SEPARATION OF
ALKAL] COMPOUNDS BY AMMONIA

THESIS FOR THE DEGREE OF M, 3.
Howard W. Eek
I934.

I ‘
0. 11‘]. 3‘0'.lz‘ ti

I, .6, #31,“ .
n‘._vllnl.r... t.

.
I 3......—

‘ .lr ‘11dl

1

 

 

..l

1.. ..
j. .1?

.«n. a .

.
a
l

.0?

n .
.\J..v

“. 1..

 

 

y_

 

 

SEPARATION OF ALKALI COMPOUNDS
BY AI-IT'JON IA

A Thesis Submitted to the Faculty
of

Michigan State College

In Partial Fulfillment of the
Requirements for the Degree
of

Master of Science

\u

By :1, t'-"
Howard W. Eck
1954,

Approved for the Department of Chemistry:

£7 ‘

0.0.0.0.0....OOOOCOOOOOOOOOOOOOOO

CD

(A

g:
:3)
I; l

-TABLE OF CONTENTS-

Page
Introduction ........................... 1
Apparatus and‘Materials ................ 5
Procedure --—-----------------------_--- 9
Results ............................... - 12
Discussion ............................. 19

Summary -------------------------------- 24

(1)

-INTRODUCTION—

It had been observed some years ago that when strong
ammonium hydroxide was added to a saturated solution of
potassium sulfate, a white precipitate was formed, but When
the strong ammonium hydroxide was added to sodium sulfate,
no precipitate was formed; and it was thought that possibly
a qualitative separation might be possible.

A search of the literature revealed that in several in-
stances this problem has been touched upon, but no one has
ever used ammonia or ammonium hydroxide as a means of sepa-
rating sodium from potassium compounds.

In a series of articles entitled "Liquid Ammonia as a
Solvent“ published in the Journal of Chemical Education by
W. Conrad Fernelius, Ohio State Univ., Warren G. Johnson,
Univ. of Chicago, Vol. V, 665-6, (June 1928), we find some
information bearing on this problem, under sub-title of
"Solubility Relationships of Liquid Ammonia.“ The authors
state that "Ammonia differs from water in its inability to
dissolve the sulfates, sulfites, the alkaline carbonates,
phosphates, and oxalates, the hydroxides of the alkaline
earth metals. The thiocyanates appear to be most soluble in
liquid ammonia.”

Similarly an article in Chemical Abstracts, 17-2998,
"Ammonia and Carbonates of the Potassium.Group" from the

Ber. 56 3., 1454-5, (1923) we find the following:

(2)

"When the anhydrous, neutral carbonate of K, Rb, or Cs
is added to an aqueous solution of NH3, of any concentration
at a temperature below 20°C, rise of temperature being pre-
vented by external cooling, at saturation two layers form,
the upper containing within 1% of all the NH3 and the lower
within 1% of all the carbonate. Water of neither layer is
at the disposal of the other layer. If to this system a lit-
tle Cu(AcO)2 is added, all the Cu, except traces, dissolves
in the lower layer. If the carbonate layer is not completely
saturated, addition of Cu salt causes the layers to coalesce,
but upon further addition of carbonate to saturation, sepap
ration of the layers again takes place, practically all the
Cu being in the lower layer. Capper can thus be removed from
NH:5 solution by shaking with solid KgCOg to saturation.”

One of the questions involved here is whether the pre-
cipitation of the alkali salts in NH3 solutions is just due
to a difference in solubility in ammonia and water solutions,
or does some change take place in chemical composition or
structure of the compound. we have a short reference on this
point found in C.A. 16:2076 which states that "Complex salts
are not formed in dilute solutions with salts of alkaline
and alkaline earth metals, because of the presence of H20.
Crystalline salts of some of these are known to exist in
NHS"

