Pressure-induced phase transition of oxygen defective perovskite srebrodolskite Ca₂Fe₂O₅

Abstract

To clarify the effect of oxygen defects on the perovskite structure under high pressure, structural changes in srebrodolskite Ca₂Fe₂O₅ were investigated by using high-pressure Raman spectroscopy and synchrotron powder X-ray diffraction analyses. The result of the high-pressure Raman spectroscopic study showed that with compression, a new Raman band appeared at 12.0 GPa. Furthermore, an additional new Raman band appeared at 16.0 GPa. The phase-transition pressure was approximately consistent with the previous research, and the intensities of these new bands became much stronger with increasing pressure. At least nine Raman bands were observable at 23.0 GPa. A high-pressure synchrotron powder X-ray diffraction study was performed up to 20.2 GPa. The obtained pressure-volume compression curve apparently deviated from the equation-of-state of srebrodolskite determined by the previous study above 9.1 GPa, at which point srebrodolskite began to transform into its high-pressure phase. The Rietveld refinement of the X-ray diffraction data at 12.6 GPa fitted with space group Pn2₁a yielded agreement factors of R_p = 1.46% and wR_p = 2.01%. The second high-pressure phase transition occurred at 14.2 GPa with the emergence of new reflections at d-spacing values of 3.938 and 1.953 Å. The powder X-ray diffraction patterns of the second high-pressure phase were characterized by three reflections appearing at approximately d-spacing values of 3.938, 2.609, and 1.953 Å. Consequently, the second high-pressure phase is likely to be composed of a new structure that is not included in the known brownmillerite-type structures. The results provide clues for understanding the physical properties of the chemically heterogeneous Earth’s mantle.

INTRODUCTION

CaSiO₃ perovskite is considered to be one of the constituents of the Earth’s lower mantle (Irifune, 1994). The calcium iron (III) oxide Ca₂Fe₂O₅ (srebrodolskite) (Chesnokov and Bazhenova, 1985) has an orthorhombic brownmillerite-type structure (Colville and Geller, 1971; Redhammer et al., 2004), which is regarded as a perovskite-related structure with ordered oxygen vacancies (Antipov et al., 2008). The defect formation mechanism in CaSiO₃ perovskites can be explained as a coupled substitution (oxygen vacancy substitution), in which two silicon cations are exchanged for two trivalent cations and the charge is balanced by the creation of one oxygen vacancy (Navrotsky et al., 2003; Kojitani et al., 2007). The incorporation of trivalent cations into the CaSiO₃ davemaoite and MgSiO₃ bridgmanite produces significant changes in the physical and chemical properties of the lower mantle (Jackson and Rigden, 1996; Mattern et al., 2005; Tsuchiya and Tsuchiya, 2006; Catalli et al., 2010; Wang et al., 2015). Therefore, it is important to investigate the high-pressure behavior of brownmillerite-type minerals to better understand Earth’s deep interior.

Brownmillerite-type structures and their variations have been well investigated (Hentschel, 1964; Colville and Geller, 1971; Kahlenberg et al., 2000; Redhammer et al., 2004; Krüger and Kahlenberg, 2005; Lazic et al., 2008). Figure 1 shows the stacking sequences of the octahedral and tetrahedral layers in the brownmillerite-type structure with a general composition of A₂B₂O₅. It consists of two fundamental building units: two types of layers composed of perovskite-type corner-sharing BO₆ octahedra and single chains of corner-sharing BO₄ tetrahedra running between the octahedral layers. The alternating stacking of the octahedral and tetrahedral layers forms a three-dimensional network in which the A cations are incorporated into the interstitial space for charge compensation. The tetrahedral layers can adopt two mirror-related configurations, denoted as left-handed (L) and right-handed (R) (Milat et al., 1993). The brownmillerite-type structure crystallizes in space group Pnma (No. 62) or I2mb (No. 46). The centric space group Pnma has alternating layers of R and L chains, whereas the acentric space group I2mb possesses only one type (R or L) of chain within the layers.

Figure 1. Brownmillerite-type structure of A₂B₂O₅, where black spheres show A cations and green and yellow polyhedra show B cations in octahedral and tetrahedral sites, respectively. Two distinct ordering patterns involving left-handed (L) and right-handed (R) tetrahedral chains yield centric space group Pnma and acentric space group I2mb.

