end{equation} (1)

  

\begin{equation} \text{4H$_{\text{ad}}$} \rightarrow \text{2H$_{2}$} \end{equation} (2)

  

\begin{equation} \text{4H$_{\text{ad}}$} + \text{2O$^{2-}$} \rightarrow \text{2H$_{2}$O} + \text{4e$^{-}$} \end{equation} (3)

  

\begin{equation} \text{C} + \text{O$^{2-}$} \rightarrow \text{CO} + \text{2e$^{-}$} \end{equation} (4)

  

\begin{equation} \text{CO} + \text{O$^{2-}$} \rightarrow \text{CO$_{2}$} + \text{2e$^{-}$} \end{equation} (5)

where H_ad means an adsorbed hydrogen atom on the metal surface of the cermet anode.

The high operating temperatures result in a problem of carbon deposition at the anode, and the subsequent deactivation leads to a loss of SOFC performance and low durability.^14–16) The mixture of nickel and yttria-stabilized zirconia (Ni/YSZ) is the most common SOFC cermet anode due to its high electrocatalytic activity for hydrogen oxidation as well as its established fabrication process.¹⁷⁾ Conventional SOFCs using the Ni/YSZ cermet anode, however, show a rapid degradation with operating time by directly feeding CH₄, which is caused by the growth and accumulation of deposited carbon on the Ni particle surface of the anode according to eq. (1). The carbon deposition can easily take place because the rate of the electrochemical oxidation of carbon (eq. (4)) is lower than that of the thermal decomposition of CH₄ (eq. (1)).¹⁴⁾ The carbon atoms assemble to form graphite and deposit on the surface of the Ni particles, and result in a deactivation of the anode.^18–21)

Modification of the Ni/YSZ anode for SOFCs can improve their long-term stability for hydrocarbon fuels.^22–24) The steam reforming of hydrocarbon fuels is thermodynamically favorable for preventing carbon deposition in SOFCs;²⁵⁾ however, the Ni/YSZ anode can be deactivated by high humidity content when internal steam reforming is employed, and the cell system needs to be accordingly more complex.^26,27) Adding an alloying element to Ni combined with the internal steam reforming is an alternative way to suppress the carbon deposition. The Ni_0.99Au_0.01/YSZ cermet was tolerant of carbon deposits at the ratio of steam to CH₄ as high as 3 in the temperature range between 700°C and 900°C.²⁸⁾ The addition of Au causes the Ni_1−xAu_x particles coarsening, leading to a decrease in the catalytically active surface sites.²⁹⁾ The addition of a low carbon solubility metal such as Ru, Rh, and Cu as an alloying component to Ni in the SOFC anode can delay carbon deposition primarily for the direct use of CH₄.^30–36) Ru and Rh are unfortunately rare and expensive metals for commercial SOFCs. Although Cu seems to be promising as the alternative alloying component, the Cu diffusion into the YSZ frame at an operating temperature causes a degradation of the Ni–Cu/YSZ cermet.^36–38) Alloying of Fe in the Ni-based anode has also been widely investigated due to its good catalytic activity for the electrochemical oxidation of hydrocarbon fuels.^39–41) By contrast, the anodic overpotential of the Ni-based anode is increased by Fe alloying,⁴²⁾ and a high content of Fe in the Ni_1−xFe_x anode provides a significant carbon deposition.⁴³⁾ A possibility of using Co combined with stabilized ZrO₂ as the cermet anode was briefly pointed out instead of using Ni.⁴⁴⁾ Separately, the cermet anodes consisting of Cu_1−xCo_x alloys and YSZ-based electrolytes are effective in retarding carbon deposition in hydrocarbon-fueled SOFCs.^45,46) A favorable effect of the Co addition on the electrochemical activity of the Ni/YSZ anode connected with a YSZ electrolyte is reported mainly for H₂ fuels.^47–49) The carbon deposition can be reduced by Co alloying in the Ni/Ce_0.8Gd_0.2O_1.9 anode for hydrocarbon fuels, accompanied by a decrease in the polarization resistance.⁵⁰⁾

