Molybdenum
E
Q
E
£
Q9
-）)
©
bQ
£
±
ſ=ſ.
<!-!
Q
£
Q
º £
63
$）.
63
© .
GÐ
C/)
Q!)
-，
É=
§§§§
§§
#########
THESIS
PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE
¿?
§§§§¡¡¡¡¡¡¡№ffſ
######§§§→
-##########
æ:）-###■■■■ ■ ■ ■
#############£; £
·##################Ē
•••• • • •• • • • ••••••••••• •
、、、、、、、、** ••••••••••••••••••••••••••••••••••••••########№ſ;##################
№
*…！！！！
ſae
--★ →
UNIVERSITY OF PENNSYLVANIA IN PARTIAL FULFILMENT
§§§
OF THE REQUIREMENTS FOR THE DEGREE
•
ºsaeae:
••••••••
*---， --★ →
OF DOCTOR OF PHILOSOPHY
*Tºț¢º， sº sega
±,±，±），
±
srae:
ſae
§§
exe:
•••
••
#######№ſſae;
ſae•----
►►####
••
**
EDWARD ELLSWORTH MARBAKER
º.ſ.fºº.（
#…§;&####
----±，±，±，±，±，±，±，±，±,±，±）,
########################
#######••###§@₪
±±:::::：:&ae
§§§:ſſ;
Ē#################################
±，±，±，±,±，±，±，±），•••----
••ſae::::：:：:：:::：：:::：:• *************
PHILADELPHIA, PA.
ſae##########
§§£（;##############
* :
###########ae•++×
---- -■**-----•:es::::::
:：：::：:：##########§§§##############È####################
&########№Ē##########5
########################
PH | LA D S L PH ! A
THE WESTMINSTER PRESS

-！！！！！！~~~*~~~~~~--~~~~~<~~--~~~~
ș*********№sae！！！！！！！！！！！！！！！！!！！！！！！！！！！！！！！！！！！!！！！！*****
|-· * →ș****- ****！，
�&
）**…
、、、、、、、、、、、、s=********&<！=）
•■=±，±，±，±,±，±，±，±，±，±，±
·■** -

Chemical Library
-N
- a q
T º
| 1 , | *_
The Separation of Tungsten from
Molybdenum *
PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE
UNIVERSITY OF PENNSYLVANIA IN PARTIAL FULFILMENT
OF THE REQUIREMENTS FOR THE DEGREE
OF DOCTOR OF PHILOSOPHY
BY
EDWARD ELLSWORTH MARBAKER
PHILADELPHIA, PA.
1914
PH LA DE L PH 1A
THE WESTMINSTER PRESS
1914
ACKNOWLEDGMENT.
The author wishes to express his sincere thanks to
Edgar F. Smith, Professor and Provost, at whose sugges-
tion this work was carried out, and to whose helpful advice
its successful completion is due.
THE SEPARATION OF TUNGSTEN FROM
- MOLYBDENUM. .
INTRODUCTION.
With the extensive introduction of tungsten and
molybdenum into industrial chemistry in the form of the
special steels and the filament of the incandescent electric
lamp, the importance of these metals has been greatly
increased from the standpoint of the analytical chemist.
In the case of the special steels the amount of the rarer
metal present has a very marked effect on the quality of
the steel in question, and it is very necessary to know the
exact amount of tungsten or molybdenum present in order
to account for the special properties and characteristics
of the steel. Tungsten and molybdenum are both used
in the manufacture of high speed tool steels, and in order
to control the amounts of the two metals present a very
wide field in the domain of analytical chemistry has been
opened up. This work has been going on for many years,
but it was not until quite recently that the subject began
to receive the almost universal attention that is now being
bestowed upon it. The reason for this special impetus is
the wide-spread and growing interest which is taking place
among chemists for the utilization of the comparatively
rare and little known elements.
When the tungsten lamp industry was inaugurated a
few years ago and the demand for pure tungsten metal
became large, the chemical analysis of tungsten and its
compounds received still greater attention. The outcome
of this is that a very extensive literature on the subject
has been produced.
It is a well known fact that molybdenum always accom-
panies tungsten in its ores, and it shows a marked tendency
to remain with the tungsten throughout the process of man-
(3)
4
ufacture. This amount of molybdenum, although small,
interferes to a great extent, and may even in some cases
entirely change the properties of the manufactured ma-
terial. This is certainly true in the case of the special steels
and in the production of tungsten filament.
For this reason, if for no other, it is necessary to know
how much of this contaminating metal is present, and hence
one of the important problems in modern analysis is to
devise a separation of the two metals which will be easy to
carry out and also of the highest accuracy. Up to the
present time there have been suggested four methods for
the attainment of this end :
Rose,”
Pechard,”
Smith and Reugenberg,”
Ibbotson and Brearley."
