Electrodeposition of Ni from low-temperature sulfamate electrolytes

Article abstract of DOI:10.1149/1.2181439

The film stress of Ni films deposited at near-ambient temperatures from sulfamate electrolytes was investigated. The particulate filtering of the electrolyte, a routine industrial practice, becomes an important deposition parameter at lower bath temperatures. At 28°C, elevated tensile film stress develops at low current densities (<10 mA cm2) if the electrolyte is filtered. Filtering at higher current densities has a negligible effect on film stress. A similar though less pronounced trend is observed at 32°C. Sulfate-based Ni plating baths display similar film stress sensitivity to filtering, suggesting that this is a general effect for Ni electrodeposition. It is shown that filtering does not significantly change either the current efficiency or the pH near the surface during deposition. The observed changes in film stress are thus attributed not to adsorbed hydrogen but instead to the effects of filtering on the formation and concentration of polyborate species due to the decreased solubility of boric acid at near-ambient temperatures.

Full text of DOI:10.1149/1.2181439

C318
Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
0
013-4651/2006/153͑5͒/C318/7/$20.00 © The Electrochemical Society
Electrodeposition of Ni from Low-Temperature Sulfamate
Electrolytes
I. Electrochemistry and Film Stress
J. J. Kelly, * S. H. Goods, A. A. Talin, and J. T. Hachman
a,
b,z
b
b
a
IBM, T.J. Watson Research Center, Yorktown Heights, New York 10598, USA
b
Sandia National Laboratories, Livermore, California 94551, USA
The film stress of Ni films deposited at near-ambient temperatures from sulfamate electrolytes was investigated. The particulate
filtering of the electrolyte, a routine industrial practice, becomes an important deposition parameter at lower bath temperatures. At
2
2
8°C, elevated tensile film stress develops at low current densities ͑Ͻ10 mA/cm ͒ if the electrolyte is filtered. Filtering at higher
current densities has a negligible effect on film stress. A similar though less pronounced trend is observed at 32°C. Sulfate-based
Ni plating baths display similar film stress sensitivity to filtering, suggesting that this is a general effect for Ni electrodeposition.
It is shown that filtering does not significantly change either the current efficiency or the pH near the surface during deposition.
The observed changes in film stress are thus attributed not to adsorbed hydrogen but instead to the effects of filtering on the
formation and concentration of polyborate species due to the decreased solubility of boric acid at near-ambient temperatures.
©
2006 The Electrochemical Society. ͓DOI: 10.1149/1.2181439͔ All rights reserved.
Manuscript submitted October 5, 2005; revised manuscript received January 10, 2006. Available electronically March 27, 2006.
The electrodeposition of nickel from sulfamate electrolytes is an
industrially important process. The sulfamate chemistry is often
Experimental
1
The composition of the electrolyte was fixed at 1.35 M
Ni͑SO NH ͒ ͑hereafter referred to as Ni-sulfamate͒, 30 g/L boric
acid ͑0.49 M boric acid͒, and 0.2 g/L sodium dodecyl sulfate ͑SDS,
a wetting agent that reduces pitting and otherwise has no effect on
the Ni͒ unless as noted. Ni-sulfamate solution was obtained as a
.12 M aqueous solution from Atotech ͑Electropure 24, Germany͒.
preferred due to low deposited film stress ͑even without S-bearing
2
-4
5
3
2 2
organic additives͒, the low sulfur content of the deposited metal,
6
-9
and the generally good mechanical properties of the Ni. Addition-
ally, the high current efficiency of the bath typically allows for high
deposition rates ͑Ͼ100 mA/cm ͒, permitting the rapid deposition of
1
3
2
3
thick, low-stress films. These electrolytes have been effectively em-
All other chemicals were from Fisher and were of ACS reagent
grade. Manufacturers generally recommend increasing boric acid
concentration from 30 to 45 g/L as the bath temperature is raised
from 32 to 50°C ͑no boric acid concentration recommendation is
made for lower temperatures͒. In the present study, a bath tempera-
ture of 28°C was typically employed, although other temperatures
were investigated as indicated. This temperature was chosen to
minimize thermal expansion and loss of dimensional fidelity in the
thick PMMA photoresist used to defined the throughmold structures
in the LIGA process. This temperature is somewhat lower than the
minimum of 32°C recommended by most Ni-sulfamate manufactur-
ers. The implications of this choice are discussed later. A pH of
3.5 ± 0.1 was used for all experiments; the typically recommended
range is from 3.5 to 4.5. Additions of sulfamic acid were made to
decrease pH as necessary.
