Effect of Oxide Additions on the Polymorphism of Tantalum Pentoxide

Abstract

The high temperature, apparently tetragonal, polymorph of tantalum pentoxide can be obtained at room temperature by quenching a specimen containing 2–5 mole percent of the following oxides:SnO₂, Ga₂O₃, Cr₂O₃, Fe₂O₃, Sc₂O₃, or MgO. All the x-ray patterns can be indexed on a body centered tetragonal cell with a ≈ 3.830 Å, c ≈ 35.68 Å. However, P₂O₅, V₂O₅, Nb₂O₅, ZrO₂, Lu₂O₃, NiO, or ZnO do not stabilize the tetragonal form at room temperature. Single crystals of scandium “stabilized” Ta₂O₅ have been grown by the Czochralski technique.

1. Introduction

Ta₂O₅ is probably the only pure metal oxide stable at high temperature in air (mp ~ 1890 °C) about which little or nothing is known of the crystal structure. The reason for this is that no suitable crystals of pure Ta₂O₅ have been prepared due to the peculiar polymorphic behavior of the material. The polymorphism of Ta₂O₅ has been discussed by several authors [1, 2, 3]¹ and fully reviewed by the present authors in a previous publication [4] together with the changes which occurred in the polymorphic behavior with the addition of TiO₂.

The low-temperature form of Ta₂O₅ has an, as yet, unknown structure related to the low-temperature form of Nb₂O₅, as well as to α − U₃O₈ and α − UO₃. On heating to a temperature of about 1360 °C [2] Ta₂O₅ exhibits an enantiotropic, sluggishly reversible, phase transition to a high-temperature form. This form apparently has a unique structure unlike any other reported compound. When quenched from above the equilibrium phase transition, the apparently tetragonal high-temperature form transforms, through a mono-clinic polymorph, to a third metastable polymorph which has an x-ray powder pattern which appears to be triclinic at room temperature [4]. These transformations result in a mechanical disruption of the crystals, making an accurate measurement of the intensities of the x-ray diffraction spots essentially an impossible task. The addition of TiO₂ lowers the equilibrium transition temperature from about 1360 °C to about 1150 to 1200 °C and the metastable tetragonal ⇄ monoclinic transition of the high-temperature form from about 950 °C to as low as 600 °C [4], However, the quality of the room temperature single crystals is not improved by this addition.

In the present study, oxides other than TiO₂, containing cations of similar size, have been added to Ta₂Or₅ in an effort to obtain the high temperature tetragonal modification of Ta₂O₅, at room temperature.

2. Materials

The general quantitative spectrochemical analyses for the Ta₂O₅ used in this study has been previously reported [4]. All other oxides used were of reagent grade or better, as described in previous publications [5, 6, 7, 8, 9]. These other oxides were P₂O₅, V₂O₅, Nb₂O₅, SnO₂, ZrO₂, Ga₂O₃, Cr₂O₃, Fe₂O₃, Sc₂O₃, Lu₂O₃, MgO, NiO, and ZnO.

3. Specimen Preparation and Test Methods

One gram batches of 95:5 and 90:10 mole ratio compositions of Ta₂O₅ and each of the additional oxides were weighed, mixed in a mechanical shaker for approximately 10 min, and pressed into disks at about 10⁴ psi. The disks were placed on Pt setters and calcined in air at appropriate temperatures from 500 to 1200 °C depending on the relative volatility of the additive, for 6 to 12 hr. Other compositions of interest were prepared in a similar manner. After the preliminary heat treatment portions of the ground specimen were placed in sealed Pt tubes and heated in the quenching furnace at approximately 1325, 1450, and 1600 °C for various periods of time. Some specimens of particular interest were heated at other temperatures.

The sealed tubes containing specimens were quenched into ice water, opened, and examined by x-ray diffraction techniques. A high angle recording Geiger counter-diffractometer and Ni-filtered Cu radiation was used in the study. The Geiger counter traversed the specimen at ¼ deg/min and radiation was recorded on the chart at 1 deg-2θ/in. The unit cell dimensions reported can be considered accurate to about ± 2 in the last decimal place listed.

4. Results and Discussion

The experimental data is shown in table 1 and is interpreted diagramatically as phase equilibria data in figures 1–4. Sufficient data has not been collected to establish the exact nature of the phase equilibria diagrams. Therefore, the system, Ta₂O₅-TiO₂ (redrawn in fig. 1), has been used as a model and the other systems have been drawn in accordance with Ta₂O₅-TiO₂, with the appropriate modifications necessary to fit the limited experimental data. No attempt has been made to delineate the metastable equilibrium relations of the H-Ta₂O₅ solid solution phase transitions as was previously done for Ta₂O₅-TiO₂ [4].