These separations are all based on different reactions,
and on the first three of them other suggestions have been
made, without however, having much effect on the work of
the original authors.
The separation of tungsten and molybdenum accord-
ing to Heinrich Rose is carried out as follows:
The alkali salts of the two metals are dissolved in an
excess of tartaric acid and a considerable amount of sul-
phuric acid is added, and then the solution is placed in a
pressure flask and the molybdenum precipitated by hydrogen
sulphide. The molybdenum sulphide is filtered off and
ignited gently to form the trioxide, which is then weighed.
In order to determine the tungsten the tartaric acid is
destroyed by repeated evaporation with nitric acid and
1 Rose, Treadwell-Hall, Analytical Chemistry, II, Third Edition,
1911, 296.
2Pechard, Ibid. 295. Comptes Rendus, 114, 173. 2
3Smith and Reugenberg, Jour. Amer. Chem. Soc. 22, 774. Tread-
well-Hall, 293.
4Ibbotson and Brearley, Chem. News, 81, 1900, 13.
5
the precipitated tungstic acid is finally filtered off, ignited
and weighed as the trioxide.
According to Treadwell and Mennicke” this method is
fairly accurate, but not to be recommended on account of
the time consumed in removing the tartaric acid.
The method of Pechard is based on the observation of
Debray" that the trioxide of molybdenum is volatilized as
MoC),2.HC1 in a stream of hydrogen chloride at a tem-
perature of two hundred degrees. The trioxide of tungsten a
is not volatile at this temperature, but a temperature of six
hundred degrees is sufficient to cause such a volatilization
to take place. This method was worked out in this labora-
tory independently by Smith and Oberholtzer, but Pechard
published his work a little earlier and hence has been given
the credit for the separation. The details of the separation
are as follows: The oxides of the two elements or their
sodium salts are weighed into a porcelain boat and the latter
is placed in a tube of moderately hard glass, of which one
end is bent vertically downward and connected with a Peli-
got tube containing a little water. The horizontal portion of
the tube is placed in a drying oven, and is connected with the
apparatus for generating hydrogen chloride. The gas is
passed through a tower containing hydrochloric acid and
another containing concentrated sulphuric acid for the pur-
pose of drying the gas completely. The molybdenum begins
to sublime as soon as the temperature reaches two hundred
degrees. At the end of an hour or an hour and a half the
sublimation is usually complete. The boat containing the
residue of tungsten trioxide is removed and weighed, after
drying in a dessicator. In case the sodium salts have been
used, the sodium chloride is first washed out, and then the
tungsten is weighed as the trioxide. The molybdenum is
determined by washing out the Peligot tube with very dilute
nitric acid, and this solution of the sublimate is evaporated
5Mennicke, Untersuchungsmethoden des Molybdaens, Vanadiums
u. Wolframs, Berlin, 1913, 126. -
6Debray, Comptes Rendus, 46, 1101.
6
to dryness in a porcelain dish. It is then dissolved in
ammonia, washed into a porcelain crucible, evaporated to
dryness, and changed into the oxide by gentle ignition, and
weighed after drying in a dessicator. -
This method has been used satisfactorily for a number
of years, and the only objection to it is that it requires spe-
cial apparatus and the most careful manipulation.
The method of Smith and Reugenberg is based on the
fact that molybdenum trioxide is soluble in concentrated
sulphuric acid, while the corresponding compound of tung-
sten is insoluble. This observation was first recorded by
Desi" and suggested by him as a possible separation. Hom-
mel, working in Treadwell's laboratory, tested the method
and found it necessary to make some changes in the pro-
cedure recommended by the authors. The details of the
improved method are as follows: The mixture of the
moist, freshly precipitated oxides is covered with concen-
trated sulphuric acid (Smith and Reugenberg recommend
acid of a specific gravity of 1.378) in a porcelain dish and
heated over a free flame. In case some of the blue oxide
of tungsten is formed, a drop or two of nitric acid is added,
and the green color disappears, leaving a tungstic acid of a
pure yellow color. After digesting for a half hour the
separation is complete. Upon cooling, the liquid is diluted
with three times its volume of water, filtered, washed with
water containing sulphuric acid, then two or three times
with alcohol, ignited and weighed as tungsten trioxide.
The molybdenum is precipitated from the filtrate by hydro-
gen sulphide under pressure, and the precipitate filtered
off and ignited gently to the trioxide, after which it is cooled
and weighed.
This method is recommended by Treadwell and Men-
nicke,” but in the experience of the author it has not given
very satisfactory results. The reason for this may have
been that in those determinations the amount of molybde-
7Desi, Jour. Chem. Soc. 74, II, 231.