S-depolarized Ni S-Rounds ͑Inco, Canada͒ contained in a Ti bas-
ket served as the anode. For unfiltered solutions, a 1 L glass beaker
was used as the reaction cell, while for filtered solutions a liquid
chromatography glass jar ͑Fisher͒ with a volume of 4 L was em-
ployed; the higher volume was necessary to accommodate the filter-
ing pump described below. It was verified that the different cell
volumes had no impact on the experimental results. Stock sulfamate
electrolytes were purified by passage through activated carbon using
an in-tank filtering pump and activated C cartridge ͑Floking, FL͒ for
various periods of time as indicated. The volume of the stock solu-
tion was 7 L. This volume was convenient for the C treatment and
also ensured that all of the solution would be used within a few
weeks of the purification. The purification occurred at the as-
received pH of about 4.5. It is shown that this initial carbon treat-
ment is important to ensure reproducible stress results. The maxi-
mum flow rate of the in-tank pump was about 10 L/min, but the use
of a particle filter reduced this somewhat. Wound, 5-␮m polypro-
pylene fiber filters were used ͑Floking͒.
ployed for years in the bulk electroforming of structural compo-
1
-8
nents, wear-resistant coatings, and stamping/embossing tools.
More recently, the sulfamate bath has become a popular choice
for the through-mask electrodeposition of microsystem components
9
using LIGA or LIGA-like processes. Since many such-fabricated
microcomponents perform mechanical functions, the microstructure
and mechanical properties of the microfabricated Ni structures are
of interest, as well as the deposition conditions that may affect them.
While the process–structure–property relationships are generally
known for Ni sulfamate baths under typical operating
1
4
1
0,11
conditions,
the successful integration of the electrodeposition
step with other steps in the LIGA process ͑or other microfabrication
processes involving plating through a lithographic mask͒ may ne-
cessitate deposition conditions that differ from standard plating
practice.
First, low bath temperatures may be preferred to minimize the
thermal expansion of the resist material ͑typically polymethyl-
methacralate or PMMA͒ used in the LIGA process. Since the lateral-
part dimensions are defined by the PMMA resist, any dimensional
errors in the resist ͑due to PMMA thermal expansion in the plating
bath or swelling from water uptake͒ is replicated by the electro-
formed part. Low electrolyte temperatures thus improve the dimen-
sional fidelity of the final LIGA parts. Additionally, low current
densities are generally employed for high-aspect-ratio features in
thick molds to avoid transport limitations. However, experimental
studies with sulfamate electrolytes under these conditions are few in
1
2
number. In this first part of a two-part study, we investigate the
film stress of sulfamate-deposited Ni at low current densities and
low bath temperatures, as well as the current efficiencies and pH
values near the deposition surface. In the second part, the effects of
these conditions on the film microstructure and mechanical proper-
ties are presented.
Film stress was measured using a bent strip method described
1
5
elsewhere. The deposition time for each current density investi-
gated was varied to keep the total charge constant, resulting in a
nominal film thickness of 1 ␮m for each measurement. This allowed
for the comparison of stress results for different current densities. A
*
Electrochemical Society Active Member.
E-mail: shgoods@sandia.gov
z
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
C319
Figure 2. Effect of a 2-h carbon treatment on nickel electrodeposition kinet-
ics on a Pt rotating disk electrode. Potentials are measured relative to satu-
rated calomel electrode ͑SCE͒ and corrected for ohmic drop. Carbon treat-
ment results in slightly lower currents at low overpotentials.
ment on the deposited film stress as a function of current density
͑
͑
Fig. 1a͒ for unfiltered 28°C solutions ͑Fig. 1a͒ and treatment time
Fig. 1b͒. Figure 1a shows that stress tends to increase monotoni-
cally with increasing deposition rate ͑current density͒; this behavior
is generally considered typical for Ni-sulfamate plating
1
0,19,20
solutions.