Effects of Co alloying on the electrochemical performance of the Ni/YSZ anodes for H₂ and hydrocarbon fuels have been reported.^47–49) However, studies on the prolonged performance stability for CH₄ in an anode-supported single cell consisting of Ni_1−xCo_x/YSZ anode, YSZ-membrane electrolyte, and La_0.8Sr_0.2MnO₃ (LSM) cathode still lack because of the limited evaluations of the cell performance mainly on the electrolyte-supported single cells. We recently reported that the Ni_0.75Co_0.25/YSZ anode combined with a YSZ electrolyte disk shows the highest cell performance for a direct supply of CH₄ and the lowest carbon deposition among the cells investigated.⁵¹⁾ In particular, the increase in the maximum power density of this electrolyte-supported cell using the Ni_0.75Co_0.25/YSZ anode for CH₄ was much higher than that for H₂ at 750°C compared with the one using the Ni/YSZ anode. We, therefore, have focused on an anode-supported cell consisting of the Ni_0.75Co_0.25/YSZ anode, YSZ membrane, and LSM cathode, which is promising to provide a higher cell performance than the electrolyte-supported cell owing to a significant reduction in the ohmic resistance of the electrolyte. In this study, we report comparisons of the prolonged stability in the cell performance for CH₄ and H₂ and the morphology of the cermet anode between the cell using the Ni/YSZ anode and the one using the Ni_0.75Co_0.25/YSZ anode.

2. Materials and Methods

Ni_1−xCo_xO (x = 0 and 0.25) precursors were prepared by dissolving Ni(NO₃)₂·6H₂O (98% purity, Nacalai Tesque) and Co(NO₃)₃·9H₂O (98% purity, Nacalai Tesque) at appropriate molar ratios in deionized water at room temperature. Each solution was mixed with citric acid anhydrous (C₆H₈O₇) (99% purity, Nacalai Tesque) to produce a mixed solution of citric acid and the Ni_1−xCo_x ions at the molar ratio of 2 to 1. Subsequently, the YSZ powder (8 mol%Y₂O₃–ZrO₂(TZ-8Y), Tosoh) was added into the solution under a continuously vigorous stirring at room temperature; the weight ratio of Ni_1−xCo_xO to YSZ was 3:2. After stirring the solution for 2 h, it was heated to 180°C with continuous stirring until the solution became a dark viscous gel by dehydration. The gel was further heated to 250°C to obtain a lump of loose dark ash. The ash was calcined at 800°C for 5 h in the air to remove the remaining hydrocarbon residues. This calcination left uniformly mixed Ni_1−xCo_xO and YSZ particles.

The prepared Ni_1−xCo_xO and YSZ particles were mixed with cornstarch as a pore former and stearic acid (95% purity, Sigma-Aldrich) as a binder, and they were thoroughly ground with a pestle in an agate mortar added with an appropriate amount of ethanol. The ratio of cornstarch to the Ni_1−xCo_xO and YSZ was 15 mass%, and that of stearic acid was 5 mass%. The ground powder was kept at 80°C, and 0.5 g of the powder filled in a molding die with an inner diameter of 16.5 mm was uniaxially pressed at 15 kN to make an anode pellet. The pellet was subsequently subjected to a two-step pre-sintering at 800°C for 2 h and 1100°C for 3 h at the heating and cooling rates of 100 K/h in every step. One side of the pre-sintered anode pellet was painted with the YSZ slurry, which was prepared by mixing 50 mass% YSZ powder and 50 mass% ethyl-cellulose co-sintered at 1400°C for 10 h to densify the electrolyte and pre-sintered anode cermet. The thickness of the sintered YSZ membrane layer was approximately 30 µm. The slurry of 70 mass% LSM (Seimi Chemical) and 30 mass% YSZ powder mixture with glycerol was coated on the YSZ membrane surface with the geometrical area of 0.28 cm² with approximately 40 µm thick. A circular Pt mesh (#100) with a 6 mm diameter spot-welded with two 0.3 mm thick Pt wires was attached to the anode surface and the cathode surface as the current collector. The prepared cells were fired at 1200°C for 3 h to obtain sufficient contact between the cathode and electrolyte.

The prepared cell, denoted as the Ni_1−xCo_x-YSZ cell, was heated at 850°C for 30 min to soften Pyrex^® glass rings used to seal the outer edge of the anode face and that of the cathode face, each with an alumina tube edge. The cell anode was reduced by H₂ at 800°C for 2 h before the cell performance evaluation. H₂ and O₂ were supplied as fuel and oxidant, respectively, at a flow rate of 20 cm³·min⁻¹. The cell was activated preliminary by passing a certain amount of electric current (preloading), which provided the cell voltage about 0.2 V for 30 min. Subsequently, the temperature was reduced to 750°C to measure the electrochemical performance using current-voltage (I-V) characteristics. After the cell performance measurement by supplying H₂, the fuel gas was switched to CH₄ (20 vol%CH₄ in He) with a total flow rate of 20 cm³·min⁻¹, and the same measurement was conducted. AC-impedance spectra of the prepared cells were recorded with a frequency response analyzer (FRA5097, NF Corp.) at the open-circuit voltage (OCV) for both the H₂ and CH₄ supplies. The frequency range was from 0.1 Hz to 1 MHz. The cell performance stability for CH₄ was investigated at a constant current density of 0.2 A for 240 h at 750°C. The electrochemical performance and its stability of the prepared cells were examined twice to confirm reproducibility.

