Enhanced oxygen storage capacity of cation-ordered cerium–zirconium oxide induced by titanium substitution

Abstract

Herein, we report on the synthesis of Ce_0.5Zr_0.5−xTi_xO₂ oxygen storage materials prepared via a solution combustion method. Ce_0.5Zr_0.4Ti_0.1O₂ showed an outstanding oxygen storage capacity (1310 μmol-O per g) at 200 °C compared to conventional κ-Ce₂Zr₂O₈ (650 μmol-O per g) due to its cation ordering and the formation of weakly bound oxygen atoms induced by Ti substitution.

---

Some metal oxides can reversibly store/release oxygen or control the oxygen concentration in the gas phase in response to changes in the temperature and oxygen partial pressure.¹ Such oxides have been actively investigated for different applications in the fields of energy conversion and environmental protection such as solid oxide fuel cells (SOFCs),² thermochemical water splitting,³ and water–gas shift reactions.⁴ Ceria (CeO₂) is generally used as an oxygen storage material in automotive exhaust catalysis in order to maintain the high purification efficiency by precisely controlling the oxygen partial pressure.^5,6

The oxygen storage/release capacity (OSC) of CeO₂ corresponding to the redox reaction between Ce³⁺ and Ce⁴⁺ is greatly enhanced by Zr substitution, reaching its maximum for Ce_0.5Zr_0.5O₂ (900 μmol-O per g at 500 °C).^1a High-temperature reduction (in H₂ or CO at 1200 °C) of Ce_0.5Zr_0.5O₂ yields the Ce₂Zr₂O₇ pyrochlore (Ce and Zr site ordered Ce_0.5Zr_0.5O_1.75), which leads to cation-ordered Ce_0.5Zr_0.5O₂ after a re-oxidation treatment.^7–9 This cation-ordered Ce_0.5Zr_0.5O₂ (referred to as κ-Ce₂Zr₂O₈) shows an outstanding OSC of 1500 μmol-O per g at 500 °C, where almost all Ce atoms in κ-Ce₂Zr₂O₈ (89%) contribute to the Ce³⁺ ↔ Ce⁴⁺ redox conversion.^9,10 Such a high OSC value is derived from the topotactic transformation of the κ-Ce₂Zr₂O₈ into Ce₂Zr₂O₇ pyrochlore assessed by the first-principles calculation method.¹¹

The cation ordering in complex oxides is often induced by cation size mismatches.^12–14 Therefore, an increase in the cation size mismatch from 1.15 (Ce⁴⁺/Zr⁴⁺) to 1.58 (Ce³⁺/Zr⁴⁺) should stabilize the cation-ordered structure of the Ce₂Zr₂O₇ pyrochlore. Thus, the substitution of Zr⁴⁺ (0.84 Å) with Ti⁴⁺ (0.74 Å) in Ce_0.5Zr_0.5O₂ would result in cation-ordered Ce_0.5Zr_0.5−xTi_xO₂ due to the expansion of the mismatch against Ce⁴⁺ (0.97 Å).¹⁵ However, conventional methods such as solid state reactions and co-precipitations are not able to achieve homogeneous Ti distribution.

In this study, we employed a solution combustion method that is characterized as a self-sustained exothermic reaction.^16–18 Technically, the combustion of fuels (e.g. urea, glycine, etc.) with oxidants (typically metal nitrates) followed by a rapid quenching allowed us to obtain homogeneous compounds with desired compositions.^19,20 Also, the Ti substitution in Ce_0.5Zr_0.5−xTi_xO₂ (0 ≤ x ≤ 0.2) induced pyrochlore-type cation ordering between the Ce and Zr sites and greatly increased the OSC values. The OSC of Ce_0.5Zr_0.4Ti_0.1O₂ (1310 μmol-O per g) was found to be approximately twice that of κ-Ce₂Zr₂O₈ (650 μmol-O per g) at a lower temperature (200 °C). Moreover, the origins of the cation ordering and enhanced OSC were investigated from the viewpoints of crystal structures and the strengths of the bonds between the oxygen atoms and the surrounding cations.

