545.15 : 546.33

THE GRAVIMETRIC DETERMINATION OF SODIUM IN THE FORM OF TRIPLE ACETATE

BY

N. SCHOORL.

Attention is drawn to the possibility of the quantitative determination of sodium gravimetrically as the triple acetate with Mg(2n) and uranyl.

A historical survey of the development of these methods is given. In view of the peculiar behaviour of the water of uystaMization in the

triple acetates a possible analytical error of 0.25% is pointed out, whereas the latter is restricted to 0.1 yo by K a h a n e ' s method (reagent with 45 yo by vol of alcohol).

Improved directions for K a h a n e ' s method are given, in which two procedures are indicated, one in the presence of little and the other in the presence of more potassium. The analytical errors, which can be applied as corrections, are indicated for both procedures.

I . The triple acetates of sodium with magnesium or zinc and uranyl of the composition NaMg[ Zn] (UO,), . Ac, . 6 H,O are very advantageous forms in which to weigh sodium for quantitative determination, since their molecular weight 1496.5 and 1537.6 amount to 65 and 67 times respectively the atomic weight of sodium.

They are definitely crystalline and can readily be collected by filtration but for quantitative determination have the apparent dis- advantage of an appreciable solubility in water. This is approximately 5 % 1). This disadvantage can however be overcome by mixing the sodium salt, dissolved in a small volume of water, with a fairly large excess of reagent (solution of magnesium- or zinc- and uranyl acetate), whereby as a consequence of the high concentration of Mg (or Zn) ion, UO, ion and acetyl ion and in view of the ion product which consists of 14 factors, the concentration of the Na ion in the mother liquor is depressed so greatly that practically all the sodium disappears from the solution. This is at any rate the case when one gives the

1) I found for the solubility a t summer room temperature of the magnesium and the zinc triple acetate 4.5 76 and 6 yo respectively. V. E r b, (2. Krist. 19, 284 (1891) mentions the solubilities 3.82% and 4.85% resp.

LIX 26

730 N. Schoorl.

crystallization sufficient time at a temperature of not more than 20' C. The precipitates are insoluble in the reagent and also very little

soluble in strong spirit (96 vol. 9%) and in acetone2) with which they are successively washed to be weighed subsequently in the air- dried state. They also stand heating to 110' C without any alteration of composition since the water of crystallization does not escape even then.

2. S t r e n g (1886) 3 ) was the first to prepare the triple acetate of sodium, magnesium and uranyl and recommended it as a micro- chemical test for sodium. He believed that the crystals contained 1.48 % Na and gave the formula NaMg (UO,) . Ac, . 9 H,O.

B 1 a n c h e t i t r e (1923) 4 ) was the first to make use this triple acetate for the quantitive determination of sodium, but his method has had little success through various errors in the instructions. He believed incorrectly that the separation of the microcrystalline precipitate was complete in half an hour and he assumed that it became anhydrous by drying at 105' C. By the compensation of these gross errors (the latter amounts to almost 8 %I ) a correct result could indeed be obtained but many writers after B 1 a n c h e t i t r e have complained of the bad results with his method. Doubt has consequently been expressed as to the constant relation of the components in the reaction product.

This doubt has been satisfactorily settled by C a 1 e y and F o u 1 k (1929) 5) . They proved by the analysis of precipitates prepared under all kinds of different conditions that the composition corresponds in every case to the relation NaAc . MgAc, . (UO,),Ac,. They determined all the components in the precipitate directly quantitatively,

") 0. R e i c h a r d, (2. Untersuch. Lebensm. 71, 501 (1936)) found for the

in water 7.6 g in absolute alcohol 0.034 g

in 50% spirit 0.72 g in acetone-alcohol 0.008 g in 75% spirit 0.21 g in reagent a t 20" 0.054 g in 96% spirit 0.02 g in reagent a t 0' O.OO0 g

Practically in agreement with the above I found for the solubility a t 20" C in strong spirit (96%) 17.5 mg for the magnesium- and 34 mg for the zinc triple acetate per 100 c m s of solvent.

The solubility in the reagent can be zero when the latter is saturated with the triple acetate a t 20" C (see the preparation). 3) A. S t r e n g. Ber. d. Oberh. Ges. f. Nat. u. Heilk. 24, 56; 2. wiss. Mikroskop.

