JANUARY 15, 1936 ANALYTICAL EDITION
Summary covery of fluoride added to apple pulp was 97.3 per cent under
The estimation of fluoride by titration with standard
thorium solution in the presence of sodium alizarin sulfonate
as indicator has been studied. The most favorable procedure
includes the use of the indicator a t a concentration of 4 x
lob6per cent, in a total volume of 50 cc., titration to match a
blank in which the end point is taken a t a very light pink
shade, and careful regulation of the p H in both blank and
sample, the most favorable p H being 3.5. This latter con-
dition is readily met by the use of the buffer system of sodium
hydroxide and chloroacetic acid at a ratio of 0.5 and total con-
centration of 0.02 M . The dissociation constant of chloro-
acetic acid in 50 per cent commercial alcohol has been found
to be 2.8 X Sodium alizarin sulfonate in this alcoholic
solution acts as an indicator for hydrogen ion over the pH
range 4.8 to 7.2 instead of 3.7 to 5.2 as in water. In a volume
of 50 cc. an average accuracy of 99 per cent has been secured
with known amounts of fluoride ranging from 57 to 760 y of
fluorine. With 5-cc. volumes approximately the same accu-
racy is possible with 6 to 90 y of fluorine.
Data are given regarding the concentrations a t which
several interfering ions have an effect. The most serious of
these are sulfate, arsenate, and phosphate, which fortunately
are left behind when the fluoride is distilled as hydrofluosilicic
acid. Sulfuric acid is entirely suitable for the distillation and
a volume of 200 cc. distilled a t 140" C. accounts for all but a
trace of the fluoride. I n the ashing of fruit samples contain-
ing fluoride for distillation it is very important not to allow the
temperature to reach above 800" C., for loss is then exces-
sive. Allowance must be made for the fluoride contained in
the lime which is added to the sample before ashing. Re-
Iodometric Determination of Copper
Adjustment of Hydrogen-Ion Concentration
WILLIAM R. CROWELL, THOMAS E. HILLIS, SIDNEY C. RITTENBERG, AND RAYMOND F. EVENSON
University of California a t Los Angeles, Los Angeles, Calif.
P ARK (3) has recently described a method of determining
copper in the presence of ferric iron and arsenic acid.
By his method an iodometric titration is carried out in a solu-
tion containing ammonium bifluoride and potassium bi-
phthalate. He states that the purpose of the bifluoride is to
suppress the action of the ferric iron on potassium iodide, and
that the biphthalate forms a buffer solution in which the pH
is about 4.0, a value sufficiently high to cause no appreciable
oxidation of iodide by the arsenic acid.
The problem of pH adjustment can be best understood if
we consider the technic involved in the process. After de-
composition of the ore with mineral acid, sufficient ammonium
hydroxide is added to neutralize the excess mineral acid,
arsenic acid, etc., precipitate the i r p , convert the copper to
the cupric ammonia complex, and produce a slight odor of
ammonia. The bifluoride and biphthalate are then added.
It is assumed that if the proper amount of biphthalate is
added, the concentrations of phthalate and of biphthalate
will be such as to produce an effective buffer solution a t the
pH desired. It is evident that an effective buffer action is
necessary to allow for a reasonable variation in the amounts
of excess ammonium hydroxide and of the other substances
which react with the weak acid. In this process the possi-
bility of the bifluoride's acting as a buffer evidently has been
overlooked. Hudleston and his co-workers (1) have shown
9
the most favorable conditions.
Acknowledgment
Acknowledgment is gratefully made of the assistance given
by Merle Randall of the Chemistry Department during the
writing of this paper, and by R. Craig of this division in the
study of the buffering properties of sodium chloroacetate in
alcoholic solution,
Literature Cited
(1) Armstrong, W. D., J . Am. Chem. Soc., 55, 1741-2 (1933).
(2) Barre, M.,Compt. rend., 151,231-4 (1910).
(3) Bowes, J. H.,and Murray, M. M., Biochem. J . , 29,102-7 (1935).
(4) Dahle, D.,J. Assoc. O$ciaZ Agr. Chem., 18, 194-7 (1935).
