Proton-conducting oxide with redox protonation and its application to a hydrogen sensor with a self-standard electrode

Abstract

In order to simplify the structure of the EMF-type hydrogen sensor using a proton conductor as the electrolyte, an electrolyte serving as a self-standard electrode in air was developed. The electrochemical properties and proton dissolution mechanism of Mn-doped CaZrO₃ were evaluated by impedance analysis, EMF measurement, IR absorption analysis and ESR measurement. Mn-doped CaZrO₃ acquires a proton from hydrogen by reduction of the manganese ion. 5 mol% Mn doped CaZrO₃ showed proton conduction in a reducing atmosphere and hole conduction in an oxidizing atmosphere. A gas concentration cell using CaZr_0.95Mn_0.05O_3−δ as the electrolyte was constructed for use as a hydrogen sensor. It was found to be dependent only on the hydrogen potential of the working electrode when air was used as the reference gas.

---

Introduction

Some perovskite-type oxides show proton conduction at high temperatures,^1–6 and have been attracting attention as key materials for electrochemical devices, such as chemical sensors,^7–10 fuel cells,^11–14 and steam electrolysis cells.¹⁵ Protons are generally incorporated in these oxides by the following mechanism:¹⁶

Due to the electrical neutrality, oxide ion vacancies are formed when trivalent cations are partially substituted for the tetravalent cation in the matrix. The uptake of water takes place on the oxide ion vacancies.

There is another proton dissolution mechanism. The proton is incorporated into the oxide doped with a transition metal ion by reduction of the transition metal ion. This phenomenon was reported for Co-doped Al₂O₃ (ref. 17–19) and Mn-doped SrZrO₃.²⁰ In this study, the protonation and conduction of CaZr_1−xMn_xO_3−δ upon exposure to a hydrogen atmosphere were evaluated. The tetravalent or trivalent state is presumed for the manganese after sintering in air during synthesis. When it is exposed to a reducing atmosphere containing hydrogen, manganese will be reduced to the trivalent or divalent state. As electroneutrality must be maintained in the CaZrO₃, the reduction of the manganese must be accompanied by incorporation of protons (redox protonation), as described by the following equations:

Reduction of the manganese totally results in the incorporation of protons into the oxide. The introduced protons are thermodynamically stable when the hydration reaction is spontaneous. The purpose of this study is to investigate if the manganese-doped oxide can be protonated based on the mechanism described above, and is analyzed by the measurements of the Electron Spin Resonance (ESR) spectra, IR absorption spectra, electrical conductivity, and electromotive force (EMF) of a gas concentration cell utilizing CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005). Moreover, we report the development of the EMF-type hydrogen gas sensor using CaZr_0.95Mn_0.5O_3−δ with a redox protonation mechanism. The EMF of the gas concentration cell using a mixed proton, oxide ion, and hole conductor as the electrolyte is represented in the form of the following integral:²¹

where E is the EMF in volts, R is the gas constant, F is Faraday's constant, and t_i and σ_i are the transport number and the conductivity of the charge carrier, i, respectively. If the conductivity of the oxide ion can be neglected in the sensor operating temperature, the EMF can be represented as follows:

Fig. 1 shows the proton transport number of CaZr_0.9In_0.1O_3−δ (ref. 22) as a typical example. According to eqn (5), the EMF of the gas concentration cell using a conventional proton conductor, such as CaZr_0.9In_0.1O_3−δ, is proportional to the gray area in Fig. 1. Therefore, the EMF depends on both the hydrogen activity of the working electrode and the reference electrode. If a conventional proton conductor was used as the electrolyte of the hydrogen sensor, the hydrogen activity on the reference electrode must be fixed. On the other hand, the EMF is proportional to the hatched area in Fig. 1 and depend only on the hydrogen activity of the working electrode if the proton transport number of the reference electrode is a low value beyond the onset of the transport number. In such case, the EMF of the mixed proton and hole conductor is represented by Schmalzried-type equation.²³

where p_⊕ is the hydrogen activity at . If air is selected as the reference gas and we can use the approximation, p′′_H2^1/n ≪ p_⊕^1/n, the EMF is represented as,

Fig. 1 Proton transport number of the proton-conducting oxide. Solid line and dashed line show proton transport number of the mixed proton and hole conductor and CaZr_0.9In_0.1O_3−δ,²² respectively.

