Protonic Transport in Layered Perovskites BaLa_nIn_nO_3n+1 (n = 1, 2) with Ruddlesden-Popper Structure

Abstract

The work focused on the layered perovskite-related materials as the potential electrolytic components of such devices as proton conducting solid oxide fuel cells for the area of clean energy. The two-layered perovskite BaLa₂In₂O₇ with the Ruddlesden–Popper structure was investigated as a protonic conductor for the first time. The role of increasing the amount of perovskite blocks in the layered structure on the ionic transport was investigated. It was shown that layered perovskites BaLa_nIn_nO_3n+1 (n = 1, 2) demonstrate nearly pure protonic conductivity below 350 °C.

1. Introduction

The rising of the environmental challenges and the increasing in the need for safe and effective devices for energy conversion and storage mark the new decade [1,2,3,4]. This makes the task of creating high-efficiency solid oxide devices for electrochemical applications very important for now. The oxygen- and proton-conducting solid oxides may be components of various devices including fuel cells and electrolyzers [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Most well-known electrolytes for medium-temperature solid oxide fuel cells (500–700 °C) are characterized by the perovskite or perovskite-related structure ABO_3−δ [22]. As an example, one of the most investigated protonic conductors, barium zirconate [23], has the structure of ordinary perovskite with a cubic unit cell (Figure 1a). The varying of the cations A and B can create the materials with a layered perovskite-related structure where the perovskite layers alternate with rock-salt layers.

The general formula of layered perovskite-related structures can be written as AO(A′BO₃)_n, where n is the number of polyhedra layers stacking in the perovskite block. When n = 1, this formula can be represented as AA′BO₄, and the compositions with this formula are characterized by the monolayered perovskite-related structure. The family of proton-conducting oxides with this type of the structure was discovered several years ago, and it is represented by the matrix compositions SrLaInO₄ [24,25,26,27,28], BaLaInO₄ [29,30,31,32,33,34,35,36,37], BaNdInO₄ [38,39,40,41,42,43], BaNdScO₄ [44]. The symmetry of these compositions is lower than cubic, and it is orthorhombic or monoclinic depending on the combination of cations in the A and B sublattices. As an example, the orthorhombic structure of the compound BaLaInO₄ is represented in Figure 1b. It was shown that various types of doping (donor and acceptor) lead to the increase in the values of ionic conductivity [45]. Moreover, the compositions based on BaLaInO₄ demonstrate almost pure protonic conductivity under wet air conditions at 450 °C [45].

The next member of the homologous series of the Ruddlesden–Popper type of structure with n = 2 can be writing in the general case as AA₂′B₂O₇ and as an example BaLa₂In₂O₇. Despite the possibility of protonic transport in the two-layered complex oxide, BaLa₂In₂O₇ is not yet described; the structure of BaLa_nIn_nO_3n+1 (n = 1, 2) is described in detail by Titov et al. [46] and Raveau et al. [47]. One of the main differences between BaLaInO₄ and BaLa₂In₂O₇ compositions is the ratio of perovskite blocks and rock-salt layers in the unit cell. This ratio is 1:1 for BaLaInO₄ and 2:1 for BaLa₂In₂O₇. It can be assumed that an increase in the ratio of perovskite blocks can lead to the increase in the ionic conductivity of solid oxide. Based on this, the goal of this work is the investigation of ionic transport (O²⁻, H⁺) in the two-layered compound BaLa₂In₂O₇ in comparison with the mono-layered compound BaLaInO₄. Proton transport is not previously reported for n = 2 RP phases in the Ba-La-In-O system.

2. Experimental

The phase BaLa₂In₂O₇ was synthesized using the solid state method. The previously dried powder reagents BaCO₃, La₂O₃, In₂O₃ were used as a starting materials. The milling of the reagents was made in an agate mortar. The calcination was performed in the temperature range from 800 to 1300 °C with the step of 100 °C and 24 h time treatments.

The XRD studies were performed using a Bruker Advance D8 (Rheinstetten, Germany) diffractometer with Cu K_α radiation over a range from 20° to 100° with a step of 0.01° and at a scanning rate of 0.5°/min. The preparation of the samples was made by their heating at 1100 °C for 5 h and following cooling in dry Ar (pH₂O = 3.5 × 10⁻⁵ atm).

