TUNGSTATES A N D MOLYBDATES OF T H E RARE EARTHS.
483
they must differ in chemical intensity. But what is chemical action? hccording to one view, it is a change in which chemical intensities are equalized, with a conversion of part of the chemical energies concerned into other energy forms ; according to the other it is a union of atoms. But how can two things, which we suppose to be exactly alike, unite ? Whence comes the difference of intensity ? By virtue of what principle is the elementary diatomic molecule formed ? If this molecule always preserved its integrity, we could regard its diatomic constitution as an elementary condition, but according to our chemical theories, the molecule in many chemical reactions is subject to division. Nevertheless, whenever the atoms are set free they conibine according to the diatomic pattern. It is thus roughly shown that the later views of chemical energy are not supported by the molecular hypothesis. Which shall we throw overboard for the sake of the other, or may we hope for a view which will embrace, possibly revise, and reconcile both ?
THE TUNGSTATES AND MOLYBDATES OF THE RARE EARTHS.' BY F A N N YR . M. HITCHCOCK. Received March
T
9.
,895,
HE estimation of molybdic and tungstic acids, and their
separation from each other have been the subject of many investigations, but notwithstanding the time and care expended the results so far attained are not perfectly satisfactory, particularly as regards their separation. T h e reactions which take place when either element i s tested alone cannot be depended upon when the two are together, each seeming to exert some modifying action on the other. Several years ago, Smith and Bradbury carried out a series of experiments on the precipitation of these two acids, chiefly with salts of the heavy metals. Their research brought out many new and interesting facts relating to the precipitation and 1 From the author's thesis presented to the Faculty of Philosophy of the University of Penn'a for the degree of Doctor of Philosophy, 1896. 2J. A n a l . A j p I . Ciicni.. September, 2892.
484
F A N N Y K . 31. I I I T C H C O C K .
T H E TCNGSTATES
estiniatioii of each acid separately. but failed to produce anything iien. i i i the way of a separation. It n a s with t!ie hope that, 1)). extending the iiivestigatioii begun by them. to the actio11 of other precipitants on these acids, additional iiiforiiialead to more definite results, tion iiiiglit be g:ithered which ~~voultl that the esperiinents were uiitlertaken, the results of which are offei-eci i i i tlie following pages. The nioly!)date solutions used in the work were made up from the sotliurii iiiolybate. Xiinlysis showed the salt to be pure and to agree n i t h the formula given, Sa,hloO,f 2H,O. T h e tuiigstate solutions were prepared from the corresponding 2H,>O. Analysis slioived that traces tuiigsteii salt, h'a,\VoO, of silica n i i d ferric iron were present, and niolybdeiiuni trioxide \vas found to the aiiiount of three and one-tenth per cent. T h i s \vas separated b y Pechard's method.' T h e allowance has been made in all calculations for the amount of iiiol!-bdic acid found in the sodium tungstate. It \vas found liest not to prepare more than 2 j O cc. of the solutioii of m d i u i i i tuiigstate at one time, as on staiicliiig it graduall) affects tlie flask, the glass heiiig attacked, ant1 a sediment appearing wliicli resembles silica, and is of noticeable amount. T h e solution of sodiuin molylidate also attacked tlie glass, but not so strongly nor so quickly as tlie sodiuiii tungstate.
+
so 111 u xr TIJNG ST.\TE .\ s I ) 31A s G.\S ESE SL-L I' X I.\T I.:. T h e solution of manganese sulphate used was a saturated one. T h e solutioii of sodium tungstate contained 27.8600 grains to t h e liter, and, if conipletely precipitated by the manganese sulphate, ten cc. of the solution should yield 0.2871 gram of manganese t u iig st a t e , M 11\V 0 .
The precipitates all showed the same character, however varied the coiiditions under which they were formed, being flocculent at first and becoming slimy 011 stailding. They filtered very slowly and were difficult to n.ash. T h e filtrations were made through asbestos i l l a Goocli crucible, and the filtrates were all tested for tungstic acid by hydrogen sulphitie i n aniinoniacal solution, the solution being subseqiieiitly acidified with hydrochloric acid. . ('*m,f!t t n d 114. 1;s: % i l i h > .Ailorg. Cheni., I , 2 6 2 : this J O L X S A I . , 15, I . (1693).
A N D MOLYBDATES O F T H E R A R E EARTHS.