Microstructural observations of the samples were made with scanning electron microscopes (SEM, SU8000 and TM3000, Hitachi High-Tech). The carbon deposition to the surface of the Ni_1−xCo_x/YSZ anode after the performance stability test supplied with CH₄ was investigated with the SU8000 combined with an energy-dispersive x-ray spectrometer (EDS, X-Max^N 80, Oxford Instruments). The sample surface was sputter-coated with Pt to provide electrical conductivity. Phase identification of the prepared powders was made by x-ray diffraction (XRD, RINT-2200, Rigaku) with the monochromated Cu-Kα emission powered at 40 kV and 30 mA. Surface areas of the prepared powders were determined using the Brunauer-Emmett-Teller (BET) method by N₂ adsorption analysis (Tristar 3000, Shimadzu).

3. Results and Discussion

3.1 Phase identification

Figure 1 shows the XRD patterns of the Ni_1−xCo_xO-YSZ (x = 0 and 0.25) powders, which were pre-sintered at 800°C for 3 h. The identification of the face-centered cubic (fcc) Ni_1−xCo_xO phase indicates complete decomposition of the nitrate ash into the metal oxides. The XRD patterns consisting of diffraction peaks from the Ni_1−xCo_xO and YSZ phases imply no appreciable reaction between the Ni_1−xCo_xO and YSZ particles. The diffraction peaks from the Ni_0.75Co_0.25O phase appeared at lower diffraction angles than those from the NiO phase; this indicates an expansion of the lattice by incorporating Co²⁺ ions into the NiO lattice, as shown in the inset in Fig. 1. The larger ionic radius of Co²⁺ (0.074 nm) than that of Ni²⁺ (0.069 nm)⁵²⁾ accounts for the lattice expansion by the partial substitution of Co²⁺ ions for Ni²⁺ ions in the NiO lattice.^53,54)

Fig. 1

XRD patterns of the (a) NiO-YSZ and (b) Ni_0.75Co_0.25O-YSZ pre-sintered at 800°C for 5 h (○: Ni_1−xCo_xO, ▼: YSZ).

Figure 2 shows the XRD patterns of the reduced NiO-YSZ and Ni_0.25Co_0.75O-YSZ pellets by H₂ at 800°C for 2 h. The prepared pellet itself is denoted as the Ni_1−xCo_xO-YSZ pellet here. The pellets were sintered at 1400°C in the air as described in the previous section. The XRD measurement was conducted for the powdered samples by grinding the pellets after the H₂ reduction. The XRD patterns reveal that the H₂ reduction of the NiO-YSZ and Ni_0.75Co_0.25O-YSZ pellets provided the Ni-YSZ and Ni_0.75Co_0.25-YSZ pellets, respectively. A comparison of the 220 reflection between the Ni and Ni_0.75Co_0.25 phases in Fig. 2 shows that the Ni_0.75Co_0.25 diffraction peaks have a lower intensity and broader peak shape than the Ni diffraction peaks. This difference can be interpreted as either insufficient crystallization or coarse grain formation in the Ni_0.75Co_0.25 alloy. The diffraction peaks of the Ni_0.75Co_0.25 phase appeared at lower diffraction angles than those of the Ni phase, implying that the unit cell size of the fcc Ni_0.75Co_0.25 alloy is larger than that of the fcc Ni metal. The difference between the atomic radius of Co (0.152 nm) and Ni (0.149 nm)⁵⁵⁾ can account for this result. In contrast, the YSZ diffraction peak intensities in the Ni_0.75Co_0.25-YSZ powder are higher than those in the Ni-YSZ powder; the 311 reflection from the YSZ phase, for example, exhibited separated Kα₁ and Kα₂ peaks indicating formation of well-crystallized YSZ grains in the Ni_0.75Co_0.25-YSZ pellet. The calcination of the dehydrated Co-nitrate precursor at 800°C resulted in forming the CoO and Co₃O₄ phases. The Co₃O₄ phase disappeared after subsequent heating the Ni_0.75Co_0.25O-YSZ pellet at 1100°C and 1400°C in the air. Since Co₃O₄ melts at 895°C and decomposes into CoO at about 950°C, Co₃O₄ probably acted as a sintering aid⁵⁶⁾ for the pressed pellet consisting of Ni_0.75Co_0.25O and YSZ.

