Synthesis of Barium Ferrates in Oxygen

Abstract

Phase relations in the system BaO—Fe₂O₃—”FeO₂” at 1 atm (1 atm = 1.013 × 10⁵ N/m²) oxygen and between 800 to 1050 °C are shown in the vicinity (high Ba portion) of the 1:1 cation-cation ratio composition. A hexagonal BaTiO₃-like phase, BaFeO_3−X (a = 5.676Å, c = 13.934Å), exists below 960 °C. A perovskite-like phase BaFeO_3−X′ occurs between 960 and 1050 °C. Single-phase tetragonal perovskite can be obtained at room temperature by quenching compositions with Ba-Fe ratios of 67:66 (a = 3.985 Å, c = 4.005 Å, 1000 °C) and 27:26 (a = 3.988 Å, c = 4.003 Å, 1000 °C) from above 960 °C. At the 17:16 ratio, single phase cubic perovskite (a = 3.994 Å, 1000 °C), can be synthesized. Influence of temperature, additions of barium, and use of Pt-containers on phase relations are discussed.

1. Introduction

Mixed alkaline earth-transition metal oxides are of particular importance because of their interesting electrical and magnetic properties. Characterization of such materials through phase equilibrium and x-ray diffraction studies is complicated by the tendency of the transition metal to exist in disproportioned oxidation states within a single phase. Variable oxygen stoichiometries, controlled by temperature and oxygen pressure result. The 1:1 cation-cation barium ferrates are excellent examples of this phenomenon. Tetravalent iron has been confirmed in phases within the BaFeO_2.5−3.0 compositional range by MacChesney et al. [1]¹, and Gallagher et al. [2], using magnetic studies and the Mossbauer technique. Some inconsistencies, however, exist concerning the nature of the phases within this compositional range. Erchak et al. [3], reported the compound ${\text{BaFeO}}_{2.62}\left({\text{Ba}}_8{\text{Fe}}_2^{4+}{\text{Fe}}_6^{3+}{\text{O}}_{21}\right)$ which was thought to be a cubic perovskite (a = 8.05 Å). Subsequently, Malinofsky and Kedesdy [4] showed that this phase is not cubic, but is actually analogous to the hexagonal polymorph of BaTiO₃ [5] with a = 5.68 Å and c = 13.86 Å. They also reported a tetragonal perovskite polymorph, admixed with the hexagonal form, with a = 3.98 Å and c = 4.01 Å. This mixture was prepared by heating the hexagonal form at 1275 °C in oxygen and slow-cooling. MacChesney, et al. [1], and Gallagher, et al. [2], prepared hexagonal forms in the BaFeO_2.74−2.95 range using oxygen pressures from 0.2 to 2400 atm and generally below 850 °C. Gushee et al. [6], reported a BaFeO_2.72 hexagonal form with a = 5.76 Å and c = 13.9 Å while Goto and Takada [7] synthesized a 2BaO·Fe₂O₃ phase with a melting point at 1370 °C in 1 atm O₂. The latter phase was reported as cubic (a = 8.07 Å) but is most likely a hexagonal barium ferrate similar to that reported by Erchak et al. [3], and originally thought to be a cubic perovskite with a′ = 2a_cub.

Mori [8, 9] reported a number of BaFeO_3−X phases with X values ranging from 0.50 to 0.10. These were prepared in air and 1 atm oxygen using quenching and slow-cooling techniques within the 25 to 1080 °C range. A hexagonal BaTiO₃-like phase was found by quenching from below 950 °C in oxygen. Above this temperature a transitory cubic form was obtained after a $\frac{1}{2}\text{-\hspace{0.17em}hr}$ heating time. With longer heating times this form converted to a triclinic form through a rhombohedral form. Heating a 2BaCO₃:Fe₂O₃ mixture at 1080 °C in oxygen for $\frac{1}{2}\text{-\hspace{0.17em}hr}$ and quenching also produced a cubic form, BaFeO_2.75, which upon further heating converted to a tetragonal phase, BaFeO_2.81, through a series of tetragonal phases with lower oxygen contents. The BaFeO_2.81 phase was not of the simple tetragonal perovskite type as its powder pattern gave evidence of superstructure lines. Final equilibration products were not indicated and difficulty was encountered in reproducing each phase, in terms of symmetry and oxygen content, by the quenching method.