This equation shows that the EMF depends only on the hydrogen activity of the working electrode.

If CaZr_1−xMn_xO_3−δ undergoes proton conduction due to reduction of the magnesium in a reducing atmosphere and no proton conduction due to oxidization of magnesium in air, a hydrogen sensor that required no reference gas having a known hydrogen concentration can be developed. When the hole transport number of the electrolyte is unity on the reference electrode of the gas concentration cell, the electric potential can be determined without depending on the hydrogen concentration. On the other hand, the electrical potential of the working electrode depends on the hydrogen concentration because electrolytes on the working electrode show a proton conduction. Therefore, materials allowing proton conduction under a reducing atmosphere and hole conduction under an oxidizing atmosphere can be the electrolyte of the hydrogen sensor with a self-standard electrode. In this study, the hydrogen sensor using air as the reference gas was developed using CaZr_0.95Mn_0.05O_3−δ as the electrolyte.

Experimental

Sample preparation

Samples of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) were prepared by a solid-state reaction method. The reagent-grade CaCO₃ (99.9%), ZrO₂ (99.9%), and MnO₂ (99.9%) powders were weighed, mixed in a zirconia mortar with ethanol and calcinated at 1573 K for 10 h in air. The calcinated powders were ground in ethanol by a ball mill for 1 h. The powder was pressed into a pellet by a cold isostatic press at 250 MPa and sintered at 1773 K for 10 h in air. The relative density of all the samples was around 98%. The X-ray diffraction (XRD) analysis of the samples gave a well-defined perovskite pattern of CaZrO₃.

ESR measurements

The ESR spectra of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) annealed in 1.9% H₂O–1% H₂–Ar at 1073 K, and annealed in 1.9% H₂O–1% O₂–Ar at 1073 K were measured at room temperature using a JEOL JES-TE100 ESR spectrometer. The microwave frequency was 9.44 GHz. For each measurement, the mass of sample was fixed at 5.8 ± 0.4 mg.

IR diffusion reflection analysis

The IR absorption spectra of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) annealed in 1.9% H₂O–1% H₂–Ar at 1073 K, and annealed in 1.9% H₂O–1% O₂–Ar at 1073 K were measured at room temperature by a Fourier transform infrared spectrometer (FT/IR-4200, JASCO). KBr powder was used as the reference.

Electrical conductivity measurements

The electrical conductivity was measured for CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) equilibrated with 1.9% H₂O–y% H₂–Ar (y = 1–98) and 1.9% H₂O–z% O₂–Ar (z = 1–98) in the temperature range of 873–1173 K. The sample had a bar shape (2.9–4.2 mm × 1.4–1.8 mm × 11.5–12.0 mm). Porous platinum electrodes were prepared on both surfaces of the bar sample by painting platinum paste (Tanaka Kikinzoku Kogyo, TR-7907) and baked at 1273 K in air for 1 h. Platinum wire was attached to the electrode as the lead. A 4-terminal ac technique was employed. The complex impedance of the samples was measured in the frequency range of 0.02 Hz to 1 MHz using an LCR meter (NF corporation: ZM2375). Two semicircles corresponding to the bulk impedance and the grain boundary impedance were observed. The resistance at the point of minimum reactance in the low frequency region was adopted as the bulk resistance of the sample.

Electromotive force measurements

The EMF of the gas concentration cells using CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) as the electrolyte was measured in the temperature range of 873–1073 K. The samples were pellets (φ 12.5 mm diameter and 0.5 mm thickness). Porous platinum electrodes were prepared on both surfaces of the pellet samples. The sample was held between aluminum tubes with pyrex glass gaskets separating the two electrode compartments. The gas concentration cell is represented by the following cell formula:

(−), Pt(p_H2(I)) or p_O2(I), p_H2O(I)|CaZr_1−xMn_xO₃|Pt(p_H2(II) or p_O2(II), p_H2O(II)), (+) (8)

The p_i in this paper represents the activity of the gas species, i, with reference to one bar in the pure state. The gases were mixtures of H₂ or O₂ and Ar containing predetermined amounts of water vapor. The amount of water vapor was controlled using a water bubbler in a thermostatic bath held at a specific temperature. The EMF measurements were performed using a digital voltmeter (Keysight Technology 34461A).

