Yu. M. Korenev, Yu. P. Simanov, Corresponding Member of the Academy of Sciences of the USSR
A. V. Novoselova
Submitted 1962-01-01 | RussiaRxiv: ru-196201.16269 | Translated from Russian

Full Text

Chemistry

Yu. M. Korenev, Yu. P. Simanov, Corresponding Member of the Academy of Sciences of the USSR
A. V. Novoselova

On the Rhombic Modification of Beryllium Fluoride

Goldschmidt (¹), on the basis of the closeness of the radii of Be²⁺ ($r = 0.34$ Å) and Si⁴⁺ ($r = 0.39$ Å), F⁻¹ ($r = 1.33$ Å) and O²⁻ ($r = 1.32$ Å), suggested that beryllium fluoride should possess a structure similar to that of silicon dioxide. The structural similarity of beryllium fluoride to silicon dioxide was first experimentally discovered by Brandenberger (²). He found that, upon decomposition of $(\mathrm{NH}_4)_2\mathrm{BeF}_4$, cristobalite-like $\mathrm{BeF}_2$ is formed, with the parameters of the tetragonal cell: $a = 6.60$ kX and $c = 6.74$ kX, $c/a = 1.02$.

In work (³) it was established that tetragonal $\mathrm{BeF}_2$ at 130° transforms into a cubic modification ($a = 6.78$ kX). This transformation is analogous to the $\alpha \rightleftarrows \beta$ transformation of cristobalite ($\mathrm{SiO}_2$).

In studying the $\mathrm{NaF}—\mathrm{BeF}_2$ system (⁴), in our laboratory a modification of $\mathrm{BeF}_2$ was discovered that is structurally similar to low-temperature quartz. This modification of $\mathrm{BeF}_2$ has a hexagonal lattice with parameters: $a = 4.72$ kX, $c = 5.18$ kX, $c/a = 1.10$.

Roy et al. (⁵) established that quartz-like $\mathrm{BeF}_2$ at 220° passes into a modification similar to high-temperature quartz. The lattice parameters of this form of $\mathrm{BeF}_2$ are: $a = 4.77$ kX, $c = 5.24$ kX, $c/a = 1.09$. In investigating the $\mathrm{BeF}_2—\mathrm{ZrF}_4$ system we discovered a new modification of $\mathrm{BeF}_2$. This modification of $\mathrm{BeF}_2$ was obtained by us from the quartz-like modification in the following way.

We prepared melts of beryllium fluoride with zirconium fluoride by driving off ammonium fluoride from a mixture of $(\mathrm{NH}_4)_2\mathrm{BeF}_4$ and $(\mathrm{NH}_4)_3\mathrm{ZrF}_7$, containing 4% $\mathrm{ZrF}_4$ (relative to the simple fluorides). The driving off of ammonium fluoride was carried out in a stream of $\mathrm{CO}_2$; the furnace temperature was at first slowly raised to 300°, at which temperature the greater part of $\mathrm{NH}_4\mathrm{F}$ was driven off; then the temperature was raised to 400–420° and the remainder of the ammonium fluoride was driven off. X-ray investigation showed that the melt obtained contains beryllium fluoride in the form of the quartz-like modification. It was of interest to determine how zirconium fluoride affects the polymorphic transformations of $\mathrm{BeF}_2$.

Earlier, in a thermographic study of the $\mathrm{NaF}—\mathrm{BeF}_2$ system (⁶), in addition to the known modifications, one more modification of $\mathrm{BeF}_2$ was found, existing in the interval 680–420°, from which, on cooling, quartz-like $\mathrm{BeF}_2$ was obtained; however, it could not be detected by X-ray methods. We carried out annealing of the fluoride mixture (4% $\mathrm{ZrF}_4$) at 590–600° in sealed quartz ampoules under a residual argon pressure of 10–20 mm Hg. To avoid interaction of quartz with the fluorides, the latter were placed in a nickel crucible. The annealing lasted 25 hours, after which the ampoule was cooled to room temperature.

X-ray investigation of the preparation obtained was carried out in capillaries made of Pyrex glass using cobalt radiation with an iron filter in an RKU-86 camera and copper radiation with a nickel...

filtered in the RKD chamber. The placement of the film was asymmetric. The photographs were measured on a comparator giving a reading accuracy of 0.05 mm. In all, three measurements were made by three different persons, and the mean value of \(1/d^{2}\) was taken from the results obtained. In the photographs (rather strongly veiled), two halos from amorphous BeF\(_2\) were clearly distinguished. The measured lines could be satisfactorily indexed in a rhombic body-centered lattice with axial parameters: \(a = 10.01 \pm 0.01\) kX; \(b = 13.07 \pm 0.02\) kX; \(c = 16.24 \pm 0.01\) kX*. The results of the indexing are given in Table 1. Lines belonging both to quartz-like and to cristobalite-like BeF\(_2\), as well as to zirconium fluoride, were not found in the photographs.

