Peter
T. Tang, M. Sc., Dr. Peter Leisner and Dr. Per Moller
Centre of Advanced
Electroplating (CAG) The
Technical University of Denmark, building 425
2800 Lyngby,
Denmark
Investigation of the properties
of electroplated nickel, using both pulse plating and conventional
direct current (DC), has lead to several interesting improvements
of deposit characteristics. Investigated properties
include; internal stress, tensile strength, yield stress,
elongation, hardness, throwing power, current efficiency
and corrosion resistance (porosity). Experiments have
been made with Watts nickel baths, sulphamate baths and
a modified Watts bath called W3.
Electrochemical deposition of nickel
from a typical electrolyte is a relatively slow process
compared to the speed at which nickel ions move in the bath.
This fact is possible to use very high current densities
at direct current conditions, and even higher current densities
using pulse plating!
The deposition of nickel takes
place through a number of intermediate steps, as indicated
below:
Studying the reactions above,
it becomes clear that the pH-value in the bath is of great
importance - since free hydrogen ions play an important
role in the deposition mechanism. The bath temperature is
also an important parameter, since the speed of the reactions
all depend on temperature (but some more than others!).
It is equally obvious that the concentration of nickel ions
near the cathode, and the current density (available electrons)
has influence on the deposition process.
Pulse plating (on/off plating)
with fast cathodic pulses (from 1 ms and shorter) creates
a new deposition mechanism.
Based on the fact that there
is a significant increase in pH near the cathode surface
during a pulse, a layer of colloidal nickel hydroxide is
formed [10]:
From this layer micelles are
formed by attaching additional nickel ions:
Nickel depositions from this
layer have a semi-bright to bright appearance, a property
usually only obtainable using additives.
The increased pH near the cathode
is created because the short pulses of high current density
(20 A/dm2 or more) attract hydrogen ions which
move much faster in the electrolyte than the OH - ions trying
to escape. As a result of this a thin film of OH is pressed
against the cathode in each pulse, enabling this alternative
mechanism of deposition.
In
this paper the following abbreviations will be used frequently:
| PC |
Pulsed
Current - Uni-polar waveform, e.g. a cathodic
pulse followed by a pause (on/off plating, see experiment
series A). |
| DC |
Direct
Current - Conventional plating at a constant
current level. |
| PR |
Periodic
Current Reversal - Bi-polar waveform, e.g. a
cathodic pulse followed by an anodic pulse (see experiment
series B). |
The
duty cycle for PC (on/off plating) is calculated as:
From this equation it becomes clear that at DC conditions
y equals 100%.
The frequency for PC and PR
is calculated as:
In this equation T represent
the total time of on cycled sequence. Time units are in
milliseconds (ms), this way the dimension of becomes Hz.
Nickel
plating is a relatively old technique for surface treatment.
The first nickel bath was formulated by C. P. Watts in 1916.
This bath is still used today, because its relatively cheap
and easy to operate and maintain.
Using DC plating hardness improvement
can be obtained in Watts baths by replacing some of the
nickel sulphate with ammonium chloride thereby creating
a bath with higher pH value. The throwing power of the Watts
bath can be increased by reducing the amount of nickel sulphate
to 30 g/l and at the same time introduce 180 g/l of sodium
sulphate. The maximum current density in these "high throwing
power" baths is much lower than in a normal Watts bath (7).
Watts Bath, compositions W1
and W2
The baths prepared were very
close to the multipurpose composition used in the industry:
300 g/l Nickel sulphate,
NiSO4 - 6H20
50 g/l Nickel chloride, NiCl2 - 6H20
40 g/l Boric acid, H3BO3
The temperature was kept at 50° C
+ 0.5° C in a 65 litre bath. The bath was a strictly
technical bath, with a pH of 4.5 and no additives what so
ever. A good air agitation was applied at all times.
The amount of nickel chloride in these
baths is a little lower than normal, this reduces the internal
stress in the deposit, but may decrease the current efficiency
and make PR difficult due to passivation during anodic pulses.
