Electrical conductivity relaxation of Sr₂Fe_1.5Mo_0.5O_6−δ–Sm_0.2Ce_0.8O_1.9 dual-phase composites

Abstract

The oxygen incorporation kinetics of Sr₂Fe_1.5Mo_0.5O_6−δ–Sm_0.2Ce_0.8O_1.9 (SFM–SDC) dual-phase composites has been investigated, at 750 °C, as a function of SDC phase volume fraction using electrical conductivity relaxation. It is shown that the oxygen re-equilibration kinetics in the range of oxygen partial pressure (pO₂) from 0.01 to 1 atm is limited by the surface exchange rate. The effective surface exchange coefficient of the composites is found to increase profoundly upon increasing the phase volume fraction of the oxide electrolyte phase SDC. The results are interpreted to reflect the synergistic oxygen incorporation at the SFM–SDC–gas triple phase boundaries (TPBs), which occurs in addition to the direct incorporation via the surface of the perovskite mixed conductor SFM. Already at a SDC phase volume fraction of 0.105, the uptake of oxygen via the synergistic TPB route (referred to as route III), following a step change in the surrounding pO₂, comprises more than 75% of the overall uptake of oxygen by the composite. It is further concluded that under the conditions of the experiments the two-phase SFM–SDC boundaries allow for a facile exchange of oxygen ions between both involved phases.

---

1. Introduction

Mixed oxide ionic–electronic conductors are of great interest for high temperature electrochemical devices. These materials are used as porous electrodes for, e.g., solid oxide fuel cells (SOFCs) and gas sensors, to extend the reaction sites beyond the triple phase boundaries (TPBs).^1,2 They can also be used as dense ceramic membranes for, e.g., integrated oxygen separation, steam reforming and partial oxidation for the conversion of natural gas into synthesis gas.^3,4 Usually, the mixed conductors include single-phase perovskite oxides like La_1−xM_xCo_1−yFe_yO_3−δ (M = Sr, Ba and Ca),⁵ Sm_0.5Sr_0.5CoO_3−δ (ref. 6) and Ba_0.5Sr_0.5Co_0.2Fe_0.8O_3−δ.⁷ The mixed conducting properties can be tailored for the so-called dual-phase composites, comprising an oxygen ion conductor and an electronic conductor such as Ag–(BaO)_0.2(Bi₂O₃)_0.8 and Pd–yttria-stabilized zirconia (YSZ).^8,9 Recently, ceramic dual-phase composites comprising an oxygen ion conductor and a mixed conductor like Ce_0.8Gd_0.2O_1.9–La_0.7Sr_0.3MnO_3−δ, Gd_0.2Ce_0.8O_1.9–La_0.8Sr_0.2Co_0.2Fe_0.8O_3−δ and Ce_0.8Sm_0.2O_1.9–La_0.8Sr_0.2CrO_3−δ have been investigated.^10,11

The oxygen surface exchange reaction plays an important role in governing the performance of electrochemical devices. It is well known that the oxygen surface exchange at the cathode of the SOFC contributes significantly to the total energy losses, particularly at low temperatures.¹² For oxygen separation membranes, the oxygen surface exchange plays a key role in optimisation of the oxygen flux. Below the so-called characteristic membrane thickness, L_c, the oxygen flux becomes predominantly limited by the surface exchange kinetics.¹³

The oxygen surface exchange reaction for a single oxide phase is characterized by a surface exchange coefficient, k, which can be determined by oxygen isotopic exchange depth profiling (IEPD) using, e.g., secondary ion mass spectrometry (SIMS).¹⁴ Also, chemical relaxation methods can be used for evaluation of k, provided that a step change in the oxygen partial pressure brings about a measurable change in either weight¹⁵ or conductivity.¹⁶ Numerous studies have been performed on, e.g., perovskites La_0.6Sr_0.4Co_1−yFe_yO_3−δ (LSCF),¹⁷ La_1−xSr_xMnO_3−δ (LSM),¹⁸ YSZ (ref. 19) and doped ceria.²⁰ Few investigations have been conducted so far on dual-phase composites. Ji et al.¹⁴ studied the oxygen surface exchange kinetics on YSZ–LSM composites, and found the apparent surface exchange coefficient of LSM to be significantly enhanced upon the addition of YSZ. In our recent studies,^21–23 we found significant enhancement of the surface exchange kinetics of mixed conducting perovskite type oxides Sr₂Fe_1.5Mo_0.5O_6−δ (SFM), LSM, and LSCF after surface coating with solid oxide electrolytes such as YSZ or samarium-doped ceria (SDC).

