Gravimetric determination of beryllium with sodium oxinate

0039-914Q/78/0401-0203SO2 00/O

Talanro,Vol. 25. pp 203-208. 0 PergamonPress.Ltd.. 1978 Printedin Great Brltam

GRAVIMETRIC DETERMINATION OF BERYLLIUM WITH SODIUM OXINATE* A. M.

HUNDEKAR,

P. UMAPATHY@

and D. N.

SEN

National Chemical Laboratory, Poona-411008, India (Received 6 April 1977. Reoised 8 July 1977. Accepted 2 October 1977)

Summary-Sodium oxinate is found to precipitate Be(H) quantitatively in the pH range 7.5-8.2. The complex has the composition BezO(CgH6N0),.2Hz0, is stable and can be weighed directly after drying at 105-110”. A method for the estimation of Be(H) and its separation from interfering elements is described. The monohydrate has been prepared from the dihydrate and characterized. The results show the presence of hydroxyl bridges in the monohydrate. Methods using various organic reagents for the direct estimation of beryllium in ores and alloys have been examined and it is found that 4-chloro2,Sdimethoxyacetoacetanilide gives the best results. A method for the determination of Be(I1) in beryl without prior separation of FeJIII) and Al(II1) is described.

There is a general lack of specific organic precipitants for the direct gravimetric determinatioti and separation of beryllium(U). Although a few chelating agents, such as p-diketones,‘d Schifl’s bases,’ &quinolinol and its derivatives,‘-” have been proposed for gravimetric determination of beryllium, on closer scrutiny they have been found to be neither specific nor precise. Selectivity has been attained in some instances by masking the interfering elements with EDTA. However, the application of these reagents to the analysis of beryl (which contains fair

amounts of aluminium) and to the analysis of alloys is complicated and involves prior separation of interfering elements. Recently 4-chloro-2,5dimethoxyacetoacetanilide has been used for the quantitative precipitation of beryllium from solutions containing Fe(II1) and Al(III), in presence of EDTA,” and is now found suitable for the direct determination of beryllium in beryl. Sodium oxinate has also been examined as a reagent for the determination of beryllium. It precipitates beryllium quantitatively in the pH range 758.2, forming a basic hydroxyquinolinate Be20(C9H6N0)2.2H20 which can serve as the weighing form. Various other reagents have also been tested.

A 25ml portion of standard solution containing approximately 10 mg of beryllium was diluted to about 125 ml with water in a 250-ml beaker. Then 60-70 ml of 1% sodium oxinate solution were added, with stirring. The pH of the resulting clear solution was adjusted to 7.5-8.0 by dropwise addition of 4M ammonia, with stirring. The bulky yellow crystalline precipitate was digested on a hot water-bath for 1 hr, collected on a weighed sintered-glass crucible (porosity No. 4), and washed with hot water till the washings were colourless (-200 ml), then dried at 105-110” for 2 hr, cooled in a desiccator’.and weighed as Be20(C9H6N0)2.2Hz0. The maximum error in determination of 3.8-10.5 mg of beryllium was 0.3%. Procedure when Al”+, Fe”+, CuZf

and Zn*+

are present

Masking of these ions with EDTA was found unsatisfactory. Aluminium was therefore separated by precipitation as the oxinate at pH 4-5 in acetate buffer or with tannin in presence of ammonium nitrate, ammonium acetate and acetic acid.16-19 Iron was removed as the oxinate or the tannin complex. Copper was removed by electrolysis20*21 or precipitation as the oxinate at pH 4-5 in acetate buffer. Zinc was separated as the quinaldinate” or the oxinate. The results were satisfactory, the maximum error being 0.7% in the determination of 10 mg of beryllium in presence of 25-50 mg of aluminium, and 1% in presence of 10-25 mg of iron(II1). The effect of pH and amount of reagent was studied. The optimum pH range is 7.5-8.0 and 150% excess of reagent is required. Composition of the precipitate

EXPERIMENTAL

Reagents

Beryllium stock solution was prepared by dissolving beryllium nitrate trihydrate, beryllium hydroxide or beryllium sulphate tetrahydrate in the appropriate acid and diluting with water, and was standardized gravimetrically.‘2~13 Sodium oxinate was prepared according to the reported methods’4*15 and used as a 1% solution in water. All other reagents were analytical grade. * N.C.L. communication

Procedure

number 2129. 203

The dried complex (m.p. 29>300”, decomp.) was analysed. Found: C, 59.9%; H, 4.8%; N, 7.5%; Be, 4.95%. Be20(C9H6N0)2~2H20 requires: C, 60.34%; H, 4.49%; N, 7.82%; Be, 5.03%. The dihydrate was dehydrated by heating at 80” under reduced pressure (- 3 mmHg) for 16 hr. A quantitative yield of the monohydrate (m.p. 306-V, decomp.) was obtained. Found: C, 63.40/ H, 4.6%; N, 8.0%; Be, 5.21%. Be20(C9H,NO)2~H20 requires: C, 63.53%; H, 4.12%; N, 8.24%; Be-, 5.30%. The results are slightly out with the accepted limits of error for elemental analysis, but seem conclusive when considered along with the other evidence.

