Table 1 shows the Rietveld refinement results for BaZr_0.1Ce_0.7Y_0.1 M_0.1O_3−δ (M = Fe, Co, Ni, and Yb). The refinement XRD profiles are shown in Figure 1, with the formation of the orthorhombic space group Imam, thus confirmed. The crystal structure involves Ce, Zr, Y, and M sharing the same lattice position. These cations are coordinated by six oxygen ions that form an octahedral geometry, as shown in Figure 2A. Each oxygen anion is bonded to two cations on both sides, leading to corner-sharing between the octahedra (Figure 2A). There are spaces between these octahedra where the large Ba²⁺ cations reside (Figures 2A,B). The Ce(ZrYM)O6 octahedron is tilted in a direction opposite to all of its nearest neighbors. Therefore, if an octahedron is tilted to the right, all of the nearest neighbor octahedra will be tilted to the left. The Ba atoms have 12 oxygen anions as their nearest neighbors (Figure 2C).

The unit cell volumes for the M = Fe and Co phases are similar to each other (Figure 1), while there is a pronounced increase for the Ni-containing oxide and also a further increase for the Yb-doped material. These observations are consistent with the large ionic radii of Ni²⁺ and Yb³⁺ (Shannon, 1976) and support the hypothesis of the substitution of the transition metal dopants in the B(Ce)-site in BZCY. However, the unit cell volume for all four phases is markedly smaller than that of the un-doped BaCeO₃, which is 339.86 ų (Knight and Bonanos, 1995), due to the effect of the small Zr⁴⁺ ion in the doped phases.

Typical AC impedance plots obtained at 484 and 837°C for BaZr_0.1Ce_0.7Y_0.1Ni_0.1O_3−δ in H₂ (+3 vol.% H₂O) are shown in Figure 3. The total resistance (bulk + grain boundary + electrode) is used to calculate the conductivity of the investigated samples. At lower temperatures (Figure 3A), three distinct arcs are observed. At higher temperatures, although it seems that the two higher frequency arcs have merged into one arc, the impedance data can still be fitted better by using three RC units, as is shown in the inset of Figure 3B. The magnitude of the constant phase elements (CPE) is larger than values normally attributed to oxygen ion conductors and is close to the capacitance reported for the proton conductors (Petit and Tao, 2013). The conductivity (σ) of the compounds was computed using Eq. 2: (2)σ=lR×A where l, R, and A are the thickness (cm), total resistance (Ω), and cross sectional area (cm²) of the cylindrical sample. Figure 4 shows the Arrhenius plots for the total conductivity of the compounds in air and H₂ + 3 vol.% H₂O after sintering at 1250, 1350, and 1550°C. Among the 1250 (Figures 4A,D) and 1350°C (Figures 4B,E) sintered samples, the Ni-doped samples show a higher conductivity. However, by increasing the sintering temperature to 1550°C, the Co-doped sample showed the highest conductivity (0.01 S cm⁻¹ at 600°C in wet H₂) (Figures 4C,F) of the new compounds synthesized in this study. The conductivity of the Co-doped sample in both air and wet H₂ was found to increase by about 10 times by increasing sintering temperature by 300°C, while the conductivity of the Ni-doped sample is suppressed by ca. 3 times. The Fe-doped sample does not show significant changes in conductivity, even after firing at 1550°C, while the BZCY-Co compound, sintered at 1550°C, shows the highest total conductivity of 2.6 × 10⁻² S cm⁻¹ at 800°C in air.

The sinterability and phase stability of the as-synthesized compounds at three temperatures (1250, 1350, and 1550°C) were also studied. To serve as a proton conductor electrolyte, materials such as BZCY-Yb and BZCY have generally been sintered at high temperatures, e.g., 1550°C, to achieve the minimum porosity (Yang et al., 2009). Co-sintering of the anode and electrolyte, especially in anode-supported SOFCs, is a common cell fabrication method. During the co-sintering of these two materials, the anode materials are exposed to a temperature high enough to achieve electrolyte sintering, and thus their stability at these temperatures is necessary. Therefore, the effect of sintering at 1550°C on the chemical stability and microstructure of the synthesized compounds was investigated. As a reference, the sintering behavior of these compounds is compared with that of BZCY-Yb.