The above results give evidence that the overall rate of oxygen surface exchange may be significantly enhanced when heterogeneous interfaces are exposed to the gas phase. In the context of SOFC cathode-based studies, these phase boundaries are referred to as triple phase boundaries (TPBs), and are considered to act as the preferred sites for oxygen incorporation into the electrolyte phase. The aim of this work is to further explore the oxygen surface exchange behaviour of composite materials. Electrical conductivity relaxation (ECR) measurements are employed for determination of the effective oxygen surface exchange coefficient of composites in the SFM–SDC system. SDC is a known solid oxide electrolyte,²⁴ while SFM is a mixed ionic-electronic conductor (MIEC), showing electronic and ionic conductivities that are comparable to those reported for the known cobalt-containing perovskite-type cathodes.^25,26 To the best of our knowledge, this is the first study using ECR for evaluation of the surface exchange kinetics of dual-phase composites.

2. Theoretical approach

2.1 Single-phase mixed conductors

Electrical conductivity relaxation (ECR) is a well-established method for evaluation of the oxygen transport parameters of mixed ionic-electronic conducting (MIEC) oxides.^16,17 The method is based on the relationship between the electrical conductivity and the oxygen concentration. The change in conductivity, which reflects the change in the oxygen concentration of the sample, after a step-wise change in the ambient pO₂ is recorded as a function of time. The obtained data are fitted to the appropriate solution of Fick's second law, assuming linear kinetics for the surface exchange reaction, and a linear relationship between the sample conductivity and the oxygen ion concentration within the applied pO₂ step change. The corresponding analytical solution is given by^16,18where g(t) is the normalized conductivity, σ(0) and σ(∞) are the initial and final conductivities, respectively, and c(0) and c(∞) are the corresponding oxygen concentrations. Parameters x, y and z are the sample dimensions, while β_m, γ_n, and ϕ_p are the positive, non-zero roots of

β_m tan β_m = L_β; γ_n tan γ_n = L_γ; ϕ_p tan ϕ_p = L_ϕ (2)

with eigenvalueswhere

The parameters obtained from fitting are the chemical surface exchange coefficient, k (m s⁻¹) and the chemical diffusion coefficient, D (m² s⁻¹). The equilibration towards the new oxygen concentration is predominantly limited by diffusion for samples with a thickness much larger than the characteristic thickness L_c, and by surface exchange when the thickness is much smaller than L_c. Values for both D and k can be obtained only when the relaxation process is under the mixed control of diffusion and surface exchange.²⁷ In the absence of diffusion limitations, the normalized conductivity transient reduces to

g(t) = 1 − exp(−t/τ) (5)

where τ = V/kS, in which S (m²) and V (m³) are the surface area exposed to the gas phase and the sample volume, respectively. It may be noted that eqn (5) follows immediately from integration of the linear surface exchange rate equation used in the derivation of eqn (1):

2.2 Dual-phase mixed conductor–ionic conductor composite

In this study, ECR measurements are conducted on SFM–SDC dual phase composites, at 750 °C, in the range of pO₂ between 0.01 and 1 atm. In this range, the oxygen stoichiometry of SDC is known to be virtually constant.²⁴ The conductivity change recorded after a pO₂ step change will thus reflect only a change in the oxygen concentration of the SFM phase. Furthermore, it is assumed that both constituent phases in the composite are randomly distributed. Fitting the normalised conductivity transient to eqn (1) yields ‘effective’ quantities for the diffusion and surface exchange coefficients. A notably faster surface exchange kinetics is observed for the composite samples than for pure SFM, which is henceforth referred to as the ‘synergistic’ effect. The results from this study indicate that, at the conditions studied and at all SDC phase volume fractions, the oxygen re-equilibration kinetics remains primarily controlled by the rate of surface exchange (see Section 4.2).