For the investigations of the electrical properties, the pressed cylindrical pellets (1300 °C, 24 h, dry air) were obtained. The pellets had a relative density ~92% (density of the sintered samples was determined by the Archimedes’ method). The ac conductivity measurements were performed by the Z-1000P (Elins, Russian Fedeartion) impedance spectrometer within the frequency range of 1–10⁶ Hz. Electrical measurements were performed using Pt paste electrodes (annealing at 1000 °C for 2 h). The temperature dependencies of conductivities were obtained in the temperature range 200–1000 °C (step 10–20 °C, cooling rate of 1°/min; the measurements were taken until a constant resistance value was established). These investigations were performed under “dry” and “wet” air or Ar. The dry gas was produced by circulating the gas through P₂O₅ (pH₂O = 3.5 × 10⁻⁵ atm). The wet gas was obtained by bubbling the gas at room temperature first through distilled water and then through a saturated solution of KBr (pH₂O = 2 × 10⁻² atm). The humidity of the gas was controlled by a Honeywell HIH-3610 H₂O-sensor. The dependencies of conductivities vs. partial oxygen pressures pO₂ were obtained by using the electrochemical method for producing different pO₂ with an oxygen pump (and sensor) from Y-stabilized ZrO₂ ceramic. Each resistance value was recorded after 3–5 h of equilibrium.

3. Results and Discussions

The phase characterization of the obtained sample BaLa₂In₂O₇ was made using XRD analysis. It was shown that all observed diffraction peaks can be related to the single phase belonging to the tetragonal symmetry (space group P4₂/mnm) (Figure 2). The lattice parameters (a = 5.914(9) Å, c = 20.846(5) Å) were in good agreement with those previously reported by Titov et al. and Raveau et al. (a = 5.915(2) Å, c = 20.86(1) Å [46]; a = 5.9141(3) Å, c = 20.8231(2) [47]). The structural parameters are presented in the

Table 1. Refined structural parameters from XRD-data.

Atom	Site	x	y	z
Ba	4f	0.2483(1)	0.2483(1)	0
La	8j	0.2687(4)	0.2687(4)	0.1851(3)
In	8j	0.2590(0)	0.2590(0)	0.4002(2)
O(1)	4g	0.806(1)	0.193(3)	0
O(2)	8j	0.183(4)	0.183(4)	0.291(1)
O(3)	8h	0	0.5	0.095(2)
O(4)	4e	0	0	0.126(1)
O(5)	4e	0	0	0.382(1)

R_p = 4.31, R_wp = 3.48, χ² = 1.73.

.

The 3D visualization of the crystal structure of the obtained phase is presented in the Figure 1c using VESTA software [48]. As can be seen (Figure 1), for the mono-layered structure (Figure 1b), the ordering in the A and A′ cations does not happen, and Ba and La atoms occupy the same positions in the crystal lattice of BaLaInO₄, and they have the same coordination number (c.n. = 9). Opposite, the ordered arrangement of A and A′ cations is observed for the two-layered complex oxide BaLa₂In₂O₇, as it was shown by Raveau et al. [47] with using high resolution electron microscopy (Figure 1c). It was confirmed that barium atoms locate in the perovskite blocks, and the lanthanum atoms are placed in the rock-salt layers [47]. In the case of the two-layered structure, the cation with the largest ionic radii (barium in this case) has the c.n. = 12 and forms polyhedra [BaO₁₂]; the cation with lower ionic radii (lanthanum) forms the polyhedra (LaO₉) with c.n. = 9. However, it was shown that both perovskite blocks and rock-salt layers are distorted, and the real coordination numbers are ten and seven for barium and lanthanum, respectively [47].

The electrical conductivity was measured by the impedance spectroscopy method. The Nyquist-plots for the investigated composition BaLa₂In₂O₇ obtained under dry air are presented in Figure 3a–c. The fitting of the spectra was made using ZView software, and the obtained results are presented in

Table 2. Results of Nyquist-plots fitting, where CPE is the constant phase element (F), and R is the resistance (kΩ∙cm).

Element	Value (700 °C)	Value (600 °C)	Value (500 °C)
CPE1	2.1 × 10−12	2.4 × 10−12	3.2 × 10−12
R1	13	36	158
CPE2	4.1 × 10−10	3.7 × 10−10	2.1 × 10−10
R2	2	8	33
CPE3	6.0 × 10−7	5.6 × 10−7	7.0 × 10−7
R3	3	7	39

. According to the fitting of the spectra (red line) with using the equivalent circuit presented in Figure 3d, three different electrochemical processes can be defined. For the calculation of electrical conductivity, the bulk resistance values R1 were used and are discussed below. It can be noted that, due to the small depression of the semicircles, the constant phase element (CPE) was used during the analysis of Nyquist plots.

The temperature dependencies of bulk conductivities obtained under dry air and dry Ar are presented in Figure 4. As can be seen, the conductivity values for the compound BaLa₂In₂O₇ are higher than for BaLaInO₄ [29] by around 0.6–0.7 order of magnitude. This confirms the previously stated assumption that the increase in the ratio of perovskite blocks in the layered structures can lead to the increase in the electrical conductivity values. However, for the correct comparison, the total conductivity values must be divided into ionic and electronic contributions.