Structure of hydrogen sensor

The performance of the hydrogen sensor using CaZr_0.95Mn_0.05O_3−δ as the electrolyte was checked by measuring the hydrogen activity in a gas. The structure of the hydrogen sensor is represented as follows:

(−), Pt(p_H2(work), p_H2O(work))|CaZr_0.95Mn_0.05O₃|Pt(p_O2(ref) = 0.21, p_H2O(ref)), (+) (9)

We labelled the two electrodes as the reference and working electrodes. Wet air (p_H2O = 0.010–0.019) was flowed to the reference electrode. The various ratios of the H₂/Ar gases were flowed to the working electrode. The EMFs were measured as a function of the hydrogen activity of the working electrode and the water vapor activity of the reference electrode.

Results

Redox protonation

In order to clarify the valence number of manganese in CaZrO₃, an ESR analysis was performed at room temperature. As shown in Fig. 2, six spectra were observed for CaZr_0.995Mn_0.005O_3−δ annealed in 1.9% H₂O–1% H₂–Ar at 1073 K, and annealed in 1.9% H₂O–1% O₂–Ar, which can be corrected to the presence of Mn²⁺.²⁴ The g value is 2.0055. The intensity of the spectrum of the sample annealed in hydrogen was greater than that of the sample annealed in oxygen. This indicated that the reduction of magnesium ion from Mn³⁺ to Mn²⁺ is due to the hydrogen treatment. CaZr_0.95Mn_0.05O_3−δ annealed in 1.9% H₂O–1% H₂–Ar show a broad signal with the g value of 2.0047. There is no difference in the g value between x = 0.05 and x = 0.005. This indicated that a broad signal of CaZr_0.95Mn_0.05O_3−δ shows the presence of Mn²⁺. A broad signal might be observed by interaction between the Mn²⁺ because the content of manganese ion is high.

Fig. 2 ESR spectrum of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005). The ESR measurements were performed at room temperature for annealed sample. (a) The powder of CaZr_0.995Mn_0.5O₃ was annealed in 1.9% H₂O–1% H₂–Ar at 1073 K. (b) The powder of CaZr_0.995Mn_0.5O₃ was annealed in 1.9% H₂O–1% O₂–Ar at 1073 K. (c) The powder of CaZr_0.95Mn_0.05O₃ was annealed in 1.9% H₂O–1% H₂–Ar at 1073 K. (d) The powder of CaZr_0.95Mn_0.05O₃ was annealed in 1.9% H₂O–1% O₂–Ar at 1073 K.

Fig. 3 shows the result of the IR diffusion reflection analysis of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005). The IR spectra were observed at 3300 cm⁻¹ which can be attributed to the O–H vibration. The absorptivity of the sample annealed in 1.9% H₂O–1% H₂–Ar was higher than that of the sample annealed in 1.9% H₂O–1% O₂–Ar. This shows the same tendency as the ESR spectra intensity. These results suggested that Mn-doped CaZrO₃ acquires a proton from hydrogen by reduction of the manganese ion as described by eqn (3).

Fig. 3 IR diffusion refraction spectrum of CaZr_1−xMn_xO₃ (x = 0.05, 0.005). The IR diffusion refraction analyses were performed at room temperature for annealed sample. (a) The powder of CaZr_0.995Mn_0.5O_3−δ was annealed in 1.9% H₂O–1% H₂–Ar at 1073 K. (b) The powder of CaZr_0.995Mn_0.5O_3−δ was annealed in 1.9% H₂O–1% O₂–Ar at 1073 K. (c) The powder of CaZr_0.95Mn_0.05O_3−δ was annealed in 1.9% H₂O–1% H₂–Ar at 1073 K. (d) The powder of CaZr_0.95Mn_0.05O_3−δ was annealed in 1.9% H₂O–1% O₂–Ar at 1073 K.

