ON REDUCTION

OF AMMONIUM
SOLUTION.

BY C. W. MILLER
(From

the Department

AND

MOLYBDATE
A. E. TAYLOR.

of Physiological
Chemistry,
vania,
Philadelphia.
Pa.)

(Received

for

IN ACID

publication,

March

University

of Pennsyl-

24, 1914.)

1 Pinoff:

Ber.

a?. d. them.

Gesellsch.,

xxxviii,
531

p. 3308.

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

Ammonium molybdate, in acid solution, is reduced to an oxide

stage, with the production of a blue color, by many substances of
biological importance.
Our attention was first attracted to this
subject through an accidental observation.
We were engaged in
the attempt to work out a technique for the nephelometric
estimation of phosphoric acid, using a heavy solution of cane sugar
for the medium of suspension.
When the mixture containing acid,
ammonium molybdate and cane-sugar was heated, a deep blue
color was rapidly produced.
The same reaction was slowly
accomplished in the cold. Further
study indicated that this
reaction was due to the levulose split off from the cane sugar by
the acidity of the solution.
Later we found that this reaction
had been described as a test for levulose by Pin0ff.l
Later study
of the reaction has yielded instructive results.
The reaction is in
some ways a rather specific one, since comparatively
few reducing
substances, otherwise active, are able to effect in acid solution the
transfer of the metal to the stage of blue oxide. A review of the
literature dealing with this metal indicates firstly that more than
one oxide stage presents a blue color, and secondly that the several
oxide stages are so unstable and interchangeable that a qualitative
relation of a particular reagent to a particular
oxide is not now
demonstrable.
The solution we have most commonly used consists of 30 grams
of ammonium molybdate and 25 cc. of sulphuric acid dissolved in
1 liter of distilled water.
The test is made by adding to a few
cubic centimeters of the reagent about an equal volume of the

532

Reduction

of Ammonium

Molybdate

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

solution of the substance to be tested (it must be free of phosphoric

acid) and standing the test tube in a water bath for from ten
minutes to a half hour.
A deep blue, somewhat lighter in color
(i.e., with less of the purple tinge) than the blue of copper sulphate
in alkaline solution, constitutes
the reaction.
On dilution, the
color fades to a greenish blue. On standing the color is apt to fade;
sometimes it changes to a pale purple.
Certain substances yield a positive reaction in a most striking
manner, using mere traces of the reducing substance.
These are
levulose (cane-sugar, raffinose, inulin), dioxy-acetone,
hydrazines,
indol and skatol, oxalic-acetic
ester, furfurol,
and aceto-acetic
ethyl ester. Many other substances yield the reaction in a distinct
but not striking fashion, with the use of a much larger amount of
Included in this group are glucose,
the substance being tested.
galactose, lactose, maltose, pyrogallol, tricresol and tryptophane.
Metol
(monomethyl-p-amino-m-cresol)
and amidol (diaminophenol) react positively but atypically.
Many substances that
possess reducing properties in other directions do not yield this
reaction.
Negative are benzaldehyde, acetaldehyde, formaldehyde
(the faintest trace of color is evolved on prolonged heating) acetone, methyl-propyl
ketone, glycerol, chloroform,
levulinic acid,
uric acid, creatine, creatinine, acrolein, pyroracemic acid, phenol,
p-nitrophenol,
hydroquinone,
salicylic acid, tyrosine,
thymol,
orcein, hydroxylamine,
vanillin, phloroglucine,
tannic acid, sulphide, sulphite and phenyhsulphate.
Some substances that do
not give the reaction in acid solution react positively in alkaline
reaction; thus, phenol, tyrosine, thymol, orcein, tannic acid, and
uric acid. Tricresol, which acts in acid solution, does not react
in alkaline reaction.
On reviewing these substances, several interesting facts appear.
Thus indol and skatol react strongly; the parent substance, tryptophane, reacts feebly. Plain phenol and tyrosine do not react.
We are however inclined to believe that derivatives
of tyrosine
and phenylalanine contained in the urine do react, since the degree
of reduction effected by urine seems to be more than can be accounted for by the indol and skatol present.
The ketone group
That
does not react-but
the ketone sugar reacts very strongly.
the primary and secondary alcoholic groups, ketone and methyl, per
se do not reduce, is seen in the negative findings with acetone and

C. W. Miller

and A. E. Taylor

533

95;

Folin
Folin
1 Lewis

and Denis:
this Journal,
xii, pp.
and Macallum:
ibid.,
xiii, p. 363.
and Nicolet:
ibid.,
xvi, p. 369.

239,

245;

xiii,

p. 469;

xiv,

p.