To establish the nature of electrical transport in BaLa₂In₂O₇, the conductivity measurements vs. pO₂ were performed (Figure 5). As can be seen (Figure 5a), in the electrolytic area (pO₂ = 10⁻¹⁶–10⁻⁵ atm), the independence of conductivity values from pO₂ was observed. Thus, the oxygen-ionic conductivity dominates in this region. For layered structures, oxygen-ionic transport is usually described in terms of anti-Frenkel disordering:

$${\mathrm{O}}_{\mathrm{O}}^{\mathrm{x}}\iff {\mathrm{V}}_{\mathrm{o}}^{••}+{\mathrm{O}}_{\mathrm{i}}^{″}$$

(1)

where ${\mathrm{O}}_{\mathrm{i}}^{″}$ is the oxygen atom in the interstitial position; ${\mathrm{V}}_{\mathrm{O}}^{••}$ is the oxygen vacancy.

The possibility of the formation of interstitial oxygen is provided by the existence of a rock-salt block and the possibility of increasing the coordination number of the cation that is located there. Accordingly, the formation of oxygen vacancies occurs in the perovskite block; the cation, octahedrally surrounded by oxygen, lowers the coordination number. That is, the presence of oxygen-deficient perovskite blocks plays an important role in the formation of oxygen-ion transport. Therefore, for such structures, it is possible to expect a significant oxygen-ionic conductivity without doping. Moreover, the measurements of the conductivity versus pO2 showed that the phases of BaLa2In2O7 are stable over a wide range of pO2.

The conductivity dependencies obtained in the dry oxidizing conditions (pO₂ = 10⁻⁵–0.21 atm) have a positive slope, which indicates the mixed ionic-electronic (hole) nature of conductivity:

$$\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.{\mathrm{O}}_2\iff {\mathrm{O}″}_{\mathrm{i}}+2h^{•}$$

(2)

$${\mathrm{V}}_{\mathrm{o}}^{••}+\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.{\mathrm{O}}_2\iff {\mathrm{O}}_{\mathrm{O}}^{\mathrm{x}}+h^{•}$$

(3)

where ${\mathrm{O}}_{\mathrm{i}}^{″}$ is the oxygen atom in the interstitial position; ${\mathrm{V}}_{\mathrm{O}}^{••}$ is the oxygen vacancy; $h^{•}$ is the hole. Obviously, both of these processes can be realized in the structure. However, it is impossible to say which process prevails. For the layered composition based on BaNdInO₄, the calculations of formation energies of point defects were made [41]. The energy of processes described by Equation (3) is about 1.2 eV. However, the data for Equation (2) are missing.

It should be noted that the conductivity values obtained in the dry Ar (blue symbols in the Figure 5) are the same as the conductivity values from the electrolytic area (black symbols in the Figure 5a). Therefore, in an argon atmosphere, it is relatively easy to measure the temperature dependence of oxygen-ionic conductivity in a wide temperature range.

It was shown earlier, for the compositions based on BaLaInO₄, the conductivity values obtained under dry Ar corresponded to the oxygen-ionic conductivity values [45]. Based on this, the oxygen-ionic transport number $t_{O^{2-}}$ can be calculated as:

$$t_{O^{2-}}=\frac{{\sigma }_{O^{2-}}}{{\sigma }_{\mathrm{tot}}}=\frac{{\sigma }_{\mathrm{Ar}}}{{\sigma }_{\mathrm{air}}}$$

(4)

The oxygen-ionic transport numbers for both BaLa₂In₂O₇ and BaLaInO₄ compositions did not change under temperature variation at pO₂ = 0.21 atm and were around 20% for each composition, and both compositions demonstrate mixed ionic-electronic (hole) conductivity under dry air, as it was shown earlier for the composition BaLaInO₄ [45].

Returning to Figure 4a, it can be seen that the values of oxygen-ionic conductivity for BaLa₂In₂O₇ were higher than for BaLaInO₄. It should the noted that the increasing in the oxygen-ionic conductivity values correlates well with the decreasing in the values of energy activation of oxygen transport (from 0.87 eV for BaLaInO₄ to 0.81 eV for BaLaInO₄). Because the increase in the oxygen-ionic conductivity correlates with the increase in the number of perovskite layers in the structure, we can suppose that the transport of oxygen ions includes the oxygen positions in the In-containing polyhedra.

The comparison of the dependencies of conductivity vs. oxygen partial pressure under dry and wet conditions is presented in Figure 5b. At the pO₂ range from 10⁻⁴ to 10⁻¹⁶ atm, the conductivity increases significantly in a wet atmosphere as a result of the appearance of a proton contribution due to the dissolution of H₂O in BaLa₂In₂O₇.