Although the existence of the hexagonal form is well documented, synthesis of a stable single-phase tetragonal perovskite or a stable, truly cubic perovskite has not been demonstrated. This study was conducted to examine the latter possibility and, therefore, considers only the existence and stabilities of phases encountered. Chemical analyses establishing Fe⁴⁺:Fe³⁺ ratios (oxygen concentration) and Mössbauer data will be reported in the future.

2. Experimental Procedure

Starting material was prepared from appropriately weighed (±0.1 mg) quantities of BaCO₃ and Fe₂O₃. These were hand mixed under acetone, packed in gold envelopes, and calcined in air at 800 °C for two weeks with periodic remixing and reheating. Gold was used as the container instead of platinum because of the tendency of uncombined BaCO₃ to react with the latter. Another problem with platinum will be discussed later. Calcined specimens were equilibrated at elevated temperatures in 1 atm (1 atm = 1.013 × 10⁵ N/m²) oxygen using the quench method. A vertical tube, resistance type, furnace, modified for vacuum and gas flow capability with end-closures and “O”-ring seals, was used for the quenching experiments. Oxygen was passed through the furnace at 10⁻⁶ m³/s flow rate. Open gold envelopes were used as containers to 1050 °C. A few experiments were conducted above 1050 °C using platinum. Temperatures were measured with Pt-Pt 10 percent Rh thermocouples calibrated against the melting points of NaCl (801 °C) and Au (1063 °C). The furnace was controlled by an a-c Wheatstone bridge controller capable of maintaining temperature to at least ±2 °C. Specimens were quenched into an ice-cooled brass cold finger which constitutes the lower portion of the furnace. X-ray patterns were made using a high-angle Geiger counter-diffractometer and nickel-filtered Cu radiation at a 1/4° 2θ/min scanning rate. Reported unit cell data are considered accurate to about ±2 in the last decimal place listed.

3. Experimental Results

A portion of the BaO-”iron oxide” pseudobinary in the vicinity (high baria portion) of the 1:1 cation-cation compound is shown in figure 1. The diagram was constructed from the data in table 1. It is emphasized that the diagram represents stability of phases present (determined by room temperature x-ray patterns) as a function of starting Ba:Fe ratio. Absolute compositions involving Ba:Fe⁴⁺:Fe³⁺ ratios belonging to the BaO—Fe2O₃—”FeO₂” ternary are not indicated.

Figure 1. Subsolidus relations at 1 atm O₂ for that portion of the BaO-”iron oxide” pseudobinary near the 1:1 barium-iron ratio.

Absolute Ba-Fe-O concentrations are not indicated.

Table 1.

Experimental data for compositions in the system BaO-”iron oxide” at 1 atm oxygen