Fig. 2

XRD patterns of the (a) Ni-YSZ and (b) Ni_0.75Co_0.25-YSZ prepared by the H₂ reduction (○: Ni_1−xCo_x, ▼: YSZ). The intensity of (a) is enlarged by 1.4 times higher than (b) to normalize the YSZ-peak height.

3.2 Microstructural characterization

Figure 3 shows cross-sectional SEM images of the Ni-YSZ and Ni_0.75Co_0.25-YSZ cells and their corresponding elemental mapping images. The Ni/YSZ and Ni_0.75Co_0.25/YSZ anodes exhibited a porous structure, which allowed a sufficient fuel gas permeation to the electrolyte layer. Figures 3(a) and 3(b) show that the grain size of Ni particles in the Ni/YSZ anode is smaller than that of Ni_0.75Co_0.25 particles in the Ni_0.75Co_0.25/YSZ anode. A comparison of the SEM images between the Ni/YSZ and Ni_0.75Co_0.25/YSZ anodes indicates that the Co addition enhanced the grain growth of the anode components during the co-sintering the anode and electrolyte. This sintering enhancement agrees with the results reported in obtaining a dense YSZ bulk by sintering with a Co₃O₄ additive.^56–58) Specific surface areas of the powdered Ni/YSZ and Ni_0.75Co_0.25/YSZ anodes were 0.34 m²g⁻¹ and 0.19 m²g⁻¹, respectively, indicating that the grain growth occurred in the Ni_0.75Co_0.25/YSZ anode is more prominent than that in the Ni/YSZ anode. Figure 3(c) shows a cross-sectional image of the Ni_0.75Co_0.25-YSZ cell, exhibiting well-bonded three layers, the anode, electrolyte, and cathode.

Fig. 3

Cross sectional SEM images and EDX mapping near the anode-electrolyte interface of (a) the Ni-YSZ cell, that of (b) the Ni_0.75Co_0.25-YSZ cell, and (c) the cross-section of the Ni_0.75Co_0.25-YSZ cell.

3.3 Electrochemical performance

Figure 4 shows a comparison of the current-voltage (I-V) and current-power (I-P) characteristics and the impedance spectra for H₂ between the Ni-YSZ and Ni_0.75Co_0.25-YSZ cells. The I-V characteristic of the Ni-YSZ cell was higher than that of the Ni_0.75Co_0.25-YSZ cell. The maximum power density (P_max) of the Ni-YSZ cell was 0.85 W·cm⁻², whereas P_max of the Ni_0.75Co_0.25-YSZ cell was 0.67 W·cm⁻². The comparison of the impedance spectra revealed that the impedance-arc intercept on the x-axis of the Ni-YSZ cell, which corresponds to the polarization resistance (Rp), was smaller than that of the Ni_0.75Co_0.25-YSZ cell. In contrast, the point of intersection on the x-axis at high frequencies, which corresponds to the ohmic resistance (Ro), did not show an appreciable difference between the Ni-YSZ and Ni_0.75Co_0.25-YSZ cells. These results show that the anodic overpotential causing the voltage drop of the Ni-YSZ cell is lower than that of the Ni_0.75Co_0.25-YSZ cell. The lower anodic overpotential of the Ni-YSZ cell is probably due to the higher specific surface area providing more active reaction sites for the anodic oxidation of H₂.

Fig. 4

(a) The cell voltage (open symbols) and power density (closed symbols) as a function of the current density, and (b) impedance spectra of the Ni_1−xCo_x-YSZ cells measured at 750°C in H₂.

Figure 5 shows the same comparison for CH₄ measured immediately after changing the fuel from H₂ to CH₄. The Ni-YSZ cell showed a voltage drop larger than the Ni_0.75Co_0.25-YSZ cell as a function of the current up to about 1.5 A·cm⁻², as shown in the I-V curves in Fig. 5(a). Conversely, the further voltage drop in the Ni-YSZ cell was smaller than that in the Ni_0.75Co_0.25-YSZ cell; the overall overpotential in the Ni-YSZ cell was lower than that in the Ni_0.75Co_0.25-YSZ cell at high current densities. In total, P_max of the Ni_0.75Co_0.25-YSZ cell (= 0.88 W·cm⁻²) decreased by 10% compared to P_max of the Ni-YSZ cell (= 0.98 W·cm⁻²). By contrast, the corresponding decrease in P_max was 22% for H₂, as shown in Fig. 4(a). This difference in the percentage of P_max decrease agrees with the difference in Rp for H₂ and CH₄ between these two cells, as shown in Figs. 4(b) and 5(b). The cell activation potential, the sum of the activation overpotentials at the anode and cathode, is caused by charge-transfer reactions at both electrodes, and it is predominant in the low current-density region causing a voltage drop from OCV. Since we used the same cathode material for the Ni_1−xCo_x-YSZ cells, the lower voltage-drop in the Ni_0.75Co_0.25-YSZ cell is probably due to an electrocatalytic-activity enhancement for the electrochemical oxidation of CH₄ at the anode.