Water molecules can be dissolved in the oxygen-deficient perovskite layers leading to the formation of proton carriers as follows:

$${\mathrm{V}}_{\mathrm{o}}^{••}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}}_{\mathrm{o}}^{\times }\iff 2{\left(\mathrm{OH}\right)}_{\mathrm{o}}^{•}$$

(5)

where ${\mathrm{V}}_{\mathrm{o}}^{••}$ is the oxygen vacancy; ${\mathrm{O}}_{\mathrm{o}}^{\times }$ is the oxygen atom in the regular position$; {\left(\mathrm{OH}\right)}_{\mathrm{o}}^{•}$ is the hydroxyl group in the oxygen sublattice.

At the same time, for layered structures containing rock-salt blocks, the second mechanism of the dissociative dissolution of water molecules is possible:

$${\mathrm{H}}_2\mathrm{O}+{\mathrm{O}}_{\mathrm{O}}^{\mathrm{x}}\iff {(\mathrm{OH})}_{\mathrm{O}}^{•}+{(\mathrm{OH})}_{\mathrm{i}}^{\prime }$$

(6)

where ${(\mathrm{OH})}_{\mathrm{O}}^{•}$ is the hydroxyl group in the regular oxygen position; ${(\mathrm{OH} )}_{\mathrm{i}}^{\prime }$ is the hydroxyl group located in the rock-salt block. This process is accompanied by the increase in the coordination number of atoms in the rock-salt block.

The increase in the conductivity in a wet atmosphere compared to a dry atmosphere is reduced with increasing temperature. In a wet atmosphere below 450 °C, the conductivity does not depend on pO₂, which reflects the dominant ionic (proton) nature of the conductivity.

Differences in the conductivities in dry and wet atmospheres at high pO₂ > 10⁻⁴ atm are not so significant compared to the plateau region (Figure 5b). This is due to the presence of holes, the concentration of which decreases in a wet atmosphere:

$$h^{•}+\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.{\mathrm{H}}_2\mathrm{O} +{\mathrm{O}}_i^{″}\leftrightarrow \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$4$}\right.{\mathrm{O}}_2+{(\mathrm{OH})}_i^{\prime }$$

(7)

where $h^{•}$ is the hole; ${\mathrm{O}}_{\mathrm{i}}^{″}$ is the oxygen atom in the interstitial position; ${(\mathrm{OH})}_{\mathrm{i}}^{\prime }$ is the oxygen-hydrogen group.

In accordance with the explanations presented above and the obtained data on conductivity-pO₂, the measurements of conductivity in a wide temperature range can be carried out by comparing the atmosphere of air and argon at different levels of humidity. This comparison is shown in Figure 4.

As can be seen (Figure 4b), the values of ionic conductivity (wet Ar) are near the same as the total conductivity values (wet air) below ~450 °C. In other words, the nature of conductivity changes from mixed ionic-hole (dry air) to predominantly ionic (wet air, T < 450 °C).

The difference between conductivity values obtained under atmospheres with different water partial pressures is well visible for the Ar atmosphere, and these temperature dependencies are presented in Figure 4c. The protonic transport numbers t_p were calculated according to the equation:

$$t_p=\frac{{\sigma }_{wet Ar}-{\sigma }_{dry Ar}}{{\sigma }_{wet}}$$

(8)

which was used earlier by Y. Zhou et al. [43] for the layered perovskites based on BaNdInO₄. It was established that for both compositions BaLa₂In₂O₇ and BaLaInO₄, the protonic transport numbers increase with a decreasing temperature and reach ~90–95% in the temperature range 300–350 °C.

Therefore, the possibility of protonic transport in the two-layered complex oxide BaLa₂In₂O₇ with the Ruddlesden–Popper structure was shown for the first time. It was proved that an increase in n in the formula BaLa_nIn_nO_3n+1 leads to the increase in the ionic conductivity. The composition BaLa₂In₂O₇ demonstrates nearly pure protonic conductivity under wet air below 350 °C. This allows for considering this complex oxide as a promising matrix composition for obtaining novel high-conductive protonic electrolytes based on it.

4. Conclusions

In this work, the complex oxide BaLa₂In₂O₇ was synthesized, and the electrical conductivity was measured for the first time. It was shown that change in the ratio between perovskite blocks and rock-salt layers in the layered perovskite-related complex oxides BaLa_nIn_nO_3n+1 (n = 1, 2) influences the ionic conductivity values. The increasing n in the formula leads to the increase in the ionic conductivity due to the increase in the share of perovskite blocks. It was proved that the complex oxide BaLa₂In₂O₇ demonstrates nearly pure protonic conductivity under wet air below 350 °C, and it can be considered as the promising compound for obtaining novel high-conductive protonic electrolytes based on it.