Fig. 1 schematically shows possible routes of oxygen incorporation into the SFM–SDC composite. The exchange route, route I, proceeds via the SFM surface. This route is similar to that observed for single-phase SFM. O₂ molecules are dissociated on the SFM surface, followed by incorporation of the formed oxygen adatoms into SFM. By analogy with eqn (6), the exchange rate is given by

where f^V_SFM and f^S_SFM are the SFM phase fractions of the volume and surface area exposed to an ambient atmosphere, respectively, and the subscript ‘I’ for the surface exchange coefficient denotes that exchange proceeds solely via route I. The normalized conductivity transient after a pO₂ step change is given by eqn (5) with τ = f^V_SFMV/(f^S_SFMSk_I).

Fig. 1 Possible routes for oxygen incorporation in SFM–SDC dual-phase composites.

The exchange route, route II, proceeds via the SDC surface. Dissociation of O₂ and incorporation of oxygen adatoms are presumed to occur on the SDC surface. The oxygen ions entering the SDC lattice are subsequently transported to the SFM phase by diffusion. As k reported for SDC is ∼10⁻⁸ cm s⁻¹ at 750 °C at a pO₂ of 0.93 atm,²⁸ which is about 3 orders of magnitude lower than that measured under similar conditions for pure SFM,²¹ the contribution of route II to the overall surface exchange kinetics of the composite can be neglected.

Oxygen exchange via route III occurs at the TPB, the triple-phase boundary between SFM, SDC and gas phase. This route is expected to play a significant if not dominant role in determining the oxygen exchange rate of SFM–SDC composites. Electronic charge carriers in mixed conducting SFM are transported to the closest TPB, facilitating charge transfer to oxygen adsorbates, while fast oxygen diffusion occurs through the SDC electrolyte and/or the SFM–SDC boundary. Without making any statement of the chemical and electronic nature of oxygen species at the surface, some widening of the TPB region could take place due to the spill-over of oxygen before incorporation into the SDC electrolyte phase. If the latter is left out of consideration and oxygen diffusion in SDC, or along grain-boundaries, is not rate-determining, the contribution of route III to the overall exchange rate should scale either with the total TPB length or with the total surface area of the heterogeneous SFM–SDC two-phase boundary.

In the following, it is assumed that the oxygen re-equilibration kinetics of SFM–SDC composites after a pO₂ step change is controlled by the oxygen surface exchange kinetics. It is further assumed that surface exchange proceeds only via routes I and III as described above. Adopting similar rate equations for routes I and III, the overall surface exchange rate can be expressed as

where k_III is the surface rate coefficient corresponding to route III. Note that k_III is normalized on the total geometrical surface area. Eqn (8) can be rewritten in terms of the effective exchange coefficient, k_eff,where

k_eff = f^S_SFMk_I + k_III (10)

The corresponding normalized conductivity transient is of the form of eqn (5) with time constant τ = f^V_SFMV/(k_effS). Note that τ decreases upon decreasing the volume fraction of the SFM phase in the sample, f^V_SFM, and upon increasing the surface exchange rate, k_eff. If only route I contributes to oxygen exchange, eqn (10) reduces to k_eff = f^S_SFMk_I. Accordingly, a rate enhancement factor can be defined as

whose ratio expresses the rate of oxygen exchange in the presence of synergistic route III relative to the exchange rate for pure SFM. The total amount of oxygen incorporated into the composite upon a step-wise increase of the oxygen partial pressure equals H_tot = f^V_SFMV[c(∞) − c(0)]. The fractional uptake of oxygen via route III is

Eqn (8) relies on a specific form for the rate equation of oxygen exchange via route III (chosen to be similar to that of route I). Alternatively, the overall exchange rate equation can be written as

where the second term on the right hand represents the contribution of route III to oxygen exchange, in which the quantity H denotes the corresponding number of moles of oxygen transferred via route III. It is easily shown that Ω_III can also be computed from the area between the experimental, normalized conductivity transient, g_exp(t), and that calculated assuming that only route I is active, g^I_calc(t):as is graphically illustrated in Fig. 2.

Fig. 2 Normalized transient conductivities after a step-wise change in pO₂: experimental conductivity, g_exp(t), and calculated conductivity for route I, g^I_calc(t). The dashed area is used to calculate the fractional uptake of oxygen via route III (in accordance with eqn (14)).