(5) Jukes, T.H., and Schmidt, C. L. A., J. Bid. Chem., 105,359-71
(1934).
(6) Kolthoff, I. M., and Furman, N. H., "Indicators," p. 97, New
York, John Wiley & Sons, 1926.
(7) Reynolds, D.S.,J.Assoc. O$ciaZ. Agr. Chem., 18,108-13 (1935).
(8) Reynolds, D. S., Ross, W. H., and Jacobs, K. D., Ibid., 11,
225-36 (1928).
(9) Sanchis, J. M., IND. ENQ.CHEM.,Anal. Ed., 6,134-5 (1934).
(10) Scott, W. W., "Standard Methods of Chemical Analysis," 4th
ed., Vol. I, p. 226,New York, D. Van Nostrand Co., 1925.
(11) Shuey, G.A., J. Assoc. Oflcial Agr. Chem., 18,156-7 (1935).
(12) Thompson, T. G.,and Taylor, H. J., IND. ENQ.CHEM.,Anal.
Ed., 5, 87-9 (1933).
(13) Vanselow. A. P..arivate communication.
(14) Willard, H. H., and Winter, 0. B., IND. ENG.CHEM.,Anal. Ed.,
5, 7-10 (1933).
(15) Winter, J. H., and Butler, L., J. Assoc. Oflcial Agr. Chem., 16,
105-7 (1933).
REO~I~F October
ID 5, 1935. Presented before the Division of Agricultural
and Food Chemistry at the 90th Meeting of the American Chemical Society,
San Francisco, Calif., Auguat 19 to 23, 1935.
that hydrogen fluoride in its aqueous solutions undergoes the
following equilibria:
H+ +F-$HF [H+l[F-J = 6.9 x 10-4 (250 c.) (1)
KI = ____
[HFI
HF + F- HF2- Kz = - [HF2-1 = 4.7 (25" C.) (2)
LHFl[F-l
The ionization constant of biphthalate a t 25" C. is 3.1 X
To obtain a p H of 4.0, the ratio of concentrations of
biphthalate to phthalate should be about 32 to 1, while the
ratio of concentrations of hydrofluoric acid to fluoride should
be about 1 to 7. At this pH the hydrofluoric acid should be
a much more effective buffer than the biphthalate, and the
p H a t which i t has maximum buffer efficiency is about 3.2.
It seems reasonable to suppose, therefore, that it should be
practicable to add such an amount of bifluoride that the con-
centrations of fluoride and of hydrofluoric acid formed would
be sufficient not only to produce the ferric complex but also
to yield simultaneously a buffer solution a t a p H between 3
and 4.
It is the purpose of the present paper to show that under
the conditions described by Park the p H a t the end point is
nearer 3.3 than 4.0, that this pH is high enough t o insure no
appreciable reaction between iodide and arsenic acid, that
the biphthalate plays practically no part in the adjustment
 10 INDUSTRIAL AND ENGINEERING CHEMISTRY VOL. 8, NO. 1
of the pH, and that it may be omitted without any material the percentage differences between the iodometric titration
effect on the accuracy or precision of the method. on the blank (performed as described in the first paragraph
under “Reagents”) and the average of the titrations in ques-
Reagents tion. The value of the blank was 25.37 CC.
The cupric sulfate solution was prepared from the salt
which had been twice recrystallized from a solution of the TABLEI. EFFECTOF IMPURITIESON TITRATION
OF COPPER
c. p. pentahydrate, and was made 0.1152 M in copper sulfate. IN BUFFERSOLUTIONS OF AMMONIUMBIFLUORIDEWITH
The concentration of this solution was determined electro- AND WITHOUT POTASSIUMBIPHTHALATE
(Copper taken, 0.1831 gram. Volume of thiosulfate required, 25.37 cc.)
lytically, and iodometrically by titration with thiosulfate
pH of Solution
solution which had been standardized against pure copper
foil. Results by the two methods agreed within less than NHiOH o?;’N
added NHlOH
0.2 per cent. Weight of Impurity Thio- t o faint added
Series Iron Arsenic NHdHFn KHCsHcOc sulfate odor in excess
The sulfuric acid, ammonium hydroxide, potassium iodide, Gram Gram Grams Grama Cc.
sodium thiosulfate, and potassium biphthalate consisted of
c. P. reagents which conformed to the standards of Murray (8).