In studying the ammonia solutions it has been found
that ammonia, up to a concentration of .4 N is dissolved

as EH4OH and above that as HHS. The results which show the

(3)

solubility of substances in ammonia are based on a certain
percent of NH5 in solution. Some work has been done on "So-
lubility of NH4 salts in ammonia" as a method for the dis-
tinguishing and separation of mono-and polybasic acids in
C.A. 20:534. In an article by E. Weitz; Z. Elektrochem. 51,
546(1925), he states that "Salts of mono-basic acids in-
crease the aqueous solubility of NH5 (they diminish the
partial pressure of NH3) while those of poly-basic acids
decrease the solubility (by increasing the partial pressure
of the mas). The influence of NH3 on the aqueous solubility
of NH4 salts makes possible the distinguishing of poly-basic
acids and the separation of mono- and polybasic acids, and
can be made a criterion between mono- and polybasicity.'

Some work directly bearing on the problem was found in
C.A. 18:5256, namely, ”Separating Alkali Metal Salts from
Aqueous Ammonia“, by C. E. Dolbear; U.S. 1,505978 Aug. 12,
in which he found that ”Mixtures of alkali metal salts are
treated with aqueous NBS solution to effect fractional sepa-
ration; for example, NaCl and KCl may be dissolved frmm
mixed Searles Lake salts (containing also Na2804, Na2003,
and borax).' '

Something of interest along this line where ammonium
oxalate is used as a reagent for distinguishing between po-
tassium and sodium salts was discovered by L. Meyerfeld.

In C.A. 20:1189 and Z. Anal. Chem. 67, 150-1 (1925), he
stated that ”Na C O is much less soluble in water than is

2 2 4

K20204. The former salt precipitates when a salt solution

(4)

of (NH4)20204 is mixed with a saturated solution of NaCl."

A method in common use for the precipitation of Li by’
the use of NH4OH and NazHPO4 is found in "Qualitative Ana-
lysis for the Rare Elements' by Noyes and Bray. They state
the following: 'Filtrate containing Li salts, H0104 and
cgsson is boiled with HNOS. Then add 02H50H and make strong-
ly ammonical with NH4OH; and add NaQHPO4. A.precipitate of
Li3P04 results." You see in this separation of L15P04.we
depend on its relative insolubility in strong ammonia solu-
tion.

Of course as stated before this difference in solubili-
ty of sodium and potassium sulfate in water and aqueous am-
monia was noted, although no definite solubility curves were
run to determine Just what was the solubility in the varying
concentrations of ammonia. It might also be possible that
other sodium or potassium compounds would exhibit like be-
havior in these tests, or in fact we might eXpect a diffe-
rence in solubility of all the compounds in the alkaline
and alkaline earth families.

If no chemical change took place in the aqueous am-
monia solution, we might then presuppose that if a diffe-
rence in solubility was noted, and if this difference was
of sufficient magnitude, an effective qualitative separation

of the compounds might be possible.

-APPARATUS AND MATERIALS-

A tank of anhydrous ammonia gas was obtained and an
apparatus for precipitation of the salts was devised. The
ammonia was liberated from the tank in the form of a gas
and the solution became warm from the bumping of the 11-
quid. A large jar of water surrounding the precipitation
tube aided in preventing a large increase in temperature.

In the top of the tube was placed a two hole rubber
stOpper with a long glass tube going to within a short
distance of the bottom, and a short tube which let the
excess gas escape from the ignition tube, that was used
for these experiments. The rubber tubing from the ammonia
tank was fastened to the curved end of the long tube and
the ammonia gas was allowed to bubble through the solution,
while the excess gas escaped through the short tube at the
top. A long piece of rubber tubing was attached to this and
the excess gas was allowed to escape into the hood.

A.more economical way would be to pass the ammonia gas
from the tube through a large trap bottle and into a jar
containing water. This would give an ammonia solution which
might be used for other purposes. The trap bottle is neces-
sary because the pressure of the gas is not constant and
will suck the ammonia solution back into the tube.