Here, it is important to note that the Ca₂(Fe_2−xAl_x)O₅ series form a complete solid solution between the Ca₂Fe₂O₅ and Ca₂Al₂O₅ end members, however they are not isostructural throughout the compositional range. The pure iron end-member srebrodolskite Ca₂Fe₂O₅ and the ion-rich region with composition of 0 ≤ x ≤ 0.56 have space group Pnma, whereas the solid solution with composition of 0.56 < x shows space group I2mb (Redhammer et al., 2004). The chemical composition of brownmillerite Ca₂FeAlO₅ represents one point with one-to-one at the ratio of Al and Fe in the solid solution. The pure aluminum end-member Ca₂Al₂O₅, which has only been obtained under high-pressure and high-temperature conditions (Aggarwal et al., 1972; Kahlenberg et al., 2000), also crystallizes in space group I2mb. Another remarkable feature is that the iron and aluminum end-members transform into incommensurately modulated structures at high temperatures (Krüger and Kahlenberg, 2005; Lazic et al., 2008; Krüger et al., 2009). The modulated high-temperature phase has a body-centered orthorhombic lattice with space group Imma(00γ)s00, in which arrangement of the R and L tetrahedral chains becomes an aperiodic sequence (Krüger and Kahlenberg, 2005; Lazic et al., 2008; Krüger et al., 2009).

The high-pressure behavior of the brownmillerite series has also been investigated by several research groups. High-pressure single-crystal X-ray diffraction (XRD) studies were performed on the three compositions in the Ca₂(Fe_2−xAl_x)O₅ solid solution series: x = 0.00, 0.37 with Pnma symmetry, and x = 0.55 with I2mb symmetry (Ross et al., 2002; Vanpeteghem et al., 2008). The results revealed that no phase transition occurred up to a maximum pressure of approximately 10 GPa. Subsequently, Li et al. (2019) conducted a high-pressure synchrotron powder XRD study on brownmillerite Ca₂FeAlO₅ with I2mb symmetry up to 26.5 GPa, and first confirmed that the phase transition occurred at 25.1 GPa. However, they could not determine the crystal structure of the high-pressure phase. Recently, Zhai et al. (2022) performed high-pressure Raman spectroscopy on the srebrodolskite Ca₂Fe₂O₅ up to 21.8 GPa. They observed a reversible phase transition at approximately 14 GPa. However, the crystal structure of the high-pressure phase of srebrodolskite has not been fully clarified yet. Table 1 summarizes the several representative crystal structures of A₂B₂O₅. Oxygen-defective perovskites with the chemical formula A₂B₂O₅ are known to exhibit a variety of crystal structures. These structures can be changed not only by cationic substitutions but also by pressure and temperature.

Table 1. The crystal structure types of oxygen defective perovskite with a chemical formula of A₂B₂O₅

Chemical formula (A2B2O5)	A2+	Ionic radii*	B3+	Ionic radii*	Space group	Crystal structure type	References
Mg2Al2O5	Mg	0.89 Å	Al	0.54 Å	Pbam	Modified ludwigite-type structure	Enomoto et al. (2009)
Mg2Cr2O5	Mg	0.89 Å	Cr	0.62 Å	Pbam	Modified ludwigite-type structure	Ishii et al. (2015)
Mg2Fe2O5	Mg	0.89 Å	Fe	0.65 Å	Cmcm	CaFe3O5-type structure	Ballaran et al. (2015)
Fe2Cr2O5	Fe	0.92 Å	Cr	0.62 Å	Pbam	Modified ludwigite-type structure	Ishii et al. (2015)
Fe2Fe2O5	Fe	0.92 Å	Fe	0.65 Å	Cmcm	CaFe3O5-type structure	Woodland et al. (2012)
Ca2Al2O5	Ca	1.12 Å	Al	0.54 Å	I2mb	Brownmillerite-type structure	Kahlenberg et al. (2000)
Ca2Cr2O5	Ca	1.12 Å	Cr	0.62 Å	I2mb	Brownmillerite-type structure	Arevalo-Lopez and Attfield (2015)
Ca2Ga2O5	Ca	1.12 Å	Ga	0.62 Å	I2mb	Brownmillerite-type structure	Kahlenberg and Shaw (2001)
Ca2Fe2O5	Ca	1.12 Å	Fe	0.65 Å	Pnma	Brownmillerite-type structure	Redhammer et al. (2004)
Ca2Co2O5	Ca	1.12 Å	Co	0.74 Å	Pbcm	Brownmillerite-type structure	Zhang et al. (2014)
Sr2Ga2O5	Sr	1.26 Å	Ga	0.62 Å	Pbca	Brownmillerite-type structure	Kahlenberg et al. (2015)
Sr2Fe2O5	Sr	1.26 Å	Fe	0.65 Å	Imma	Brownmillerite-type structure	Greaves et al. (1975)
Sr2Co2O5	Sr	1.26 Å	Co	0.74 Å	Imma	Brownmillerite-type structure	Sullivan et al. (2011)

^* The ionic radii of the A²⁺ and B³⁺ cations are eight and six coordination values, respectively (Shannon, 1976).

In this study, high-pressure Raman spectroscopy and synchrotron powder XRD analyses were performed on a pure iron end-member srebrodolskite Ca₂Fe₂O₅. Herein, we report the high-pressure behavior of srebrodolskite and discuss the pressure response of the brownmillerite structure.