The Ce_0.5Zr_0.5−xTi_xO₂ samples were prepared from aqueous solutions containing both glycine as a fuel and the respective metal nitrates as oxidants. The solutions with cation ratios of Ce : Zr : Ti = 0.5 : (0.5 − x) : x (x = 0, 0.1, 0.2, 0.3, 0.4, and 0.5) were burned by heating up to 400 °C, resulting in light-yellow powders (experimental details are provided in the ESI†). The specific surface areas (SSAs) of the Ti-containing samples (x = 0.1 : 8.29 m² g⁻¹, x = 0.2 : 7.48 m² g⁻¹) were less than half of the amount of x = 0 (17.08 m² g⁻¹). However, these SSAs were much higher than that of κ-Ce₂Zr₂O₈ (0.35 m² g⁻¹), which was obtained by a high-temperature reduction of Ce_0.5Zr_0.5O₂ at 1200 °C followed by oxidation at 500 °C (Table 1).

Table 1 Physical characteristics of Ce_0.5Zr_0.5−xTi_xO₂ and κ-Ce₂Zr₂O₈

Compound	SSA^a (m^2 g^−1)	Space group	Cell volume^b (Å)	Cation ratio^c (mol%)			OSC^d (μmol-O per g)
				Ce	Zr	Ti	200 °C	400 °C	600 °C
a Measured by BET analysis. b Calculated from the lattice parameters based on the unit cell of Ce0.5Zr0.5O2 (Z = 4). c Measured by ICP-OES. d Estimated from the weight losses in the TGA curves. e Prepared by the reduction of Ce0.5Zr0.5O2 at 1200 °C followed by oxidation at 500 °C.^9 f Feed ratio in the synthesis.
Ce0.5Zr0.5O2	17.08	Fm3̄m	146.167(5)	50.2	49.8	—	910	1070	1120
Ce0.5Zr0.4Ti0.1O2	8.29	F4̄3m	144.729(3)	50.3	40.0	9.7	1310	1530	1530
Ce0.5Zr0.3Ti0.2O2	7.48	F4̄3m	143.598(4)	50.2	30.1	19.7	1370	1550	1590
κ-Ce2Zr2O8^e	0.35	P213	145.758(5)	(50)^f	(50)^f	—	650	1530	1560

---

A cubic phase was observed in the X-ray diffraction (XRD) patterns of x = 0, 0.1, and 0.2 (Fig. 1a). An unknown impurity with a reflection at around 2θ = 26.7° was found, but its intensity was less than 1% of the main cubic phase. Additional phases (CeO₂ and unknown) were found in the patterns of x = 0.3, 0.4, and 0.5 (Fig. S1, ESI†). The main phase of x = 0 was characterized by a cubic fluorite structure (Fm3̄m) with a = 5.2682(2) Å, which was in agreement with the reported space group (Fm3̄m) and lattice parameter (a = 5.269) of Ce_0.5Zr_0.5O₂.²¹ In contrast, the main phases of x = 0.1 and 0.2 showed additional peaks originating from their superstructures assigned to a 2 × 2 × 2 supercell based on Ce_0.5Zr_0.5O₂. The cell volumes calculated from the lattice parameters showed a linear decrease with an increase in the value of x (Fig. 1b and Table 1). The volume reduction indicated the formation of Ce_0.5Zr_0.5−xTi_xO₂ solid solutions (0 ≤ x ≤ 0.2) because the ionic radius of tetravalent Ti (0.74 Å) is smaller than that of Zr (0.84 Å).¹⁵ Elemental analysis using inductively coupled plasma-optical emission spectrometry (ICP-OES) supported the formation of the solid solution since the measured compositions of the obtained Ce_0.5Zr_0.5−xTi_xO₂ samples were equal to the feed ratios of Ce : Zr : Ti = 5 : 5 : 0, 5 : 4 : 1, and 5 : 3 : 2 for x = 0, 0.1, and 0.2, respectively (Table 1).

Fig. 1 (a) XRD patterns of Ce_0.5Zr_0.5−xTi_xO₂. Black, red, and blue solid lines represent the patterns of x = 0, 0.1, and 0.2, respectively. Hashtags (#), asterisks (*), and black squares (◆) indicate peaks arising from the main cubic phase, their superstructure, and unknown impurity, respectively. (b) Cell volumes of Ce_0.5Zr_0.5−xTi_xO₂ based on the unit cell of Ce_0.5Zr_0.5O₂ (Z = 4).

