The inhibition of anomalous codeposition of iron-group alloys using cyclic voltammetry

Article abstract of DOI:10.1149/1.1511753

Binary iron-group deposits, including Co-Ni, Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni alloys, with composition equal to that of the plating baths, were electroplated using cyclic voltammetry (CV) in suitable potential regions. Three processes, corresponding to cath

Full text of DOI:10.1149/1.1511753

Journal of The Electrochemical Society, 149 ͑11͒ C615-C622 ͑2002͒
C615
0013-4651/2002/149͑11͒/C615/8/$7.00 © The Electrochemical Society, Inc.
The Inhibition of Anomalous Codeposition of Iron-Group
Alloys Using Cyclic Voltammetry
,z
*
**
Chi-Chang Hu and Allen Bai
Department of Chemical Engineering, National Chung Cheng University, Chia-Yi 621, Taiwan
Binary iron-group deposits, including Co-Ni, Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni alloys, with composition equal to that of the plating
baths, were electroplated using cyclic voltammetry ͑CV͒ in suitable potential regions. Three processes, corresponding to cathodic
deposition, double-layer response, and anodic dissolution, were found on the CV curves of deposition. The dissolution of newly
deposited metal atoms caused the codissolution of an adsorbed monohydroxide layer in the anodic dissolution process, resulting
in the inhibition of anomalous codeposition for these iron-group alloys. This proposal was further confirmed by an electrochemical
impedance spectroscopic study. The anodic stripping of linear sweep voltammetry and the X-ray diffraction patterns demonstrated
the poor-crystalline or amorphous nature of Zn-Ni alloys in different phases.
© 2002 The Electrochemical Society. ͓DOI: 10.1149/1.1511753͔ All rights reserved.
Manuscript submitted January 16, 2002; revised manuscript received May 6, 2002. Available electronically October 9, 2002.
Nickel, cobalt, iron ͑i.e., iron-group metals͒, and their alloys are
important engineering materials in many applications because of
their unique magnetic, corrosion-resistance, wear-resistance, and
electrocatalytic properties.1-10 These functional materials are usually
prepared by electroplating because this technique is often much
cheaper and simpler than other methods. The physicochemical prop-
erties of these materials are seriously affected by their composition
and structure,2,6,9-13 thus, reliable control of these properties for the
iron-group alloys is important for their wide applications.
The electroplating of iron-group binary alloys ͑e.g., Co-Ni,
Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni͒ is generally recognized as a
codeposition of the anomalous type5,6,11-20 because the composition
of the more active metals within the deposits is higher than that in
the plating baths. Several models reasonably and quantitatively de-
scribe these anomalous phenomena. The formation of Fe(OH)2 due
to the hydrogen evolution side reaction was proposed to depress Ni
deposition for the Fe-Ni alloys.15 The preferred adsorption of more
active metal monohydroxides ͓denoted as M(OH)ϩ] on the deposits,
inhibiting the adsorption of less active metal ions, was also proposed
to account for this anomaly.12,14,15,18,21-25 Moreover, in plating baths
with a constant pH, the sequence of metals with respect to increas-
ing the equilibrium concentration of M(OH)ϩ was expected to be:
the literature, mechanisms of the anomalous codeposition are not
well understood. Also, the compositions and structures of these in-
dustrial alloys are not controllable.
It is desirable for the deposited alloys to have the same compo-
sition as the corresponding precursor baths because the concentra-
tions of metal ions in the plating baths are controllable. This require-
ment can be met by the equilibrium codeposition14 because having
the alloys in chemical equilibrium with the baths should result in
electroplated alloys with compositions equal to that of the baths.
However, the electroplating of iron-group alloys typically exhibits
anomalous codeposition. To circumvent this anomaly, these alloys
can be normally electroplated from their molten salt systems.31 But
the procedures for this kind of deposition are more complicated, and
control of the deposition conditions is difficult. Recently, pulse
and pulse-reverse electroplating reduced the anomaly of these
alloys,32-35 but the compositions of these alloys are still difficult to
predict from the compositions of the plating baths.
In this study, a unique method ͓cyclic voltammetry ͑CV͔͒ was
proposed to deposit the Co-Ni, Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni
alloys with compositions approximately equal to that of their corre-
sponding acidic chloride baths, although the potential cycling
method is not normally employed in electroplating. Reasonable ex-
planations responsible for this novel phenomenon as well as the
mechanisms of anomalous codeposition were further clarified. Some
important information for the pulse-reverse plating of iron-group
alloys with compositions equal to that of the plating baths has been
deduced from this work. This information is important in developing
practical and suitable conditions for electroplating iron-group alloys
with a controllable composition. Note that no additive was used in
the chloride baths because additives could be codeposited in the
alloys, seriously affecting their physicochemical properties.2,11,36
The physicochemical properties ͑e.g., crystalline, magnetic, anticor-
rosive, morphologic, and electrocatalytic properties͒ of these alloys
are probably different from those prepared by dc, pulse, or pulse-
reverse electroplating ͑these are to be systematically investigated in
our next study͒.