Composition	Heal treatments				X-ray diffraction analysesa
	Initialb		Finalc
	Temp.	Time	Temp.	Time
Ba:Fe ratio	°C	hr	°C	hr
1:1	800	335
			800	70	Hex-BaFeO3-X	a = 5.676
						c= 13.934
			910	70	Hex-BaFeO3-X	a = 5.676
						c= 13.934
			935	45	Hex-BaFeO3-X	a = 5.676
						r= 13.934
			950	70	Hex-BaFeO3-X	a = 5.676
						c= 13.934
			965	45	Hex-BaFeO3-X + BaFeO3-X’d
			975	70	BaFeO3-X’
			1000	70	BaFeO3-X’
			1050	70	BaFeO3-X’
67:66	800	335
			910	70	Hex-BaFeO3-Xe
			935	45	Hex-BaFeO3-Xe
			950	70	Hex-BaFeO3-X + perovskite
			965	45	t. p.f
			975	70	t. p.	a = 3.984
						c = 4.005
			1000	70	t. p.	a = 3.985
						c = 4.005
			1050	70	t. p.	a = 3.987
						c = 4.004
27:26	800	335
			975	70	t. p.	a = 3.986
						c = 4.003
			1000	70	t. p.	a = 3.988
						c = 4.003
			1050	70	t. p.	a = 3.992
						c = 4.003
17:16	800	335
			910	70	Hex-BaFeO3-X + Ba2FeO4
			935	45	Hex-BaKeO3-X + Ba2FeO4
			950	70	c. p.g + Hex-BaFeO3-X
			965	45	c. p.
			975	70	c. p.	a = 3.993
			1000	70	c. p.	a = 3.994
			1050	70	c. p.	a = 3.997
69:62	800	335
			800	70	Hex-BaFeO3-X + Ba2FeO4
			910	70	Hex-BaFeO3-X + Ba2FeO4
			935	45	Hex-BaFeO3-X + Ba2FeQ4
			950	70	c. p. + Ba2FeO4 + Hex-BaFeO3-Xh
			975	70	c. p. (a = 3.992) + Ba2FeO4
			1000	70	c. p. (a = 3.993) + Ba2FeO4
			1050	70	c. p. (a = 3.996) + unknown phase “X”
7:6	800	335
			910	70	Hex-BaFeO3-x + Ba2FeO4
			935	45	Hex-BaFeO3-X + Ba2FeO4
			950	70	c. P. + Ba2FeO4 + Hex-BaFeO4
			975	70	c. p.+ Ba2FeO4
			1000	70	c. p. + Ba2FeO4
			1050	70	c. p. + unknown phase “X”
71:58	800	335
			910	70	Hex-BaFeO3-X + Ba2FeO4
			950	70	c. p. + Ba2FeO4
			975	70	c. p. + Ba2FeO4
			1000	70	c. p. + Ba2FeO4
			1050	70	c. p. + unknown phase “X”

^aPhases identified are given in the order of amount present at room temperature. The phases are not necessarily those present at the temperature to which the specimen was heated.

^bHeated in air in Au.

^cHeated in 1 atm oxygen in Au.

^dPerovskite-like phase.

^eBa₂FeO₄ in insufficient amount to detect.

^ft. p. = tetragonal perovskite.

^gc. p. = cubic perovskite.

^hIncomplete reaction.

The l:l hexagonal BaTiO₃-like phase, BaFeO_3−X (a = 5.676 Å, c = 13.934 Å), exists at low temperatures as previously observed by others. An indexed powder pattern is provided in table 2. MacChesney et al. [1], reported the hexagonal form, BaFeO_2.82 and BaFeO_2.84, at low temperatures in 1 atm O₂ and air, respectively. The value of X for the BaFeO_3−X in this study is, therefore, probably within the 0.20−0.15 range. Above approximately 960° C the phase transforms to a perovskite-like form, probably by further reduction of part of the available Fe⁴⁺ to Fe³⁺. The powder pattern, given in table 3, bears a superficial resemblance to a tetragonal perovskite but is rather diffuse and contains a number of extra lines. This phase may be represented as BaFeO_3−X′, where 0.5 > X′ > 0 and X′ > X_hex, where X_hex refers to the oxygen deficiency in the hexagonal form. It is not Ba₂Fe₂O₅, (BaFeO_2.5, X′ = 0.5), a brownmillerite-like compound also reported by MacChesney et al. [1]. Within this temperature range, however, a Ba2Fe₂O₅-like phase was found to exist in air. Although the transformation near 960 °C in this study correlates well with that observed by Mori [9] (> 950 °C), a definitive correlation of phases could not be made. It seems very likely, however, that Malinofsky and Kedesdy [4] observed the BaFeO_3−X′ of this study or one of the tetragonal phases reported by Mori when they heated and slow-cooled the hexagonal form from 1275 °C in oxygen.

Table 2.

