- Email: [email protected]
electrolyte interface in solid oxide fuel cell
Journal of Power Sources 412 (2019) 695–700
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Investigation of oxide ion flux at cathode/electrolyte interface in solid oxide fuel cell
T
Tsuyoshi Nagasawaa,∗, Katsunori Hanamurab,∗∗ a b
School of Engineering, Department of Systems and Control Engineering, Tokyo Institute of Technology, 2-12-1, O-okayama, Meguro-ku, Tokyo, 152-8550, Japan School of Engineering, Department of Mechanical Engineering, Tokyo Institute of Technology, Japan
H I GH L IG H T S
ion flux at SOFC cathode/electrolyte interface is quantitatively investigated. • Oxide isotope labeling and reaction quenching by He gas impinging jet is conducted. • Oxygen O distribution at 973 K is obtained by SIMS and analyzed by diffusion equation. • 22–31% of overall cathode reaction occurs at the cathode/electrolyte interface. • 18
A R T I C LE I N FO
A B S T R A C T
Keywords: Solid oxide fuel cell Composite cathode Quenching Oxygen isotope SIMS Oxide ion flux
Oxide ion flux at cathode/electrolyte interface of solid oxide fuel cell (SOFC) is investigated through quenching reaction and oxygen isotope labeling. A YSZ (yttria-stabilized zirconia) electrolyte-supported cell with LSM (strontium-doped lanthanum manganite)/YSZ porous cathode is operated by supplying 18O2 at 973 K and abruptly quenched to room temperature by a direct helium gas-impinging jet to the cell. The 18O concentration distribution in the cross section of the cathode/electrolyte interface is obtained by secondary ion mass spectrometry (SIMS) with a spatial resolution of 50 nm. From the analysis of oxygen isotope diffusion profiles in YSZ electrolyte, oxide ion flux incorporated from a cathode/electrolyte interface to an electrolyte is first estimated. The obtained flux 1.01–1.43 × 10−3 mol m−2 s−1 at a current density of 0.09 A cm−2 indicates that 22–31% of the overall electrochemical reaction occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode under the present experimental condition.
1. Introduction The electrochemical performance of porous composite electrodes in solid oxide fuel cells (SOFCs), such as Ni/YSZ (yttria-stabilized zirconia) anode or LSM (lanthanum strontium manganite)/YSZ cathode, is dominated by electrochemical reaction at the active sites and transport of gases, oxide ion, and electron in the porous structure [1–3]. For precise understanding and prediction of an electrode polarization behaviour, numerous efforts have so far been made to explore 1-D to 3-D spatial distributions of the reaction rates and transport paths and to link those with electrode performance through numerical simulations [4–8]. In contrast, quantitative information about those distributions in porous electrodes has not been experimentally obtained so far, hindering the clear validation and understanding of the numerical studies. Atomic labeling using oxygen isotope has been widely introduced to
∗
investigate oxide ion transport and reactions in various oxide materials and systems [9–21]. One of the major directions in atomic labeling research is to obtain oxygen transport and reaction parameters such as tracer diffusion coefficient and surface exchange coefficient [9–15]. The obtained parameters are mainly discussed with oxide ionic conductivity and catalytic activity of oxide materials. Other interests are to visualize active reaction sites and electrode reaction pathways by using thin film electrodes on solid oxide electrolyte [16–19]. The detailed electrode reaction mechanism can be discussed through this method. In addition, Horita et al. [20,21] applied oxygen isotope labeling to a practical SOFC, offering the visualization of overall oxide ion flows from cathode to anode sides. However, neither active reaction sites nor quantitative oxide ion flux in a practical SOFC porous electrodes was investigated in the previous oxygen isotope labeling studies. Based on these circumstances, the authors have recently developed
Corresponding author. Corresponding author. E-mail addresses: [email protected] (T. Nagasawa), [email protected] (K. Hanamura).
∗∗
https://doi.org/10.1016/j.jpowsour.2018.12.013 Received 20 September 2018; Received in revised form 3 December 2018; Accepted 5 December 2018 0378-7753/ © 2018 Elsevier B.V. All rights reserved.
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Fig. 1. (a) Schematic of power generation equipment of SOFC single cell combined with quench system and the geometry of the electrolyte-supported cell used in the experiment. SIMS observation areas are indicated by blue rectangles (OCV and Biased). (b) Terminal voltage versus elapsed time with a constant current density of 0.09 A cm−2 at 973 K. Humidified mixture of H2 − Ar and pure oxygen were supplied as a fuel and oxidant, respectively. (c) 18O concentration (c18O ) mappings of the interface between the LSM/YSZ cathode and YSZ electrolyte at OCV and biased conditions. For biased condition, two images denoted by Biased #1 and #2 from different positions in the same cell were taken. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
1.6 vol%. Initially, normal O2 was supplied to the cathode side and the cell was operated for 10 min under a constant current density of 0.09 A cm−2. Then, the normal O2 was switched to 18O2 (> 97.5%), and the cell was operated for 70 s. Finally, the cell was abruptly quenched to a room temperature by the helium gas-impinging jet. Once 18 O2 is switched to He gas, a space between the nozzle tip and electrode surface is purged by He gas within 10–20 ms, which is much shorter than quench duration of 1–2 s. After quenching to room temperature, tracer diffusion coefficient and surface exchange coefficient of YSZ are reduced by 10 orders of magnitude or more [15], making it possible to stop the movement of oxygen atom in YSZ lattice. As shown in Fig. 1(b), the cell showed an almost stable terminal voltage of about 0.42 V at 0.09 A cm−2 until quenching reaction. The dwell time in 18O2 is one of the critical factors to get useful 18O distribution image because too long dwell time results in uniform 18O distribution in YSZ due to long-range 18 O diffusion. In our prior work, 3 min exposure to 18O2 at 1073 K gave almost uniform 18O distribution in YSZ and ScSZ phases [22]. On the other hand, Fig. 1(b) shows that at least 30 s is needed to stabilize terminal voltage again after the normal O2 is switched to 18O2. Therefore, 70 s, longer enough than 30 s, was selected as a dwell time in this study. The 18O distribution in a cross section of the quenched cathode was obtained by secondary ion mass spectroscopy (CAMECA, NanoSIMS 50L) with a spatial resolution of 50 nm. Before the SIMS analysis, pores of the sample were infiltrated with an epoxy resin; then, the cross section of the cathode was polished using a cross section polisher to obtain a smooth surface. As shown in Fig. 1(a), SIMS images at the interface between the cathode and electrolyte were taken from both reference (for OCV) and working (for Biased) electrodes.