The flow of NHS from the tank must be started to pres
vent the solution in the ignition tube from being pulled

into the rubber hose upon opening the valve in the ammonia

(6)

tank. This ammonia contains a small amount of water and
could be dried by means of metallic sodium, but such re-
finement is not necessary here because we only need a satu-
rated.ammonia solution. However, if the solubility of the
substance in liquid ammonia was wanted, this drying step
would be of utmost importance. As the substances which we
are most interested in, in this experiment, are insoluble
in liquid ammonia, this piece of work will deal entirely
with aqueous ammonia solutions of lower concentrations,
which do not call for Special pressure apparatus.

The apparatus pictured in this thesis consists of dewar
tubes, into Which a two hole stepper was placed, and the
ignition tubes inserted in these. The dewar tubes contained
solutions which tend to lower the temperature of the solu-
tion during the precipitation process with ammonia gas. In
some cases small quantities of liquid ammonia were placed
in the dewar tubes. The solution in the dewar tube might be
precooled water, liquid ammonia, solid carbon dioxide in
alcohol, liquid air, or any other solution which would lower
the temperature to the desired level. In liquid ammonia
where the solubilities are determined at the boiling point
of the liquid ammonia, it is advisable to have the outer
container filled with the liquid NH5,'so that a constant
temperature may be maintained. If temperaturesbelow the
boiling point of ammonia are desired, solid carbon dioxide
in alcohol would be very good, or if extremely low tem-

peratures that would solidify ammonia are desired, liquid

('7)

air can be used in the dewar tubes.

It will be remembered that the ammonia should be let

 

LIQUID gamma PRECIPITATION APPARATUS

in, in the form of a gas, with the first apparatus pre-
viously described. However, where the dewar tubes are used,
the liquid ammonia.may be added, and when the excess gas
volatilizes, the concentration of the ammonia can be de-
termined at the desired temperature, usually 25° C.
The:materials used in this work consisted mainly of
the purified salts of the alkaline and alkaline earth
elements. In the preliminary experiments all the available
sodium and potassium salts were used along with the
chlorides of the alkaline earths, magnesium, and ammonium.

The sulfates of the Li, NH4, and Mg; and the nitrates of

(8)

NH4, alkaline earths, and magnesium; and the carbonates of \
Li, and NH4 were used in addition to the sodium and potas-
sium salts. The sodium sulfate and the potassium sulfate
used in the solubility experiment were purified by recrys-
tallization from water three times. They were dried in the
centrifuge. Coarse crystals of from one to three mm. in

diameter were used.

-PROCEDURE-

Test tubes which held about 15 cc were filled to
within 3 cm. of the top. Rubber stoppers were then placed
in the ends of the tubes, and fastened over these a thin
wide rubber band or tubing to hold the stoppers in and
keep the water out. These were then wired to a revolving
wheel and placed in a thermostat for a few hours. The
rate of rotation was about 40 revolutions per minute. The
process was stOpped and the test tubes were brought to the
top, stoppers removed, and the contents removed in two
portions by filtering pipettes.

These pipettes were provided with tips connected by
means of rubber tubes. Cotton was placed in these tips and
the tips removed after the pipettes were filled. The con-
tents were weighed and then evaporated in small porcelain
dishes, covered with watch glasses, on a steam bath. The
residue was cautiously heated over a bunsen flame, very low
at first and increasing to full intensity in one hour. The
residue was kept covered during heating to constant weight,
which was usually attained very shortly. ‘

In order to determine the strength of the N35 solution
a pipette was inserted into it, and the NHS passed directly
into a standard acid solution. To determine the amount of
salt dissolved in the NHS solution, one must either obtain
a separate portion or evaporate the neutralized solution,

and then volatalize the ammonium salts.

(10)

A 10 cc pipette was constructed by putting two 5 cc
pipettes together and then calibrating it so that it would
deliver two 5 cc portions. The first portion was evaporated
to dryness and the weight of dissolved salt in 5 cc of
solution obtained, and from this the solubility was de-
termined in grams per 1000 cc. The other five cc was de-
livered into a flask containing a standard acid, and the
excess acid titrated, and from this the concentration of
the ammonia calculated. The solubility of the salt was
then expressed in grams per 100 cc, and the concentration
as a certain percent NHB solution.