3, 129-130 (1886). 4 ) A. B l a n c h e t i e r e , Bull. soc. chim. [4], 33, 807-818(1923). 5) R. C a l e y and C. W. F o u l k , J. Am. Chem. Soc. 51, 1664-1674 (1929)

magnesium triple acetate per 100 cms solvent (temp. not stated)

in 40% spirit 0.75 g in acetone 0.004 g

The gravimetric determination of sodium in the form, etc. 731

with the exception of the water of crystallization since this does not escape even at 110" C, whereas at a still higher temperature (125- 130') decomposition already begins to occur with the evolution of acetic acid. They found 6% H20 for the water of crystallization by deducting all the other constituents.

They arrived at satisfactory quantitative determinations of sodium on the basis of the above given formula, with the following stipulations:

a considerable excess of reagent to be added to the sodium salt dissolved in a small volume of water:

either after repeated shaking and reversal during the beginning of the precipitation, allow the liquid to stand a further 24 hours or vigorous (mechanical) shaking or stirring for 30-40 minutes:

the temperature not higher than 20' during the precipitation: the precipitate to be collected in a G o o c h filter crucible, washed

with 5 cm3 strong spirit (95 %) each time, whereby for each 5 cm3 1 mg must be added to the precipitate as a correction on account of the solubility of the precipitate and finally drying for W hour at 105-110° C.

Meanwhile M i h o 1 i c (1920) 6 ) was the first to find six molecules of water of crystallization in the magnesium triple acetate while B a r b e r and K o 1 t h o f f (1929) 7) replaced magnesium by zinc 8 ) and also assume the formula NaZn( UO,), . Ac, . 6 H,O for the precipitate with which their determinations agree. Their method has perhaps become the most familiar one, although the replacement of magnesium by zinc has not been found to be quite necessary (see above: C a 1 e y and F o u 1 k) and magnesium is much more general as the accompanying metal for sodium than zinc.

K a h a n e (1930) 9) has improved B l a n c h e t i k r e ' s method in a quite different way, namely by adding alcohol (45 vol. %) to the reagent. He believes he obtains the following advantages thereby: reduction of the solubility of the reaction product: improvement of the velocity of crystallization, whereby quantitative precipitation is attained in % hour, when one shakes occasionally: non-attachment of the

S. M i h o 1 i c, Bull. Acad. Sci. Zagrat. 1920, 1 6 2 3 ; c 1921, W. 683. He did not actually succeed in the quantitative determination of sodium and believed that the triple acetate would be inconstant in composition.

?) H. B a r b e r and I. M. K o l t h o f f . J. Am. Chem. SOC. 50, 1625-1631 (1928); also Chem. Weekblad 26, 294-298 (1929).

See also B a r r e n s c h e e n and M a s s i n e r . Biochem. 2. 189, 3 0 0 3 1 3 (1927), who employ the zinc triple acetate for the quantitative determination of sdiurn.

O) C. K a h a n e , Bull. SOC. chim. [4 ] , 47, 382-404 (1930).

732 N. Schoorl.

crystals to the wall of the glass: decrease of the necessary excess of uranium salt. For the composition of the precipitate under the conditions of the

precipitation (at least 2.5 cm3 of reagent to at most 1 mg Na, dissolved in at most 1 cm3 water, so that the alcohol concentration becomes 32 vol. %) K a h a n e now finds NaMg( UO,),Ac, . 8 H,O, also after drying at 110' C and indeed on the basis of the determination of the residue on ignition which consists of % Na,U20, + MgU,O,.

This method of K a h a n e has been applied with success to agricultural products by v a n K a m p e n and W e s t e n b e r g (1932) 10). in which much potassium is previously removed as perchlorate if necessary.

3. I have studied the water of crystallization in the triple acetates in more detail 11). One can determine the amount of this quantitatively in a relatively simple way by drying at 80' C, in the presence of phosphorus pentoxide in vacuo to constant weight and then weighing the residue with due precautions.

It was then found that the precipitates of B l a n c h e t i e r e (magnesium triple acetate) and of K o 1 t h o f f (zinc triple acetate) contained the following amounts of water, expressed in percentages and in mols and as a function of the relative water vapour-pressure indicated at the time of the determination.

Magnesium triple acetate

Na Mg(U0Zh. Acq = 1388.5 Relative

humidity

-

0 0.34 0.68 0.74 0.98

Zinc triple acetate Na zn(U023. Acq = 1429.6

7.99 8 25 8.39 8.61 8.79

I

6.17 7.77 6.17 6.36 8.12 6.45 6.47 8.22 6.53 6.64 8.35 6 64 6.78 8.55 6.80

O/O water calc. I 0/0 water calc. I rnols. water mols. water on dry subst. I I on dry subst.