Watts bath, composition W3
To improve the current efficiency
during anodic pulses thereby making throwing power improvements
possible (avoiding partly passivation of the sample), the
following bath (called W3) was prepared:
250 g/l Nickel sulphate,
NiSO4 - 6H20
100 g/l Nickel chloride, NiCl2 - 6H20
40 g/l Boric acid, H3BO3
The temperature
was 55°C + 0.5° C and pH 4.0 - as in the other Watts
baths air agitation was applied.
Two identical
baths (S1 and S2), each of 65 litre, with the following
composition were used:
600 g/l Nickel sulphamate,
Ni(NH2SO3)2 4H2O
10 g/l Nickel chloride, NiCl2 - 6H2O
40 g/l Boric acid, H3BO3
The temperature was 55°C +
0.5° C. Since a surface active additive is absolutely necessary
in sulphamate baths, the wetting agent PBA-1 was used according
to prescription. This wetting agent is design for sulphamate
bath without air agitation.
The bath was used with a constant
filter system (pump) and stirring with a submerged propel
at 500 rpm. In sulphamate baths it is essential to use sulphur
alloyed anodes to ensure a reasonable current efficiency
and to avoid decomposition of the sulphamate ions. To avoid
internal stress introduced by chloride ions, the content
of nickel chloride is usually reduced as much as possible
[7] - preferably 10 g/l or less.
Sulphamate baths are more sensitive
towards pollution than the other nickel baths. Organic compounds,
introduced through insufficient cleaning or from additives
used in other baths, makes bright and brittle coatings by
decomposition of the sulphamate ions. Metal pollution will
[7]; increase internal stress in the coatings (iron >
200 ppm) reduce the ductility (copper or zinc > 10 ppm)
or make burned corners (aluminum > 6 ppm). Finally chromium
and lead can reduce hardness and current efficiency even
at very small concentrations.
To be able to determine mechanical
properties such as yield stress, tensile strength and elongation
a tensile test was performed using a specially designed tensile
test rod. The test rod was plated directly on an aluminum
bar masked in such a way that only the rod and a current thief
were exposed to the nickel bath.
After plating the rod and the
thief can be removed from the aluminum substrate simply
be bending the substrate and pulling the nickel film. Because
of the well known poor adhesion between aluminum and nickel,
it is possible to do this without damaging the sample.
An
important property for any plating process is the distribution
of the material. With the aluminum bar technique for the
throwing power can be obtained immediately after the plating
simply be measuring the weight of the tensile test rod and
the current thief. The total area of the tensile test rod
is 0.27 dm2 and the area of the current thief
is 0.19 dm2. Since plating takes place on both
sides of the aluminum the total plateable area is
2 x (0.19 + 0.27) = 0.92 dm2.
The distribution ratio is then
calculated as:
Using the above expression a
distribution ratio or throwing power of 1.0 is obtained
when the coating thickness is the same everywhere on the
test panel (perfect distribution) and less than 1 if the
coating is thicker on corners and edges (this is usually
the case).
Hardness is measured on the
handles of the tensile test rods before the tensile
test itself. Hardness is measured using standard Vickers
hardness equipment with a weight of 50 g (usually referred
to as HV 50 or HV 0.05). For each sample 4 impressions have
been made creating an average hardness value. Given the
nickel films from both the front and back of the aluminum
substrate (rod1
and rod2) each hardness value is actually based
on 8 measurements.
The tensile test rod is placed
in the test machine (Instron) and a standard tensile test
curve is produced. The elongation is then calculated according
to standard procedures just as the test rod has been designed
according the Danish (and international) standards for tensile
test (8).
Because of random errors in
the nickel coating the tensile test rod is generally known
to break before the maximum elongation has been reached.
Scratches and holes in the base metal (aluminum) will also
create weaknesses in the test rod and eventually cause a
premature break. For this reason the highest elongation
value is always used as the result, and experiments have
frequently been repeated to make sure that the obtained
value was correct.