Electrochemical properties

Fig. 4(a) shows the temperature dependence of the bulk conductivity. The conductivity of CaZr_0.95Mn_0.05O_3−δ in 1.9% H₂O–1% H₂–Ar was higher than that of the conventional hydrogen sensor material, CaZr_0.9In_0.1O_3−δ. The activation energy was estimated from the data at 873–1073 K. The values are 0.66 eV and 0.63 eV for CaZr_0.95Mn_0.05O_3−δ and CaZr_0.995Mn_0.005O_3−δ, respectively which is in agreement with the activation energy of the conductivity for CaZr_0.9In_0.1O_3−δ.²² The conductivity of CaZr_1−xMn_xO₃ (x = 0.05, x = 0.005) in 1.9% H₂O–1% O₂–Ar was lower than that in 1.9% H₂O–1% H₂–Ar. The activation energy of the conductivity in 1.9% H₂O–1% O₂–Ar was 1.10 eV and 1.23 eV for CaZr_0.95Mn_0.05O_3−δ and CaZr_0.995Mn_0.005O_3−δ, respectively, which is higher than that in 1.9% H₂O–1% H₂–Ar. This indicated that the predominant charge carrier in 1.9% H₂O–1% H₂–Ar and 1.9% H₂O–1% O₂–Ar differs for CaZr_1−xMn_xO_3−δ.

Fig. 4 Electrical conductivity of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005). (a) Temperature dependence of the electrical conductivity. (b) The hydrogen activity dependence of the electrical conductivity. The solid line, the dotted line and the dashed line show the total conductivity, the proton conductivity and the hole conductivity, respectively.

Fig. 4(b) shows the hydrogen activity dependence of the bulk conductivity at 1073 K. As shown in Fig. 4(b), the conductivity was dependent on the hydrogen activity in the reducing atmosphere (H₂O/H₂). The hydrogen activity dependence of the conductivity is reduced in the oxidizing atmosphere (H₂O/O₂). These dependencies are discussed based on the defect equilibrium in the next section.

In order to clarify the charge carriers of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005), the EMF of a gas concentration cell was measured in the temperature range of 873–1073 K. According to eqn (4), the EMF of a gas concentration cell can be arranged using the water equilibrium (H₂O = H₂ + ½O₂) as follows

or

When the ionic transport number is unity, the EMFs of the hydrogen concentration cell at a constant water vapor pressure obey the following equation:

When the transport number of the oxide ion is ignored, the EMFs of the water vapor concentration cell at a constant hydrogen activity are not observed according to eqn (10). Fig. 5(a) shows the EMF of the hydrogen concentration cell at a constant water vapor activity. As shown Fig. 5(a), the observed EMF was in good agreement with the calculated value by eqn 12. Moreover, the EMFs of the water vapor concentration cell at a constant hydrogen activity were not observed as shown in Fig. 5(b). Therefore, the oxide ion conductivity of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) could be disregarded as compared to the proton conductivity. The proton transport number of CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) is unity under the hydrogen atmosphere. On the other hand, the EMFs of the oxygen concentration cell using CaZr_0.95Mn_0.05O_3−δ as the electrolyte cannot be observed in the oxidizing atmosphere (H₂O/O₂) as shown in Fig. 5(c). Therefore, the dominant charge carrier was found to be the hole in the oxidizing atmosphere. The EMFs of the oxygen concentration cell using CaZr_0.995Mn_0.005O_3−δ as the electrolyte were slightly observed. This indicated that CaZr_0.995Mn_0.005O_3−δ shows a mixed proton and hole conduction in the oxidizing atmosphere.

Fig. 5 EMFs of the gas concentration cell using CaZr_1−xMn_xO_3−δ (x = 0.05, 0.005) as the electrolyte. (a) Hydrogen concentration cell. The dotted lines show the calculated value by Nernst equation when the ion transport number is unity (b) water vapor concentration cell. The dotted lines show the calculated value by Nernst equation when the oxygen ion transport number is unity (c) oxygen concentration cell. The dotted lines show the calculated value by Nernst equation when the ion transport number is unity.