The triple acetates thus possess not exactly 6 molecules of water of crystallization but somewhat more; this excess is dependent on the humidity of the atmosphere in which the salt is weighed.

10) B. v a n K a m p e n and H. W e s t e n b e r g , Verslag, Landb. Onderz.

11) N. S c h o or 1, Rec. trav. chim. 59, 305 (1940). NO. 38. E, 21-29 (1932).

The gravimetric determination of sodium in the form, etc. 733

When the water of crystallization is completely removed from the triple acetates as indicated above, the crystals were found not to desintegrate into a fine conglomerate but (under the microscope) to show a pseudomorph of the original shape. The 6 molecules of water bound with an unmeasurably small vapour pressure are thus "facultative water of crystallization" (case I1 of Chem. Weekblad 30, 361 (1933)). while in addition almost 1 mol. Aq more can be taken up at different vapour pressures as "zeolithic increment" (case IV, 1.c.).

It follows from this that in the "air-dried" state the weight of the triple acetates can exhibit a fluctuation of about 0.25 %. according as one carries out the weighing in very dry air (rel. humidity 0.15) or in very humid air (rel. humidity 0.85) and that the weight in air of average humidity corresponds very closely with the salts with 6% mols Aq.

The magnesium triple acetate, prepared by K a h a n e ' s method and in fact by adding a tenfold volume of reagent (with 45 vol. % alcohol) to a solution of sodium chloride in water, was found to contain not only water but also ethyl alcohol in the crystals. The weight of this K a h a n e salt in the air-dried state does not change by drying it at 105-1 lo', while the analysis by drying at 80' C in the presence of P,O, in vacuo and by determining the ethyl alcohol quantitatively gave:

O / O volatile o/o volatile matterlz) of which alcohol 4.8 O/o = apprcx 1.5 mols. 1 water 6.3 O,o = approx 5 mols. matter on dry subst. 10.0 11.1

It is remarkable that this K a h a n e triple acetate shows a much smaller zeolithic increment:

Rel. humidity 0 0.32 0.60 0.95

after 2 hrs. at 100°C then in the air

O/o volatile matter

9.95 10 0 10.05 10.15 9.85 9.95

Thus on weighing K a h a n e ' s magnesium triple acetate in very dry or very humid air there occurs no greater difference in weight than 0.1 %, while it can also be dried at 100' C or for % hbur at 110' C without changing more than 0.1 % in weight.

I*) This content of "volatile matter" would correspond to 8.5 mol. Aq while K a h a n e (1.c.) also has hesitated whether to attribute 8 or 9 mol. Aq to the salt.

734 N. Schoorl.

4. I have therefore given the preference to the precipitation of sodium by K a h a n e ' s method to be carried out according to the following directions:

Reagent ( K a h a n e ) : 100 g cryst. magnesium acetate (MgAc, . 4 H,O) , 32 g cryst. uranyl acetate (UO, . Ac, . 2 H,O) . 20 cm3 of glacial acetic acid are dissolved in 300 cm3 water, then 500 cm3 spirit (90 vol. % ) or 470 cm3 strong spirit (96 vol. % ) added and made up with water to 1 litre.

The solution is allowed to stand for several days at a temperature of 20' C to clarify and the top liquid is then poured off through a filter. Since the reagents employed contain sufficient sodium as impurity, the sediment consists of triple acetate, with which the reagent is thus saturated at 20' C.

It contains 45 vol. % ethyl alcohol and is preferably stored in a Jena glass flask and must not be subjected to direct sunlight.

QUANTITATIVE DETERMINATION. Isf method in which at most 4 mg Na (10 mg NaCl) may be

present and at the same at the most 1 mg K (2 mg KCl). The neutral aqueous solution is evaporated, finally on a water-bath

in a 25 cm3 wide necked E r l e n m e y e r flask, to 1 cm3. After cooling, it is mixed with 10 cm3 of reagent, reversed for some minutes as soon as the crystallization begins, later shaken from time to time and then allowed to stand until the next day.