Since it is extremely difficult
to determine exactly where the elastic elongation ends,
it is almost impossible to measure yield stress without
huge dispersion of the results. It is legal, and still within
the standard, to use a line parallel to elastic elongation
line at 0.1% elongation, and then read the yield stress
where this line meets the tensile curve.
The tensile strength is read
at the highest point of the tensile curve - in this case
always the point reached just before breaking. The highest
available value of identical experiments will be used as
the result.
For each experiment the total
charge Q has been measured. This value can the, with the
total weight of the deposited nickel, be used for calculating
the current efficiency (0):
in which:
F
z
MNi
WNi |
96440.4°C/mole
(Faraday's number)
atom
charge (for Ni2+ z equals 2)
molar weight
for nickel (58.71 g/mole)
total weight
of deposited nickel |
The total charge measured in each
experiment (using a coulomb meter) is not entirely reliable
since the meter is not designed for fast alternating currents.
This has resulted in somewhat low current efficiency levels
compared to the almost 100% that nickel deposition usually
exhibits.
For corrosion measurements the
kesternick(1)test has been used according to standard (ASTM
G87-84). The panels are steel sheets (10 x 15 cm) which
are electrocleaned, activated (dry acid) and plated successively
in the same nickel baths as the aluminum bars mentioned
above. After each cycle in the chamber, the panels are evaluated
(red rust) according to standard (ISO 4540) on a scale from
0 to 10. At 10 no visible red rust is seen.
Internal stress measurements
have been done directly in the plating baths using a dilatometer.
Following a specific procedure [9] including strict temperature
control, it is possible to measure the stress building up
in the nickel film - while this is being plated onto a copper
strip. Additional information on internal stress is available
in the electroforming session, when my colleague Mr. Michael
Eis presents his paper; "Measurements of Mechanical Stress
in Plated Coatings for Electroforming".
The experiments have been divided
into series, based on the waveforms used. In each series
parameters and levels have been varied according to the
L4 and L9 orthogonal arrays of the
Taguchi statistical method [9,12]. The series are included
in the appendix Series A pulse patterns have been used for
both Watts baths (W1 and W2) and sulphamate baths (S1 and
S2). Internal stress measurements as well as corrosion tests
have also been conducted using this series of experiments.
Series B has only been used
for Watts baths and series C only for sulphamate baths.
A corrosion test was conducted
using the pulse patterns referred to as series A, on both
Watts bath (W2) and Sulphamate bath (S2). 16 identical steel
panels (10 x 15 cm) was prepared, plated and tested as described
(experimental details).
| Sample name |
Plating Technique |
Thickness
Xmin (um) |
Rating
(red rust) |
| WA1 |
DC,
2 A/dm2 |
5.6 |
6 |
| WA2 |
DC,
6 A/dm2 |
6.3 |
8 |
| WA3 |
PC,
2 A/dm2 |
5.5 |
9.5 |
| WA4 |
PC,
6 A/dm2 |
6.5 |
10 |
| SA1 |
DC,
2 A/dm2 |
5.9 |
2 |
| SA2 |
DC,
6 A/dm2 |
7.0 |
3.5 |
| SA3 |
PC,
2 A/dm2 |
6.2 |
4 |
| SA4 |
PC,
6 A/dm2 |
7.5 |
4.5 |
Table 1: Corrosion results for
pulse plated panels from series A (see appendix). Samples
W are Watts bath and S are sulphamate.
The thickness measurements were
performed using x-ray equipment. This thickness values in
table 1 are for the middle of the panel where the coating
is most thin. Although the thickness in not the same on
all the panels, it is possible to compare experiments 1
with 3 and 2 with 4 directly. Doing that it becomes obvious,
especially for Watts nickel, that the pulse plated panels
have higher rating numbers than the DC plated panels.
Internal stress
building up in the deposit, is a very important parameter
for the quality and the number of possible applications
for any coating. Nickel coatings, primarily from Watts baths,
are known to have high internal stress values, but these
have been reduced using pulse plating.
It
should be pointed out that the lowest values (around 60
N/mm2) are obtained using a relatively high anodic
current in the anodic pulse. High Qa/Qc
ratios take longer time, and might create rough coatings.