X-ray diffraction powder data for the hexagonal BaTiO₃-like phase, BaFeO_3−X^a

dobsb	I/I0c	$\frac{1}{d_{\text{ calc }}^2}$	$\frac{1}{d_{\text{obs}}^2}$	hkld
4.017	9	0.0620	0.0620	102
3.376	23	.0877	.0878	103
2.838	100	.1238	.1242	104
		.1242		110
2.421	13	.1705	.1706	113
		.1707		201
2.322	15	.1854	.1855	006
2.318	17	.1861	.1861	202
2.173	21	.2119	.2119	203
2.009	28	.2479	.2479	204
1.8433	12	.2943	.2943	205
		.2948		211
1.7954	8	.3103	.3102	212
1.7251	7	.3361	.3361	213
1.6388	28	.3721	.3724	214
		.3725		300
1.5457	8	.4185	.4186	215
		.4188		303
1.4837	e2	.4538	.4543	118
		.4549		304
1.4765	5	.4586	.4587	109
1.4209	7	.4952	.4953	208
1.4190	12	.4966	.4966	220
1.3573	5	.5429	.5428	223
1.3389	e3	.5564	.5578	1,0,10
		.5579		306
		.5586		312
1.3084	6	.5844	.5842	313
1.2696	8	.6204	.6204	314
1.2242	3	.6673	.6673	401
1.2106	6	.6820	.6824	226
		.6828		402
1.1881	4	.7085	.7085	403
1.1589	4	.7446	.7446	404
1.1245	3	.7909	.7909	405
1.0727	8	.8691	.8691	410

^aPrepared at 910 °C, 1 atm O₂.

^bInterplanar spacing, A.

^cObserved intensity.

^dIndexed on the basis of a hexagonal cell with a = 5.676 Å and c = 13.934 Å.

^eBroad.

Table 3.

X-ray diffraction powder data for the perovskite-like phase, BaFeO_3−X′^a

line	db	I/I0c
1	4.09	d2
2	4.01	6
3	3.98	7
4	2.91	d9
5	2.86	d,e7
6	2.82	100
7	2.31	4
8	2.30	15
9	2.02	d3
10	2.00	9
11	1.989	15
12	1.789	4
13	1.781	4
14	1.633	10
15	1.627	18
16	1.413	11
17	1.408	d8

^aPrepared at 1000 °C, 1 atm O₂.

^bInterplanar spacing, Å.

^cObserved intensity.

^dBroad.

^eAppears only after long exposure to air.

Hexagonal BaFeO_3−X does not appear to form solid solutions in the high baria portion of the system below 960 °C. Changes in oxygen stoichiometry may occur as a function of temperature to 960 °C. Cell parameters, however, do not change from 800 to 960 °C and it is, therefore, expected that change in oxygen concentration within this temperature range is limited. Below approximately 943 °C, the phase assemblage consists of hex-BaFeO_3−X and Ba₂FeO₄. The latter phase is extremely sensitive to moisture and tends to deteriorate when exposed to air and during x-ray analysis. The compound was made from a 4BaCO₃:Fe2O₃ mixture at 900 °C in air and at 900 °C in 1 atm O₂. A portion of its powder pattern, indexed by analogy with K₂SO₄ is given in table 4. The phase is orthorhombic with a = 5.93 Å, b = 10.36 Å, and c = 7.64 Å.

Table 4.

X-ray diffraction powder data for the compound, Ba₂FeC₄^a

dobsb	I/I0c	$\frac{1}{d_{\text{ calc }}^2}$	$\frac{1}{d_{\text{obs}}^2}$	hkld
4.279	10	0.0544	0.0546	021
		.0549		111
3.824	12	.0685	.0684	002
3.581	12	.0779	.0780	012
3.474	64	.0828	.0829	121
3.211	33	.0970	.0970	102
3.149	21	.1010	.1009	031
3.069	100	.1058	.1062	022
		.1063		112
2.985	91	.1123	.1122	130
2.964	71	.1138	.1138	200

^aPrepared at 900 °C in air.

^bInterplanar spacing, Å.

^cObserved intensity.

^dIndexed on the basis of an orthorhombic cell with a = 5.93 Å, b = 10.36 Å, c = 7.64 Å.