8Mennicke, Untersuchungsmethoden, 126.
num was extremely small as compared with the tungsten
present.
The method of Ibbotson and Brearley is based on the
fact that hydrochloric acid has a different action on the
lead salts of tungsten and molybdenum. The ignited salts
are both soluble in concentrated hydrochloric acid, but on
dilution, the tungsten is precipitated in the form of the
trioxide. The details of the method are as follows: The
two metals are present in the form of their sodium salts.
A slight excess of lead acetate solution is added, and the
precipitate allowed to settle. The precipitate is then filtered
off and washed with hot water containing a few drops of
acetic acid. The filter and its contents are ignited, and
after cooling transferred to a beaker and then as many ce.
of concentrated hydrochloric acid are added as there are
centigrams of the combined lead salts. This is heated at the
boiling point of the acid until the solution is complete.
From two to three times the volume of boiling water are
then added, and the tungsten is precipitated. This is
allowed to settle and then the clear supernatant liquid is
decanted through a filter. The residue in the beaker is
treated with about half the original quantity of concentrated
hydrochloric acid and the same procedure carried out as
before. As a rule, three acid treatments are enough for a
complete separation. The precipitate of tungstic acid on .
the filter is ignited and weighed. The molybdenum in the
filtrate is reprecipitated by the addition of ammonia in the
form of lead molybdate, and this is filtered off, washed with
very dilute acetic acid, ignited and weighed.
This method is not mentioned by Treadwell, and Men-
nicke does not recommend it.” The author tried it several
times and the results were quite satisfactory, although it
is a time-consuming method.
In 1911, Mdivani” announced a new method for the
determination of tungsten which was based on the observa-
9Mennicke, Ibid, 126.
10Mdivani, Bull. Soc. Chem. 9, Ser. 4, 1911, 122.
8
tion of Anthon” that when a tungstate solution is treated
with a solution of Stannous chloride a yellow precipitate of
normal tin tungstate, SnwC), is formed, which is changed
to the blue oxide of tungsten, W., Os, on the addition of con-
centrated hydrochloric acid. Mdivani added to the tung-
state solution a solution of Stannous chloride in concen-
trated hydrochloric acid and then boiled for a few minutes.
The blue oxide was precipitated and, after settling, was
filtered off, washed with hot water, dried, ignited to tung-
sten trioxide and weighed as such.
It was suggested that this work be repeated, and the
method tried in the presence of molybdenum with the idea
of perhaps effecting a new separation of tungsten and
molybdenum. The experimental work which was carried
out will be detailed in the part of the paper which follows.
ExPERIMENTAL PART.
I. An effort to reproduce the results of Mdivani.
Mdivani made up his solution for analysis by dissolv-
ing two grams of pure tungsten trioxide in strong ammonia
and diluting to one liter. He treated 50 cc. of this solution,
which contained 0.1 g of WOa, with 20 cc. of a solution
containing 50 g. crystallized Stannous chloride in 200 cc.
concentrated hydrochloric acid. The blue precipitate which
was thrown down was boiled up for a few minutes and then
allowed to settle, after which it was washed by decantation
with hot distilled water and finally ignited and weighed as
WO,.
A sample of pure tungstic acid, which had been pre-
pared for the purpose of making ductile tungsten lamp fila-
ment, was weighed out and dissolved in strong ammonia.
The solution was not complete, showing that the purity of
the substance was doubtful; but it was filtered and made up
to one liter. Samples of this solution were analyzed by
11Anthon, Dammer, Handbuch, Vol. 3, 1893, 659, Jour. Prac.
Chem. 9, 341.
9
evaporating to dryness three times with hydrochloric and
nitric acids, baking at about 120° for fifteen minutes after
each evaporation, then washing on to a filter with hot
ammonium nitrate solution, and finally igniting and weigh-
ing. Fifty ce. of the solution were thus found to contain
0.1422 g. WOa. This value was obtained as the average
of ten closely agreeing determinations.
A solution of stannous chloride was made up according
to the directions of Mdivani. Twenty ce. of the tin solution
were added to 50 cc. of the ammonium tungstate solution,
and the resulting blue precipitate boiled up for a few min-
utes and then washed according to the directions of Mdi-
vani. The pure water recommended for washing was found
to be inadequate because it caused the formation of basic
tin salts which could not be washed out and hence gave rise
to high results. This was easily remedied by washing the
precipitates with a solution of hydrochloric acid made by
adding 50 cc. of strong acid to a liter of boiling distilled
water. The washed precipitates were thus entirely free
from tin.
