A study of the Mohr and Volhard methods of determination of chloride content in brines

a m m
o f th e m m
mmm®
and m im m
of oetebmimatiom
OF CHLORIDE m m M M T
m
m m m
A
Thesis
Presented to the Departsent a t Chemistry
Coffersity of Southern California
fa partial fulfillment
of the
acquirements for the
Degree of Master of Science
By
Charles.6« Carlson
Cun# #* 1939
UMI Number: EP41514
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI
UMI EP41514
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106 - 1346
d *0
T h i s thesis, w r i t t e n by
Charles G. Carlson
u n d e r the d i r e c t io n o f h ..-Lf F a c u l t y C o m m it t e e ,
a n d a p p r o v e d by a l l it s m e m b e r s , has been
presented to a n d accepted by the C o u n c i l on
G r a d u a t e S t u d y a n d Research in p a r t i a l f u l f i l l ­
m ent o f the re q u ire m e n ts f o r the degree o f
Master of Science
Dean
Secretary
D a te .
F a c u lty Com m ittee
\atrman_
AT*
CONTENTS
INTRODUCTION. *.......... *..............
I
MOHR'S METHOD..
......
7
EXPERIMENTAL............
7
Materials
Apparatus.
......
7
........
Determination of the Chloride
Content in the Brines
.......
Determination of Chloride by
the Mohr Method.......
VOLHARD'S METHOD---- .......................
The Volhard Method of Chloride
Determination
.............
EXPERIMENTAL.
..........
8
9
II
16
16
28
Materials.............
22
Apparatus .............
23
Determination of Chloride
Content by Volhard's Method;......
23
GRAVIMETRIC CHECK
....
29
SUMMARY AND CONCLUSIONS. ...
30
BIBLIOGRAPHY
33
1
A STUDY OF THE MOHR AND VOLHARD
METHODS OF DETERMINATION
OF CHLORIDE CONTENT
IN BRINES
INTRODUCTION
In the volumetrie determination of chloride content
in weak and strong brines the methods of Mohr and Volhard
are fairly rapid, simple and accurate, and therefore have
been very popular*
However, since both methods are based
on the formation of a characteristic color at the end­
point of a titration, the accuracies of the methods de­
pend upon the completeness of these titrations*
Before
the two methods can be compared for their relative
accuracies and the concentration range in which each is
best suited, a thorough review of the original methods
and the variations that have been suggested since then
are necessary in order to find the simplest and most
accurate form of each method*
First, the method of Mohr which was first given in
18561 will be taken up*
In a neutral or weakly acid
solution in which the chloride content is to be deter­
mined advantage is taken of the different solubilities
F* Mohr, Liebig's Ann*, 97, 335 (1856)*
2
of the two precipitates —
silver chloride and silver
chromate»
The solubility of silver chrornate in grams
-3
per 100 grams of water is 2*5 x 10 * That of silver
-4
chloride is 1*5 x 10 , or, in other words, the silver
chromate is almost twenty times as soluble as silver
chloride*
Therefore it is impossible for red silver
chromate to form permanently in a solution of chromate
and chloride until practically all of the chloride has
been precipitated*
The procedure is to prepare a stan­
dard silver nitrate solution of approximately 0*1N
strength*
A one per cent solution of potassium chromate
is used as the end-point indicator*
If it is desired to
express the chloride concentration in parts per million
(as is done in most water and brine analyses) a Known
volume of the test solution is placed in a Nessler tube,
one ml* of the one per cent potassium chromate solution
is added and then, with stirring, the standard silver
nitrate solution is added at the rate of about one ml*
per second, pausing every three or four ml* to stir the
solution*
As the end-point is approached more time and
stirring are necessary to cause the red color to fade*
During this last part of the reaction the silver nitrate
must be added drop at a time, and at the very end it is
advisable to split drops in half*
The first appearance
of a permanent red color indicates that all of the
3
chloride has been taken out of solution*
It is always
necessary to run a blank experiment using distilled water
and the chromate indicator so as to determine how mueh
of the silver nitrate is necessary to produce the per­
manent red shade in the absence of chloride.
A small
amount of calcium carbonate mixed with the solution
gives about the same turbidity as does the silver chloride
in the regular titration, and thus the matching of the two
colored solutions is made easier.