the imaging technique of active sites in porous electrodes through quenching SOFC reaction using a helium gas impinging jet and oxygen isotope labeling [22]. In this prior work, active sites of LSM/ScSZ (scandia-stabilized zirconia) porous cathode were qualitatively visualized at 800 °C in microstructure scale based on the 18O diffusion into LSM induced by overpotential. However, quantitative discussion of oxide ion flux incorporated to electrolyte phase (YSZ, ScSZ) is still lacking due to fast 18O diffusion in the electrolyte phase and resulting uniform 18O distribution [22]. In this communication, we first investigated quantitative oxide ion flux incorporated from the cathode/ electrolyte interface based on the analysis of the 18O diffusion profile in YSZ electrolyte at 700 °C. 2. Experimental In the experiment, a YSZ-electrolyte supported cell with a Ni/YSZ anode and LSM/YSZ (LSM:YSZ = 50:50 wt%) cathode was used. The fabrication process is given in our previous work [22]. As shown in Fig. 1(a), both the anode and cathode were separated into two parts to prepare reference electrodes. Pt wires for electric lead were attached to working electrodes, and thermocouple to anode reference using a silver paste. The cell was set to the power generation equipment with a watercooled nozzle for helium gas impinging jet as shown in Fig. 1(a). In this study, the cell was operated in a mixture of H2 (15 mL min−1) and Ar (15 mL min−1), which was humidified by a bubbler, for the anode side and in pure O2 (30 mL min−1) for the cathode side at 973 K. Although most researches and commercial SOFC stacks use air as an oxidant, pure oxygen was used in this study because oxide ion flux behaviour in pure O2 is a standard information to understand those in any reduced oxygen partial pressure including air composition. The observed OCV of 1.15 V and Nernst equation indicates that the supplied fuel composition is H2: Ar: H2O = 49.2: 49.2: 696
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Fig. 2. (a,b) High-resolution phase and c18O mappings of the cathode/electrolyte interface at (a) OCV and (b) Biased. c18O profiles in Line A and B are also shown in (a) and (b), respectively. (c) c18O line profiles (4 μm form cathode/electrolyte interface to inside YSZ electrolyte) denoted by Line 1–5 of (a) OCV and (b) Biased. Linear fitted curves of each profile and coefficient of variation (CV) of the gradient (dC18O/ dx ) are also shown.
3. Results and discussion
Fig. 2(a) and (b) show high-resolution phase and c18O mappings. The phase mappings were created based on the signal of Mn16O− (for LSM) and Zr16O− (for YSZ). As is the case in our previous study [22], 18O diffusion to LSM bulk was only observed at the region near the cathode/ electrolyte interface under biased condition, indicating that reaction sites near the interface are more electrochemically active than those at other regions. In addition, c18O in the electrolyte is uniform near the cathode/electrolyte interface at OCV while inhomogeneous at biased condition, which is related to the non-uniform oxide ion flux incorporated from the interface. 18O flux from the interface is proportional to c18O gradient dc18O/ dx . Here, c18O line profiles of YSZ electrolyte with 4 μm-length from cathode/electrolyte interface, denoted by Line 1–5 in Fig. 2(a) and (b), were linearly fitted as shown in Fig. 2(c). At OCV, the mean value, standard deviation (σ), and coefficient of variation (CV = σ/mean) of the gradient dc18O/ dx of 5 fitting curves are −2.32 × 10−3 μm−1, 2.79 × 10−4 μm−1, and 12.0%, respectively. In the case of biased condition, the mean, σ, and CV are −9.86 × 10−3 μm−1, 1.795 × 10−3 μm−1, and 18.2%, respectively. The increased CV at biased (with oxide ion flux) compared to OCV (without oxide ion flux) indicates that the oxide ion flux incorporated
3.1. Secondary ion mappings Fig. 1(c) shows 18O concentration c18O (= I(18O−)/[I(16O−) + I ( O−)]) mappings of the interface between the LSM/YSZ cathode and YSZ electrolyte at OCV and biased conditions. For biased condition, two images denoted by Biased #1 and #2 from different positions in the same cell were taken. Here, I(16O−) and I(18O−) indicate the SIMS signal intensity of each ion. Prior to the calculation of c18O image, binarization processing is performed on the 16O− image to remove all signal from pore phases. As shown in Fig. 1(c), c18O ≈ 0.04 of YSZ phases in cathode part is higher than that in the electrolyte (maximum c18O ≈ 0.025 at the surface) at OCV. This is because the surface to volume ratio of YSZ phase in cathode part is much larger than that in electrolyte part. At biased condition, c18O in YSZ phase is around 0.12–0.16 at cathode or electrolyte surface, which is higher than that at OCV because of additionally-incorporated 18O by the electrochemical reaction. Both in OCV and biased conditions, c18O in the electrolyte gradually decrease from cathode/electrolyte interface to anode side. 18
697
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Table 1 Oxygen tracer diffusion coefficient D* and surface exchange coefficient k* of YSZ electrolyte at 973 K obtained in this study and other references [14,15]. All literature values were measured at non-biased condition. Measurement
Sample
D* [m2 s−1] (at 973 K)
k* [m s−1] (at 973 K)
OCV Biased #1 Biased #2 Naito et al. [14]
8mol%Y2O3-ZrO2 Polycrystalline (with LSM/YSZ cathode) 9.1mol%Y2O3-ZrO2 Polycrystalline 2.8–16mol%Y2O3ZrO2 Single crystal, Polycrystalline
9.70 × 10−13 1.15 × 10−12 1.45 × 10−12 4.0 × 10−13
2.45 × 10−9 1.50 × 10−8 2.08 × 10−8 1.4 × 10−9
2.1 × 10−13 – 4.0 × 10−12
2.0 × 10−11 – 2.8 × 10−8
Knöner et al. [15]
side. Therefore, incorporated 18O itself does not move along the electrochemical potential gradient of oxide ion, ∇μ˜ O2 − , but moves along the 18 O concentration gradient ∇c18O . On the other hand, the one order enhancement of apparent k* indicates that 18O flux incorporated from the cathode/electrolyte interface to the electrolyte is greatly increased by applying the current. The detail of the 18O flux is quantitatively discussed in the next section. 3.3. Oxide ion flux incorporated from the cathode/electrolyte interface con(C/E) ex(C/E) , J18O , and As shown in Fig. 3(b), three kinds of 18O fluxes, J18O ion(C/E) 18 J18O contribute to a total O flux at the cathode/electrolyte interex(C/E) face. J18O is a18O flux by a16O/18O exchange at a free electrolyte ion(C/E) is an isotope oxide ion (18O2−) flux by an electrosurface and J18O
chemical reaction at a LSM/YSZ/gas triple phase boundary (TPB) on the electrolyte surface. In addition, c18O of YSZ phase in the cathode is generally higher than that of the electrolyte due to the higher surface to volume ratio of YSZ phase in the cathode, as mentioned previously. con(C/E) J18O is a18O flux by the c18O difference of YSZ phase between con(C/E) ex(C/E) cathode and electrolyte part. Here J18O and J18O flow at both OCV ion(C/E) exists only at biased. Therefore, the and biased conditions while J18O total 18O flux incorporated from the cathode/electrolyte interface to the Biased(C/E) OCV(C/E) electrolyte at OCV and biased conditions, J18O and J18O are described as follows.
Fig. 3. (a) Three (OCV, Biased #1, Biased #2) c18O line profiles inside YSZ electrolyte denoted by the arrows in Fig. 1(c). Analytical solution of diffusion equation is fitted to each data. The obtained and literature values of D* and k* are summarized in Table 1. (b) Three kinds of 18O fluxes incorporated from the cathode/electrolyte interface to the electrolyte. Obtained 18O fluxes are summarized in Table 2.
from the cathode/electrolyte interface is non-uniform in the observed region of Fig. 2(b). 3.2.