Ͼ
Ͼ
Ͼ
Co Ni, resulting in the anomalous codeposition of
Zn
Fe
iron-group alloys.26 A model involving two-step reaction mecha-
nisms of adsorbed monovalent intermediate ions ͑i.e., competitive
adsorption of Mϩ) for the electrodeposition of Fe and Ni as single
metals was combined to form a predictive model for the codeposi-
tion of Fe-Ni alloys.27 There is also a model considering the diffu-
sion effect coupled with the homogeneous dissociation reactions of
water and monohydroxides describing the anomalous codeposition
of iron-group alloys.23 Recently, the codeposition of Co-Ni and
Fe-Ni showed more mass-transfer effects than the Fe-Co codeposi-
tion although M(OH)ϩ was also assumed to be the important
charge-transfer species.28 The effects of bath pH and complexing
agents on the composition of Fe-Ni deposits were examined; the
deposition became less anomalous when the pH of the plating bath
was increased from 3 to 5.29 More recently, the incorporation of
oxygen within Fe-Ni deposits was attributed to the precipitation and
occlusion of Ni(OH)2 in the growing deposit.30 A model considering
the adsorption and diffusion effects of M(OH)ϩ was proposed to be
extendable to Fe-Ni, Co-Ni, and Fe-Co alloys.26 Although many
models ͑including predictive and supportive models͒ are proposed in
Experimental
The Co-Ni, Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni deposits were elec-
troplated on commercial pure ͑99.5%͒ 1 ϫ 2 cm Cu plates. These
Cu plates were first cleaned with trichloroethylene, rinsed with pure
water, and then anodized at 40 mA cmϪ2 in a 0.1 M NaOH solution
for 10 min. After anodizing, these plates were cathodically polarized
at 40 mA cmϪ2 in another 0.1 M NaOH solution for 1 min, vibrated
in an ultrasonic bath for 5 min, acid-cleaned with 0.1 M HCl for 2
min, and rinsed with pure water. The exposed geometric area of
these pretreated Cu substrates was equal to 1 cm2. The other sur-
faces were insulated with a poly͑tetrafluoroethylene͒ coating before
codeposition. The deposition baths contained the metal chlorides in
specified ratios ͑i.e., 10:0; 8:2; 6:4; 4:6; 2:8; and 0:10). The total
* Electrochemical Society Active Member.
** Electrochemical Society Student Member.
z E-mail: chmhcc@ccu.edu.tw
Downloaded on 2015-01-15 to IP 132.239.1.231 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C616
Journal of The Electrochemical Society, 149 ͑11͒ C615-C622 ͑2002͒
concentration of metal ions was kept constant at 0.3 M in the plating
bath with a pH of 2 ͑adjusted with 1 M HCl͒. The Cu substrates
were plated with the alloys using CV in different potential ranges
͑from 0 to Ϫ1.2, from Ϫ0.2 to Ϫ1.4, from Ϫ0.2 to Ϫ1.4, from Ϫ0.6
to Ϫ1.6, and from Ϫ0.55 to Ϫ1.6 V for Co-Ni, Fe-Co, Fe-Ni, Zn-
Fe, and Zn-Ni, respectively͒. The electroplating was interrupted
when 30 CV cycles had been performed.
CV for the electroplating of alloys was performed with an elec-
trochemical analyzer system, CHI 633A ͑CH Instruments, USA͒.
The CV scan rate was kept at 20 mV sϪ1. All experiments were
carried out in a three-compartment cell. An Ag/AgCl electrode ͓Ar-
genthal, 3 M KCl, 0.207 V vs. a standard hydrogen electrode at
25°C͔ was used as the reference, and a piece of platinum gauze with
an exposed area of 4 cm2 was employed as the counter electrode. A
Luggin capillary, whose tip was set at 2 mm from the surface of the
working electrode, was used to minimize errors due to iR drop in the
electrolytes.
The average composition of all deposits was measured using an
energy dispersive X-ray analysis ͑EDX͒ spectroscope with standards
at three points coupled with a scanning electron microscope ͑JEOL
JSM35͒. The EDX data were also confirmed by composition data
obtained with a glow discharge spectrometer ͑GDS-750A, LECO,
USA͒. To avoid oxygen adsorption on the deposits from the air, the
electroplated samples were rinsed with pure water and then dried
and kept in a vacuum oven (Ͻ10Ϫ3 bar) at room temperature be-
fore the EDX and GDS analysis. The X-ray diffraction ͑XRD͒ pat-
terns were measured at 4° minϪ1 through a Rigaku X-ray diffrac-
tometer using a Cu target. All solutions were prepared with pure
water produced by a reagent water system ͑Milli-Q SP, Japan͒ at 18
M⍀ cm. All reagents not otherwise specified are Merck, G.R. The
solution temperature was maintained at 25°C with an accuracy of
0.1°C by using a water thermostat ͑Haake DC3 and K20͒.
Results and Discussion
CV.—The cyclic voltammograms of the first, third, and 30th
cycles for Co-Ni codeposition, where Co:Ni ϭ 8:2 in the plating
bath, are represented in Fig. 1a-c, respectively. On the negative
sweep ͑the dashed line͒ in Fig. 1a, minor anodic currents occurred at
potentials positive of ca. Ϫ0.1 V, probably due to the oxidation/
dissolution of the Cu substrate. When potentials were located be-
tween Ϫ0.1 and Ϫ0.8 V, double-layer responses occurred. When
the potentials were negative of Ϫ0.8 V, cathodic currents due to the
codeposition of Co-Ni sharply increased with the negative shift in
potential. On the following positive sweep, cathodic codeposition of
Co-Ni also occurred in a similar potential region, while the deposi-
tion rate ͑cathodic currents͒ on the positive sweep was somewhat
more rapid than that on the negative sweep at potentials between
Ϫ1.0 and Ϫ0.75 V. This phenomenon can be attributed to the sig-
nificant effect of Co-Ni nucleation on the negative sweep.37,38 An-
odic currents presumably due to the dissolution of Co-Ni com-
menced at ca. Ϫ0.3 V on the positive sweep, which is more
negative than that of Cu oxidation/dissolution. From a comparison
Figure 1. Cyclic voltammograms for Co-Ni codeposition measured at
20 mV sϪ1 in a deposition bath containing 0.24 M CoCl2 and 0.06 M NiCl2
with initial pH of 2 at 25°C. ͑a͒ First, ͑b͒ third, and ͑c͒ thirtieth cycles.
