Organic amalgams: substances with metallic properties composed in part of non-metallic elements ..

til 301

LIBRARY

UNIVERSITY OF CALIFORNIA,

RECEIVED BY EXCHANGE

Class

The University of Chicago

Founded by JOHN D. ROCKEFELLER

Organic Amalgams: Substances with Me

tallic Properties Composed in Part

of Non- Metallic Elements

A DISSERTATION

SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE

SCHOOL OF SCIENCE IN CANDIDACY FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEM

WILLIAM CABLER MOORE

EASTON, PA.:

PRESS OF THE ESCHENBACH PRINTING CO.
1911

The University of Chicago

Founded by JOHN D. ROCKEFBLLER

Organic Amalgams: Substances with Me

tallic Properties Composed in Part

of Non-Metallic Elements

A DISSERTATION

SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE

SCHOOL OF SCIENCE IN CANDIDACY FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY

WILLIAM CABLER MOORE

EASTON, PA.:

PRESS OF THE ESCHENBACH PRINTING CO.
1911

Organic Amalgams: Substances with Metallic

Properties Composed in Part of

Non-Metallic Elements.

A. Introduction.

A great deal of evidence has been accumulated showing that complex
radicals forming the positive ions of salts are metallic in nature, but
hitherto the nearest approach to the actual isolation of such substances
in the free state has been found in the single case of ammonium amalgam.
At the suggestion of Dr. Herbert N. McCoy, to whom the author wishes
to express his thanks for the many courtesies shown him during the prog-
ress of the work, an investigation having for its purpose the preparation of
amalgams of positive radicals was taken up, with the hope that substances
with more metallic properties than ammonium amalgam could be isolated.

The following detailed description of the experimental work is an ac-
count of the results achieved by the author, and likewise constitutes a
portion of a joint paper by Dr. McCoy and himself, published in the
Journal of the American Chemical Society for March, 19 n.1

B. Preparation of Tetramethyl Ammonium Amalgam.

The first attempts at the preparation of tetramethyl ammonium amal-
gam were made by the action at o° of 8 per cent, sodium amalgam on a
25 per cent, aqueous solution of tetramethyl ammonium chloride. Hy-
drogen was given off, but no new product resulted. The electrolysis of
aqueous solutions of the chloride under various conditions of temperature
and concentration gave always colloidal mercury at the mercury cathode,
but no amalgam could be isolated. It was thought that an amalgam
might be formed by the action of the current and be rapidly decom-
posed by the water. Accordingly, absolute alcohol was substituted
for water as solvent and proved wholly satisfactory.

The first successful preparation of tetramethyl ammonium amalgam
was accomplished by the electrolysis of a saturated solution of tetra-
methyl ammonium chloride in absolute alcohol, using a platinum anode
and a mercury cathode. The temperature was — 10°. By means of a
potential-reducing device, the direct current of the lighting circuit (no
volts) was cut down to 18 volts, and the current was allowed to run at
this pressure for thirty minutes. As prepared in this way, the amalgam

1 /. Am. Chem. Soc., 33, 273 (1911).

236757

is a stiff mixture of shining crystals of the amalgam and the excess of
mercury; it is but very slightly inflated,1 silvery white in color and de-
cidedly metallic in appearance. In contact with alcohol it is moderately
stable, but reacts violently with water.

Since this first preparation was made, the amalgam has been prepared
by this method many times and many of its properties studied. How-
ever, since in using a platinum anode chlorine is evolved, the method has
been improved by the substitution of an anode of silver-plated platinum
gauze.

In the most convenient form of apparatus for the production of the
amalgam, a small Gooch funnel is used as the electrolytic chamber. This
is supported by passing the stem through a one-hole rubber stopper
closing the tubulure of an inverted four-liter bell jar, which serves as an
ice chamber. The whole apparatus is supported at an appropriate height,
by means of a ring and a universal clamp, on an iron stand. A piece of
rubber tubing fastened to the stem of the Gooch funnel, and closed by a
pinchcock, affords a convenient outlet for the amalgam after it is formed.
The anode consists of a piece of silver-plated platinum gauze rolled into
cylindrical form. This gauze is welded to a stout platinum wire which
passes between the edge of the funnel and the rubber stopper closing the
funnel. The negative wire, insulated by a glass capillary tube, enters
the funnel through one leg of a T tube. The upper end of this T tube,
where the wire enters, is closed by means of universal wax or a glass cap
and a piece of rubber tubing, while a calcium chloride tube is attached
to the other leg of the T to exclude moisture. A thistle tube, reaching
almost to the mercury cathode, is inserted into the second hole in the
rubber stopper, and serves for the introduction of mercury or of the
electrolyte.