It is also clear from Fig. 1a that activated C treat-
ment lowers stress for all current densities, and that in the current
investigation, treatments longer than 4 h yield only a marginal de-
crease in film stress over the entire range of current densities. Figure
1
b more clearly illustrates the effect of activated C treatment at the
two different current densities examined in the present work. Such
carbon treatment removes organic contaminants that may be present
in the as-received electrolyte. In order to obtain reproducible stress
results when varying other deposition parameters, all electrolytes
hereafter studied received a 4-h C treatment.
2
0
Figure 1. ͑a͒ Effect of carbon purification treatment on deposited film stress
for Ni deposited from a Ni sulfamate electrolyte at 28°C. Removal of elec-
The polarization behavior of treated and untreated electrolytes
was investigated to obtain possible insight on the mechanism of the
treatment’s effect on stress. Figure 2 shows the effect of C treatment
on deposition polarization behavior using a Pt rotating disk elec-
trolyte impurities shifts film stress towards less tensile values. ͑b͒ Film stress
for 3 and 15 mA/cm2 as a function of carbon treatment time. The largest
effect is observed after just 2 h of treatment. Error bars represent the standard
deviation of five measurements.
2
trode. For current densities less than about 15 mA/cm , deposition
is somewhat inhibited for the C-treated electrolytes. Other rotation
rates were considered, but the degree of inhibition did not seem to
be a strong function of rotation rate. This behavior is somewhat
counterintuitive, as organics ͑that are presumably removed via puri-
fication͒ are usually thought as inhibiting species for metal deposi-
tion. A possible explanation is that the removed contaminants are
reduced at the cathode and their removal results in a lower measured
current. Another possibility is a more efficient adsorption of other
electrolyte species at the cathode ͑such as sulfamate or borate
constant current source accurate to 0.1 mA was used to deposit the
Ni films for stress measurements. Polarization curves were obtained
using a Pine rotator and Pt disk electrode ͑New Jersey͒ with an
Autolab PGSTAT 12 potentiostat ͑Holland͒.
The pH near the cathode surface was investigated using a Au
mesh working electrode ͑Internet, Inc., Minneapolis, MN͒ and flat-
bottomed pH electrode ͑Omega͒; this technique has been discussed
16-18
2
1,22
in some detail in the literature.
During the measurement, the
ions͒
that may act as weak deposition inhibitors.
bulk solution was quiescent; in between measurements, the solution
was well-mixed by a stir bar. Although no stirring was used during
the measurement, it was verified that bulk stirring had no effect on
the pH signal. Reported pH values correspond to the average value
in a stable plateau that the pH reached after approximately 1 min.
The mesh size was 500, corresponding to a Au wire diameter of
about 17 ␮m spaced about 33 ␮m apart. The validity of this ap-
proach in capturing conditions at the cathode surface has been dis-
Electrolyte filtering effects.— The effect of adding a 5-␮m par-
ticle filter to the bath circulation pump is shown in Fig. 3. Stress at
low current densities is dramatically increased with filtering at 28°C
2
͑
about a factor of 6 increase for 3 mA/cm ͒. The increase in film
stress becomes less pronounced with increasing temperature, as
shown for 32 and 50°C in Fig. 4 and 5, respectively. Although the
data in Fig. 3 show that filtering is an important variable at 28°C,
this bath temperature is below the minimum recommended tempera-
ture of 32°C. But Fig. 4 and 5 show that at temperatures within the
recommended operating range, stress may be affected by simple
particle filtering when using low current densities. These results sug-
gest that if low bath temperatures are desired along with low depo-
16-18
cussed previously
present study.
and we believe it is a useful technique for the
Results and Discussion
Carbon treatment.— Figure 1a and b shows the effect of C treat-
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C320
Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
Figure 5. Deposited Ni film stress for filtered and unfiltered electrolytes at
0°C. Filtering causes a small increase in film stress at 50°C. Plotted values
are the average of two measurements that were within a few percent of one
another.
Figure 3. Deposited Ni film stress for filtered and unfiltered electrolytes at
8°C. Particulate filtering of the electrolyte causes elevated tensile stresses at
low current densities. Error bars represent the standard deviation of five
measurements.
5
2
sition rates, on–off pulse plating with high peak current densities
effect of the deposition temperature on stress as the result of simple
particle filtration, especially at low current densities.