Table 1.

Experimental data

System	Composition mol percent	Heat treatment				X-ray diffraction analyses c
		Initial a		Final b
		Temp. °C	Time hr	Temp. °C	Time hr
Ta2O5-P2O5	95:5	300	1
		500	1
		d1000	10
				1331	16	0 519O25(ss) + L-Ta2O5 phase(s)
				1457	16	PTa9O25(ss) + Htri-Ta2O5
				1585	4.5	PTa9O25(ss) + Htri-Ta2O5
	90:10	300	1
		500	1
		d1000	10
				1331	16	PTa9O25 + PTaO5 (trace)
				1457	16	PTa9O25 + PTaO5 (trace)
				1585	4	PTa9O25 + PTaO5 (trace)
Ta2O5-V2O5	95:5	1000	10	1333	19	VTa9O25(ss) + L-Ta2O5 phase(s)
				1444	65	VTa9O25(ss) + H-Ta2O5(ss)
				1587	5	Htri-Ta2O5 + Hmon-Ta2O5
	90:10	500	16
		1000	16			VTa9O25 + unknown phase
				1247	6	VTa9O25
				1390	16	VTa9O25 + L-Ta2O5 phase(s)e
				1444	0.667	VTa9O25 + L-Ta2O5 phase(s)e
				1597	4	Htri-Ta2O5(ss) + liquid
				1608	3.75	Htri-Ta2O5(ss) + Hmon-Ta2O5(ss)+ liquid
Ta2O5-Nb2O5	95:5	1000	10
				1333	19	L-Ta2O5 phase(s)
				1444	65	Htri-Ta2O5(ss)+L-Ta2O5 phase(s) (trace)
	90:10	1000	10	1331	16	L-Ta2O5 phase(s)
				1444	65	HtriTa2O5(ss)+L-Ta2O5 phase(s) (trace)
Ta2O5-SnO2	99:1	1200	10
				1445	16	Htri-Ta2O5(ss) + L-Ta2O5 phase(s) (trace)
				1571	4	Htri-Ta2O5(ss)
	98:2	1200	10
				1446	16	Htri-Ta2O5(ss) + Htet-Ta2O5(ss)
	97:3	1200	10
				1446	16	Htri-Ta2O5 + Htet-Ta2O5(ss)
				1571	4	Htri-Ta2O5 + Htet-Ta2O5(ss)
	95:5	1000	10	1325	16	Htet-Ta2O5(ss); a = 3.830, c = 35.68
				1448	16	Htet-Ta2O5(ss); a = 3.830, c = 35.68
				1660	2	Htet-Ta2O5(ss) + Htri-Ta2O5(ss)f
				1769	17	Htet-Ta2O5(ss)+ Htri-Ta2O5(ss)f
		1769	17	1463	65	Htet-Ta2O3(ss)+ Htri-Ta2O5(ss)f
	90:10	1000	10	1325	16	Htet-Ta2O5(ss)+ SnO2(ss)
				1448	16	Htet-Ta2O5(ss)+ SnO2(ss)
Ta2O5-ZrO2	95:5	1000	10
				1326	19	L-Ta2O5, phase(s)
				1449	19	Htri-Ta2O5(ss)+ Hmon-Ta2O5(ss)
				1586	4	Htri-Ta2O5(ss) + Hmon-Ta2O5(ss)
	90:10	1000	10
				1327	19	L-Ta2O5 phase(s) + H-Ta2O5(ss)
				1449	19	Hmon-Ta2O5(ss)
				1592	4.5	Hmon-Ta2O5(ss)
		1225	168			L-Ta2O5 phases(s)
	75:25	1000	10	1523	16	Hmon-Ta2O5(ss) + Ta2O5 · 6ZrO2
Ta2O5-Ga2O3	95:5	1000	10
				1300	16	L-Ta2O5 phase(s)+H-Ta2O5(ss) + GaTaO4-wolframite type g
				1444	64	Htet-Ta2O5(ss) + GaTaO4-ixiolite type
				1586	4	Htet-Ta2O5(ss)+GaTaO4-rutile type
				1602	20	Htet-Ta2O5(ss)f
	90:10	1000	10	1300	16	L-Ta2O5 phase(s) + GaTaO4.-wolframite type
				1444	64	Htet-Ta2O5(ss) + GaTaO4-ixiolite type
				1602	4	Htet-Ta2O5(ss) + GaTaO4-rutilet ype
Ta2O5-Cr2O3	95:5	300	1
		d 1000	10
				1320	16	H-Ta2O5(ss) + L-Ta2O5 phase(s) + CrTaO4-rutile typeg
				1444	16	Htet-Ta2O5(ss)+ Hmon-Ta2O5(ss)+CrTaO4-rutile typeg
				1591	44	Htet-Ta2O5(ss) + CrTaO4-ruti]e type (trace)f
				1602	4	Htet-Ta2O5(ss) + CrTaO4-rutile type
	90:10	300	1
		d 1000	10
				1320	16	L-Ta2O5 phase(s) + H-Ta2O5(ss)+CrTaO4-rutile typeg
				1448	16	Htet-Ta2O5(ss)+ CrTaO4-rutile type
				1597	4	Htet-Ta2O5 (ss) + CrTaO4-rutile type