204

A. M. HUNDEKAR,P. UMAPATHVand D. N.

SEN

Table 1. Gravimetric determination of beryllium in pure Be(H) solutions with organic reagents

Reagent

Be found,* mg

Be, %t

Be, msS

9.76 9.17 9.76 9.93 9.27

4.91 8.11 6.03 6.17

3.99 3.94 3.96 3.47 3.41

9.16 9.72 9.48

5.62 7.29

4.18 3.91 3.74

9.76 9.17

4.91 Ij 4.925

3.92 3.94, 4.04

Hydroxide Pyrophosphate Tannin’8s’g Cupferron6 Acetoacetanilide’ bis-salicyladehydeethylenediamine’ 2-Methyl-8-quinolino1’0 8-Hydroxyquinolineg 4-Chloro-2,5-dimethoxyacetoacetanilide” Sodium oxinate

* Be taken, 9.16 mg. t In beryl. $ In Be-Cu alloy, 3.98 mg of Be taken. 11 With 2 g of EDTA added as masking agent. fi After removal of interfering elementsVarious other organic reagents, including 4-chloro-2,5dimethoxyacetoacetanilide, were used for analysis of pure beryllium solutions by the procedures described in the literature. The results are presented in Table 1. Determination of beryllium in beryl

Finely powdered beryl (0.5 g) was fused with 15 g of sodium carbonate in a platinum crucible for about 15 min. The melt was cooled and taken up with hydrochloric acid (1:l). The solution was evaporated to dryness, the residue treated with 2 ml of perchloric acid and 5 ml of concentrated sulphuric acid, and the solution evaporated to dryness. The residue was dissolved in hot water and silica was filtered off, washed, ignited, and treated with hydrofluoric and sulphuric acids. The residue was fused with potassium bisulphate, the cooled melt was dissolved in water and the solution added to the main filtrate, which was then made up to 500 ml. The beryllium content was determined on 50-ml aliquots with various reagents. The results are presented in Table 1. Determination of beryllium in Be-Cu alloy

The alloy (2 g) was dissolved in 50 ml of nitric acid (1: 1) by gentle heating. The solution was boiled for 5-10 min, cooled and diluted to 500 ml. The beryllium content was determined on 50-ml aliquots. As copper is quantitatively precipitated by 4-chloro-2,5-dimethoxyacetoacetamlide at pH 6.G7.3, I’ the copper was separated electrolytically before precipitation of the beryllium. The results are shown in Table 1.

RESULTS AND DISCUSSION Although

8-quinolinol

has been proposed

workers8*g as a gravimetric tion of beryllium, concerning reported

there

the nature quantitative

by earlier

reagent for the determinais considerable

of the precipitate. precipitation

in

controversy

radius (0.31 A) beryllium might be expected not to form a complex with 8-quinolinol. The procedures using Squinolinol were therefore examined and rather variable results (+ 3% error) were obtained. Further, acetic acid was found to interfere with the estimation, necessitating the use of undue excess of the reagent. During washing of the beryllium complex with dilute ammonia (1:lOO) to remove excess of the reagent, losses of the complex itself, due to dissolution, are observed (Table 2). It was for these reasons that the water-soluble sodium salt of 8-quinolinol was considered, and it was found to give consistently satisfactory results. However, prior separation of Al(III), Fe(III), Zn(II), Cu(II), etc. is necessary. Acetate, sulphate, chloride etc. do not interfere. Elemental analysis, infrared spectroscopy, thermogravimetry and differential thermal analysis clearly show that the precipitated beryllium complex has the composition Be20(CgH6N0)2 .2Hz0. Thermogravimetric and DTA curves for the complexes, BezO(CgH6N0)2.2Hz0 and BezO(CgHsNO)z.H,O are shown in Figs. 1 and 2. The dihydrate was also found to be formed when beryllium oxinate was prepared by precipitation from homogeneous solution,24 and it was postulated that on loss of water during pyrolysis of the compound, an unstable intermediate was formed which immediately decomposed to form Table 2. The effect of washing with dilute ammonia solution (1:lOO) in estimation of beryllium with I-quinolinol (0.3528 g of complex used)