Figures 5A–C show the effect of the sintering temperature on the microstructure of the BZCY-Fe compound. While the sintering process has already started at 1250°C, the dense sample is obtained at 1550°C. The morphology of the BZCY-Ni and BZCY-Co compounds after sintering at different temperatures is shown in Figures 5D–I, respectively. Both samples show a dense morphology even at 1250°C, confirming the effect of Ni and Co as effective sintering aids. By increasing the sintering temperature, the grain size increases considerably. On the other hand, the sintering of the BZCY-Yb compound only begins at 1550°C, and after 12 h of sintering at this temperature; the sample still has some porosity (Figure 5L). Comparing Figures 5J,K, it can be seen that the 1350°C sintering temperature has caused some grain growth, but without noticeable sintering of the BZCY-Yb sample. Also, considering Figures 5H,I, a secondary phase (marked with the letter “S”) is observable in the BZCY-Co sample. Observation of a BaCoO₃ phase in a similar system has been reported in the literature (Yang et al., 2011). Based on the SEM image, the secondary phase seems to wet the BZCY-Co grains very well, running into the grain-boundaries. Similar to the BZCY-Co sample, the Ni-doped compound also shows a secondary phase (Figure 5F) after firing at 1550°C in air. However, no secondary phase is observed for the Fe- and Yb-doped samples, even after firing at 1550°C.

The chemical stability of the Ni-, Co-, and Fe-doped BZCY-M compounds was studied in CO₂/N₂ (1:2 ratio) and in 30 ppm H₂S/H₂ and results are shown in Figures 6 and 7, respectively. To eliminate the effect of the synthesis method on the stability, the BZCY compound was also synthesized using a similar method (at 1350°C). The BZCY compound shows a minimal conversion to BaCO₃, while the 1350°C synthesized doped BZCY-M compounds start to gain weight in the temperature range of 450–500°C (Figure 6A). The decomposition process accelerates at around 600°C and, at 800°C, reaches maximum values of 6, 12, 14, and 17% conversion to BaCO₃ for the BZCY-Fe, -Ni, -Co, and -Yb-doped samples, respectively. Interestingly, except for the BZCY sample, the CO₂ stability trend of the BZCY-M compounds follows the reverse pattern of the conductivity trend. While the BZCY-Yb and BZCY-Co samples show some of the highest conductivity values among the compounds examined here, they show the highest CO₂ uptake and thus the lowest chemical stability. Similarly, the BZCY-Fe compound, with the lowest conductivity, illustrates a lower CO₂ uptake and better stability. On the other hand, after the samples were fired at 1550°C for 12 h, the chemical stability in CO₂ has increased significantly, with the BZCY-Ni sample not gaining any measurable weight. The percent conversion of the Fe- and Co-doped samples to BaCO₃ is also now less than 2%, while the Yb-doped sample still exhibits greater decomposition, with a close to 6% conversion to BaCO₃.

The chemical stability of the as-synthesized compounds was also studied in 30 ppm H₂S in a dry H₂ atmosphere, with the ex situ XRD results are shown in Figure 7. After heating the BZCY, BZCY-Fe, Co, and Ni compounds in powder form at 800°C for 24 h in 30 ppm H₂S balanced with dry H₂, no impurity phase is observed in the BZCY-Fe and -Co samples. In the BZCY-Ni XRD pattern, however, a weak peak for metallic Ni is observed [joint committee on powder diffraction standards (JCPDS) card no.: 45–1027], which could be related to the ex-solution of Ni after reduction in dry H₂ at 800°C. The BZCY synthesized in our lab showed small impurity peaks, however, after treatment in 30 ppm H₂S/H₂ at 800°C for 24 h no new phases were observed (Figure 7).

Previous studies of the crystal structure of BaCeO₃ at a range of temperatures has shown that a transition from the space group Pnma to Imam occurs at 290°C (Knight, 1994) or 300°C (Genet et al., 1999). This transition has been identified, based on the disappearance of the 201 and 221 peaks (Genet et al., 1999), which further confirms the space group Imam. In addition to the 201 and 221 peaks, another distinct Pnma peak, which is absent, is 131 that should be just below the 221 peak. The observation of the Imam space group is interesting, as in the parent compound (BaCeO₃), the occurrence of an Imam structure is triggered by the increase in temperature, while in BaZr_0.1Ce_0.7Y_0.1M_0.1O_3−δ, this space group is observed at room temperature (Figure 1).

This change in crystal symmetry as a result of doping has been observed previously, where a large amount of Y³⁺ has replaced Ce⁴⁺ in BaCeO₃ (Malavasi et al., 2008). The BaCe_0.8Y_0.2O_3−δ compound, with 20% Y³⁺ doping, has been reported to have a monoclinic I2/m space group at room temperature (Malavasi et al., 2008), whereas, for BaCe_0.9Y_0.1O_3−δ with 10% doping, a Pnma structure was reported at room temperature (Malavasi et al., 2009). The effect of doping has also been observed for Nd-doped BaCeO₃ (Knight, 1999), where 5% doping has been shown to change the crystal symmetry from orthorhombic, Pnma, to tetragonal, P4/mbm, while 10% doping results in the formation of the cubic space group Pm3m. It appears that the disorder prompted by doping can lead to a similar outcome as the thermal effect. Also, the unit cell volume increases from 330.97(5) ų for the Fe-doped sample to 335.35(6) ų for the Ni-doped sample and reaches a maximum of 337.26(7) ų in the Yb-doped sample (Figure 1).