18
O diffusion profile analysis in YSZ electrolyte
Three (OCV, Biased #1, Biased #2) c18O line profiles inside YSZ electrolyte denoted by the arrows in Fig. 1(c) are shown in Fig. 3(a). In this study, polycrystalline YSZ (8mol%Y2O3-ZrO2) was used as an electrolyte. In all cases, typical monotonically decreasing diffusion profiles are observed, which can be analyzed by 1-dimensional diffusion in a semi-infinite media. As shown in Fig. 3(a), the analytical solution of diffusion equation [12] is well fitted to each profile, which gives oxygen tracer diffusion coefficient D* and surface exchange coefficient k*. The D* and k* obtained here and other references [14,15] are summarized in Table 1. At OCV, obtained D* = 9.70 × 10−13 m2 s−1 and k* = 2.45 × 10−9 m s−1 are relatively close to those of polycrystalline YSZ (9.1mol%Y2O3) by Naito et al. [14], and within the range of summarized data sets by Knöner et al. (2.8–16mol%Y2O3-ZrO2, single crystal and polycrystalline) [15]. When the current is applied, D* increased by a factor of only 1.2 to 1.5 compared to that of OCV while k* was enhanced by one order of magnitude, as shown in Table 1. The weak enhancement of D* means that the enhancement of 18O diffusion in YSZ bulk by the current is small. In a YSZ, oxide ion moves through vacancy diffusion mechanism [23]. In a fuel cell operation, when one oxide ion is incorporated to an electrolyte form a cathode, one oxide ion is emitted to an anode accompanied by a generation of oxygen vacancy in the electrolyte. Subsequently the generated oxygen vacancy moves from anode to cathode
OCV(C/E) ex(C/E) con(C/E) J18O = J18O + J18O
(1)
Biased(C/E) ex(C/E) con(C/E) ion(C/E) J18O = J18O + J18O + J18O
(2)
Biased(C/E) OCV(C/E) and J18O [mol m−2 s−1] corAs shown in Fig. 3(a), J18O responds to a c18O gradient at cathode/electrolyte interface ( x = 0 ), which can be described as follows. OCV(C/E) ⎡ ∂c18O ∗ J18O = −COYSZ DOCV ⎢ ∂x ⎣
OCV
⎤ ⎥ x=0 ⎦
Biased(C/E) ⎡ ∂c18O ∗ J18O = −COYSZ DBiased ⎢ ∂x ⎣
(3) Biased
⎤ ⎥ x=0 ⎦
(4)
∗ ∗ Here DOCV and DBiased [m2 s−1] represents the oxygen tracer diffusion coefficient at OCV and biased condition, respectively. COYSZ (= 9.41 × 104 mol m−3) is an absolute concentration of total oxygen atom in a YSZ calculated from the lattice structure of 8.3mol%Y2O3-ZrO2 Biased(C/E) (lattice constant: 5.135 Å [24]). The calculated J18O is one order of OCV(C/E) as summarized in Table 2. magnitude larger than J18O OCV(C/E) At OCV, J18O is also described by the surface exchange coeffi∗ cient kOCV [m s−1] as follows:
gas OCV(C/E) ex(C/E) con(C/E) ∗ J18O = J18O + J18O = COYSZ kOCV [c18O − c18O
where 698
gas c18O
is a relative
18
OCV x=0 ]
(5)
O concentration in gas phase (= 0.975). In the
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
coefficient, and 18O diffusion length into account in order to get useful O distribution image, as mentioned in section 2. Experimental.
Table 2 18 O flux incorporated from the cathode/electrolyte interface to the electrolyte estimated by diffusion profiles. Total18O2− flux at given current density is also shown.
18
4. Conclusions
Value [mol m−2 s−1]
Flux
Oxide ion flux at the interface between LSM/YSZ cathode and YSZ electrolyte of electrolyte-supported SOFC was investigated at 973 K through quenching reaction by helium gas-impinging jet and oxygen isotope labeling. From the analysis of 18O diffusion profiles in the quenched YSZ electrolyte measured by SIMS, it was disclosed that by applying a current, oxygen tracer diffusion coefficient D* increased by a factor of only 1.2–1.5 while surface exchange coefficient k* was enhanced by one order of magnitude. The greatly enhanced k* comes from oxide ion flux incorporated from the cathode/electrolyte interface to the electrolyte. Based on the 18O profiles, quantitative oxide ion flux incorporated from the cathode/electrolyte interface was first estimated. The obtained flux 1.01–1.43 × 10−3 mol m−2 s−1 at a current density of 0.09 A cm−2 indicates that 22–31% of the overall electrochemical reaction occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode under the present experimental condition.
2.18 × 10−4
OCV(C/E) J18O
1.21 × 10−3 1.63 × 10−3 1.01 × 10−3 1.43 × 10−3
Biased(C/E) J18O ion(C/E) J18O
(Biased (Biased (Biased (Biased
#1) #2) #1) #2)
gas = 0.975 ) 4.58 × 10−3 (i = 0.09 A/cm2, c18O
ion(all) J18O
above expression, it seems to be difficult to separate the contribution con(C/E) ex(C/E) from J18O and J18O precisely. Therefore, in order to calculate ex(C/E) con(C/E) ion(C/E) + J18O J18O in Eq. (2), J18O at biased condition is roughly estimated below. Both in OCV and biased condition, surface exchange coefficient on ex(C/E) free YSZ surface, which determines J18O , should be the same. In con(C/E) OCV(C/E) addition, J18O at OCV is not more than J18O (= 2.18 × 10−4 mol m−2 s−1) from Eq. (5). As shown in Fig. 3(b), the con(C/E) is the difference of c18O ( Δc18O ) between driving force of J18O cathode and electrolyte part. From c18O line profiles of Line A and B in Fig. 2(a) and (b), Δc18O at OCV (≈ 0.015) and biased condition (≈ 0.03) con(C/E) is the same order of magnitude. Therefore, J18O at biased condition is also estimated to be less than or equal to 10−4 order of magnitude. Biased(C/E) is 10−3 order of magnitude, rough estimation of Because J18O ex(C/E) con(C/E) J18O + J18O at biased condition (10−4 order of magnitude) does ion(C/E) not strongly affect J18O . con(C/E) ex(C/E) and J18O at biased Based on the above discussion, sum of J18O ∗ condition was approximately estimated using kOCV as follows. Biased
ex(C/E) con(C/E) [J18O + J18O ]
gas ∗ = COYSZ kOCV [c18O − c18O
Biased x=0 ]
Acknowledgements This work was partly supported by the Japan Society for the Promotion of Science KAKENHI Grant Numbers 16K14171 and 16J03479, and The Hattori-Hokokai Foundation. SIMS analysis was conducted at the Advanced Characterization Nanotechnology Platform of the University of Tokyo, supported by the Nanotechnology Platform of the Ministry of Education, Culture, Sports, Science and Technology, Japan. The authors would like to thank Dr M. Takeuchi for support with the SIMS measurement.
