CoFeRh alloys. Part 1. Electrodeposition of Rh and nonmagnetic CoFeRh alloy

Article abstract of DOI:10.1016/j.electacta.2007.10.045

The electrochemical behavior of Rh(III) species in CoFe solution containing RhCl₃, NH₄Cl, H₃BO₃, CoSO₄, FeSO₄, saccharin, and NaLS (Na lauryl sulfate) has been investigated. The electrochemistry of Rh(III) species is influenced by each of the compounds present in CoFe plating solution, but especially by addition of saccharin and H₃BO₃ to the RhCl₃-NH₄Cl solution. The nucleation and growth of Rh on GC (glassy carbon), Ru, and Cu electrodes from NH₄Cl solution was studied using the potentiostatic current-transient methods. The results support a predominantly progressive nucleation of Rh on all three-electrode surfaces. The nucleation kinetic parameters ANo (steady state nucleation rate) and Ns (saturation nuclear number density) were found to vary with potential and are electrode-dependent in order: GC > Ru～Cu. The electrodeposited Rh films obtained from NH₄Cl solution and nonmagnetic CoFeRh film obtained from CoFe solution were characterized in terms of the following properties: morphology, surface roughness, crystal structure and chemical composition. The origin of light elements found in Rh and CoFeRh films (O, Cl, S, C, N) was discussed.

Full text of DOI:10.1016/j.electacta.2007.10.045

Available online at www.sciencedirect.com
Electrochimica Acta 53 (2008) 2483–2493
CoFeRh alloys
Part 1. Electrodeposition of Rh and
nonmagnetic CoFeRh alloy
∗
Ibro Tabakovic , Jiao-Ming Qiu, Steve Riemer,
Ming Sun, Vlad Vas’ko, Mark Kief
Seagate Technology, Research and Development, Bloomington, MN 55435, USA
Received 18 June 2007; received in revised form 5 September 2007; accepted 13 October 2007
Available online 26 October 2007
Abstract
The electrochemical behavior of Rh(III) species in CoFe solution containing RhCl
3
4 3 3 4 4
, NH Cl, H BO , CoSO , FeSO , saccharin, and NaLS (Na
lauryl sulfate) has been investigated. The electrochemistry of Rh(III) species is influenced by each of the compounds present in CoFe plating
solution, but especially by addition of saccharin and H BO to the RhCl -NH Cl solution. The nucleation and growth of Rh on GC (glassy carbon),
Ru, and Cu electrodes from NH Cl solution was studied using the potentiostatic current-transient methods. The results support a predominantly
3
3
3
4
4
progressive nucleation of Rh on all three-electrode surfaces. The nucleation kinetic parameters ANo (steady state nucleation rate) and Ns (saturation
nuclear number density) were found to vary with potential and are electrode-dependent in order: GC > Ru∼Cu. The electrodeposited Rh films
4
obtained from NH Cl solution and nonmagnetic CoFeRh film obtained from CoFe solution were characterized in terms of the following properties:
morphology, surface roughness, crystal structure and chemical composition. The origin of light elements found in Rh and CoFeRh films (O, Cl, S,
C, N) was discussed.
©
2007 Elsevier Ltd. All rights reserved.
Keywords: Electrodeposition of Rh and nonmagnetic CoFeRh; Nucleation and growth; Foreign elements in Rh and CoFeRh films
1
. Introduction
of low cost, high yield, and ease for thick film deposition, etc.
However, the growth of rhodium has a strong dependence on
electrode materials such as vitreous or glassy carbon [5,6], poly-
crystalline or single crystal gold [8–10], and pyrolitic graphite
electrode [11]. The co-deposition of impurity chemicals also has
a crucial influence on the physical performance of the final prod-
ucts. Therefore, a detailed study of rhodium electrodeposition
is desired.