3. Experimental

All chemicals were of analytical grade and obtained from Sinopharm Chemical Reagent Co. (China). SDC was prepared by oxalate co-precipitation.²⁹ Cerium and samarium nitrates (99.95%) were used as the cation sources, and ammonia carbonate as the precipitant. The precipitate was calcined, at 600 °C, in air for 2 h. X-ray powder diffraction (XRD, Rigaku TTR-III) confirmed that the desired fluorite structure was formed. No evidence was found for second phase impurities. SFM was synthesized by a microwave-assisted combustion method.^21,25 Sr(NO₃)₂ (99.5%), Fe(NO₃)₃·9H₂O (98.5%) and (NH₄)₆Mo₇O₂₄·4H₂O (99.0%) were used as the metal precursors. Glycine (99.95%) and citric acid (99.95%) were used to assist the combustion process. The as-prepared ash was calcined, at 1000 °C, for 5 h under air to form the SFM powder with the perovskite structure.

SDC and SFM powders were mixed by ball milling for 2 h. Powders with SDC weight percentages 15, 30, 45, 60 and 75 wt% (which correspond to SDC phase volume fractions 0.105, 0.223, 0.353, 0.501 and 0.667, respectively) were prepared. These were uniaxially pressed (320 MPa) into rectangular bars, and fired at 1350 °C for 5 h under air to achieve dimensions of about 20 × 4.8 × 0.6 mm³. Both the density measured by the Archimedes method and the geometrical density of the samples were above 96% of theoretical value. The surface morphology of the samples was investigated by scanning electron microscopy (SEM, JSM-6700F). The length of the SFM–SDC heterogeneous phase boundary (TPB length) and the surface area of both phases were determined using Image-plus Software (Media Cybernetics Company). For each sample, images from at least five different areas of the sample were used to obtain average data.

The electrical conductivity was measured using the four-point method with a digital multimeter (Keithley, 2001-785D). Electrical contacts were applied by means of silver wires (0.10 mm diameter) and silver paste (both from Sina-Platinum Metals Co., Ltd.), fired at 600 °C in air.

For ECR experiments, the sample was mounted in a quartz tube, heated to 750 °C, and equilibrated at this temperature and at a pO₂ of 0.01 atm for 1 h prior to measurements. The heating rate was 3 °C min⁻¹. Next, the oxygen partial pressure of the gas was changed abruptly, and the change in conductivity recorded until the new equilibrium was established. Different O₂–N₂ gas mixtures were used to subject the sample to step changes in pO₂. Gas switches from 0.01 to 0.1 atm, and from 0.1 to 1 atm, at a gas flow rate of 200 ml min⁻¹ were used for the experiments. The gas switches were realized in less than 1 s. Data acquisition was carried out until the relative change in conductivity was less than 0.5% within 20 min.

4. Results and discussion

4.1 Surface microstructures

Fig. 3 shows backscattered electron SEM images of the SFM–SDC composites with different volume fractions of the SDC phase. The two constituent phases are apparent with different grey levels. The lighter grey particles are grains from SDC, while the darker ones are grains from SFM (the coloration is proportional to the average atomic number of the phase). No defects or pinholes are detected. The grain-boundaries between the two phases are clean, suggesting that no obvious solid state reaction occurs between SDC and SFM. This observation is in good agreement with a previous study based on XRD reported by Liu et al.²⁶

Fig. 3 SEM surface images of SFM–SDC composites with different SDC phase volume fractions: (a) 0, (b) 0.105, (c) 0.223, (d) 0.353, (e) 0.501, and (f) 0.667.

The SEM micrographs in Fig. 3 show that the grains of both phases in the composite are homogeneously dispersed. The average grain size for SDC increases slightly from a value of 1.0 μm to 1.7 μm upon increasing its volume fraction, f^V_SDC, from 0.105 to 0.667. Concurrently, the average grain size of SFM decreases from about 3 to 1 μm. For f^V_SDC = 0.105, the SDC surface grains are completely isolated. When the SDC content increases, more and more SDC grains are connected to each other. Table 1 lists values (normalised per unit area) of the surface areas of SFM and SDC and the TPB length derived from the SEM micrographs using image analysis. The TPB length is plotted as a function of the surface area fraction of the SDC phase, f^V_SDC, in Fig. 4. It can be noted from this figure that a maximum TPB length is found at f^V_SDC ≈ 0.5, as expected, given the similar ranges of particle sizes for SFM and SDC.