Fig. 5

(a) The cell voltage (open symbols) and power density (closed symbols) as a function of the current density and (b) AC-impedance spectra of the Ni_1−xCo_x-YSZ cells at 750°C in CH₄.

The stepwise bond-dissociation reactions of CH₄ are expressed as follows:   

\begin{equation} \text{CH$_{4}$ (g)} \rightarrow \text{CH$_{3}{}^{*}$} + \text{H$^{*}$} \end{equation} (6)

  

\begin{equation} \text{CH$_{3}$ (g)} \rightarrow \text{CH$_{2}{}^{*}$} + \text{H$^{*}$} \end{equation} (7)

  

\begin{equation} \text{CH$_{2}$ (g)} \rightarrow \text{CH$^{*}$} + \text{H$^{*}$} \end{equation} (8)

  

\begin{equation} \text{CH (g)} \rightarrow \text{C$^{*}$} + \text{H$^{*}$} \end{equation} (9)

Here, the symbols (g) and ^* mean the gaseous and radical states, respectively. Recently, the density functional theory (DFT) method has been used to reveal catalytic reaction mechanisms occurring at the catalyst surface and the stepwise bond dissociation energies of adsorbed molecules. The activation energy (E_a) of each reaction in the CH₄ dissociation dominates the rate-determining step. The DFT method revealed that the Ni(111) surface provides the rate-determining step as eq. (9) for the CH₄ dissociation, whereas the Ni–Co(111) surface provides it as eq. (6).⁵⁹⁾ This DFT calculation shows that E_a of the CH₄ dissociation for Ni(111) is 1.36 eV (= 131 kJ·mol⁻¹) and that for Co(111) is 1.29 eV (= 124 kJ·mol⁻¹). These calculation results suggest that the Co addition as an alloying element to the Ni-YSZ cermet can decrease the CH₄ dissociation energy and impede the carbon deposition according to eq. (9). This favorable effect by the Co addition accounts for the lower activation overpotential in the Ni_0.75Co_0.25-YSZ cell, as shown in Fig. 5(a).

3.4 Prolonged cell-performance stability in CH₄

We compared the prolonged cell-performance stability for dry CH₄ feeding at 750°C between the Ni-YSZ and Ni_0.75Co_0.25-YSZ cells. The current was fixed at 0.2 A (= 0.71 A·cm⁻²) at 750°C; we assumed that the anodic overpotential is predominant at this current. Figure 6 shows variations of the terminal voltage measured at 0.5 A (= 1.77 A·cm⁻²), Ro, and Rp as a function of time. The terminal voltage of the Ni-YSZ cell decreased from 0.51 V to 0.30 V for 240 h, whereas that of the Ni_0.75Co_0.25-YSZ cell decreased from 0.49 V to 0.47 V. The smaller terminal-voltage decrease of the Ni_0.75Co_0.25-YSZ cell shows improved cell-performance stability. This improvement corresponded to a smaller increase in Rp of the Ni_0.75Co_0.25-YSZ cell, ΔRp = 0.28 Ω·cm², for 240 h. In contrast, the increase in Rp of the Ni-YSZ cell was ΔRp = 0.77 Ω·cm². Rp of the Ni-YSZ cell decreased slightly to about 25 h, but it increased monotonically after that. This result showed the same behavior in at least two measurements for the Ni-YSZ cell, suggesting that the initial carbon deposition, according to eq. (9), decreases the polarization resistance of the Ni/YSZ anode from covering the surface of Ni particles. The further supply of CH₄ to the Ni/YSZ anode probably resulted in a significant amount of carbon deposition accompanying the cell performance decrease owing to reducing the number of active sites for the anodic oxidation of CH₄. This behavior is similar to the result for the Ni/YSZ cermet exposed to CH₄ reporting that carbon deposition is promoted in the operation at 800°C near OCV with increasing the cermet layer thickness.⁶⁰⁾ The Ni-YSZ and Ni_0.75Co_0.25-YSZ cells showed almost the same and constant ohmic resistance, Ro, at least for 240 h. This result suggests that there were no appreciable microstructural changes of the anode and the interface contact between the anode and the electrolyte layer during the cell-performance stability test.