Ta2O5-Fe2O3	99:1	1200	10
				1457	16	Htri-Ta2O5(ss)
				1585	4	Htri-Ta2O5(ss)
	98:2	1200	10
				1445	16	Htet-Ta2O5(ss)
				1468	65	Htet-Ta2O5(ss)
				1650	67	Htri-Ta2O5(ss)f
				1725	0.5	L-Ta2O5 phase(s)
		d 1445	16	h 1000	1	Htet-Ta2O5(ss)+ L-Ta2O5 phase(s)
				h 1000	16	Htet-Ta2O5(ss)+ L-Ta2O5 phase(s)
				h 1100	16	L-Ta2O5 phase(s)
				h 1200	60	L-Ta2O5 phase(s)
				h 1250	10	L-Ta2O5 phase(s)
				h 1300	10	L-Ta2O5 phase(s)
				1641	2	Htet-Ta2O5(ss)
	97:3	1200	10
				1445	16	Htet-Ta2O5(8s) + FeTaO4-rutile type
				1581	4	Htet-Ta2O5(ss) + FeTaO4-rutile type
	95:5	1000	10
				1327	16	Htet-Ta2O5(ss) + L-Ta2O5 phase(s) +FeTaO4-rutile typeg
				1442	16	Htet-Ta2O5(ss) + FeTaO4-rutile type
				1619	19	Htet-Ta2O5(ss) + FeTaO4-rutile type
	90:10	1000	10
				1327	16	L-Ta2O5 phase(s) + Htet-Ta2O5(ss)+FeTaO4-rutile typeg
				1442	16	Htet-Ta2O5(ss) + FeTaO4-rutile type
Ta2O5-Sc2O3	98:2	1200	10	1650	72	Htet-Ta2O5(ss)
				1734	19	Htet-Ta2O5(ss)
				1737	2	Htet-Ta2O5(ss)
	97:3	1200	10
				1737	0.5	Htet-Ta2O5(ss)
	95:5	1000	10
				1327	16	Htet-Ta2O5(ss)+L-Ta2O5(ss)+ScTaO4-AlNbO4 typeg
				1448	16	Htet-Ta2O5(ss) + ScTaO4-AlNbO4 type
				1606	4	Htet-Ta2O5(ss) + ScTaO4-AlNbO4 type
	90:10	1000	10
				1327	16	L-Ta2O5 phase(s) + Htet Ta2O5(ss) + ScTaO4-AlNbO4 typeg
				1448	16	Htet-Ta2O5(ss)+ ScTaO4-AlNbO4 type
				1608	4.5	Htet-Ta2O5(ss)+ ScTaO4-AlNbO4 type
Ta2O5-Lu2O3	95:5	1200	10
				1640	20	Htri-Ta2O5(ss)+ LuTaO4
	90:10	1200	10
				1640	19	Htri-Ta2O5(ss)+LuTaO4
Ta2O5-MgO	95:5	1000	10
				1327	16	L-Ta2O5 phase(s) + MgTa2O6-trirutile type
				1445	16	H “30:1” Ta2O5(ss) + MgTa2O6-trirutile type
				1608	4.5	Htet-Ta2O5(ss)+MgTa2O6-trinitile type (trace)
	90:10	1000	10
				1327	16	L-Ta2O5 phase(s) + MgTa2O6-trirutile type
				1445	16	H “30:l”-Ta2O5(ss) + MgTa2O6-trirutile type
				1597	4.5	Htet-Ta2O5(ss)+MgTa2O6-trirutile type
Ta2O5-NiO	95:5	1000	10
				1331	65	L-Ta2O5 phase(s) + NiTa2O6-trirutile type
				1445	16	L-Ta2O5 phase(s) + H-Ta2O5(ss) + NiTa2O6-trirutile typeg
				1583	16	H “30:1”-Ta2O5(ss) + L-Ta2O5 phase(s)f
				1597	4.5	H “30:1”-Ta2O5(ss) + L-Ta2O5 phase(s)
	90:10	1000	10	1331	65	L-Ta2O5 phase(s) + NiTa2O6-trirutile type
				1444	16	L-Ta2O5 phase(s) + H-Ta2O5(ss) + NiTa2O6-trirutile typeg
				1559	16	H-Ta2O5(ss)+L-Ta2O5(ss) + NiTa2O6-trirutile typeg
				1570	16	H-Ta2O5(ss)+L-Ta2O5(ss) + NiTa2O6-trirutile typeg
				1590	4	H-Ta2O5(ss)+L-Ta2O5(ss) + NiTa2O6-trirutile typeg
Ta2O5-ZnO	95:5	1000	10	1331	65	L-Ta2O5 phase(s) + ZnTa2O6-columbite type
				1444	16	L-Ta2O5 phase(s) + H-Ta2O5(ss) + ZnTa2O6-columbite typeg
				1590	4	H “30:1”-Ta2O5(ss) + ZnTa2O6-rutile type
	90:10			1333	19	L-Ta2O5 phase(s) + ZnTa2O6-columbite type
				1444	16	L-Ta2O5 phase(s) + H-Ta2O5(ss)4-ZnTa2O6-columbite typeg
				1585	4.5	H “30:1”-Ta2O5(ss) + ZnTa2O6-rutile type

