CHEMISTRY

V. K. TRUNOV, L. M. KOVBA, E. I. SIROTKINA

X-RAY DIFFRACTION STUDY OF DOUBLE OXIDES OF SOME TRANSITION METALS

(Presented by Academician V. I. Spitsyn, July 5, 1963)

In the present work the formation of double oxides was investigated in the interaction of β-Nb₂O₅ with WO₃, MoO₃, V₂O₅ and of α-Ta₂O₅ with V₂O₅.

The investigation was carried out by us by the method of X-ray phase analysis. The samples were photographed in CuKα radiation in a focusing camera with a bent germanium single crystal as monochromator (¹). Annealing of mixtures of stoichiometric amounts of the initial oxides was carried out at temperatures from 700 to 1200°. Mixtures containing more than 80 mol.% WO₃ and all mixtures with MoO₃ were annealed in sealed quartz ampoules to prevent intensive evaporation of these oxides.

Table 1

X-ray phase analysis of samples in the β-Nb₂O₅—WO₃ system

* I — solid solution of WO₃ in β-Nb₂O₅; II — WO₃·4Nb₂O₅; III — WO₃·2Nb₂O₅; IV — 7WO₃·4Nb₂O₅; V — Nb₂O₅·3WO₃; VI — Nb₂O₅·nWO₃; VII — solid solution of Nb₂O₅ in WO₃.

In the β-Nb₂O₅—WO₃ system we found the following compounds (Table 1): WO₃·4Nb₂O₅ (20 mol.% WO₃), WO₃·2Nb₂O₅ (33.3 mol.% WO₃), 7WO₃·4Nb₂O₅ (63.6 mol.% WO₃), 3WO₃·Nb₂O₅, Nb₂O₅·nWO₃ (n ≃ 48—50).

In samples with a content of more than 90 mol.% WO₃, homogenization proceeds with difficulty, and on the X-ray diffraction patterns of all samples containing more than 90 mol.% WO₃, in addition to the lines of the Nb₂O₅·nWO₃ phase, there are present lines of WO₃ with a somewhat changed axial ratio, and in a number of samples also 3WO₃·Nb₂O₅. The weakest lines of impurity phases are on the X-ray diffraction pattern of the sample with 98 mol.% WO₃; from the results of phase analysis it may be concluded that n ≃ 48—50. The Nb₂O₅·nWO₃ phase cannot be considered a solid solution of Nb₂O₅ in WO₃, since between WO₃ and Nb₂O₅·nWO₃ there exists a heterogeneous region, and the change in the WO₃ lattice on passing to Nb₂O₅·nWO₃ occurs discontinuously.

On the X-ray diffraction pattern of the phase Nb₂O₅·\(n\)WO₃ there is a large number of superstructure lines that are not observed for WO₃. The type of superstructure could not be established from the powder X-ray pattern.

The results obtained differ substantially from the data of Goldschmidt (²), who probably took the double oxides WO₃·4Nb₂O₅ and WO₃·2Nb₂O₅ to be a solid solution of WO₃ in β-Nb₂O₅, and the phases 7WO₃·4Nb₂O₅ and 3WO₃·Nb₂O₅ to be a single compound. The phase Nb₂O₅·\(n\)WO₃ was apparently regarded by him as a solution of Nb₂O₅ in WO₃.

In the investigation of the system β-Nb₂O₅—MoO₃, the samples were annealed in sealed ampoules at temperatures of 700 and 1100° (see Table 2). In samples annealed at 1100°, only three phases are present: β-Nb₂O₅, MoO₃, and 4Nb₂O₅·MoO₃, the X-ray pattern of which is identical with the X-ray pattern of 4Nb₂O₅·WO₃. The sample with 20 mol. % MoO₃ contained only the phase 4Nb₂O₅·MoO₃. We also carried out annealing of samples at 700°. On the X-ray patterns of samples with 66.7, 50, and 33.3 mol. % MoO₃, lines of three phases were present: MoO₃, 4Nb₂O₅·MoO₃, and a phase isostructural with the compound 2Nb₂O₅·WO₃. Homogenization of the sample with 33.3 mol. % MoO₃ proceeds with great difficulty (annealing at 800° for 14 days did not lead to homogenization of the sample), and therefore the phase 2Nb₂O₅·MoO₃ could not be obtained in pure form; however, in view of the isostructurality of this phase with the phase 2Nb₂O₅·WO₃, it may be considered that its composition is expressed by the indicated formula. The sample with 75 mol. % MoO₃ contained only the phase Nb₂O₅·3MoO₃, the X-ray pattern of which proved not to be identical with the X-ray pattern of Nb₂O₅·3WO₃.