The results of four determinations were as follows:
WO, present WO, found
0.1422 0.1392
0.1492 0.1386
0.1422 0.1390
0.1422 0.1387
In order to determine the cause of these low results,
the filtrate from one of the determinations, which was per-
fectly clear and colorless, was evaporated to dryness. When
the bulk of the solution had become very small a separation
of the blue oxide of tungsten was noticed. This showed
that either the precipitation had not been complete or that
the acid of the wash water had caused some of the oxide
to pass into solution. Subsequent results showed that the
IO
latter was not the case. The small amount of blue oxide
which came out on evaporation was transferred to a filter
and washed with the dilute acid until the filtrate gave no
test for tin on the addition of ammonia. The filter and its
contents were then ignited, but instead of obtaining a small
amount of yellow WO, there was present a mixture of the
yellow oxide and a white substance which proved to be
oxide of tin. This points to the fact that a combination
had taken place between the tin and the tungsten, because
all of the tin had been washed out of the precipitate so far
as the tests on the filtrate showed. In order to eliminate the
small amount of tin, the residue was carefully treated with
ammonium chloride, as recommended by Rammelsberg”,
which causes the tin to volatilize as stannic acid leaving the
tungstic acid behind. This residue was then weighed and
this weight added to that of the main portion of the oxide
brought the result into agreement with the theoretical. In
Subsequent analyses it was found that the ignition with
ammonium chloride must be carried out with the utmost
care or some of the tungstic acid will be volatilized with the
Stannic acid.
While several results were obtained by this method,
which were fairly good, the method can hardly be recom-
mended.
In order to test the method more nearly as Molivani has
recommended, it was decided to prepare some absolutely
pure tungsten trioxide by the method of Smith and Exner”.
The idea of this was that a sample of pure oxide might be
weighed out and made up to a definite volume and the
weight of the oxide taken as the criterion of the solution
without having recourse to any method of analysis. This
was not done, however, because several other observa-
tions were made which rendered such a procedure unneces-
Sary.
12.Rammelsberg, Chem. News, 9, 25.
13Smith and Exner, Proc. Amer. Phil. Soc. 43, 1904, 123.
I [
II. The action of Mdivani's reagent on a mixture of
ammonium tungstate and ammonium molybdate.
When the quantity of the tin reagent was added to a
solution of ammonium molybdate a brown precipitate was
formed which rapidly dissolved in the excess of the reagent
to form a solution having a characteristic Orange color,
due to the reduction of the MoC), to Mo,Oa. This is very
different from the reaction which takes place in the tung-
state solution, hence it seemed possible to effect a separa-
tion by means of Mdivani's reagent.
A solution of molybdenum trioxide in ammonia was
made up and analyzed by the method of Chatard", in which
the molybdenum is precipitated by lead acetate and weighed
as lead molybdate. Ten cc. of this solution were found to
contain 0.1164 g. MoC)a.
A solution was made up containing 0.1164 g. MOOa
and 0.1422 g. WOa. This was treated with 20 cc. of the tin
solution and the resulting precipitate of W.O. treated as
outlined before. The precipitate was washed with the acid
wash water until the washings gave no test for tin. The
filter and its contents were ignited and weighed. A fairly
close value for the tungsten was obtained.
It was assumed at this point that the precipitate of
tungstic acid contained no molybdenum, but a careful test
was made to determine whether the filtrate containing the
molybdenum contained any tungsten. This test was carried
out as follows: An excess of metallic zinc was added to
the solution and the tin was precipitated as metal. The
solution changed in color from Orange to green. The solu-
tion was decanted from the tin, which was present in the
form of a porous cake, and the tin washed several times
with hot water. The clear green solution was then evapo-
rated on the hot plate, during which process the color
changed to brown again. When the bulk of the solution
had been decreased to about 150 cc, a few drops of nitric
1+Chatard, Chem. News, 24, 175.
I 2
acid were added. The solution rapidly became colorless,
showing that the molybdenum had been oxidized to the form
of molybdate. Ammonia was then added in great excess.
The precipitate of zinc hydroxide which first appeared dis-
solved in excess, leaving a small precipitate of iron and tin
hydroxides. This was filtered out and reprecipitated. The
filtrate was added to the main solution and the whole treated
with acid to the point of neutrality. A small amount (3-5
cc.) of acetic acid was then added, followed by an excess of
lead acetate. A precipitate of lead molybdate was formed
which was allowed to settle, then was filtered, washed with
dilute acetic acid, ignited and weighed. This weight proved
to be lower than the theoretical, but the loss could not be
accounted for, although the precipitate of tin metal, the
iron and tin precipitate, and the filtrate from the lead molyb-
date were all tested with zinc and sulpho-cyanide, but no
test for molybdenum was obtained.
The ignited residue of lead molybdate was then tested
with hydrochloric acid, according to the method of separa-
tion suggested by Ibbotson and Brearley”. The solution of .
the residue was complete, but on dilution no tungstic acid
separated out, thus showing that the molybdenum was not
contaminated with any tungsten and that So far as the tung-
sten was concerned the separation was complete.