^a(Initial Heat Treatment) − All specimens were calcined as pressed disks on Pt foil at the indicated heat treatment with heating and cooling rates of approximately 4 °C/min unless otherwise specified.

^b(Final Heat Treatment) − All specimens were quenched in sealed Pt tubes from the indicated temperature, unless otherwise specified.

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

L-Ta₂O₅ phase(s) − One or more phases with an x-ray powder diffraction pattern similar to the low-temperature form of Ta₂O₅.

H_tri-Ta₂O₅ − Triclinic metastable distortion of the high-temperature form.

H_mon-Ta₂O₅ − Monoclinic metastable distortion of the high-temperature form related to the triclinic form by $~a\sqrt{2}$ and $~b\sqrt{2}$.

Htet-Ta₂O₅ − The tetragonal high-temperature form of Ta₂O₅.

H”30:1” Ta₂O₅ − A metastable monoclinic form of Ta₂O₅ solid solution similar to that reported as “30:1” in the Ta₂O₅:TiO₂ system [4].

H-Ta₂O₃ − The high-temperature form of Ta₂O₅ in such poor crystallinity or in-sufficient amounts that the exact symmetry could not be determined.

ss − Solid solution.

^dSpecimen was sealed in a large Pt tube to minimize possible loss of more volatile component.

^eProbably a result of reduction of the addition oxide.

^fProbably due to change of specimen composition by the addition oxide reacting with Pt tube.

^gPresence of three phases indicates non-equilibrium.

^hSpecimen powder was prepared for x-ray diffraction on a Pt bar, heated, and subjected to x-ray examination without further disturbance of the specimen. Each succeeding heat treatment had also had the previous treatments.

Figure 1. Ta₂O₅-rich region of Ta₂O₅-Me₂O₅ systems, as deduced from limited quenching and x-ray diffraction data.

• − experimental data points

H-Ta₂O₅ − high temperature form of Ta₂O₅

L-Ta₂O₅ − low temperature form of Ta₂O₅

ss − solid solution

(c) See references [13, 14] for more complete phase diagrams of Ta₂O₅-Nb₂O₅ system.

Figure 4. Ta₂O₅ rich regions of Ta₂O₅-MeO systems, as deduced from limited quenching and x-ray diffraction data.

• − experimental data points

H-Ta₂O₅ − high temperature form of Ta₂O₅

L-Ta₂O₅ phase(s) − one or more phases with an x-ray diffraction powder pattern similar to the low temperature form of Ta₂O₅

ss − solid solution

T − trirutile-type structure

R − rutile-type structure

C − columbite-type structure.

Figure 2. Ta₂O₅-rich regions of Ta₂O₅-MeO₂ systems, as deduced from limited quenching and x-ray diffraction data.

• − experimental data points

▼ − specimen reheated at lower temperature than original heat treatment (see table 1)

H-Ta₂O₅–high temperature form of Ta₂O₅

L-Ta₂O₅ phase(s) − one or more phases with an x-ray powder diffraction pattern similar to the low temperature form of Ta₂O₅

ss − solid solution

(a) Portion of Ta₂O₅-TiO₂ system redrawn from reference [4].

Figure 3. Ta₂O₅-rich regions of Ta₂O₅-Me₂O₃ systems, as deduced from limited quenching and x-ray diffraction data.