Fitted up in this manner, the procedure for making a "run" with the
apparatus is as follows: The mercury to serve as a cathode is run into
the electrolysis chamber through the thistle tube; the solution of the
electrolyte is then introduced, the bell jar having previously been filled
nearly to the top of the Gooch funnel with crushed ice or a freezing mix-
ture. The current is then turned on, the potential and strength being
regulated so that about 0.25 ampere passes through the cell, under ordinary
circumstances. After running for 10—15 minutes, the current is inter-

1 When properly prepared, it shows at temperatures below 4- 10° absolutely
no tendency to become inflated, in which respect it differs markedly from ammonium
amalgam.

rupted; the excess of mercury run off through the delivery tube at the
bottom of the apparatus, into a dry dish; the amalgam is then run off
into a second dry dish and used as desired, while the excess of mercury
is poured, through the thistle tube, back into the electrolytic cell. In
this way a number of successive runs can be made with but little inter-
ruption.

C. Properties of Tetramethyl Ammonium Amalgam.

Physical Properties. — As prepared above, at temperatures of from
— 10 to — 5° the amalgam is a pasty, crystallin mass about the color of
metallic zinc. While at o° or at temperatures slightly above o° the
crystallin structure is not so pronounced, yet on several occasions almost
solid lumps have been observed in the amalgam formed at o°. It is
somewhat lighter than mercury, as it always floats on the latter. There
is evidence that more than one active phase exists in the amalgam. The
surface tension of the amalgam is less than that of mercury. There is
little tendency, from — 10 to o°, toward inflation of the amalgam, but
when the temperature of a sample, washed with absolute alcohol and
partially dried with filter paper, was allowed to rise slowly, the volume
increased and at 25° was about twice the volume at o°.

Stability of the Amalgam. — When the temperature of the clean amalgam
is kept near o° it is quite stable, but on allowing the temperature to rise
it decomposes very rapidly. Trimethylamine is one of the products
of this decomposition, as is shown by the powerful fish-like odor. The
other products have not been fully investigated; it is hoped to make a
further study of this point.

General Chemical Properties. — If the amalgam is run from the elec-
trolytic cell directly into a clean, dry flask, it becomes covered with a
gray-white crust, probably tetramethyl ammonium hydroxide, as it
reacts with water to form an intensely alkaline solution.

The Reaction of Water on the Amalgam. — When allowed to come in
contact with water a violent action takes place; hydrogen is formed and
the whole of the mixture becomes ink-black, rapidly turning gray, and a
gray, colloidal solution is the final result. Le Blanc1 probably got the
same substance by the electrolysis of aqueous solutions of tetramethyl
ammonium chloride, using a mercury cathode — a result we have also
obtained under similar conditions. We studied the phenomenon further
as follows.

The amalgam was formed as usual, in the apparatus described above.
1 Le Blanc, loc. cit.

After the current had been allowed to run for some time, the excess of
mercury was drawn off into a clean, dry dish. The amalgam was then
drawn off into a second small porcelain dish, quickly washed several
times with absolute alcohol, dried with filter paper, and the dish held
in the hand, against the inside edge of a 600 cc. beaker. A fine stream
of water was now forcibly directed against the upper surface of the amal-
gam, near the lip of the dish, in such a way that the colloidal black de-
posit was swept into the beaker without carrying along globules of mer-
cury. The washing with water was continued until little or no further
action between the amalgam and the water was noticed. When the
main portion of the "black deposit" had been swept into the beaker,
in nearly every case a black, nearly solid residue, small in amount, was
noticed floating on the globule of mercury left. This black substance
acted less vigorously with water than the pasty amalgam, but still gave
a "colloidal black deposit" with water. An explanation of this fact
will be taken up later. The above outlined process was repeated a number
of times, until sufficient material had been accumulated ; the black residue
was then filtered off through a weighed Gooch crucible, dried at 108—
no0 and weighed; the filtrate was reserved for further experimentation.
This weighed residue was then dissolved in hot nitric acid, repeatedly
evaporated with concentrated hydrochloric acid until nitric acid was
practically completely eliminated, then diluted, filtered free from asbestos
fibers, and the mercury precipitated by hydrogen sulfide. This pre-
cipitate was washed, dried, washed several times with carbon disulfide,
dried at 108-110° and weighed. In three analyses in which from 0.4 to
0.8 g. of substance was taken the percentages of mercury found were
98 .06, 97.93, 96.98.

These low but consistent values of the mercury content of the black
deposit mean that the substance is in all probability the pure element,
the low values being due either to (i) water absorbed to a slight degree,
or (2) to a little occluded hydrogen. Less plausible would be the as-
sumption that (i) the deposit is a mixture of mercury and mercurous
oxide, or (2) that it is a compound containing the tetramethyl ammonium
radical. The deposit after drying generally aggregated into actual
globules, and the portion which did not do so became silvery white in
color; this disposes of the first assumption, and there is no other evidence
in favor of the second. The mercury itself, used in the preparation of
the particular amalgams from which the black deposit was obtained for
analysis, was, by analysis, 99.86 per cent. pure. The nature of the im-
purity was not determined.