(6)
References
ion(C/E) J18O ,
which are Substitution of Eqs. (4) and (6) into (2) gives summarized in Table 2. The total 18O2− flux incorporated from whole gas ion(all) = c18O i/2F at given current density i [A m−2] is also cathode J18O shown in this table. Under the assumption that whole cathode area contributes to the electrochemical reaction equivalently, the ratio of the ion(C/E) ion(all) (= 1.01–1.43 × 10−3 mol m−2 s−1) and J18O (= obtained J18O 4.58 × 10−3 mol m−2 s−1) indicates that 22–31% of the overall electrochemical reaction (oxide ion incorporations to YSZ phase) occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode in the present experimental condition. In a state-of-the-art 3-dimensional numerical simulation of LSM/ YSZ porous cathode (LSM:YSZ = 50:50 wt%) based on the FIB-SEMobserved actual microstructure by Miyoshi et al. [25], charge transfer current distribution in the cathode thickness direction was calculated at 850 °C. In their result, 22.5% of the overall electrochemical reaction occurs at the region within 1.25 μm from the cathode/electrolyte interface, which is relatively close to the scale of one LSM particle on the electrolyte as shown in Fig. 2(b). As shown here, present work offers the first experimental quantification of oxide ion flux inside a SOFC porous cathode, making it possible to compare experimental and numerical reaction distributions. As a future direction, the correlation between oxide ion flux from a cathode/electrolyte interface to an electrolyte and several parameters such as current density, cathode overpotential and impedance, oxygen concentration, and operating temperature can be experimentally investigated by using the current method. Such information can provide both deeper understanding of electrode reaction mechanisms and guidance for electrode design optimization. In addition, other state-of-the-art cathode materials, e.g., (La,Sr)CoO3 (LSC), (La,Sr) (Co,Fe)O3 (LSCF), or (Ba,Sr) (Co,Fe)O3 (BSCF), can be studied, though operating temperature and dwell time in 18O2 should be carefully selected taking oxide ionic conductivity, tracer diffusion
[1] J. Hanna, W.Y. Lee, Y. Shi, A.F. Ghoniem, Fundamentals of electro- and thermochemistry in the anode of solid-oxide fuel cells with hydrocarbon and syngas fuels, Prog. Energy Combust. Sci. 40 (2014) 74–111. [2] T. Nagasawa, K. Hanamura, Prediction of overpotential and effective thickness of Ni/YSZ anode for solid oxide fuel cell by improved species territory adsorption model, J. Power Sources 353 (2017) 115–122. [3] D. Chen, W. Bi, W. Kong, Z. Lin, Combined micro-scale and macro-scale modelling of the composite electrode of a solid oxide fuel cell, J. Power Sources 195 (2010) 6598–6610. [4] P. Costamagna, P. Costa, V. Antonucci, Micro-modelling of solid oxide fuel cell electrodes, Electrochim. Acta 43 (1998) 375–394. [5] J.H. Nam, D.H. Jeon, A comprehensive micro-scale model for transport and reaction in intermediate temperature solid oxide fuel cells, Electrochim. Acta 51 (2006) 3446–3460. [6] M. Kishimoto, H. Iwai, M. Saito, H. Yoshida, Three-dimensional simulation of SOFC anode polarization characteristics based on sub-grid scale modeling of microstructure, J. Electrochem. Soc. 159 (3) (2012) B315–B323. [7] T. Carraro, J. Joos, B. Rüger, A. Weber, E. I-Tiffée, 3D finite element model for reconstructed mixed-conducting cathode: I. Performance quantification, Electrochim. Acta 77 (2012) 315–325. [8] A. He, Y. Kim, N. Shikazono, Three-dimensional numerical simulation of solid oxide fuel cell cathode based on lattice Boltzmann method with sub-grid scale models, Int. J. Hydrogen Energy 42 (2017) 21886–21900. [9] I. Yasuda, K. Ogasawara, M. Hishinuma, T. Kawada, M. Dokiya, Oxygen tracer diffusion coefficient of (La, Sr)MnO3 ± δ, Solid State Ionics 86–88 (1996) 1197–1201. [10] J.A. Kilner, S.J. Skinner, H.H. Brongersma, The isotope exchange depth profiling (IEDP) technique using SIMS and LEIS, J. Solid State Electrochem. 15 (5) (2011) 861–876. [11] A. Esquirol, J. Kilner, N. Brandon, Oxygen transport in La0.6Sr0.4Co0.2Fe0.8O3-δ/ Ce0.8Ge0.2O2-x composite cathode for IT-SOFCs, Solid State Ionics 175 (2004) 63–67. [12] R.A. Souza, J.A. Kilner, J.F. Walker, A SIMS study of oxygen tracer diffusion and surface exchange in La0.8Sr0.2MnO3+δ, Mater. Lett. 43 (2000) 43–52. [13] E.S. Raj, A. Atkinson, J.A. Kilner, Oxygen diffusion studies on (Y2O3)2(Sc2O3)9(ZrO2)89, Solid State Ionics 180 (2009) 952–955. [14] H. Naito, N. Sakai, T. Otake, H. Yugami, H. Yokokawa, Oxygen transport properties in ZrO2-CeO2-Y2O3 by SIMS analyses, Solid State Ionics 135 (2000) 669–673. [15] G. Knöner, K. Reimann, R. Röwer, U. Södervall, H.-E. Schaefer, Enhanced oxygen diffusivity in interfaces of nanocrystalline ZrO2•Y2O3, Proc. Natl. Acad. Sci. Unit.
699
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Electrochem. Solid State Lett. 13 (12) (2010) B135–B138. [21] T. Horita, M. Nishi, T. Shimonosono, H. Kishimoto, K. Yamaji, M.E. Brito, H. Yokokawa, Visualization of oxide ionic diffusion at SOFC cathode/electrolyte interfaces by isotope labeling techniques, Solid State Ionics 262 (2014) 398–402. [22] T. Nagasawa, K. Hanamura, Microstructure-scaled active sites imaging of a solid oxide fuel cell composite cathode, J. Power Sources 367 (2017) 57–62. [23] H. Tagawa, Solid Oxide Fuel Cell and Global Environment, Agne-Shofu-Sya, 1998 (in Japanese). [24] H. Hayashi, T. Saitou, N. Maruyama, H. Inaba, K. Kawamura, M. Mori, Thermal expansion coefficient of yttria stabilized zirconia for various yttria contents, Solid State Ionics 176 (2005) 613–619. [25] K. Miyoshi, T. Miyamae, H. Iwai, M. Saito, M. Kishimoto, H. Yoshida, Exchange current model for (La0.8Sr0.2)0.95MnO3 (LSM) porous cathode for solid oxide fuel cells, J, Power Sources 315 (2016) 63–69.