Table 1 Microstructural characteristics of SFM–SDC composites

SDC volume fraction f^VSDC	SFM volume fraction f^VSFM	SDC surface area fraction f^SSDC	SFM surface area fraction f^SSFM	Sample thickness z/10^−4 m	TPBlength (per unit area)/10^6 m^−1
0	1	0	1	6.2 ± 0.2	0
0.105	0.895	0.103 ± 0.005	0.897 ± 0.005	6.4 ± 0.3	0.505 ± 0.023
0.223	0.777	0.183 ± 0.011	0.817 ± 0.011	6.6 ± 0.1	0.851 ± 0.041
0.353	0.647	0.314 ± 0.004	0.686 ± 0.004	6.1 ± 0.1	1.38 ± 0.018
0.501	0.499	0.423 ± 0.025	0.577 ± 0.025	6.1 ± 0.1	1.744 ± 0.032
0.667	0.333	0.669 ± 0.021	0.331 ± 0.021	6.2 ± 0.2	1.46 ± 0.055

---

Fig. 4 Triple phase boundary length, TPB_length, of SFM–SDC composites as a function of SDC phase volume fraction, f^V_SDC. The dashed line is drawn to guide the eye.

4.2 Electrical conductivity

Fig. 5 shows the conductivity of SFM–SDC composites, at 750 °C, at different oxygen partial pressures. As expected from the intrinsic conductivities of both constituent phases, the conductivity of SFM–SDC composites drops upon increasing the SDC content. At a fixed volume fraction of the SDC phase, the conductivity decreases with decreasing pO₂, which is due to the p-type conductivity of SFM. At the maximum volume fraction of SDC in this study, f^V_SDC = 0.667, the microstructure of the composite still offers percolation for electrical conduction. The corresponding conductivity is found to be almost one order of magnitude higher than that found for pure SDC.²⁴

Fig. 5 Electrical conductivity of SFM–SDC composites, at 750 °C, as a function of SDC phase volume fraction, f^V_SDC, measured at different oxygen partial pressures.

4.3 Electrical conductivity relaxation

Fig. 6a and b show the time dependences of the normalised conductivity for pure SFM and SFM–SDC composites, following pO₂ step changes from 0.01 to 0.1 atm, and from 0.1 to 1. The data for the entire series were fit assuming pure surface-controlled equilibration kinetics (eqn (5)). Fitting instead with eqn (1), assuming the mixed control of diffusion and surface exchange, gives eigenvalues L_α, L_β, and L_γ significantly smaller than 0.03. As discussed by Den Otter et al.,²⁷ if this criterion is met the relaxation process is predominantly controlled by the surface exchange kinetics. Error plots of D and k obtained from fitting showed that for all compositions investigated in this work, the goodness-of-fit by means of the minimum standard error of regression (SER) is invariant to the value of D (see ESI†).

Fig. 6 Normalized conductivity relaxation transients for pO₂ step changes: (a) 0.01 → 0.1 atm and (b) 0.1 → 1 atm. The right-hand figures show a magnification of the transients in the initial time period. The dashed lines represent the fits to eqn (5).

Effective exchange coefficients, k_eff, obtained from the fitting procedure as a function of f^V_SDC are plotted in Fig. 7. The results show that the exchange rate is significantly enhanced by the dispersion of SDC grains into the SFM host. For the pO₂ step change 0.1 → 1 atm, k_eff increases from 3.7 × 10⁻⁷ m s⁻¹ for pure SFM to 16.8 × 10⁻⁷ m s⁻¹ for the composite with f^V_SDC = 0.667. Fig. 8 shows the calculated enhancement of the exchange rate relative to that contributed via route I (see Section 2) by grains of SFM as a function of f^V_SDC. The representative parameter Λ is found to increase with increasing f^V_SDC (besides some inflection at f^V_SDC ≈ 0.35), reaching a value of 13.9 at f^V_SDC = 0.667. Already at f^V_SDC = 0.105, the uptake of oxygen via route III (calculated using eqn (12)) comprises more than 75% of the overall uptake of oxygen by the composite, as can be judged from the data presented in Fig. 9. At the highest values of f^V_SDC in this work, these percentages are in excess of 90%. Fig. 10 compares corresponding values of Ω_III calculated using eqn (12) and (14). The good agreement obtained supports underlying assumptions made in derivations of both equations. Finally, as can be judged from the data presented in Fig. 7–9, similar results are obtained for pO₂ step changes 0.01 → 0.1 atm and 0.1 → 1 atm. Numerical values of k_eff, k_III, Λ, and Ω_III evaluated from fitting the conductivity relaxation data for both pO₂ steps are provided in the ESI.†

Fig. 7 Effective surface exchange coefficients, k_eff, of SFM–SDC composites as a function of SDC phase volume fraction, f^V_SDC, obtained from fitting experimental data (Fig. 6) to eqn (5). Dashed lines are drawn to guide the eye.