It was found in some other determinations carried out
in the same way that the tungsten weighed high, although
all the tin had been removed by careful washing. The high
results must then have been due to a small amount of molyb-
denum left in the tungsten. The precipitate of W.O. was
dissolved in ammonia containing a little hydrogen peroxide
and then reprecipitated with a second portion of the tin
solution. The precipitation at this point was never com-
plete, because the filtrate had a blue color. This was evapo-
rated to dryness, the W2O, washed out and ignited with
ammonium chloride and finally weighed. This weight was
15Ibbotson and Brearley, Chem. News, 81, 1900, 13.
I 3
added to that of the main precipitate and the sum of the
two gave a somewhat low result, due probably to unavoid-
able losses in the process of analysis.
A series of results follows:
WO, present - WO, found
0.1422 0.1426
0.1422 0.| 4 |4
0.1422 0.1429
0.1422 0.1434
0.1422 0.1432
0.1422 0.1404
§ § reprecipitated.
0.1422 0.1414 -
0.1422 0.1434
0.1422 0.1428
0.1422 0.1415
Average 0.1422
These results show that a separation of tungsten from
molybdenum by this method is possible, although they are
not particularly concordant. The molybdenum was not
determined, because at this time there was no method avail-
able for the separation of the molybdenum from extremely
large amounts of tin.
III. The action of Mdivani's reagent on a mixture of
sodium tungstate and sodium molybdate.
Molivani states in his paper that the quantitative pre-
cipitation of tungsten from solutions of the tungstates of
Sodium and potassium can be carried out as well as that
from solutions of the ammonium salts.
The solution of sodium tungstate used in this test was
made from a portion of the salt which had been prepared
by fusing together the theoretical amounts of pure WO,
and Na2CO, and crystallizing the product. The solution
I4
was found to contain 0.1489 g, WO, in each 10 cc. The
solution of sodiurn molybdate was made by dissolving a
sample of a salt, which had been made by Schuchardt, and
afterwards recrystallized several times. This solution con-
tained 0.1200 g. MoC), in each 10 cc.
A sample of the tungstate solution was measured out
and treated with the usual amount of Mdivani’s reagent.
The blue precipitate appeared and was filtered off. The
filtrate was colored blue, showing that the precipitation had
not been complete. Several attempts were made to improve
this precipitation by adding more reagent than usual and
also by the addition of ammonium chloride, but without
any favorable results.
A solution containing both tungstate and molybdate
was then treated in the usual manner with the tin solution
in order to determine the effect of the presence of sodium
molybdate. The blue precipitate of W.O., came down well
and was weighed after filtering and washing. The filtrate
in this case was orange colored, on account of the molybde-
num present. A series of ten determinations was carried
out and the results were so concordant and near the theo-
retical that it shows conclusively that for the amounts of
tungstate and molybdate present a separation of tungsten
and molybdenum can be effected by means of an acid tin
chloride solution. The results follow :
MoC), present WO, present WO, found
0.1200 0.1489 0.1492
0.1200 0.1489 0.1489
0.1200 0.1489 0.1490
0.1200 0.1489 0.1486
0.1200 0.1489 0.1486
0.1200 0.1489 0.1489
0.1200 0.1489 0.1492
0.1200 0.1489 0.1486
0. | 200 0.1489 0.1489
0.1200 0, 1489 0.1489
I 5
As in the case of the separation of tungsten and molyb-
denum using the ammonium salts, the molybdenum was not
determined.
The tungsten trioxide obtained in the foregoing analy-
ses was tested for molybdenum by the method of Debray"
by passing over it dry hydrogen chloride at a temperature of
200°. A slight sublimate of the compound MoCa2HCl was
formed, showing that a little molybdenum was present.
On subsequent analysis by means of the lead acetate method
of Chatard it was found that the molybdenum contained in
the tungsten trioxide from six determinations amounted to
less than one milligram of MoCA.
The fact that the separation takes place so easily is so
at variance with the usual experience of analysts of tungsten
compounds that it seemed very remarkable that the molyb-
date actually aids the precipitation of the tungsten, as it will
be remembered that the precipitation of W.O., from the
tungstate alone is never complete. This was so peculiar that
the effect of other salts on the precipitation was tried. The
results of these tests were as follows:
Salt present WO, present WO, found
0.1 g. K., SO, O. 1489 0.148 |
0.1 g. K., SO, 0.1489 0.1488
K, CrO, 0. || 4 || 5 0.14|4
0.2 g. NaCl 0.1489 0.1487
0.2 g. NaCl 0.1489 0.1494
The filtrate from the tungsten in all of the foregoing
determinations, except that containing chromium, was per-
fectly colorless. These results show that an accurate esti-
mation of the tungsten in sodium tungstate can be made
by the addition of any of the salts mentioned in the table.