Discussion

Defect structure of Mn-doped CaZrO₃

As shown in Fig. 4(b), the conductivity increased with an increase in the hydrogen activity. This might be attributed to the proton concentration that increased with a reduction in the manganese ion. We considered the defect structure of CaZr_1−xMn_xO_3−δ based on the equilibrium between the point defect and the ambient atmosphere to explain the hydrogen partial pressure dependence of the conductivity. If the relative charge carriers in the crystals are limited to the substitutional defect of magnesium ions on the zirconium ion sites, , interstitial proton , the oxide ion vacancy, , and the hole, h˙, the electroneutrality is maintained as:

When CaZr_1−xMn_xO_3−δ was exposed to hydrogen gas, the magnesium ions were reduced from and a proton was incorporated into the crystal as described by eqn (3). Although functions as the acceptor dopant, protons cannot be incorporated into the crystal under the oxidization atmosphere. On the other hand, lead to the protonation under the reducing atmosphere. The ionic radius is in order of Mn³⁺ (0.0645 nm) < Zr⁴⁺ (0.0720 nm) < Mn²⁺ (0.0830 nm).²⁵ A proton might be incorporated due to expansion of the octahedral oxide around Mn²⁺.

It is assumed that are compensated with protons and holes, respectively, and the oxide ion vacancy can be neglected. The electroneutrality can be represented as follows:

Since Henry's law holds for low defect concentrations, the equilibrium constant, K_red, for eqn (3) can be expressed as:

From eqn (14)–(17) based on the Nernst–Einstein relation, the proton and conductivities can be represented as follows:

where μ_i, F and V_oxide are the mobility of the charge carrier, i, Faraday's constant and the molar volume of the oxide, respectively. By setting , μ_h˙ and K_red as the fitting parameter, the proton, and hole, the total conductivities were estimated from eqn (18) and (19). Fig. 4(b) shows the estimated conductivity. The solid line, the dotted line and the dashed line show the total conductivity, the proton conductivity and the hole conductivity, respectively. The estimated total conductivity was in good agreement with the measured value for the set of parameters shown in Table 1.

Table 1 Parameters for the estimation of partial conductivity at 1073 K

x	Kred	[cm^2 V^−1 s^−1]	μh˙ [cm^2 V^−1 s^−1]
0.05	1.8	1.3 × 10^−6	3.0 × 10^−8
0.005	1.8	4.5 × 10^−7	2.0 × 10^−9

---

The transport number of the protons estimated from the partial conductivity is shown in Fig. 6. As shown in Fig. 6, the proton transport number decreased with an increase in the manganese content in the oxidizing atmosphere. As shown in Fig. 5(c), the EMF of the oxygen concentration cell using CaZr_0.95Mn_0.05O_3−δ as the electrolyte was not observed. The EMF of CaZr_0.95Mn_0.05O_3−δ might not change with the change in the hydrogen potential in air. Therefore, CaZr_0.95Mn_0.05O_3−δ is a potential material for a hydrogen sensor that use air as the reference gas.

Fig. 6 Proton transport number of Mn doped CaZrO₃. Solid line and dotted line show proton transport number of CaZr_0.95Mn_0.05O_3−δ and CaZr_0.995Mn_0.005O_3−δ, respectively.

Hydrogen sensor with self-standard electrode

As shown in Fig. 6, the protonic transport number of CaZr_0.95Mn_0.05O_3−δ was unity under a hydrogen atmosphere and fell below 0.2 in air. The EMF change from the hydrogen potential change in air can be neglected because the protonic transport number is a sufficiently low value in air. The hydrogen sensor with a self-standard electrode using CaZr_0.95Mn_0.05O_3−δ as the electrolyte and air as the reference gas (cell (9)) was developed. The changes in the EMF of the cell (9) are shown in Fig. 7(a). As shown in the figure, the EMF clearly changed with the hydrogen potential change. The time constant of the EMF change was 90 s when the hydrogen potential was changed from 0.1 to 0.98. This value was in good agreement with the time constant of EMF of the gas concentration cell using proton conducting alumina.²⁶ The time constant of the EMF change is due to the chemical potential change on the surface of the working electrode because it is based on the EMF generated across the electrolyte. Therefore, the same time constant of the EMF change might show for CaZr_0.95Mn_0.05O_3−δ with the redox protonation mechanism.