Fig. 6

The resistances and terminal voltage at a current of 0.5 A as a function of time at 750°C in CH₄ for the Ni_1−xCo_x-YSZ cells.

Figure 7 shows the I-V curves, P-V curves, and impedance spectra measured at 0 h, 120 h, and 240 h. The cell performance of the Ni-YSZ cell significantly decreased with time; P_max decreased by 40% for 240 h, whereas P_max of the Ni_0.75Co_0.25-YSZ cell decreased by 5%. The I-V curves of the Ni-YSZ cell exhibited a significant decrease with time compared to those of the Ni_0.75Co_0.25-YSZ cell, as shown in Fig. 7(a). The difference in this cell-performance decrease between these two cells corresponds to the increase in Rp, as shown in Fig. 7(c). These results suggest that the Ni_0.75Co_0.25/YSZ anode is more tolerant in carbon deposition than the Ni/YSZ anode to the exposure of CH₄.

Fig. 7

(a) I-V curves, (b) P-V curves and (c) AC-impedance spectra of the Ni-YSZ and Ni_0.25Co_0.75-YSZ cells at 750°C in CH₄ for 0 h, 120 h, and 240 h.

Figure 8 shows EDX spectra obtained from the surfaces of the Ni/YSZ and Ni_0.75Co_0.25/YSZ anodes exposed to CH₄ for 240 h. The SEM images show the whole analyzed areas. The carbon peak intensity that appeared in the Ni/YSZ anode was higher than that in the Ni_0.75Co_0.25/YSZ anode, indicating that the carbon deposition was reduced by using the Ni_0.75Co_0.25/YSZ anode. Carbon can be formed at the Ni/YSZ cermet by thermal decomposition of CH₄ according to eq. (1), which proceeds more favorably at temperatures higher than about 700°C on the Ni metal surface.^14–16) The carbon reacts with O²⁻ ions transported through the electrolyte and produce CO according to eq. (4), which contributes to suppress the carbon deposition. The produced CO, however, strongly adsorbs to the Ni metal surface resulting in deactivation of the anodic oxidation reactions. The desorption energy of CO on the Co metal surface (161 kJ·mol⁻¹) was lower than that on the Ni metal surface (173 kJ·mol⁻¹).⁶¹⁾ This report indicates that the CO desorption from the Ni–Co alloy surface occurs more favorably than that from the Ni metal surface, which results in the carbon deposition decrease accompanying the smaller Rp change in the Ni_0.75Co_0.25-YSZ cell than the Ni-YSZ cell during the prolonged operation time.

Fig. 8

SEM micrographs and EDX spectra of (a) Ni-YSZ and (b) Ni_0.75Co_0.25-YSZ anode surfaces after durability test at 750°C in CH₄ for 240 h.

4. Conclusions

We fabricated two anode-supported SOFCs, consisting of the Ni_1−xCo_x/YSZ (x = 0, 0.25) anode substrate, YSZ film electrolyte, and LSM/YSZ cathode. The single cell using the Ni_0.75Co_0.25/YSZ anode (the Ni_0.75Co_0.25-YSZ cell) showed a stable prolonged cell-performance at 750°C for direct CH₄ feeding compared to the one using the Ni/YSZ anode (the Ni-YSZ cell), although the initial cell performance of the Ni_0.75Co_0.25-YSZ cell was lower than that of the Ni-YSZ cell. The initial cell performance difference is considered to be responsible for the active surface area difference. The prolonged cell-performance stability for CH₄, however, is improved by incorporating Co into Ni in the Ni/YSZ anode due to an enhancement of catalytic activity for the electrochemical oxidation of CH₄ accompanying a retardation effect of carbon deposition.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Number 18H01746. Nicharee Wongsawatgul would like to acknowledge the financial support by a scholarship in the form of a grant from the Pacific Rim Green Innovation Hub Project of Nagaoka University of Technology (NUT) supported by MEXT. We acknowledge M. Takahashi for assistance with the EDX analysis measurement. The advanced methane-utilization research project of NUT supported by MEXT provided some of the analytical instruments.