The filtrates from each o the above experiments and a number of
others were titrated with standard hydrochloric acid; the results were
calculated to N(CH3)4 and the ratio, N(CH3)4 to Hg, was determined
from these results, and from the weight of the colloidal mercury. In the
table below, in the experiments in which the values "calculated weight,
N(CH8)4C1" and "actual weight, N(CH3)4CP" are given, the solutions
after titration were actually evaporated to dryness, and the residues
weighed. These residues were then combined, purified and analyzed
as shown below.

TABLE I.

Expt. Cc. 98% N(CFr3)4 Wt. Calc. wt.

No. o.i N HC1. Wt. colloidal Hg. calc. A. res. N(CH3)4C1.

270 4.00 0.1329 0.0297 1.66 ... ...

271 2.50 0.0515 0.0186 1.02

272 9.50 0.2303 0.0705 I. 21

273 5.82 0.4496 0.0431 3.87

274 12.72 0.2797 . 0.0943 i -io
277 7-12 0.3037 0.0527 2.13

279 18.18 0.3629 0.1348 0.99 0.230 0.199

282 n. ii 0.7946 0.0823 3-57

283 15-43 I-3451 0.1143 4.35 0.180 86 K 10
285 26.17 0.3713 0.1938 0.71 0.294 °61 68
288 12.28 0.4161 0.0910 1.69 0.126 ' .o 9

291 35-5° 0.4607 0.2630 0.60 0.395 °9 'ozz

292 23.71 0.4983 0.1757 1-05 0.252 o£ "

In the column headed "A" is given the number of atoms of colloidal
mercury formed for each tetramethyl ammonium ion set free. It is
seen that this ratio is not constant. Four of the thirteen values, how-
ever, are close to unity, which may indicate a possible compound of the
formula HgN(CH3)4. Since the colloid analyzed to 98 per cent, mercury,
this per cent, of the actual weight of the colloid is used in the calculations.

In several of the experiments in which the colloidal mercury was not
used for analysis, it was weighed on dried filter papers. In some cases
the deposit was coagulated by the addition of sodium chloride and in
others by the addition of concentrated hydrochloric acid. The latter,
however, was abandoned as a coagulant since it dissolved some of the
deposit, though not much. In experiments 270, 271, 272, 273, phenol-
phthalein was the indicator used, while in the rest of the experiments
methyl orange was used.

The values in column "A" are to be taken as approximate only, as
it was always difficult to remove every trace of the colloid from the

8

small dish to the beaker by means of wash water. Again, there was al-
ways a slight decomposition of the amalgam not due to the action of
water.

As stated above, the residues of supposed tetramethyl ammonium
chloride left on evaporating to dryness the solutions in experiments 279,
283, 285, 288, 291 and 292 were combined and analyzed. This analysis
was made as follows: The combined residues were dissolved in water,
the solution evaporated to dryness, the resulting mass taken up with
absolute alcohol and precipitated with ether. From the mother liquor
of this ethereal solution, a second crop of crystals was obtained and
this was recrystallized from absolute ethyl alcohol. These samples were
dried, and weighed portions dissolved in water and titrated with deci-
normal silver nitrate, using potassium chromate as indicator. 0.1290
g. of salt of first crystallization required 11.82 cc. of o.i N AgNO3; 0.1044
g. of salt of second crystallization required 9.55 cc. o.i N AgNO3. Per
cent. Cl calculated for N(CH3)4C1, 32.36. Found: I, 32.27; II, 32.47.
The results of these two analyses, with the close agreement between the
actual and calculated weights of N(CH3)4C1, show conclusively that the
amalgam contains the tetramethyl ammonium group.

Two analyses of the gas evolved when the amalgam reacts with water
were made as follows: A quantity of the amalgam was run out into a
small dish and quickly washed with absolute alcohol and dried as much
as possible by means of filter paper. It was then decomposed by water
and the gas evolved collected. This process was repeated until an amount
of gas sufficient for analysis was obtained. The gas was then trans-
ferred to a eudiometer filled with mercury, measured, mixed with air,
measured, exploded, and the contraction noted. Concentrated sodium
hydroxide was now admitted to the tube, but no contraction resulted,
hence no carbon dioxide was formed on explosion. In the first analysis
i i.oo cc. of the unknown gas were used, and the contraction of the mixture
on explosion was 15.30 cc., equivalent to 92.73 per cent, of hydrogen;
while in the second case 6.50 cc. was the contraction when 4.6 cc. of the
unknown gas were used, indicating 94.13 per cent, of hydrogen present.