In order to determine the availability of the method
16Debray, Comptes Rendus, 46, 1101.
I 6
various amounts of sodium tungstate and sodium molybdate
were taken. The results are shown in the following table:
MoC), present WO, present WO, found
0. | 200 0.0089 0.0089
0.1200 0.0 || 49 0.0 || 5 |
0.1200 0.0596 0.0592
0.1200 0.0900 0.0900
0.1200 0.1200 0.1200
0.1200 0.1788 0.1786
0.1200 0.2 100 0.2 || 0 |
0.1200 0.2400 0.2397
0.| 200 0.2700 0.2700
0.1200 0.3000 0.2998
0.0 || 2 | 0. 500 0.1496
0.024 | 0.|500 0.1497
0.0482 0.1500 0.|500
0.0724 0. |500 0.1504
0.0965 0.1500 0.1499
The Solution of sodium tungstate which was used in
the last ten of these determinations was made from Kahl-
baum sodium tungstate which had been recrystallized once.
It contained 0.1500 g. WO, in each 10 cc.
The above results show that a separation of tungsten
and molybdenum can be effected regardless of the amount
of the two metals present.
IV. The determination of the molybdenum in the filtrate
from the blue oaſide.
The filtrate from the tungsten in this separation con-
tains all the molybdenum and a large amount of tin in the
form of stannous and stannic chlorides. This offered a
problem which required considerable study. The first
I 7
efforts made were in the direction of a separation of the
tin from the molybdenum and a subsequent determination
of the molybdenum gravimetrically in the form of lead
molybdate.
The separation was carried out as was mentioned
before, that is, by the precipitation of the tin from the solu-
tion by means of metallic zinc. The solution was removed
from the tin, and the tin washed carefully with hot water,
and then the solution containing molybdenum and zinc was
evaporated down to a small bulk, oxidized with nitric acid
and then treated with an excess of ammonia. The zinc
hydroxide which was thrown down dissolved in excess and
the molybdenum precipitated by lead acetate after the solu-
tion had been almost neutralized with HCl and finally
treated with a small excess of acetic acid. The precipitate
of lead molybdate was allowed to settle and was then filtered
and washed with hot dilute acetic acid, and finally weighed.
In every case the result was low, and no reason for this
condition has been found. Tests were carefully made at
every stage of the operation where any loss might occur,
and all of these tests for molybdenum proved to be nega-
tive.
Since the foregoing procedure failed to solve the prob-
lem of determining the molybdenum, a careful review of
the literature was taken up in the hope that a method for
determining molybdenum in a hydrochloric acid solution
had been devised. Nothing of this nature was found in
connection with any gravimetric method, but the discussion
of the volumetric methods seemed to offer a means of solv–
ing the problem.
Randall" has shown that when a molybdate solution is
passed through a Jones reductor and the reduced solution
caught in a ferric ammonium Sulphate solution, decolorized
by phosphoric acid, that the ferric iron oxidized the molyb-
denum, and the ferrous iron thus formed may be titrated
with standard potassium permanganate. This investigator
17 Randall, Amer. Jour. Sci. (4) 24, 1907, 313.
I 8
proved that the molybdenum is completely reduced to the
form of Mo.Oa. All the work which has been done on this
method of détermination, and it has been the subject of
much investigation, has been done with the molybdenum
acidified with sulphuric acid. In this case the molybdenum
was acid with HC1. It was thought probable that the dis-
advantages of the HCl might be counteracted by the use of
the solution known as “titrating solution,” and by means
of which it is possible to make a permanganate titration in
the presence of HC1. --
The solutions used in this part of the work were as
follows:
An approximately tenth normal solution of potassium
permanganate, which had been standardized against ferrous
ammonium sulphate and also oxalic acid. The iron value of
the solution was multiplied by the factor 0.857, which gives
the value of the solution in units of MoCa.
A solution of 100 g. ferric ammonium sulphate in 1 1.
water.
A solution containing 70 cc. concentrated HCl per liter.
The titrating solution consisted of 90 g. crystallized
manganous sulphate, 650 cc. distilled water, 175 cc. syrupy
phosphoric acid, and 175 cc. concentrated sulphuric acid.
The reductor was made by fusing a stop-cock on to the
constricted end of a piece of wide glass tubing, and a sort
of funnel made at the other end. A piece of glass wool
was introduced into this tube, followed by a column of 20
mesh zinc (Baker's Analyzed), which had been amalga-
mated”, 36 cm. in length. The diameter of the column
was 1.5 cm. Another piece of glass wool was placed on top
and then the apparatus was filled with distilled water and
aſ the air withdrawn by means of a filter-pump.