Rietveld refinement of the synchrotron XRD pattern of Ce_0.5Zr_0.4Ti_0.1O₂ was performed using a F4̄3m model in which Ce and Zr(Ti) atoms were ordered along the 〈110〉 direction (Fig. 2a). The high reliability of the refined data (R_wp = 6.46%, goodness of fit = 0.80, Table S1, ESI†) suggested that the predicted model for Ce_0.5Zr_0.4Ti_0.1O₂ was reasonable. The crystal structure is shown in Fig. 2a. The electron diffraction (ED) patterns of the [100], [110], and [111] zone axes for Ce_0.5Zr_0.4Ti_0.1O₂ (x = 0.1) were indexed to the space group F4̄3m (Fig. 2b), which was consistent with the structural model used in the Rietveld refinement. Moreover, the absence of 110, 211, and 330 reflections excluded the space group P2₁3, which could be derived from the slight displacements of some O sites from the ideal positions as observed in κ-Ce₂Zr₂O₈.¹⁰ This indicates that the structure of Ce_0.5Zr_0.4Ti_0.1O₂ was less distorted than that of κ-Ce₂Zr₂O₈.

Fig. 2 (a) Structural characterization of Ce_0.5Zr_0.4Ti_0.1O₂ obtained via Rietveld refinement of the synchrotron XRD data at room temperature. The red crosses, green solid line, and blue solid line represent the observed, calculated, and different intensities, respectively. The black ticks show the peak positions. The inset displays the crystal structure of Ce_0.5Zr_0.4Ti_0.1O₂ with the space group F4̄3m. The orange, green, blue, and red spheres represent Ce, Zr, Ti, and O, respectively. (b) ED patterns for Ce_0.5Zr_0.4Ti_0.1O₂ projected along the [100], [110], and [111] axes, indexed constantly to the F4̄3m space group.

The results from the XRD and ED measurements suggested that pyrochlore-type cation ordering in the Ce_0.5Zr_0.5−xTi_xO₂ fluorite was observed above a cation size mismatch (r_Ce⁴⁺/r_{Zr⁴⁺(Ti⁴⁺)}) of 1.18, where r_{Zr⁴⁺(Ti⁴⁺)} denotes the weighted average of ionic radii for Zr⁴⁺ and Ti⁴⁺. In contrast, cation ordering in the A_0.5B_0.5O_1.75 defect fluorite (e.g. Gd_0.5Zr_0.5−xCe_xO_1.75 and Y_0.5Ti_0.5−xZr_xO_1.75) is generally observed above 1.23,^12,13 which is slightly larger than the value for the Ce_0.5Zr_0.5−xTi_xO₂ fluorite (1.18). This difference suggests that the absence of oxygen defect enabled cation ordering in the A_0.5B_0.5O₂ fluorite with a smaller cation size mismatch.

The OSC values of Ce_0.5Zr_0.5−xTi_xO₂ (0 ≤ x ≤ 0.2) were estimated from the weight losses recorded using thermogravimetric analysis (TGA) during the atmosphere switching between 5% O₂/N₂ and 5% H₂/N₂ every 5 min (Fig. 3a).

Fig. 3 (a) TGA curves of Pt/Ce_0.5Zr_0.5−xTi_xO₂ upon switching between 5% O₂/N₂ and 5% H₂/N₂ every 5 min at various temperatures. (b) Estimated OSCs. Black, red, and blue lines and columns represent Pt/Ce_0.5Zr_0.5O₂, Pt/Ce_0.5Zr_0.4Ti_0.1O₂, and Pt/Ce_0.5Zr_0.3Ti_0.2O₂, respectively.

The weight losses and gains of the samples corresponded to the following reactions:

Ce_0.5Zr_0.5−xTi_xO₂ + δH₂ → Ce_0.5Zr_0.5−xTi_xO_2−δ + δH₂O (1)

Ce_0.5Zr_0.5−xTi_xO_2−δ + δ/2O₂ → Ce_0.5Zr_0.5−xTi_xO₂ (2)