These experiments should be regarded as qualitative only; no correc-
tions were made for the vapor pressure of water; the experiments were
made simply to demonstrate that hydrogen is the gas evolved when the
amalgam is treated with water, and that no gas containing carbon was
produced.

Reaction of the Amalgam with Aqueous Solutions of Various Salts. — A

cold concentrated solution of ammonium chloride was treated with tetra-
methyl ammonium amalgam. Ammonium amalgam, as shown by the
inflation peculiar to this substance, was formed. There were absolutely
no indications of colloidal mercury. Part of the same tetramethyl ammo-
nium amalgam reacted in the usual way with water. A saturated solu-
tion of potassium chloride was treated with tetramethyl ammonium
amalgam at room temperature. There was only a trace of colloidal
mercury formed. After the mixture had stood for three minutes, the
residual mercury was washed quickly six or seven times with water and
then allowed to stand in contact with water. There was absolutely no
indication of the presence of tetramethyl ammonium amalgam, but
there was a slow reaction between the mercury and the water, gas being
evolved. That potassium amalgam had been formed was proved by
acidifying the aqueous solution with hydrochloric acid, adding a few drops
of chloroplatinic acid and evaporating to dryness; the residue, when
treated with water, remained for the most part undissolved; the yellow
crystals were undoubtedly potassium chloroplatinate. A solution of
sodium chloride, approximately three times normal strength, was treated
with tetramethyl ammonium amalgam. No colloidal mercury was
formed, but a quiet action ensued between the solution and the amalgam.
This was allowed to continue for four minutes, then the mercury was
quickly washed twelve times with cold water and covered with water.
Gas was evolved for more than two hours. The resulting solution was
alkaline to litmus — conclusive proof that sodium had replaced the tetra-
methyl ammonium group in the amalgam. Concentrated solutions of
cesium and rubidium chlorides in water an 1 thrice normal potassium
chloride were each treated with parts of the same lot of tetramethyl am-
monium amalgam. With the cesium chloride, violent action occurred
with the formation of colloidal mercury, but the action was less violent
than when pure water alone is used. With the rubidium chloride, only a
little colloidal mercury was formed, and with the potassium chloride, none
at all. When the experiment was repeated, using normal solutions of each
salt in water, the results were in the same order, the only difference being
that with potassium chloride a little colloidal mercury was formed, less,
however, than with rubidium chloride.

An aqueous solution of copper sulfate of unknown concentration was
treated with tetramethyl ammonium amalgam. The action was fairly
violent; only a little colloidal mercury was formed. There was efferves-
cence, and after standing fifteen minutes, the mercury was covered with

10

bubbles of gas. At this stage, very thin, copper-red crystals were noticed
floating on the surface of the supernatant liquid. The mercury itself
had a coppery color. It was washed seven or eight times with water,
and showed every indication of the presence of copper amalgam. Part
of it was dissolved in nitric acid, the resulting solution evaporated and
ammonium hydroxide added to the greenish residue. The characteristic
blue color of the cupri-ammonium ion was seen while some of the original
mercury from which the amalgam had been formed gave absolutely no
indications of the presence of copper on similar treatment.

Mercuric chloride in aqueous solutions was also treated with tetra-
methyl ammonium amalgam; the reduction to the mercurous state was
more rapid than when pure mercury acts on mercuric chloride. Colloidal
mercury was formed. The flocculent gray precipitate was filtered off
and tested with ammonium hydroxide; the characteristic black color
which this reagent affords with mercurous chloride was produced.

Normal aqueous hydrochloric acid acts vigorously on t.etramethyl
ammonium amalgam, only a little colloid, however, being formed. We
Jhave also shown1 that the amalgam precipitates copper and zinc from
^alcoholic solutions of their nitrates. All these experiments go to show
the complete analogy between the tetramethyl ammonium radical and
an Ordinary alkali metal, and will be referred to again after the con-
sideration of further experimental work.

The Action of Alcohol on the Amalgam. — One of the first properties of
the amalgam to be studied was the reaction with alcohol. The speed
of this reaction was determined, approximately, at o°, as follows: The
amalgam was made in the usual way; it was washed four times with
absolute alcohol and then immediately poured into absolute alcohol in a
conductivity cell at o°. The resistance of the solution was determined
at definit intervals and from the values for this resistance at the various
times the speed of the reaction was calculated, applying the formula for
a unimolecular reaction, k = i/t loge a /(a — A). The mixture was, of
course, stirred to insure uniformity of concentration as far as possible.
If W is the resistance at the time t, and Wa is the final resistance, it is
easy to show that k = i/t \oge W(W — Wa), since the initial resistance,
that of pure alcohol, is very high. Using this formula, /, , the rate of for-
mation of the tetramethyl ammonium ion when the amalgam is in contact
with absolute ethyl alcohol is calculated from the results below:
1 McCoy and Moore, Science, 30, 315 (Sept. 3, 1909).