Fig. 7 Performance of hydrogen sensor. (a) EMF responses to the hydrogen composition. (b) Relation between EMF and hydrogen activity of the working electrode. (c) Relation between EMF and water vapor activity of the reference electrode.

Fig. 7(b) show the EMFs as a function of the hydrogen activity on the working electrode. The dashed line shows the calculated EMF based on the protonic transport number by eqn (5). The calculated value was in good agreement with the measured value. Moreover, Fig. 7(c) show the EMFs as a function of the water vapor activity on the reference electrode. The EMF was found to be independent of the water vapor pressure in air. The EMF of the hydrogen sensor using CaZr_0.95Mn_0.05O_3−δ as the electrolyte was found to depend only on the hydrogen potential of the working electrode when air was used as the reference gas. The results obtained in this study suggested that CaZr_0.95Mn_0.05O_3−δ is a potential material as an electrolyte for an EMF-type hydrogen sensor with a self-standard electrode.

Conclusion

In this study, the electrochemical properties of the Mn-doped CaZrO₃ were examined for development of a hydrogen sensor. An ESR spectra were observed for the sample annealed in hydrogen and annealed in oxygen, which can be corrected to the presence of Mn²⁺. The intensity of the spectra of the sample annealed in hydrogen was greater than that of the sample annealed in oxygen. Moreover, the IR spectra observed at 3300 cm⁻¹ can be attributed to the O–H vibration. The absorptivity shows the same tendency of the ESR spectra intensity. These results suggested that the Mn-doped CaZrO₃ acquires a proton from hydrogen by the reduction of Mn.

Based on the electromotive force of the gas concentration cell, it was found that the Mn-doped CaZrO₃ shows proton conduction in a reducing atmosphere and hole conduction in an oxidizing atmosphere. The EMF of a gas concentration cell using CaZr_0.95Mn_0.05O_3−δ as the electrolyte was found to depend only on the hydrogen potential of the working electrode when air was used as the reference gas. CaZr_0.95Mn_0.05O_3−δ was found to be a potential material for a hydrogen sensor that used air as a reference gas. The present sensor is an advanced useful tool to monitor the hydrogen atmosphere in industrial process at high temperature.

Acknowledgements

This work was supported by JSPS KAKENHI Grant Number 25709072 (Grant-in-Aid for Young Scientists (A)).