485
A . Experiment r.-Ten cc. of the sodium tungstate solution were diluted to 250 cc. with distilled water, brought to boiling, and fifty cc. of the manganese sulphate solution added. A voluminous precipitate formed at once, which was allowed to settle, then filtered off, washed with boiling water, dried in an air-bath a t 180' C., cooled in a desiccator for three hours, and weighed. Wt. obtained.
Wt. calculated.
0.2310 gram.
0.2871
Difference.
4.0561gram.
T h e filtrate gave a precipitate of tungsten trisulphide.
Experiment 2.-Ten cc. of the sodium tungstate solution were treated as before, but the precipitate was allowed to stand for two days, then filtered cold, and washed with hot water. Wt. obtained. 0.2779
gram.
Calculated.
Difference
0.2872
4.0092
T h e filtrate gave a precipitate of tungsten trisulphide.
Expen'ment 3.-Ten cc. sodium tungstate solution were treated as before ; the precipitate allowed to stand for twelve hours, then filtered cold, and washed with cold water. Wt. obtained. 0.2667
gram.
Calculated. 0.2871
Difference. -0.0204
Experiment #.-Ten cc. sodium tungstate solution were treated as before ; the precipitate was allowed to stand for twelve hours, then brought to boiling, filtered hot, and washed with a hot solution of ammonium nitrate. Wt. obtained.
Calculated.
Difference.
0.2110gram.
0.2871
-0.0693.
B. Experiment r.-Ten cc. sodium tungstate solution were diluted to 250 cc. with distilled water, and five grams of amnionium nitrate added ; the whole was then brought to boiling, and fifty cc. of the manganese sulphate solution were added; the precipitate was allowed to subside, filtered off, and washed with water containing ammonium nitrate ; the filtrate was perfectly clear. Wt. obtained.
Calculated.
Difference.
0.2351 gram.
0.2871
4.0520
Expen'ment 2.-Ten cc. sodium tungstate solution were treated T h e precipitate was allowed to stand over night
a s in B r .
486
F A N N Y R . &I. H I T C H C O C K .
T H E TUNGSTATES
before filtering, was filtered cold, and washed with cold water containing amnioniuni nitrate. W t olitainod.
Calculated.
0.2393 gram.
0.2871
Difference.
--0.0478
Both filtrates gave tungsten trisulphide. C. Exflevitiienf z .-Ten cc. sodium tungstate solution were diluted to 2 5 0 cc. with distilled water, and eighty cc. of ninetyfive per cent. alcohol \vere added ; then fifty cc. manganese sulphate solution. T h e solution was heated, and the precipitate forined slon.ly. I t was voluminous, white and flocculent, beconiing granular on standing ; it was filtered off while still hot, washed with thirty-three p e r cent. alcohol, and ignited a t a low temperature. W t . obtained.
Calculated.
0.2666gram.
0.2871
Difference.
-0.0164
Expevinienf 2.-Ten cc. sodium tungstate solution were treated as in C z. T h e precipitate was allowed to stand for several days before filtering, was filtered cold, and washed with thirty-three per cent. alcohol as before. Wt. obtained. 0.2707 gram.
Calculated.
0.287I
Difference. -0.OZOj
A brief inspection of the analytical results given in detail above shows beyond reasonable doubt that the precipitation of tungstic acid from its salts cannot be well hoped for when manganese salts are used a s the precipitants. T h e most varying conditions failing to give any promise of even approximate success, further experiments with manganese salts were abandoned, and attention was turned to uranium salts. Trials were made with uranium acetate, uranyl nitrate, and uranyl chloride. With uranium acetate no precipitate formed in an aqueous solution of sodiuiii tungstate, whether cold or hot, and an excess of either the tungstate or the acetate produced no effect. T h e addition of a few drops of acetic acid caused no change. Alcohol produced a slight opalescence, but though the solution stood for several days no precipitate formed. When uranyl nitrate was used a precipitate formed at once, which was pale yellow in color, flocculent, and quick to subside. T h e precipitation seemed to take place equally well in the presence of alcohol, and in that of ammonium salts.
A N D MOLYBDATES OF
THE R A R E EARTHS.
457
T h e solution of sodium tungstate used contained five grams in a liter, and ten cc. would therefore yield 0.0883+ gram of uranyl tungstate, if conipletely precipitated. T h e atomic mass of uranium was taken a s 239.6. A . Twenty cc. sodium tungstate solution were diluted with twenty cc. of distilled water and five cc. of a solution of uranyl nitrate added, the latter being in slight excess. This was followed by twenty-five cc. of a saturated solution of ammonium chloride, and the whole heated nearly to boiling for two hours, after which the precipitate was filtered off and washed first with water contailling aniniouiuin chloride, then with pure water. T h e filtration was made through Schleicher & Schull's 590 filterpaper, and the precipitate was ignited strongly with free access of air. Wt. ohtained. 0.1j28 gram.