II

TABLE II.

Experiment No 33. Experiment No. 35.

Time, t

Resistance

Time, t

Resistance

(minutes).

(ohms) W.

k.

(minute.6).

(ohms) W.

O

6l,6oo

0

15,400

u-5

840

0.0238

3

399

o.

21-5

308

0.0492

6

300

o.i

3i-5

264

o . 0456

9

250

O.i

41-5

234

0.0472

12

215

O.i

5i-5

205

0.0752

16

194

O.(

61.5

201

19

i?3

O.I

7i-5

2OI

24

150

0.(

81.5

201

29

139

O.I

34

129

O.I

40

117

O.I

83

"5

.

.

100

US

.0609

D. The Electrode Potential of the Electrode: Tetramethyl Ammonium

Amalgam, Half-normal Tetramethyl Ammonium Chloride

in Absolute Alcohol, and in Water.

The electromotive force of the electrode: tetramethyl ammonium
amalgam, 0.5 N N(CH3)4C1 in absolute alcohol, was measured, at o°,
against a decinormal calomel electrode. Measurements were made
by a compensation method and were carried out both while the primary
or polarizing current was running and after the amalgam had been formed
and the polarizing current was cut off.

In order to make the measurements while the polarizing current was
running, it was necessary very rapidly to shift the connections of the
amalgam electrode from the polarizing circuit on the one hand to the
measuring circuit on the other. Le Blanc1 used for this purpose an
electric tuning fork, but in the present work a remodeled electric bell
gave better results as a vibrator. The gong and clapper were removed,
and a platinum contact point passed through the clapper stem, so that
when the armature of the bell was in motion it made contacts in such a
way that the amalgam electrode was alternately in connection with the
polarizing circuit and the measuring circuit. When the vibrator was at
rest, permanent connection was made between the amalgam electrode
and the measuring circuit.

Since the electrode potential of the amalgam was measured against a
decinormal calomel electrode, both while the polarizing current was
running and after it was cut off, provision had to be made for the anode
1 Le Blanc, he. cit.

12

of the polarizing circuit. This anode was made of silver wire, and was
run between the stopper and the edge of the glass half-cell serving as
container for the mercury forming the cathode. Connection between this
mercury and the negative terminal of the polarizing circuit (through the
circuit shifter) was made by means of a platinum wire sealed into a glass
capillary rilled with mercury. This capillary passed through a hole in the
rubber stopper. The glass half-cells containing the amalgam and calomel
electrodes were immersed in crushed ice, while the capillary portions of
the cells extended over the edge of the ice bath into the "bridge" solu-
tion of decinormal potassium chloride. The mercury used in the prep-
aration of the decinormal calomel electrode, and as a cathode in the
amalgam cell, was recently purified by distillation in a vacuum. A
D'Arsonval galvanometer was used in the measuring circuit. In all
instances the "zero" method of measurement was used.

The compensating current in the measuring circuit was derived from
two storage cells; the polarizing current was taken from the lighting
circuit of the laboratory and by means of a variable resistance in parallel
with the electrolytic apparatus the voltage was regulated without difficulty.

TABLE III. — ELECTROMOTIVE FORCE OP THE AMALGAM-CALOMEL ELECTRODE CELL,

ATO°.

A. Polarizing current on.

B. Polarizing cut rent off.

Expt. No.: 205.
Time. E. M. F.
Mins. volts.

206.

E. M. F.

volts.

209. Expt. No
E. M. F. Time,
volts. Mins.

.: 205.
E. M. F.
volts.

206.
E. M. F.
volts.

209.

E M. F.
volts.

0

0.229

0.240

0.247

0

2.6111

2.990

2.630

8

2.928

1.5

2.628

2.622

10

. . .

2.908

2.681

2

. . .

2.671

2.618

ii

2.664

4

2 .629

2.648

2.618

12

2.807

2.908

2.664

6

2.628

2.621

2.618

13

2.681

8

2.605

2 .617

14

2.97

9

2.624

2.622

15

2.807

. . .

10

2.617

16

. . .

