º gº
Sºº c.
ºx. ޺
º
§3.
º
* . g
§ §§ §
sº
º
d
jº
º º
§:
ensile Strength of Electrolytic
Copper on a Rotating Cathode
fº
&
A THESIS
PRESENTED ro THE FA Cu1.TY OF THE GRADUATE SCHOOL OF CORNELL
UNIVERSITy FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
BY
§
:
Kºś
º
#
*:::::
§ -
º:
§

* : * * *
* * * * * * *
* t
} : * *
* * * * * * , , ,
* * * : . . . . ;
* * * *
*
A e
. . . . . . . . ;
* : * * * * , , , , ;
* * * . . . . . .
e
* { M :
!!!!"; ; ; , , ...,
::::::::::::
* * * *
* * * * *
{
:::::
e :*** * * **,
'''''' . . . . .....
* * * * * * *
* * * * * * .
!!!” tº:*:
A n
* * *
* * * * * *
* * * * * *
R
* : * > . . . .x:
*::::::: ; ; , , i.
| sº
4 : * *
3.

•neinical Library
– e
2- - \
C e
ºpºv" W
Tensile Strength of Electrolytic
Copper on a Rotating Cathode
A THESIS
PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF CORNELL
UNIVERSITY FOR TEIE DEGREE OF DOCTOR OF PEIILOSOPHY
By
CHARLES WILLIAM BENNETT
[Reprinted from the Journal of Physical Chemistry, Vol. 16, p. 294 (1912)]
TENSILE STRENGTH OF ELECTROLYTIC COPPER ON
A ROTATING CATHODE
BY C. W. BENNETT
As early as 1865,” it was noticed that during electrolysis,
if the cathode were rotated, a higher current density could be
used. Among those actually making use of this was Wilde
who patented a process in 1875, using for cathode a vertical
iron cylinder which was rotated. The current density used
in this process was never more than 20 amperes per square
foot.
The next important commercial application of this
principle was the Elmore process. In this, the cathode was a
, mandrel, rotating vertically, over which agate burnishers
rotated, keeping the deposited copper tubes smooth, and of
constant thickness throughout. This process is used com-
mercially in Europe at present in the manufacture of seamless
copper tubes. The current density used is not more than
30 amperes per Square foot. The copper thus obtained has a
tensile strength of from 36,000–50,000 pounds per square
inch, depending, it is stated, on the speed of rotation.” How-
ever, a part of this increase in tensile strength, over that of
copper precipitated on a stationary cathode, is most likely
due to the increase in the rate of burnishing, for we have
here a true analogue to the process of rolling. Due to the
increased rate of rotation, the tube passes under the agate
burnisher faster, thus giving more and more an approxima-
tion to the cold rolled copper, with its correspondingly higher
tensile strength. The Dumoulin process substituted a
burnisher of sheepskin for the agate of the Elmore process.
It was claimed that the animal fat insulated the projections,
thus tending to give a more even surface. The current
density was run up to 40 amperes per square foot. No
* See “Electrochemical and Metallurgical Industry,” 6, 412 (1908), for
review of these processes.
* “Electrochemical and Metallurgical Industry,” 3, 83 (1905).
\ºy
:
Electrolytic Copper on a Rotating Cathode 295
mention is made of the strength of the deposit. This was
tried commercially in England but failed completely.
Emerson in 1899 patented a process" for making a copper
wire by plating copper on a rotating mandrel, wound around
with a spiral of insulating material. The strip between the
insulating material was then pulled off and drawn down.
In 1899 the same man patented a process” for making copper
bars. A copper strip was wound spirally around a cathode
which was rotated. The strip was thickened to a bar by
depositing the copper on it while rotating. A large (sic)
cathode was rotated slowly.
S. O. Cowper-Cowles patented” practically the same
process, using a smaller cathode and rotating it more rapidly.
The United States Patent called for a process using a cathode
moving “at such a rate of speed that will cause the hydrogen
bubbles to be thrown off from the metal deposited on the
cathode, and cause such friction between the metal deposited
on the cathode and the electrolyte as to yield tough and
smooth deposits.”