Pt-loaded samples (1 wt% on a metal basis) were used to quickly saturate their weight changes. As shown in Fig. 3 and Table 1, Pt/Ce_0.5Zr_0.4Ti_0.1O₂ and Pt/Ce_0.5Zr_0.3Ti_0.2O₂ showed considerably higher OSCs than Pt/Ce_0.5Zr_0.5O₂ at every temperature. The fact that Pt/Ce_0.5Zr_0.4Ti_0.1O₂ and Pt/Ce_0.5Zr_0.3Ti_0.2O₂ exhibited nearly equal OSCs at 600 °C (δ = 0.219 and 0.221, respectively) suggested that Ti would not directly contribute to their OSCs via a Ti⁴⁺ ↔ Ti³⁺ redox. This assumption was also supported by the irreducible nature of Ti⁴⁺ in this temperature range.²² Fractions of reducible Ce⁴⁺ ([Ce³⁺]/[Ce³⁺] + [Ce⁴⁺]) based on the theoretical maximum δ value of 0.25 (Table 1) were calculated to be 66.2%, 87.7%, and 88.4% for Pt/Ce_0.5Zr_0.5O₂, Pt/Ce_0.5Zr_0.4Ti_0.1O₂, and Pt/Ce_0.5Zr_0.3Ti_0.2O₂, respectively. These values clearly indicate that the high OSCs of Pt/Ce_0.5Zr_0.4Ti_0.1O₂ and Pt/Ce_0.5Zr_0.3Ti_0.2O₂ originate from the increase in the fraction of reducible Ce⁴⁺.

Ce L₃-edge X-ray absorption near edge structure (XANES) spectra of Pt/Ce_0.5Zr_0.5O₂ and Pt/Ce_0.5Zr_0.4Ti_0.1O₂ at 600 °C in 3% O₂/He (Fig. 4a) showed two absorption peaks, at 5730 and 5737 eV (marked as A₁ and A₂), which were assigned to the Ce⁴⁺ species. These results were in agreement with the CeO₂ (Ce⁴⁺) reference spectrum (Fig. S2, ESI†).²³ In 3% H₂/He, only one absorption peak, at 5726 eV (marked as B), assigned to the Ce³⁺ species (shown in the Ce(NO₃)₃·6H₂O (Ce³⁺) reference) was observed, which indicated the reduction of Ce⁴⁺ species to Ce³⁺. Moreover, in 3% H₂/He the absorption of Pt/Ce_0.5Zr_0.4Ti_0.1O₂ at B was higher than that of Pt/Ce_0.5Zr_0.5O₂, while in 3% O₂/He both samples showed almost the same absorption intensities. This result indicated that the fraction of reducible Ce⁴⁺ in Pt/Ce_0.5Zr_0.4Ti_0.1O₂ was higher than that in Pt/Ce_0.5Zr_0.5O₂, which was in agreement with the TGA results (Fig. 3b). The transient response of the peak height of the absorption edge during the switching between 3% H₂/He and 3% O₂/He (Fig. 4b and Fig. S3, ESI†) suggested that the Ce⁴⁺ reduction rate in Pt/Ce_0.5Zr_0.4Ti_0.1O₂ was higher than that in Pt/Ce_0.5Zr_0.5O₂. For both samples, the intensities that increased in 3% H₂/He immediately returned to the original intensities when the atmosphere was changed to 3% O₂/He. This demonstrates that the reduced Ce³⁺ species were reversely re-oxidized to Ce⁴⁺ with high reproducibility.

Fig. 4 (a) Ce L₃-edge XANES spectra of Pt/Ce_0.5Zr_0.5O₂ (black lines) and Pt/Ce_0.5Zr_0.4Ti_0.1O₂ (red lines) in 3% O₂/He (dashed lines) and 3% H₂/He (solid lines) at 600 °C. Peaks A₁ and A₂ were assigned to Ce⁴⁺, whereas peak B was assigned to Ce³⁺. (b) Time-resolved absorption edge intensities in Ce L₃-edge XANES spectra during the atmosphere switching between 3% O₂/He and 3% H₂/He every 5 min at 600 °C. Data were collected every 0.2 s for 3 min after changing the atmosphere. Black and red lines represent Pt/Ce_0.5Zr_0.5O₂ and Pt/Ce_0.5Zr_0.4Ti_0.1O₂, respectively.

The high OSCs of Pt/Ce_0.5Zr_0.4Ti_0.1O₂ could have originated from the topotactic transformation of the Ce_0.5Zr_0.4Ti_0.1O₂ into Ce_0.5Zr_0.4Ti_0.1O_1.75 pyrochlore. Bond valence sums (BVSs)^24,25 are often used to estimate the bond strengths of specific anion sites in crystals.^26,27 The respective BVSs of all the oxygen sites in Ce_0.5Zr_0.4Ti_0.1O₂ (marked as O1–O6; Fig. S4, ESI†) were obtained, and the OZr(Ti)₄ tetrahedron sites (O3 site: −1.61, O4 site: −1.37) were found to have larger values than the other O sites (from −1.82 to −2.46; Table S2, ESI†). The larger BVS values of the O3 and O4 sites suggested that the oxygen atoms at these sites were easily released, which caused the topotactic transformation of the Ce_0.5Zr_0.4Ti_0.1O₂ (Ce⁴⁺) into Ce_0.5Zr_0.4Ti_0.1O_1.75 (Ce³⁺) pyrochlore. These conclusions were in agreement with the results from κ-Ce₂Zr₂O₈ (Ce⁴⁺), which was topotactically reduced to the Ce₂Zr₂O₇ (Ce³⁺) pyrochlore with the release of oxygen atoms at the O3 and O4 sites (Fig. S4, ESI†).²⁸