States Am. 100 (7) (2003) 3870–3873. [16] T. Kawada, T. Horita, N. Sakai, H. Yokokawa, M. Dokiya, J. Mizusaki, A novel technique for imaging electrochemical reaction sites on a solid oxide electrolyte, Solid State Ionics 131 (2000) 199–210. [17] T. Horita, K. Yamaji, N. Sakai, Y. Xiong, T. Kato, H. Yokokawa, T. Kawada, Imaging of oxygen transport at SOFC cathode/electrolyte interfaces by novel technique, J. Power Sources 106 (2002) 224–230. [18] A.K. Opitz, M. Kubicek, S. Huber, T. Huber, G. Holzlechner, Thin film cathodes in SOFC research: how to identify oxygen reduction pathways? J. Mater. Res. 28 (16) (2013) 2085–2105. [19] J.H. Shim, J.S. Park, T.P. Holme, K. Crabb, W. Lee, Y.B. Kim, X. Tian, T.M. Gür, F.B. Prinz, Enhanced oxygen exchange and incorporation at surface grain boundaries on an oxide ion conductor, Acta Mater. 60 (2012) 1–7. [20] T. Horita, T. Shimonosono, H. Kishimoto, K. Yamaji, M.E. Brito, Y. Hori, H. Yokokawa, Visualization of oxygen ionization and flows in solid oxide fuel cells,
700
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Investigation of oxide ion flux at cathode/electrolyte interface in solid oxide fuel cell
T
Tsuyoshi Nagasawaa,∗, Katsunori Hanamurab,∗∗ a b
School of Engineering, Department of Systems and Control Engineering, Tokyo Institute of Technology, 2-12-1, O-okayama, Meguro-ku, Tokyo, 152-8550, Japan School of Engineering, Department of Mechanical Engineering, Tokyo Institute of Technology, Japan
H I GH L IG H T S
ion flux at SOFC cathode/electrolyte interface is quantitatively investigated. • Oxide isotope labeling and reaction quenching by He gas impinging jet is conducted. • Oxygen O distribution at 973 K is obtained by SIMS and analyzed by diffusion equation. • 22–31% of overall cathode reaction occurs at the cathode/electrolyte interface. • 18
A R T I C LE I N FO
A B S T R A C T
Keywords: Solid oxide fuel cell Composite cathode Quenching Oxygen isotope SIMS Oxide ion flux
Oxide ion flux at cathode/electrolyte interface of solid oxide fuel cell (SOFC) is investigated through quenching reaction and oxygen isotope labeling. A YSZ (yttria-stabilized zirconia) electrolyte-supported cell with LSM (strontium-doped lanthanum manganite)/YSZ porous cathode is operated by supplying 18O2 at 973 K and abruptly quenched to room temperature by a direct helium gas-impinging jet to the cell. The 18O concentration distribution in the cross section of the cathode/electrolyte interface is obtained by secondary ion mass spectrometry (SIMS) with a spatial resolution of 50 nm. From the analysis of oxygen isotope diffusion profiles in YSZ electrolyte, oxide ion flux incorporated from a cathode/electrolyte interface to an electrolyte is first estimated. The obtained flux 1.01–1.43 × 10−3 mol m−2 s−1 at a current density of 0.09 A cm−2 indicates that 22–31% of the overall electrochemical reaction occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode under the present experimental condition.
1. Introduction The electrochemical performance of porous composite electrodes in solid oxide fuel cells (SOFCs), such as Ni/YSZ (yttria-stabilized zirconia) anode or LSM (lanthanum strontium manganite)/YSZ cathode, is dominated by electrochemical reaction at the active sites and transport of gases, oxide ion, and electron in the porous structure [1–3]. For precise understanding and prediction of an electrode polarization behaviour, numerous efforts have so far been made to explore 1-D to 3-D spatial distributions of the reaction rates and transport paths and to link those with electrode performance through numerical simulations [4–8]. In contrast, quantitative information about those distributions in porous electrodes has not been experimentally obtained so far, hindering the clear validation and understanding of the numerical studies. Atomic labeling using oxygen isotope has been widely introduced to
∗
investigate oxide ion transport and reactions in various oxide materials and systems [9–21]. One of the major directions in atomic labeling research is to obtain oxygen transport and reaction parameters such as tracer diffusion coefficient and surface exchange coefficient [9–15]. The obtained parameters are mainly discussed with oxide ionic conductivity and catalytic activity of oxide materials. Other interests are to visualize active reaction sites and electrode reaction pathways by using thin film electrodes on solid oxide electrolyte [16–19]. The detailed electrode reaction mechanism can be discussed through this method. In addition, Horita et al. [20,21] applied oxygen isotope labeling to a practical SOFC, offering the visualization of overall oxide ion flows from cathode to anode sides. However, neither active reaction sites nor quantitative oxide ion flux in a practical SOFC porous electrodes was investigated in the previous oxygen isotope labeling studies. Based on these circumstances, the authors have recently developed
Corresponding author. Corresponding author. E-mail addresses: [email protected] (T. Nagasawa), [email protected] (K. Hanamura).
∗∗
https://doi.org/10.1016/j.jpowsour.2018.12.013 Received 20 September 2018; Received in revised form 3 December 2018; Accepted 5 December 2018 0378-7753/ © 2018 Elsevier B.V. All rights reserved.
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Fig. 1. (a) Schematic of power generation equipment of SOFC single cell combined with quench system and the geometry of the electrolyte-supported cell used in the experiment. SIMS observation areas are indicated by blue rectangles (OCV and Biased). (b) Terminal voltage versus elapsed time with a constant current density of 0.09 A cm−2 at 973 K. Humidified mixture of H2 − Ar and pure oxygen were supplied as a fuel and oxidant, respectively. (c) 18O concentration (c18O ) mappings of the interface between the LSM/YSZ cathode and YSZ electrolyte at OCV and biased conditions. For biased condition, two images denoted by Biased #1 and #2 from different positions in the same cell were taken. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
1.6 vol%. Initially, normal O2 was supplied to the cathode side and the cell was operated for 10 min under a constant current density of 0.09 A cm−2. Then, the normal O2 was switched to 18O2 (> 97.5%), and the cell was operated for 70 s. Finally, the cell was abruptly quenched to a room temperature by the helium gas-impinging jet. Once 18 O2 is switched to He gas, a space between the nozzle tip and electrode surface is purged by He gas within 10–20 ms, which is much shorter than quench duration of 1–2 s. After quenching to room temperature, tracer diffusion coefficient and surface exchange coefficient of YSZ are reduced by 10 orders of magnitude or more [15], making it possible to stop the movement of oxygen atom in YSZ lattice. As shown in Fig. 1(b), the cell showed an almost stable terminal voltage of about 0.42 V at 0.09 A cm−2 until quenching reaction. The dwell time in 18O2 is one of the critical factors to get useful 18O distribution image because too long dwell time results in uniform 18O distribution in YSZ due to long-range 18 O diffusion. In our prior work, 3 min exposure to 18O2 at 1073 K gave almost uniform 18O distribution in YSZ and ScSZ phases [22]. On the other hand, Fig. 1(b) shows that at least 30 s is needed to stabilize terminal voltage again after the normal O2 is switched to 18O2. Therefore, 70 s, longer enough than 30 s, was selected as a dwell time in this study. The 18O distribution in a cross section of the quenched cathode was obtained by secondary ion mass spectroscopy (CAMECA, NanoSIMS 50L) with a spatial resolution of 50 nm. Before the SIMS analysis, pores of the sample were infiltrated with an epoxy resin; then, the cross section of the cathode was polished using a cross section polisher to obtain a smooth surface. As shown in Fig. 1(a), SIMS images at the interface between the cathode and electrolyte were taken from both reference (for OCV) and working (for Biased) electrodes.