1
2
0.2
0.2
0.0
0.0
2.0
2.0
0.0
0.0
25.35
25.39
3.3
... ...
4.0
The iron and arsenic added as impurities were supplied from 3
4
0.2
0.0
0.0
0.2
2.0
2.0
1 .o
0.0
25.39
25.34
...
3.3
4.0
...
solutions of c. P. ferric nitrate and arsenic acid. These solu- 5
6
0.0
0.0
0.2
0.2
2.0
2.0
0.0
1.0
25.36
25.35
...
... 3.9
3.9
tions contained approximately 0.1 gram each of iron and of 7 0.2 0.2 2.0 0.0 25.37 3.3
arsenic in each cubic centimeter of reagent. 8 0.2 0.2 2.0 0.0 25.37 ... 4:o
9
10
0.2
0.2
0.2
0.2
2.0
2.0
1.0 25.35 3.4
... ...
The ammoninm bifluoride was Merck’s Purified grade, 11 0.3 0.2 3.0
1.0
0.0
25.37 3.9
25.35 3.3
The starch solution was made from soluble starch prepared Average per cent error without biphthalate -0.04: with biphthalate,
“according to Lintner.” -0.02. Average deviatjon of each result from’ mean without biphthalate,
=t0.06 per cent; with biphthalate, k0.06 per cent.
Experimental Procedure and Results
Table I shows results of titrations of solutions of copper TABLE11. ANALYSISOF A MIXTURECONTAININGCUPRIC
sulfate containing iron and arsenic as impurities. In series SULFIDEAND ARSENOPYRITEBY THE BIFLUORIDE
1, 2, 4, 5, 7, 8, and 11 ammonium bifluoride, NHaFH2,alone METHOD
(Per cent of copper present, 14.72. Bifluoride added 2.0 grams. Conoen-
was used, and in series 3, 6, 9, and 10 potassium biphthalate tration of thiosulfate. 0.1022 N . j
was also present. The procedure was as follows: Weight Volume of pH at
Run of Sample Thiosulfate End Point Copper
To 25.00 cc. of copper sulfate solution containing 0.1831 gram Grams cc. %
of copper were added 5 cc. of a solution containing the impurities
as designated in Table I and 5 cc. of concentrated sulfuric acid.
This was followed by concentrated ammonium hydroxide until
the blue copper complex began to appear. Then 6 N ammonium
h droxide was added dropwise until the solution smelled faintly
ofammonia, and 2.0 grams of ammonium bifluoride and, in the Average percentage error $0.01.
runs indicated, 1.0 gram of potassium biphthalate were dissolved Average deviation of eac; result from mean, A0.07 per cent,
in the mixture. As soon as solution was complete, 10 cc. of
3 M potassium iodide were added and titration with thiosulfate
was carried out. This procedure is essentially the same as that Table I1 shows results of analyses of a mixture containing
described by Park except in the cases in which the biphthalate approximately 22 per cent of c. P. cupric sulfide and 10 per
was omitted. The amount of ammonium bifluoride added cor- cent of copper-free arsenopyrite mixed with finely ground
responded to 1 gram for each 0.1 gram of iron. unglazed porcelain which gave no test for iron when treated
pH measurements were made by means of the quinhydrone with hydrochloric and nitric acids. The cupric sulfide was
electrode on separate solutions containing in 100 cc. (the thoroughly mixed with the unglazed porcelain powder and a
approximate end-point volume) the same amounts of the number of samples accurately weighed out. Some of the
same constituents that were present before the addition of samples were analyzed for copper and the values thus ob-
the iodide. The iodide was omitted because it reduces the tained were used as a basis on which to determine the error
oxidized form of the quinhydrone and causes the pH deter- of the method. To each of the other samples was added
minations to be too high. Park in describing his procedure about 0.1 gram of arsenopyrite, the whole thoroughly mixed,
states that “the pH values of the solutions a t the end point and the copper determined by the bifluoride method.