Rao et ~1.’

the pH range 7.168.12 and said that the beryllium oxinate was Be(CgH6NO),. Motojima, however, proposed the formula Be20(CgH6N0)2 .2H20.’ In both procedures the complex may be weighed as such, but Rao ,et d8 apparently preferred to ignite it to BeO. Magee and Gordonz3 reported that because of its small ionic

Wash solution used, ml 50 100 150 200

Wt. of complex after washing and drying, g

Loss, g

0.3402 0.3305 0.3278 0.3221

0.0126 0.0223 0.0250 0.0301

Determination

1000 7

of beryllium

205

DTG _

5,, , , ,\ IO

20

30

40 Time.

50

60

70

60

Fig. 1. TG, DTG and DTA curves for 200 mg of Be20(C9H.sN0)2.2H20. no units for DTA and DTG. beryllium oxide. Our results confirm that the dihydrate does not pass through a monohydrate stage on heating on the thermobalance, but also show that a thermally stable monohydrate can be formed by moderate heating under reduced pressure but decomposes on heating at about 240” and atmospheric pressure. This monohydrate was characterized from its elemental analysis, infrared spectrum and TG and

90

loo

min

DTA sensitivity l/10;

DTA data. The OH band at 3315 cm-’ was observed but other OH modes could not be assigned unequivocally because of the absorption bands of the ligand. On heating, the compound shows a loss corresponding to removal of water of hydration, accompanied by decomposition of the complex, suggesting that the water is co-ordinatively attached, probably in the form of hydroxyl bridges. Beryllium compounds and

206

A. M. HUNDEKAR,P. UMAPATHYand D. N. SEN

I

I

I

I

I

I

I

1

I

I -

DTG

60-

70-

SO-

9OI

IO

20

30

40 Time,

SO

60

70

Fig. 2. TG, DTG and DTA curves for 200 mg of BezO(CgHsNO)z.H20. units for DTA and DTG. chelates containing such hydroxyl bridges have been reported.25-28 Thqir presence in the monohydrate is

confirmed by p.m.r. studies. The p.m.r. spectrum of the compound in trifluoroacetic acid (Fig. 3) shows a broad absorption centred at -3.477 arising from hydroxyl protons [aromatic protons are observed at l.Ck (2H and 4H) and at 2.177 (3, 5, 6, and 7H)] and furthermore, there is no change in the spectrum

90

90

100

min

DTA sensitivity l/10; no

after 48 hr (no exchange of the hydroxyl protons with protons of the acid group), indicating the probable presence of bridged hydroxyl groups. The hydroxylproton absorption disappears on addition of small amounts of DzO to the sample, indicating that the monohydrate reverts to thd dihydrate complex (which presumably contains no hydroxyl bridges) so the structure is (A) rather than (B). This is in agreement

20-l

Determination of beryllium HZ 400

300

f

1

200 I

100 I

0 I r-H*

2.177 D

8

1

1

,

,,iII‘,111I11Ilt,l.ltll.~~lllllt~~l~~l~~~.~,

6.0

7.0

6.0

5.0

4.0 rmn.

3.0

2.0

I.0

0

8

Fig. 3. Proton magnetic resonance spectra of Be,0(C9H,NO), . Hz0 in trifluoroaeetic acid (A) immediately after preparation, (B) after 24 hr, (C) after 48 hr, (D) after deuteration. with the opinion

expressed by Sastri and Prasad.24

t%o t Ox 2e

Be -

OX~,/~~?&ZZ-OX

O--BeZOx

YOH/

t

H 0 *

2

“20

(6)

(A)

These studies indicate that the monohydrate have the structure: OxIZ%Be

/OH\

Ee”C

may

ox

\OH/

The TG and DTA data show that the dihydrate is thermally stable up to 150-160”; accelerated weight loss occurs at 250-500”. Formation of Be0 is complete at 840”. The weight loss at 200-220” suggests volatilization of one mole of water per mole of complex. Further loss of water on continued heating results in decomposition of the complex. The decomposition pattern for the monohydrate is similar to that for the dihydrate and shows that the compound, once it has been prepared as described, is stable up to about 240”. The results for analysis of beryl (Table 1) clearly show that of the reagents examined only 4-chloro-2,5dimethoxyacetoacetanilide gives consistently satisfactory results. Prior removal of Al(II1) and Fe(III) is not required. However, in the analysis of Be-Cu alloy, since the reagent precipitates copper(H) at pH 6.fV7.3, prior separation of copper is necessary. Cupferron, salicylaldehyde-ethylenediamine, tannin,