The higher conductivity of the Ni-doped sample at lower sintering temperatures, compared to other samples studied here, may be related to the higher concentration of oxygen vacancies (in air) introduced into the crystal structure upon substitution of Ni²⁺ at the B-site of the perovskite material. The oxygen vacancies present in the structure could be transformed to proton charge carriers (H⁺) upon exposure to water in wet H₂ (Eq. 1). In most cases, the activation energy associated with the conductivity of these materials in air is higher than in wet H₂. While it is expected that some electronic conductivity is present in both atmospheres, the lower activation energy in wet H₂ could be attributed to differences in the charge carriers, i.e., protons (H⁺) in wet H₂ compared to oxygen ions (O²⁻) in air. Due to their smaller size, proton transport has lower activation energy, providing a higher conductivity at lower temperatures, compared to oxygen ions.

Increasing the sintering temperature enhances the total conductivity of the BZCY-Co, as shown in Figure 8. This could be related to the formation of the Co-rich secondary phase (Figure 5H). Yang et al. (2011) have shown an increase in conductivity of the BaZr_0.1Ce_0.7Y_0.2−xCo_xO_3−δ (0 ≤ x ≤ 0.2) compound with an increase in the Co content, attributing the higher conductivity values to the formation of a BaCoO₃-based phase. The BaCoO₃ phase has a higher electronic conductivity compared to the parent BZCY phase.

To understand the chemical composition of the secondary phase, EDX line scans were collected across the grains, with the results shown in Figure 9. The inset in Figure 9A shows that the secondary phase has a high concentration of Co, but a decreased Ce content. The Ba concentration remains almost the same in both the primary and secondary phases, suggesting the formation of a BaCoO₃-based perovskite. This result is also consistent with WDS study (not shown). It has been shown that, in BaZr_0.1Ce_0.7Y_0.2−xCo_xO_3−δ, a BaCoO₃-based phase is observed in the XRD pattern when the Co content is about 10 mol% (Yang et al., 2011). However, in contrast to our SEM images (Figures 5H,I and 9A) that clearly show the formation of the secondary phase, XRD results (Figure 1B) do not show the BaCoO₃ peaks after synthesis at 1350°C. This could be related to the lower concentration of the secondary phase in the present work. Yang et al. prepared single-phase BaZr_0.1Ce_0.7Y_0.2−xCo_xO_3−δ after three times sintering and subsequent ball-milling at 1100°C for 10 h. However, in this study the BZCY-Co single-phase was obtained after firing at 1350°C. Calcination at 1350°C may have caused some Co evaporation from the compound, as blue color associated with Co was observed on the alumina crucible. This may have decreased the Co content and lowered the concentration of the BaCoO₃ phase below the XRD detection limit.

In contrast to the Co-doped sample, the conductivity of the BZCY-Ni compound decreases with increasing sintering temperature. At 1550°C, the sub-micron sized NiO particles can be observed at the grain-boundaries (Figure 5F), with their chemical composition confirmed by EDX analysis. A subsequent EDX line scan study on these phases confirmed that Ni and oxygen are the only elements present (Figure 9B). The ex-solution of metals at high temperatures from perovskite materials upon reduction has been reported recently (Neagu et al., 2013). This may have caused the lower conductivity of the BZCY-Ni sample after sintering at 1550°C, compared to the sample fired at two lower temperatures (Figure 8).

The enhanced tolerance of the samples fired at 1550°C toward CO₂ exposure may have been caused by Ba evaporation during heat treatment at higher temperature, which makes these compounds Ba deficient and hence less prone to reaction with CO₂. Although the particle size of the powders used for the two tests is very similar, the samples that were heat treated at 1550°C have larger grain sizes and fewer grain-boundaries, as shown in Figure 5.

In summary, the BZCY-Co compound, studied here, shows a very promising electrical conductivity compared to other compounds arising from the same perovskite family (Figure 10), especially at intermediate temperatures (0.01 S cm⁻¹ at 600°C in wet H₂). Thus, BZCY-Co is considered to be a very promising candidate anode material for SOFCs. The higher sintering temperature of 1550°C enhances the total conductivity of the BZCY-Co compound, due to the formation of a BaCoO₃-based secondary phase. However, the highest conductivity within the BZCY-Ni samples is observed for the sample fired at the lowest sintering temperature of 1250°C. Ni ex-solution from the BZCY-Ni sample is observed after sintering at 1550°C for 12 h and seems to be responsible for the conductivity loss, as a result of bulk composition change. Also, sintering at the higher temperature of 1550°C enhances the stability of the compound in the CO₂-containing environment.