EXPERIMENTAL METHODS

Sample preparation

Commercially available CaCO₃ (purity >99.99%, Fujifilm Wako Pure Chemical Co., Inc., Osaka, Japan) and Fe₂O₃ (purity >99.0%, Fujifilm Wako Pure Chemical Co., Inc.) were used as the starting materials. The samples were synthesized using the solid-state reaction described by Vanpeteghem et al. (2008). First, CaCO₃ and Fe₂O₃ were mixed with ethanol in an agate mortar in a molar ratio of 2:1. The mixture was then transferred to a platinum crucible and heated at 1000 °C for 17 h in a resistance furnace. The sample was then cooled to 25 °C in the furnace with the cooling rate of 10 °C/min. The resulting powder was ground and pressed into pellets. They were introduced into the furnace and sintered at 1200 °C for 20 h. After cooling to 25 °C with the cooling rate of 10 °C/min, the pellets were thoroughly ground and pressed again. The new pellets were placed in a furnace and sintered at 1200 °C for 50 h. This procedure was repeated three times. Finally, single black crystals with metallic luster were obtained. The products were characterized using a high-resolution CuKα X-ray powder diffractometer (D8 Advance, Bruker AXS Inc., Karlsruhe, Germany) equipped with the Bruker LynxEye detector used to minimize the Fe fluorescence background excited by the Cu radiation. It revealed that the product consisted only of srebrodolskite single crystals ranging in size from 10 to 20 µm.

High-pressure experiments

High-pressure experiments were performed at 25 °C using a symmetric-type diamond anvil cell (DAC) with a 300-µm culet size. A 200-µm-thick steel gasket was pre-indented to a thickness of 80 µm and holed for a sample chamber with a diameter of 150 µm. The sample was loaded into the sample chamber with a few ruby chips of approximately 2 µm in size. A 16:3:1 methanol:ethanol:water mixture was used as the pressure-transmitting medium. The pressure was determined before and after each measurement at five points in the sample chamber using the pressure-induced shift of the R₁ ruby luminescence line (Mao et al., 1986). As shown later in the result section, XRD patterns showed the broadening of the diffraction peaks above 10 GPa, indicating that the pressure-transmitting medium subjected non-hydrostatic stress to the sample. In the study, therefore, non-hydrostatic stress was applied to the sample above 10 GPa.

Raman spectroscopy

High-pressure Raman spectroscopy was performed using a micro-Raman spectrometer (NRS-5100, Japan Spectroscopic Co., Tokyo, Japan) equipped with a grating of 1800 lines/mm and a high-sensitivity cooled CCD detector. Raman spectra were collected in backscattering geometry using a green laser operating at 532.12 nm. The incident laser power was maintained below 5.4 mW. A 20× objective lens was used to focus the laser beam onto a spot of approximately 10 µm. The Raman spectra were obtained with an integration time of 2 min, and five spectra were averaged. The resolution of the spectra was 2.1 cm⁻¹. The spectrometer was calibrated using the auto calibration tool with the emission lines from a neon lamp. The laser beam produced no visible damage to the sample surface during the measurements. Least-squares peak-fitting software, Peak-Fit (AISN Software Inc., Chicago, USA), was used to analyze the Raman spectral data. Data smoothing and baseline fitting were conducted using the Savitzky-Golay method and the hyperbolic model, respectively. The band positions were determined by fitting the Raman peaks to a Lorentzian peak shape function. In this study, the determination of whether a Raman peak originated from the sample was based on the reliability of the peak-fitted peak position and full width at half maximum (FWHM) value.

Synchrotron powder XRD measurement

First, single crystals of srebrodolskite were ground by adding ethanol to an agate mortar. The resulting powder was dispersed in a beaker of ethanol. Large particles were precipitated in ethanol. The suspension was removed and dried. The fine powder obtained was inserted into the sample chamber of the DAC. High-pressure synchrotron powder XRD measurements were performed using a BL18C beamline at the Photon Factory (PF), High-Energy Accelerator Research Organization (KEK), Japan. The monochromatic incident X-ray beam was collimated to 100 µm in diameter. The wavelength used was determined as λ = 0.6179 (1) Å using the Si standard (NIST SRM 640e). An imaging plate (IP) was used as the X-ray detector. The typical exposure time was 30 min. Two-dimensional XRD images were integrated as a function of 2θ to provide conventional one-dimensional profiles using the programs IPAnalyzer and PDIndexer (Seto et al., 2010). Lattice parameters were determined by least-squares fitting of the obtained XRD data using the PDIndexer. A third-order Birch-Murnaghan equation of state was determined from the experimental pressure-volume data using EosFit7-GUI program (Gonzalez-Platas et al., 2016). The XRD data at 12.6 GPa was analyzed by the Rietveld method using GSAS software (Larson and Von Dreele, 2004). The structural parameters were refined by the least-squares methods using the diffraction data in the 2θ range from 1.8 to 22°. The background was modeled with 36 terms using the Chebyschev polynomial function. The peak shape was modeled using the pseudo-Voigt function described by Howard (1982) and Thompson et al. (1987). The peak cut-off was set at 0.1% of the peak maximum. All atoms were refined with individual isotropic displacement parameters.