Fig. 8 Exchange rate enhancement parameter, Λ, a function of SDC phase volume fraction. (For the definition of Λ, see eqn (11).) Dashed lines are drawn to guide the eye.

Fig. 9 Fractional uptake of oxygen via route III (relative to the overall oxygen uptake) as a function of SDC phase volume fraction, f^V_SDC. The parameter Ω_III was calculated using eqn (12). Dashed lines are drawn to guide the eye.

Fig. 10 Comparison of fractional oxygen uptake parameters, Ω_III, calculated using eqn (12) and (14). A perfect agreement is obtained when the points coincide with the line y = x.

4.4 Relationship of oxygen surface exchange with microstructures

In the following, correlation of the rate of exchange via route III with microscopic data of both constituent phases in the SFM–SDC composites is attempted. As discussed in the theoretical section (Section 2.2), the contribution of route III to the overall exchange rate is expected to increase with extension of the geometrical TPB length. Fig. 11 shows a plot of the normalized value of k_III (per unit TPB length) as a function of the SDC volume fraction, f^V_SDC. It is clearly seen from this figure that such an analysis suggests that already at comparatively low values of f^V_SDC, the TPB becomes less effective. The normalized value of k_III is found to drop almost a factor of 2 beyond f^V_SDC ≈ 0.25. At present, we do not have a clear-cut explanation for this behaviour. The observations can neither be accounted for by an effective width of the TPB, nor by taking into consideration issues related to percolation of the SDC phase. Furthermore, it cannot be ruled out that oxygen exchange on SFM–SDC composites is controlled by multiple rate determining steps, whose proportion varies with the volume fractions of the constituent phases.

Fig. 11 Surface exchange coefficient k_III (normalized per unit TPB length) as a function of SDC phase volume fraction, f^V_SDC. Dashed lines are drawn to guide the eye.

Noting the absence of diffusion limitations in the oxygen re-equilibration kinetics of SFM–SDC composites, another possibility we have considered is the variation of exchange rate k_III on change of the (bulk) heterogeneous two-phase boundary area. The extent of interfacial area between both phases increases upon increasing the SDC phase volume fraction, f^V_SDC, below ∼0.5, and decreases above this value. The impact on k_III, however, would be the most significant as the SDC fraction reaches the percolation threshold. Qualitatively, we expect a percolative ionic conductive pathway above f^V_SDC ≈ 0.3.³⁰ Since a profound increase of k_III is not observed at the latter volume fraction, the possibility of a relationship between k_III and two-phase boundary area was abandoned. It may, however, be inferred from the present results that the two-phase boundary area between SFM and SDC enables facile exchange of oxygen ions between both phases. This would be consistent with data from our previous study where we found the oxygen exchange rate of SFM to be significantly enhanced after surface coating with SDC.^21–23

5. Conclusions

The present study clearly demonstrates that the oxygen incorporation kinetics of SFM–SDC composites is significantly enhanced relative to that observed for the pure perovskite mixed conductor SFM. Already at the minimum SDC phase volume fraction of 0.105 in this work, the oxygen incorporation kinetics of the composite appears to be dominated by the synergistic TPB route (referred to as route III). The corresponding uptake of oxygen under the conditions studied comprises 75% of the overall oxygen uptake, increasing to values over 90% upon further increasing the phase volume fraction of SDC in the composite. It must, however, also be concluded that additional studies are required to fully understand the relationship between the effective surface exchange coefficient and the microstructure of the composites. Particularly useful would be to study the effect of grain size of both constituent phases (e.g., at a fixed composition) on oxygen incorporation kinetics.

Acknowledgements

This work was financially supported by the Ministry of Science and Technology of China (2012CB215403).