REFERENCES

• 1)  S.C. Singhal and K. Kendall: High Temperature Solid Oxide Fuel Cells, (Elsevier Advanced Technology, Oxford, 2003) pp. 1–22.
• 2)   W.  Wang,  C.  Su,  Y.  Wu,  R.  Ran and  Z.  Shao: Chem. Rev. 113 (2013) 8104–8151.
• 3)   M.  Rokni: Energy 61 (2013) 87–97.
• 4)   Z.L.  Zhan and  S.A.  Barnett: Science 308 (2005) 844–847.
• 5)   A.  Atkinson,  S.  Barnett,  R.J.  Gorte,  J.T.S.  Irvine,  A.J.  McEvoy,  M.  Mogensen,  S.C.  Singhal and  J.  Vohs: Nat. Mater. 3 (2004) 17–27.
• 6)   S.  Park,  J.M.  Vohs and  R.J.  Gorte: Nature 404 (2000) 265–267.
• 7)   J.  Liu and  S.A.  Barnett: Solid State Ionics 158 (2003) 11–16.
• 8)   E.N.  Armstrong,  J.W.  Pwark and  N.Q.  Minh: ECS Trans. 45 (2012) 499–507.
• 9)   R.  Muccillo,  E.N.S.  Muccillo,  F.C.  Fonseca and  D.Z.  de Florio: J. Electrochem. Soc. 155 (2008) B232–B235.
• 10)   P.  Kaparaju,  I.  Buendia,  L.  Ellegaard and  I.  Angelidakia: Bioresour. Technol. 99 (2008) 4919–4928.
• 11)   C.H.  Ting and  D.J.  Lee: Int. J. Hydrogen Energy 32 (2007) 677–682.
• 12)   M.S.  Fountoulakis and  T.  Manios: Bioresour. Technol. 100 (2009) 3043–3047.
• 13)   N.  Sanphoti,  S.  Towprayoon,  P.  Chaiprasert and  A.  Nopharatana: J. Environ. Manage. 81 (2006) 27–35.
• 14)   W.  Yin and  S.S.C.  Chuang: Catal. Commun. 102 (2017) 62–66.
• 15)   N.  Shah,  D.  Panjala and  G.P.  Huffman: Energy Fuels 15 (2001) 1528–1534.
• 16)   R.C.  Cantelo: J. Phys. Chem. 28 (1924) 1036–1048.
• 17)   T.M.  Gür: Prog. Energy Combust. Sci. 54 (2016) 1–64.
• 18)   M.  Mogensen and  K.  Kammer: Annu. Rev. Mater. Res. 33 (2003) 321–331.
• 19)   T.  Kim,  G.  Liu,  M.  Boaro,  S.-I.  Lee,  J.M.  Vohs,  R.J.  Gorte,  O.H.  Al-Madhi and  B.O.  Dabbousi: J. Power Sources 155 (2006) 231–238.
• 20)   T.  Iida,  M.  Kawano,  T.  Matsui,  R.  Kikuchi and  K.  Eguchi: J. Electrochem. Soc. 154 (2007) B234–B241.
• 21)   H.  Sumi,  P.  Puengjinda,  H.  Muroyama,  T.  Matsui and  K.  Eguchi: J. Power Sources 196 (2011) 6048–6054.
• 22)   A.  Hagen,  R.  Barfod,  P.V.  Hendriksen,  Y.L.  Liu and  S.  Ramousse: J. Electrochem. Soc. 153 (2006) A1165–A1171.
• 23)   A.  Rismanchian,  J.  Mirzababaei and  S.S.C.  Chuang: Catal. Today 245 (2015) 79–85.
• 24)   M.D.  Gross,  J.M.  Vohs and  R.J.  Gorte: J. Mater. Chem. 17 (2007) 3071–3077.
• 25)   K.  Sasaki and  Y.  Teraoka: J. Electrochem. Soc. 150 (2003) A878–A884.
• 26)   R.A.  Budiman,  T.  Ishiyama,  S.S.  Liu,  K.D.  Bagarinao,  H.  Kishimoto,  K.  Yamaji and  T.  Horita: ECS Trans. 91 (2019) 1973–1978.
• 27)   D.  Mogensen,  J.D.  Grunwaldt,  P.V.  Hendriksen,  K.  Dam-Johansen and  J.U.  Nielsen: J. Power Sources 196 (2011) 25–38.
• 28)   I.  Gavrielatos,  V.  Drakopoulos and  S.G.  Neophytides: J. Catal. 259 (2008) 75–84.
• 29)   D.K.  Niakolas,  J.P.  Ouweltjes,  G.  Rietveld,  V.  Drakopoulos and  S.G.  Neophytides: Int. J. Hydrogen Energy 35 (2010) 7898–7904.
• 30)   V.  Sariboğa and  F.  Öksüzömer: Appl. Energy 93 (2012) 707–721.