In the first part of this work we describe the electrodeposi-
tion of Rh from NH4Cl solution and nonmagnetic CoFeRh alloy
from a CoFe plating bath containing RhCl3, NH4Cl, H3BO3,
CoSO4, FeSO4, saccharin and Na-lauryl sulfate (NaLS). The
nonmagnetic CoFeRh alloy described in this work contain only
small amount of magnetic Co and Fe elements, i.e. ∼4 wt.%.
A crucial requirement for the addition of rhodium into CoFe
deposit is the minimum dilution of the magnetic moment in
the product. This dilution is mainly due to the increase of
non-magnetic light elements present in the CoFeRh deposit
as “impurities”, which will be shown in the second part of
Rhodium is an important element in industrial applications.
It has been widely exploited as catalysts in fuel cells for the
production of hydrogen-free carbon monoxide, in automobile
catalytic converters, and in chemical reactors for hydrocarbon
conversion [1,2]. Recently rhodium was considered a poten-
tial barrier and seed material for copper in the interconnect
metallization for integrated circuits [3,4]. The incorporation
of appropriate amount of rhodium into CoFe alloys provides
much enhanced corrosion resistance than pure CoFe, which
serves as the dominant writer material for magnetic recording
head technology because of its high saturation magnetization.
Rhodium can be produced using either physical vapor deposi-
tion approaches or electrodeposition. The latter has advantages
∗
Corresponding author. Tel.: +1 952 402 8874.
E-mail address: ibro.m.tabakovic@seagate.com (I. Tabakovic).
0
013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2007.10.045

2
484
I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
this work. For example, Co Fe Rh1 film showed the high-
to deposit Rh films from solution adjusted at pH 2.0 failed due
to the delaminating of Rh films during electrodeposition caused
possibly by high stress from hydrogen embrittlement. The elec-
trodeposition at pH 4.0 produced compact and shiny Rh films.
The plating rate of 0.028 ␮m/min was achieved by electrodepo-
36
61
est obtainable magnetic moment of 2.4 T while Co28Fe Rh12
60
film gave Bs = 1.5 T.
The purpose of the present work is to study the effects of
the compounds present in the CoFe plating solution on rhodium
deposition, including its nucleating and growth, chemical com-
position, as well as the properties of rhodium thin films. It
is known that Rh(III) forms various octahedral complexes in
2
sition of Rh at 0.5 mA/cm in thickness ranges 0.1–2.0 ␮m.
2.2.2. Electrodeposition of nonmagnetic CoFeRh alloy
from CoFe plating solution
(
3−n)+
aqueous solution as [RhCln(H2O)6
]
[12]. Amazingly,
−n
althoughthesubstitutionofligandsisveryslowprocessinhomo-
geneous solution, only one reduction peak was observed for the
Electrodeposition of CoFeRh from a CoFe plating solution
containing 1 mM RhCl , 0.3 M NH Cl, 0.4 M H BO , 0.09 M
3
4
3
3
3
-electron reduction of Rh(III) to Rh(0) [5,6]. The proposed
4 4
FeSO , 0.041 M CoSO , 3 mM saccharin, 0.1 mM NaLS at pH
2.0 was a carried out at constant current density of 5 mA/cm
with a plating rate of 0.005 ␮m/min.
2
possibility for such a voltammetric behavior involves ligand
substitution catalyzed by electron transfer. However, it is also
possible that the products of 1-electron or 2-electron reduction
processes precipitate in the CoFeRh deposit during electrode-
position. This would increase the amount of the compounds
containing rhodium with increased amount of light elements
in CoFeRh deposit, which would result in a reduction of its
saturation magnetization.