2
͑
Ͼ15 mA/cm ͒ and low duty cycles may be a good approach for
avoiding electrolyte handling sensitivities.
Even though such difficulties can be avoided by employing
higher current densities, determining the source of the electrolyte
handling sensitivity remains of interest. The most obvious possibil-
ity is that the filtering removes boric acid from the electrolyte at low
temperatures. However, measurements of boric acid content in fil-
tered electrolytes indicated this is not the case. It was also verified
that removing the SDS wetting agent did not affect the behavior
shown in Fig. 3. After discounting these possibilities, a series of
experiments was carried out with different electrolytes to determine
the source of the electrolyte handling sensitivity.
We first considered the boric acid content of the bath. Results of
stress behavior from a Ni-sulfamate bath with only 5 g/L of boric
acid are shown in Fig. 7. Filtering at 28°C causes only a modest
increase of stress at low current densities, as opposed to baths with
The effect of temperature and particle filtering is summarized in
Fig. 6a, which shows the difference in stress, ⌬␴, between filtered
and unfiltered baths for 28, 32, and 50°C as a function of deposition
current density. Values for ⌬␴ are calculated by subtracting the film
stress for unfiltered baths from film stresses in filtered baths ͑using
Fig. 3-5͒. Not surprisingly, for low current densities
2
͑
Ͻ10 mA/cm ͒ ⌬␴ diminishes with increasing bath temperature. In
Fig. 6b, we show another interpretation of the data in Fig. 3 and 4 by
plotting the difference in stress values between electrolytes at 28 and
3
2°C for unfiltered and filtered baths as a function of current den-
sity. We consider these two temperatures since filtering has the larg-
est effect on film stress in this temperature range. Differences in film
stress between the two temperatures for unfiltered baths are rela-
tively modest. For filtered baths, Fig. 6b shows the remarkably large
3
0 g/L boric acid. The fact that the filter may be used with only a
minor increase in stress indicates that the elevated stress behavior
observed with filtering is not due to a contaminant from the filter
itself. Instead, the slight increase observed upon lowering the boric
acid content of the bath suggests that the filter-dependent stresses
are strongly influenced by the boric acid bath content.
The effect of boric acid concentration is shown another way by
plotting the difference in film stress between 30 and 5 g/L boric acid
contents for unfiltered and filtered electrolytes in Fig. 8. The effect
of boric acid content seems to be particularly sensitive to filtering, as
the difference in film stress for unfiltered baths is not strongly de-
pendent on current density. For filtered baths, however, a large in-
2
crease in stress is apparent for the 30-g/L bath at 3 mA/cm . Figure
8
makes clear the important role of the boric acid content of the bath
at low temperatures, discussed in more detail below.
We also considered the importance of the anion of the Ni salt in
the film stress behavior by studying a Ni sulfate electrolyte. Figure 9
shows film stresses for a 1.35 M NiSO electrolyte with 0.2-g/L
4
SDS and 30-g/L boric acid at a pH of 3.5; a similar increase in film
stress is observed upon filtering at 28°C ͑about a factor of 4 at
2
3
mA/cm ͒, indicating that the identity of the anion paired with the
2
+
Ni does not play a large role in elevated stresses observed upon
filtering. In fact, for the sulfate electrolyte, the increase in film stress
upon filtering is larger than that for the sulfamate electrolyte. These
results indicate that the electrolyte handling sensitivity is not a char-
acteristic particular to the sulfamate ion and is instead a more gen-
eral effect related to the use of boric acid.
Figure 4. Deposited Ni film stress for filtered and unfiltered electrolytes at
3
2°C. The increase in stress observed at low current densities caused by
electrolyte filtering is less pronounced at 32°C. Error bars represent the
standard deviation of five measurements.
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
C321
Figure 7. Deposited Ni film stress for filtered and unfiltered electrolytes with
only 5 g/L boric acid at 28°C. Deposition conditions are the same as those
used in Fig. 3, except for the lower boric acid content. With a lower boric
acid content, filtering has only a modest effect on film stress values.
Interestingly, the difference in stress is only apparent at low cur-
rent densities. Hoare showed that for low overpotentials boric acid is
2
5,26
essential for Ni deposition in sulfate electrolytes.