Table 2

X-ray phase analysis of samples in the system β-Nb₂O₅—MoO₃*

* I — β-Nb₂O₅; II — 4Nb₂O₅·MoO₃; III — MoO₃; IV — 3MoO₃·Nb₂O₅; V — 2Nb₂O₅·MoO₃.
** The samples were annealed in sealed quartz ampoules.

The structures of the oxides 7WO₃·4Nb₂O₅ and 3MoO₃·Nb₂O₅, as well as the structure of 3WO₃·Nb₂O₅, are probably built on the basis of the ReO₃ structure, with the difference that the double oxides 3WO₃·Nb₂O₅ and 7WO₃·4Nb₂O₅ have a similar superstructure, from which the superstructure of 3MoO₃·Nb₂O₅ differs considerably. All lines of the X-ray pattern of 3WO₃·Nb₂O₅ (³) and most lines of the X-ray pattern of 7WO₃·4Nb₂O₅ are indexed assuming a tetragonal unit cell, the \(c\) axis of which is analogous to the lattice period of ReO₃, while the \(a\) axis is the lattice period of ReO₃ increased by \(\sqrt{10}\) times (i.e., corresponds to the 310 direction of the ReO₃ cell).

A certain number of weak lines of the X-ray pattern of the compound 7WO₃·4Nb₂O₅ are not indexed assuming a unit cell with such identity periods; the cell is rhombic, pseudotetragonal. If the cell period of ReO₃ is denoted by \(a'\), then the cell periods of 7WO₃·4Nb₂O₅ are: \(a \approx a'\sqrt{10}\); \(b \approx 3a'\sqrt{10}\) and \(c \approx a'\), i.e. \(a = 12.219 \pm 0.006\) Å; \(b = 36.66 \pm 0.02\) Å and \(c = 3.940 \pm 0.002\) Å.

The lines of the X-ray pattern of the compound 3MoO₃·Nb₂O₅ are indexed assuming a tetragonal unit cell, the \(c\) axis of which is equal to the period \(a\) of the ReO₃ cell, while the direction of the \(a\) axis is apparently close to the 210 direction of the ReO₃ cell (\(a = 23.12 \pm 0.01\) Å; \(c = 3.995 \pm 0.005\) Å;

\(z = 9\)). The results of indexing the X-ray diffraction pattern of \(3\mathrm{MoO}_3 \cdot \mathrm{Nb}_2\mathrm{O}_5\) are given in Table 3.

Table 3

Results of indexing the X-ray diffraction pattern of \(3\mathrm{MoO}_3 \cdot \mathrm{Nb}_2\mathrm{O}_5\)

In the absence of appreciable displacement of atoms from the subcell nodes, the 630 line of \(3\mathrm{MoO}_2 \cdot \mathrm{Nb}_2\mathrm{O}_5\) should have been the brightest among the \(hk0\) lines. Such a regularity in the intensities of \(hk0\) reflections is indeed observed for the oxides \(3\mathrm{WO}_3 \cdot \mathrm{Nb}_2\mathrm{O}_5\), \(7\mathrm{WO}_3 \cdot 4\mathrm{Nb}_2\mathrm{O}_5\), and \(3\mathrm{WO}_3 \cdot \mathrm{Ta}_2\mathrm{O}_5\). In the case of \(3\mathrm{MoO}_3 \cdot \mathrm{Nb}_2\mathrm{O}_5\), the brightest line is 540, which indicates a more considerable displacement of metal atoms from the subcell nodes than in the analogous tungsten compounds.

Table 4

Interplanar spacings for the phase \(\mathrm{Nb}_2\mathrm{O}_5 \cdot n\mathrm{WO}_3\)

The bright lines of the X-ray diffraction pattern of the \(\mathrm{Nb}_2\mathrm{O}_5 \cdot n\mathrm{WO}_3\) phase are indexed on the assumption of a monoclinic subcell with \(a = 5.277\ \text{Å}\), \(b = 5.172\ \text{Å}\), \(c = 3.832\ \text{Å}\), \(\beta = 87^\circ 50'\). The presence of superstructure lines indicates an ordered arrangement of Nb and W atoms. The interplanar spacings for the \(\mathrm{Nb}_2\mathrm{O}_5 \cdot n\mathrm{WO}_3\) phase are given in Table 4.