2.96

12

1.984

2 .617

18

2.626

14

1.948

19

3.034

2.502

15

2.615

21

2.611

19

1.062

22

3.034

20

2.017

24

2.611

22

2 .604

25

2.990

25

1.958

0.477

27

2.611

2.990

26

2.041

28

2.630

30

2.031

29

2.630

36

2.031

30

2.630

37

0.503

47

2.001

13

In making a measurement, the electromotive force of the two storage
cells was first balanced against that of a cadmium cell. The polarizing
current was then allowed to flow continuously through the hall-cell in
which the amalgam was to be formed, for several minutes. The con-
nections having in the meanwhile been changed from the cadmium cell
to the amalgam calomel electrode cell, the electric vibrator was set in
motion, and the e. m. f . of the cell : Amalgam, 0.5 N N(CH3)4C1 in absolute
alcohol, o.i N KC1, o.i N calomel electrode, balanced against that of the
two storage cells at intervals. After making measurements in this way
for some minutes, the primary or polarizing circuit was broken, and the
e. m. f. of the cell again measured at intervals. After completing this
set of measurements, the current of the storage cells was again balanced
against that of the cadmium element and the average of the two values
of the electromotive force of the storage cells used in calculating
the electromotive force of the amalgam calomel electrode cell.
These two values of the electromotive force of the storage cell were
always close together. The detailed results of these experiments are
given in Table III.

In the series of values given when the polarizing current was cut off,
the value for zero time is taken as the same as the final value obtained
just before the polarizing current was cut off.

Fig. i .

In addition to the potential measurements above, the potential of an
amalgam cell using water as the solvent for tetramethyl ammonium
chloride, of half -normal concentration, was determined. This set of
measurements was likewise made against a decinormal calomel electrode,
at zero, in a manner exactly similar to that above. The results of this
experiment, with the polarizing current on (Series A) and off (Series B)
are given below :

TABLE IV.

Time, min o 2 2.5 4 6 8 10 12 14

Series A 0.219 2.507 ... 2.507 2.452 2.571 2.46 2.46 2.448

Time, min o 2. 5 34 5 6 7 8 9

Series B 2.359 1.99 1.831 1.844 1.811 1.763 1.752 1.736 1.714

Time, min 16 18 20 22 24-27 ... ...

Series A 2.387 2.387 2.373 2.352 2.359

Time, min 10 n 12 14 17 19 ...

Series B 1.699 1.678 1.637 J-596 i-59i 0.063

The curves for two experiments, Nos. 209 and 233, the one in which
alcohol was the solvent, and the other, in which water was the solvent,
are shown in Fig. i.

E. Discussion of the Potential Measurements.

On examining the two curves, it is seen that their forms are very inter-
esting. In 209 B, where the electrolyte was half-normal tetramethyl
ammonium chloride, in absolute alcohol, the potential remained constant
for some tune after the polarizing current was interrupted. This con-
stant value was about 2.62 volts; after awhile, however, the potential
dropped rapidly to a new value near 2.00 volts.

In 233 B, with water as the solvent for the half-normal tetramethyl
ammonium chloride, the average value of the potential, while the current
was on, was about 2.4 volts, but on cutting off the polarizing current the
potential dropped to about 2.0 volts immediately, then gradually fell
to i .6 volts, and in the end quickly dropped to almost zero.

The horizontal or nearly horizontal portions of these curves can have
but one interpretation; a definit phase of the amalgam gives rise to a
definit potential; and when this particular phase is exhausted, if there
is a different phase possible, the potential drops to that of the new phase.
Hence, in 209 (as in 205 and 206, as will be seen by plotting curves for
these experiments) there were formed at least two phases: one may be
the decomposition product of the other, or they may be coincident in the
amalgam, the more active disappearing first.

15

Now, Wilsmore and Johnson1 have recently shown that for the elec-
trode potential of metal, using as a solvent, for the salt of the metal,
liquid ammonia at — 35-5°, a higher value is obtained than when water is
the solvent. They account for this fact on the assumption that the
degree of ionization is less in liquid ammonia than in water solutions of
equivalent concentrations. This explains why, in the present instance,
the electrode potential of one of the phases stable in contact with ab-
solute alcohol solutions of tetramethyl ammonium chloride is 2.6 volts,
while in contact with an aqueous solution of the same salt in equivalent
concentration the potential averaged about 2.40 volts. The two values,
2.6 volts in alcoholic solution and 2.4 volts in aqueous solutions of tetra-
methyl ammonium chloride, are obviously due then to the same phase,
but when the polarizing current is cut off where water is the solvent the
potential immediately falls because this phase is very reactive towards
water. Likewise, the phase showing a potential of 2.0 volts in experi-
ment 209 must be the same which shows a potential of 1.74 volts (average)
in experiment 233, but this phase is likewise less stable in contact with
water than alcohol as is shown by the slowly but constantly changing
values in 233 B.

The high polarization values in experiments 205 and 206 are probably
due to a still more active phase which did not appear in experiment 209.

These experiments do not afford the only proof at hand that more than
one phase of the amalgam is formed. Throughout this work it has been
noticed that immediately under the anode the amalgam formed has a
more solid, more crystallin appearance and a darker color than the rest
of the amalgam; in experiment 300, a portion of amalgam practically
solid was formed. Again, when the amalgam is treated with water the
first action is violent and the whole of the reaction mixture becomes
inky-black, quickly turning to the gray color of ordinary colloidal mercury.
On treating the amalgam with water in a flat dish and washing away the
colloid by means of a stream of water as fast as it is formed a darker,
more solid phase, floating on the mercury, has been noticed. While
this phase still reacts with water to form the colloidal mercury, it is less
active than the main portion of the amalgam.