Notes and references

1. I. Riess, M. Godickemeier and L. J. Gauckler, Solid State Ionics, 1996, 90, 91–104.
2. J. J. Sprague, O. Porat and H. L. Tuller, Sens. Actuators, B, 1996, 35, 348–352.
3. J. Sunarso, S. Baumann, J. M. Serra, W. A. Meulenberg, S. Liu, Y. S. Lin and J. C. Diniz da Costa, J. Membr. Sci., 2008, 320, 13–41.
4. Q. Y. Jiang, S. Faraji, D. A. Slade and S. M. Stagg-Williams, Membr. Sci. Technol., 2011, 14, 235–273.
5. J. W. Stevenson, T. R. Armstrong, R. D. Carneim, L. R. Pederson and W. J. Weber, J. Electrochem. Soc., 1996, 143, 2722–2729.
6. C. R. Xia, W. Rauch, F. L. Chen and M. L. Liu, Solid State Ionics, 2002, 149, 11–19.
7. Z. P. Shao and S. M. Haile, Nature, 2004, 431, 170–173.
8. T. H. Lee, Y. M. L. Yang and A. J. Jacobson, Solid State Ionics, 2000, 134, 331–339.
9. C. S. Chen, B. A. Boukamp, H. J. M. Bouwmeester, G. Z. Cao, H. Kruidhof, A. J. A. Winnubst and A. J. Burggraaf, Solid State Ionics, 1995, 76, 23–28.
10. V. V. Kharton, A. V. Kovalevsky, A. P. Viskup, A. L. Shaula, F. M. Figueiredo, E. N. Naumovich and F. M. B. Marques, Solid State Ionics, 2003, 160, 247–258.
11. J. X. Yi, Y. B. Zuo, W. Liu, L. Winnubst and C. S. Chen, J. Membr. Sci., 2006, 280, 849–855.
12. M. de Ridder, R. G. van Welzenis, H. H. Brongersma and U. Kreissig, Solid State Ionics, 2003, 158, 67–77.
13. H. J. M. Bouwmeester, H. Kruidhof and A. J. Burggraaf, Solid State Ionics, 1994, 72, 185–194.
14. Y. Ji, J. A. Kilner and M. F. Carolan, Solid State Ionics, 2005, 176, 937–943.
15. M. Katsuki, S. R. Wang, K. Yasumoto and M. Dokiya, Solid State Ionics, 2002, 154–155, 589–595.
16. J. A. Lane and J. A. Kilner, Solid State Ionics, 2000, 136–137, 997–1001.
17. H. J. M. Bouwmeester, M. W. Den Otter and B. A. Boukamp, J. Solid State Electrochem., 2004, 8, 599–605.
18. I. Yasuda and M. Hishinuma, J. Solid State Chem., 1996, 123, 382–390.
19. K. Sasaki and J. Maier, Solid State Ionics, 2003, 161, 145–154.
20. Y. L. Wang, Y. Wang and C. R. Xia, J. Electrochem. Soc., 2012, 159, F570–F576.
21. L. Zhang, Y. Q. Liu, Y. X. Zhang, G. L. Xiao, F. L. Chen and C. R. Xia, Electrochem. Commun., 2011, 13, 711–713.
22. Y. Wang, L. Zhang and C. R. Xia, Int. J. Hydrogen Energy, 2012, 37, 2182–2186.
23. T. Hong, L. Zhang, F. L. Chen and C. R. Xia, J. Power Sources, 2012, 218, 254–260.
24. M. Mogensen, N. M. Sammes and G. A. Tompsett, Solid State Ionics, 2000, 129, 63–94.
25. Q. Liu, C. H. Yang, X. H. Dong and F. L. Chen, Int. J. Hydrogen Energy, 2010, 35, 10039–10044.
26. Q. Liu, X. H. Dong, G. L. Xiao, F. Zhao and F. L. Chen, Adv. Mater., 2010, 22, 5478–5482.
27. M. W. den Otter, H. J. M. Bouwmeester, B. A. Boukamp and H. Verweij, J. Electrochem. Soc., 2001, 148, J1–J6.
28. P. S. Manning, J. D. Sirman and J. A. Kilner, Solid State Ionics, 1997, 93, 125–132.
29. D. Ding, B. B. Liu, Z. N. Zhu, S. Zhou and C. R. Xia, Solid State Ionics, 2008, 179, 896–899.
30. A. Bunde and W. Dieterich, J. Electroceram., 2000, 5, 81–92.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ta12787g

---