The crystalline precipitate is then collected on a glass filter crucible 3G3, during which the filtrate is used to rinse the precipitate quantitatively from the flask, washed twice with 2 cm3 of reagent, twice with 2 cm3 of strong spirit and finally once with 2 cm3 acetone. In each case the precipitate i s each time covered with the wash-liquid and this is removed after 5-10 minutes by gentle suction. The filter crucible with the precipitate is dried in the air or for 5-15 minutes in an oven at 100-105". It is weighed by cooling completely.

There is a small negative error, which has appeared in determinations of accurately known amounts of sodium and by relating the result to the precipitate freed from water and alcohol (see above). This must be added as a correction to the result 13) as follows:

Wt. precip. Correction Wt. precip. Correction 261 mg 3 mg 100 mg' 0.5 mg 250 .. 2.9 .. 80 .. 0.3 ., 200 ,. 2.1 .. 50-60 ,, 0.2 .. 150 ,, 1.3 ., 20-30 ,, 0.1 .,

I*) One can interpolate linearly between the amounts of precipitate and the corrections indicated above.

The gravimetric determination of sodium in the form, etc. 735

2nd method in which at most 4 mg Na (10 mg NaCl) may be present as well as at most 10 mg K (19 mg KCl).

As in the first method the neutral aqueous solution is reduced by evaporation to 5 cm3, this is mixed with 5 cm3 of spirit (90 vol. %) and then with 10 cm3 of reagent. It is further treated in the same way as in the first method.

Here there is a larger negative error, which must be added as a correction 1 4 ) , namely:

Wt. precip. Correction Wt. precip. Correction 250 mg 10 rng 70 mg 0.5 mg 200 .. 7.3 ., 60 ., 0.4 .. 150 .. 4.5 .. 50 .. 0.5 ., 100 ., 1.6 .. 40 .. 0.6 .. 90 ,* 1.2 ., 30 ., 0.8 ,* 80 .. 0.8 11 20 .. 1.0 ..

10 9. 1.2 ,I

The course of the error curue (see fig.) requires some explanation. There is a constant negative error through the solubility of the

14) One can interpolate linearly between the amounts of precipitate and the corrections indicated above.

736 N. Schoorl.

precipitate, which is caused in the first method by the added 1 cm3 of water, in the second method by the added 10 cm3 of 45 vol. o/o spirit and increased in both cases by the wash-liquids. There is a positive error through occlusion and adsorption of mother liquor, proportional to the amount of precipitate, as a result of which the error curve in the second method (and possibly also in the first method at still smaller amounts of precipitate) rises to a maximum at 60 mg.

Finally there is a fairly considerable negative error through exhaustion of the reagent, which error, from about 50 mg precipitate onwards increases rapidly with larger amounts 16) of precipitate formed.

Calculation of the analysis.

K a h a n e * s triple acetate contains 10 96 of “volatile matter”( water and alcohol: see above) and 90 % NaMg (UO,),Ac, . ( = 1388.5). On this basis the ratio

-- - =0.01490 Na

weighed precip. + correction 67.12

-- - * =0.03787 NaCl

weighed precip. + correction 26.40

Disturbing factors. All neutral sodium salts can be used except the phosphate. Calcium and magnesium salts may be present but not lithium and not too much potassium (see above).

Phosphate. if need be, is removed by warming the solution with basic magnesium carbonate (magnesia alba) for an hour on a water- bath reversing repeatedly, then allowing it to stand over night, filtering, then washing and evaporating.

Much potassium (according to v a n K a m p e n and W e s t e n- b e r g, 1.c.) is removed with an amount of perchloric acid sufficient for the precipitation as KClO,, and the solution is evaporated on the water-bath. The KC104 is filtered off, then washed with a 1 % solution of perchloric acid, the solution is neutralized with magnesium oxide (magnesia usta) and evaporated to dryness, the residue is taken up in water, filtered and if necessary evaporated again.

la) The amount of uranyl (the constituent which is first exhausted) in 10 cma of K a h a n e * s reagent corresponds theoretically with 5.75 mg Na (14.6 mg NaCI) in the form of triple acetate.

The gravimetric determination of sodium in the form, etc. 737

Applications. The first method can in general be applied directly. to drinking water. Also to the ash of body fluids (remove phosphate) and anywhere where sodium greatly predominates over potassium. For botanical material, where more potassium than sodium is present,

the second method can frequently still be of service.

U t r e c h t, Pharmaceutical Laboratory of the University, Jan. 1940.

(Received Febr. 7th 1940).