The amount necessary
to color the blank must be subtracted from the total
silver nitrate used in the main titration*
It is very important that the titration be run in
neutral or weakly acid solutions.
Acid solutions should
be neutralized by calcium carbonate or should be partly
neutralized by ammonium hydroxide and then kept almost
neutral by addition of ammonium acetate.
It is necessary
to remove the acid because silver chromate is soluble in
acid, and it is necessary to remove any hydroxide because
of the formation of brown silver oxide.
It is suggested
that the Mohr method gives best results for small amounts
of chloride in concentrated solutions.
Large volumes
cause low accuracy*
Meldrura and Forbes2 claim that even though the Mohr
method is rapid and simple, the results obtained can only
^MSldrum and Forbes, wThe Volumetric Determination of
Chloride*
J. of Chem. Education, Vol. 5, 1928.
4
be considered as approximately correct tinder the most
favorable conditions*.
They further claim that there are
three sources of error in the method, namelyt
(1) The
temperature effect of the solubility of silver chromate.
At zero degrees the solubility of silver chromate is
*00015; at 70 degrees it is .008 grams per 100 grams of
water, or at 70 degrees the precipitate is roughly fifty
times as soluble as it is at 0 degrees.
(2) A lack of
sensitiveness of the end-point due to too great acidity
or alkalinity of the solution.
(3) The adsorption of
soluble chloride by the precipitated silver chloride.
The first source of error may be almost completely
avoided by running the titration in a cold solution and
by taking care that the temperature of the blank solution
is identical with that of the main test solution.
Similarly the second source of error may be omitted
by a careful adjustment of the ptt of the solution*
Error number three, however, may not be avoided so
easily.
Upon addition of the silver nitrate solution to
the test chloride solution, the silver chloride that is
precipitated first comes down in a finely divided condi­
tion.
But as the end-point is reached coagulation begins
to take place and the silver chloride forms in curds*
It
is at this point that Keldrum and Forbes claim that ad­
sorption takes place.
If, by some means, this adsorption
5
could be avoided either by preventing coagulation or by
getting the adsorbed chloride back into the solution they
suggested that better results might be expected.
The variation, offered by Meldrum and Forbes is as
followst
Mohrrs method is followed as given up to the end­
point.
At this point the solution containing the
curded
precipitate is boiled for five minutes, cooled to 20
degrees 0*, or lower, and
then titrated again
with silver
nitrate to a faint end-point*
Lottermoser and Lorenz3 offer a modification of the
Mohr method which retards or prevents coagulation of the
silver chloride precipitate and thus, they claim, the
adsorption of chloride on the curded precipitate is pre­
vented.
They found that in the titration of strong
chloride solutions of Li, Na, K, Mg, and Ca the results
were unsatisfactory.
"The error increased in the above
series in agreement with increased coagulating power."
By addition of from five to ten ml. of a one per cent
agar solution coagulation and curding of the silver
ehloride precipitate could be avoided and a sharper
end-point could be brought about*
lottermoser and Lorenz, Kolloid Z ., 68, p. 201-3 C1934).
6
By combining these two modifications with the original
method given by Mohr, four different modifications of the
Mohr method were laid out*
(1)
They are t
The original titration as given by Mohr is
followed exactly*
Potassium chromate is added to
the test solution and the resulting solution is
titrated to a faint red end-point with *1 N silver
nitrate*
(2)
The titration is run to a faint end-point;
the solution containing the curded precipitate
is boiled for five minutes, cooled and then ti­
trated again with silver nitrate to a faint end­
point*
(3)
To the test solution one ml* of a one per
cent solution of agar is added along with the
potassium chromate indicator*
Silver nitrate
is added until the end-point is reached.
(4)
The test solution, to which has been added
agar and potassium chromate, is titrated to an
end-point and then boiled for five minutes after
which silver nitrate is again added until a
faint red end-point appears*
Test runs were made by these four variations on
solutions of chloride of different concentrations.
7
EXPERIMENTAL (MOHR'S METHOD)
MATERIALS
The fallowing reagents were prepared from so-called
‘'chemically pure1* chemicals:
Silver nitrate solution:
16.9889 grams of fresh, crys­
talline silver nitrate were dissolved in a liter of
distilled water.
The solution was .1 N.