Using a cathode twelve inches in diameter rotating IOOO
R. P. M., and an electrolyte of 12.5 percent copper sulphate
and 13 percent sulphuric acid, at about 70°C, with a current
density of 200 (preferably 170) amperes per square foot,
copper foil was obtained which was even stronger than the
best cold worked copper. The results would have been more
conclusive if a thicker deposit had been precipitated and
tested. The liability to error will be apparent when the
thickness of an average sample is considered. This is given :
Dimensions Square Tensile strength per
Inches inch square inch
I. Io9 X O.OO6 O. OO66 51,000 pounds
* U. S. Patent No. 395,773, January 8, 1899.
* U. S. Patent No. 638,917, I899.
* English Patent No. 26,724, 1898; U. S. Patent No. 644,029, February
2O, IQOO.
* “Mineral Industry,” 9, 229 (1900).
277939
296 C. W. Bennett
These results, as given, mean absolutely nothing in them-
selves, as will be seen below in the work on brass and bronze.
Five reasons are given for rotating the cathode.
(1) The electrolyte is stirred and impoverishment is
prevented. -
(2) The copper is burnished by the friction with the
electrolyte. e
(3) Foreign matter is eliminated, thus preventing “tree
formation.” * •
(4) Air bubbles are brushed away, thus preventing
“nodule formation.’’
(5) Thickness of deposit is uniform.
There is little doubt that all of these factors enter in and
tend to increase the tensile strength, and to enhance the
character of the deposit. However, it seems highly im-
probable that the factors given above are the only ones
entering into the equation. For this reason, it was deemed
expedient to make some experiments, to see if other factors
could be found, and to find their true relation to the tensile
strength of the deposit. Then the principle was to be ap-
plied to the brasses and bronzes, with the hope that alloys
of high tensile strength would be obtained.
With the copper the various factors were studied by
holding others constant and varying one for a series of runs.
In this way the effect of speed of rotation, current density,
concentration of electrolyte, and temperature were deter-
mined. The apparatus consisted of an electrode holder de-
signed to rotate continuously at any speed up to 6000 R. P.
M., and carry 300 amperes. This has been described in a
previous paper."
For currents up to 150 amperes, the IIo volt circuit,
from the university power house, was used. For the higher
current, a motor generator set was used.
The test pieces were prepared by the general method
outlined in the previous paper referred to above. These
pieces ran from ooºo-o.O60 inch thick. The actual deposit
* Jour. Phys. Chem., 16, 287 (1912).
Electrolytic Copper on a Rotating Cathode 297
before turning down was much thicker. The measurements
of the cross section were taken with micrometer calipers,
reading directly to O.OOI inch. These pieces were then
broken in an Olsen testing machine. Five or more tests
were made, and their average taken as the true tensile strength.
During the runs it was found necessary to burnish the
deposit once or twice. A mechanical burnisher was not
desirable, for it would be open to the objection that the
copper was being “rolled” as in the Elmore process. There-
fore it was deemed best to stop the run once or twice, de-
pending on the relative rate of stirring and on the current
density, and burnish with emery paper. This was done by
holding the paper against the rotating tube. The surface
was then washed, treated with a strong solution of potassium
cyanide to remove grease, and then with I : 1 nitric acid
solution to slightly roughen the surface and ensure the ad-
herence of the next layer of copper deposited.
In general, if a solution be stirred during crystallization,
the crystals resulting are smaller than those from the same
solution without stirring, because more nuclei are formed.
In depositing a metal, then, if it be precipitated directly
in the crystalline state, we shall expect to get smaller crystals
if the solution be stirred vigorously. However, the pre-
cipitation of the metal in the crystalline state directly, is not
at all probable. It is likely that the metal comes down in
a condition analogous to a “melt,” and then crystallizes
from this. By rotating the cathode the uncrystallized ma-
terial is agitated and smaller crystals result. The force
of the rotation tends to move the material, forcing it to de-
velop new crystal centers, and in this way prevents the growth
of large crystals. Hence in precipitating copper, smaller
crystals were expected from the run where the rotation was
rapid.
