I 846
LLOTD C. DAK1F:LS.
freezing point, both of which results coiilirtn views held or conclusions arrived at by others. 4. I t has then been proved t h a t the preparation of absolute sulphuric acid may rightly be based 011 the rninirnurii specilic electrical contluc~tivit>~ or on the inaxiniuni freezing point. j . I t has been shown t h a t for the analysis of 9 j per cent. or more coilcentrated sulphuric acid, the electrical conductivity yields results whose error never seriously exceeds o m i per cent., and is generall). less. 6 . There has been arranged a table of specific electrical conductilities of sulphuric acid which may be used for analytical purposes. Further work on the conductivity and other physical properties of sulphuric acid is now in progress. I t is also hoped t h a t sulphur trioxide may be more full!. investigated both a s to physical and chemical prolwrties. UNIVERSITY OF
MICHIGAN,A N N A R B O R ,
September, 1508.
-
[CONTRIBUTION FROM THE JOHN
HARRISON I,ABORATORY
OF CHEMISTRY.]
DERIVATIVES OF COMPLEX INORGANIC ACIDS ALUMINICOTUNGSTATES AND ALUMINICO-PHOSPHOTUNGSTATES. BY L L O Y D C. DARIELS. Keceived August 3 1 , rgah'.
Some years ago B a k e and Smith' described an ammonium and n silver ~ ~ . I4I~.Ig,O. ~~~ ammonium aluniinico-tungstate, 3 ( N € r , ) , ( . ) . , ~ l , o , . y ~ ~ 'and 2 I (NH,),O,qX1,0,.36WO,, in which the ratio of aluminum oxide to tungstic trioxide was I : 9. I t seemed well to prepare other salts to learn whether this ratio was preserved in a series of such derivatives. It was also desired to isolate the free aluminico-tungstic acid. ;I solution of i t was, indeed, obtained, hut upon evaporation, even at rooni temperature, decomposition occurred. \Yhen the solution was ncu tralized lvith ammonium hydroxide. a salt identical with t h a t from which the acid was liberated, reappeared. Aluminico-tungstic acid reacted acid to methyl orange, and in the cold displaced carbon dioxide from solutions of sodium bicarbonate. I t is soluble in j o per cent. or stronger alcohol, b u t on standing in such a solution the tungstic acid, or part of i t , is reduced, a blue solution, b u t no precipitation, resulting. 'This acid was prepared b y the method employed by Wolcott Gibbs and later workers, in liberating free coniples inorganic acids; i. c . , b y treating the mercurous salt with a little less than the amount of hydrochloric acid reqtiirecl to precipitate nll of the ntercurj1
2
From the author's thesis for tlic PI1.D degree IS
J O Y R N A L , 25, I??().
DERIVATIVES OF COMPLEX INORGANIC ACIDS, ETC.
1847
as mercurous chloride, shaking well, and filtering off the mercurous chloride, and the excess of the mercurous complex salts. Having established the stability of the ratio A1,0, to c)WO, in several salts, the introduction of pentoxides into the complex was made, obtaining new complexes. The analysis of complex inorganic salts has alltays been a matter of great difficulty. I t will be noticed how often dry gas niethods were used in this study to decompose the compounds. This was because they were much simpler on account of not requiring the admixture of large quantities of foreign matter. Tungsten compounds almost always require repeated fusions for complete decomposition, and sometimes a second repetition of a fusion with carbonate or acid sulphate is not enough. The greatest difficulty arising in the dry gas decompositions was the adherence of the residue to the glaze of the porcelain boat. This could usually be avoided by keeping down the temperature of the furnace to as low a point as possible. When any of the gaseous products of the decomposition were to be caught for analysis, as, for instance, the arsenious chloride in the arsenic compounds, the anterior end of the combustion tube was bent into a u, with a bulb at the bottom and one at the anterior side of the U.
Experimental Part.
Copfiel- A lzcnnkico- Tungstate, 2CuO.A 120,. 9 WO,. I 6 ‘l,H,O .- This salt was made in two ways: by adding to a I O per cent. solution of copper sulphate, a concentrated solution of anmonium aluminico-tungstate, and by precipitating the cupric oxide from an ammoniacal solution with ammoniunt aluminico-tungstate, ~(NH,),O.A1,O,.9~~O3. I n the first case, a slimy, “bird’s-egg blue” sediment forms at the bottom of the vessel. If the copper sulphate solution be diluted about six times, no precipitate, or very little precipitate, is formed. One or two drops of concentrated hydrochloric acid are sufficient to prevent precipitation, and acid of that strength even redissolves the precipitate on warming. This acid solution is, however, not of such a deep blue as the copper sulphate solution from which i t was made, b u t has an extremely pale blue tint. If the copper sulphate solution be warmed before precipitating, the copper salt formed is of much darker blue and is flocculent in appearance. Filtering this off while hot, more of the pale blue sediment separates from the filtrate on cooling. All attempts t o get the salt crystalline by slow evaporation resulted in the separation of the pale blue sediment. Whether precipitated hot or cold, the copper compound, on filtering off and drying in a vacuum desiccator, becomes an indigo-colored powder which dissolves in large amounts of water, and separates out in the pale blue form if evaporated slowly, or as the darker blue floccules upon evaporation on a water-bath.