Calculated.
Difference.
0.1767
0.0039
T h e filtrate gave no trace of tungsten trisulphide when tested with hydrogen sulphide, and from the appearance of the precipitate after ignition some reduction had evidently taken place. B. Twenty cc. sodium tungstate solution diluted with twentyfive cc. distilled water, were precipitated with five cc. of the uranyl nitrate solution. T h e precipitate was allowed to stand for one hour, was then filtered in the cold, and washed with cold water, on a tared filter which had been dried at 150' C. T h e precipitate was dried at the same temperature in an air-bath for two hours, and weighed a t constant weight. Wt. obtaitied.
Calculated.
Difference.
0.1770 gram.
0. I 767
f 0.0003
T h e filtrate, when tested, gave no trace of tungsten trisulphide. I t may h.e added that when the ammonium chloride used in the first experiment was replaced by alcohol, the precipitation seemed to be equally complete, as there was no evidence of the presence of tungsten when the filtrate was tested with hydrogen sulphide. To ascertain whether the tungstic acid could be determined volumetrically, the following experiment was made : h solution was prepared containing fourteen grams of uranyl nitrate in one liter of water. Twenty cc. sodium tungstate solu-
4ss
l:.\SSS K . ?,I. i I I T C I I C O C I ; .
T I I E TI.SGdT.\TES
tion ivere cliliitetl \vit!i tiventy-five cc. distilled \ v a t u , arid tlie nitrate solutiotl introduced from a hurette, the tuiigstate solution lxiiiq coiistaiitly stirred. and tlie end.reactiori ascertained 1,te-tiiig \\.it11 potassiuiii ferrocyaniclc oil n porcehiii plate. 'I'o\varcl the In.;t tlie reaction is rather slo\\.. so that a 1x-on.n crilor iiixy l ~ eol,t:iined \\'it11 the potassiuni fcrroc),ariitie before tlie tungsten is full). precipitated. If. hoivel-er. the solution be nllti~,vetlt o staiitl a fe\v secoiids aiter each adtlitioii of the uranyl iiitrate solution, 110 coloration is obtained on testing, Li!iles:j the reactiori is coiii1>lete. 13). a little care tlie end-reaction caii be iiiatle estreinel). close. Upoii suljstitutiiig a solution of uranyl chloritle for the nitrate, equally good results were o l h n e t l . SO I> I I ' 31 3101,Y1%D.iTI1 .\S D Y K .IT IL-JI S.\ LTS .
X solutioii \vas prepared contaiiiing 8.j I grains of sodium iliol!.btlate to the liter, aiid the sniiie salts of iiraiiiuiii Ivere tried tliat had been used i i i tlie experiiiieiits with tungstic acid. iIrith uranium acetate, sotliuni molybdate behaved similarly to sotliuiii tungstate, 110 precipitate being obtained b y aril. method tried. IVitIi the iiitrate of urniiiuiii, howei.er, a precipitate is foriiied at once \vliicli redissolves 011 stirring, though not cornpletel>., n siiinll quaiitity of soiiie flocculent illaterial remainiiig uiidissolxwl : lmiliiig dit1 not lriiig clo\vri any further precipitate. arid ail exyc of the iiitrate hat1 no effect. T h e uranyl nitrate solution \\'as the same as that used with sodium tungstate, and contained fourteen graiiis to the liter. 0 1 1 trying a coiiceiitratetl soluticxi of tiraiil.1 nitrate a dense precipitate formed at oiice. \vliich, 011 dilution, dissolved i i i part, but very imperiec t 1>.. .'rddition of alcohol does iiot cliaiige the lxliavior in either a cold or a hot solution. -1solution containing tweiity-five cc. of a saturated solution of a n i i ~ i o n i ~ icliloritle, ii twenty cc. of the sodium iiiolybtlate solution, atid tnwity-five cc. of water gave no periiiaiient precipitate with uranyl nitrate, hut when twenty cc. of the sodiuiii molybdate solution were diluted with twenty-five cc. ivater, arid the uranyl nitrate solution added in excess, the addition of ammoniuiii chloride brought down the whole of the niolybdeiiuiii. \fritli uranyl chloride, a precipitate was obtained at once,
A N D 3 I O L T B D A T E S O F THE R.IRE
E.\RTHS.