It is also known that the tensile strength of steel, copper,
etc., is increased by rolling. Rolling, it is generally admitted,
does nothing more than break down crystal aggregates,
giving a more finely crystalline mass. Hence, with the
298 C. W. Bennett
precipitated copper an increase of tensile strength was ex-
pected with a decrease of crystal size, as the rotation was
increased. When this was tried the results showed that the
theory was correct. - .
A solution containing 20 percent CuSO,5H,O, and 12
percent H.SO, was used. The temperature at starting was
35° C. This was desirable, for trial showed that this was the
temperature maintained throughout a run at the current
density used, i. e., 500 amperes per square foot. The de-
posits were treated alike, and every precaution was used to
keep all conditions, save speed of rotation, constant. Re-
sults for four runs are tabulated below:
CURRENT DENSITY 500 AMPEREs PER SQUARE FOOT
Revolutions Tensile strength in Tensile strength in
per Voltage vertical direction | direction of rotation
minute Pounds per sq. in. Pounds per sq. in.
I75O 3.2 37,000 *
25OO 3.8 49,000 4 I,OOO
35OO 4. 7 5 I,OOO 51,000
5500 * 58,000 66,OOO
See Fig. I for curves representing these results. The voltage
curve is discussed later.
R. P. N.1. in Hundreds
I º ſ f | l l l | .
[5 20 25 3O 35 4 O A5 5O 55
1–1 i 1. f 1
Fig. I

Electrolytic Copper on a Rotating Cathode 299
A gradual increase in the tensile strength, with increased
rotation of the cathode, is seen from the curves. There has
been therefore a gradual decrease in the size of the crystals
due to the increased rate of agitation of the uncrystallized
matrix. *
The observations given in the last column were taken
to show that the theory, assigning the increase in tensile
strength to the mechanical deformation of the crystals or
particles of copper, by the friction against the solution, is
untenable. This theory states that the particles are drawn
out in the direction of rotation, and a fibrous interlacing
mass is obtained. According to this, the tensile strength
would be greatest in the direction of the lamination. The
test pieces in the form of rings were cut off the bottom of the
tube, while it was in the lathe. These were broken by apply-
ing pressure in the direction of a tangent. Fig. 2 gives a
- • , * • * - *
A/A/A ºf — S7A: Azi– 24444/v7&Z2 23 S.4142 Colºr
/~ - -
tº- —# *is——ºft-l- V : " ' \ la
gº £º * º \ *
NS Lºri Air-º-º:#
*A. e g * W g
- sº #. ++ ole || || || ** { }
22:24:52. *N* S$ ºff- ii W
* * -* - - I Aro 7& S7 Aºravg
- º/Z.Z. Ar
- -3-1-3#— --! &P * CoA-A-aff
3& <! .. 3 x 4'szze, Aay ... •
Žo-zzae e-zae & cºszoa "o-32 ºzacz. Scaevº, #2Oaws. Z2/sc-S7A & Z.
A £423/v7& D
Z//v/- — S7&z.–2 P42.4/V7 & P
Fig. 2
sketch with detail of the apparatus used for these tests.
This was prepared in the following way: A disk of steel
1" diameter and 9/16" thick was sawed through with a hack
a "... " .. 3
º
* *
& ,”
:
:

3OO C. W. Bennett
saw, in the middle, giving two half discs. Holes were drilled
in these, as shown in the figure, a little to the right and left
of the centers of gravity of the sections. These were then
fastened in two forks or brackets by pins, as shown. The
brackets could then be held in the testing machine, the ring
of copper slipped over the disc sections, the brackets pulled
apart and the ring broken.
The breaking load was then divided by two, thus assum-
ing that it was evenly divided. In the last run, at 55oo
R. P. M., the difference is a bit greater than the experimental
error and needs a little comment. At this rapid rate of
rotation the solution is very violently agitated, and very
fine bubbles of air are carried down into the solution. These
are to some extent, most likely, entrapped and occluded
in the copper. Now the bottom end of the cathode would
receive the least amount of this air, and here the most com-
pact deposit would be found, and consequently the tensile
strength would be highest. This is the most probable ex-
planation, for, on examination, the texture of the deposit
at the end proved to be finer. Anyway, if the theory were
true, all the measurements should be higher than those on the
vertical strip, by an amount depending on the speed of rota-
tion. Since they do not show this general tendency it may
be concluded that the deformation of the crystals is not a
factor.