I 848
LLOYD C . D A K I E I S .
. i n ainrnoniacal solutioii o!' cupric. o s i d c , oii addition ol' ;Ltlinioliilltn aluminico-tungstate solutioit gives I: (lark 1)lur flocculent preci1)itate indistinguishable b y c c h r froiii t Iicl 1)rccil)itatc fortilet1 i l l 11ot col)l)cr sulphate solutions. This 1)recil)itatc is p r e s e r \ ~ di l l very tlilute solutioiih. l it is of tile siiliic Filtered off, washed with ( I to 3 ) aiiiiiioiiiaq ~ i t tlrietl, color as the copper conipouiid from :t co1qwr su1l)hatc solution. 1 t :ilso shows the same reactions on dissolving i i i much ivnter u i t l cixporatiiig slowl!~, or rapidly, as inentionetl above. 'I'he precipitate iliatle i i i this ivay, washed with ammonia, :ml tlrietl i l l i t \.~icuLiiii-tlesiccator, \\.:~s analyzed as follows: One gram of the sample was dissolved iti \vater to ivhich a little sulphuric acid had been added, and a stream 01 hydrogen sulphide passctl through tlie solution, until all the copper was precipitated. I:or fear a little tungsten might precipitate ivitli tlie copper sulphide, it vxs re dissolved in hydrochloric acid, the hydrogen sulphide boiled oii' and the solution made slightly aminoniacal, aiid Iiytlrogcn sulphide again passet1 through m t i l the copper was a11 dowil. 'l'his cupric sulphide was washed with water containing aiiinioiiiitiii sulphide awl ;itiitiioiii:i. I t \y:is dried, separated from the filter, the latter hunied in a weighed crticible into which the cupric sulphide \vas then placed, :tiid ignited to cupric oside. A drop of nitric ncitl was atldetl after cooling-, a n d tIie residual cupric oxide again ignited a i i t l iveigllc(1. The filtrate irorri the first cupric sulphide ~)rwipitntioii\vas evaporated i i i a platinum crucible and fusetl Ivitli acid soditim sulphate, tlie fused mass being extracted with hot ( I t o ,i) nitric acid to which a little hydrochloric acid was added. ?'he filtrate froni tlie tungsten trioxide so formed was analyzed, as usual, for aluiiiinitiin. 'l'hc alumina was fused with sodiuiii bisulphate to remove any tungsten trioxide adhering to i t . I t \\';is then reprecipitated a n d reireighecl. ..Ihalf gram of the sul)stance heated oii a weighed ivatch-glass t o 20oo in ari air-bath gave the water content. -4 portion, tested in a qmiititative \vay for ammonia, showed none. Calculated: zcUO, 6.01; A1,0,, 3.86 ; 9\1'0,, 78.94; 16%H,O, 11.19. Found: zCUO,6.07; :112()3, 3.80, 3 , 8 ? ; 9\VOa, ;8.;1;IG):H,O, 11.39. WO, is by difference.
R a ~ i z w ~. ~ ~ Z ~ ~ ~ ~ ~ } Z ~ C [ ) -81~uO.rll,O,.pI.f~O~,.~ZI,~~.--This ~ ' ~ [ ~ ~ ~ ~ , \ ~ ~ ~ ~ ( , , conpound is a white powder precipitated from barium chloride solution by ammonium aluniinico-tungstate. It is so insoluble t h a t dilute sulphuric acid gives hut a very slight precipitate in the filtrate froni it. I t is not, however, precipitated from a solution of barium chloride contailling more than traces of hydrochloric acid, b u t acids do not dissolve it after drying. 'I'lie dried powder is somewhat dccoml)osed on ljoiliiig
DERIVATIVES OF COMPLEX INORGANIC ACIDS, ETC.