Strong ammonium hydroxide solution, approximately
28% HHS, was placed in contact with solid K2304 and
shaken for a period to saturate the solution. Ten cc
of this solution was then drawn off and evaporated in a
weighed crucible. The residue represented the solubility
of K2304 in 28% HHS.

After solubility curves were determined on K2304
and Na2804 in varying concentrations of NHS, it was then
necessary to determine how the other substances behaved
in ammonia solutions. What was wanted in these qualita-
tive tests was a comparison between the solubility of
the salt in water and its solubility in ammonia solu-
tions. Lower concentrations did not provide sufficient
precipitate for comparison with the K2804 precipitate,
so that only strong NH3 solutions and YH3 gas were used

as precipitating agents. The former was added to a

 

(ll)

saturated solution of the salt at 25° C, and as stated
before, a precipitate was obtained with K2304. New many
of the salts did not precipitate as the resulting con-
centration of the ammonia was not sufficient to cause

5 gas to the
saturated solution of their salt, a copious precipitate

precipitation; but upon the addition of NH

was obtained. Of course the concentration of the re-
sulting NH3 solution depended on the salt being used for
the experiment, but was approximately 32% NHS at 250 C.

Ten cc of the saturated solution of the salt was used
in every case, and the temperature maintained at 25° C, so
that any precipitate formed was due to a change in solu-
bility of the salt, and not due to a change in solubility
with temperature. Since the volume of the solution in-
creased upon the addition of NH5, the amount of precipi-
tate was used as a basis of comparison.

The steps in the precipitation process were then as
follows: first, preparation of the saturated solution of
the salt in water, noting its solubility at 25° C; second,
treatment of 10 cc of this solution with strong ammonium

hydroxide, 28% NH , observing the amount of precipitation;

5
third, treatment of another ten cc with NHS gas, noting
the amount of the precipitate; fourth, comparison of this

precipitate with the one produced with K2304.

(12)

-RESULTS-

We have noted previously a difference in the behavior
of K2304, when treated with strong ammonium hydroxide, in

contrast to the behavior of Na $04, when similarly treated.

2
Now from this we might presuppose that K2804 was insoluble
-in strong NH3 solutions, and that Na2304 was soluble in
these solutions. This wOuld be then a fine scheme for the
separation of potassium from sodium compounds, by first
converting them to sulfates, and then precipitating with
K85; and only the K2304 would precipitate in the strong
ammonium hydroxide. However by passing HHS gas into the
saturated solutions of the salts, we also got a precipitate
with Na2304. In other words increasing the concentration
of the NH3 decreased the solubility of the Nazso4 to the
extent that it also precipitated like the K2304, and the
amount in each case was approximately the same. But we
found that K2804 was much less soluble in water than Na2804,
and considering the same amount of precipitate in each
case with NH3 gas, the relative solubility showed the
K2304 to be also much more insoluble in NH than the
Na2304,

As was expected, we found that practically all of the
salts eXhibited a difference in solubility in water and

in strong NH3 solutions. Also in some cases the solubility

of the salt was greater in saturated ammonia solutions

(13)

than in water, but in most cases the addition of the NH3
gas decreased the solubility of the salt. No general rule
could be stated but in most cases potassium salts seemed
to be more insoluble in saturated NH3 solutions than the
correSponding sodium salts. However, as will be seen by
looking at the table, there were many notable exceptions
where the Opposite was true. Specific cases where the
potassium salt was more insoluble in water than the cor-
reSponding sodium salt may be seen in the table, but the
same salts in saturated NHS solutions exhibited a reversal
in that the sodium salt was more insoluble in this solu-
tion than the corresponding potassium salt.

Since qualitative tests depend on the bulk of the
precipitate formed, the number 20 found in the table under
solubility in 32%‘NH3, would mean twice as much precipi-
tate formed in this case as with the K2304, which is
represented by 10 in the table. A precipitate with the
rating of 2 would be only one fifth as much as the K2304,
and so on through the table.