Potassium chromate solutions:
Three solutions of potas­
sium chromate containing one, two, and five grams
of the salt in one hundred ml* of distilled water
were prepared giving solutions of one, two, and
five per cent.
Agar solution:
One gram of agar was dissolved partially
in one hundred ml. of distilled water giving a
solution of approximately one per cent.
Chloride solutions:
Five solutions of one liter each
were prepared containing roughly ten, five, two,
two-tenths, and two-hundredths grams of sodium
chloride per liter.
These gave solutions of twenty,
two hundred, two thousand, five thousand, and ten
thousand parts per million.
8
APPARATUS
The silver nitrate was run into the chloride solution
from a fifty-ml. burrette.
The agar and chromate solutions
were measured roughly in a ten-ral* graduated cylinder*
Exact portions of the chloride solutions were measured out
in ten, twenty-five, and fifty-ml* pipettes*
The twenty,
two hundred, two thousand, and five thousand ppm. titra­
tions were carried out in fifty-ml. Nessler tubes; the
ten thousand and twenty ppm. titrations were run in a
one hundred ml. flask.
9
DETERMINATION OF THE CHLORIDE CONTENT IN THE BRINES
Before the aetuaL determination was started, a
series of blanks was run to determine the amount of sil­
ver nitrate solution necessary to produce the faint red
color at the titration end-point*
Twelve blanks were
man for each of the five solutions, using approximate
amounts of distilled water equal to the volume of the
solution at the end of the titration*
The four above-
mentioned modifications of the titration procedure were
carried out using one ml* of the one, two and five per
cent potassium chromate solutions, respectively*
Small
amounts of calcium carbonate, varying in amount so as to
correspond to the particles of white silver chloride
precipitate, were added in order to approximate the con­
ditions met in a real titration*
A blank was run before
each set of titrations and the end-point of the titration
was made to match the color of the blank.
Collectively,
the amounts of silver nitrate necessary to produce the
desired end-point color are presented in a table as
followst
10
TABLE I
Cl content in ppm.
_________________
20(100) ml. sample
200(50) ml* sample
2000(25) ml. sample
5000(25) ml. sample
10,000 (20)ml. sample
Amount of Ag NOg in ml.
1% Or Oa
Zf» Qr 04
5% Or 0A
.10
.10
.08
With agar
.08
.08
.07
Without agar
.18
.17
.17
Boil with agar
.1C
.14
.12
Boil without agar
.04
.04
.04
With agar
.03
.04
.03
Without agar
.09
.09
.07
Boil with agar
.08
.07
.07
Boil without agar
.04
.04
.04
With agar
.03
.03
.03
Without agar
.10
.09
.10
Boil with agar
.08
.08
.08
Boil without agar
.04
♦04
.04
With agar
.03
.04
.03
Without agar
.10
.10
.10
Boil with agar
.07
.08
.07
Boil without agar
.03
.04
.03
With agar
.03
.03
.03
Without agar
.10
.10
.09
Boil with agar
.08
.08
.08
Boil without agar
II
Determination of Chloride by the Mohr Method
One hundred ml* of the approximately twenty parts per
million chloride solution were placed in a ISO ml* flask.
One ml* of the one per cent chromate solution was added
and the solution was titrated by addition of the .1 N
silver nitrate solution from a burette until the first
appearance of a permanent red color*
to match a blank*
The color was made
Four titrations were made using the
one per cent chromate solution*
Mext, the same procedure
was followed using the two, and finally the five per cent
chromate solutions.
The amount of silver nitrate neces­
sary to complete the titration in each case was recorded
and the amount of silver nitrate necessary to produce the
required blank coloration was subtracted, giving the total
re quired amount *
The whole series of titrations was next repeated, but
this time one ml* of the one per cent solution of agar was
added along with the chromate indicator.
A third series of titrations was run, this time agar
was excluded but after the first appearance of the per­
manent red color the solution was boiled for five minutes
during which time the red color disappeared;
The solution
was then cooled in a running water bath and titrated again
with the silver nitrate solution to a permanent end-point.
12
The same procedure was followed with solutions containing
one ml. of the one per cent solution of agar*
This gave twelve titrations for each of the four
modifications of Mohr*s method.