were determined by means of the quinhydrone electrode.” The procedure used on the blanks was essentially the same
If this means that potassium iodide was present, that condi- as that employed in the standardization of the thiosuhte
tion alone might account for the fact that the authors’ values by copper foil. Runs were also made on the blank samples,
are so much lower than his. using the procedure described below for the mixture after
In the cases in which ammonium hydroxide was added until the addition of the arsenopyrite. The two sets of values
a faint odor was produced, the p H of the solution was approxi- agreed within 0.1 per cent.
mately 3.3. In order to obtain a pH of 4.0 it was found The procedure used on the cupric sulfide arsenopyrite
necessary to add about 2 cc. of G N ammonium hydroxide mixture was as follows:
more than that required to produce a faint odor of ammonia. Fifteen cubic centimeters of concentrated nitric acid were
This was introduced after the addition of the bifluoride. The added to the sample, evaporated to 5 cc., 10 cc. of concentrated
quinhydrone potential was always read after the addition of hydrochloric acid and 10 cc. of 18 N sulfuric acid added, and the
the bifluoride. When the biphthalate was used, it was added whole was evaporated to dense white fumes. In runs 1, 2, and
3, 10 cc. of concentrated nitric acid and 10 cc. of concentrated
after the bifluoride and its addition caused no recognizable hydrochloric acid were then added, and the solution was again
change in the potential of the solution. In the cases in evaporated to dense white fumes and diluted with 20 cc. of water.
which the p H was 3.3, as well as those in which it was 4.0, In runs 4, 5, and 6 the fuming was followed by the addition of
the end points were sharp and after the titration was com- 20 cc. of water and 10 cc. of saturated bromine water, and the
solution boiled until all bromine fumes were removed. These
pleted there was no further liberation of iodine for 15 minutes two types of procedure were found to be necessary to insure
or more. complete oxidation of the arsenic. After the addition of the
The values designated as “average per cent error” represent water and boiling, the procedure was the same as in the titration
JANUARY 15, 1936 ANALYTICAL EDITION
of the copper sulfate solutions previously described. In the
present case it was not necessary to filter off the insoluble residue
before making the copper titration.
Summary and Conclusions
I n the Park method of determining copper in the presence
of as much as 0.3 gram of iron and 0.2 gram of arsenic, the
potassium biphthalate may be omitted without any appre-
ciable effect on the accuracy or precision of the results.
The addition of biphthalate has no material effect on the
pH of the solution.
The p H at the end point is nearer 3.3 than 4.0 and yet the
end point is practically permanent.
11
To insure complete oxidation of an ore containing sulfide,
iron, and arsenic, treatment with nitric acid alone is not
sufficient. A double treatment with nitric and hydrochloric
acids or a single treatment with the two acids followed by
one with saturated bromine water is found necessary.
Literature Cited
(1) Hudleston, J. Chem. Soc., 125, 260, 1122, 1451 (1924); 127, 1122
(1926).
(2) Murray, “Standards and Tests for Reagent Chemicals,” New
York, D. Van Nostrand Co., 1920.
(3) Park, IND.ENG.CHEM.,Anal. Ed., 3, 77 (1931).
RECEIVED
October 15, 1935.

The Determination of Rhenium

I. Qualitative
LOREN C. HURD, University of Wisconsin, Madison, Wis.

INCE the discovery of rhe- Although a number of papers have ap- acid by weight it has been the
S nium in 1925 (28) there have
appeared a number of papers
peared dealing wholly or in part with reac-
tions of rhenium of value in the qualita-
author’s experience as well as
that of others (9) that precipita-
dealing wholly or in part with tion is not quantitative, In light
various qualitative tests for the t h e detection of the element, no mention of this rather peculiar condition
identification of the element. has been made of its place in the conven- it was thought desirable to ascer-
Its place in the Noyes and Bray tional scheme. Data are presented which tain where in the conventional
system was studied by Kao and indicate that in the Prescott and Johnson scheme of analysis the element
( I 6 ) who found that system the element will concentrate with would be c o n c e n t r a t e d and
s i g n i f i c a n t concentration was what precautions must be taken
effected in the tellurium group. arsenic* Early work in the has been in order to insure a clean-cut
These authors reported that dur- critically examined and evaluated and separation,
ing the course of the analysis of several new tests are reported. New con- A stock solution was prepared to
the group, tellurium and rho- firmatory tests are described. which were added 5 mg. Der 50 ml.