The existence of these various phases may explain why no constant
values for the ratio of the colloidal mercury to the tetramethyl ammonium
group have been found, as each run has obviously been a mixture of a
number of phases in varying proportions.

1 Wilsmore and Johnson, Elektrochem. Z., 14, 203.

i6

It will now be of interest to compare our results with those of Lewis
and Kraus,1 just published, on the potentials of sodium and sodium
amalgam, although we made no serious attempt at accurate definition
of the potentials measured, while they attained a remarkably high degree
of accuracy. For the cell: Sodium amalgam (0.206 per cent.), normal
Na ion, normal calomel electrode, Lewis and Kraus found an electro-
motive force of 2.1525 volts at 25° and for metallic sodium, normal Na
ion, normal electrode, 2.9981 volts. In experiment 233, Table IV, we
found for the cell: tetramethyl ammonium amalgam, 0.5 normal ion,
o.i normal KC1, decinormal electrode, a maximum electromotive force
of 2.57 volts at o°. This would be equivalent to about 2.73 volts for
conditions as in the experiments of Lewis and Kraus, excepting the
concentration of the amalgam. The latter was not known in our experi-
ment, but was probably in the neighborhood of i per cent. It is, there-
fore, clear that the potential of our amalgam is decidedly greater than
that of sodium amalgam of the same concentration, which conclusion is
in harmony with the very much greater activity of our amalgam toward
water and alcohol.

Reuter2 found that the potential of potassium is 0.4 volt higher than
that of sodium at — 80°, which would indicate a difference of 0.6 volt at
25°. As a rough estimate, then, the potential of our amalgam should
be about the same as that of potassium. From the observed behavior
toward water, however, we should expect the potential to exceed that of
potassium, but to fall somewhat short of that of rubidium. It is of
passing interest, though of doubtful significance, that the molecular
weight of tetramethyl ammonium, 74.1, falls between the atomic weight
of potassium, 39.1, and rubidium, 85.45.

F. Monomethyl Ammonium Amalgam.

The polarization measurements of Le Blanc3 indicated that an amalgam
was formed by the electrolysis of an aqueous solution of monomethyl
ammonium chloride. We had no difficulty in obtaining the amalgam.
The electrolysis took place at ordinary temperature, with 0.25 ampere.
The readily formed amalgam resembled ammonium amalgam in its
properties. When a solution of dry monomethyl ammonium chloride in
absolute alcohol was electrolyzed (experiment 55) at ordinary tempera-
ture, at o° and at — 9° amalgamation again occurred. At — 9° the

1 THIS JOURNAL, 32, 1459 (1910).

2 Z. Elektrochem., 8, 801 (1902).

3 Loc. cit.

17

amalgam formed was pasty. The potential of the cell: monomethyl
ammonium amalgam, 0.5 N monomethyl ammonium chloride, o.i N KC1,
o.i N calomel electrode was measured, exactly as in the method above
for the tetramethyl ammonium amalgam. The solvent in 229 was water
and in 231 was absolute alcohol. "Series A" in each case contains
measurements made while the polarizing current was on; "Series B"
after the polarizing current had been cut off.

TABLE V. — ELECTROMOTIVE FORCES IN VOLTS.

Time, min. 22QA.. 2296. Time, min. 23iA 2318.

0 0.023 2.369 o 0.161 1-952

1 ... I . 646

2 1.655 2 1.372

4 1.666 4 1-314

6 ... i . 666 6 ... i . 340

7 ... 1-670

8 1.318

9 1-623

10 1.323

ii ... 1-569 12 2.126 1.311

13 2.099 I-556

15 2.099 1-504 M 2.082 1.260

17 2.153 1.512 16 1.889 0-387

19 2.153 1.481 18 1-915 0.300

21 2.335 1.481 20 1-915 0.269

23 2.282 1.471 22 1-933

25 1.471 24 1.924

27 2.479 1.460 27 1.952

28 2.369

30 1-455

33 ..- 1-085

34 0.842

On examining these results, it is seen that here the potentials in con-
tact with alcohol are less than those where water is the solvent; it was
found that a similar result was obtained with ammonium amalgam; in-
deed the remarkable fact was discovered that ammonium amalgam is
much more reactive toward absolute alcohol than toward water.