4SY
which, like that produced by uraiiyl nitrate, redissolved on stirring, as fast as it formed. On standing, a pale lenlon yellow precipitate appeared, which on heating went into solution. With the addition of an escess of uranyl chloride to the cold solution, a precipitate forms siniilar in appearance to the one just described, but on heating i t does not redissolve. Heated for an hour it g r o m denser, and gradually becomes crystalline. These crystals are insoluble i n water, and those which I obtained were washed first by decaiitation, then on the filter with cold water, dried on the air-bath at 128' C. (without any alteration in appearance), and analyzed. T h e molybdic acid was determined by heating the salt i n n current of liydrogen chloride gas. T h e ?\Io0,.2HC1 caine off readily at a low heat, tlie salt i n the boat changing in color from a brilliant golden yellow to a light green, then to a golden brown, and finall? to a dark green, alniost to a black. This blackish residue when treated with coticetitrated hytlrochloric acid did not dissolve ; with aqua regia it dit1 slo\vly tlissolve to a deep yellow solution from which tlie uranium was precipitated by aninioniuni hydroxide ; the precipitate after strong ignition had a metallic appearance. It was dissolved i l l concentrated nitric acid, evaporated to dryness, and gradually heated to a dull red heat. T h e brownish yellow powder obtained w a s weighed as uranium trioxide. A slight loss occurred when the salt was heated in tlie streniii of hydrogen chloride gas. as a little was carried over niecliaiiically, and heated so strongly to the walls of the tube that it could not be removed. T h e weight of material taken for analysis was one_tenth of a gram. T h e weights obtained were as follows : RIolybdenuni trioxide = 0.0333 gram = 33.3 per cent. Uranium trioxide = 0.0620 gram = 62.0 per cent. These values correspond very closely to the foriiiula UrOJIoO,, when the radicle MOO, may be regarded as replaciiig the two chlorine atonis i n tlie uranyl chloride. T h e tlieoretical aiiiounts required for this formula are RIolybdenum trioxide = 0.0334 gram = 33.4 per cent. Uranium trioxide = 0.0666 gram = 66.6 per cent.
490
F A N N Y K.
nx.
HITCHCOCK.
T H E TUXGSTATES
M'ater of crystallization does not seem to be present ; o m tenth of a gram of the substance was dried over sulphuric acid for two days, and then in the air-hat11 for two hours at a teniperature of 128' C. T h e loss i i i Tveight was eight-tenths of a niilligrain. On strong ignition the color c1ian:;etl at first a light green, then to a golden l)rowii, and on cooling lwcanle orice more bright yellow. By this ignition a loss i n weight occurred nniounting to nine and three-tenths niillligranis ; a second ignition caused 110 further loss. Both before and after ignition the salt is readily soluble in hydrochloric acid, apparently without tlecoiiipositiori. as no reaction for uranium is given by the solution \vlieii tested with potassium ferrocyanide. The perfect precipitation of tungstic acid by urarlyl nitrate from the sodium tuiigstate solution, and the iioii-l"ecipitatioi1 of iiiolyldic acid by the same reagent, encouraged a hope that a separation of tlie two acids might b? possible through its ineans. A s potassium ferrocyanide indicated the end reaction YO sharply between sodium tungstate and uranyl nitrate, soine tests were made with it as an indicator, withsolutions of sodium molybdate and uranyl nitrate. Observing the same precautions as before the elid reaction wab equally sharp. I t \vas observed, liowever, that the red color imparted to tlie potassium ferrocyanide by a drop of tlie nioll-bdate solution as sooil as tlie uranyl iiitrate was in excess, gradually disappeared. and \vas not restored b y the further addition of uranyl salts. T h e conditions were next T-arid b y introducing tlie potassium ferrocyariitle into tlie iiiolybdate solution instead of niakitig the tests 011 a porcelaiii plate as heretofore. 'I'nwit!. cc. of sodium molybdate were diluted Lvitli twenty-five cc. of water, and fi\-e cc. of potassiuiii ferrocyaiiitle added. T h e uranyl nitrate solution nxs then run in) the red color appearing w l ~ e n33.3 cc. of the uranyl iiitrate had been used ; an excess of six or seven cc. were added, and a deep red solution formed, but 110 precipitate appeared. .liter standing for two hours the red color had disappeared, ant1 the solution had become yellow. X portion of this \vas tested with potassium ferrocyanide, but 110 reaction \vas obtained for uraniuiii.
A N D MOLYBDATES O F T H E RARE EARTHS.