Interestingly, the OSC value of Pt/Ce_0.5Zr_0.4Ti_0.1O₂ (1310 μmol-O per g) at a lower temperature (200 °C) was approximately twice that of Pt/κ-Ce₂Zr₂O₈ (650 μmol-O per g, Fig. 5a). In contrast, the OSC values of both compounds above 400 °C were almost the same (1530–1560 μmol-O per g; Table 1 and Fig. S5, ESI†). Prolonging the reduction period (120 min) for Pt/κ-Ce₂Zr₂O₈ at 200 °C gave an OSC value of 740 μmol-O per g (Fig. S6, ESI†), which is still smaller than that of Pt/Ce_0.5Zr_0.4Ti_0.1O₂. This suggests that another factor other than SSA would determine the high OSC of Pt/Ce_0.5Zr_0.4Ti_0.1O₂. As shown in Fig. 5b and c, the BVS value of the O4 site in Ce_0.5Zr_0.4Ti_0.1O₂ (−1.37) was larger than that in κ-Ce₂Zr₂O₈ (−1.68), while the O3 sites in both had similar BVS values (−1.61 for Ce_0.5Zr_0.4Ti_0.1O₂ and −1.59 for κ-Ce₂Zr₂O₈). The larger BVS value of the O4 site in Ce_0.5Zr_0.4Ti_0.1O₂ indicates that the oxygen atoms at this site were more weakly bound to the surrounding cations. The presence of such weakly bound oxygen atoms in Ce_0.5Zr_0.4Ti_0.1O₂ would decrease the oxygen vacancy formation energy, leading to a high OSC even at a lower temperature (200 °C).

Fig. 5 (a) TGA curves of Pt/Ce_0.5Zr_0.4Ti_0.1O₂ (black dashed line) and Pt/κ-Ce₂Zr₂O₈ (red solid line) during the atmosphere switching between 5% O₂/N₂ and 5% H₂/N₂ every 5 min at 200 °C. BVS values of the O3 and O4 sites in (b) Ce_0.5Zr_0.4Ti_0.1O₂ and (c) κ-Ce₂Zr₂O₈.

In conclusion, Ce_0.5Zr_0.5−xTi_xO₂ (0 ≤ x ≤ 0.2) were successfully synthesized via the solution combustion method. The substitution of Ti in Ce_0.5Zr_0.5O₂ induced pyrochlore-type cation ordering due to the increased cation size mismatch between the Ce and Zr sites, which caused a significant enhancement of the OSC. The most remarkable observation for the cation-ordered Ce_0.5Zr_0.4Ti_0.1O₂ was its high OSC (1310 μmol-O per g) even at a lower temperature (200 °C), which was twice that of conventional κ-Ce₂Zr₂O₈ (650 μmol-O per g). The high OSC at low temperature was explained by the formation of weakly bound oxygen atoms. This study demonstrated that an appropriate compositional control of Ce_0.5Zr_0.5O₂ allows for the induction of pyrochlore-type cation ordering without a high-temperature reduction. Future investigations of the various applications and catalytic activity of these cation-ordered materials synthesized via the solution combustion strategy would be of high interest.

The authors would like to thank Mr Shimotsu (Hitachi Power Solutions Co., Ltd, Ibaraki, Japan) for the support with the ED measurements and Dr Towata (Aichi Synchrotron Radiation Center, Aichi, Japan) for the support with the synchrotron XRD measurements.

Conflicts of interest

There are no conflicts of interest to declare.