As tetramethyl ammonium amalgam was obtained very soon after this
work was undertaken, almost immediately an extensive search for amal-
gams of other radicals was begun, with the hope that some might be found
which would be more stable. Of eighteen additional radicals studied, only
one, monomethyl ammonium, gave an amalgam which could be isolated.
The others gave, in general, negative results. The salts so studied were:

i8

dimethyl ammonium chloride, trimethyl ammonium chloride, monoethyl
ammonium chloride, tetraethyl ammonium chloride, propyl ammonium
chloride, butyl ammonium chloride, iodornethyltrimethyl ammonium
iodide, aniline hydrochloride, dimethylaniline hydrochloride, phenylene-
diamine hydrochloride, phenyldiazonium chloride, pyridine hydrochloride,
tetraethylphosphonium iodide, tetramethylstibonium iodide, trimethyl-
sulfmium iodide, hydroxylamine hydrochloride and hydrazine hydro-
chloride.

G. Summary of Results.

1. Tetramethyl ammonium amalgam has been prepared by the elec-
trolysis in the cold of solutions of tetramethyl ammonium chloride in
absolute alcohol, using a mercury cathode.

2. The amalgam exhibits certain physical properties of metals to a
high degree. It has a crystallin structure more or less pronounced under
certain conditions. It is lighter than mercury, but does not expand or
become inflated at or below 10°, as does ammonium amalgam.

3. While fairly stable at low temperatures, near 20° it decomposes
rapidly, yielding trimethylamine as one of the decomposition products.

4. In contact with cold air, it becomes coated over with a white alkaline
crust, due to oxidation.

5. Water reacts violently on the amalgam, producing hydrogen, colloidal
mercury, and tetramethyl ammonium hydroxide.

A study of this reaction affords evidence that more than one active
phase exists in the amalgam.

6. The amalgam acts on aqueous solutions of ammonium, sodium,
potassium and copper salts, and alcoholic solutions of copper and zinc
salts, the tetramethyl ammonium group replacing these metals in the
salts and setting free the metals themselves or forming the amalgams of
these metals. With rubidium and cesium salts the action is more
violent than with potassium salts of equivalent concentration, but there
is replacement even in the case of the cesium salts. The tetramethyl
ammonium radical possesses a solution tension comparable with that of
potassium, but probably less than that of rubidium, and considerably
less than that of cesium.

7. The rate of formation of the tetramethyl ammonium ion from the
amalgam, in contact with absolute alcohol, is about 5 per cent, per minute
at o°.

8. The electrode potentials of the amalgam in contact with water
solutions and in contact with alcoholic solutions of tetramethyl ammo-

19

nium salts have been measured against a decinormal calomel electrode
at o°.

2.6 volts and 2.0 volts were the values of the electromotive force of the
amalgam calomel electrode cell found in contact with alcoholic solutions;
2.4 and 1.7 volts were found in contact with aqueous solutions of tetra-
methyl ammonium salts as the values of these potentials. The two
values in each case probably correspond to two phases, thus substantia-
ting other evidence that more than one phase of the amalgam is present
in the amalgam. The curves for these potentials have been plotted.

9. Besides the extensive work upon tetramethyl ammonium amalgam,
search has been made for other possible amalgams. Of the substances
investigated, the monomethyl radical yields an amalgam, and the potential
of this amalgam against a decinormal calomel electrode has been measured.
This amalgam, like ammonium amalgam, is less stable in contact with
alcohol than with water. The dimethyl ammonium radical, according
to potential measurements, may possibly form an amalgam under certain
conditions. Tetraethyl ammonium probably forms a very unstable
amalgam. Of all the other substances investigated, a number gave faint
indications of amalgam formation, but none gave as positive results as
the three substances mentioned above.

10. As set forth in the introduction, we have proposed the hypothesis
that complex radicals, constituting the positive ions of salts, if elec-
trically neutrali?ed by the introduction of electrons, equal in number
to the valence of the ion, will be substances having metallic properties.
We have, however, given some reasons why it may be impossible to isolate
such radicals. While our experiments have not yet led to the complete
isolation of complex radicals, we have gotten a new amalgam of one such
radical which is far more stable than either ammonium amalgam or mono-
methyl ammonium amalgam. Though this substance is a compound of
carbon, hydrogen and nitrogen on the one hand and mercury on the
other, it has true metallic properties. Inasmuch as ordinary binary
alloys with true metallic properties are formed only from components
which are both true metals we are warranted, we think, in concluding
that the organic radicals in our amalgams are in the metallic state and,
therefore, that it is possible to prepare composit metallic substances from
non-metallic constituent elements.

THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW

AN INITIAL FINE OF 25 CENTS

WILL BE ASSESSED FOR FAILURE TO RETURN
THIS BOOK ON THE DATE DUE. THE PENALTY
WILL INCREASE TO SO CENTS ON THE FOURTH
DAY AND TO $1.OO ON THE SEVENTH DAY
OVERDUE.

SEP 251934
FEB 2R 1977

<< t

LD 21-100w-71'33