Co-Ni alloys prepared by a dc method ͑data from Ref. 14͒ on the
composition of the plating baths for purposes of comparison. Note
that the Co-Ni alloys prepared by CV without the anodic dissolution
region are less anomalous than those fabricated by using dc electro-
plating, where an anomaly associated with the Co-Ni codeposition is
still obvious. This depression in anomaly may be attributed to a
smaller diffusion effect of metal ions in the potential cycling
deposition23,27,28 because the electrodeposition of Co-Ni alloys
within the potential range of the cathodic deposition process on the
CV curves passes the electron-transfer and mass-transfer control re-
gions. However, the composition of the Co-Ni alloys deposited by
potential cycling with the anodic dissolution region is very close to
that of the plating baths, demonstrating that the anodic process effi-
ciently depresses the anomaly of the Co-Ni codeposition ͑see curve
2͒. The preferred adsorption of more active metal monohydroxides
͓i.e., M(OH)ϩ] on the deposits is widely recognized to be respon-
sible for the anomalous codeposition of iron-group alloys; thus, the
anodic process in the CV curves is expected to damage the adsorbed
layer of M(OH)ϩ. Note the low pH and high ClϪ concentration of
Ͻ
of Fig. 1a-c, three processes, cathodic deposition (E
Ϫ0.75 V),
Ϫ0.3 V), and anodic dis-
Ϫ0.3 V), are clearly seen on the CV curves for the
Ͻ
Ͻ
E
double-layer response (Ϫ0.75 V
Ͼ
solution (E
Ni-Co codeposition. Also, the anodic dissolution of Co-Ni alloys at
Ͼ
E
Ϫ0.3 V increased somewhat with an increase in the number of
deposition cycles, as indicated by the gradual increase in anodic
charge integrated from both positive and negative sweeps of these
CV curves. Because the cathodic charge on every CV curve is much
larger than its corresponding anodic charge, Co-Ni alloys can be
continuously deposited on the substrate using CV.
The dependence of the Co content in the Co-Ni alloys, prepared
by the CV method without the anodic dissolution region ͑the upper
potential limit of CV, EU ϭ Ϫ0.4 V, labeled as छ͒ and CV with
the anodic dissolution region (EU ϭ 0 V, labeled as ᭝͒, on the
composition of the plating baths is shown in Fig. 2. In addition,
curve 1 in Fig. 2 represents the dependence of the Co content in
Downloaded on 2015-01-15 to IP 132.239.1.231 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, 149 ͑11͒ C615-C622 ͑2002͒
C617
Figure 2. The Co content ͑Co atom %͒ of Co-Ni deposits vs. the composi-
tion of Co2ϩ in the deposition baths; the Co-Ni alloys were prepared by ͑1,
͒ a dc method ͑see Ref. 14͒; ͑2, ¯᭝¯͒ CV between Ϫ1.2 and 0 V;
and ͑3, --छ--͒ CV between Ϫ1.2 and 0.4 V at 20 mV sϪ1
.
the plating solutions. The Co-Ni alloys should be readily dissolved
during an anodic process in such media. Accordingly, the anodic
process inhibiting the anomalous codeposition of iron-group alloys
is reasonably attributed to the dissolution of the newly deposited
metals, rendering the codissolution of the adsorbed M(OH)ϩ layer.
To support this concept, the codeposition of Fe-Co, Fe-Ni, Zn-
Fe, and Zn-Ni alloys was also carried out under similar conditions.
Typical cyclic voltammograms with the anodic dissolution region
for electroplating these alloys are shown in Fig. 3. Although the CV
potential windows for depositing these alloys are different from each
other, these CV curves generally exhibit similar features to those in
Fig. 1 ͑with the exception of Zn-Fe codeposition͒; three processes
corresponding to cathodic deposition, double-layer response, and
anodic dissolution are clearly seen. Thus, the discussion for Fig. 1
and 2 should be applicable for the electroplating of these alloys. The
results for composition for all the alloys are shown in Table I. The
average thickness of these deposits ranged from ca. 7.2 to 15.0 ␮m,
and these films are of uniform composition because no outlier is
found in the composition measurements. The first column in this
table represents the ion concentration in atomic percentage ͑atom %͒
of the more active metals in the plating baths. The remaining col-
umns represent the metal content of the more active metals in the
alloys. An examination of the data in Table I shows that the absolute
deviation of metal content from the compositions of the plating
baths is generally smaller than or equal to 6 atom % for all alloys.
These results confirm the proposal that the anodic process in the
CVs inhibited the anomalous codeposition of iron-group alloys.
The preceding results unearth facile changes in the composition
of the Co-Ni, Fe-Co, Fe-Ni, Zn-Fe, and Zn-Ni alloys wrought by the
CV potential region, indicating that the composition of the alloys
should be linked to the potential window of CV. To gain an under-
standing of this relationship, the composition of Co-Ni alloys depos-
ited in different potential ranges and the number of sweeps of the
CV ͑see the E Ϫ t diagrams in Fig. 4͒ were measured, and the
results are shown in Table II. Note that at the end of period 1,
neither Co nor Ni was found on the substrate. In period 2, nucleation
and growth of Co-Ni alloys occurred, and the Co/Ni ratio in the fifth
column of Table II was equal to 8.2, showing the anomalous prop-
erty. At the end of period 3, the loading of the Co-Ni alloy obviously
increased because the clear decrease in Cu content in the fourth
column indicates an increase in coverage of the Co-Ni alloy on the
substrate. At the end of period 4, a significant amount of the Co-Ni
deposit was dissolved during this positive sweep region ͑i.e., the Cu
content increased significantly͒, and the Co/Ni ratio of this deposit
Figure 3. The final cycle of cyclic voltammograms of iron-group alloy
codeposition in their chloride baths with initial pH of 2: ͑a͒ Fe-Co, ͑b͒ Fe-Ni,
͑c͒ Zn-Fe, and ͑d͒ Zn-Ni alloys. CV curves were measured at 20 mV sϪ1
and 25°C.
Table I. Composition of deposition baths and Co-Ni, Fe-Ni,
Fe-Co, Zn-Fe, and Zn-Ni alloys.
a
M ͑atom % in alloys͒
M2ϩ ͑atom % in
a
plating baths͒
Zn-Ni
Co-Ni
Fe-Ni
Fe-Co
Zn-Fe
0
20
0
19
0
23
0
25
0
23
0
23
40
43
42
46
42
40
60
63
62
66
61
65
80
83
80
85
80
82
100
100
100
100
100
100
a M2ϩ and M are indicative of the more active metals.
Downloaded on 2015-01-15 to IP 132.239.1.231 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C618
Journal of The Electrochemical Society, 149 ͑11͒ C615-C622 ͑2002͒
Also note that from a comparison of the Cu content in the fourth
column, the loading of Co-Ni codeposits on the substrate gradually
increased with continuous potential cycling. This result reveals that
the rate of dissolution of Co-Ni alloys in the anodic dissolution
range is slower than that of the deposition in the cathodic plating
region, which is consistent with the CV results in Fig. 1.
From the preceding results and discussion, a reasonable reaction
scheme for the electroplating of iron-group alloys by means of CV is
proposed in Fig. 5, where M represents Zn, Fe, Co, and Ni. In the
cathodic deposition region on the negative sweep, most metal ions
can be ‘‘normally’’ plated on the deposit through an adsorbed inter-
mediate ͑i.e., M2ϩ → Maϩds ϩ e; see position B in Fig. 5a and b͒ of
different metal ions.27 When potentials are more negative, the diffu-
sion effect of different metal ions becomes more significant. How-
ever, due to the side reaction of hydrogen evolution, the newly
formed OHϪ near the interface between the deposit and electrolyte
should favor the formation of M͑OH͒ϩ.12,14,15,18,21-26,28-30 This ef-
ϩ
fect enhances the adsorption of M͑OH͒ with a less noble property
ϩ
͓i.e., M͑OH͒ads], favoring the anomalous codeposition of iron-group
alloys ͑see position A in Fig. 5b͒. Because the degree of anomaly is
Figure 4. The E Ϫ t plots for the first three cycles of Co-Ni codeposition in
a plating bath containing 0.24 M CoCl2 and 0.06 M NiCl2 with initial pH of
2.
ϩ
believed to be directly proportional to the coverage of M͑OH͒
,
ads
anomalous codeposition of iron-group alloys may significantly oc-
cur when the potentials are located in the cathodic deposition region
on the positive sweep ͑Fig. 5b͒. However, this anomalous phenom-
enon is depressed by the simultaneous dissolution of newly depos-
was 4.9, reducing the anomaly of codeposition. These results are
consistent with the voltammetric curves shown in Fig. 1, supporting
the proposal that the anodic reaction is a dissolution process, which
caused the simultaneous dissolution of the newly deposited Co-Ni
atoms and the adsorbed M(OH)ϩ layer. Because nucleation, growth,
and dissolution of Co-Ni alloys occurred during the first cycle, the
procedure of anomalous codeposition may not be well established in
this cycle. However, from the composition variation of the deposit
during the second and third cycles ͑see periods 5-8͒, mechanisms for
the potential cycling codeposition of iron-group alloys with compo-
sitions close to those of the plating baths can be reasonably ex-
plained by the following. First, from a comparison of the Co/Ni ratio
between the ends of periods 6 and 7, the anomaly is obviously lower
at the end of a negative sweep ͑period 7͒ in comparison with the
corresponding previous positive sweep ͑period 6͒. This phenomenon
ϩ
ited metals and M͑OH͒ads occurring in the anodic dissolution region
on both positive and negative sweeps ͑Fig. 5c͒. Moreover, the more
active metals should be more readily dissolved in the anodic disso-
lution process, which may also depress the anomalous codeposition.
These three processes are repeated during the potential cycling pro-
cess, and the anomalous codeposition of iron-group alloys is
inhibited.
ϩ
ads
In this work, the model involving the M͑OH͒ intermediate is
preferred for the following reasons. (i) Several hydroxides, e.g.,
Ni͑OH͒2 , Co͑OH͒2 , can be ‘‘cathodically’’ deposited on the sub-
strate due to the hydrogen evolution reaction.39,40 Thus, the forma-
ϩ
ads
tion of M͑OH͒ or even M͑OH͒ intermediates during the cathodic
2
deposition of the iron-group alloys is reasonable. (ii) In Fig. 2, the
anomaly of Co-Ni alloys prepared by CV without the anodic disso-
lution region is still clearly found, while this effect can be com-
pletely depressed by the anodic dissolution process. The two-step
model involving the preferred adsorption of more active metal
ions27 does not seem to reflect this anodic effect. (iii) The oxygen
content within the iron-group alloys changed from Ͻ5 atom % to
ca. 8-11 atom % when the pH was increased from 2.0 to 3.0, which
has been confirmed by the GDS analysis. This result further supports
ϩ
is attributable to the damage of the adsorbed M͑OH͒ layer in the
anodic dissolution process on the positive and following negative
sweeps, favoring the ‘‘normal’’ codeposition of Co-Ni alloys in the
cathodic deposition region on the negative sweeps of CV. Second,
when normal Co-Ni deposition occurred, the process corresponding
to the anomalous codeposition ͓the formation and adsorption of
M͑OH͒ϩ] would commence simultaneously. Therefore, in the ca-
thodic deposition region on the positive sweep of the CV, anomalous
codeposition became significant, inhibiting the continuous deposi-
tion of the more noble metals ͑e.g., Ni for Co-Ni alloys͒. This point
of view is also supported from a comparison of the Co/Ni ratio
between the ends of periods 5 and 6 ͑or between periods 7 and 8͒.
ϩ
ads
the formation of M͑OH͒ or even M͑OH͒ intermediates due to the
2
localized increase in OHϪ during the cathodic deposition of the
iron-group alloys. Therefore, the reaction scheme proposed in Fig. 5
is believed to be reasonable and acceptable.
Electrochemical impedance spectra.—The formation of an ad-
sorbed layer consisting of monohydroxides is expected to be de-
tected in electrochemical impedance spectroscopic ͑EIS͒ spectra be-
cause this technique provides a kinetic model ͑i.e., the equivalent
circuit͒, allowing one to efficiently attribute a clear physical mean-
ing to the parameters of interest. Thus, this method is believed to be
a good technique for decoupling the complicated adsorption and
deposition processes in the anomalous codeposition. Note that if the
dc potentials of the electrode were kept too negative during the EIS
measurements, the values of parameters for the adsorbed layer
should change significantly with measuring time because a more
complete adsorbed layer is expected to form with increasing depo-
sition time. Thus, the EIS spectra were measured at Ϫ0.6 V for the
deposits after cathodic deposition at Ϫ1.1 V for 5, 10, 20, and 40
min and then under the open-circuit state for 30 s. To further confirm
the proposed mechanism inhibiting the anomalous codeposition of
iron-group alloys, an EIS spectrum was measured at Ϫ0.6 V after
Table II. Composition of Co-Ni alloys prepared by CV in a depo-
sition bath containing 0.24 M CoCl2 and 0.06 M NiCl2 with initial
pH of 2. The number in the first column indicates the deposition
periods as shown in Fig. 4.
Composition in alloys ͑atom %͒
Step
Co
Ni
Cu
Co/Ni
1
2
3
4
5
6
7
8
0
0
100
84.3
—
8.2
4.2
4.9
4.7
10.5
4.4
5.2
14.0
30.5
14.7
34.4
37.7
38.8
56.3
1.7
7.2
3.0
7.3
3.6
8.9
10.8
62.3
82.3
58.3
58.7
52.3
32.9
Downloaded on 2015-01-15 to IP 132.239.1.231 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, 149 ͑11͒ C615-C622 ͑2002͒
C619
Figure 5. The deposition scheme of iron-group alloys using CV.
the deposit cathodically deposited at Ϫ1.1 V for 40 min was anodi-
cally oxidized at Ϫ0.1 V for 2 min and then under the open-circuit
state for 30 s. These experimental impedance data measured from
the plating solution are shown as Nyquist diagrams ͑in the complex
plane͒ in Fig. 6 as curves 1 ͑᭢͒, 2 ͑᭝͒, 3 ͑᭹͒, 4 ͑छ͒, and 5 ͑᭺͒.
In Fig. 6b, an impedance arc of ca. one-fifth circle is found in the
high frequency region on curves 1-4. The impedance of this high
frequency process is slightly decreased with the polarization time at
Ϫ1.1 V. This small impedance arc becomes clearer with increasing
the polarization time at Ϫ1.1 V. This arc is attributable to the
double-layer process because this process should be influenced by
the thickness and/or coverage of the adsorbed monohydroxide layer,
which is expected to increase with the polarization time of dc plat-
ing. However, in the low frequency region, there is an obvious im-
pedance arc, and a more obvious decrease in impedance ͑both real
and imaginary parts͒ was found by prolonging the deposition time of
dc electroplating. This result demonstrates that the low frequency
impedance arc is predominately governed by the adsorbed monohy-
droxide layer, probably due to the charge-transfer ͑faradaic͒ process
of the metal ions within the adsorbed layer because the barrier for
electron transfer for the adsorbed cations should be lower than that
for free solvated cations. If this is true, the formation of the adsorbed
monohydroxide layer should promote subsequent metal deposition.
Note that the adsorbed layer mainly consists of the more active
metal monohydroxides,12,14,15,18,21-26,28-30 which favors the adsorp-
tion of similar cations but inhibits the adsorption of dissimilar cat-
Figure 6. Nyquist impedance spectra measured at Ϫ0.6 V for a Ni-Co de-
posit formed at Ϫ1.1 V for ͑1,᭢͒ 5; ͑2, ᭝͒ 10; ͑3,᭹͒ 20; ͑4,छ͒ 40 min; and
then ͑5, ᭺͒ anodized at Ϫ0.1 V for 2 min in 0.15 M CoCl2 and 0.15 M
NiCl2 with pH of 2. The solid lines represent the fitting results according to
the equivalent circuit represented in Fig. 7.
Downloaded on 2015-01-15 to IP 132.239.1.231 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C620
Journal of The Electrochemical Society, 149 ͑11͒ C615-C622 ͑2002͒
Table III. The best-fit values of the equivalent circuit elements in
Fig. 7 for the impedance data shown in Fig. 6.
Polarization time at Ϫ1.1 V ͑min͒
Element
5
10
20
40
40b
Rct (⍀ cm2)
Cf (␮F cmϪ2
␣
48.9
195.6
1.0
94.3
86.8
0.71
1.71
4.12
46.8
237.6
1.0
82.8
97.2
0.71
2.01
3.85
26.0
453.0
0.98
9.6
165.0
0.72
1.99
4.13
25.5
558.7
0.96
9.2
171.9
0.73
1.73
4.9
69.1
240.0
)
0.89
4.02
92.5
0.85
1.64
3.2
Rad (⍀ cm2)
T (␮F cmϪ2
␣
)
␣
Rs (⍀ cm2)
Errora ͑%͒
Figure 7. The equivalent circuit for the impedance spectra shown in Fig. 6.
a The maximal error of the modulus between the measured and fitted
data for these spectra.
ions. Thus, the degree of anomaly should be further enhanced by the
lower barrier of electron transfer for the adsorbed monohydroxides
in the deposition of iron-group alloys. Note that the impedance re-
sponses of a Ni-Co deposit are very similar in shape to that of curve
1 when this deposit plated at Ϫ1.1 V for 40 min is oxidized at
Ϫ0.1 V for 2 min ͑see curve 5͒. This result indicates that the anodic
dissolution of the surface metal atoms causes the codissolution of
monohydroxide species and results in a decrease in coverage of the
monohydroxide layer. Based on the these results and discussion, one
can propose that the changes in the impedance responses with po-
larization time at Ϫ1.1 V are due to the variation in coverage/
thickness of the monohydroxide layer.
b The data from the spectrum of the deposit with anodizing at Ϫ0.1 V
for 2 min.
the mean errors of modulus and phase are, respectively, smaller than
1.0% and 0.3°, indicating a good fitting of the experimental data.
In Table III, the bulk solution resistance Rs is relatively low and
approximately independent of the dc deposition process, which is a
strong function of the distance between the tip of the Luggin capil-
lary and the deposit. The ionic charge-transfer resistance Rad de-
creases gradually with increasing dc plating time at Ϫ1.1 V, and a
sharp decrease is clearly observed at polarization times у20 min.
This phenomenon supports the proposal that a longer period of dc
deposition at Ϫ1.1 V should result in a complete coverage of the
adsorbed monohydroxides. Note the significant difference from 1 for
␣, indicating that Cdl at the deposit-electrolyte interface does not
behave as an ideal capacitor. This behavior may be attributed to the
electrochemical properties of the adsorbed monohydroxides as well
as the porous nature of the Ni-Co deposit. Thus, a constant-phase
element ͑CPE͒, T, is employed to replace the Cdl element. A CPE is
usually defined as 1/T(j␻)␣, where T is a constant with dimensions
(␮F cmϪ2)␣, which has been associated with the roughness of
electrode surfaces and caused rotation of the impedance semicircles
by an angle of (1 Ϫ ␣)90°.41-43 Note that T obviously increases as
the polarization time becomes longer. This phenomenon is attributed
to two reasons. First, the double-layer capacitance of the monohy-
droxide film is much larger than that of a metal deposit. This phe-
From the preceding results and discussion, an equivalent circuit
corresponding to the anomalous deposition of iron-group alloys is
proposed here. The analysis of Z(␻) against the polarization time of
dc plating at Ϫ1.1 V is based on the following equations
Z ␻͒ ϭ R ϩ Z ␻͒
͓1͔
͓2͔
͑
͑
I
s
1
1
ϭ
ϩ j␻Cdl
Z ␻͒
R
ad ϩ Zf
͑
I
1
1
ϭ
ϩ j␻Cf
͓3͔
Z ␻͒
͑
Rct
f
Here ␻, Rs , ZI , Zf , Rad , Rct , Cdl , and Cf are the angular fre-
quency, solution resistance, impedance of the deposit-electrolyte in-
terface, impedance of the adsorbed layer, adsorption resistance at the
adsorbed hydroxide layer-electrolyte interface ͑i.e., ionic charge-
transfer resistance͒, electron-transfer resistance of adsorbed mono-
hydroxides, double-layer capacitance, and faradaic pseudocapaci-
tance of the adsorbed layer, respectively.
ϩ
nomenon may be due to the ionic property of the adsorbed M͑OH͒
film, favoring the adsorption of charged species. Second, the surface
roughness of the deposits should increase with dc plating time, re-
sulting in an increase of the surface area of the deposits ͑i.e., double-
layer capacitance͒. However, Cf , the faradaic pseudocapacitance of
the adsorbed layer, does exhibit the ideal-capacitor impedance re-
sponses because ␣ of Cf is approximately equal to 1 for curves 1-4.
Note that the faradaic capacitance of Cf is much larger than that of
T ͑double-layer capacitance͒ and increases with the deposition time
of the dc plating at Ϫ1.1 V. This result implies that Cf is attribut-
The equivalent-circuit model for the anomalous deposition of
iron-group alloys, where every element has its physical meaning, is
shown in Fig. 7. Zf represents the faradaic process in the adsorbed
layer, which can be modeled as a pseudocapacitor (Cf) in parallel
with a resistor (Rct) of an electron-transfer reaction ͑i.e., the metal
deposition process͒. ZI indicates the impedance at the deposit-
electrolyte interface. This is reasonably modeled as a double-layer
capacitor (Cdl) in parallel with an ionic charge-transfer resistor
ϩ
able to the underpotential deposition of M͑OH͒ and/or metal ions,
which is believed to exhibit capacitive-like behavior.44,45 Because a
thicker layer of adsorbed monohydroxides is believed to form at a
longer time of dc deposition at Ϫ1.1 V, Cf is reasonably indicative
of the faradaic pseudocapacitance of the adsorbed film. The decrease
in electron-transfer resistance (Rct) with the thickness of the ad-
sorbed layer supports the proposal that the formation of the adsorbed
monohydroxide layer promotes the following metal deposition.
ϩ
(Rad) corresponding to the adsorption of M͑OH͒ at the deposit-
electrolyte interface, which is in series with Zf . The formation of the
adsorbed monohydroxide layer should favor the adsorption of simi-
lar monohydroxide and/or metal ions; therefore, the resistance of
adsorption is expected to decrease with the thickness of the adsorbed
layer, resulting from a longer period of dc deposition at Ϫ1.1 V.
In Fig. 6, the corresponding solid lines represent the best-fit
curves in terms of the equivalent circuits of Fig. 7. The good fitting
indicates that the equivalent-circuit models reasonably represent the
anomalous deposition of iron-group alloys. The best-fit values of the
equivalent-circuit elements are listed in Table III. The data in the last
row in Table III show the maximal error of the modulus. Note that
Linear sweep voltammetry (LSV) stripping and XRD pat-
terns.—Because the phase information of Zn-Ni alloys can be quali-
tatively identified by the anodic stripping analysis of the alloys in
an aqueous solution containing ethylenediaminetetraacetic acid
͑EDTA͒,46-48 the Zn-Ni codeposits were subjected to anodic strip-
Downloaded on 2015-01-15 to IP 132.239.1.231 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, 149 ͑11͒ C615-C622 ͑2002͒
C621
Figure 8. Linear sweep voltammograms of anodic stripping at 1 mV sϪ1
for Zn-Ni alloys with Zn content of ͑1͒ 100, ͑2͒ 80, ͑3͒ 60, ͑4͒ 40, ͑5͒ 20, and
͑6͒ 0 atom %. The LSV curves were measured in a solution containing
20 g dmϪ3 Na2SO4 and 20 g dmϪ3 EDTA with pH of 2.5.
ping by using LSV in a solution containing 20 g dmϪ3 Na2SO4 and
20 g dmϪ3 EDTA ͑pH 2.5͒. Typical LSV curves measured at
1 mV sϪ1 and 25°C for the anodic stripping of Ni-Zn alloys are
shown in Fig. 8. On curve 1, the anodic peak (A1) located between
ca. Ϫ1.05 and Ϫ0.65 V is attributed to the anodic dissolution of Zn.
From a comparison of curves 1-6, the voltammetric charge of peak
A1 decreases with increasing Ni content within the alloys. This phe-
nomenon was also observed by Karwas and Hepel,46 indicating the
preferential dissolution of Zn from the ␩ ϩ ␥ phases. Note that two
additional peaks ͑labeled A2 and A3) with a shoulder (A3Ј) are
clearly found when the Zn content varies from 80 to 40 atom %. The
existence of these stripping peaks reveals the formation of Zn-Ni
alloys in various phases.46-48 The relative amount of Zn-Ni alloys in
different phases is also affected by their composition on the basis of
their respective voltammetric charges. On curve 5, a plateau due to
the overlap of peaks A3 and AЈ3 and an additional peak (A4) are
clearly found, while both peaks A1 and A2 are not found. Peak A4
may be attributed to the anodic stripping of Ni. However, no sen-
sible current in the potential region of A4 is found on curve 6 ͑i.e.,
the LSV of the pure Ni deposit͒. Accordingly, Ni should be not
significantly dissolved in the stripping solution at potentials negative
of ca. 0.15 V Note the obvious increase in anodic currents at poten-
tials positive of ca. 0.05 V on curves 1-4. This is attributed to the
anodic dissolution of Cu because this sharp increase in anodic cur-
rents is most obvious on curve 1 ͑the LSV curve of pure Zn deposit͒.
From these results and discussion, Zn-Ni alloys with various phases
can be electroplated by CV from the deposition baths containing
their chloride precursors in different ratios, which is believed to be
suitable for the other alloys prepared in this work.
Figure 9. XRD spectra of Zn-Ni alloys with Zn content of ͑1͒ 80, ͑2͒ 60, ͑3͒
40, and ͑4͒ 20 atom %.
crystalline or amorphous Zn-Ni alloys of different phases are be-
lieved to electroplate on the Cu substrate by the potential cycling
method.
Conclusions
The anomalous codeposition of binary Co-Ni, Fe-Co, Fe-Ni, Zn-
Fe, and Zn-Ni alloys was efficiently inhibited by the CV technique
with cathodic deposition, double-layer, and anodic dissolution pro-
cesses. The dissolution of the freshly deposited alloys and the co-
ϩ
dissolution of the adsorbed M͑OH͒ layer in the anodic dissolution
region depressed the propagation of more active metals and inhib-
ited the anomalous codeposition of the iron-group alloys. On the
basis of the EIS results and equivalent-circuit model, the formation
of an adsorbed monohydroxide layer promoted subsequent metal
deposition, favoring the anomalous codeposition of iron-group al-
loys by the dc electroplating technique. The results of anodic strip-
ping of LSV and XRD spectra revealed that Zn-Ni alloys of several
phases with amorphous or poor-crystalline structures could be elec-
troplated on the Cu substrate using CV.
The crystalline structures of these alloys have also been checked
using XRD, and their XRD spectra are shown in Fig. 9. Diffraction
peaks of Cu are obviously found in these spectra, indicating that
these deposits should be thinner than 20 ␮m,49 which has been
confirmed to be between 7.2 and 15.0 ␮m and is good for magnetic
recording applications. The intensities of the Zn-Ni alloy peaks are
generally very low, indicating a poor crystalline quality of the Zn-Ni
alloys in different phases, although diffraction peaks corresponding
to Ni, NiZn3 , and NiZn can be identified in these spectra. Because
the mean thickness of these deposits ranged from ca. 7.2 to 15.0
␮m, constructive interference of X-rays should take place if the
deposit is of a good crystalline structure. Accordingly, poor-
Acknowledgments
The financial support of this work by the National Science Coun-
cil of the Republic of China is gratefully acknowledged.
National Chung Cheng University assisted in meeting the publication
costs of this article.
Downloaded on 2015-01-15 to IP 132.239.1.231 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C622
Journal of The Electrochemical Society, 149 ͑11͒ C615-C622 ͑2002͒
25. J. Vaes, J. Fransaer, and J.-P. Celis, J. Electrochem. Soc., 147, 3718 ͑2000͒.
26. K. Y. Sasaki and J. B. Talbot, J. Electrochem. Soc., 147, 189 ͑2000͒.
27. M. Matlosz, J. Electrochem. Soc., 140, 2272 ͑1993͒.
References
1. H. Saito, H. Hashimoto, and K. Inomata, Appl. Phys. Lett., 60, 2436 ͑1992͒.
2. W. Schwarzacher and D. S. Lashmore, IEEE Trans. Magn., 32, 3133 ͑1996͒.
3. C.-C. Hu and C.-Y. Weng, J. Appl. Electrochem., 30, 499 ͑2000͒.
4. C.-C. Hu and A. Bai, J. Appl. Electrochem., 31, 565 ͑2001͒.
5. Y.-P. Lin and J. R. Selman, J. Electrochem. Soc., 140, 1299 ͑1993͒.
6. S. S. Djokic and M. D. Maksimovic, in Modern Aspects of Electrochemistry, No.
22, J. O’M. Bockris, Editor, p. 417, Plenum Press, New York ͑1992͒.
7. H. V. Venkatasetty, J. Electrochem. Soc., 117, 403 ͑1970͒.
8. J. Bielinski and J. Przyluski, Surf. Technol., 9, 53 ͑1979͒.
9. T. Osaka, Electrochim. Acta, 44, 3885 ͑1999͒.
10. T. Osaka, M. Takai, K. Hayashi, K. Ohashi, M. Saito, and K. Yamada, Nature
(London), 392, 796 ͑1998͒.
11. E. M. Kakuno, D. H. Mosca, I. Mazzaro, N. Mattoso, W. H. Schreiner, M. A. B.
Gomes, and M. P. Cantao, J. Electrochem. Soc., 144, 3222 ͑1997͒.
12. K.-M. Yin, J. Electrochem. Soc., 144, 1560 ͑1997͒.
28. K. Y. Sasaki and J. B. Talbot, J. Electrochem. Soc., 142, 775 ͑1995͒.
29. T. M. Harris and J. St. Clair, J. Electrochem. Soc., 143, 3918 ͑1996͒.
30. S. Godad and T. M. Harris, J. Electrochem. Soc., 145, 3699 ͑1998͒.
31. P.-Y. Chen and I-Wen Sun, Electrochim. Acta, 46, 1169 ͑2001͒.
32. D. L. Grimmett, M. Schwartz, and K. Nobe, J. Electrochem. Soc., 137, 3414
͑1990͒.
33. D. L. Grimmett, M. Schwartz, and K. Nobe, J. Electrochem. Soc., 140, 973 ͑1993͒.
34. B. N. Popov, K.-M. Yin, and R. E. White, J. Electrochem. Soc., 140, 1321 ͑1993͒.
35. H. Schultz and M. Pritzker, J. Electrochem. Soc., 145, 2033 ͑1998͒.
36. I. Shao, P. M. Vereecken, R. C. Cammarata, and P. C. Searson, Abstract 603, The
Electrochemical Society Meeting Abstracts, Phoenix, AZ, Oct 22-27, 2000.
37. M. Paunovic and M. Schlessinger, Fundamentals of Electrochemical Deposition, p.
107, John Wiley & Sons, Inc., Singapore ͑1998͒.
13. E. Beltowska-Lehman and A. Riesenkampf, Surf. Technol., 11, 349 ͑1980͒.
14. A. Brenner, Electrodeposition of Alloys, Vol. 1-2, Academic Press, New York
͑1963͒.
38. D. Pletcher, in A First Course in Electrode Processes, p. 174, The Electrochemical
Consultancy, Romsey, England ͑1991͒.
39. D. F. Pickett and J. T. Maloy, J. Electrochem. Soc., 125, 1026 ͑1978͒.
40. Y. Sasaki and T. Yamashita, Thin Solid Films, 334, 117 ͑1998͒.
41. Impedance Spectroscopy Emphasizing Solid Materials and Systems, J. R. Mac-
donald, Editor, John Wiley & Sons, Singapore ͑1987͒.
42. L. Bai, L. Gao, and B. E. Conway, J. Chem. Soc., Faraday Trans., 89, 235 ͑1993͒.
43. L. Bai, L. Gao, and B. E. Conway, J. Chem. Soc., Faraday Trans., 89, 243 ͑1993͒.
44. B. E. Conway, J. Electrochem. Soc., 138, 1539 ͑1991͒.
45. B. E. Conway, Electrochemical Supercapacitors, Kluwer-Plenum Publishing, p. 17.
New York ͑1999͒.
46. C. Karwas and T. Hepel, J. Electrochem. Soc., 135, 2033 ͑1988͒.
47. V. I. Dybkov and O. V. Duchenko, J. Phase Equilib., 19, 434 ͑1998͒.
48. S. Swathirajan, J. Electrochem. Soc., 133, 671 ͑1986͒.
15. H. Dahms and I. M. Croll, J. Electrochem. Soc., 112, 771 ͑1965͒.
16. S. Biallozor and M. Lieder, Surf. Technol., 21, 1 ͑1984͒.
17. M. Lieder and S. Biallozor, Surf. Technol., 26, 23 ͑1985͒.
18. M. Ramasubramanian, S. N. Popova, B. N. Popov, R. E. White, and K.-M. Yin, J.
Electrochem. Soc., 143, 2164 ͑1996͒.
19. K.-M. Yin and S.-L. Jan, Surf. Coat. Technol., 79, 252 ͑1996͒.
20. K.-M. Yin and C.-C. Lee, J. Chem. Tech. Biotechnol., 70, 337 ͑1997͒.
21. J. O’M. Bockris, D. Drazic, and A. R. Despic, Electrochim. Acta, 4, 325 ͑1961͒.
22. L. T. Romankiw, in Electrodeposition Technology, Theory and Practice, L. T. Ro-
mankiw, Editor, PV 87-17, p. 301, The Electrochemical Society Proceedings Se-
ries, Pennington, NJ ͑1987͒.
23. S. Hessami and C. W. Tobias, J. Electrochem. Soc., 136, 3611 ͑1989͒.
24. K.-M. Yin, J.-H. Wei, J.-R. Fu, B. N. Popov, S. N. Popova, and R. E. White, J.
Appl. Electrochem., 25, 543 ͑1995͒.
49. C.-C. Hu and A. Bai, Mater. Chem. Phys., 77, 215 ͑2002͒.
Downloaded on 2015-01-15 to IP 132.239.1.231 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).