References

1. T. Takahashi and H. Iwahara, Solid state ionics: protonic conduction in perovskite type oxide solid-solutions, Rev. Chim. Miner., 1980, 17, 243–253.
2. H. Iwahara, T. Esaka, H. Uchida and N. Maeda, Proton conduction in sintered oxides and its application to steam electrolysis for hydrogen production, Solid State Ionics, 1981, 3/4, 359–363.
3. H. Iwahara, H. Uchida, K. Ono and K. Ogaki, Proton conduction in sintered oxides based on BaCeO₃, J. Electrochem. Soc., 1988, 135, 529–533.
4. S. Shin, H. H. Huang, M. Ishigame and H. Iwahara, Protonic conduction in the single crystals of SrZrO₃ and SrCeO₃ doped with Y₂O₃, Solid State Ionics, 1990, 40/41, 910–913.
5. T. Yajima, H. Kazeok, T. Yogo and H. Iwahara, Proton conduction in sinter oxides based on CaZrO₃, Solid State Ionics, 1991, 47, 271–275.
6. H. Iwahara, T. Yajima, T. Hibino, K. Ozaki and H. Suzuki, Protonic conduction in calcium, strontium and barium zirconates, Solid State Ionics, 1993, 61, 65–69.
7. T. Yajima, H. Iwahara, N. Fukatsu, T. Ohashi and K. Koide, Measurement of hydrogen content in molten aluminum using proton conducting ceramic sensor, Keikinzoku, 1992, 42, 263–267.
8. T. Yajima, K. Koide, H. Takai, N. Fukatsu and H. Iwahara, Application of hydrogen sensor using proton conductive ceramics as a solid electrolyte to aluminum casting industries, Solid State Ionics, 1995, 79, 333–337.
9. N. Kurita, N. Fukatsu, S. Miyamoto, F. Sato, H. Nakai, K. Irie and T. Ohashi, The measurement of hydrogen activities in molten copper using an oxide protonic conductor, Metall. Mater. Trans. B, 1996, 27, 929–935.
10. N. Fukatsu, N. Kurita, K. Koide and T. Ohashi, Hydrogen sensor for molten metals usable up to 1500 K, Solid State Ionics, 1998, 113–115, 219–227.
11. N. Ito, M. Iijima, K. Kimura and S. Iguchi, New intermediate temperature fuel cell with ultra-thin proton conductor electrolyte, J. Power Sources, 2005, 152, 200–203.
12. N. Ito, S. Aoyama, T. Masui, S. Matsumoto, H. Matsumoto and T. Ishihara, Electrochemical analysis of hydrogen membrane fuel cells, J. Power Sources, 2008, 185, 922–926.
13. C. Duan, J. Tong, M. Shang, S. Nikodemski, M. Sanders, S. Ricote, A. Almansoori and R. O'Hyre, Readily processed protonic ceramic fuel cells with high performance at low temperatures, Science, 2015, 349, 1321–1326.
14. Y. Okuyama, K. Okuyama, Y. Mizutani, T. Sakai, Y. S. Lee and H. Matsumoto, Proton transport properties of La_0.9Sr_0.1Yb_0.8In_0.2O_3−δ and its application to proton ceramic fuel cell, Int. J. Hydrogen Energy, 2014, 39, 20829–20836.
15. T. Sakai, S. Matsushita, H. Matsumoto, S. Okada, S. Hashimoto and T. Ishihara, Intermediate temperature steam electrolysis using strontium zirconate-based protonic conductors, Int. J. Hydrogen Energy, 2009, 34, 56–63.
16. C. Wagner, Die Löslichkeit von Wasserdampf in ZrO₂–Y₂O₃ Mischkristallen, Ber. Bunsenges. Phys. Chem., 1968, 72, 778–781.
17. R. Müller and Hs. H. Günthard, Spectroscopic study of reduction of nickel and cobalt ion in sapphire, J. Chem. Phys., 1966, 44, 365–373.
18. K. Eigenmann and Hs. H. Günthard, Hydrogen incorporation in doped α-Al₂O₃ by high temperature redox reactions, Chem. Phys. Lett., 1971, 12, 12–15.
19. P. Balmer, H. Blum, M. Forster, A. Schweiger and Hs. H. Günthard, Co^z+ α-Al₂O₃: cobalt and proton ENDOR, infrared and optical spectra, the nature of charge compensator and the structure of the defect (Co²⁺, H⁺): α-Al₂O₃, J. Phys. C: Solid State Phys., 1980, 13, 517–534.
20. H. Matsumoto, T. Sakai, Y. Kawasaki, Y. Sato and T. Ishihara. Abstract of SSPC-15, 2010, p. 24.
21. C. Wagner, Advance in Electrochemistry and Electrochemical Engineering, 4 Interscience, NY, 1966, pp. 1–64.
22. N. Kurita, N. Fukatsu, K. Ito and T. Ohashi, Protonic conduction domain of indium-doped calcium zirconate, J. Electrochem. Soc., 1995, 142, 1552–1559.
23. H. Schmalzried, Ionen-und Electronenleitung in binären Oxiden und ihare Untersuchung mittels EMF-Messungen, Z. Physik. Chem., 1963, 38, 87–102.
24. G. Bacquet, J. Dugas and C. Escribe, The system ZrO₂–CaO studied by the electron spin resonance of Mn²⁺ ion, J. Solid State Chem., 1976, 19, 251–261.
25. R. D. Shannon, Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751–767.
26. Y. Okuyama, N. Kurita, A. Yamada, H. Takami, T. Oshima, K. Katahira and N. Fukatsu, A new type of hydrogen sensor for molten metals usable up to 1600 K, Electrochim. Acta, 2009, 55, 470–474.

---