• − experimental data points

▼ − specimens reheated at lower temperatures than original heat treatment (see table 1)

H-Ta₂O₅ high temperature form of Ta₂O₅

L-Ta₂O₅ phases – one or more phases with an x-ray diffraction powder pattern similar to the low temperature form of Ta₂O₅

ss − solid solution

R − rutile-type structure

I − ixiolite-type structure

W − wolframite-type structure

A − AlNbO₄-type structure.

From the results reported for the system Ta₂O₅-TiO₂ [4], it can be concluded that any specimen containing some form of H-Ta₂O₅ solid solution was probably of tetragonal symmetry at the temperature from which the specimen was quenched. Just as in the Ta₂O₅-TiO₂ system, this tetragonal phase could not be quenched to room temperature in the systems involving P₂O₅, V₂O₅, Nb₂O₃, ZrO₂, Lu₂O3, NiO, or ZnO. However, very good apparently tetragonal x-ray diffraction powder patterns could be obtained from systems involving SnO₂, Ga₂O₃, Cr₂O₃, Fe₂O₃, Sc₂O₃, and MgO added to Ta₂O₅ in amounts less than 5 mol percent

The heat treatment of the single phase apparently tetragonal specimen containing 2 mol percent Fe₂O₃ (table 1) a = 3.830, c = 35.68 Å indicates that this phase is not stable below about 1300 °C where it reverts to the low temperature polymorph. Nevertheless, it was assumed that single crystals of a size sufficient for structure analyses could be made in any of the second group of materials by annealing a specimen at high temperatures for a long period of time in a sealed platinum tube and subsequently quenching to room temperature. As can be seen from table 1 (footnote f) this cannot be done for systems involving SnO₂, Ga₂O₃, Cr₂O₃, or Fe₂O₃, since the latter components apparently react with the Pt container over a period of time. The specimen becomes enriched in Ta₂O₅ and “destabilized.”

Therefore only Sc₂O₃ and MgO can be used to synthesize crystals of tetragonal H-Ta₂O₅ solid solution. Small crystals of 98Ta₂O₅:2Sc₂O₃ were succesfully grown in the solid state using sealed or open Pt tubes at about 1735 °C. Preliminary single crystal x-ray patterns, made with the Burger precession camera and Weissenburg camera, indicated that the crystals were of sufficient size for single crystal structure determination. However, accurate measurement of intensities revealed serious inconsistencies from one crystal to the next.

It should be noted that all of the additional oxides which tend to stabilize the high-temperature form are those with cations generally occurring in octahedral coordination. Another group of oxides containing smaller cations were found to stabilize phases structurally similar to the low-temperature form of Ta₂O₅. These experiments are reported in Part III of this series [10].

5. Crystal Growth

Attempts to grow larger crystals of Htet-Ta₂O₅-Fe₂O₃ solid solution by the Czochralski method were unsuccessful as there was an apparent slow reaction with the iridium container. Therefore efforts were directed toward growing a Sc stabilized crystal. The initial crystal growth attempts were made using 2 mol percent Sc₂O₅ in the melt. At this level the amount of Sc incorporated into the crystal was insufficient to stabilize the high form. The concentration of Sc₂O₃ in the melt was increased to 4 mol percent and this was sufficient for the crystals to be stabilized in the high form in the “as grown condition.” The actual amount of Sc incorporated in the crystals is not known.

For all compositions there were some general problems encountered which seem noteworthy and give a general indication of the unusual nature of this material. The first problem encountered was in the preliminary melting or filling of the iridium crucibles with either partially sintered pellets or powder. As the material melted it wet the iridium and the surface tension would pull the unmelted central core up in the container until some of the contents would flow over the sides. This problem was overcome by increasing the size of the container from ¾ in ID to 1½ in ID. The second problem was encountered when the Ta₂O₅ remaining in the iridium container was solidified at the termination of a growth attempt. Evidently Ta₂O₅ undergoes a volume expansion on freezing and this expands the container after each run, thus drastically shortening the container’s useful lifetime. The third problem was one which is common to those materials which are poor thermal conductors, particularly at high temperatures. In such a material, growth across the melt surface occurs, which insulates the top of the melt. “Melt off” occurs and a hollow bowl like crystal is grown. A similar behavior has been noted for the growth of TiO₂ [11]. Due to the difficulties mentioned the largest crystals obtained were of the order of 1 × 2 mm long and about 1 inch in diameter. A small cleavage fragment of one of these crystals has been used to determine the structure of H-Ta₂O₅ and the results of this structure determination will be reported elsewhere [12].