In order to check up the inverse ratio of crystal size
and tensile strength this series was polished, etched with
ammonium hydroxide (I : 3), and the crystal sizes and
shapes noted under the microscope. It was hoped that the
measurements of the actual size of the crystals could be
obtained, but they were so small that the results would have
meant nothing had they been taken. However, in passing
along the series rapidly a decrease in the crystal size could be
detected. When they were mixed, one could arrange the
series by noting the crystal size under the microscope. No
difference in the shape of the crystals could be seen, as would
be expected according to the deformation theory. Again
Electrolytic Copper on a Rotating Cathode 3OI
the structure of the piece, from planes at right angles to each
Other, appeared to be the same. With the higher rates of
rotation the deposit was smoother and better, showing that
there was more efficient stirring and hence less impoverish-
Inent.
In Fig. I, a curve is given, showing the voltage drop
across the cell at varying speeds of rotation. An astonish-
ingly large increase in voltage with increase in rotation is
noted. The increase was more than was anticipated. It
may be due to an increased resistance of the cell, or a back
electromotive force. Increased resistance of the cell may be
caused by an increase in the resistance of the brush contact,
an air film on the cathode, or a tendency for the electrolyte
to concentrate in the Outer portion of the cell by being sub-
jected to centrifugal force. These tendencies would increase
with increased rotation. The back electromotive force may
be due to frictional electricity, from the friction of the cathode
against the liquid, Or a more rapid solubility of the cathode
than the anode. An increase in relative solubility of the
cathode may be caused by the fact that the crystals at this
electrode are much smaller than those of cast copper, or in-
creased stirring, tending to equalize the concentration of the
liquid about the cathode, and aiding diffusion, would allow
the cathode to dissolve more rapidly than the anode, and thus
set up a back electromotive force. Lastly the increased
relative solubility of the cathode and anode may be due to
a combination of the last reason with another. There may
be an impoverishment with respect to copper ions in the
film about the cathode. This then would give the cell,
Cu | x — y Cuº & Cu | Cu,
or a concentration cell, manifesting its electromotive force
in a direction opposite the charging voltage. When now, the
cathode is rotated, the increase in solubility may be greater
than the reverse tendency, caused by stirring the solution
faster, and consequently the final effect would be a total
increase in the solubility of the cathode, and hence a higher
3O2 C. W. Bennett
back electromotive force. Which and how many of these
factors enter in, determining the increase in voltage, cannot
be said at present. A detailed study of this, however, has
been begun, and it is hoped that results will be forthcoming.
This seems to be an important point in explaining some of the
results gotten with revolving cathodes. These will be dealt
with in detail in a future paper.
The variation of current density was then studied. It
is generally known that rapid crystallization from a solution
or “melt” gives smaller crystals than that which takes
place more slowly. If now in depositing a metal the current
density be increased, the amount of metal to crystallize in
unit time is increased, and hence the crystals should be smaller.
With the smaller crystals goes more even loading stress and
hence higher tensile strength. Therefore, for any given set
of conditions the tensile strength ought to increase as the
current density is increased. The upper limit will depend
on the rate of stirring, other things being equal. If this be
true, the maximum tensile strength would be found at a
higher current density, if the stirring were more vigorous.
To show this, two series of observations were made, one at
2500 R. P. M., another at 5500 R. P. M. These were not,
however, run under the same conditions. Due to the large
current used, the temperature rise was found to be different
in the series of the first run. To decrease this difference a
higher initial temperature was used in the second series.
The effect of this increase had been anticipated, and will be
discussed under temperature, below. The desired point,
however, is illustrated, that is, that the maximum tensile
strength is obtained at a higher current density where the rate
of rotation and hence rate of stirring is higher. The solution
was the same as was used before. In the first series, the
initial temperature was that of the room, in the last it was
about 50° C. Results are tabulated below.
Electrolytic Copper on a Rotating Cathode
3O3.
2500 revolutions per minute 55oo revolutions per minute
Amperes Tensile Amperes Tensile
per strength per strength
square Voltage pounds square Voltage pounds
foot per Sq. in. foot - per sq. in.