the imaging technique of active sites in porous electrodes through quenching SOFC reaction using a helium gas impinging jet and oxygen isotope labeling [22]. In this prior work, active sites of LSM/ScSZ (scandia-stabilized zirconia) porous cathode were qualitatively visualized at 800 °C in microstructure scale based on the 18O diffusion into LSM induced by overpotential. However, quantitative discussion of oxide ion flux incorporated to electrolyte phase (YSZ, ScSZ) is still lacking due to fast 18O diffusion in the electrolyte phase and resulting uniform 18O distribution [22]. In this communication, we first investigated quantitative oxide ion flux incorporated from the cathode/ electrolyte interface based on the analysis of the 18O diffusion profile in YSZ electrolyte at 700 °C. 2. Experimental In the experiment, a YSZ-electrolyte supported cell with a Ni/YSZ anode and LSM/YSZ (LSM:YSZ = 50:50 wt%) cathode was used. The fabrication process is given in our previous work [22]. As shown in Fig. 1(a), both the anode and cathode were separated into two parts to prepare reference electrodes. Pt wires for electric lead were attached to working electrodes, and thermocouple to anode reference using a silver paste. The cell was set to the power generation equipment with a watercooled nozzle for helium gas impinging jet as shown in Fig. 1(a). In this study, the cell was operated in a mixture of H2 (15 mL min−1) and Ar (15 mL min−1), which was humidified by a bubbler, for the anode side and in pure O2 (30 mL min−1) for the cathode side at 973 K. Although most researches and commercial SOFC stacks use air as an oxidant, pure oxygen was used in this study because oxide ion flux behaviour in pure O2 is a standard information to understand those in any reduced oxygen partial pressure including air composition. The observed OCV of 1.15 V and Nernst equation indicates that the supplied fuel composition is H2: Ar: H2O = 49.2: 49.2: 696
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Fig. 2. (a,b) High-resolution phase and c18O mappings of the cathode/electrolyte interface at (a) OCV and (b) Biased. c18O profiles in Line A and B are also shown in (a) and (b), respectively. (c) c18O line profiles (4 μm form cathode/electrolyte interface to inside YSZ electrolyte) denoted by Line 1–5 of (a) OCV and (b) Biased. Linear fitted curves of each profile and coefficient of variation (CV) of the gradient (dC18O/ dx ) are also shown.
3. Results and discussion
Fig. 2(a) and (b) show high-resolution phase and c18O mappings. The phase mappings were created based on the signal of Mn16O− (for LSM) and Zr16O− (for YSZ). As is the case in our previous study [22], 18O diffusion to LSM bulk was only observed at the region near the cathode/ electrolyte interface under biased condition, indicating that reaction sites near the interface are more electrochemically active than those at other regions. In addition, c18O in the electrolyte is uniform near the cathode/electrolyte interface at OCV while inhomogeneous at biased condition, which is related to the non-uniform oxide ion flux incorporated from the interface. 18O flux from the interface is proportional to c18O gradient dc18O/ dx . Here, c18O line profiles of YSZ electrolyte with 4 μm-length from cathode/electrolyte interface, denoted by Line 1–5 in Fig. 2(a) and (b), were linearly fitted as shown in Fig. 2(c). At OCV, the mean value, standard deviation (σ), and coefficient of variation (CV = σ/mean) of the gradient dc18O/ dx of 5 fitting curves are −2.32 × 10−3 μm−1, 2.79 × 10−4 μm−1, and 12.0%, respectively. In the case of biased condition, the mean, σ, and CV are −9.86 × 10−3 μm−1, 1.795 × 10−3 μm−1, and 18.2%, respectively. The increased CV at biased (with oxide ion flux) compared to OCV (without oxide ion flux) indicates that the oxide ion flux incorporated
3.1. Secondary ion mappings Fig. 1(c) shows 18O concentration c18O (= I(18O−)/[I(16O−) + I ( O−)]) mappings of the interface between the LSM/YSZ cathode and YSZ electrolyte at OCV and biased conditions. For biased condition, two images denoted by Biased #1 and #2 from different positions in the same cell were taken. Here, I(16O−) and I(18O−) indicate the SIMS signal intensity of each ion. Prior to the calculation of c18O image, binarization processing is performed on the 16O− image to remove all signal from pore phases. As shown in Fig. 1(c), c18O ≈ 0.04 of YSZ phases in cathode part is higher than that in the electrolyte (maximum c18O ≈ 0.025 at the surface) at OCV. This is because the surface to volume ratio of YSZ phase in cathode part is much larger than that in electrolyte part. At biased condition, c18O in YSZ phase is around 0.12–0.16 at cathode or electrolyte surface, which is higher than that at OCV because of additionally-incorporated 18O by the electrochemical reaction. Both in OCV and biased conditions, c18O in the electrolyte gradually decrease from cathode/electrolyte interface to anode side. 18
697
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Table 1 Oxygen tracer diffusion coefficient D* and surface exchange coefficient k* of YSZ electrolyte at 973 K obtained in this study and other references [14,15]. All literature values were measured at non-biased condition. Measurement
Sample
D* [m2 s−1] (at 973 K)
k* [m s−1] (at 973 K)
OCV Biased #1 Biased #2 Naito et al. [14]
8mol%Y2O3-ZrO2 Polycrystalline (with LSM/YSZ cathode) 9.1mol%Y2O3-ZrO2 Polycrystalline 2.8–16mol%Y2O3ZrO2 Single crystal, Polycrystalline
9.70 × 10−13 1.15 × 10−12 1.45 × 10−12 4.0 × 10−13
2.45 × 10−9 1.50 × 10−8 2.08 × 10−8 1.4 × 10−9
2.1 × 10−13 – 4.0 × 10−12
2.0 × 10−11 – 2.8 × 10−8
Knöner et al. [15]
side. Therefore, incorporated 18O itself does not move along the electrochemical potential gradient of oxide ion, ∇μ˜ O2 − , but moves along the 18 O concentration gradient ∇c18O . On the other hand, the one order enhancement of apparent k* indicates that 18O flux incorporated from the cathode/electrolyte interface to the electrolyte is greatly increased by applying the current. The detail of the 18O flux is quantitatively discussed in the next section. 3.3. Oxide ion flux incorporated from the cathode/electrolyte interface con(C/E) ex(C/E) , J18O , and As shown in Fig. 3(b), three kinds of 18O fluxes, J18O ion(C/E) 18 J18O contribute to a total O flux at the cathode/electrolyte interex(C/E) face. J18O is a18O flux by a16O/18O exchange at a free electrolyte ion(C/E) is an isotope oxide ion (18O2−) flux by an electrosurface and J18O
chemical reaction at a LSM/YSZ/gas triple phase boundary (TPB) on the electrolyte surface. In addition, c18O of YSZ phase in the cathode is generally higher than that of the electrolyte due to the higher surface to volume ratio of YSZ phase in the cathode, as mentioned previously. con(C/E) J18O is a18O flux by the c18O difference of YSZ phase between con(C/E) ex(C/E) cathode and electrolyte part. Here J18O and J18O flow at both OCV ion(C/E) exists only at biased. Therefore, the and biased conditions while J18O total 18O flux incorporated from the cathode/electrolyte interface to the Biased(C/E) OCV(C/E) electrolyte at OCV and biased conditions, J18O and J18O are described as follows.
Fig. 3. (a) Three (OCV, Biased #1, Biased #2) c18O line profiles inside YSZ electrolyte denoted by the arrows in Fig. 1(c). Analytical solution of diffusion equation is fitted to each data. The obtained and literature values of D* and k* are summarized in Table 1. (b) Three kinds of 18O fluxes incorporated from the cathode/electrolyte interface to the electrolyte. Obtained 18O fluxes are summarized in Table 2.
from the cathode/electrolyte interface is non-uniform in the observed region of Fig. 2(b). 3.2.