dium could be precipitated in the of each of the follo&i metals:
presence of rhenium by reducing Hg(ic), Pb, Cu, Cd, As(oG), As(ic),
with hydrazine hydrochloride and sodium bisulfite in hydro- Sb, Sn(ic), Cr(ic), Fe(ic), AI, Ni, Co, Zn, Mn, Ca, Sr, Ba, Mg, K,
chloric acid solution. Rhenium was subsequently precipi- and Na. Fifty-milliliter portions of this solution were carried
through the qualitative scheme of Prescott and Johnson (19).
tated as the sulfide from the filtrate. Each group precipitate was washed thoroughly, digested with
The insolubility of rhenium heptasulfide was one of the sodium hydroxide to expel all ammonia, oxidized with hydrogen
characteristic properties early reported. Although this was peroxide, acidified with sulfuric acid, and treated with nitron
acetate as described by Geilmann and Voigt (7).
a controversial subject for several years (Id) it has been In the case of solutions containing chromium and manganese
definitely established by several workers, notably Geilmann it was necessary to add a few drops of alcohol to the acid solution
and Weibke (9),that the sulfide is quantitatively precipitated prior to the nitron precipitation in order to reduce compounds of
when a perrhenate solution containing as high as 33 per cent higher valence which yield precipitates with the reagent. Alco-
hol in low concentrations is without influence in the nitron pre-
of hydrochloric acid by weight is treated with hydrogen cipitation (SI). Certain normal constituents of complete group
sulfide. This separation is now used extensively in the precipitates were excluded because it was found that they yielded
analysis of rhenium. For reasons to be discussed below, the insoluble nitron derivatives. For this reason tungstates, molyb-
sulfide precipitation is not applicable to all rhenium-contain- dates, palladium, gold, chloroplatinates, and germanium were not
included in the stock solution.
ing samples. A series of blank determinations using the stock solution was
carried through the Prescott and Johnson separation and each
TABLEI. RHENIUM FOUND BY PRESCOTT AND JOHNSON group or subgroup analyzed according to the nitron method.
SYSTEM No precipitate was obtained in any case. Fifty milligrams of
(50 mg. of Re added) rhenium as KRe04 were then added to a 50-ml. portion of the
Sample 1 Sample 2 Sample 3 Sample 4 stock solution and the separation and analysis repeated. Group 1
Mg. % Mg. % Mg. % MQ. % reagents yielded no precipitate. When hydrogen sulfide was
First analysis 1.5 3 . 0 4 . 3 8 . 6 4 . 7 9 . 4 3 . 5 7.0
passed into the 0.25 N hydrochloric acid solution for second
Second analysis 1 . 4 2 . 8 6 . 4 12.8 1 1 . 9 2 3 . 8 16.1 3 2 . 2 group precipitation, some trouble was experienced in obtaining a
Third analysis 44.7 8 9 . 4 2 9 . 9 5 9 . 8 2 9 . 1 5 8 . 2 2 8 . 8 5 7 . 6
- - - - _ _ - ~ flocculent precipitate. It filtered without difficulty, however,
Total 47.6 95.2 4 0 . 6 81.2 45.7 91.4 48.4 96.8
and was analyzed (SO) for its rhenium content (first analysis,
Table I).
The filtrate from the first precipitation with hydrogen sulfide
was evaporated to a volume of 8 ml. for the arsenic precipitation.
Prescott and Johnson System During the course of the concentration sulfur and rhenium sulfide
separated from the solution. This was removed by filtration
Although rhenium heptasulfide is quantitatively precipi- and analyzed for rhenium (second analysis, Table I). Samples
tated from solutions containing a relatively high concentra- 2,3, and 4 were allowed to stand overnight before filtering.
The solutions were then acidified and saturated with hydrogen
tion of hydrochloric acid, the precipitation takes place slowly. sulfide to precipitate the arsenic. The mixed sulfides were
I n solutions containing less than 4 per cent of hydrochloric analyzed for rhenium (third analysis, Table I).