He proposed
that Ni cations formed a complex with boric acid adsorbed at the
cathode through which low-overpotential deposition occurred. In the
case of Cu deposition from acid sulfate electrolytes, sulfate ions
2
7,28
have been considered as weak inhibitors at low deposition rates.
Since at low deposition overpotentials film nucleation and growth is
2
7
more readily affected by adsorbed species, we believe that polybo-
rate ions may act in a similar fashion under these conditions.
Electrolyte chemistry.— The effect of filtering on polarization is
shown in Fig. 10. At 28°C, a broad peak starting at −0.5 vs SCE is
evident before the onset of Ni deposition at ϳ−0.88 V for both
Figure 6. ͑a͒ Film stress difference between deposits from unfiltered and
filtered baths, ⌬␴ for different bath temperatures and current densities ͑val-
ues for ⌬␴ are calculated by subtracting the film stress for unfiltered baths
from film stresses in filtered baths using the data from Fig. 3-5͒. Higher bath
temperatures lead to lower ⌬␴ values. ͑b͒ Difference in film stress values for
deposits obtained in 32 and 28°C baths, both filtered and unfiltered. Filtering
filtered and unfiltered electrolytes. This broad peak has been attrib-
25,26
uted to hydrogen evolution on the Pt electrode surface.
The
current density for filtered electrolytes at low current densities was
2
enhances the difference in deposited film stress at 3 mA/cm between the
two bath temperatures.
Since the bath temperature and boric acid content strongly influ-
ence the stress for films deposited in filtered baths, we postulate that
polyborate ions may be formed during the passage of the electrolyte
through the filter material; these polyborate ions would in turn act as
weak inhibitors at low deposition rates. At 30°C, boric acid has an
aqueous solubility of 6.23 wt% ͑ϳ62 g/L͒, but the solubility is de-
pendent not only on the ionic strength of the solution but also the
23
nature of its dissolved ions. Upon cooling to about 28°C, the elec-
trolyte turns from translucent to opalescent; nickel sulfamate solu-
tion concentrate is always translucent, even at room temperature.
Boric acid, when concentrated, has a tendency to form linkages with
2
3,24
itself, resulting in long chains of polyborate ions.
This precipi-
tation was seen as a white solid collecting on the filter at 28°C; no
solid was visible at higher temperatures. Polyborate formation
should be less favored at higher electrolyte temperatures and lower
boric acid bath contents as the boric acid solubility increases. These
trends are consistent with the observed trends in stress behavior.
Another scenario is a large change in the surface pH for the filtered
baths due to a change in the borate chemistry. We revisit this possi-
bility later when we discuss the results of surface pH measurements.
Figure 8. Difference in Ni film stress values for electrolytes having 30 and
5
g/L boric acid for filtered and unfiltered baths ͑at 28°C͒. The difference is
not a strong function of current density, except for filtered baths, where a
2
large difference observed at 3 mA/cm subsides at higher current densities.
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C322
Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
a
Table I. Measured CE for unfiltered and filtered electrolytes.
Current density
and temperature
CE% ͑Unfiltered͒
CE% ͑Filtered͒
2
3
1
3
1
mA/cm 28°C
77.0 ± 1.1
95.6 ± 0.9
76.4 ± 0.7
96.8 ± 0.8
79.9 ± 1.6
94.2 ± 0.5
77.8 ± 0.6
95.9 ± 0.8
2
5 mA/cm 28°C
2
mA/cm 32°C
2
5 mA/cm 32°C
a
Current efficiencies measured by anodic stripping for various Ni elec-
trodeposition conditions.
gen is present at this deposition rate, but the nearly identical
2
3
mA/cm CE values for both unfiltered ͑77.0 ± 1.1%͒ and unfil-
tered ͑79.9 ± 1.6%͒ baths are not consistent with the very different
stress behaviors of these baths ͑see Fig. 3͒. This is an interesting
situation, since stressed Ni deposits at low current densities have
2
9
been associated with adsorbed hydrogen inhibiting growth; the
similar CEs for both low stress and high stress would suggest that
the amount of hydrogen generated is the same in both cases.