I849
ten or fifteen minutes with concentrated hydrochloric acid, nitric acid or aqua regia. Complete decomposition of this compound for analysis proved difficult. Three successive fusions with sodium and potassium carbonate mixture failed to remove the tungstic acid completely, and an accidental test of one of the filtrates from the barium carbonate showed the presence of barium in considerable quantity. Abandoning the alkaline fusion, an effort was made to separate the barium as barium sulphate, by fusing with sodium acid sulphate and leaching with dilute acid to remove aluminium, and then with ammonia to remove tungsten trioxide. This also proved fruitless, since two fusions failed to remove as much as half of the tungstic acid. Alumina also remained in the bariuni sulphate. Next, i t was sought t o remove the tungstic acid by dry hydrochloric acid gas, and it was found t h a t heating to a red heat in the gas does remove a large part of the tungstic acid, but not enough to give anything like quantitative results. By first reducing in dry hydrogen, and then passing dry chlorine at a low temperature over the boat, the tungsten is removed as oxychloride with ease and almost quantitatively, in one operation. One repetition of this treatment leaves a pure white barium chloride alumina mixture in the boat, and if it has not been heated to too high a temperature the boat will not be stained. The barium chloride may then be leached out with water and the alumina weighed. On fusing the alumina with sodium bisulphate, no traces of tungstic acid were found. Barium sulphate was also absent. It was also (observed t h a t hydrogen sulphide passed over the barium aluminico-tungstate in a boat, in a combustion tube heated very gently, converts the tungsten trioxide to tungsten trisulphide, or partly to a lower sulphide depending on the temperature. The contents of the boat treated M ith dilute hydrochloric acid yield barium chloride free from alumina, and almost free from tungsten trioxide, the traces of the latter being removed by evaporating with hydrochloric acid and taking up with the same dilute acid. The tungsten sulphide is removed from the alumina by an ammonium sulphide solution. This did not give very good results for alumina, but checked well with the method mentioned above for barium. Calculated: 8Ba0, 34.64 ; A~LO,, 2.89 ; 9WO8, 58 87 ; 7H,O, 3.56 Found: 8Ba0, 34.80, 34.76; A1,0,, 2.82, 2.73; gWO,, 58 77, 58.90; 7H,O, 3.61 WO, is by difference. The first results are by the first method, the second by the second.
Mercuroits .4luminico-Tungstate, 5Hg20.A1,0,.pT/t’0,. - When precipitated in the cold, this is a very pale yellow precipitate of a flocculent nature, much like silver bromide. It darkens somewhat on boiling,
1850
LLOYD C . DANIELS.
and becomes a powder with the lemon-yellow color of tungsten trioxide on drying. It is much more soluble in water than the barium salt, but is not prevented from tieing precipitated b y the presence of a considerable amount of nitric acid. I t is not redissolved to any considerable extent by the addition of 3 cc. ol' concentrated nitric acid to 50 cc. of the supernatant liquor. I t is, however, solulile in ( I to 5) nitric acid. The mercurous oxide in this coinpouiid was determined by precipitation as iiiercurous chloride from :L sulphuric acid solution. The precipitate was dried a t I o j c to I i o 0 . and weighed on a tared filter. The mercury and water were driven off together a t a low red heat, and the value of mercurous oxide subtracted from the loss in weight gave the water content. The alumina \vas determined as in the copper salt. Calculated: jHg,O, 48.25 ; -41203, 2.9j ; gTVO,, 48.40 Found: 5Hg,O, 48.42, 48.38; .41,0,, 2 . 8 2 , ?.8o; 9n'O,, 48.76, 18.82 M'O, is b y difference.
Z i m .-1lu1iiitiico-7'lc?lg~~tut(', x r / 2 Z ? ~ 0 . d 4 1 2 09I/VO,, 3. 8H,O; %no.A1203. 9bt7o3.20/I2O..--In a solution of zinc oxide in ammonia a solution of ammonium aluminico-tungstate produces a line white powdery precipitate. This is soluble in much water lree from ammonia, and evaporation, slow or rapid, yields a light flocculent residue. The precipitate is inimediately soluble in extremely dilute mineral acids or acetic acid. I n hot concentrated mineral acids i t dissolves to a light yellow solution which deposits tungstic acid 011 boiling awhile or upon standing in the cold. While attempting to dry on a water-bath, the mass was melted t o a yellowish white oil, a white residue floating 011 its surface. On continued heating on the bath the oil dried u p to a vitreous mass, easily friable. On treating this mass with water, the white portion remained undissolved, while the glassy portion went into solution in very little water. 'l'he white residue, iiltered off, pro\,ed to be nearly pure zinc , no more zinc oxide, sepaoxide, while the same yellon. oil :IS l ~ f o r e but rated out. zinc sulphate solution, even saturated, gives no precipitate with amnioriium alurninico-tungstate unless ammonia be added. But a very concentrated zinc chloride solution on treatment with a concentrated solution of amnionium aluminico-tungstate leads to the separation of a yellow oil which settles and sooii solidifies a t the hottom of the vessel in an amorphous mass. l'ouring off the excess of zinc chloride and rinsing quickly with a little water two o r three times, the residue is dissolved in a little water. I t appears to be in ever!. lvay like the soluble zinc salt obtained by the other method. .Inalysis proved this to be the case, but it \vas difficult to reniove all of the zinc chloride and still have enough of the aluminico-tungstate to work with. The zinc was determined both in the soluble and insoluble compounds
1851
DERIVATIVES OF COMPLEX INORGANIC ACIDS, ETC.