A determination was carried out as follows: A portion
of the sodium molybdate solution was measured out and
treated with about 20 cc. of Molivani's reagent. This gave
rise to an orange colored solution, to which was added
18Blair, Chemical Analysis of Iron, Sixth Edition, 1906, 97.
I9
about five grams of pure 20 mesh zinc. The tin was depos-
ited on the zinc as metal, and the green supernatant liquid
was poured off, the tin washed with hot water, and then the
solution and washings were heated to about 60°.
The reductor flask was charged with 20 cc. of the
ferric alum solution and 20 cc. of the titrating solution.
Fifty ce. of the HCl heated to 60° were passed through the
reductor, followed by the solution containing the molyb-
denum, and then 150 cc. of the warm HCl and finally 150
cc. hot distilled water.
The flask was removed from the reductor and the
titration was carried out in the usual way with the perman-
ganate solution. The molybdenum solution before entering
the reductor is green; while in the reductor the reduction
is completed, although there is no noticeable change in color.
When the green solution enters the ferric solution the color
changes to red. This red solution is titrated with perman-
ganate, but the red color offers no obstacle because, as the
permanganate is added, this color disappears, and toward
the end of the titration the color of the solution is a faint
bluish green in which the end-point is very easily seen. The
end-point has to be taken very quickly, as it is not at all
permanent. However, the titration is one which is very
easily carried out.
A number of such determinations were carried out, and
the results were as follows: *
MoC), present MoOa found
0.1500 0.1505 -
0.1500 0.1505
0.1470 0.1476
0.1500 0.1498
0.1500 0.1498
0.1350 0.1354
0.1635 0.1632
0.1200 0.1 198
0.1050 0.105 l
0.1800 0.1801
2O
The amount of tin present in each sample was 2.6 g.
This gave no test for molybdenum when dissolved in acid
and treated with sulpho-cyanide.
These results show that the separation of tin from
molybdenum by means of metallic zinc is quantitative and
that the method may be used to determine the molybdenum
in the filtrate from the tungsten in the method of separa-
tion under consideration.
V. The separation of tungsten from molybdenum.
In this part of the work the two processes which have
been outlined in the third and fourth parts of this paper
were combined, and both constituents of the mixture of
sodium tungstate and molybdate were determined.
It will be seen from the results that quite large amounts
have been taken, this having been done to find out the limita-
tions of the method. It was found that the separation
works very well with quantities up to and including O.3000
g. of each constituent. More than this amount gives rise
to high results unless an inordinate amount of time is con-
sumed in washing out the molybdenum from the tungsten.
The high results can be more quickly corrected by treating
the ignited residue of WOs with as many ce. of concen-
trated HCl as there are centigrams and boiling for a few
minutes. Then this is diluted with three times its volume of
hot water and allowed to settle. This process removes the
contaminating molybdenum and leaves the tungsten pure.
The filtrate containing the small amount of molybdenum is
of course added to the main molybdenum solution. There .
seem to be no limits to the applicability of the molybdenum
titration. When large amounts of molybdenum were pres-
ent it was found advisable to make up to 250 cc. and then
work with 50 cc, portions. The latter method proved to be
a great factor in reducing the time of the determination.
2 I
The time factor of the separation depends greatly on
the amounts of material that are being used. When mod-
erate samples are used the whole process, involving the
determination of both the constituents can be carried out in
about three hours. Of this time the greater part is con-
Sumed in washing the tungsten precipitate and in evapo-
rating the filtrate containing the molybdenum to a bulk
which can be easily handled.
The results of ten complete separations follow :
WO, MoC), WO, MoC),
present present found found
0.1005 0.1005 0.100 0.10 || 0
0.1500 0.1500 0.1496 0.1498
0.2000 0.2000 0.2004 0.1995
0.2520 0.2505 0.2526 0.2499
0.30 || 5 0.3000 0.30 || 6 0.2997
0.3500 0.3500 0.3508 0.3502
0.4005 0.4005 0.4008 0.3995
0.4500 0.4500 0.4508 0.4495
0.4995 0.4995 0.4999 0.4994
0.3000 0.3000 0.3000 0.2996
SUMMARY.
Tungsten and molybdenum, in the form of their sodium
salts, may be quickly and easily separated by means of a
solution of crystallized stannous chloride in hydrochloric
acid, known as Molivani’s reagent. The analysis is carried
out in the following manner:
To a boiling solution containing the tungstate and
molybdate of sodium in appropriate amounts add the tin
solution (50 g. SnCl2.H2O dissolved in 200 cc. concen-
trated HCl) in the proportion of 20 cc. for each 0.15 g.