Although electrolyte handling appeared to have a negligible ef-
fect on CE at low deposition rates, it was thought that further insight
into the origin of its effect on stress could be gained by studying the
pH near the deposition surface. In order to consider this possibility,
surface pH measurements were made using a flat-bottomed pH elec-
Figure 9. Deposited Ni film stress for filtered and unfiltered 1.35 M Ni
sulfate electrolytes with 30 g/L boric acid. Filtering causes higher tensile
stress at low current densities for sulfate electrolytes as well.
always lower than that for unfiltered electrolytes. This behavior was
extremely reproducible and may be related to inhibition of the hy-
drogen evolution or the Ni deposition reactions. Such inhibition is
consistent with a species more strongly adsorbed at the cathode in
filtered electrolytes. At higher current densities, the difference in
polarization between the two baths disappears. This behavior was
not a strong function of electrode rotation rate. A similar effect is
noted at 32°C, but the broad peak due to hydrogen was not visible.
Current efficiencies ͑CEs͒ were measured by depositing onto a Pt
rotating disk electrode, transferring the electrode to a 0.2 M HCl
1
6-18
trode and Au mesh as previously described in the literature.
Three different schemes for bringing the Au mesh close to the
flat-bottomed pH electrode were considered. In an initial attempt,
the Au mesh was mounted on the electrode with adhesive. A second
approach entailed mounting the Au mesh in a plastic fixture that
could hold the pH electrode. Both approaches were complicated by
the fragility of the assembly and the trapping of bubbles on the Au
mesh.
The third and ultimately most successful approach entailed em-
ploying a loop of 304 L stainless steel wire to press the Au mesh
against the pH electrode for electrical contact. Only the ring at the
bottom touching the mesh was exposed, while the rest of the wire
was insulated with shrink wrap and ran up against the side of the
electrode where electrical contact was made to the cathode lead. The
assembly was mechanically robust and allowed bubbles to escape
from the mesh. While all three methods gave similar results, the data
shown are from the wire-mesh assembly, which gave the best repro-
ducibility and the least noise. Two runs were made for each condi-
tion to ensure reproducibility. We calculated an effective area for the
stripping solution, and galvanostatically stripping the Ni deposit at
2
1
0 mA/cm . Five runs were made for each experimental condition,
with the average and standard deviation reported. The end point is
evident as the potential jumps after the Ni is completely dissolved
and oxygen evolution occurs. Table I shows the results of measured
CEs. There appears to be no large difference in CE between filtered
and unfiltered electrolytes at either temperature for the two deposi-
tion current densities considered. The low current density of
2
2
3
mA/cm has a CE somewhat below 80%, while at 15 mA/cm the
2
CE approaches ϳ97%. The CE value at 3 mA/cm suggests hydro-
1
6
Au mesh as previously described. Before each run, approximately
2
0
.5 ␮m of Ni was deposited on the Au mesh at 20 mA/cm . After
waiting a few minutes for the measured pH to stabilize at the bulk
pH value, the desired deposition current was set. After each mea-
surement, the plated Ni was cleaned from the Au mesh by gently
stripping the plated Ni in a 50%-by-volume H SO solution.
2
4
An example of the pH behavior is shown in Fig. 11. The curve
near the lower part of the figure is the pH electrode response during
2
deposition at 15 mA/cm from a Ni sulfamate electrolyte filtered at
a
Table II. Measured pH for different current densities and filter-
ing conditions.
Current density
2
͑
mA/cm ͒
pH unfiltered
pH filtered
3
5
0
5
3.61 ± 0.01
3.62 ± 0.02
3.69 ± 0.02
3.65 ± 0.02
3.57 ± 0.05
3.73 ± 0.11
3.89 ± 0.02
3.87 ± 0.02
1
1
a
Figure 10. Polarization sweeps for Ni deposition from filtered and unfiltered
sulfamate electrolytes at 28 and 32°C. Deposition current is suppressed at
low current densities for filtered baths.
Measured pH values near the cathode surface during Ni electrodepo-
sition for various conditions. Reported pH value corresponds to the
stable plateau reached after several seconds of electrodeposition.