Above 943 °C, phase assemblages are quite interesting, as true perovskite-like phases appear. At the 67:66 Ba:Fe ratio single-phase tetragonal perovskites may be quenched from 960 to 1050 °C. The indexed powder pattern of a typical tetragonal phase is given in table 5. Single-phase tetragonal perovskites can also be quenched from above 960 °C at the 27:26 Ba:Fe ratio. Tetragonal perovskites at the 67:66 ratio transform to cubic on heating near 75 °C while those at the 27:26 ratio transform near 45 °C. Single phase cubic perovskites are obtained at room temperature at the 17:16 Ba:Fe ratio. An indexed pattern of a typical cubic phase is given in table 6. Figure 2 shows the variation of room temperature-determined cell parameters with temperature and composition from 975 to 1050 °C. Compositions more BaO-rich than 17:16 yield cubic perovskite solid-solution plus Ba₂FeO₄ between 943° and about 1025 °C. Above 1025 °C an unknown phase exists with the perovskite and is characterized by the appearance of lines with d = 4.19, 3.84, 3.66, 3.23, 3.12, and 2.97 Å in the low angle region. Variation of cell parameters as a function of cation-cation ratio and temperature in the limited homogeneity region is a complex feature which can be deciphered only with knowledge of accurate Fe⁴⁺:Fe³⁺ ratios. At a fixed ratio, cell expansion with increasing temperature is probably a result of increased reduction of Fe⁴⁺ to Fe³⁺ with a corresponding decrease in oxygen content. Changes with cation-cation ratio at fixed temperature may be much more complicated than merely being due to variation in barium content. Increasing barium appears to cause expansion of a_tet and contraction of c_tet and a_cubic at fixed temperature in the single-phase region. Furthermore, it is highly unlikely that a stable perovskite structure can accommodate interstitial barium. At 1000 °C, the single-phase cubic perovskite, for example, contains an excess of about 1.3 mol percent BaO (in terms of BaO:Fe₂O₃) and its composition might be described as Ba₁₇Fe₁₆O_N where N = 49−X. This may be more appropriately written as BaFe_16/17O_N/17 which now indicates iron and oxygen vacancies rather than barium interstitials. The value of X will, of course, depend on the Fe⁴⁺:Fe³⁺ ratio in the phase. Starting at the BaFeO_3−X′ phase, increasing amounts of barium at constant temperature will tend to increase the concentration of iron and oxygen vacancies. Increasing iron and O⁼ vacancies should tend to expand the structure. Although unit cell volumes increase initially with increasing barium content, the apparent a cubic contraction associated with single phases from near the 17:16 ratio to the two phase region suggests a second mechanism operating simultaneously. A plausible hypothesis involves the general equilibrum

$$2{\text{Fe}}^{3+}+\frac{1}{2}{\text{O}}_2\to 2{\text{Fe}}^{4+}+{\text{O}}^{2-}.$$

Physically, this mechanism would operate in the following manner. As iron and O⁼ vacancies are created by addition of barium, the concentration level of the O⁼ vacancies would be decreased if some of the available Fe³⁺ is oxidized to Fe⁴⁺. The total number of iron vacancies would not be affected but the Fe⁴⁺:Fe³⁺ ratio, and consequently O⁼ concentration, would increase. Cell contraction, as observed, might result.

Table 5.

X-ray diffraction data for the tetragonal perovskite at the 67:66 Ba-Fe ratio^a

dobsb	I/I0c	$\frac{1}{d_{\text{ calc }}^2}$	$\frac{1}{d_{\text{obs}}^2}$	hkld
4.003	8	0.0624	0.0624	001
3.981	12	.0630	.0631	100
2.824	100	.1253	.1254	101
2.818	90	.1259	.1259	110
2.304	29	.1883	.1884	111
2.002	13	.2494	.2494	002
1.9922	26	.2519	.2520	200
1.7892	4	.3124	.3124	102
1.7836	7	.3142	.3143	201
1.7823	5	.3148	.3148	210
1.6321	18	.3754	.3754	112
1.6281	35	.3772	.3772	211
1.4124	13	.5013	.5013	202
1.4090	14	.5037	.5038	220
1.3349	4	.5612	.5612	003
1.3284	4	.5667	.5667	300
1.2657	6	.6242	.6242	103
1.2606	10	.6290	.6293	301
		.6296		310
1.2064	4	.6871	.6872	113
1.2021	5	.6920	.6921	311
1.1524	7	.7531	.7531	222
1.0684	6	.8760	.8760	213
1.0666	10	.8791	.8791	312
1.0655	8	.8809	.8809	321

^aPrepared at 1000 °C, 1 aim O₂.

^bInterplanar spacing, Å.

^cObserved intensity.

^dIndexed on the basis of a tetragonal cell with, a = 3.985 Å, c = 4.005 Å.

Table 6.