At Qa/Qc values around 25% good reliable
coatings are obtained with internal stress values below
100 N/mm2.
For a normal Watts
bath (like W1 and W2) and for the sulphamate bath, throwing
power depends on the current density in the cathodic pulse
- and on the cathodic current density only (see figure 3).
Even at high frequencies it is the
current density in the cathodic pulse and not the average
current density that controls the throwing power.
Using PR plating (like in series
B) it should be possible to dissolve nickel on corners and
edges with a short anodic pulse at a high current density.
In figure 2 it has been shown that the internal stress can
be reduce using PR plating, but unfortunately throwing power
is not improved in a normal Watts bath because the sample
will passivate during the anodic pulse (especially in the
high current density regions, like corners etc.). To increase
the throwing power, it is therefore necessary to change
the bath composition. By replacing some of the nickel sulphamate
with nickel chloride (this bath is called W3), the increase
chloride concentration will make it possible to dissolve
nickel at higher current densities than in the Watts bath,
enabling the throwing power improvement mentioned above.
|
ic
(A/dm2)
|
ia
(A/dm2)
|
Tc
(ms) |
Ta
(ms) |
rav. |
| E1 |
2 |
10 |
500 |
30 |
0.64 |
| E2 |
2 |
4 |
50 |
5 |
0.56 |
| E3 |
2 |
- |
- |
- |
0.53 |
| E4 |
2 |
10 |
500 |
20 |
0.60 |
| E6 |
2 |
10 |
500 |
50 |
0.66 |
| E7 |
1 |
10 |
960 |
40 |
0.70 |
| E9 |
2 |
16 |
480 |
30 |
0.59 |
Table 2: Experiments with bath
W3 to improve throwing power.
The ratio rav. represents throwing power (see
below).
A series of experiments called
series E was carried out using steel panels (5.0 x 7.5 cm)
as the substrate. Experiment E3 was simply DC plating at
2 A/dm2. Each panels weight was measured
before and after plating and the average nickel thickness
calculated. The coating thickness in the middle (where
the coating is most thin) was then measured using x-ray
equipment and an expression for throwing power was calculated:
When this rav. Value
becomes 1 the distribution of nickel is perfect (e.g. the
coating thickness is the same everywhere on the panel).
In table 2 E7 is better than the others because the current
density in each cathodic pulse is only 1 A/dm2
while it is 2 in the other experiments.
Comparing experiments E3 and E6 it
is clear that the throwing power has been improved 23 %,
i.e. using pulse plating the thickness in the middle of
the panel can be increased 23% given the same total charge
(Q=3000 c). When plating with nickel for corrosion protection
purposes, it is not the average thickness but the thickness
at the most thin point of the coating that is important.
If this thickness is increased 23% the total amount of nickel
used can be reduced 23%!
As a side effect of the PR plating
in the W3 bath, the appearance of the nickel coatings can
also be improved. While E3 (DC) was rather dull and grey,
experiments E1, E6 and E9 were semi-bright and E2 and E4
light grey.
It has been claimed by Paatsch
[4] and others [1], that the hardness of nickel coatings can
be significantly increased by high frequency PC plating 9f
> 50 Hz):
|
T
on
(ms)
|
T
off
(ms)
|
Y
(%)
|
i
on
(A/dm2)
|
Micro hardness
(HV 0.01) |
| - |
- |
100 |
5 |
191 |
| 1 |
1 |
5 |
1 |
2 |
| 9.9 |
10 |
50 |
10 |
279 |
| 0.5 |
2 |
20 |
25 |
259 |
| 2.5 |
10 |
20 |
25 |
367 |
| 0.1 |
1 |
9 |
55 |
261 |
| 1 |
10 |
9 |
55 |
385 |
| 1 |
20 |
5 |
104 |
389 |
Table 3: Micro hardness (Vickers 10
g) for on/off plating in Watts nickel [1]. For all tests
the average current density was 5 A/dm2.