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
C323
Figure 11. Measurement of pH as a function of time near cathode surface
during Ni electrodeposition from sulfamate electrolytes at 28°C. The curve
near the top of the plot illustrates the rapid pH rise during electrodeposition
when no boric acid is present, while the second curve at the bottom of the
figure shows the relatively small pH rise occurring when boric acid is present
Figure 12. Response of pH near the cathode surface for filtered and unfil-
tered 28°C Ni sulfamate baths ͑both with 30 g/L boric acid͒ during elec-
trodeposition at 3 mA/cm . Despite the large differences in film stress ob-
served for these deposition conditions ͑see Fig. 3͒, both filtered and
unfiltered baths exhibit similar surface pH behavior.
2
͑
both baths were filtered in this case͒.
ences in pH near the surface are observed. We thus postulate that at
low temperatures, electrolyte handling sensitivity is due to the for-
mation of polyborates.
2
8°C. The curve near the upper part of the figure is the response
from a Ni sulfamate electrolyte with no boric acid. Boric acid has
been shown to suppresses hydrogen evolution and stabilize the pH
near the cathode during Ni and Ni alloy electrodeposition.
1
7,28,30,31
Thus, when no boric acid is present, the pH rises to about 6. This pH
rise for the boric acid-free bath is consistent with previous work
Acknowledgment
The authors thank Dorrance McLean for her support with respect
to electrodeposition. Sandia is a multiprogram laboratory operated
by Sandia Corporation, a Lockheed Martin Company, for the United
States Department of Energy under contract no. DE-AC04-
with Watts solutions with no boric acid, where pH values between 5
and 7 were attained, depending on Ni2+ concentration.
16,32
The sec-
ond curve in Fig. 11 shows the pH rise associated with the filtered
2
bath at 28°C for 15 mA/cm . A fairly modest rise from 3.5 to about
9
4AL85000.
3
.8 was observed. This behavior is reasonable and in accord with
4,16
previous measurements using Watts and Ni sulfamate solutions.
Sandia National Laboratories assisted in meeting the publication costs of
Considering the measurement sensitivity shown in Fig. 11, the pH
electrode assembly is clearly capable of detecting changes in pH
near the surface if they occur.
this article.
References
1
. D. Baudrand, Met. Finish., 94͑7͒, 15 ͑1996͒.
2. J. L. Marti, Plating, 53͑1͒, 61 ͑1966͒.
. Y. Tsuru, M. Nomura, and F. R. Foulkes, J. Appl. Electrochem., 30, 231 ͑2000͒.
A summary of all pH measurements is given in Table II. Re-
2
sponse curves for depositions at 3 mA/cm using filtered and unfil-
3
tered solutions are shown in Fig. 12. These two conditions are most
interesting given the strong differences in film stress previously ob-
served in these cases. Figure 12 shows that pH changes for both
conditions are similar and fairly small, from 3.5 to about 3.6. Upon
close comparison of Tables I and II, it is apparent that the filtering-
induced changes in CE and pH are consistent; at low current densi-
ties, a slight rise in the CE and a smaller increase in pH is noted for
filtered baths, while at higher current densities, a lower CE and a
larger increase in pH is observed. Electrolyte handling does not have
a large effect on surface pH, even under conditions that lead to large
changes in film stress. This would suggest that adsorbed hydrogen
does not play a role in higher observed film stresses at low deposi-
tion rates in filtered baths. We thus attribute such changes in the
deposited material not to adsorbed hydrogen but to changes in other
adsorbates, most likely polyborates.
4. Y. Tsuru, M. Nomura, and F. R. Foulkes, J. Appl. Electrochem., 32, 629 ͑2002͒.
5. J. W. Dini and H. R. Johnson, Thin Solid Films, 54, 183 ͑1978͒.
6
7
8
. J. L. Marti and G. P. Lanza, Plating, 56͑4͒, 377 ͑1969͒.
. J. W. Dini and H. R. Johnson, Surf. Technol., 4, 217 ͑1976͒.
. B. E. Jacobson and J. W. Sliwa, Plat. Surf. Finish., 66͑9͒, 42 ͑1979͒.
9. T. E. Buchheit, D. A. LaVan, J. R. Michael, T. R. Christenson, and S. D. Leith,
Metall. Mater. Trans. A, 33, 539 ͑2002͒.
0. W. H. Safranek, The Properties of Electrodeposited Metals and Alloys, 2nd ed.,
American Electroplaters and Surface Finishers Society, U. S. A. ͑1986͒.