b y precipitation from acetic acid solutions with hydrogen sulphide. The zinc sulphide'was pure white, and contained no sulphides of tungsten, though the solution took on a distinct bluish tint, showing partial reduction of the tungsten trioxide. The other constituents were determined as in the analysis of the copper compound. INSOLUBLE COMPOUND. Mol. wt. 1%
I
9 8
ZnO.. . . . . . . . . 1 2 2 A1,0,. . . . . . . . . 1 0 2 WO,. . . . . . . . . . 2088 H,O.. . . . . . . . . 144
Found.
Calc.
4.97 4.13 85 .05 5.87
5.09 4.10
5.11
4.04
.... ....
84.85
5.96
2455
SOLUBLE COMPOUND. Found. Mol. wt.
ZnO.. . . . . . . . 8 1 . 4 I A1,0,. . . . . . . . 102 9 WO,. . . . . . . . . 2088 20 H,O.. . . . . . . . 360 I
Calc.
I.
la.
11.
3.09 3.88 79.36 13.68
3.20 3.92 79.37 '3.51
3.14 3.95
3.28 3.83
.... ....
....
....
2631
The results in column I1 are from the sample made from zinc chloride. It did not have the same water content as the other sample, but the zinc oxide and alumina contents are calculated up to this water content. Both these samples were merely air-dried. The sample I1 gave a distinct test for chlorine. Ammonium A luminico-Phosphotungstate, 9 (NH,) ,O.aA l,O,.4P2O,. 9 WO,. qH,O.-When a concentrated solution of orthophosphoric acid, or a solution of ammonium phosphate, containing excess of ammonia, is added to a solution of ammonium aluminico-tungstate, a white powder separates and the solution becomes slightly yellow in color. The white powder is an ammonium aluminico-phosphotungstate Lnd can be purified by solution in a large quantity of water. I t is more soluble in hot than in cold water, b u t the difference is not great. After separation from solution i t redissolves with some difficulty, requiring long boiling and more than enough water to make a saturated solution; nevertheless, i t dissolves completely. The mother liquor from which this compound separates probably contains one or more other aluniinico-phosphotungstates since qualitative tests show the presence of the three constituents, b u t no definite conipound could be obtained from it by slow evaporation. If phosphoric acid is used, the liquor is acid after the white powder has deposited. More of the white powder may be obtained on the addition of ammonia
LLOYD C . DASIELS.
1852
in excess. Sil\.er nitrate adtlecl to the liquor, iii ai1 attempt to free it from excess of phosphoric acitl, :tilt1 t o lcarti iC some rcci-ystallizable silver salt or salts could he oht:tiiir>(l,1)rcc’iliitatctl dniost the entire mass, and attempts to dissolve out aiicl recr!,stallize ivere not successful. The whole precipitate was soluble in ainnioiiia. The ai-rinioniuni aluiniiiico-l,liosl,hotuiigstate obtained above was recrystallized and gave the follo\ving tests : LVith silver nitrate, a yellow precipitate, soluble in ammonia, insoluble in acetic acid, b u t soluble iri dilute iiitric acid. This precipitate was too nearly insoluble iii water t o :illon. of crystallization, and attempts to crystallize it from aniinoniacal solutioii gave negative results. Direct sunlight darkens it sloivl>- t o a deep brown. An analysis of it is appended. LVith copper sulphate, a n apple-green precipitate, beconling pale blue on heating, or when the precipitation occurs in hot solution. Deep blue precipitate froni aiiiinoriiacal solutioii, insoluble in large excess of atnmoiiia. Both are soluble in dilute nitric acid. Barium, calcium, zinc ant1 cxliiiiuni salts give white powdery precipitates, soluble i n dilute acid. ilnalysis: Thc amiiioniuni oxide and water were determined as usual. The compound heated to incipieiit redness in a boat, iii a stream of hydrogen sulphide, carrying vapors of carbon bisulphide from having been passed over the surface of cold carbon bisulphide in a flask, gave all the tungsten as sulphide, and the alumina and phosphoric acid were dissolved out with acid. I t is notable t h a t the alumina here is soluble in moderately dilute hydrochloric acid, whereas it dissolved little, or not at all, in the case of the ulumiiiico-tungst3tes, showing t h a t the acid must be in direct union with the alumina in the comples, or t h a t it is in such close contact t h a t 011 separation of tungstic acid i t goes into combination. ’The alumina atitl phosphoric acid were separated by inolybdate solution in one-half of the scilutioii, aiid the ptiosplioric :tcicl determiiied as niagnesiuni pyro1)11osphate, ’1‘0 tlic other l i d i of ilie solution a little disodium phosphate and acetic acid and sotlitiiii acetate were added, aiid the aluininium phosphate filterctl, washed with acetic acid. arid weighed as d u niinii.iiii phosphate. Mol. wt.