• 31)   E.W.  Park,  H.  Moon,  M.S.  Park and  S.H.  Hyun: Int. J. Hydrogen Energy 34 (2009) 5537–5545.
• 32)   T.  Takeguchi,  R.  Kikuchi,  T.  Yano,  K.  Eguchi and  K.  Murata: Catal. Today 84 (2003) 217–222.
• 33)   T.  Hibino,  A.  Hashimoto,  M.  Yano,  M.  Suzuki and  M.  Sano: Electrochim. Acta 48 (2003) 2531–2537.
• 34)   J.J.  Strohm,  J.  Zheng and  C.  Song: J. Catal. 238 (2006) 309–320.
• 35)   D.B.  Ingram and  S.  Linic: J. Electrochem. Soc. 156 (2009) B1457–B1465.
• 36)   S.  McIntosh,  J.M.  Vohs and  R.J.  Gorte: J. Electrochem. Soc. 150 (2003) A470–A476.
• 37)   S.  Jung,  C.  Lu,  H.P.  He,  K.  Ahn,  R.J.  Gorte and  J.M.  Vohs: J. Power Sources 154 (2006) 42–50.
• 38)   A.  Ringuedé,  J.A.  Labrincha and  J.R.  Frade: Solid State Ionics 141–142 (2001) 549–557.
• 39)   T.  Ishihara,  J.W.  Yan,  M.  Shinagawa and  H.  Matsumoto: Electrochim. Acta 52 (2006) 1645–1650.
• 40)   W.  Shi-zhong and  G.  Jie: Electrochem. Solid-State Lett. 9 (2006) A395–A398.
• 41)   H.  Kan and  H.  Lee: Catal. Commun. 12 (2010) 36–39.
• 42)   X.C.  Lu and  J.H.  Zhu: J. Power Sources 165 (2007) 678–684.
• 43)   H.  Zhu,  W.  Wang,  R.  Ran,  C.  Su,  H.  Shi and  Z.  Shao: Int. J. Hydrogen Energy 37 (2012) 9801–9808.
• 44)   N.Q.  Minh: J. Am. Ceram. Soc. 76 (1993) 563–588.
• 45)   A.  Benyoucef,  D.  Klein,  C.  Coddet and  B.  Benyoucef: Surf. Coat. Technol. 202 (2008) 2202–2207.
• 46)   S.-I.  Lee,  K.  Ahn,  J.M.  Vohs and  R.J.  Gorte: Electrochem. Solid-State Lett. 8 (2005) A48–A51.
• 47)   T.  Guo,  X.  Dong,  M.M.  Shirolkar,  X.  Song,  M.  Wang,  L.  Zhang,  M.  Li and  H.  Wang: ACS Appl. Mater. Interfaces 6 (2014) 16131–16139.
• 48)   A.  Ringuedé,  D.  Bronine and  J.R.  Frade: Electrochim. Acta 48 (2002) 437–442.
• 49)   C.M.  Grgicak,  M.M.  Pakulska,  J.S.  O’Brien and  J.B.  Giorgi: J. Power Sources 183 (2008) 26–33.
• 50)   C.-K.  Cho,  B.-H.  Choi and  K.-T.  Lee: J. Alloy. Compd. 541 (2012) 433–439.
• 51)   N.  Wongsawatgul,  S.  Chaianansucharit,  K.  Yamamoto,  M.  Nanko and  K.  Sato: Ceramics 3 (2020) 114–126.
• 52)   R.D.  Shannon: Acta Crystallogr. Sec. A 32 (1976) 751–767.
• 53)   S.  Kuboon and  Y.H.  Hu: Ind. Eng. Chem. Res. 50 (2011) 2015–2020.
• 54)   S.  Bošković and  M.  Stevanović: J. Mater. Sci. 10 (1975) 25–31.
• 55)   T.L.  Achmad,  W.  Fu,  H.  Chen,  C.  Zhang and  Z.-G.  Yang: J. Alloy. Compd. 694 (2017) 1265–1279.
• 56)   G.S.  Lewis,  A.  Atkinson and  B.C.H.  Steele: J. Mater. Sci. Lett. 20 (2001) 1155–1157.
• 57)   M.  Hartmanová,  F.  Hanic,  D.  Tunega and  K.  Putyera: Chem. Zvesti 52 (1998) 12–15.
• 58)   G.C.T.  Silva and  E.N.S.  Muccillo: Solid State Ionics 180 (2009) 835–838.
• 59)   H.  Liu,  B.  Wang,  M.  Fan,  N.  Henson,  Y.  Zhang,  B.F.  Towler and  H.G.  Harris: Fuel 113 (2013) 712–718.
• 60)   V.  Alzate-Restrepo and  J.M.  Hill: Appl. Catal. A 342 (2008) 49–55.
• 61)   C.  Liu,  T.R.  Cundari and  A.K.  Wilson: J. Phys. Chem. C 116 (2012) 5681–5688.