Exactly the same proce­
dure was followed with fifty ml. samples of the 200 ppm.
chloride solution, twenty-five ml. samples of the 2000
ppm. solution, twenty-five ml. samples of the 5000 ppm.
solution, and twenty ml. samples of the 10,000 ppm.
solution.
The amounts of silver nitrate necessary to
titrate to the end-point were recorded in Table II.
13
TABLE II
2% CrOd
Cr04
Amt.
Dev. Am t.
5£ Cr01
Dev. Amt.
Dev
20 ppm . So In.
With agar
.04
.04
.38
.03
o
•
to
o«
Without agar
.37
.01
.39
.02
.42
.02
Boil with agar
.40
.01
.38
.02
.36
.01
Boil without agar
.40
.01
.37
.01
.40
.03
With agar
1.76
.01
1.77
.02
1.76
.01
Without agar
1.77
.02
1.76
.01
1.77
.03
Boil with agar
1.76
.03
1.75
.00
1.74
.00
Boil without agar
1.76
.01
1.75
.00
1.76
.01
With agar
17.20
.02
17.22
.01
17.20
.01
Without agar
17.18
.02
17.14
.02
17.14
.02
Boil with agar
17.20
.04
17.20
►05
17.18
.02
Boil without agar
17.15
.01
17.13
.02
17.15
.02
With agar
42.74
.01
42.73
HI
O
•
42.74
o•
Without agar
42.56
.00
42.55
.01
42.57
o
o•
Boil with agar
42.72
.01
42.74
.02
42.72
.01
Boil without agar
42.52
.01
42.51
o
o•
42.53
.01
With agar
41.03
.02
41.03
.01
41.05
.02
Without agar
40.83
.04
40.85
.02
40.82
to
o•
Boil with agar
41.03
.02
41.01
.01
41.03
.02
Boil without agar
40.79
.02
40.84
.03
40.80
.01
200 ppm. SoIn.
2000 ppm. SoIn.
5000 ppm. SoIn.
O
10,000 ppm. So In.
14
By combining a critical laboratory study with a thorough
investigation of the literature, of the Mohr method of
chloride determination, the following facts were deduced;
1*
In the determination of chloride content in
brines of a concentration of 2000 ppm. and
higher there is a decided difference in the
amounts of silver nitrate needed to produce
an end-point in the cases when agar was added
and when agar was omitted.
The difference
increases with increased coagulation, which,
in turn, is controlled by the concentration
of the chloride ion.
Addition of agar pre­
vents curdling of the silver chloride, thus
preventing adsorption, and therefore gives more
accurate results.
2.
In concentrations of 200 ppm. and lower with
respect to chloride concentration there is no
appreciable adsorption due to coagulation of
the silver chloride precipitate and therefore
the addition of agar to the solution produces
no better results in concentrations of 200
ppm. and lower.
3.
In concentrations of 2000 ppm. chloride and
higher the addition of agar to the solution
allows a quicker end-point*
The reason is
15
that curdling is prevented; this makes impos­
sible the formation of large curds of red silver
chromate at the point where the silver nitrate
enters the solution*
The usual time needed to
convert these curds of silver chromate over
into the Less soluble silver chloride is not
necessary*
4*
Mdition of agar to solutions of 2000 ppm* and
higher produces a sharper end-point.
The end­
point, when once reached, is sharp, does not
fade due to the breaking up of curds and thus
liberation of unacted-upon chloride*
5*
Results showed, in agreement with the predic­
tions of Meldrum and Forbes, that both solutions
containing agar and those not containing agar do
require a small additional amount of AgKOj to
produce an end-point after the solution is
boiled*
However, a similar amount of additional
silver nitrate is also necessary to produce an
end-point in the blank solutions after boiling.
Therefore, boiling the solution does not improve
the method*
£.
There is no evidence to show that a quicker,
sharper or more accurate end-point is produced
by varying the concentration of the chromate
16
indicator between the one, two and five per cent
solutions.
VOLHAHD»S METHOD
The Volhard Method of Chloride Determination
Hie Volhard method of chloride determination was
4
originally given by Charpentier in 187G, and is primarily
a method to determine silver by titrating the solution
containing silver and ferric iron with a standard solu­
tion of potassium thiocyanate*
Later (1874 and 1878)
Volhard presented the method in complete form,
and it
is because of his publications that the method now bears
his name*
As in the case of the Mohr method, Volhard*s method
is based on the difference in solubility of two silver
precipitates, white silver thiocyanate, and red ferric
thiocyanate.