Table 1

Results of indexing the X-ray pattern of BeF\(_2\)

No. \(I\) \(1/d^{2}_{\mathrm{exp}}\) \(hkl\) \(1/d^{2}_{\mathrm{calc}}\) No. \(I\) \(1/d^{2}_{\mathrm{exp}}\) \(hkl\) \(1/d^{2}_{\mathrm{calc}}\)
1 med. br. 0.0605 004 0.0606 26 weak med. 0.4208 604 0.4195
2 v. v. v. weak 0.0670 123 0.0675 27 v. v. weak 0.4433 624 0.4429
3 v. v. weak 0.0759 114 0.0764 28 v. v. weak 0.4602 615 0.4595
4 v. v. weak 0.0833 024 0.840 29 v. v. weak 0.4781 358 0.4785
5 med. 0.0936 040 0.0936 30 weak sh. 0.4952 606 0.4953
6 med. weak 0.1002 204 0.1005 31 v. v. weak 0.5140 644 0.5131
7 v. v. weak 0.1099 312 0.1107 32 v. v. weak 0.5291 547 0.5285
8 med. 0.1240 224 0.1239 33 v. v. v. weak 0.5449 194 0.5444
9 v. v. weak 0.1337 240 0.1335 34 v. v. weak 0.5536 ? ?
10 v. weak 0.1469 323 0.1472 35 v. v. weak 0.5696 660 0.5695
11 med. 0.1546 044 0.1542 36 v. v. weak sh. 0.5872 2.0.12 0.5857
12 v. weak sh. 0.2080 325 0.2079 36 v. v. weak sh. 0.5872 646 0.5889
13 v. v.weak 0.2239 161 0.2244 37 v. v. v. weak 0.6085 725 (?) 0.6067
14 med. 0.2425 008 0.2426 38 v. v. weak 0.6279 581 0.6274
15 v. v. weak 0.2540 501 0.2530 39 v. v. v. weak 0.6483 811 0.6477
16 v. v. weak 0.2715 064 0.2712 40 v. v. v. weak 0.6791 673 0.6796
17 v. v. weak sh. 0.2847 503 (?) 0.2833 41 v. v. v. weak 0.6975 727 0.6976
18 v. v. weak 0.2960 354 0.2964 42 v. v. weak 0.7164 1.11.0 0.7178
19 weak med. sh. 0.3058 435 0.3069 43 v. v. weak 0.8637 862 0.8642
20 v. v. v. weak 0.3190 426 0.3193 44 v. weak 0.9192 5.6.11 0.9184
21 med. 0.3358 048 0.3362 45 v. weak 0.9426 3.0.15 0.9425
22 weak 0.3498 417 0.3510 46 v. weak 0.9707 0.0.16 0.9702
23 weak sh. 0.3688 611 0.3685 47 v. v. v. weak 1.0046 6.10.4 1.0045
24 v. v. v. weak 0.3855 462 0.3853 48 v. v. v. sh. 1.2785 6.8.12 1.2791
25 weak 0.3986 239 0.3995 49 v. v. v. sh. 1.3998 10.2.10 1.3994

The modification of BeF\(_2\) found by us lies precisely in the temperature interval in which, according to work \((^{7})\), the existence (by analogy with SiO\(_2\)) of a tridymite-like modification of beryllium fluoride is assumed, as is seen from the following scheme given in that article:

\[ \beta Q — \mathrm{BeF}_{2} \ \underset{}{\overset{220^\circ}{\rightleftarrows}}\ \alpha Q — \mathrm{BeF}_{2} \ \underset{}{\overset{420—450^\circ}{\rightleftarrows}}\ T\ (?)\,\mathrm{BeF}_{2} \ \underset{}{\overset{\sim 680^\circ}{\rightleftarrows}}\ \alpha C — \mathrm{BeF}_{2} \ \underset{}{\overset{\sim 800^\circ}{\rightleftarrows}}\ \text{melt.} \]

\[ \alpha C — \mathrm{BeF}_{2} \ \underset{1130^\circ}{\downarrow}\ \beta C — \mathrm{BeF}_{2} \]

According to Gibbs \((^{8})\), low-temperature \(\alpha\)-tridymite (SiO\(_2\)) crystallizes in the rhombic system with parameters: \(a = 9.88\) kX; \(b = 17.1\) kX; \(c = 16.3\) kX.

The results of indexing the X-ray pattern of Gibbs’s tridymite, presented in Shishakov’s book \((^{9})\), as well as our photograph of dinas refractory, indicate that the lattice is body-centered.

* The small accuracy achieved in determining the parameters is explained by the indistinctness of most of the lines in the photographs.

The prominent role of reflections with all even indices permits one to speak of the presence of a subcell with a volume equal to \(1/8\) of the main cell. In considering the data of our table for the indexing of rhombic BeF\(_2\), one can be convinced of its analogy with tridymite with respect to the mode of centering and the presence of a subcell. As for the absolute values of the axial parameters for the two substances being compared, then, along with a difference between the values of \(b\), for the axes \(a\) and \(c\) we have a close similarity.

The temperature range of existence, as well as many structural properties, of the BeF\(_2\) modification obtained by us make it possible to call it tridymite-like.

Moscow State University
named after M. V. Lomonosov

Received
18 VII 1962

REFERENCES

  1. V. H. Goldschmidt, Skr. Norske Vidensk.-akad. Oslo, Mat.-Naturvid. Klasse, 8 (1926); Geochemische Verteilungsgesetze der Elemente, 8 (1927).
  2. F. V. Brandenberger, Schweiz. min. u. petrogr. Mitt., 12, 243 (1932).
  3. D. F. Kireev, A. V. Novoselova, Yu. P. Simanov, DAN, 107, No. 6, 837 (1956).
  4. A. V. Novoselova, M. E. Levina et al., ZhOKh, 14, issue 6, 385 (1944).
  5. D. M. Roy, R. Roy, E. F. Osborn, J. Am. Ceram. Soc., 36, No. 6, 185 (1953).
  6. A. V. Novoselova, M. E. Levina, M. P. Savel’eva, ZhNKh, 3, 2562 (1958).
  7. A. V. Novoselova, Usp. khim., 28, issue 1, 33 (1959).
  8. R. E. G. Gibbs, Proc. Roy. Soc., 113, 351 (1927).
  9. N. A. Shishakov, Problems of the Structure of Silicate Glasses, Publishing House of the Academy of Sciences of the USSR, 1954, p. 44.

Submission history

Yu. M. Korenev, Yu. P. Simanov, Corresponding Member of the Academy of Sciences of the USSR