2-methyl-S-quinolinol and sodium oxinate give quite satisfactory results in the determination of beryllium from pure Be(I1) solution (Table l), but unfortunately these reagents are not selective. Masking with EDTA gives a certain degree of selectivity but prior separation of some elements is still necessary. We found tannin, acetoacetanilide and cupferron to give rather unsatisfactory results for analysis of beryl and Be-& alloy. Beryllium acetoacetanilide, although crystalline and thermally stable, is markedly soluble in hot water; the reportedz9 experimental error is 3%. Beryllium cupferronate is soluble on heating, the reagent is sensitive to light, heat and atmospheric oxygen, and ignition to Be0 is necessary. REFERENCES 1. J. Das and S. Banerjee, Z. Anal. Chem., 1961, 184, 110. 2. E. S. Przheval’skii and L. M. Moiseeva, Zh. Analit. Khim., 1960, 15, 117. -3. Liu Shao-Ling and Yin T’an-Cheng, Acta. Chem. Sinica, 1962, UI, 20; Re$ Zh. Rhim., 1963, 22639. 4. E. Uhlemann and R. Fritzsche, Z. Anorg. Allgem. Chem., 1964, 321, 19. 5. E. S. Przheval’skii and L. Moiseeva, Vestn., Mosk. Gos. Univ., Ser. Mat. Mekh., Astron., Fiz., Xhim., 1959, No. 1, 203. 6. J. Das and S. Banerjee, Z. Anal. Chem., 1962, 189, 183. I. B. R. Singh and S. Kumar, Indian J. Chem., 1912, 10, 663. 8. C. L. Sastri, G. Sriramulu and B. S. V. R. Rao, J. Sci. Ind. Res. India, 1955, 14% 171. 9. K. Motojima, Nippon Kagaku Zasshi, 1956, 77, 95; Chem. Abstr., 1957, 51, 14474. 10. Idem, BuII. Chem. Sot. Japan, 1956, 29, 29.

208

A. M. HUNDEKAR, P. UMAPATHY and D. N.

11. A. M. Hundekar, P. Umapathy and D. N. Sen, J. Indian Chem. Sot., 1975, 52, 71. 12. A. I. Vogel, A Textbook of Quantitative Inorganic Analysis, 3rd Ed., p. 517. Longmans, London, 1962. 13. J. Hure, M. Kremer and F. Le Berquier, Anal. Chim. Acta, 1952, 7, 37. 14. R. G. W. Hollingshead, Oxine and its Derivatives, Vol. II, p. 548. Butterworfhs, London, 1954. 15. W. 0. Foye and J. R. Marshall, .I. Pharm. Sci., 1964, 53, 1338. 16. W. F. Hiilebrand, G. E. F. Lundell, H. A. Bright and J. I. Hoffman, Applied Inorganic Analysis. p. 122 and references therein. Wiley, New York, 1953. 17. N. H. Furman (ed.) Standard Methods of Chemical Analysis, 6th Ed., p. 160. Van Nostrand, New York, 1962. 18. W. R. Schoeller and A. R. Powell, The Analysis of Minerals and Ores of the Rarer Elements, 3rd Ed. pp. 55-60 and references therein; Griffin, London, 1955.

SEN

19. L. Erdey, Grauimetric Analysis, Part II, p. 614. Pergamon, New York, 1965. 20. E. Halls, Ind. Chem., 1941, 17, 120. 21. V. G. Goryushina, Zavodsk. Lab., 1955, 21, 148. 22. P. Ray and A. K. Mojumdar, Z. Anal. Chem., 1935, 100, 324. 23. R. J. Magee and L. Gordon, Talanta, 1963, 10, 851. 24. M. N. Sastry and T. P. Prasad, ibid., 1967, 14, 481. 25. A. Kakihana and L. G. Sillen, Acta Chem. Stand., 1956, 10, 985. 26. B. Care11 and A. Olin, ibid, 1961, 15, 1875. 27. A. W. Thomas and S. H. Miller, .I. Am. Chem. Sot., 1936, 58, 2526. 28. R. M. Klein and J. C. Bailar. Jr.. Inora. Chem.. 1963. 2, 1187. and L. R. Batsanova, The Analytical 29. A. V. Novoselova Chemistry of Beryllium, Israel Program for Scientific Translation, Jerusalem, 1968. ,I”