No permanent formation of red ferric thio­
cyanate is possible until practically all of the silver
has been precipitated as white silver thiocyanate*
As
applied to chloride determination, the method is as
follows
4P. Charpentier, Bull. Soc. Ing. Civ* France, 135, 385 (1870)
5J.. Volhard, J. pradt. Chem. (8), 9 , 817 (1874);
Eieb'igFs AnHTT 190, 1 (1878)
%illebrand and Lundell, Applied Inorganic Analysis, New
York, p. 164 (1989) .
Treadwell-Hall, Analytical Chemistry, New York, Vol. II
p. 654 (1938)
17
To a nitric acid solution of chloride containing a
few ml* of ferric ammonium alum, standard silver nitrate
is added until complete precipitation of the chloride has
taken place*
A few ml* in excess of the silver nitrate
standard are added.
The excess silver is titrated with
standard potassium thiocyanate, and by subtracting the
excess silver nitrate found in this titration from the
total amount added, the amount of silver equal to the
amount of chloride in the solution is found.
From this
the total amount of chloride may be calculated.
Volhard stated that the results were good in cases
where fairly concentrated solutions of chloride were
analyzed, but that too high values were obtained in
solutions of low concentration.
7
Dreschsel, Rosanoff and Hill claimed that in its
simple form the Volhard method gave unsatisfactory re­
sults.
Because of the fact that silver chloride is more
soluble than silver thiocyanate it is evident that the
precipitated silver chloride in the solution will react
with the ferric thiocyanate, giving solid silver thio­
cyanate and ferric chloride according to the reaction;
3 Ag Cl + Fa (CHS)3 a 3 Ag CNS + Fe CI3 *
Therefore,
they claim, it is impossible to get a permanent end-point
that will not fade on stirring until after a considerable
7G* Dreschsel, 2. anol. Chem., 16, 351 (1877).
18
amount of potassium thiocyanate has been added in excess.
To overcome the action of the silver chloride of the
ferric thiocyanate several variations of the Volhard
method have been suggested.
Dreschsell, Rosanoff and Hill suggested the following
method to overcome the effect of the silver chloride:
An excess of silver nitrate standard solution is
added to the chloride test solution in a 200 ml. measur­
ing flask; it is made acid with nitric acid and the re­
sulting solution thoroughly shaken until the liquid above
the precipitate clears on standing a few seconds.
The
measuring flask is filled up to the 200 ml. mark with
distilled water, the solution is then filtered through a
dry filter.
The first ten or fifteen ml. of the filtrate
are rejected; 50 or 100 ml. samples of the filtrate are
treated with the ferric alum indicator.
The excess silver
nitrate is then titrated by the standard potassium thio­
cyanate solution in the absence of the interfering silver
chloride.
In the calculation account is taken of the
fraction of the solution used for titration*
Rothmund and Bugstaller 8 claim that filtering is
not necessary.
Instead they suggest that, after the
addition of the excess silver nitrate, the solution be
^Rothmund and Bugstaller, Z. anorg. Chem., 63, p. 333 (1909)
48, p. 79 (1909)
19
heated moderately, cooled and then shaken until, upon
standing a few seconds, the supernatant liquid is clear.
The indicator is then added and the solution is titrated
with the standard thiocyanate solution.
Coagulation of
the silver chloride by heating, they claim, prevents its
action on the ferric thiocyanide.
q
Caldwell and Moyer have also suggested a method
that excludes the necessity of filtering off the silver
chloride precipitate.
If one ml. of nitrobenzene be
added for each mg. of chloride present, and the resulting
mixture shaken vigorously in a stoppered bottle, the
nitrobenzene coats the particles of silver chloride pre­
cipitate and prevents the precipitate from reacting with
the ferric thiocyanate.
The nitrobenzene also inhibits
the effect of light on silver ehloride, preventing the
darkening of the precipitate and thus the end-point is
improved.
Because the nitrobenzene is heavier than water
the mass of silver chloride and nitrobenzene sinks to the
bottom and gives a very clear solution on top.
The Dreschsell modification requires an extra opera­
tion of filtering and therefore takes longer.
However,
the addition of nitrobenzene as suggested by Caldwell and
Moyer, and the procedure given by Rothmund and Bugstaller,
9Caldwell and Moyer, 2T. anal. Chetn:., 99, p. 258-*69 (1934)
20
in which the solution is gently heated, require very little
extra time and, if accurate, would be the more desirable
methods*
These two modifications in addition to the original
as given by Volhard, give three rapid and simple variations:
(1)
Volhard*s original method Ls followed as given.
The chloride is precipitated with a standard
silver nitrate solution and a few ml* of silver
nitrate in excess are added.
Ferric alum is
next added as an indicator and the excess silver
nitrate is titrated with a standard potassium
thiocyanate solution*
The excess silver is
subtracted from the total amount, giving a re*^
mainder of silver nitrate equivalent to the
amount of chloride originally present*
(2)
As in the original method the chloride is pre­
cipitated with standard silver nitrate and a few
ml* excess are added.
At this point the solution
containing the precipitate is gently heated,
causing complete coagulation of the precipitate
and a clear supernatant liquid.
Ferric alum in­
dicator is added and the excess silver nitrate
is titrated with standard potassium thiocyanate.
21
(3)
Nitrobenzene is added to the test solution at
the start along with the ferric alum*
The ti­
tration is carried out in a stoppered bottle
and after the addition of the excess silver ni­
trate the stopper is put in the bottle and the
solution is shaken vigorously for about thirty
seconds, after which the particles of silver
chloride which have become surrounded by a film
of nitrobenzene sink to the bottom leaving a
clear solution*
The excess silver is next ti­
trated with standard potassium thiocyanate.
Tests were run on the same chloride solutions as were
used in the above mentioned Mohr titrations.
22
EXPERIMENTAL (VOLHAKD*S METHOD)
MATERIALS
The following reagents were prepared;
Silver nitrate solution:
16.9889 grams of fresh, crystal­
line silver nitrate were dissolved in a liter of
distilled water giving a solution of *1 N*
Potassium thiocyanate solution:
Approximately 5*86 grams
of potassium thiocyanate crystals were dissolved in
a liter of distilled water*
This solution was
standardized against the . I N silver nitrate solu­
tion*
The solution was *0489 M«
Eerric ammonium sulfate solution:
About ten grams of
ferric ammonium sulfate crystals were placed in a
flask with about 100 ml. of distilled water and the
solution was stirred until saturated*
The clear
liquid was decanted off, and concentrated nitric
acid was added until a greenish yellow color appeared.
Nitrobenzene;
Stockroom nitrobenzene was used*
Sodium chloride solutions;
Eive solutions of one liter
each were prepared containing about ten, five, two,
two-tenths, and two-hundredths grams of sodium
chloride per liter.
The solutions contained twenty,
two hundred, two thousand, five thousand, and ten
thousand parts per million of chloride.
23
APPARATUS
Th© two standard solutions, silver nitrate and potas­
sium thiocyanate, were added to the test solution from
fifty ml* burettes*
flasks*
The titrations were run in stoppered
Exact portions of the chloride solutions were
measured in ten, twenty-five and fifty-ml. pipettes*
The
ferric alum indicator and the nitrobenzene were measured
roughly in a ten-ml* graduated cylinder*
The silver
chloride solution was heated over a small gas flame and
cooled in a running water bath*
DETERMINATION OF CHLORIDE CONTENT BY VOLHARD»S METHOD
Fifty ml* of the two hundred parts per million
chloride solution were measured out in a pipette and run
into a 250^ml* measuring flask*
Eight to ten drops of
concentrated nitric acid were added and the chloride
precipitated with the standard silver nitrate solution*
Three or four ml* of the standard in excess were run
into the solution*
The bottle was stoppered and the
solution was shaken vigorously for about thirty seconds.
Next one ml* of the ferric alum indicator was added to the
solution and the excess silver nitrate was titrated with
the standard potassium thiocyanide until a reddish-brown
end-point was reached.
A false end-point which faded
after about thirty seconds usually appeared approximately
24
one drop from the true end-point.
using this method*
Three runs were made
The results were recorded in Table
III*
Next fifty ml. of the same solution were pipetted
into a Pyrex flask, the solution was precipitated with
an excess of the standard silver nitrate solution and
then the precipitate was coagulated by placing the flask
over a small flame for twenty or thirty seconds.
The
solution was cooled in a running water bath, one ml. of
the indicator was added and the excess silver nitrate
was titrated with the standard potassium thiocyanate
solution.
Three runs were made using this method.
Fifty ml. of the same solution were run into a
250-ml. measuring flask*
Five ml. of nitrobenzene were
added and the chloride precipitated with the standard
silver nitrate solution.
The bottle was stoppered and
the solution was shaken very vigorously for thirty
seconds.
After standing for a few seconds one ml. of
the ferric alum indicator was added and the excess silver
nitrate was titrated with the standard thiocyanate solu­
tion.
Three runs were also made using this method.
In the same way, 10Q ml. of the 20 ppm., fifty ml.
of the 2000, and 5000, ppm. solutions, and twenty ml. of
the 10,000 ppm. solution were treated and titrated and
the chloride content determined by the three variations
25
of \TQlhard*s method*
Ten ml. of nitrobenzene were added
in the 5000 and 10,000 ppm. runs.
Results of the titra­
tions are as followst
TABLE III
Total AgNO^
Deviation
20 ppm. SoIn.
Ag Cl treated with
nitrobenzene
.24
.10
.30
Avg.
Soln. heated
.21
.22
.15
.10
Avg.
Ag Cl untouched
+.0$
-.01
-.OS
.18
.13
.03
.23
Avg.
4*03
-.11
+.09
-.00
-.10
4.10
.13
200 ppm. Soln.
Ag Cl treated with
nitrobenzene
1.75
1.71
1.66
Avg.
Soln. heated
1.71
1.59
1.50
1.50
Avg.
Ag Cl untouched
(Table continued)
4.06
-.03
-.03
1.53
1.50
1.26
1.11
Avg.
4.04
-.00
-.05
1.29
+.21
—.03
-.18
26
TABLE III (continued)
Total AgNOg
Deviation
2000 ppm* Soln*
Ag 01 treated with:
nitrobenzene
17*11
17.17
17*16
Avg.
Soln. heated
17.15
17.13
16.99
17.05
Avg*
Ag 01 untouched
+.07
-.07
-.01
17.06
16.20
16.55
16.15
Avg.
-.04
f*02
+ .01
-.10
+ .25
-.15
16*30
5000 ppm. Soln*
Ag 01 treated with
nitrobenz ene
42*72
42.69
42.59
Avg*
Soln* heated
42.66
42.30
42.37
42*33
Avg.
Ag 01 untouched
-.03
+ .04
£.oq
42*33
42*10
42.26
42.1&
Avg.
+ .06
+ .03
-.07
-.07
+.09
-.03
42.17
10,000 ppm * Soln*
Ag 01 treated with
nitrobenzene
41.32
41.18
41.22
41.24
+ .08
— .06
-.02
27
TABLE III (continued)
Total AgNQ-g
Soln* heated
Deviation
40.90
40.62
40.56
Avg.
Ag Cl untouched
40.69
40.82
40.77
40.61
Avg.
+.21
-.07
-.13
+.09
+.04
-.12
40.73
A thorough study of these results reveals the follow­
ing factst
1.
With each of the four chloride solutions the
addition of nitrobenzene succeeded in rendering
the interfering silver chloride inactive.
Much
less, if any, of the silver chloride changed
over into the less soluble silver thiocyanate
and therefore with each concentration the amount
of equivalent silver nitrate required was greater
in the case where nitrobenzene was added.
The
method in which nitrobenzene is added gives the
best results.
2.
In all cases the average deviation was less in
the case where nitrobenzene was added than it
was when the solution was heated to cause coagu­
lation, and in every case was less than it was
when the silver chloride was left untouched.
28
Therefore, the determination of the end-point
is more exact in the case where nitrobenzene is
added*
3*
The uncertainty and fading of the end-point,
the Large deviation, and the Low resuLts prove
that unless the siLver chioride is rendered in­
active the method is unsatisfactory*
4.
The Low results indicate that heating the solu­
tion does not make the silver chloride inactive
and although the results are much better than
in the case where the silver chloride is left
untouched, the method is still unsatisfactory*
5*
nitrobenzene inhibits the darkening of the
silver chloride precipitate in light, thus
causing the color at the end-point to be noted
more easily.
The nitrobenzene is easily ob­
tained, only a small amount is required and the
modification takes only a few seconds.
6*
Being heavier than water nitrobenzene does not
form a layer on top of the solution*
Instead,
it collects the silver chloride particles, sinks
to the bottom and leaves a much clearer solution.
29
GRAVIMETRIC CHECK
A gravimetric check of chloride content was made on
the 200, 2000 and 5000 ppm. chloride solutions.
The
chloride was precipitated with an excess of silver ni­
trate and the precipitate was collected and weighed in a
Gooch crucible.
This check gave an accurate determination
of the chloride content, and the Mohr and Volhard methods
were checked against the gravimetric results.
The results, expressed in equivalent weight of
sodium chloride per liter and in ml. of silver nitrate
equivalent to 50 ml* of the chloride solutions, are as
follows:
TABLE IV
200 ppm. Na 01 Solution —
.2036 grams per liter or
1.74 ml.
of AgN03
2000 ppm. Na 01 Solution —
2.0158 grams per liter or 17.20 ml.
of AgN03
5000 ppm. Na 01 Solution —
4.9923 grams per liter or 42.64 ml.
of AgN03
30
SUMMARY AND CONCLUSIONS
A thorough study of the results derived from the
various modifications of the Mohr and Yolhard methods
points decisively to the facts that:
Cl)
Ih the Mohr
titration method addition of agar to the chloride solu­
tion gives better results and an easier and sharper end­
point.
(Although the results were not noticeably better
in the 200 and 20 ppm. solutions the end-point was sharper
and addition of agar is, for that reason, desirable*)
(2)
In the Yolhard method the addition of nitrobenzene
gave the best results in every case*
The end-point was
sharper and the average deviation was less in the cases
where nitrobenzene was used*
Therefore in comparing the
Mohr and Yblhard methods the agar modification of the
Mohr method and the nitrobenzene modification of the
Yblhard method will be used.
31
TABLET Y
Amt. of AgNOg
20 ppm* solution
Mohr method
.41
Yolhard method
.21
200 ppm* solution
Mohr method
1.76:
Yolhard method
1.76
Gravimetric method
1*74
2000 ppm* solution
Mohr method
17.21
Yolhard method
17.15
Gravimetric method
17.20
5000 ppm. solution
Mohr method
42*74
Yolhard method
42.66
Gravimetric method
42.64
10,000 ppm. solution
Mohr method
41.03
Yolhard method
41*24
33
These results show that in determining the chloride
content of solutions from 30 to 2000 ppm. chloride content
either method may be used.
However, the results show that
the Mohr method is the preferable method because by this
method duplication of results is more certain*
The average
deviation is less in each case due to a more exact end-point.
In concentrations of 5000 ppm. and higher the Volhard.
method gives much the better results.
Even with the addi­
tion of agar to the solution there still seems to be ad­
sorption of chloride in the silver chloride precipitate
in these higher concentrations and this seems to give
results that are too small.
As in the case of less con­
centrated solutions the Mohr method gives less average
deviation.
In concentrations of 2000 ppm. and below, the Volhard
method gave too low results.
This leaves the Mohr method
as the most accurate method of chloride determination in
the less concentrated solutions.
33
B I B L I O G R A P H Y
Caldwell and Moyer, Z » anal. Cham., 99, p. 258-69 (1934)
Gharpentier, P., Bull. Soc. Ing. Giv., France, 135, 325
Cl870)
Dreschsel, G., Z. anal. Ghem., 16, 351 (1877)
Hillebrand and Lundall, Applied Inorganic Analysis, Hew
York, p. 164 (1929)
Lottermoser and Lorenz, Kolloid Z ., 68, p. 201-3 (1934)
Meldrum and Forbes, “The Volumetric Determination of
Chloride,“ J.. of Chem. Education, Vol. 5, 1928
Mohr, G. F., Liebig*s Ann., 97, 335 (1856)
Rothmund and Bugstaller, Z. anorg. Ghem., 63, p. 333 (1909)
48, p. 79 (1909)
Treadwell-Hall, Analytical Ghemistry, New York, Vol. II,
p. 654 Cl932)
Volhard, J., J. pradt. Chem. (2), 9, 217 C1874)