2.3. Characterization of deposit
The chemical analysis of Rh and CoFeRh films were carried
out by using X-ray fluorescence (XRF), energy dispersive X-ray
spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS),
Rutherford back scattering (RBS) and Auger electron spec-
troscopy (AES) techniques. The structure and morphology of
Rh films were studied by using X-ray diffraction (XRD), trans-
mission electron microscopy (TEM), transmission high-energy
electron diffraction (THEED), scanning electron microscopy
2
. Experimental
2
.1. Voltammetric experiments
The instrumentation for voltammetric experiments has been
(SEM) and atomic force microscopy (AFM) techniques.
described previously [13]. For the voltammetry, the three-
electrode cell was used with a Pt-mesh anode and a Radiometer
saturatedcalomelelectrode(SCE), asareferenceelectrode. Note
that the cathodic currents in voltammograms have a positive val-
ues. Three kinds of electrodes were used as cathodes, i.e. glassy
carbon electrode (GCE), Cu and Ru. A GCE disc (area 0.07
3
3
3
. Results and discussion
.1. Voltammetry study
.1.1. The electrochemical behavior of Rh(III) in NH4Cl
2
or 0.2 cm ) was employed for cyclic voltammetry (CV), linear
solution
Fig. 1 shows cyclic voltammograms of 0.3 M NH4Cl solution
with2.8 mMRhCl3 (Fig. 1a)andwithoutRhCl3 present(Fig. 1b)
at pH 2.0. The first cathodic wave with Ep = −310 mV vs. SCE
corresponds to the 3-electron reduction of Rh(III) to Rh [5]. On
sweep voltammetry (LSV), chronoamperometry, and rotating
disc electrode voltammetry (RDE) in a 100 ml three-electrode
cell. The surface of GCE disc was polished before each measure-
ment with alumina (0.3 and 0.05 ␮m grades successively) on a
moist polishing cloth followed by subbing on emery paper and
washing the electrode with acetone and water. The voltammetry
on a sputtered alumina coated AlTiC 6 in. round wafers with a
˚
1
000 A of Cu or Ru layer were performed by enclosing the sub-
2
strate (1 cm ) within an electrochemical cell to the electrolyte
solution.
2
.2. Electrodeposition of Rh and nonmagnetic CoFeRh
films
The Rh and CoFeRh films, used for the characterization of
the deposit, were electrodeposited on 6 in. round alumina coated
˚
AlTiC wafers using 1000 A Cu seed layer. The thickness of films
were in the range of 0.1–2 ␮m and they were electrodeposited
using a standard paddle cell configuration [14] with a 1000 Oe
external magnetic field at constant current density.
2
.2.1. Electrodeposition of Rh from NH4Cl solution
Electrodeposition of Rh was carried out from 1 mM RhCl3 to
.3 M NH4Cl solution adjusted to pH 4.0. Notably, the attempts
Fig. 1. Cyclic voltammogramms for a solutions in aqueous 0.3 M NH4Cl at pH
2
2.0, scan rate: 20 mV/s, electrode: GCE (area 0.07 cm ) (a) with 2.8 mM RhCl3,
0
(b) without RhCl3.

I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
2485
the reverse scan there is a characteristic “nucleation” loop, i.e.
where the cathodic current is higher on reverse scan than on the
forward one. In accordance with literature data, there was never
any anodic peak, which could be interpreted in terms of stripping
of the metal [4–11]. At a more negative potential there is second
reduction wave at Ep = −507 mV vs. SCE which is attributed to
+
the H reduction on the Rh surface formed along the first wave.
This is followed by an oxidation peak at the reverse scan with a
peak at Ep = −348 mV vs. SCE which corresponds to the oxida-
tion of adsorbed hydrogen at the rhodium surface [5]. Rhodium
+
is a catalyst for hydrogen evolution and the overpotential for H
reduction at Rh-surface is less negative by about 600 mV than
+
the overpotential for H reduction at a GCE surface (Fig. 1b).
The cathodic peak potential, corresponding to the reduction
of Rh(III) to Rh, shifted negatively with the increase of the sweep
rate, v (Fig. 2). The shape factor, Ep − Ep/2, corresponds to the
irreversible case according to the equation:
Fig. 3. RDE voltammograms of 2.8 mM RhCl3 in aqueous 0.3 M NH4Cl solu-
tion adjusted to pH 2.0, scan rate 5 mV/s, rotation rate: 500 rpm, electrode: GCE
2
with area 0.2 cm . Rotation rates: 100, 250, 500, 1000, and 1500 rpm.