RESULTS AND DISCUSSION

High-pressure Raman spectroscopy

The result of the high-pressure Raman spectroscopy of srebrodolskite up to 23 GPa was given in Figure 2 and Supplementary Table S1 (Tables S1 and S2 are available online from https://doi.org/10.2465/jmps.230422). The Raman spectra at low pressure were almost consistent with the previous study (Zhai et al., 2022), but the only difference was that the main Raman band observed at 700 cm⁻¹ showed an asymmetrical spectral shape with a shoulder at the low frequency side (Fig. 2a). With increasing pressure up to 12.0 GPa, a new Raman band appeared at 645 cm⁻¹. With further increasing pressure to 16.0 GPa, an additional new Raman band appeared at 513 cm⁻¹. Since this Raman band was also observed by Zhai et al. (2022), it cannot be considered that it is derived from pressure medium. The phase-transition pressure was approximately consistent with Zhai’s results. Pressure dependence of Raman band positions for the observed each vibrational mode is presented in Figure 2b. The Raman bands observed before the phase transition showed pressure dependencies similar to those exhibited by Zhai et al. (2022); however, the intensities of the new Raman bands emerging after the phase transition became much stronger than those observed by Zhai et al. (2022). Although three Raman bands derived from the high-pressure phase were observed by Zhai et al. (2022), at a pressure of 23.0 GPa, at least nine Raman bands were observable in the present study (Fig. 2).

Figure 2. (a) Evolution of Raman spectra of srebrodolskite during the compression process. The intensity is normalized in each profile. (b) The pressure dependence of Raman band positions for the observed vibrational modes.

High-pressure XRD analysis

Figure 3 shows the evolutions of the high-pressure XRD pattern of srebrodolskite during compression and decompression processes. These results indicate that srebrodolskite underwent a reversible phase transition from srebrodolskite with Pnma structure to its high-pressure phase. All reflections of srebrodolskite at 1.7 GPa can be indexed to a brownmillerite-type structure with space group Pnma. We assumed that the srebrodolskite structure survived at higher pressure. When all high-pressure XRD patterns were indexed to the brownmillerite-type structure with space group Pnma, the obtained lattice parameters are given in Table S2. The resulting P-V compression curve is shown in Figure 3c. A third-order Birch-Murnaghan equation of state fitted to the P-V data between 1.0 and 8.0 GPa led to bulk modulus K₀ = 127.8 (9.7) GPa and its pressure derivative K’₀ = 5.2 (2.3). From high-pressure single-crystal XRD measurements of srebrodolskite, the fit of the P-V data yielded bulk modulus K₀ = 127.0 (5) GPa and its pressure derivative K’₀ = 5.99 (13) (Ross et al., 2002). Below 8.0 GPa, our bulk modulus and its pressure derivative were exactly consistent with previous reports (Ross et al., 2002), but they deviated from the equation-of-state determined by Ross et al. (2002) above 9.1 GPa. This suggests that srebrodolskite began to transform into its high-pressure phase at 9.1 GPa. The structural changes during a phase transition are typically determined by the symmetry relationships between the maximal non-isomorphic subgroup and its supergroup. The space group Pnma has seven non-isomorphic subgroups: Pn2₁a (No. 33), Pnm2₁ (No. 31), P2₁ma (No. 26), P2₁2₁2₁ (No. 19), P112₁/a (No. 14), P2₁/n11 (No. 14), and P12₁/m1 (No. 11). We fitted the XRD data at 12.6 GPa to these candidates by the Rietveld technique, because it seemed to be relatively stable after the phase transition. As a result, a structure model of the space group Pn2₁a yielded the most reliable structural parameters R_p = 1.46% and wR_p = 2.01% in the least-squares fitting of the Rietveld refinement (Table 2 and Fig. 4). The refined atomic coordinates and selected bond lengths for the high-pressure phase of srebrodolskite are given in Tables 3 and 4, respectively. The high-pressure phase transition of srebrodolskite accompanied by a decrease in symmetry from space group Pnma to Pn2₁a occurs at 9.1 GPa, and continued up to 12.6 GPa.

Figure 3. (a) Evolution of the synchrotron powder X-ray diffraction pattern of srebrodolskite during (a) the compression and (b) decompression processes. The intensity is normalized in each profile. The new peaks are marked by inverse triangles. (c) Pressure-volume compression profile of srebrodolskite at a pressure range of 0-20 GPa. The red solid line shows the third-order Birch-Murnaghan equation-of-state fitted to the P-V data between 1.0 and 8.0 GPa, and the black dashed line represents that determined by Ross et al. (2002).