3OO 2.5 6O,OOO 34O 2.7 34,OOO
4OO 3.3 68,000 5OO 3. 8 5O,OOO
5IO 3.9 4O,OOO IOOO 7. I 4I,OOO
I IOO 7.2 35,000 I6OO II . I 32,OOO
I7OO I 2 .. 2 I4,OOO 24OO 2O. 3 28,OOO
* * * 4OOO 27.5 I3,OOO
These results are shown graphically by curves in Fig. 3.
[ſ)
(*)
§
}*
(\!
Fºs
&
}*
H.
19
}*}*--
o
|-
Current Densitu in Amps. per sq. ft.
_l | l | g | Ps fºr s |
| L
| |
3OO 4OO 5CO
*T5:5--T55
Fig. 3
The first three points were checked, that there might be
no doubt as to the form of the curves. The tensile strength
reaches a maximum and then drops off. This is due, no
doubt, to the inefficient stirring, at the higher current
densities. The current density was run up to a limit where
the deposit was noticeably bad, due to the separation, prob-
ably, of cuprous Oxide. The sample at 24OO amperes per
square foot was very ductile, could be bent and worked
nicely. No trace of oxide or other impurity was noted.
The sample at 4000 amperes per Square foot was brittle, but

3O4. C. W. Bennett
worked nicely in the lathe giving a perfect copper surface.
The fracture, however, showed a reddish brown color.
Upon examination under the microscope a gradual
decrease in the size of the crystals with increase in current
density could be noticed.
A few runs were made to show that variation of the
concentration of the electrolyte makes little difference in the
tensile strength. The following experiments were made,
varying first the concentration of copper sulphate and then
that of the sulphuric acid. *
CURRENT DENSITY 500 AMPERES PER SQUARE FOOT.
Percent CuSO4.5H2O Percent H,SO, |PO º: †he
I2 I5 6O,OOO
2O I5 58,000
25 & I5 55,000
I5 I 2 6O,OOO
I5 25 57,000
The last one of each series is a bit low, but this is due
to the slightly elevated initial temperature required to hold
the copper sulphate in solution. The current efficiency was
not so good where a large percentage of acid was used.
Now as to temperature variation, which has been antici-
pated above. If copper wire be drawn while cold through a
die, the effect is a decrease in ductility and an increase in
tensile strength. This is caused by the breaking down of the
large crystals giving smaller ones embedded more or less in
amorphous material. The distance of shear through the
crystal is relatively short, and hence the elongation is short,
or in other words the ductility is slight. The tensile strength
goes up, for the material is worked more or less toward the
amorphous end, i. e., the crystals become smaller and under
stress show a more even loading. If the hard drawn copper
wire be annealed, the tensile strength goes down to about
30,000 pounds per Square inch as a limit, and the ductility
is increased. The crystals become larger, the distance of
Electrolytic Copper on a Rotating Cathode 3O5
shear before they break is longer and hence the ductility is
increased. By depositing copper on a rotating cathode,
from a solution at room temperature, it is possible to duplicate
the cold worked copper, getting a hard compact sample with
a very slight ductility, and a high tensile strength. Now by
raising the temperature of the solution and hence that of the
uncrystallized copper deposited, it ought to be possible to
anneal the metal, while it is being precipitated. Or looked
at from the other point of view, crystallization from a hotter
solution should give larger crystals, other things being equal.
The following illustrates the principle admirably.
CURRENT DENSITY 500 AMPERES PER SQUARE FOOT
Temperature Tensile strength
25°C 63,000 pounds per square inch
50° C 49,000 pounds. per Square inch
75° C 3O,OOO pounds per Square inch
This is graphically illustrated by a curve, Fig. 4.
lſ)
cº)
§
}*
8
§
H.
R
|-
º
}*|-km.1.}* *
Temperature of Electrolyte
| l | L– | i |
O
25° 5O° 75°
The last sample, especially, was very ductile and soft, and very
similar in properties to the annealed copper wire.