4
8
Ep − Ep/2 =
mV
(1)
αcnα
fusion occur. This behavior may be attributed to the nucleation
process as was reported for electrodeposition of Ag [16], Co
where nα is the number of electrons transferred up to, and includ-
ing, the rate determining step [15]. The average value of αcnα
was 1.99 and if nα = 3.0 then the αc = 0.66. The peak poten-
tial shifts cathodically for about 30/αcnα mV for each decade
increase in v. All of the diagnostic criteria fit to an irreversible
electrode reaction and the following equation for the peak cur-
[17] and Au [18]. An alternative explanation involving adsorp-
tion phenomena competing with the diffusion process could be
given.
The reduction of Rh(III) species in the steady state diffusion
regime was studied at GCE by RDE voltammetry. The rotation
rate, ω, of GCE was varied from 100 to 1000 rpm. A well-formed
reduction waves are obtained and the limiting currents, iL, are
proportional to the square root of the rotation rate, confirming
that the deposition is mass transport controlled (Fig. 3). Extrap-
◦
rent density at 25 C is:
5
1/2
1/2 1/2
ip = (2.99 × 10 )n(αcnα) CoD
v
(2)
The plot, ip–v1/2, is linear (see inset in Fig. 2), but does
not pass through the origin. The linearity is expected for reduc-
tion process under mass transport control. The current function,
1/2
1/2
olating the plot iL vs. ω to ω = 0 using the best-fit line
(
2
R = 0.998) leads to the zero current density value (see inset
1
/2
1/2
ip/v Co, decreases with increase of v by about 20% in the
range of studied sweep rates. Similar voltammetric behavior
of RhCl3 was observed on Ru and Cu electrodes (not shown
here) in the same solution. The apparent diffusion coefficient
in Fig. 3 where line is taken as a lead for eye). Inset in Fig. 3
is constructed by taking the limiting currents at the potential of
−
300 mV vs. SCE. Under these conditions the Levich equation
is satisfied [15]:
−
5
2
calculated using Eq. (2) is 4.58 × 10 cm /s. The decrease of
the current function, ip/v1 Co, with an increase of v
/2
1/2
(plot
2/3 −1/6
1/2
iL = 0.62nFD
ν
C ω
o
(3)
not shown here), along with the intercept higher than zero in
2
1/2
where i is limiting current density (A/cm ), F is the Faraday
ip–v
plot, indicates that additional processes other than dif-
L
2
constant, n = 3, ν is the kinematic viscosity (0.011 cm /s), Co
−6
3
is the concentration of RhCl3 (2.8 × 10 mol/cm ). From the
slope of Levich’s equation the diffusion coefficient for Rh(III)
−
6
2
species was evaluated as D = 5.9 × 10 cm /s. This value is
very close with value reported in the literature, i.e. D = 6.9 × 10
−
6
2
cm /s, determined also by RDE technique [6]. These values
are nearly an order of magnitude lower than those obtained using
LSV technique performed in quiescent solution. Similar differ-
ences between the D-values for Au(NH3)2 reduction obtained
through experiments run in quiescent (chronoamperometry) and
stirred (RDE) solutions have been reported [19]. We may assume
that weakly-adsorbed electroactive species exist on the electrode
during the time scale of LSV measurements. The existence of
adsorbed species gives rise to an apparent higher concentra-
tion of Rh(III) electroactive species at the electrode surface,
which can explain the higher D-value. It is also possible that
the weakly-adsorbed species are removed due to the forced con-
+
Fig. 2. Linearsweepvoltammogramsof2.8 mMRhCl3 inaqueous0.3 MNH4Cl
solution at pH 2.0. Electrode: GCE (area 0.07 cm ), scan rates: 10, 20, 50, 75,
2
1
00, and 150 mV/s.