1
11. J. W. Dini, Electrodeposition: The Materials Science of Coatings and Substrates,
Noyes Publications, Norwich ͑1993͒.
12. C.-S. Lin, K.-C. Peng, P.-C. Hsu, L. Chang, and C.-H. Chen, Mater. Trans., JIM,
4
1͑7͒, 777 ͑2000͒.
1
1
3. J. W. Dini and H. R. Johnson, Thin Solid Films, 54, 183 ͑1978͒.
4. A. Ruzzu and B. Matthis, Microsyst. Technol., 8 116 ͑2002͒.
15. J. J. Kelly, N. Yang, T. Headley, and J. Hachman, J. Electrochem. Soc., 150, C445
͑2003͒.
1
6. J. Ji, W. C. Cooper, D. B. Dreisinger, and E. Peters, J. Appl. Electrochem., 25, 642
1995͒.
͑
1
1
7. H. Deligianni and L. T. Romankiw, IBM J. Res. Dev., 37͑2͒, 85 ͑1993͒.
8. P. Kern, C. Bonhote, and L. T. Romankiw, in Magnetic Materials, Processes, and
Devices VII and Electrodeposition of Alloys, S. Krongelb, L. T. Romankiw, J.-W.
Chang, Y. Kitamoto, J. W. Judy, C. Bonhote, G. Zangari, and W. Schwarzacher,
Editors, PV 2002-27, p. 328, The Electrochemical Society Proceedings Series,
Pennington, NJ ͑2003͒.
Conclusions
We have shown that at low electrolyte temperatures, boric acid
acts as an inhibitor if the electrolyte is handled in such a way as to
promote the formation of polyborates. Particle filtering of electro-
lytes thus becomes an important consideration at low bath tempera-
tures. Higher film stresses are observed at low current densities
Ͻ5 mA/cm ͒ in low-temperature, filtered baths. Despite the large
changes in film stress for these conditions, no concomitant differ-
19. S. A. Watson, Nickel Sulphamate Solutions, NiDI Technical Series 10052, Nickel
Development Institute, Toronto ͑1989͒.
2
2
0. N. V. Mandich and D. W. Baudrand, Plat. Surf. Finish., 89͑9͒, 68 ͑2002͒.
1. J. L. Bubendorff, L. Cagnon, V. Costa-Kieling, J. P. Bucher, and P. Allongue, Surf.
Sci., 384, L836 ͑1997͒.
2
͑
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C324
Journal of The Electrochemical Society, 153 ͑5͒ C318-C324 ͑2006͒
2
2
2
2. E. Chassaing, M. Joussellin, and R. Wiart, J. Electroanal. Chem. Interfacial Elec-
trochem., 157, 75 ͑1983͒.
3. R. A. Smith and R. B. McBroom, Encyclopedia of Chemical Technology, 4th ed., p.
28. S. Wu, J. Lipkowski, T. Tyliszczak, and A. P. Hitchcock, Prog. Surf. Sci., 50͑1–4͒,
227 ͑1995͒.
29. J. Amblard, I. Epelboin, M. Froment, and G. Maurin, J. Appl. Electrochem., 9, 233
3
65, John Wiley & Sons, New York ͑1992͒.
͑1979͒.
4. F. A. Cotton, G. Wilkinson, C. A. Murillo, and M. Bochmann, Advanced Inorganic
Chemistry, 6th ed., p. 131, John Wiley & Sons, New York ͑1999͒.
5. J. P. Hoare, J. Electrochem. Soc., 133, 2491 ͑1986͒.
30. T. Fritz, H. S. Cho, K. J. Hemker, W. Mokwa, and U. Schnakenberg, Microsyst.
Technol., 9, 87 ͑2002͒.
31. J. Horkans, J. Electrochem. Soc., 126, 1861 ͑1979͒.
32. Y. Tsuru, R. Takamatsu, and K. Hoskawa, J. Surf. Finish. Soc. Jpn., 44͑1͒, 39
͑1993͒.
2
2
2
6. J. P. Hoare, J. Electrochem. Soc., 134, 3102 ͑1987͒.
7. R. Winand, Electrochim. Acta, 39, 1091 ͑1994͒.
Downloaded on 2014-09-08 to IP 128.123.35.41 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).