Notes and references

1. (a) M. Yashima, Catal. Today, 2015, 253, 3; (b) M. Karppinen, H. Yamauchi, S. Otani, T. Fujita, T. Motohashi, Y.-H. Huang, M. Valkeapää and H. Fjellvåg, Chem. Mater., 2006, 18, 490; (c) S. Remsen and B. Dabrowski, Chem. Mater., 2011, 23, 3818; (d) T. Motohashi, Y. Hirano, Y. Masubuchi, K. Oshima, T. Setoyama and S. Kikkawa, Chem. Mater., 2013, 25, 372; (e) A. Demizu, K. Beppu, S. Hosokawa, K. Kato, H. Asakura, K. Teramura and T. Tanaka, J. Phys. Chem. C, 2017, 121, 19358.
2. Z. Xu, Z. Qi and A. Kaufman, Power Sources, 2003, 115, 40.
3. W. C. Chueh, C. Falter, M. Abbott, D. Scipio, P. Furler, S. M. Haile and A. Steinfeld, Science, 2010, 330, 1797.
4. C. M. Kalamaras, D. D. Dionysiou and A. M. Efstathiou, ACS Catal., 2012, 2, 2729.
5. A. F. Diwell, R. R. Rajaram, H. A. Shaw and T. J. Truex, Stud. Surf. Sci. Catal., 1991, 71, 139.
6. J. Kašpar and P. Fornasiero, J. Solid State Chem., 2003, 171, 19.
7. S. Otsuka-Yao-Matsuo, T. Omata, N. Izu and H. Kishimoto, J. Solid State Chem., 1998, 138, 47.
8. A. Suda, Y. Ukyo, H. Sobukawa and M. Sugiura, J. Ceram. Soc. Jpn., 2002, 110, 126.
9. Y. Nagai, T. Yamamoto, T. Tanaka, S. Yoshida, T. Nonaka, T. Okamoto, A. Suda and M. Sugiura, Catal. Today, 2002, 74, 225.
10. H. Kishimoto, T. Omata, S. Otsuka-Yao-Matsuo, K. Ueda, H. Hosono and H. Kawazoe, J. Alloys Compd., 2000, 312, 94.
11. H.-F. Wang, Y.-L. Guo, G.-Z. Lu and P. Hu, Angew. Chem., Int. Ed., 2009, 48, 8289.
12. D. P. Reid, M. C. Stennett and N. C. Hyatt, J. Solid State Chem., 2012, 191, 2.
13. C. Heremans, B. J. Wuensch, J. K. Stalick and E. Prince, J. Solid State Chem., 1995, 117, 108.
14. G. King and P. M. Woodward, J. Mater. Chem., 2010, 20, 5785.
15. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
16. B. D. Mukri, U. V. Waghmare and M. S. Hegde, Chem. Mater., 2013, 25, 3822.
17. B. Murugan, A. V. Ramaswamy, D. Srinivas and C. S. Gopinath, Chem. Mater., 2005, 17, 3983.
18. A. A. Saad, W. Khan, P. Dhiman, A. H. Naqvi and M. Singh, Electron. Mater. Lett., 2013, 9, 77.
19. M. S. Hegde, G. Madras and K. C. Patil, Acc. Chem. Res., 2009, 42, 704.
20. A. Varma, A. S. Mukasyan, A. S. Rogachev and K. V. Manukyan, Chem. Rev., 2016, 116, 14493.
21. S. F. Pal'guev, S. I. Alyamovskii and Z. S. Volchenkova, Russ. J. Inorg. Chem., 1959, 4, 1185.
22. B. Xu, H. Y. Sohn, Y. Mohassab and Y. Lan, RSC Adv., 2016, 6, 79706.
23. J. El Fallah, S. Boujana, H. Dexpert, A. Kiennemann, J. Majerus, O. Touret, F. Villain and F. Le Normand, J. Phys. Chem., 1994, 98, 5522.
24. I. D. Brown and D. Altermatt, Acta Crystallogr., Sect. B: Struct. Sci., 1985, 41, 244.
25. N. E. Brese and M. O’Keeffe, Acta Crystallogr., Sect. B: Struct. Sci., 1991, 47, 192.
26. A. Gupta, A. Kumar, U. V. Waghmare and M. S. Hegde, J. Phys. Chem. C, 2017, 121, 1803.
27. N. Wakiya, K. Shinozaki, N. Mizutani and N. Ishizawa, J. Am. Ceram. Soc., 1997, 80, 3217.
28. S. N. Achary, S. K. Sali, N. K. Kulkarni, P. S. R. Krishna, A. B. Shinde and A. K. Tyagi, Chem. Mater., 2009, 21, 5848.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, XRD patterns, XANES spectra, crystal structures, TGA curves, crystallographic information, and BVS parameters. See DOI: 10.1039/c8cc01308j

---