In this work at the current densities above 500 amperes
per square foot, no attempts were made to cool the solution
during the progress of the run. The large rise in temperature
explains why the tensile strength at the higher current densities

3O6 * C. W. Bennett
º
was always low. The heat developed by the resistance of the
solution keeps the temperature up, and anneals the deposit
while forming. Under these conditions the maximum tensile,
strength is that of soft drawn copper, 3O,OOO–35,000 pounds
per square inch. The samples above 500 amperes per square
foot, Fig. 3, were all ductile except the last ones of the two
series, given at 2500 and 5500 R. P. M., respectively. Here
the deposit was noticeably bad.
It was thought advisable to get an approximation of the
current efficiency with the rotating cathode at a high current
density. There being no evolution of gas at either electrode
this was gotten by taking the ratio of the anode loss and
cathode gain. Consequently, a run was made, using IOOO
amperes per square foot, while the cathode was rotated 5500
R. P. M. The results follow : -
Weight of anodes before 2582 grams
Weight of anodes after 2529 grams
Anodes loss 53.O grams
Weight of cathode after 869 grams
Weight of cathode before 816.2 grams
Cathode gain 52.8 grams
53 = 99.6 current efficiency.
The efficiency therefore is practically IOO percent.
Some runs were made with varying amounts of gelatine
(*/,-5 grams per liter of solution) to see if the character of the
deposit would be improved. Good conditions otherwise,
that is, 400 amperes per Square foot, at 2500 R. P. M., were
used. The deposits, however, were very hard, and so brittle
they could be crushed between the fingers. The brittleness
was no doubt due to the mechanical deposition of traces of
gelatine, with the copper. The gelatine is undoubtedly
present here as a colloid. It is possible that the particles of
copper become covered with a surface film of the gelatine,
and are thus prevented from adhering to each other. Thus
a “brick and mortar” structure would result, which could
Electrolytic Copper on a Rotating Cathode 3O7
easily be broken down. This is of importance in accounting
for the results gotten with bronze, for here, too, there is cer-
tainly colloidal tin oxide present in the solution.
A run was made with copper nitrate and nitric acid, the
concentrations of which were equivalent to I5 percent
CuSO4·5H,O and 12 percent H.SO,. A current density of
500 amperes per square foot, at 2500 R. P. M., was used.
The deposit, however, was dark, hard and brittle.
In an attempt to apply the principle of rapid stirring
and high current density, to the alloys of copper, bronze was
studied. The first question was that of a solution from which
to deposit the copper and tin. The solution recommended
by Curry" was first tried. The solution contained 15 grams
of CuSO4·5H,O, 28 grams of SnC,O, 5 grams of H.C.O., and 55
grams of (NHA),C,C), per liter of water. The anodes were
90 percent copper and IO percent tin. The current density
varied from 20–2OO amperes per square foot, but no satisfac-
tory deposit could be obtained. At first bronze would be
deposited, then the deposit became copper-rich until finally
only copper was precipitated. CuCl2.2H,O was substituted
for CuSO4·5H,O with only greater complications, for here we
introduced the possibility of cuprous chloride. The con-
ductivity of the solution was increased by adding ammonium
chloride, but the character of the deposit was not enhanced.
There are two possibilities here. The tin may go in as
stannous tin and be oxidized to the stannic condition, from
which it would not be precipitated, or the tin may dissolve
and become colloidal. If the latter be true, a copper-rich
deposit should be obtained from the beginning, if the solution
were boiled before the deposition was started. Here the tin
in solution would be changed entirely to the colloidal state,
and only copper would deposit. When this was tried the
results indicated that the tin went bad, due to the formation
of the colloid. A solution was made up as given above.
One portion was electrolyzed, cold for a minute, and another,
after being boiled, was run under identical conditions. A
* Jour. Phys. Chem., Io, 515 (1906).
3O8 C. W. Bennett
bronze was obtained in the first case, while in the latter
practically a pure deposit of copper resulted.
On account of the low conductivity of this solution it was
deemed wiser to study the alkaline tartrate, for use in this
precipitation. When the solution recommended" was tried,
no deposit at all could be obtained. The sodium hydroxide
concentration was high enough so that presumably it alone
was decomposed. Some preliminary runs were made then,
on a small scale, to determine the proper concentration of the
various substances. Solutions containing copper sulphate
and sodium stannate in the proportion of 90 parts of copper
to IO parts of tin, with enough sodium potassium tartrate
to dissolve the whole easily, with varying amounts of sodium
hydroxide, were electrolyzed with an anode of 90 percent
copper, and Io percent tin. These were run at a moderate
current density using a rotating perforated platinum electrode.