Solubilities in water are expressed in grams per 100
cc. In some cases only a symbol is used because information
on its solubility is unavailable, since presumably its
solubility has not been determined as yet. The symbol 5
represents soluble, 81. s. slightly soluble, and v.s. very
soluble. This solubility for each salt is the first num-
ber in each table.

The second term in the table eXpresses the solubility

(14)

of the salt when 28%'NH is added to it. The symbol L

5
represents light precipitation, M medium precipitation,
and H heavy precipitation. Where no symbol was used in
this space no precipitate was obtained, and this was
true in most cases.

The third number represents the amount of precipitate

solution, taking the K

obtained in the 52% NH 804 pre-

3 2

cipitate equal to 10.

Table of Solubilities and Relative Precipitations pf Salts

 

 

 

 

 

 

 

 

 

K Na Li NH4 Ca Ba Sr Hg
92 4
F
O 2
54 56 79 57 75 56 55 55
Cl M M
40 2 O O 55 o O 85
65 48
Br
20 0
144 179
I
l
7 101
010
5 s o
7 55
BrO
5 lo
8 9
IO:5 M H
415 85
ll 25 25 75 50
SO M L H

4
lO 10 so so 45

 

 

 

 

 

 

 

 

 

 

 

 

 

K Na Li III—I4 Ca Ba lfg
51 50
HSO H
4 90 50
27
so
3 so
8. 81.8.
HSO:5 M
60 30
s. 16
s H
60 l
52 as 192 45 9 65
N0 L H
5 15 o o l o 95
75 46
NO,
2 50 s
193 11
PO H
4 #0 95
v.s. 7
31304 H M
95 85
25 85
H o M H
2P 4 60 100
55 22 1.5 100
co m
3 4. 45 10 o
25 10
H603
s 45
S. B.
3105 H
o 100

 

(16)

 

 

 

 

 

 

 

 

 

 

 

 

K Na Li EH Ca Ba Sr 158‘
62 4'7
CrO4 L
80 O
12 64
Cr 0 M
2 7 6O 95
72 3.
CN
25 5O
69 v.s.
CNS
0 O
) 55
K F CN
3 e( 5 10
28
K4Fe(CN)6 H
55 4
C O H M
2 4 4o 15
2.2 1.7
HO O
2 4 10 2
. 1 1.8 s.
50
C4H406 80
26
NaKC4H406 100
.5

(1'7)

 

 

 

 

 

 

K Na Li NH4 Ca Ba Sr Mg
256 124
H o
02 3 2 o o
167 91
c H o
6 5 7 70 70
8.
AsO
v.s.
HAso
3 95
3.
302 5
6
Hno4 H
50

 

In the precipitation experiments with strong ammonium
hydroxide and NH3 gas, we discovered how the salts behaved
in these solutions in comparison with aqueous solutions. The
relative amount of precipitation was the criteria of separa»
tion in each pair of the sodium and potassium salts. After
observing the amount of precipitate formed for qualitative
tests, it was also of interest to determine just how much
of the salt was still left in solution after complete pre-
cipitation.

Since the sulfates are very common and easily produced
in solution, it was thought advisable to run solubility
curves on the sodium and potassium sulfates as a basis for

comparison. We wanted to determine the solubility of these

(18)

salts in different concentrations of ammonia, and plot on
a graph the solubility of the salt in grams per 100 cc of
solution against the percent ammonia in the solution.

The following table shows the results of the solu-
bility experiments, and the accompanying graph illustrates
clearly the decreasing solubility of the salts with in-

creasing ammonia concentrations.

Table of Solubilities in Varying Ammonia Concentrations

 

% HHS Solubility of K2804 Solubility of Na2304
o 11.660g. 25.662g.
5.6 5.660” 20.020~
8.4 5.875" 17.225"
11.2 2.654n 15.547"
14 1.577n 11.407"
16.8 .947" 6.987“
19.6 .601“ 4.555"
22.4 .589" 5.006"
25.2 .240" 1.885"
28 .145" 1.227“

32.5

.025”

.886"

 

 

 

‘ MICHIGAN STATE COLLEGE ° -

 

DEPARTMENT OF MATH ENATIC.

(19)

-DISCUSSION-

Perhaps the most interesting aspects of this work
relate to the distinguishing features which tend to show the
behavior of the different salts during the ammonia precipi-
tation process. If we look at the Table of Solubilities and
Relative Precipitations of the Salts, we observe many inter-
esting features among the salts. The important points of
separation will now be pointed out.

NazF is much more insoluble in water than K2F2, and

2
we observe that the same condition holds in 52%’NH3,

namely, the Na F, precipitates while the K

F. remains in
2 2 3

2
solution.

The chlorides offer an interesting example of the be-
havior with NHS‘ The KCl precipitates freely, the NaCl
very slightly, LiCl and NH4Cl remain in solution, CaCl2
precipitates freely, while BaCl2 and Sr012 show no pre-
cipitate; an HgClg almost completely precipitates, which
undoubtedly is due to the formation of magnesium hydroxide.
From this we can see that large amounts of KCl can be pre-
cipitated in solutions containing NaCl, LiCl, or NH4Cl,
simply by passing NH:5 gas into the saturated solution of
these salts. Also CaCl2 can be separated from Ba012 and
SrCl2 in a like manner, the CaClZ even precipitating in
a saturated solution treated with 28% NHS.

KBr is more soluble in water than NaBr, yet when HHS

gas is added, solubility is reversed, and KBr precipi-

(20)

tates. This would tend to show that the potassium salts

seem to be more insoluble in the 52% HH5 solution than

the correSponding sodium salts.

However, the iodates tend to show quite the Opposite
behavior. The NaIO3 is a little more soluble in water than
K103, but in 52%NI—I:5 the NaIOsprecipitates almost complete-

ly, while the K10 shows only slightly above normal precipi-

5
tation.

Since tha Ca, Ba, Sr, sulfates are almost completely
insoluble in water, we neglected them because they are
also insoluble in NH5' In the 52:?25'NH3 we find precipitates
in all the remaining cases. The Li2804 is more insoluble

than the M2304 in 3275 NH and they have the same solu-

5
bility in water. The (NH4)2804 seems to be the only one
of the ammonium salts that precipitated.

The bisulfates are interesting when treated with 28%
KHz, since they have approximately the same solubility in
water. The KHSO4 shows a heavy precipitate, while the
Na‘HSO4 shows no precipitate.

We expected that all the nitrates would be soluble in
52% HH5’ but even here the KNO3 is difficultly soluble to
the extent that it predipitates. Again the Mg(OH)2 is
formed as an insoluble precipitate from the Mg(N03)2.

The normal phoSphates are very interesting. The K PO

5 4

is very much more soluble in water than Na3P04. K3PO4 is

a deliquescent substance, and when treated with 28% NH5'

or when NH3 gas is passed into the saturated solution of

(21)

the salt, a two phase system is produced, which on long
standing does not break down, as long as the HHS content
is maintained. The upper phase. contains most of the ammo-
nia, while the lower phase contains most of the potassium
phOSphate. Now if we take NaéPO4 which isn't so soluble

in water as the K PO and treat the saturated solution with

5 4'
28% .53, or with NH5 gas, we get almost complete precipi-

tation.
LiQCOS is sparingly soluble in water and is very in-
soluble in the 52% III-{3.
Na SiO

2 3 5

NHS solution, while K28133 shows no tendency to precipitate.

precipitates completely with NH or with 28%

K20r04 is more soluble in water than NagCr04, but K20r04
precipitates heavily when NH3 is added, while NagCrO4 shows
no evidence of precipitation. When dichromates are used,
both the Na and K salts precipitated from their saturated
solutions.

Both of the cyanides precipitate while the thio-cyanates
appear to be the most soluble substances in NH3. Although
the K4Fe(CN)6 is less soluble in water than KSFe(CN)6, the
former precipitates much more heavily with NH3 than the
later, even giving a heavy precipitate of K4Fe(CN)6 When
treated with 28% NH3 solution.