18
O diffusion profile analysis in YSZ electrolyte
Three (OCV, Biased #1, Biased #2) c18O line profiles inside YSZ electrolyte denoted by the arrows in Fig. 1(c) are shown in Fig. 3(a). In this study, polycrystalline YSZ (8mol%Y2O3-ZrO2) was used as an electrolyte. In all cases, typical monotonically decreasing diffusion profiles are observed, which can be analyzed by 1-dimensional diffusion in a semi-infinite media. As shown in Fig. 3(a), the analytical solution of diffusion equation [12] is well fitted to each profile, which gives oxygen tracer diffusion coefficient D* and surface exchange coefficient k*. The D* and k* obtained here and other references [14,15] are summarized in Table 1. At OCV, obtained D* = 9.70 × 10−13 m2 s−1 and k* = 2.45 × 10−9 m s−1 are relatively close to those of polycrystalline YSZ (9.1mol%Y2O3) by Naito et al. [14], and within the range of summarized data sets by Knöner et al. (2.8–16mol%Y2O3-ZrO2, single crystal and polycrystalline) [15]. When the current is applied, D* increased by a factor of only 1.2 to 1.5 compared to that of OCV while k* was enhanced by one order of magnitude, as shown in Table 1. The weak enhancement of D* means that the enhancement of 18O diffusion in YSZ bulk by the current is small. In a YSZ, oxide ion moves through vacancy diffusion mechanism [23]. In a fuel cell operation, when one oxide ion is incorporated to an electrolyte form a cathode, one oxide ion is emitted to an anode accompanied by a generation of oxygen vacancy in the electrolyte. Subsequently the generated oxygen vacancy moves from anode to cathode
OCV(C/E) ex(C/E) con(C/E) J18O = J18O + J18O
(1)
Biased(C/E) ex(C/E) con(C/E) ion(C/E) J18O = J18O + J18O + J18O
(2)
Biased(C/E) OCV(C/E) and J18O [mol m−2 s−1] corAs shown in Fig. 3(a), J18O responds to a c18O gradient at cathode/electrolyte interface ( x = 0 ), which can be described as follows. OCV(C/E) ⎡ ∂c18O ∗ J18O = −COYSZ DOCV ⎢ ∂x ⎣
OCV
⎤ ⎥ x=0 ⎦
Biased(C/E) ⎡ ∂c18O ∗ J18O = −COYSZ DBiased ⎢ ∂x ⎣
(3) Biased
⎤ ⎥ x=0 ⎦
(4)
∗ ∗ Here DOCV and DBiased [m2 s−1] represents the oxygen tracer diffusion coefficient at OCV and biased condition, respectively. COYSZ (= 9.41 × 104 mol m−3) is an absolute concentration of total oxygen atom in a YSZ calculated from the lattice structure of 8.3mol%Y2O3-ZrO2 Biased(C/E) (lattice constant: 5.135 Å [24]). The calculated J18O is one order of OCV(C/E) as summarized in Table 2. magnitude larger than J18O OCV(C/E) At OCV, J18O is also described by the surface exchange coeffi∗ cient kOCV [m s−1] as follows:
gas OCV(C/E) ex(C/E) con(C/E) ∗ J18O = J18O + J18O = COYSZ kOCV [c18O − c18O
where 698
gas c18O
is a relative
18
OCV x=0 ]
(5)
O concentration in gas phase (= 0.975). In the
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
coefficient, and 18O diffusion length into account in order to get useful O distribution image, as mentioned in section 2. Experimental.
Table 2 18 O flux incorporated from the cathode/electrolyte interface to the electrolyte estimated by diffusion profiles. Total18O2− flux at given current density is also shown.
18
4. Conclusions
Value [mol m−2 s−1]
Flux
Oxide ion flux at the interface between LSM/YSZ cathode and YSZ electrolyte of electrolyte-supported SOFC was investigated at 973 K through quenching reaction by helium gas-impinging jet and oxygen isotope labeling. From the analysis of 18O diffusion profiles in the quenched YSZ electrolyte measured by SIMS, it was disclosed that by applying a current, oxygen tracer diffusion coefficient D* increased by a factor of only 1.2–1.5 while surface exchange coefficient k* was enhanced by one order of magnitude. The greatly enhanced k* comes from oxide ion flux incorporated from the cathode/electrolyte interface to the electrolyte. Based on the 18O profiles, quantitative oxide ion flux incorporated from the cathode/electrolyte interface was first estimated. The obtained flux 1.01–1.43 × 10−3 mol m−2 s−1 at a current density of 0.09 A cm−2 indicates that 22–31% of the overall electrochemical reaction occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode under the present experimental condition.
2.18 × 10−4
OCV(C/E) J18O
1.21 × 10−3 1.63 × 10−3 1.01 × 10−3 1.43 × 10−3
Biased(C/E) J18O ion(C/E) J18O
(Biased (Biased (Biased (Biased
#1) #2) #1) #2)
gas = 0.975 ) 4.58 × 10−3 (i = 0.09 A/cm2, c18O
ion(all) J18O
above expression, it seems to be difficult to separate the contribution con(C/E) ex(C/E) from J18O and J18O precisely. Therefore, in order to calculate ex(C/E) con(C/E) ion(C/E) + J18O J18O in Eq. (2), J18O at biased condition is roughly estimated below. Both in OCV and biased condition, surface exchange coefficient on ex(C/E) free YSZ surface, which determines J18O , should be the same. In con(C/E) OCV(C/E) addition, J18O at OCV is not more than J18O (= 2.18 × 10−4 mol m−2 s−1) from Eq. (5). As shown in Fig. 3(b), the con(C/E) is the difference of c18O ( Δc18O ) between driving force of J18O cathode and electrolyte part. From c18O line profiles of Line A and B in Fig. 2(a) and (b), Δc18O at OCV (≈ 0.015) and biased condition (≈ 0.03) con(C/E) is the same order of magnitude. Therefore, J18O at biased condition is also estimated to be less than or equal to 10−4 order of magnitude. Biased(C/E) is 10−3 order of magnitude, rough estimation of Because J18O ex(C/E) con(C/E) J18O + J18O at biased condition (10−4 order of magnitude) does ion(C/E) not strongly affect J18O . con(C/E) ex(C/E) and J18O at biased Based on the above discussion, sum of J18O ∗ condition was approximately estimated using kOCV as follows. Biased
ex(C/E) con(C/E) [J18O + J18O ]
gas ∗ = COYSZ kOCV [c18O − c18O
Biased x=0 ]
Acknowledgements This work was partly supported by the Japan Society for the Promotion of Science KAKENHI Grant Numbers 16K14171 and 16J03479, and The Hattori-Hokokai Foundation. SIMS analysis was conducted at the Advanced Characterization Nanotechnology Platform of the University of Tokyo, supported by the Nanotechnology Platform of the Ministry of Education, Culture, Sports, Science and Technology, Japan. The authors would like to thank Dr M. Takeuchi for support with the SIMS measurement.