In the systems \(\beta\)-\(\mathrm{Nb}_2\mathrm{O}_5\)—\(\mathrm{V}_2\mathrm{O}_5\) and \(\alpha\)-\(\mathrm{Ta}_2\mathrm{O}_5\)—\(\mathrm{V}_2\mathrm{O}_5\), we established the existence of isostructural compounds \(\mathrm{V}_2\mathrm{O}_5 \cdot 6\mathrm{Me}_2\mathrm{O}_5\) (14.3% \(\mathrm{V}_2\mathrm{O}_5\)) or \(3\mathrm{V}_2\mathrm{O}_5 \cdot 17\mathrm{Me}_2\mathrm{O}_5\) (15 mol.% \(\mathrm{V}_2\mathrm{O}_5\)), where \(\mathrm{Me}=\mathrm{Nb}\) or \(\mathrm{Ta}\). All samples with other compositions (10, 20, 50, 90% \(\mathrm{V}_2\mathrm{O}_5\)) are two-phase, and their X-ray diffraction patterns contained lines of the starting oxides present in excess. The oxide \(\mathrm{V}_2\mathrm{O}_5 \cdot 6\mathrm{Nb}_2\mathrm{O}_5\) corresponds to the \(\beta'\)-\((\mathrm{Nb}, \mathrm{V})_2\mathrm{O}_5\) phase found by Goldschmidt, who assigns to it a homogeneity range of 10–25 mol.% \(\mathrm{V}_2\mathrm{O}_5\). According to our data, the phases \(\mathrm{V}_2\mathrm{O}_5 \cdot 6\mathrm{Me}_2\mathrm{O}_5\) have a very narrow homogeneity range.

The present work practically completes the investigation of binary systems formed from the pentoxides of vanadium, niobium, and tantalum and the trioxides of molybdenum, tungsten, and uranium. Although the structures of these compounds have not been sufficiently studied, even the available data make it possible to divide these compounds into several groups. Uranyl vanadates, molybdates, and tungstates have a complex structure and are probably uranyl salts. The oxides $UMeO_5$ and $UMeO_{5+x}$ ($Me = V, Nb, Ta, Mo, W$) and $UMe_3O_{10} — UMe_3O_{10+y}$ ($Me = V, Nb, Ta$) are built on the basis of the structure of $\alpha$-$UO_3 — U_3O_8$ and the $\alpha$ modifications of niobium and tantalum pentoxides. The ordered arrangement of metal atoms leads to the appearance of various superstructures. Several compounds ($Nb_2O_5 \cdot nWO_3$, $Nb_2O_5 \cdot 3WO_3$, $Nb_2O_5 \cdot 3MoO_3$, $Ta_2O_5 \cdot 3WO_3$, $7WO_3 \cdot 4Nb_2O_5$) are built on the basis of the $ReO_3$ structure. The ordered placement of metal atoms at the nodes of a primitive pseudocubic lattice, displacement of atoms from these ideal positions associated with deviations of the compositions from $MeO_3$, and distortion of the oxygen octahedra around the metal atoms lead to the appearance of superstructures. The Debye patterns of the compounds $V_2O_5 \cdot 6Me_2O_5$ ($Me = Nb, Ta$), $4Nb_2O_5 \cdot MeO_3$, $2Nb_2O_5 \cdot MeO_3$ ($Me = W, Mo$) somewhat resemble the Debye pattern of $\beta$-$Nb_2O_5$. Their structures can probably be regarded as intermediate between $\beta$-$Nb_2O_5$ and oxides built on the basis of the $ReO_3$ structure. All three latter groups of compounds are characteristic representatives of double oxides.

Moscow State University
named after M. V. Lomonosov

Received
28 VI 1963

CITED LITERATURE

¹ Yu. P. Simanov, V. K. Trunov et al., collection New Machines and Instruments for Testing Metals, Moscow, 1963. ² H. J. Goldschmidt, Metallurgia, 62, No. 373 (1960). ³ L. M. Kovba, V. K. Trunov, DAN, 147, 622 (1962).