After running for at least 24 hours and after a deposit of
constant composition was obtained, the deposits were com-
pared, the solution giving the best bronze was analyzed, and
this concentration used. This solution contained 22 grams
of CuSO,.5H,O, 5.5 grams of Na,SnO, I5O grams of
NaKC, H,O, and 5 grams of NaOH per liter. The anode
analyzed 89 percent copper, the cathode 93.5 percent copper.
From this and a consideration of the final tin content, a
good part of the tin went into the electrolyte in the unavail-
able form, most likely colloidal. The deposit on the platinum
was brittle. However, about 50 grams of a compact deposit
was obtained, which showed very nearly the same composi-
tion throughout. The solution above being used, runs were
made with 6, 18, and 30 amperes per Square foot, but the
deposits were too brittle to get a measurement of the tensile
strength. The relative amounts being the same, the solution
was made five times more concentrated, in order to increase
the conductivity, and runs were made at I50 and 250 amperes
per square foot. The deposits were no better, however.
The trouble here is most likely due to the colloidal material.
* McMillan: “A Treatise on Electrometallurgy,” p. 287.
Electrolytic Copper on a Rotating Cathode 3O9
This may be mechanically deposited with the bronze and
prevent adherence, as the gelatine with the copper. Colloidal
tin or tin oxide is no doubt present, for the tin content in the
solution gradually goes up, showing that it is going into solu-
tion in an unavailable form. Tests applied for stannic tin
were negative. When this is taken up again, the question
of a solution is the first one to be answered. A systematic
search should be made for a solution of high conductivity,
where the chances for the formation of colloidal material and
stannic tin are negligible, and for One, from which a deposit
of constant composition can be gotten. Any attempt to use
the solution given above is useless if a bronze with good
physical properties is desired.
The next attempts at the application of the foregoing
principle of rapid stirring and high current density were the
study of brass. The solution used here was one of the double
cyanides of copper and zinc with excess of potassium cyanide.
This solution, 85 grams of ZnSO,.7H,O, 35 grams of Cu(CN),
and 130 grams of KCN per liter, was electrolyzed with brass
anodes containing 60 percent copper and 40 percent zinc.
The current density varied from 12–150 amperes per square
foot, with a rotation of 2500 R. P. M. Only one deposit
was tough enough to test. A measurement was obtained on
the run at 75 amperes per Square foot. The average of two
tests gave 9,500 pounds per Square inch. Two pieces were
annealed at 800° C for 8 hours and quenched, and then
tested. Curiously the tensile strength went up. *The average
of two measurements gave 18,OOO pounds per square inch.
There has been a change in the piece other than the growth
of the crystals, otherwise the tensile strength would have
gone down. The fracture of the unannealed piece was inter-
spersed with brownish black specks and lines which dis-
appeared when the piece was annealed. It is possible that
the piece as deposited had interspersed through it, probably
as a film around the crystals, Some decomposition product
of the cyanide solution, which was driven out by annealing
at the temperature used. At least, the curious reversal must
|
S.-- 3 + 6 º'
3 9015 O7505 2590
Z
<C
Ø
I
C
>
LL-
O
>–
|-
(/)
CC
ULI
>
Z
1-D
|
ſaeg！！！！！！！！！：·e-****， ， ， ， ，· ·º ，！！！！！！！
：：:：：：：：：：::-：- ！， （~~~~ ..." --.* * · · · _ _ - __，_|• • • •: ~~） *----
e- - -----+---+--+ → →→→→• • • • • ***

*: *） º.º.
§§
.**
č№&&
:№ſ
·ſ-№-3№
*
№ž
►
**
§§¿
§§§§
§§§§?
ș$ſa，~ſae;· •%%*（?：）?（.*¿¿.¿-- ĶĒģ> ·¿??¿*¿¿.*·§§§§§· ºs º ae|-§§
:。
%% #4）
§§
Ř
º; º
º, ºr ºf rºº
*
§§§
*: *）: