Effect of surface pH on electrodeposited Ni films

Article abstract of DOI:10.1149/1.2196672

Ni metal was potentiostatically electrodeposited on a vertical plane cathode in a Watts bath in order to determine the effect of the cathode surface pH on the Ni film microstructure. Polarization curves and current efficiencies were measured at pH values of 1.5, 3.4, and 5.5. The surface pH value (pHs) was estimated from the measured partial current density for hydrogen gas evolution based on a steady-state one-dimensional mass transfer model. Any species relating to the buffering function through the dissociation reactions (HSO4-, H3 BO3, Ni4 (OH) 4 4+, H3 O+, OH-) were taken into account. The pHs abruptly rose to above 6 as the partial current density for H2 gas evolution increased. The preferred orientation of electrodeposited Ni thin film was plotted in electrode potential- pHs diagram. It was found that the transition boundary between {110} and {100} preferred orientations was located along a ridge 500 mV below the H+ H2 equilibrium potential line. This relationship suggests that the dissolved hydrogen atoms in Ni metal are partly responsible for the evolution of structural texture of the Ni films.

Full text of DOI:10.1149/1.2196672

C502
Journal of The Electrochemical Society, 153 ͑7͒ C502-C508 ͑2006͒
0
013-4651/2006/153͑7͒/C502/7/$20.00 © The Electrochemical Society
Effect of Surface pH on Electrodeposited Ni Films
a
a, ,z
b,
b,
M. Motoyama, Y. Fukunaka, * T. Sakka, * and Y. H. Ogata *
a
Department of Energy Science and Technology, Kyoto University, Yoshida-honmachi, Sakyo,
Kyoto 606-8501, Japan
b
Institute of Advanced Energy, Kyoto University, Uji, Kyoto 611-0011, Japan
Ni metal was potentiostatically electrodeposited on a vertical plane cathode in a Watts bath in order to determine the effect of the
cathode surface pH on the Ni film microstructure. Polarization curves and current efficiencies were measured at pH values of 1.5,
s
3
.4, and 5.5. The surface pH value ͑pH ͒ was estimated from the measured partial current density for hydrogen gas evolution based
on a steady-state one-dimensional mass transfer model. Any species relating to the buffering function through the dissociation
−
4+
+
−
s
reactions ͑HSO , H BO , Ni ͑OH͒ , H O , OH ͒ were taken into account. The pH abruptly rose to above 6 as the partial
4
3
3
4
4
3
current density for H gas evolution increased. The preferred orientation of electrodeposited Ni thin film was plotted in electrode
potential-pH diagram. It was found that the transition boundary between ͕110͖ and ͕100͖ preferred orientations was located along
a ridge 500 mV below the H /H equilibrium potential line. This relationship suggests that the dissolved hydrogen atoms in Ni
2
s
+
2
metal are partly responsible for the evolution of structural texture of the Ni films.
©
2006 The Electrochemical Society. ͓DOI: 10.1149/1.2196672͔ All rights reserved.
Manuscript submitted December 19, 2005; revised manuscript received February 27, 2006. Available electronically May 17, 2006.
Electrodeposited Ni metal films are used for protective coatings
on carbon steels. Such films also have been applied to microelectro-
lored materials is carried out, at least on the current density coordi-
nate. The overpotential concept may replace the qualitative coordi-
nate of current density for noble metals, whereas the simultaneous
hydrogen gas evolution must be taken into account for the elec-
trodeposition of iron group transition metals. Thus, the pH value at
the cathode surface may provide an additional coordinate.
1,2
mechanical systems ͑MEMS͒
as well as magnetic recording
3
,4
5
heads and ultrathin films for magnetic sensors. The textured
structure of electrodeposited Ni metal films is known to evolve as a
result of nucleation and growth processes, depending on the electro-
lytic conditions, e.g., the electrolyte composition, pH, bath tempera-
ture, current density, stirring condition, and organic additives. The
specific structure determines the physicochemical and mechanical
properties of Ni films. Much more effort to improve the microcrys-
talline structure of high-area electrodeposited films is also necessary
in the field of advanced energy science. Doing so leads to a highly
functionalized catalytic electrode by tailoring a unique interface that
is required for efficient energy conversion.
It is the present objective to understand the electrodeposition
2
+
reaction mechanism of Ni ions onto a vertical plane substrate im-
mersed in a stagnant electrolyte. One of the key issues is to correlate
the surface pH value to the orientation index of the electrodeposited
film.
Experimental
−
1
Ni was electrodeposited from a Watts bath containing 262 g L
6
-17
−
1
−1
In spite of many works,
the mechanism of Ni electrodeposi-
NiSO ·6H O, 45 g L NiCl ·6H O, and 38 g L H BO ͑pH 3.4͒
4 2 2 2 3 3
2
+
tion has not been thoroughly understood. Because the Ni /Ni sys-
tem has a more negative equilibrium potential ͑E = −0.23 V͒ than
the H /H couple, H2 gas evolution inevitably accompanies Ni
at room temperature. The pH value was adjusted by adding H SO
2 4
o
or NaOH aqueous solutions. The vertical cathode was fabricated
from a rolled silver sheet and the anode from a nickel sheet. The
effective surface area of both electrodes was 1 ϫ 1 cm. The vertical
plane anode surface was installed parallel to the cathode. Prior to
each experiment, the substrate was ultrasonically cleaned sequen-
tially in acetone, ethanol, and distilled water for 15 min. The grain
size of Ag substrate was less than 1 ␮m. Electrodeposition was con-
ducted potentiostatically at potentials from E = −0.8 to −1.6 V with
respect to a Ag/AgCl, KCl sat. ͓E = +0.20 V vs standard hydrogen
electrode ͑SHE͔͒ reference electrode. The bulk pH values were var-
ied from 1.5 to 5.5.
+
2
metal electrodeposition. The pH value in the neighborhood of the
+
cathode increases as the H ions are reduced to evolved H gas.
2
The Centre National de la Recherche Scientifique ͑CNRS͒ re-
search group investigated the correlations between the microstruc-
ture of Ni films electrodeposited from a Watts bath and the electro-
6-10
lytic conditions.
They considered the inhibiting influences of
hydrogen adatoms ͑Hads͒, H2 molecules, and the stabilization of
Ni͑OH͒ on the cathode surface. The texture evolution phenomena
2
were then discussed. Nakahara and Felder examined electrodepos-
1
5
Working and counter electrodes were configured face to face on
the inner walls of the electrolysis cell. The reference electrode was
placed close to the silver working electrode using a Luggin capillary.
ited Ni films with transmission electron microscopy ͑TEM͒. They
emphasized the possible formation of metal hydrides as well as the
inclusion of hydroxide.
−
2
The amount of charge passed was maintained at 150 C cm .
It would correspond to 35–50 ␮m film thickness under the presump-
tions of uniformly and densely precipitated films without any void
fraction and current efficiency over 70%. After experiments, the
electrodeposited Ni specimens were cleaned carefully with distilled
water, dried, and weighed. The partial current density for Ni elec-
trodeposition was calculated from the mass of Ni electrodeposits,
which was determined gravimetrically and was directly converted
to current efficiency. The crystal orientation of the electrodeposited
film was examined by X-ray powder diffraction using Co K␣
radiation.
Dahms and Croll proposed the concept of the surface pH in order
to confirm their model that nickel hydroxide works as an inhibitor in
16
the case of Fe–Ni alloy electrodeposition. Deligianni and Roman-
kiw then measured the pH in the neighborhood of the cathode and
demonstrated that the surface pH value was actually higher than that
17
in the bulk solution.
The texture structure evolution on a polycrystalline substrate was
18
examined in Europe after Fischer. A classification diagram with
two coordinates of current density divided by metal-ion concentra-
1
9
tion and inhibition intensity was proposed by Winand. His diagram
may provide a guide to control the microstructural and morphologi-
cal variations of an electrodeposited film. Unfortunately, his chosen
coordinate is qualitative. It would be preferable to describe them in
a quantitative manner when the electrochemical processing of tai-
Results and Discussion
Polarization curve and current efficiency.— Figure
the polarization curve measured in the solutions at pH 1.5, 3.4, and
.5. The current density increased almost linearly from 10 to
1
shows
5
−
2
*
Electrochemical Society Active Member.
E-mail: fukunaka@energy.kyoto-u.ac.jp
160 mA cm at pH 1.5. In the higher pH ͑3.4, 5.5͒ solutions, the
z
current density increases with decreasing potential almost in a par-
Downloaded on 2015-03-17 to IP 132.174.255.116 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 ͑7͒ C502-C508 ͑2006͒
C503
Figure 1. Polarization curves and current efficiencies during electrodeposi-
tion in a Watts bath.
allel way among electrolytes with three pH values in the potential
range from −0.8 to − 1.4 V. The current density is larger at the
lower pH. However, the tendency of the current density to increase
is significantly suppressed at E = −1.6 V.
Because the equilibrium potential of the electrochemical reaction
Ni2+ + 2e = Ni is E = −0.23 V vs SHE, H gas evolution always
−
o
2
accompanies Ni electrodeposition from a thermodynamic perspec-
tive. The current efficiency is less than 100% within the range ex-
amined in this work ͑see Fig. 1͒. Ni metal electrodeposition is con-
ventionally carried out in a solution at such high pH as to restrict H2
gas evolution, but doing so induces more nickel hydroxide forma-
tion under a higher pH condition. A Watts bath for the industrial Ni
metal electrodeposition is hence adjusted to pH 3–5 and boric acid is
always added as a pH buffer.
Figure 2. Variations of the partial current densities for Ni electrodeposition,
i , and H gas evolution, i , with the electrode potential.
Ni
2
H
2
hypothesis of the semi-infinite model in direction to anode, because
the natural convection is induced significantly at high current den-
Figure 1 illustrates that the current efficiency is 76% at E
2
6
=
−0.8 V and then increases to 84% at E = −0.9 V in the lowest pH
sity. The effect of the counter electrode must be taken into account
͑
1.5͒ solution. This means that the partial current density for Ni
because the anode–cathode distance is only 1 cm. With an effective
electrodeposition increases more rapidly than that for H evolution.
concentration boundary layer thickness ͑␦͒ of 100–200 ␮m, the lim-
2
In the more negative potential range, the current efficiency remains
−2
iting current density is calculated to be 0.13–0.26 A cm by the
8
3
4–87%. In contrast, a nearly 100% efficiency is maintained at pH
.4 and 5.5, which are close to the pH value for industrial electro-
following equation
c
plating. These pH solutions show almost the same results as 97–99%
iL = zFD
͓2͔
efficiency over E = −0.8 to − 1.2 V, but the efficiency decreases to
͑1 − t͒␦
9
0% at E = −1.6 V.
The limiting current density on a vertical plane cathode im-
This calculated limiting current density still looks high. Sup-
pressing the dependence of i on electrode potential at E = −1.6 V
may be understood not only in the mass-transfer mechanism, but
also in the other chemical transfers that are described later.
mersed in a semi-infinite stagnant electrolyte is calculated by Eq. 1
20-22
from the similarity profile principle.
1
y
/4
Sh = 0.499Ra
y
͓1͔
Figure 2 shows the partial current densities for Ni electrodeposi-
tion ͑i ͒ and H gas evolution ͑i ͒. The iNi at pH 5.5 is slightly
Ni
2
H
2
where Shy and Ray are Sherwood number and Rayleigh num-
ber, respectively. No H2 gas evolution is presumed. The li-
lower than the other ones, particularly at potentials more positive
than E = −1.4 V as the pH dependence of i is similarly demon-
−
2
miting current density ͑i ͒ of about 0.30 A cm is given by
L
2+
strated in Fig. 1. That may be caused by lowed Ni concentration
−
1
2+
⌰
= c͑=1.19 mol L ͒ ͑see Appendix A͒. Numerical values of Ni
ions are listed in Table I. It is a little questionable to accept the
4+27,28
due to the complexes formation as Ni ͑OH͒
as well as the
4
4
−
1
difference of the electrolyte conductivity ͑pH 1.5: 0.059 S cm ,
pH 3.4: 0.051 S cm , pH 5.5: 0.050 S cm ͒. The current density
iNi͒ continues to increase even above 0.10 A cm at pH 1.5.
The H gas evolution rate has a significant role in enhancing the
−
1
−1
−
2
͑
Table I. Numerical values for Ni2+ ions used in the calculation.a
2
2
9-31
cathodic mass-transfer coefficient
natural convection.
because it promotes the
Reference
D ͑cm2 s−1͒
t ͑-͒
6.9 ϫ 10−6
The iH2 at pH 3.4 and 5.5 is always smaller than at pH 1.5. Such
16
23
24
25
0.4
a difference is partly caused by the pH dependence of an equilibrium
−
1
−1
−2
+
␮
͑g cm
s
͒
1.77 ϫ 10
electrode potential of H /H system.
2
3
1.23 ϫ 102
␣
͑cm mol−1͒
Preferred orientation.— Figure 3 shows the X-ray diffraction
͑XRD͒ patterns of the electrodeposited Ni thin films. The XRD pat-
terns change abruptly when E decreases from −0.9 to − 1.05 V at
a
Viscosity and densification coefficient are estimated in the electrolyte
solution containing only NiSO4.
Downloaded on 2015-03-17 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C504
Journal of The Electrochemical Society, 153 ͑7͒ C502-C508 ͑2006͒
Figure 3. XRD patterns of Ni films electrodeposited at E = ͑a͒ −0.8, ͑b͒
0.9, ͑c͒ −1.05, ͑d͒ −1.2, ͑e͒ −1.4, and ͑f͒ −1.6 V. ͓The bulk pH values are
A͒ 1.5, ͑B͒ 3.4, and ͑C͒ 5.5.͔
−
͑
pH 1.5. The peak intensity ratios observed at E = −0.8 V are similar
to those in the Joint Committee on Powder Diffraction Standards
͑
JCPDS͒ powder pattern. The 220 plane peak becomes predominant,
and the other peaks almost disappear, in the potential range below
−
1.05 V.
This kind of microstructural transition in the diffraction pattern
also appears when E decreases from −1.05 to − 1.2 V at pH 3.4.
However, the 200 plane peak becomes predominant before the
strong 220 planes peak appears at pH 3.4. The Ni thin film deposited
at pH 5.5 is characterized by the relatively strong 311 plane peak at
E = −0.8 V. Thus, the 220 plane peak generally appears in the high
overpotential range in all pH solutions.
Figure 5. ͑A͒ Texture distribution of electrodeposited Ni thin film in the
3
3
2+
E–pH diagram reported in Ref. 6. The activity coefficient of Ni is ap-
3
4
proximated as 0.04. The broken line indicated by ͑H͒ shows the equilib-
+
rium potential of H /H2. ͑B͒ Distributions of the current efficiency and the
b
preferred orientation in the E–pH diagram. Each inset number from 77 to 99
The preferred orientation of an electrodeposited film was quan-
titatively discussed by conveniently defining the orientation index of
is assigned to the current efficiency ͑%͒. The dotted lines indicate the bound-
aries between the different preferred orientation domains with shadow areas.
32
hkl plane ͑Mhkl͒.
o
I
I
h k l ͚j Ih k l
i
i i
j j j
M
=
͓3͔
values greater than 1 and less than 1 represent the preferential ori-
entation and the suppressed orientation compared with the powder
pattern, respectively.
h k l
i i
i
o
h k l ͚j Ih k l
i
i i
j j j
where Ihkl and Io are XRD intensities for the hkl plane of an elec-
When the solution pH is 1.5, all crystal planes show M of nearly
1 at E = −0.8 V. M220 increases abruptly to 8 at E = −1.05 V and is
thereafter saturated. As a result, the preferential growth of the other
planes are completely suppressed more negative than E = −1.2 V.
At pH 3.4, only the ͕100͖ plane is preferentially oriented at E
= −0.8 to − 1.05 V. It is replaced by the ͕110͖ plane in the poten-
tial range below −1.2 V. The orientation indices of the other planes
always remain near zero.
hkl
trodeposited film and the JCPDS ͑04-0850͒ powder pattern, respec-
tively.
Figure 4 shows the preferred orientation index variation of the Ni
films as a function of electrode potential. If all crystal planes are
oriented in all directions with the same probability as a powder
pattern, then the M value is equal to 1. The hkl planes exhibiting M
The ͕311͖ preferred orientation with M311 = 2–3 is seen at
E = −0.8 and −0.9 V at pH 5.5. The other planes are not completely
suppressed. The ͕100͖ preferred orientation appears only at
E = −1.05 V.
Appearance of the ͕110͖ preferred orientation is inevitable as the
electrode potential decreases, regardless of pH. It is often pointed
out that the adsorbents, e.g., hydrogen adatoms and nickel hydrox-
ide, strongly affect the preferred orientation of electrodeposited Ni
films.
The preferred orientation variations were plotted with the current
efficiency contour map onto the E-pH diagram in Fig. 5B. They had
6
-10
already been reported by Froment et al.
in a similar electrolyte
−
1
−1
composition ͑300 g L NiSO ·7H O, 35 g L NiCl ·6H O, and
4
2
2
2
1
4
0 g L− H BO ͒ ͑see Fig. 5A͒. The ͓110͔ orientation appears twice
3 3
A
B
with increased cathodic overpotential ͓͑110͔ and ͓110͔ denote the
͓110͔ axis orientations at lower and higher cathodic overpotentials,
respectively͒. The ͓210͔ and ͓211͔ orientations are separated by the
Figure 4. Variations of the orientation indices of Ni films with the electrode
potential. ͓The bulk pH values are ͑A͒ 1.5, ͑B͒ 3.4, and ͑C͒ 5.5.͔
Downloaded on 2015-03-17 to IP 132.174.255.116 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 ͑7͒ C502-C508 ͑2006͒
C505
͓
͓
100͔ one in the low- and high-pH regions. They described that the
110͔ , ͓211͔, ͓210͔, and ͓110͔ orientations are attributed to the
iH2
F
ץC
H A
n
ץCH+
ץCOH−
A
B
=
ͯ
DH
+
ͯ
−
ͯ
DOH
−
ͯ
+
ͯ
nD
ͯ
H A
n
ץx
ץx
ץx
x=0
x=0
x=0
inhibitions by Hads, Ni͑OH͒ , H molecules, and both Ni͑OH͒ and
H molecules, respectively, and that the ͓100͔ texture was evolved
by the crystal growth free from any adsorbed species.
2
2
2
͓
8͔
2
8
-11
where iH2 is the current density for H2 gas evolution due to Re-
actions 4 and 5; F is Faraday constant; D
diffusion coefficients of H+ ions, OH− ions, and buffering species
H A, respectively; C is the concentration of species ͑i͒; and x is the
distance normal to the electrode plane.
Figure 5B was constructed with the following results. The ͕311͖
+
, DOH
−
, and D
are
H A
H
n
preferred orientation is located nearest to the region where Ni͑OH͒
is stable. The electrodeposited sample with a ͕311͖ preferred ori-
entation showed a gray-colored and smooth surface. The current
efficiency is close to 100% at ͑E, pH ͒ = ͑−0.80 V, 5.5͒ where
the ͕311͖ orientation exists. The ͕100͖ orientation area surrounds the
2
3
3
n
i
b
We can substitute COH
−
and C
in Eq. 8 by using the relations
H A
n
͕
311͖ area. A random orientation is prevalent in the lowest pH and
CH
+
COH
−
= KW
͓9͔
the most positive potential corner.
A
n
H
In the present work, the ͓110͔ orientation was not confirmed
because the electrodeposition was not conducted at more positive
potentials than E = −0.8 V. The ͕311͖ orientation is found in the
domain where the ͓211͔ orientation prevails in Fig. 5A. The random
orientation at ͑E, pH͒ = ͑−0.80 V, 1.5͒ appears in the domains of
the ͓210͔ orientation in Fig. 5A. The ͓110͔ orientation becomes pre-
dominant with increased the cathodic overpotential in Fig. 5A and
B. The definition of the crystal preferred orientation plane is slightly
different between the reports by Froment et al. and this paper. It
must be emphasized, however, that the transition boundary lines
between the ͓100͔ and ͓110͔B orientations similarly locate in both
E-pH diagrams.
Along the pH axis from ͑−0.80 V, 5.5͒ to ͑−0.80 V, 1.5͒, the
current efficiency decreases from 99 to 76%. The ͕311͖ orientation
changes through the ͕100͖ orientation and finally reaches a random
orientation at the lowest current efficiency of 77%.
Additionally, this ͕311͖ orientation changes to the ͕110͖ orien-
tation through the ͕100͖ orientation domain along the E axis
from ͑−0.80 V, 5.5͒ through ͑−1.6 V, 5.5͒ and then toward the
C
+
CAn− = KH ϫ CHnA
͓10͔
and
C = C
+ CAn−
͓11͔
H A
n
where KW is the autoprotolysis constant of water; K , the dissocia-
tion constant of buffer; C, total concentration of buffer components;
, the concentration of undissociated buffer acid; CAn−, the con-
centration of buffer anion. Eq. 8 is now rearranged as
H
C
H A
n
n−1
iH2
ץCH+
DOH−KW ץCH+
H+ ץCH+
CC KH
2
=
DH
+
+
+ n D
H A
n
2
H
n
2
F
ץx
C
ץx
͑KH + C +͒ ץx
H
+
͓12͔
Equation 12 must be integrated over the distance between the elec-
trode surface ͑x = 0, CH = C ͒ and the outer edge of the “diffu-
s
H
+
+
b
H
sion layer” ͑x = ␦H , CH
+ + = C ͒. Appendix B describes how to de-
termine the diffusion layer thickness.
+
͑
9
−1.6 V, 1.5͒ corner. The current efficiency gradually decreases from
1 to 87% along this trajectory line.
Figure 5B indicates that the transition of the preferred orientation
As the first-order approximation, we introduce an identical dif-
fusion layer thickness for all species in order to integrate Eq. 12.
This approximation gives Eq. 13
plane may not perfectly accommodate the current efficiency contour
map in the two-dimensional E–pH map.
b
H
s
H
iH2
F
C
+
− C
+
b
H
s
H
␦H
+
= DH
+
͑C
+
− C
+
͒ + DOH
−
KW
b
H
s
H
C
+C
+
Surface pH.— Dahms and Croll tried to understand the anoma-
b
H
n
s
n
lous codeposition of Fe–Ni alloys by introducing the possibility of
͑
C
+
͒ − ͑C
+
͒
16
H
nickel hydroxide formation as an inhibitor. They evaluated the pH
value at the cathode surface based on the diffusion model combined
with hydrolysis reactions of metal ions. The pH at the cathode sur-
face remains an interesting subject for the researchers to examine
+ nD K C
͓13͔
H A
H
b
s
H
n
n
n
͕
KH + ͑CH
+
͒ ͖͕K + ͑C
+
͒ ͖
H
where it is assumed for the simplicity of calculation that the varia-
tion of buffer species concentration C with x is negligibly small.
This assumption is acceptable because the diffusion layer thickness
is only 100–200 ␮m and the partial current density is relatively
small.
2
8,35-37
the electrodeposition mechanism of transition metals.
The pH
value at the cathode surface changes due to the overall reactions
+
−
2
H O + 2e = H ͑g͒ + 2H O
͓4͔
͓5͔
3
2
2
In this system, the candidate species for the buffer acids are
and
−
4
4+
4+
4
HSO , H BO , and Ni ͑OH͒ . It was reported that Ni ͑OH͒ is
3
3
4
4
4
−
−
2
H O + 2e = H ͑g͒ + 2 OH
+
2+
2
2
predominant rather than NiOH among the Ni complexes in the
2
7,28
Both processes may take place at nickel electrodes.
concentrated solution.
The first species stems from H2SO4 as the
pH adjuster only when the pH is set to 1.5. Thus, n, HnA, and A
represent one, HSO , and SO , respectively. The amount of neutral
4 4
species H2SO4 is assumed to be zero due to the complete dissocia-
tion of the first step reaction ͑see Appendix C͒.
The third term on the right side in Eq. 13 cannot be applied to
H BO and Ni ͑OH͒ because those dissociation reactions do not
n−
The diffusion process tends to compensate the pH variations at
+
−
2−
the interface. The H O and buffer species must be considered. Any
3
species reacting with H O or OH− exhibits a buffering function
+
3
through the reactions as
+
n−
nH + A = H A
͓6͔
4+
4
n
3
3
4
In addition, hydrolysis reactions
follow the type of Eq. 6.
Men+ + mOH = Me͑OH͒ ͑mn −m͒+͑Me = Ni, n = 2, m = 2͒ ͓7͔
−
The dissociation mechanism of H3BO3 is rather complicated.
The role of boric acid to catalyze the Ni electrodeposition reaction
3
8-41
must be considered.
The diffusion of all buffering species can be treated in the same
way by taking into account the one-dimensional diffusion normal to
in a nickel sulfate solution was often pointed out.
In the present
paper, only the buffering action of boric acid is considered. The
3
3
literature describes the following ten equilibrium reactions in the
aqueous system
the cathode surface. In the subsequent treatment, the species H A
n
therefore represents hydrolysis products as well.
−
+
KHm = 10−9.21
H BO = H BO + H
͓14͔
͓15͔
For a steady-state one-dimensional diffusion process inside a par-
ticular concentration boundary layer thickness normal to a plane
electrode, we can express the flux as
3
3
2
3
H BO = HBO + H+ KHm = 10−12.70
−
2−
2
3
3
Downloaded on 2015-03-17 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C506
Journal of The Electrochemical Society, 153 ͑7͒ C502-C508 ͑2006͒
Table II. Diffusion coefficients used in the calculation.
2
Species
H+
Diffusion coefficient ͑cm s−1͒
Reference
−5
5.0 ϫ 10
42
42
16
43
28
−
−5
OH
HSO4
1.0 ϫ 10
−
−5
1.1 ϫ 10
−
5
H BO
1.1 ϫ 10
4.3 ϫ 10
3
3
4
4
+
−6
Ni ͑OH͒
4
ץC
ץCH BO−
ץCHB O−
ץCB O2−
4
7
H BO
3
3
2
3
4
7
nD
= − D
ͩ
+
+ 2
ͪ
H BO
H BO
3 3
3
3
ץx
ץx
ץx
ץx
͓
25͔
−
1
where the total concentration of H BO ͑C = 0.62 mol L ͒ is as-
3
3
sumed constant at the cathode surface ͑x = 0͒ and in the bulk phase
x ജ ␦H ͒ as described previously.
The molarity-based stability constants of Ni ͑OH͒ ͑KN1͒ and
͑
+
4
4
+
4
+
28
NiOH ͑KN2͒ are listed below
4
+
1
/4Ni ͑OH͒ = Ni2+ + OH− KN1 = 10−7.17
͓26͔
͓27͔
4
4
NiOH = Ni + OH− KN2 = 10−4.14
+
2+
Figure 6. Concentration distribution of the soluble species from boric acid.
+
The concentration of NiOH is always small enough to be neglected
within the range of pH examined in this paper, which coincides
2
8
2
−
3−
with Murthy et al.
HBO = BO3 + H+ KHm = 10−13.80
͓16͔
͓17͔
͓18͔
͓19͔
͓20͔
͓21͔
͓22͔
͓23͔
3
The diffusion flux of 4D
contained in Eq. 8 includes CNi2+ through iNi as follows
4+͑ץC
4+/ץx͒x=0 potentially
4
Ni ͑OH͒
Ni ͑OH͒
4
4
4
4
H BO = H B O + 5H O KHm = 10−2.55
3
3
2
4
7
2
−
7
+
ץCNi2+
ץCNi ͑OH͒4+
H B O = HB O + H KHm = 10−4.00
iNi͑1 − t͒
4
4
2
4
7
4
=
ͯ
DNi2+
ͯ
+
ͯ
4D
4+
ͯ
Ni ͑OH͒4
4
2
F
ץx
ץx
x=0
x=0
−
7
+
4
H BO = HB O + H + 5H O KHm = 10−6.54
3 3 4 2
͓
28͔
HB O = B O + H+ KHm = 10−9.00
−
7
2−
7
4
4
and
4
2
7
−
+
CNi2+KW
4
H BO = B O + 2H + 5H O KHm = 10−15.55
3 3 4 2
C
4+
4
=
ͩ ͪ
͓29͔
Ni ͑OH͒
4
C K
+
H N1
B O + 5H O = 4H BO + 2H KHm = 10−21.31
2
−
−
3
+
4
7
2
2
The ionic flux due to the migration mechanism is approximated
Then the concentrations of H and Ni at x = 0
, C 2+͒ are given by the optimal solutions of Eq. 8 and 28, and
2
0-22
+
2+
B O + 5H O = 4HBO + 6H+ KHm = 10−72.11
2
7
−
2−
3
Ni
by i t/zF.
4
2
s
s
͑
C
+
H
Ni
KHm ͑m = 1, ... ,10͒ is the stability constant where the unit of
concentration is molarity. Figure 6 shows the concentration distribu-
tion of soluble species from H BO depending on the pH neglecting
the activity coefficients.
the mass balance equation of H BO . The diffusion coefficient of
each chemical species is listed in Table II. The diffusivity of
Ni ͑OH͒ was taken from the literature.
3
3
3
3
4+
4
28
4
s
Figure 7 shows the variations of the surface pH ͑pH ͒. When
The total concentration of H BO added in the electrolyte solu-
3
3
b
s
−
1
the bulk pH ͑pH ͒ is 1.5, the pH gradually increases with
tion is 0.62 mol L . It is clearly observed that H BO barely dis-
3
3
sociates up to about pH 5. Above this pH value, the concentration of
H BO starts to decrease significantly and HB O from Eq. 19
the H2 gas evolution rate and then abruptly jumps to 6 above
−
7
−2
iH = 12 mA cm like a typical pH titration curve.
3
3
4
2
b
s
gradually increases.
At pH 3.4, the pH increases immediately after i is generated,
H
2
−
3
2−
7
b
s
H BO is generated by Eq. 14 above pH 5.5. B O stemming
from Eq. 20 and 21 appears above pH 6.3. A constant concentration
of H B O is maintained up to pH 5. It decreases corresponding to
the behavior of H BO due to Eq. 17, though it does not have a
in contrast to the case of pH 1.5. The pH already reaches 6.0
even at iH2 = 1 mA cm . H BO must dissociate to a significant
extent near the cathode surface in such a case, according to Fig. 6.
At pH 5.5, three stages are noticed in the variation of the preferred
2
4
−
2
3 3
2
4
7
b
3
3
s
buffering function.
orientation in Fig. 5. The pH gradually increases and reaches 6.8
Consequently, the buffer effect of H BO3 must be considered
based on the simultaneous dissociating reactions. Thus, it needs to
be rewritten as the appropriate form for H BO including the several
buffer reactions. The primary reactions to be considered are Eq. 14,
3
at E = −1.4 V.
Figure 7 demonstrates that the transition to the ͕110͖ preferred
orientation is completed within the iH2 of few mA cm in all pH
−2
3
3
solutions.
1
9, and 21.
The suppression of the iNi at E = −1.6 V at pH 3.4 and 5.5 in
Fig. 2 is probably due to the shift of the ionic equilibrium at the
cathode surface to reduce the concentration of Ni and generate
Therefore, n can be written by
2
+
ץC
ץC
−
3
ץC
ץC
−
7
ץC
B O
4
2−
7
H BO
2
HB O
4
4+
4
2
OH
n = −
−
− 2
͓24͔
Ni ͑OH͒ or Ni͑OH͒ . The ionic products of C 2+ and C
−
at
4
2
Ni
ץC
H BO
3
H BO
3
H BO
−16
3
−3
3
3
3
3
x = 0 are calculated to be about 10 ͑mol L ͒, taking the activity
34
As a result
coefficients into consideration. This is equivalent to the solubility
Downloaded on 2015-03-17 to IP 132.174.255.116 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 ͑7͒ C502-C508 ͑2006͒
C507
Figure 7. Relation between electrode po-
s
tential and pH calculated electrode poten-
tial. ͓The bulk pH values are ͑A͒ 1.5, ͑B͒
.4, and ͑C͒ 5.5.͔
3
product of Ni͑OH͒ of 10−16–10
−14 16,28,33,37
.
Moreover, the conduc-
introduces the surface coverage concept quantitatively combined
with the surface concentrations of chemical species in the next step
of our research.
2
tivity variation due to the electrolyte composition may also partici-
pate in the phenomenon appearing in Fig. 2.
Figure 8 rearranges Fig. 5 by introducing the concept of the
s
surface pH value ͑pH ͒. The ͕311͖ orientation domain is shrunk
Conclusion
toward the higher pH region. It is found that the transition boundary
between the preferred orientations ͕100͖ and ͕110͖ is located about
00 mV below the H /H2 equilibrium potential line ͑␩H2
500 mV͒. Moreover, a ridge in the current efficiency contour map
is apparently running parallel to this transition boundary line. It is
reported that hydrogen overpotential is roughly 200 mV on a Ni
Ni metal was electrodeposited on the vertical plane cathode in
three different pH solutions. The polarization curve showed a clear
pH dependence. The current efficiency varied from 70 to 83% at
pH 1.5. H gas evolution might enhance the cathodic mass-transfer
rate of Ni ions significantly at pH 1.5. However, it was always
close to 100% and slightly depressed at potentials more negative
than E = −1.4 V at pH 3.4 and 5.5.
+
5
=
2
2+
44,45
metal substrate.
A significant rate of iH2 thus introduces a tre-
mendous amount of dissolved H atoms ͑into Ni metal͒. This deduc-
tion reaction probably results in the transition from ͕100͖ to ͕110͖,
which may be due to the residual strain around the interstitial H
atoms. This appears to support the postulate that the texture evolu-
tion phenomena of the transition metals are partly influenced by the
In the low-overpotential range, the preferred crystal orientation
of electrodeposited Ni films varied depending on the pH. The ͕110͖
preferred orientation only appeared in the more negative potential
range, regardless of pH.
s
The surface pH at the cathode ͑pH ͒ was then calculated using
H gas evolution as the competitive reaction on the depositing Ni
2
the measured partial current density for H gas evolution based on a
2
metal cathode surface.
one-dimensional mass-transfer model that included the buffer ac-
B
Amblard et al. ascribed the ͓110͔ texture to both adsorbed spe-
s
tion. The calculated pH jumped abruptly above 6 as i increased.
H
8
-11
2
cies of H molecule and Ni͑OH͒ ,
but their discussion was es-
2
2
The transition of the preferred orientation was plotted with a
sentially obscure. Another implicated mechanism caused by nickel
s
current efficiency contour line in the E–pH map. The transition
13
adatoms was also proposed. The soundness of such deductions is
further beyond the present characterization technique. The present
contribution to this problem is to introduce the surface pH concept
in comparison with the past qualitative discussion. The kinetic
model on the competitive reaction during Ni electrocrystallization
boundary of the preferred orientation from ͕100͖ to ͕110͖ was lo-
cated almost parallel with the equilibrium H2 gas evolution line.
This behavior may support the postulate that the dissolved hydrogen
atoms in Ni metal are partly responsible for controlling the texture
structure evolution.
Acknowledgments
The authors thank Dr. F. McLarnon, Dr. S. Nakahara, and Dr. S.
Kikuchi for their valuable discussions throughout the present works.
Part of this work was carried out under financial assistance given to
Y.F. by the Ministry of Education, Science, Sports, and Culture
͑
project no. 15360402͒, for which the authors are grateful.
Kyoto University assisted in meeting the publication costs of this article.
Appendix A
The Rayleigh number can be written by
g␣cy3
Ray =
͓A-1͔
␯
D
where g, ␣, c, y, ␯, and D are gravitational acceleration ͑cm s−2͒, densification coeffi-
3
−1
−1
cient ͑cm mol ͒, bulk concentration ͑mol L ͒, vertical distance from lower edge
2
−1
2
−1
Figure 8. Distributions of the current efficiency and the preferred orientation
͑cm͒, kinematic viscosity ͑cm s ͒, and diffusivity ͑cm s ͒, respectively. The Sher-
wood number was
s
in the E–pH diagram.
Downloaded on 2015-03-17 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C508
Journal of The Electrochemical Society, 153 ͑7͒ C502-C508 ͑2006͒
5
. X. Liu and G. Zangari, IEEE Trans. Magn., 37, 1764 ͑2001͒.
iy͑1 − t͒y
zFD⌰
Shy =
͓A-2͔
6. M. Froment and G. Maurin, J. Microsc., 7, 39 ͑1968͒.
7
. M. Froment and J. Thevenin, Electrochim. Acta, 20, 877 ͑1975͒.
where iy, t, z, F, and ⌰ are local current density at y ͑A cm−2͒, transference number ͑-͒,
8. J. Amblard, M. Froment, and N. Spyrellis, Surf. Technol., 5, 205 ͑1977͒.
9. J. Amblard, I. Epelboin, M. Froment, and G. Maurin, J. Appl. Electrochem., 9, 233
͑1979͒.
10. J. Amblard, M. Froment, G. Maurin, N. Spyrellis, and E. T.-Souteyrand, Electro-
chim. Acta, 28, 909 ͑1983͒.
−
1
valency number ͑-͒, Faraday constant ͑C mol ͒, and difference of the surface concen-
−
1
tration from the bulk concentration ͑mol L ͒, respectively. The ⌰ at a given iy can be
evaluated by using the correlation equation of Eq. 1.
Appendix B
11. C. Kollia and N. Spyrellis, Surf. Coat. Technol., 57, 71 ͑1993͒.
1
1
1
2. E. Gómez, R. Pollina, and E. Vallés, J. Electroanal. Chem., 386, 45 ͑1995͒.
3. H. S. Karayannis and G. Patermarakis, Electrochim. Acta, 40, 1079 ͑1995͒.
4. A. M. El-Sherik, U. Erb, and J. Page, Surf. Coat. Technol., 88, 70 ͑1996͒.
The cathodic mass-transfer rate is enhanced by H2 gas bubble evolution. Fukunaka
et al. measured the variations of the mass-transfer coefficient, rising velocity of gas
bubbles, and thickness of bubble dispersion layer along a vertical direction with current
3
1
15. S. Nakahara and E. C. Felder, J. Electrochem. Soc., 129, 45 ͑1982͒.
6. H. Dahms and I. M. Croll, J. Electrochem. Soc., 112, 771 ͑1965͒.
density. The measured ionic mass-transfer coefficient was analyzed by applying the
3
0
1
additivity rule of micro- and macromixing proposed by Alkire and Lu. Micromixing is
induced by the microconvective flow of electrolyte during bubble growth and detach-
ment on the cathode surface. Macromixing refers to the turbulent natural convection
induced by the density difference between the electrolyte at the cathode surface and the
bulk electrolyte. The diffusion layer thickness is given by
17. H. Deligianni and L. T. Romankiw, IBM J. Res. Dev., 37, 85 ͑1993͒.
18. H. Fischer, Angew. Chem., Int. Ed. Engl., 8, 108 ͑1969͒.
19. R. Winand, Trans. Inst. Min. Metall., Sect. C, 84, 67 ͑1975͒.
20. C. Wagner, J. Electrochem. Soc., 95, 161 ͑1949͒.
2
1. C. R. Wilke, M. Eisenberg, and C. W. Tobias, J. Electrochem. Soc., 100, 513
1953͒.
22. Y. Fukunaka, T. Minegishi, N. Nishioka, and Y. Kondo, J. Electrochem. Soc., 128,
274 ͑1981͒.
3. A. H. DuRose, Plat. Surf. Finish., 48 ͑1977͒.
4. S. Alamelu and C. V. Suryanarayana, Acta Chim. Acad. Sci. Hung., 20, 7 ͑1959͒.
5. R. Wu, M. Oliazadeh and A. M. Alfantazi, J. Appl. Electrochem., 33, 1043 ͑2003͒.
DNi2+
͑
␦
Ni2+
=
͓B-1͔
km
1
where km = kb + kln; kb and kln are the mass-transfer coefficient due to micromixing and
the mass-transfer coefficient under laminar natural convection, respectively. The latter is
assumed to be governed by the void fraction during H2 evolution.
Within the range of iH2 examined in this study, the possibility of introducing tur-
bulent natural convection is discarded because of log Rax ഛ 5.5. The ratio of the diffu-
2
2
2
26. K. Denpo, S. Teruta, Y. Fukunaka, and Y. Kondo, Metall. Trans. B,
14B, 633
͑1983͒.
+
2+
2+
sion layer thickness of H to that of Ni is approximated to the case of Cu
2
2
7. C. C. Streinz, A. P. Hartman, S. Motupally, and J. W. Weidner, J. Electrochem.
Soc., 142, 1084 ͑1995͒.
8. M. Murthy, G. S. Nagarajan, J. W. Weidner, and J. W. Van Zee, J. Electrochem.
␦
H+ = ␴␦Ni2+
1,46,47
͓B-2͔
These equations are used to estimate
3
where ␴ denotes the ratio ␦H+/␦Ni2+ = 1.25.
the diffusion layer thickness of the ions.
Soc., 143, 2319 ͑1996͒.
2
3
3
9. V. A. Ettel, B. V. Tilak, and A. S. Gendron, J. Electrochem. Soc., 121, 867 ͑1974͒.
0. R. Alkire and P.-Y. Lu, J. Electrochem. Soc., 126, 2118 ͑1979͒.
1. Y. Fukunaka, K. Suzuki, A. Ueda, and Y. Kondo, J. Electrochem. Soc., 136, 1002
͑1989͒.
Appendix C
Adding sulfuric acid of 2 ϫ 10 ͑mol L−1͒ in a Watts bath reduces the pH by 1.5.
−
2
+
−1.5
−3.4
−2
−1
32. K. S. Willson and J. A. Rogers, Tech. Proc. Am. Electroplaters Soc., 51, 92 ͑1964͒.
33. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions, Pergamon
Press, London ͑1966͒.
34. M. El Guendouzi, A. Mounir, and A. Dinane, J. Chem. Thermodyn., 35, 209
͑2003͒.
Then the H concentration increment is 10
− 10
= 3.1 ϫ 10 ͑mol L ͒. As-
+
suming that the H concentration increment contributed by
H2SO4 = HSO −4 + H+
͓C-1͔
is 2 ϫ 10−2 ͑mol L−1͒, the additional amount of 1.1 ϫ 10−2 ͑mol L−1͒ must be com-
3
3
5. M. Matlosz, J. Electrochem. Soc., 140, 2272 ͑1993͒.
6. Y. Fukunaka, S. Aikawa, and Z. Asaki, J. Electrochem. Soc., 141, 1783 ͑1994͒.
pensated as follows
HSO −4 = SO 24 − + H+
Boric acid does not release H ions in this pH region according to Fig. 6. Finally, the
͓C-2͔
37. N. Zech and D. Landolt, Electrochim. Acta, 45, 3461 ͑2000͒.
3
3
8. J. Horkans, J. Electrochem. Soc., 126, 1861 ͑1979͒.
9. J. P. Hoare, J. Electrochem. Soc., 133, 2491 ͑1986͒.
+
−
−1
40. J. P. Hoare, J. Electrochem. Soc., 134, 3102 ͑1987͒.
4
4
1. M. Y. Abyaneh and M. Hashemi-Pour, Trans. Inst. Met. Finish., 72, 23 ͑1993͒.
2. S. Hessami and C. W. Tobias, J. Electrochem. Soc., 136, 3611 ͑1989͒.
References
1
2
. N. V. Myung, D.-Y. Park, B.-Y. Yoo, and P. T. A. Sumodjo, J. Magn. Magn. Mater.,
65, 189 ͑2003͒.
. G. Gubbiotti, G. Carlotti, S. Tacchi, Y.-K. Liu, C. Scheck, R. Schad, and G. Zan-
43. J. K. Park and K. J. Lee, J. Chem. Eng. Data, 39, 891 ͑1994͒.
44. H. Ezaki, M. Morinaga, and S. Watanabe, Electrochim. Acta, 38, 557 ͑1993͒.
45. J. Tamm, L. Tamm, and P. Vares, Russ. J. Electrochem., 36, 1174 ͑2000͒.
46. Y. Fukunaka, K. Denpo, M. Iwata, K. Maruoka, and Y. Kondo, J. Electrochem.
Soc., 130, 2492 ͑1983͒.
2
gari, J. Appl. Phys., 10, 10J102 ͑2005͒.
3
4
. X. Liu, P. Evans, and G. Zangari, IEEE Trans. Magn., 36, 3479 ͑2000͒.
. K. P. Larsen, A. A. Rasmussen, J. T. Ravnkilde, M. Ginnerup, and O. Hansen, Sens.
Actuators, A, 103, 156 ͑2003͒.
47. K. Denpo, T. Okumura, Y. Fukunaka, and Y. Kondo, J. Electrochem. Soc., 132,
1145 ͑1985͒.
Downloaded on 2015-03-17 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).