The problem
with the results in table 3, is that the current density
in each pulse is relatively high - especially considering
the frequency. From a production point of view a pulse of
100 A/dm2 in 1 ms followed by a 20 ms pause (no
current) is not easy to obtain in large bath with normal
sized samples!
Using lower current densities,
air agitation and a total are of 1 dm2 it is
not possible to obtain these impressing hardness improvements.
Small improvement
is possible using pulse plating (both PC and PR). At low
frequencies (from 10 to 35 Hz) hardness seems to increase
when the anodic pulse time (or pause time) Ta
is longer than the cathodic pulse time Tc (i.e.
when Tc/Ta is less than 1).
This relationship between pulse
time ratios and hardness also appeared in the sulphamate
bath experiments (series A and C).
The current efficiency ranged from
90 to 100 per cent of the theoretical value. As could be expected,
lowest when the current density was high, and close to 100%
at low current densities.
Tensile strength and yield stress
are important parameters for all metal coatings. Experiments
made with both Watts bath and the sulphamate bath show that
the tensile strength and yield stress depends on the hardness
of the coating.
This is perhaps
not surprising, but having established the relationship
it becomes a lot easier to optimize mechanical parameters
in the future, since it will be sufficient to investigate
hardness improvements in order to get an impression of tensile
strength etc.
The elongation results are not
ready for publication yet, but they are not expected to
have any relation to hardness values.
The corrosion resistance provide
by pulse plated nickel is much better than that of conventional
nickel. The internal stress in Watts nickel coatings can
be reduced more than 50% compared to DC plating, when periodic
current reversal (PR) plating is used.
Using a modified Watts bath, it is
possible to improve the throwing power of the bath almost
25%. This can, in combination with the improved corrosion
resistance mentioned above, reduce the total amount of nickel
needed for a specific degree of corrosion protection.
Hardness,
and the other mechanical properties depending on the hardness
such as tensile strength and yield stress, can be improved
using pulse plating, but there might be some technical problems
due to high frequencies.
The investigation of pulse plated
nickel coatings and their properties is far from finished.
The work will continue, and I hope to be able to present
more results, at the conference in June, than what is presently
available.
1.
W. Kleinekathofer et al: "die Eigenschaften von mit pulsierendem
Gleichstrom (pulse Plating) abgeschiedenem Nickel" Metalloberflache
(9) 1982.
2. W. Kleinekathofer &
Ch. J. Raub: "Die abscheidung von Nickel mit pulsierendem
Strom" Surface Technology (7) 1978.
4. W. Paatsch: "Galvanotechnik
mit Strompulsen - Teil 1: Nickelabscheidung". Metalloberflache
(40) 1986.
5. Tai-Ping Sun, C.C.
Wan & Y.M. Shy: "Plating with Pulsed and Periodic-Reverse
Current". Metal Finishing May 1979.
6. R.C.V. Piatti
El al. Electrochim. Acta, (14) 1969 pp.541.
7. Dr. S.A. Watson:
"Compendium on Nickel Electroplating and Electroforming".
Nickel Development Institute, Technical reports 10047
- 10055 1989.
8. Dansk Standard
10 110: "Metalprovning, trackprovning". 2 udg. 1968.
9. Yugo Kimoto, Michael
Eis & P. Torben Tang: "Dilatomater Manual". Internal
report, CAG 920911-59 1992.
10. N.A. Kostin Et al: "Mechanism
of Brighteneing of Nickel Coatings in Pulsed Electrolysis".
Elektrokhimiya (18) 2 pp. 210-214 1982.
11. "Theory and Practice of
Pulse Plating", J - Cl. Puippe (ed.) Publ. By the American
Electroplaters and Surface Finishers Society (AESF), Orlando
1986.
12.
Peter Leisner: "Pulse Plating - Ph.D thesis". Technical
University of Denmark, PI 92.24-A 1992.
1. The kesternick
test is a moist sulphur dioxide corrosion test in which
the panels are exposed for 8 hours to the corrosive atmosphere
and then washed and dryed for 16 hours (one cycle is then
24 hours).