Table 2. Crystal data and refinement results for the high-pressure phase of srebrodolskite at 12.6 GPa

Crystal formula	Ca2Fe2O5
Chemical formula weight	271.85
Crystal system	Orthorhombic
Space group	Pn21a
Unit cell parameter a (Å)	5.2289(4)
b (Å)	14.2548(9)
c (Å)	5.4291(4)
Unit cell volume V (Å3)	404.67(2)
Z	4
Rp	0.0146
wRp	0.0201

Figure 4. Rietveld refinement of X-ray diffraction pattern at 12.6 GPa fitted with space group Pn2₁a converging to agreement factors R_p = 1.46% and wR_p = 2.01%.

Table 3. Atomic coordinates and isotropic atomic displacement parameters for the high-pressure phase of srebrodolskite at 12.6 GPa

Atom	Wycoff position	x	y	z	Uiso
Ca1	4a	0.488(8)	0.110(3)	0.015(5)	0.01
Ca2	4a	0.526(4)	0.892(2)	0.962(4)	0.01
Fe1	4a	0.008(7)	−0.004(2)	−0.007(7)	0.01
Fe2	4a	0.955(2)	0.2401(12)	0.9443(18)	0.01
O11	4a	0.231(2)	0.998(2)	0.242(9)	0.03
O12	4a	0.736(2)	0.0312(19)	0.745(2)	0.03
O21	4a	0.110(8)	0.143(4)	0.044(2)	0.03
O22	4a	1.033(2)	0.856(5)	0.915(2)	0.03
O31	4a	0.628(6)	0.247(5)	0.865(6)	0.03

Table 4. Selected bond lengthes (Å) for the high-pressure phase of srebrodolskite at 12.6 GPa

This study (12.6 GPa)						Srebrodolskite (Redhammer et al., 2004)
Ca1-	O21	2.04(7)	Ca2-	O22	2.11(7)	Ca-	O2	2.3226
	O12	2.23(10)		O31	2.41(7)		O3	2.3452
	O31	2.24(8)		O11	2.45(7)		O1	2.4300
	O12	2.26(9)		O12	2.55(7)		O1	2.4820
	O11	2.43(9)		O11	2.64(8)		O1	2.4844
	O11	2.42(9)		O22	2.64(11)		O2	2.5413
	O21	2.52(8)		O22	2.71(11)		O1	2.7358
	O21	3.14(8)		O12	2.74(8)		O2	3.0000
Average		2.41	Average		2.53	Average		2.543
Fe1-	O11	1.79(10)				FeM-	O1 (× 2)	1.9614
	O12	1.83(10)					O1 (× 2)	1.9691
	O12	2.02(10)					O2 (× 2)	2.1201
	O11	2.04(10)				Average		2.017
	O22	2.05(8)
	O21	2.18(7)
Average		1.99
Fe2-	O21	1.69(6)				FeT-	O2 (× 2)	1.8430
	O31	1.77(4)					O3	1.9127
	O22	1.82(8)					O3	1.9140
	O31	1.91(4)				Average		1.878
Average		1.80

At 14.1 GPa, new reflections emerged at 2θ of 9.0° and 18.2 (Fig. 3a). It has been well known that the brownmillerite-type structure, exhibiting Pnma symmetry, undergoes a structure change to the I2mb space group through the ionic substitution of large ions with smaller ions [e.g., Ca₂(Fe,Al)₂O₅, Redhammer et al., 2004; Ca₂(Fe,Ga)₂O₅, Stahl et al., 2019; Ca₂Mn(Ga,Al)O₅, Abakumov et al., 2005]. If the unit cell volume of srebrodolskite Ca₂Fe₂O₅ contracts with pressure, it could induce a structural change resembling the effect of substituting larger ions with smaller ions in the brownmillerite-type structure. Since the space group Pn2₁a belongs to the non-isomorphic subgroups of not only space group Pnma but I2mb, under compression, it is likely that the brownmillerite-type structure with Pnma symmetry undergoes a phase transition into I2mb symmetry through the intermediate Pn2₁a symmetry. The disappearance of the Bragg reflections for 131 and 151 (h + k + l = 2n + 1) can be used to identify the phase transition from Pnma to I2mb symmetry in the brownmillerite-type structure. With increasing pressure, the intensities of 131 and 151 reflections steadily decreased and eventually almost disappeared at 20.2 GPa (Fig. 3a). In this case, extra reflections that cannot be indexed based on an orthorhombic body-centered lattice may be identified as incommensurate satellite reflections. In the study, however, no satellite reflections clearly emerged. It is important to note that the Bragg reflections, except for h + k + l = 2n + 1, must remain even if the srebrodolskite transforms from Pnma to I2mb. However, as shown in Figure 3a, most of all reflections except three reflections appearing at approximately 2θ of 9.0, 13.6, and 18.2° tended to weaken and disappear with increasing pressure (Fig. 3a). Therefore, it cannot be concluded that above 14.1 GPa the high-pressure structure of srebrodolskite is not an I2mb symmetry. Indeed, it may be improbable that the alternating L and R tetrahedral layer arrangement transforms into a single type of layer arrangement solely with compression. Therefore, srebrodolskite may transform into other brownmillerite-type structures or their derivatives. In the present study, XRD data indicated the phase transition pressure at 9.1 GPa, but high-pressure Raman spectroscopic study suggested the phase transition pressure at 12.0 GPa. Concerning the difference in the phase transition pressure between XRD measurement and Raman spectroscopic analysis, the XRD is derived from the diffraction phenomenon of interatomic spacings of a crystal, whereas the Raman spectroscopy is ascribed to bond vibrations between constituent atoms. Since each observation originates from a different physical phenomenon, they do not necessarily show exactly the same phase transition pressure.

Figure 5 shows the diversity of the crystal structure of the oxygen-defective perovskite with the chemical formula A₂B₂O₅ as the ionic radius changes. A high-pressure phase transition leads to a structure characterized by higher r_c/r_a ratios due to the well-established higher compressibility of anions compared to cations (Fukunaga and Yamaoka, 1979). Consequently, it is expected that the high-pressure phase will be located towards the right and upper region of the diagram (Fig. 5). This implies that srebrodolskite may transform to Imma structure with pressure. We examined whether the high-pressure structure of srebrodolskite can be fitted with the Imma structure, but it could not be fitted to the XRD pattern of the high-pressure phase of srebrodolskite. Other structures depicted in the diagram (Fig. 5) were also examined to fit to the XRD pattern, but no satisfactory results were obtained. Li et al. (2019) investigated the structural changes in brownmillerite Ca₂FeAlO₅ with I2mb symmetry under high pressure. A high-pressure phase transition of brownmillerite was observed at 25.1 GPa, but the structure of the high-pressure phase of brownmillerite has not yet been determined. Brownmillerite Ca₂FeAlO₅ may also undergo a high-pressure phase transition to a different structure from that of known brownmillerite-type structures.

Figure 5. The diversity of the crystal structure of A₂B₂O₅ as the ionic radius changes.

ACKNOWLEDGMENTS

We thank two anonymous reviewers for their constructive comments and suggestions, which led to significant improvement in this manuscript. The high-pressure synchrotron powder XRD measurements were performed with the approval of the Photon Factory Program Advisory Committee (Proposal Nos. 2017G120, 2019G026, and 2020G064). This study was partially supported by JSPS KAKENHI Grant Number JP20K04124.

SUPPLEMENTARY MATERIALS

Supplementary Tables S1 and S2 are available online from https://doi.org/10.2465/jmps.230422.

REFERENCES

•  Abakumov,  A.M.,  Kalyuzhnaya,  A.S.,  Rozova,  M.G.,  Antipov,  E.V., et al (2005) Compositionally induced phase transition in the Ca₂MnGa_1−xAl_xO₅ solid solutions: Ordering of tetrahedral chains in brownmillerite structure. Solid State Sciences, 7, 801-811.
•  Aggarwal,  P.S.,  Gard,  J.A.,  Glasser,  F.P. and  Biggar,  G.M. (1972) Synthesis and properties of dicalcium aluminate, 2CaO·Al₂O₃. Cement and Concrete Research, 2, 291-297.
•  Antipov,  E.V.,  Abakumov,  A.M. and  Istomin,  S.Y. (2008) Target-aimed synthesis of anion-deficient perovskites. Inorganic Chemistry, 47, 8543-8552.
•  Arevalo-Lopez,  A.M. and  Attfield,  J.P. (2015) Crystal and magnetic structures of the brownmillerite Ca₂Cr₂O₅. Dalton Transactions, 44, 10661-10664.
•  Ballaran,  T.B.,  Uenver-Thiele,  L. and  Woodland,  A.B. (2015) Complete substitution of Fe²⁺ by Mg in Fe₄O₅: The crystal structure of the Mg₂Fe₂O₅ end-member. American Mineralogist, 100, 628-632.
•  Catalli,  K.,  Shim,  S.H.,  Prakapenka,  V.B.,  Zhao,  J.Y., et al (2010) Spin state of ferric iron in MgSiO₃ perovskite and its effect on elastic properties. Earth and Planetary Science Letters, 289, 68-75.
•  Chesnokov,  B.V. and  Bazhenova,  L.F. (1985) Srebrodolskite Ca₂Fe₂O₅ - A new mineral. Zapiski Vsesoyuznogo Mineralogicheskogo Obshchestva, 114, 195-199.
•  Colville,  A.A. and  Geller,  S. (1971) The crystal structure of brownmillerite, Ca₂FeAlO₅. Acta Crystallographica, B27, 2311-2315.
•  Enomoto,  A.,  Kojitani,  H.,  Akaogi,  M.,  Miura,  H. and  Yusa,  H. (2009) High-pressure transitions in MgAl₂O₄ and a new high-pressure phase of Mg₂Al₂O₅. Journal of Solid State Chemistry, 182, 389-395.
•  Fukunaga,  O. and  Yamaoka,  S. (1979) Phase transformations in ABO₄ type compounds under high pressure. Physics and Chemistry of Minerals, 5, 167-177.
•  Gonzalez-Platas,  J.,  Alvaro,  M.,  Nestola,  F. and  Angel,  R. (2016) EosFit7-GUI: a new graphical user interface for equation of state calculations, analyses and teaching. Journal of Applied Crystallography, 49, 1377-1382.
•  Greaves,  C.,  Jacobson,  A.J.,  Tofield,  B.C. and  Fender,  B.E.F. (1975) A powder neutron diffraction investigation of the nuclear and magnetic structure of Sr₂Fe₂O₅. Acta Crystallographica, B31, 641-646.
•  Hentschel,  G. (1964) Mayenit, 12CaO·7Al₂O₃, und brownmillerit, 2CaO·(Al,Fe)₂O₃, zwei neue minerale in den kalksteinschlüssen der lava des Ettringer Bellerberges. Neues Jahrbuch für Mineralogie, Monatshefte, 1964, 22-29.
•  Howard,  C.J. (1982) The approximation of asymmetric neutron powder diffraction peaks by sums of Gaussians. Journal of Applied Crystallography, 15, 615-620.
•  Irifune,  T. (1994) Absence of an aluminous phase in the upper part of the Earth’s lower mantle. Nature, 370, 131-133.
•  Ishii,  T.,  Kojitani,  H.,  Fujino,  K.,  Yusa,  H., et al (2015) High-pressure high-temperature transitions in MgCr₂O₄ and crystal structures of new Mg₂Cr₂O₅ and post-spinel MgCr₂O₄ phases with implications for ultrahigh-pressure chromitites in ophiolites. American Mineralogist, 100, 59-65.
•  Jackson,  I. and  Rigden,  S.M. (1996) Analysis of P-V-T data: Constraints on the thermoelastic properties of high-pressure minerals. Physics of the Earth and Planetary Interiors, 96, 85-112.
•  Kahlenberg,  V.,  Fischer,  R.X. and  Shaw,  C.S.J. (2000) Rietveld analysis of dicalcium aluminate (Ca₂Al₂O₅)—A new high pressure phase with the brownmillerite type structure. American Mineralogist, 85, 1061-1065.
•  Kahlenberg,  V. and  Shaw,  C.S.J. (2001) Ca₂Ga₂O₅: a new high pressure oxogallate. Zeitschrift für Kristallographie, 216, 206-209.
•  Kahlenberg,  V.,  Goettgens,  V.,  Mair,  P. and  Schmidmair,  D. (2015) High-pressure synthesis and crystal structures of the strontium oxogallates Sr₂Ga₂O₅ and Sr₅Ga₆O₁₄. Journal of Solid State Chemistry, 228, 27-35.
•  Kojitani,  H.,  Katsura,  T. and  Akaogi,  M. (2007) Aluminum substitution mechanisms in perovskite-type MgSiO₃: an investigation by Rietveld analysis. Physics and Chemistry of Minerals, 34, 257-267.
•  Krüger,  H. and  Kahlenberg,  V. (2005) Incommensurately modulated ordering of tetrahedral chains in Ca₂Fe₂O₅ at elevated temperatures. Acta Crystallographica, B61, 656-662.
•  Krüger,  H.,  Kahlenberg,  V.,  Petříček,  V.,  Phillipp,  F. and  Wertl,  W. (2009) High-temperature structural phase transition in Ca₂Fe₂O₅ studied by in-situ X-ray diffraction and transmission electron microscopy. Journal of Solid State Chemistry, 182, 1515-1523.
• Larson, A.C. and Von Dreele, R. (2004) General Structure Analysis System (GSAS). Los Alamos National Laboratory Report, LAUR 86-748.
•  Lazic,  B.,  Kruger,  H.,  Kahlenberg,  V.,  Konzett,  J. and  Kaindl,  R. (2008) Incommensurate structure of Ca₂Al₂O₅ at high temperatures - structure investigation and Raman spectroscopy. Acta Crystallographica, B64, 417-425.
•  Li,  Z.M.,  Yin,  Y.,  Rumney,  J.D.,  Shieh,  S.R., et al (2019) High-pressure in-situ X-ray diffraction and Raman spectroscopy of Ca₂AlFeO₅ brownmillerite. High Pressure Research, 39, 92-105.
•  Mao,  H.K.,  Xu,  J. and  Bell,  P.M. (1986) Calibration of the ruby pressure gauge to 800 kbar under quasi-hydrostatic conditions. Journal of Geophysical Research, 91, 4673-4676.
•  Mattern,  E.,  Matas,  J.,  Ricard,  Y. and  Bass,  J. (2005) Lower mantle composition and temperature from mineral physics and thermodynamic modelling. Geophysical Journal International, 160, 973-990.
•  Milat,  O.,  Krekels,  T.,  Van Tendeloo,  G. and  Amelinckx,  S. (1993) Ordering principles for tetrahedral chains in Ga- and Co- substituted YBCO intergrowths. Journal de Physique I, 3, 1219-1234.
•  Navrotsky,  A.,  Schoenitz,  M.,  Kojitani,  H.,  Xu,  H.W., et al (2003) Aluminum in magnesium silicate perovskite: Formation, structure, and energetics of magnesium-rich defect solid solutions. Journal of Geophysical Research: Solid Earth, 108, 2330.
•  Redhammer,  G.J.,  Tippelt,  G.,  Roth,  G. and  Amthauer,  G. (2004) Structural variations in the brownmillerite series Ca₂(Fe_2−xAl_x)O₅: Single-crystal X-ray diffraction at 25 °C and high-temperature X-ray powder diffraction (25 °C ≤ T ≤ 1000 °C). American Mineralogist, 89, 405-420.
•  Ross,  N.L.,  Angel,  R.J. and  Seifert,  F. (2002) Compressibility of brownmillerite (Ca₂Fe₂O₅): effect of vacancies on the elastic properties of perovskites. Physics of the Earth and Planetary Interiors, 129, 145-151.
•  Seto,  Y.,  Nishio-Hamane,  D.,  Nagai,  T. and  Sata,  N. (2010) Development of a software suite on X-ray diffraction experiments. Review of High Pressure Science and Technology, 20, 269-276.
•  Shannon,  R.D. (1976) Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallographica, A32, 751-767.
•  Stahl,  Q.E.,  Redhammer,  G.J.,  Tippelt,  G. and  Reyer,  A. (2019) Structural and spectroscopic characterization of the brownmillerite-type Ca₂Fe_2−xGa_xO₅ solid solution series. Physics and Chemistry of Minerals, 46, 271-298.
•  Sullivan,  E.,  Hadermann,  J. and  Greaves,  C. (2011) Crystallographic and magnetic characterisation of the brownmillerite Sr₂Co₂O₅. Journal of Solid State Chemistry, 184, 649-654.
•  Thompson,  P.,  Cox,  D.E. and  Hastings,  J.B. (1987) Rietveld refinement of Debye-Scherrer synchrotron X-ray data from Al₂O₃. Journal of Applied Crystallography, 20, 79-83.
•  Tsuchiya,  T. and  Tsuchiya,  J. (2006) Effect of impurity on the elasticity of perovskite and postperovskite: Velocity contrast across the postperovskite transition in (Mg,Fe,Al)(Si,Al)O₃. Geophysical Research Letters, 33, L12S04.
•  Vanpeteghem,  C.B.,  Angel,  R.J.,  Zhao,  J.,  Ross,  N.L., et al (2008) The effect of oxygen vacancies and aluminium substitution on the high-pressure properties of brownmillerite-structured Ca₂F_e2−xAl_xO₅. Physics and Chemistry of Minerals, 35, 493-504.
•  Wang,  X.L.,  Tsuchiya,  T. and  Hase,  A. (2015) Computational support for a pyrolitic lower mantle containing ferric iron. Nature Geoscience, 8, 556-559.
•  Woodland,  A.B.,  Frost,  D.J.,  Trots,  D.M.,  Klimm,  K. and  Mezouar,  M. (2012) In situ observation of the breakdown of magnetite (Fe₃O₄) to Fe₄O₅ and hematite at high pressures and temperatures. American Mineralogist, 97, 808-1811.
•  Zhai,  S.M.,  Dai,  B.,  Xue,  W.H.,  Rumney,  J.D., et al (2022) Pressure- and temperature-dependent Raman spectra of Ca₂Fe₂O₅ oxygen defect perovskite. Spectrochimica Acta, Part A: Molecular and Biomolecular Spectroscopy, 279, 121436.
•  Zhang,  J.J.,  Zheng,  H.,  Malliakas,  C.D.,  Allred,  J.M., et al (2014) Brownmillerite Ca₂Co₂O₅: synthesis, stability, and re-entrant single crystal to single crystal structural transitions. Chemistry of Materials, 26, 7172-7182.