The acetates offer an example where neither the Na
or K salts precipitate, but unlike the thio-cyanates, do

not increase their solubility in HHS solution.

The precipitate of KMnO4 which is formed in the pre-

cipitation process with NH changes on standing to HnOZ.

5

In every case we might take each salt.and run its
solubility in concentrations of ammonia from 0% to 100%,
and after getting all this data, plot curves which show
the behavior of the salts when NHS is added. In most
cases, we have pointed out the decreasing solubility with
increasing ammonia content, and as the sulfates are used
in this work, reference to the graph illustrates this
point.

K2304 is an anhydrous salt and the solubility curve
is smooth, decreasing with increases in HHS content.

Na2804 in crystalline form exists as a decahydrate
and as a heptahydrate, and also there is an anhydrous
form. Now the undissolved salt is always in equilibrium
with the dissolved salt, and if anhydrous Na2304 is the
solid phase, then Na2304 in solution exists in equili-
brium with this. If we start with Na2304(H20)10 and
add NH3 removing the amount of water available to the salt,
we might expect to pass through Na2304(H20)7, and finally

in strong ammonia solutions, only the Na SO4 anhydrous

2
would exist. Now in water, Na2804(H20)lo is soluble to the
extent of about 26 grams per 100 cc, while Na2SO4(H20)7

is about 44 grams per 100 cc soluble, while anhydrous
Na2304 is even more soluble. However, when NHS is added,
the solubility of the salt decreases, and if the curve's
deflection is extrapolated to the zero line, the appro-

ximate location of the Na2SO4(H20) and anhydrous NaZSO

7 4

could be located. The curve of the Na2304 thus differs

from K2804 because of the possible hydrates formed.

-SUMMARY—

This work is unique in that it Opens an entirely
new field for Speculation and investigation in inorganic
chemistry. Here we have a precipitating agent, namely
ammonia gas, which does not depend on common ion action,
or formation of a new compound such as complex ammonia
salts. The observed precipitation is due solely to a
change in solubility from water to strong aqueous ammonia
solutions.

Of course,when the solubility of the salt decreases
in the solution, precipitation takes place. If the solubi-
lity remains the same, or is actually increased upon the
addition of ammonia, no precipitation of the salt is ob-
served. Thiocyanates seemed to increase their solubility
in aqueous ammonia, acetates remained the same, while
most of the other salts of the alkaline metals showed a
tendency to decrease in solubility and precipitate. Each
salt seemed to exhibit a definite solubility in aqueous
ammonia, usually different from that in water, and this
could not be predicted. Usually the potassium salt was
more insoluble than the sodium salt, yet in some cases,
the opposite was true. This followed generally the
solubility order in water, the potassium salt being
slightly more insoluble than the correSponding sodium salt.

Since the ammonia used does not enter into chemical

union with the salts, it may be used over in industrial

(25)

applications of this method of precipitation. For example,
saturated salt brines containing chlorides, bromides,
nitrates, or chlorates may be treated with ammonia gas,
and the resulting precipitates, largely potassium salts,
separated from the bulk of the brine. The ammonia gas
could be volatilized by heating the brine and used again
to treat a new sample. Precipitates rich in potassium
salts could then be further treated. Any magnesium salts
would in this process produce the insoluble hydroxide,
which could be filtered off from the water soluble po-
tassium salts.

Analytical schemes, where large samples are available,
in which only a flame test for potassium or sodium salts
is used, could further improve the tests by using some of
the things presented in this work. The residue in the
form of sulfates, containing sodium and potassium salts,
after the volatilization of ammonium salts, could be dis-
solved in a small amount of water, such as 5 cc, to make a
saturated solution, and then treated with strong ammonium
hydroxide. The presence of a white precipitate would indi-
cate the presence of potassium in the final residue.

It is unnecesary to elucidate further on this scheme
of precipitation, since numerous applications would pre-

sent themselves in pursuance of this research.

 

 

31293 024