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

glycerol.
But when a primary alcoholic group is attached to a
ketone group, as in levulose or dioxy-acetone, the reaction is proThe aldehyde group does not react strongly, as evinounced.
denced by the faint reaction with the aldoses as well as with the
simple aldehydes.
Hydrazine
reacts strongly,
as sulphate or
chloride; hydrasine reacts best in alkaline reaction, phenyl hydrazine, however, best in acid solution.
The hydrazines act energetically in the cold. They not only reduce the molybdate from the
ammonium salt, but also from the complex combination
with
phosphoric acid, upon which behavior we have founded a method
for the calorimetric estimation of traces of phosphoric acid.
It is interesting to contrast these results with those reported by
Folin2 and Lewis3 with the use of the tungsten and molybden reaThe reduction in acid
gents of Folin, used in alkaline reaction.
solution is much more selective and is effected by fewer substances.
The Folin calorimetric method for the estimation of uric acid rests
upon the fact that if the phenol and indol bodies are removed from
the urine, uric acid alone remains to react with the Folin reagent,
with the production of the blue color that is then measured by
com.parison with a standard.
These benzene derivatives
are
removed by evaporating the urine to dryness in the presence of
oxalic acid and extracting the residue with a mixture of ether and
methyl alcohol. Now our reagent reacts with these benzene derivWhen this reagent is applied to
atives, but not with uric acid.
the extract and residue respectively of the Folin method, the extract
is found to give a positive result and the residue gives no reactioni.e., the residue is free of the reacting benzene derivatives, thus
proving the operation of the Folin procedure to be quantitative
and reliable.
The reaction of normal urine to our reagent is due,
so far as we are able to determine, solely to derivatives of tryptophane, tyrosine and phenylalanine, most largely to indol and skatol.
Normal blood serum, freed of protein, does not give a demonstrable reaction, or at most the merest trace, comparable to that
effected by a small amount of glucose.
Upon the occasion of the
last meeting of the Society of Biological Chemists, when Professor
J. J. Abel of Johns Hopkins University
demonstrated
his method

534

Reduction

of Ammonium

Molybdate

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

we were able to apply the test to the

of so-called vividiffusion,
dialyzation fluid, which gave a reaction far more pronounced than
blood serum, and much more marked than the reduction effected
by a solution of glucose of the concentration
present.
From this
it may be inferred that the named aromatic substances are not
present in the blood except in the most minute traces, that they are
promptly and almost completely eliminated-into
the urine normally-into
the diffusion fluid of the vividiffusion
experiment.
The availability of this reagent for determining the presence of
levulose in the urine depends upon the fact that the latter substance
gives a far more striking reaction than either normal urine or glucose solutions.
The reducing power of urine in the natural state
cannot be studied satisfactorily
because of the confusing effect of
the reaction between the phosphates and the molybdate.
The
phosphates may be conveniently
removed as follows.
To 5 cc.
of urine are added 10 cc. of a 4 per cent solution of lead acetate.
After the mixture has been well agitated, 15 cc. of a 6 per cent solution of sodium sulphate are added, the whole well mixed and
filtered.
The filtrate is practically
free of phosphates and lead.
Five cc. of this filtrate are then mixed with 15 cc. of a 5 per cent
solution of sulphuric acid and the solution thus obtained represents
a dilution of the original urine of about 24:l.
Thisconsiderable
dilution is recommended because the colors are more easily compared when not too intense.
The reducing power of normal urine
may now be studied by mixing in a test tube one part of the final
diluted solution with one part of the reagent and placing the tube
in a boiling water bath for ten minutes.
Normal urine yields first
a faint green, becoming deeper, finally a moderate blue with a
greenish tinge. If comparable results are desired, time and
temperature
must be held constant.
Now if we dilute the contents of this tube, containing the reaction
of normal urine, about twenty-five
times (a total dilution of the
urine of about 600:1), the color will practically
disappear.
If
however levulose be present, as in the clinical test for hepatic
functionation,
a green or blue color will persist to a much greater
dilution.
Proceeding as follows, a urine containing as little as
0.5 per cent of levulose will yield a positive reaction.
One part of
the mixture of filtrate and 5 per cent sulphuric acid, described
above, is diluted with 25 parts of distilled water; and one part of

C. W. Miller

and A. E. Taylor

535

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

this mixture is added to an equal volume of the reagent and tested

and heated, as described, in a water bath. Working with such
high dilutions, the time in the bath may be extended to twenty
or twenty-five
minutes before normal urine will give more than a
very faint coloration, while the levulose-containing
urine yields a
goodly blue. In making the test, a control with a normal urine
should be run through with the test. A trace of glucose in the
urine does not seriously interfere with the test for levulose, since
glucose does not react as strongly as do the normal aromatic derivatives of the urine.
To what extent this test is capable of clinical
application, has not been determined.
Instead of the above-described clarification with lead, the phosphates of the urine may
be removed with barium hydrate or ferric chloride plus ammonia;
in either case, the filtrate is practically free of phosphates.
We have hnally compared the reduction due to the aromatic
substances of the urine, extracted with ether-methyl-alcohol
of
Folin, with the colors developed in the oxidation tests for indican,
using especially the bimolecular oxidation with isatine.
We have
not found the reactions to be parallel. On the one hand, the molybdate reagent reacts with other aromatic derivatives than indol and
skatol; on the other hand, the oxidation and extraction of indol is
rarely quantitative.
It is however possible that our reagent, or
the reagent of Folin, might be developed to serve for the colorimetric estimation of the indol and phenol derivatives contained in
urine.

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

ARTICLE:
ON REDUCTION OF AMMONIUM
MOLYBDATE IN ACID SOLUTION
C. W. Miller and A. E. Taylor

Access the most updated version of this article at

http://www.jbc.org/content/17/4/531.citation
Find articles, minireviews, Reflections and
Classics on similar topics on the JBC Affinity
Sites .
Alerts:
When this article is cited
When a correction for this article is posted
Click here to choose from all of JBC's e-mail
alerts
This article cites 0 references, 0 of which can be
accessed free at
http://www.jbc.org/content/17/4/531.citation.full.ht
ml#ref-list-1

Downloaded from http://www.jbc.org/ by guest on January 9, 2016

J. Biol. Chem. 1914, 17:531-535.