X-ray powder diffraction data for the cubic perovskite at the 17:16 Ba-Fe ratio^a

dobsb	I/I0c	$\frac{1}{d_{\text{ calc }}^2}$	$\frac{1}{d_{\text{obs}}^2}$	hkld
3.994	13	0.0627	0.0627	100
2.824	100	.1254	.1253	110
2.306	23	.1880	.1880	111
1.9972	34	.2507	.2507	200
1.7862	8	.3134	.3134	210
1.6308	37	.3761	.3760	211
1.4122	17	.5014	.5015	220
1.3315	4	.5641	.5641	300
1.2631	12	.6268	.6268	310
1.2043	5	.6895	.6895	311
1.1530	6	.7522	.7522	222
1.0675	13	.8775	.8775	321
0.99858	3	1.0029	1.0029	400
.96875	3	1.0656	1.0656	410/322
.94146	5	1.1283	1.1282	411/330
.91635	3	1.1910	1.1909	331
.89313	5	1.2536	1.2537	420

^aPrepared at 1000 °C, 1 atm O₂.

^bInterplanar spacing, Å.

^cObserved intensity.

^dIndexed on the basis of a cubic cell with a = 3.994, Å.

Figure 2. Variation of room temperature-determined cell parameters with temperature and composition for cubic and tetragonal barium-iron perovskites.

Data pertinent to temperatures above 1050 °C could not be obtained in this study because of problems with container materials. Above the melting point of Au, platinum tubes were used as containers for equilibration runs. In gold, hex-BaFeO_3−X is not stable above 960 °C and 1 atm O₂. In platinum, however, a hexagonal BaFeO_3−X-like phase does exist above 960 °C and persists at elevated temperatures to at least 1300 °C. The phase is normally admixed with perovskite-like material but tends to predominate with increasing firing time. Consistent and reproducible equilibria data are difficult to obtain. It is hardly likely that hex-BaFeO_3−X converts at 960 °C to perovskite-like BaFeO_3−X′ which in turn transforms at some temperature > 1050 °C back to a hex-BaFeO_3−X″ phase. This is apparent, especially in view of the data of MacChesney et al. [1], suggesting the existence of an orthorhombic brownmillerite-like Ba₂Fe₂O₅ which might be expected to form at sufficiently elevated temperatures in O₂. Goto and Takada [7] using Pt containers and 1 atm O₂ reported a cubic 2BaO·Fe₂O₃ phase with a melting point near 1370 °C. As pure Ba₂Fe2O₅ is not cubic and the reported a = 8.07 Å parameter is approximately 2 × a_cubic perovskite, a value which may be used to index a hexagonal BaTiO₃-like phase, the compound was probably erroneously identified. It is likely that the phase observed was of the hex-BaFeO_3−X type. More interesting, however, is the stability of the compound to a congruent melting point in 1 atm O₂ and the exclusion of a stability range for a perovskite-like BaFe_3−X′. These reported data and the inconsistent data above 1050 °C in this study have but one common denominator, platinum containers. It is, therefore, suggested that in the presence of Pt, a reaction occurs resulting in the stabilization of a hexagonal BaTiO₃-like phase containing Pt⁴⁺. Hexagonal phases containing Pt⁴⁺ are not unusual and have been reported by Dickinson et al. [10], Blattner et al. [11], and Stratton [12]. If platinum does indeed tend to stabilize the hexagonal form, it is not surprising that the cubic and tetragonal high temperature perovskites herein reported have not been previously observed, as most experiments involving synthesis of materials are conducted using Pt-containers. Furthermore, chemical analyses for Fe⁴⁺:Fe³⁺ ratios, when Pt is present, would lead to erroneous oxygen ion concentrations.

4. Summary

Phase relations in the BaO-”iron oxide” pseudobinary in the vicinity of the 1:1 cation-cation ratio were determined in 1 atm O₂ and between 800 to 1050 °C. Below 960 °C a hex-BaTiO₃-like phase, BaFeO_3−X exists. Above 960 °C, this phase transforms to a perovskite-like BaFeO_3−X′. Specimens quenched from the 960 to 1050 °C range and having Ba:Fe ratios of 67:66 and 27:26 yield single-phase tetragonal perovskite at room temperature. Single-phase cubic perovskite can be quenched from this temperature range at the 17:16 ratio. Hex-BaTiO₃-like phases were observed in specimens heated above 1050 °C in Pt-containers. These are considered to be products of a Pt-specimen reaction.