9 (NHJ20 . . . . . . . 468 2 .4l2OJ. . . . . . . . . . m.+ 4 l’20c,. ......... jOY \VO2.. . . . . . . . . 2 0 8 8 1 ~ 3€I2(). . . . . . . . . ?,\.+
Calc. 1.7.
11
,iT 2 15
05
,58.(1.! (1
57
---
13.30 5.55 l j
.G6
5 0 .07 6.3‘)
Found.
13.24 5.62
1.7’73 59.03 (6.39)
\ V o 3 is by dirierencr.
. % ~ / w I . ~ ~ ~ ~ / ~ ~ i ~ ~ i ~ [ j - f ~ / /3-Alr?(,._..41,(,,,.4~”,0,.gIT.0,.6tl,O.--~).\~~//[if~~?~!/.\~[~~~,, The silver \vas tleterniinetl b y precipitatitig lroni dilute nitric acid solu-
DERIVATIVES OF COMPLEX INORGANIC ACIDS, ETC.
18.53
tion with hydrochloric acid, and the alumina and phosphoric acid as in the ammonium salt, after passing hydrogen sulphide over the compound resting in a boat. Mol. wt.
. . . . . . . . . . . . . . 928 A1,0,. . . . . . . . . . . . . . . . 204 P,O,. . . . . . . . . . . . . . . 568 WO, . . . . . . . . . . . . . . . 2088
4 Ag,O.. 2
4 9 6 H,O.. . . . . . . . . . . . . .
128
Calc.
Found.
23.70 5.21 14.50 53.33 2.26
23.82 5.06 14.29 53.67 3.20
B a r i u m tllzimi,rtico-Yhosphotungstat~, qRa0.aA1,0,.qPz0,.91il'0,..~~~~.This is a powdery white precipitate formed on adding barium chloride solution in excess to a solution of the ammonium salt of the complex. I t is very slightly soluble in water, b u t very dilute hydrochloric or nitric acid dissolves i t immediately, even after drying. The sample used for analysis was washed well with ammonium chloride solution to remove any basic barium salt t h a t came down with it, and then with water. The barium oxide was determined o&e by direct precipitation as sulphate from an acid solution of the complex salt, and, on the other two occasions, in the acid solution resulting from the separation of the tungsten trioxide, as tungsten sulphide, as in the ammonium and silver salts. The alumina and phosphoric acid were determined as in the ammonium salt of the complex. Found. Mol. wt.
4 BaO.. . . . 564 2 A1,0,. , . . 204 4 P,O, . . . . . 568 9 WO,. . . . . 2088 54 3 H,O.. . . .
Calc.
I.
11.
16.02 5.86 16.03 60.56 1.55
16. 28
15.98 5.79 15.91 60.75
..... ..... .....
.....
.....
15.90 5.93
15.88 60.72 (by diff.) . . . . . 1.67
3478
Zinc Alumin ico - Phosphotungstate, ,5Zn0.2A1,O3.4P,06.pW0,. IIH,O. -This salt is a white powdery precipitate, soluble in dilute acids and in a large quantity of concentrated ammonia when ammonium chloride is present. From this solution, by spontaneous evaporation in the air, hard, semi-transparent scales deposit. The zinc was determined as sulphide in acetic acid solution of the complex and then converted into oxide with dilute nitric acid. I t was contaminated with some tungstic acid which separated on evaporation in the crucible, so after baking it was taken up with strong nitric acid and a drop of hvdrochloric acid, and diluted with two volumes of water. The tungstic acid having been filtered off, the solution was again evaporated, the residue ignited, and the zinc oxide weighed. The alumina and phosphoric oxide u-ere determined as in the analysis of the ammonium salt.
1854
L L O \ ' D C. DANII