22
WO, present. This amount of the reagent should be care-
fully regulated, because less will not effect a complete sep-
aration, and more will be that much harder to handle in the
determination of the molybdenum. The dilution of the solu-
tion for analysis should be kept between 60 cc. and 300 cc.,
depending on the amount of tungsten present. After boil-
ing a few minutes allow the blue precipitate of W.O. to
settle and then pour the orange supernatant liquid through
a filter. Wash the precipitate thoroughly by decantation
with hot 5 per cent. HCl until the washings give no test
for molybdenum by the zinc and sulpho-cyanide method.
Then bring the precipitate on the filter and allow to drain.
If the amount of precipitate is small the filter may be
ignited wet, but 1ſ the amount is large it is better to heat it
until the greater part of the water has been driven off.
Place the filter in a porcelain crucible and ignite gently to
WOa. Cool in a dessicator and weigh.
The filtrate and wash water are united and the whole
evaporated to a small bulk. If the amount of tungsten and
molybdenum taken has been large, cool and dilute to 250 cc.
and use aliquot portions of 50 cc. To this solution add from
five to ten grams of 20 mesh zinc. The tin will come out
as metal. Pour off the green solution when the precipita-
tion of the tin is complete (this takes about ten minutes),
and wash the cake of tin several times with hot water.
Unite the filtrate and wash water and heat to 60°. Pass
through the reductor in the order named, 50 cc. 2.5 per
cent. HCl, the solution containing the molybdenum, 150 cc.
2.5 per cent. HCl, and finally 150 cc. hot distilled water.
These pass into the flask of the reductor, which has been
previously charged with 20 cc. 10 per cent. ferric ammonium
sulphate and 20 cc. of the “titrating solution”. This solu-
tion is then titrated with approximately tenth normal potas-
sium permanganate and the amount required calculated to
MoC),.
If it is desired to determine the amount of tungsten in
23
a sodium tungstate solution, add to the solution a quantity of
sodium chloride solution, and proceed as when molybdenum
is present. (See page 15.)
It may be stated that a fairly accurate separation of
tungsten and molybdenum may be effected when they are
present as ammonium salts, but the method is not to be
highly recommended.
2
<C
Ø
I
C
>
LL-
O
| | Off
CC
LLI
>
Z
TO
3 9015 07506 0213
|
Makers
Gaylord Bros.
Syracuse, N Y
PAT. JAN. 21, 1908.
·----！!！！№№5！！！！！！！！SEĒ►►-
：§§§！- ™-3№， №8：-！№.（~~~~·º=<！--·
:::ĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒĒsae§§§§Êt§ĒÈĒĒĒ§§§§ĒĒĒĒĒĒĒĒĒ№.§§！！！）:
·-·-*** -ae，：·-№：№
§§-ſaeg§§ĒĒĒĒĒĒĒĒĒĒĒĒĒĒaes，~~：----،
ſ.•-！-，-§Ē§=§§ĒĒĒĒĒĒĖĖĘĒĒĖʧ§
aeae aetersaeSae:！-ſae：§§§§）:£§§£zsī£§ĒĒĒĒĒĒĒĒĒ§！§:№Ē-ĒĢĒ-·- -
~~--，ºjſae• •5№•=.■■■ſaeae----：！！！！！！！!-№sae
！ſae·Færaes№.ſae№ssºs，（№ssae·****-§§Ē№s:-----·ſae·№=•.§§§§----
№:：:<■■*<ĶĒĒĒĒĒĒĒĒĒĒĒ§§§§§§§§！§§№ſae§ĒĢĒĒĒĒĒĒĒĖ§§§
|－[…]-§§-§§·
，*è «…--ſae！！!！!•№::：:::ſs:£§!• -§§§§§§№§:§©®±±±§№ae，- -
~~~~===№№ţ###ȧ§§§§§§§§
· -- -------- -5==· · · · **********~~~. --~~~~*~~ º.º., I-EVOERS~~ ~~~~::~:sſaewr-
→
· ~~~~<！=~ş:35

#:
::::::::
#
!
º
id
º
º
::
º
#
:
:
º
#
g
;
:
;
º
#
§
º
§
;
§
º
§
º
*
º
º
º
;
§
Ç
- º &
ãº
§§ ####### t
º
§ ; § º
sº
º
ºf
t #
º ; # # §
;
ºft#.
§ §
|
*
º
º
§: § §§ §: 8.
#
#
: # i t
#! # ; |
#
#:
#######
§§
; §§ # §
º º: {{
#
º
:;
º
#
º#
-]
#
º
f
{j
;
#
:
#:- º
i
º
º§
#:
#
º
:
ſ::
i
*
j
i.
#
;
4
;
i
;
;
#
sº #:
º
§ w
sº º
º
§§§
:
:
*
t