(6)
References
ion(C/E) J18O ,
which are Substitution of Eqs. (4) and (6) into (2) gives summarized in Table 2. The total 18O2− flux incorporated from whole gas ion(all) = c18O i/2F at given current density i [A m−2] is also cathode J18O shown in this table. Under the assumption that whole cathode area contributes to the electrochemical reaction equivalently, the ratio of the ion(C/E) ion(all) (= 1.01–1.43 × 10−3 mol m−2 s−1) and J18O (= obtained J18O 4.58 × 10−3 mol m−2 s−1) indicates that 22–31% of the overall electrochemical reaction (oxide ion incorporations to YSZ phase) occurs at the cathode/electrolyte interface, while the remaining 69–78% of those proceeds inside the porous cathode in the present experimental condition. In a state-of-the-art 3-dimensional numerical simulation of LSM/ YSZ porous cathode (LSM:YSZ = 50:50 wt%) based on the FIB-SEMobserved actual microstructure by Miyoshi et al. [25], charge transfer current distribution in the cathode thickness direction was calculated at 850 °C. In their result, 22.5% of the overall electrochemical reaction occurs at the region within 1.25 μm from the cathode/electrolyte interface, which is relatively close to the scale of one LSM particle on the electrolyte as shown in Fig. 2(b). As shown here, present work offers the first experimental quantification of oxide ion flux inside a SOFC porous cathode, making it possible to compare experimental and numerical reaction distributions. As a future direction, the correlation between oxide ion flux from a cathode/electrolyte interface to an electrolyte and several parameters such as current density, cathode overpotential and impedance, oxygen concentration, and operating temperature can be experimentally investigated by using the current method. Such information can provide both deeper understanding of electrode reaction mechanisms and guidance for electrode design optimization. In addition, other state-of-the-art cathode materials, e.g., (La,Sr)CoO3 (LSC), (La,Sr) (Co,Fe)O3 (LSCF), or (Ba,Sr) (Co,Fe)O3 (BSCF), can be studied, though operating temperature and dwell time in 18O2 should be carefully selected taking oxide ionic conductivity, tracer diffusion
[1] J. Hanna, W.Y. Lee, Y. Shi, A.F. Ghoniem, Fundamentals of electro- and thermochemistry in the anode of solid-oxide fuel cells with hydrocarbon and syngas fuels, Prog. Energy Combust. Sci. 40 (2014) 74–111. [2] T. Nagasawa, K. Hanamura, Prediction of overpotential and effective thickness of Ni/YSZ anode for solid oxide fuel cell by improved species territory adsorption model, J. Power Sources 353 (2017) 115–122. [3] D. Chen, W. Bi, W. Kong, Z. Lin, Combined micro-scale and macro-scale modelling of the composite electrode of a solid oxide fuel cell, J. Power Sources 195 (2010) 6598–6610. [4] P. Costamagna, P. Costa, V. Antonucci, Micro-modelling of solid oxide fuel cell electrodes, Electrochim. Acta 43 (1998) 375–394. [5] J.H. Nam, D.H. Jeon, A comprehensive micro-scale model for transport and reaction in intermediate temperature solid oxide fuel cells, Electrochim. Acta 51 (2006) 3446–3460. [6] M. Kishimoto, H. Iwai, M. Saito, H. Yoshida, Three-dimensional simulation of SOFC anode polarization characteristics based on sub-grid scale modeling of microstructure, J. Electrochem. Soc. 159 (3) (2012) B315–B323. [7] T. Carraro, J. Joos, B. Rüger, A. Weber, E. I-Tiffée, 3D finite element model for reconstructed mixed-conducting cathode: I. Performance quantification, Electrochim. Acta 77 (2012) 315–325. [8] A. He, Y. Kim, N. Shikazono, Three-dimensional numerical simulation of solid oxide fuel cell cathode based on lattice Boltzmann method with sub-grid scale models, Int. J. Hydrogen Energy 42 (2017) 21886–21900. [9] I. Yasuda, K. Ogasawara, M. Hishinuma, T. Kawada, M. Dokiya, Oxygen tracer diffusion coefficient of (La, Sr)MnO3 ± δ, Solid State Ionics 86–88 (1996) 1197–1201. [10] J.A. Kilner, S.J. Skinner, H.H. Brongersma, The isotope exchange depth profiling (IEDP) technique using SIMS and LEIS, J. Solid State Electrochem. 15 (5) (2011) 861–876. [11] A. Esquirol, J. Kilner, N. Brandon, Oxygen transport in La0.6Sr0.4Co0.2Fe0.8O3-δ/ Ce0.8Ge0.2O2-x composite cathode for IT-SOFCs, Solid State Ionics 175 (2004) 63–67. [12] R.A. Souza, J.A. Kilner, J.F. Walker, A SIMS study of oxygen tracer diffusion and surface exchange in La0.8Sr0.2MnO3+δ, Mater. Lett. 43 (2000) 43–52. [13] E.S. Raj, A. Atkinson, J.A. Kilner, Oxygen diffusion studies on (Y2O3)2(Sc2O3)9(ZrO2)89, Solid State Ionics 180 (2009) 952–955. [14] H. Naito, N. Sakai, T. Otake, H. Yugami, H. Yokokawa, Oxygen transport properties in ZrO2-CeO2-Y2O3 by SIMS analyses, Solid State Ionics 135 (2000) 669–673. [15] G. Knöner, K. Reimann, R. Röwer, U. Södervall, H.-E. Schaefer, Enhanced oxygen diffusivity in interfaces of nanocrystalline ZrO2•Y2O3, Proc. Natl. Acad. Sci. Unit.
699
Journal of Power Sources 412 (2019) 695–700
T. Nagasawa, K. Hanamura
Electrochem. Solid State Lett. 13 (12) (2010) B135–B138. [21] T. Horita, M. Nishi, T. Shimonosono, H. Kishimoto, K. Yamaji, M.E. Brito, H. Yokokawa, Visualization of oxide ionic diffusion at SOFC cathode/electrolyte interfaces by isotope labeling techniques, Solid State Ionics 262 (2014) 398–402. [22] T. Nagasawa, K. Hanamura, Microstructure-scaled active sites imaging of a solid oxide fuel cell composite cathode, J. Power Sources 367 (2017) 57–62. [23] H. Tagawa, Solid Oxide Fuel Cell and Global Environment, Agne-Shofu-Sya, 1998 (in Japanese). [24] H. Hayashi, T. Saitou, N. Maruyama, H. Inaba, K. Kawamura, M. Mori, Thermal expansion coefficient of yttria stabilized zirconia for various yttria contents, Solid State Ionics 176 (2005) 613–619. [25] K. Miyoshi, T. Miyamae, H. Iwai, M. Saito, M. Kishimoto, H. Yoshida, Exchange current model for (La0.8Sr0.2)0.95MnO3 (LSM) porous cathode for solid oxide fuel cells, J, Power Sources 315 (2016) 63–69.
States Am. 100 (7) (2003) 3870–3873. [16] T. Kawada, T. Horita, N. Sakai, H. Yokokawa, M. Dokiya, J. Mizusaki, A novel technique for imaging electrochemical reaction sites on a solid oxide electrolyte, Solid State Ionics 131 (2000) 199–210. [17] T. Horita, K. Yamaji, N. Sakai, Y. Xiong, T. Kato, H. Yokokawa, T. Kawada, Imaging of oxygen transport at SOFC cathode/electrolyte interfaces by novel technique, J. Power Sources 106 (2002) 224–230. [18] A.K. Opitz, M. Kubicek, S. Huber, T. Huber, G. Holzlechner, Thin film cathodes in SOFC research: how to identify oxygen reduction pathways? J. Mater. Res. 28 (16) (2013) 2085–2105. [19] J.H. Shim, J.S. Park, T.P. Holme, K. Crabb, W. Lee, Y.B. Kim, X. Tian, T.M. Gür, F.B. Prinz, Enhanced oxygen exchange and incorporation at surface grain boundaries on an oxide ion conductor, Acta Mater. 60 (2012) 1–7. [20] T. Horita, T. Shimonosono, H. Kishimoto, K. Yamaji, M.E. Brito, Y. Hori, H. Yokokawa, Visualization of oxygen ionization and flows in solid oxide fuel cells,
700