49 1
T h e addition of uranyl nitrate to another portion showed no excess of the potassiuni ferrycyanide. On substituting uranyl chloride for the nitrate the red solution was not obtained, but a reddish brown precipitate canie down which was much lighter in color than that given by the uranyl salts alone. Twenty cc. sodium tungstate solution were niixed with five cc. of the sodium molybdate solution, and diluted with twentyfive cc. of water : the uranyl nitrate solution was next run i n from a burette, and the end reaction ascertaiiied by potassiuni ferrocyanide on a porcelain plate. After standing for an hour the precipitate was filtered off in the cold, and washed with thirty-three per cent. alcohol. It was hoped that the filtrate would contain the molybdenum, but examination showed that the greater part of it had been precipitated with the tungsten. A second solution was prepared and titrated in the sanie manner a s before ; the precipitate with the supernatant fluid was then heated for an hour at a temperature of 60" C., filtered while hot, and washed with hot water containing thirty-three per cent. of alcohol. T h e molybdic acid was found i n both filtrate and precipitate, though more went into the filtrate than in the first case. On substituting uranyl chloride for the nitrate and repeating the titration, only a slight precipitate of the tungsten formed, the greater part dissolving with the molybdate, as fast as formed. An excess of uranyl chloride brought down both the tungsten and the molybdenum completely. I t was eviderit that no separation could be hoped for by means of uranium salts. There seems to be a tendency on the part of the members of the chroniiuni sub-group to form conipounds with each other which are not readily broken up, and which probably contain complex radicles of a s yet unknown constitution. This is very strongly indicated by the behaviour of molybdic acid with uranyl salts. and the failure of the resulting compound to react in solution, for tiranium with potassium ferrocyanide. As the chief object of my im~estigationswas to ascertain the manner i n which the rare earths would act with tungstates arid
492
I ' X X S Y R . Jl. H I T C H C O C K .
THE TI.SCST.ITBS
molybclates, these were now taken up and studied in detail. X s tlie quantity at iiiy disposal of a n y 011e of the rare earths used was necessnri1)- ver). liniitetl the material had to be worked o \ w frequently, ant1 a great deal of time w a s i n coilsequence consuiiied b y this preparatory work. S O U I L711 3101.E'13 1) .\'F 1; .\ 1;D C E R 11;11 SUI, I' € I AT E .
T h e solutioii of sodium niolybdate used contailled 1 3 . 2 4 grams to the liter, so that teu cc. of the solution should give 0.1628 graii? of cerium iiiolyMate, Ce,( MOO,)^ if fully precipitated. T h e solutioii of cerium sulphate used was saturated. ~.ti3et.iiiroif.-Fi~e cc. of the sodium molybdate solution were diluted to I jo cc. \vith distilled water, and ten cc. of the ceriuni sulphate solutioii added in the cold. -4 voluniiiious, white gelatiiious precipitate formed at once. This stood for eighteen hours, when it had become granular in appearance anti \vas ).ellow in color. I t w a s filtered cold. washed with cold n-ater, ignited and weighed. T h e theoretical weights are all calculated for the formula C q ( MOO,),,. I V t . ol>taillt.d.
Cnlculnted.
Uiff ereiice.
0.0760 graiii.
0.08 I4
--u.ooj~
Expeviiuciif 2.-Five cc. of the sodium molybdate solution diluted to I j o cc. with distilled ivater \Yere brought to l)oiliiig, wheii ten cc. of the cerium sulphate solution were added. A white precipitate formed a t o w e , soon becoming curtly, and then granular, the color changing to ).ellow. T h i s was allov-ed to stand for twelve hours, was tlicri filtered, and washed with hot water, ignited arid \veiglied. \Vt. o1,tniried.
grain. / k , b c Y i i ) z t w f ~.-'I'eii O.Ojj2
Cnlcrilated.
U.OSIJ
1)iffereiicr.
--0.0062
cc. sodium molybdate solutioii were diluted to 500 cc. n i t h distilled \vater, arid fifteen cc. of tlie cerium sulphate solutiori added iii the cold. T h e precipitate after standi;ig for twelve hours was filtered on asbestos in a Coocli crucible, ignited and weighed. \Vt
01)tniiied.
0.1596 grain.
Ca1cIll;lted.
Ijifferciice.
0.1629
--o,oo~~ 3
E.c/ci-iu/rfLf4.-This was a duplicate of No. 3, excepting that the filtration took place as soon as tlie precipitate hat1 settled. \VI. 0 . 1j
ol)tniiie