2
486
I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
Fig. 5. RDE voltammograms for aqueous 2.8 mM RhCl3 solution adjusted
to pH 2.0, scan rate 5 mV/s, rotation rate: 500 rpm, electrode: GCE with
2
area 0.2 cm (a) 0.3 M NH4Cl, (b) 0.3 M NH4Cl + 3 mM saccharin, (c) 0.3 M
NH4Cl + 0.4 M H3BO3, (d) 0.3 M NH4Cl + 0.4 M H3BO3 + 3 mM saccharin, (e)
0.3 M NH4Cl + 0.4 M H3BO3 + 0.1 M FeSO4 + 0.04 M CoSO4 + 3 mM saccha-
rin.
surface coverage of saccharin on Cu electrode. The reverse of
the potential after first peak shows characteristic “nucleation”
loop for reduction of Rh(III) to Rh occurring on free sites of
Cu electrode in the potential range from −290 to −360 mV vs.
SCE.
3
.1.3. The electrochemical behavior of Rh(III) in CoFe
solution
Five RDE voltammograms recorded on a GCE for a series
of solutions are shown in Fig. 5. The voltammogram of 2.8 mM
RhCl3 in 0.3 M NH4Cl solution shows two waves with half-wave
potentials around −270 and −450 mV vs. SCE (Fig. 3a). The
first wave corresponds to the Rh(III) reduction to Rh and the sec-
+
ond is the reduction of H on Rh surface formed along the first
wave. Addition of saccharin shifts Rh(III) reduction potential
toward more negative values (Fig. 5b). The reduction of Rh(III)
in the presence of saccharin possibly starts at the potentials at
the foot of the voltammetric wave giving rise to the Rh deposit
which catalyses proton reduction. Therefore, the limiting current
Fig. 4. Cyclic voltammograms for 2.8 mM RhCl3 solution in aqueous 0.3 M
NH4Cl with and without saccharin present at pH 2.0, electrodes: GCE
2
2
2
(
2
area 0.07 cm ), Ru (area 1 cm ), and Cu (area 1 cm ) electrodes, scan rate:
0 mV/s.
+
density in Fig. 2b is a result of Rh(III) and H reduction in the
vection during polarization of the rotating disc electrode giving
rise to the lower D-value.
potential range from −400 to −800 mV vs. SCE. The addition
of boric acid into the RhCl3-NH4Cl solution significantly low-
+
ers the limiting current for H reduction in the potential range
3
.1.2. Effect of saccharin on electrochemical behavior of
studied (Fig. 5c). Similar effect of boric acid on limiting cur-
rent at a Pt electrode in Na2SO4 solution was explained through
a mechanism involving a net decrease in active surface area
caused by the adsorption of boric acid [20]. The addition of sac-
charin into the RhCl3-NH4Cl-H3BO3 solution causes the shift of
Rh(III) half-wave potential from −270 mV to −450 mV vs. SCE
(Fig. 5d). The RDE voltammogram obtained in the complete
CoFe plating solution shows one single wave with a half-wave
potential around −500 mV vs. SCE and higher limiting current
(Fig. 5e).
Rh(III) in NH4Cl solution
The cyclic voltammograms recorded at three different elec-
trodes (GCE, Ru, Cu) in 2.8 mM RhCl3–0.3 M NH4Cl solution
with and without saccharin addition are shown in Fig. 4. The
shapes of voltammograms obtained without presence of saccha-
rin is the same for all three electrodes with peak potentials of
−
311 mV (GCE), −305 mV (Ru), and −323 mV (Cu) vs. SCE,
respectively. Addition of saccharin shifts the Rh(III) reduction
potential toward more negative values, presumably due to the
surface coverage of adsorbed saccharin molecules. The peak
corresponding to the reduction of Rh(III) species disappeared
completely in voltammograms recorded at GCE and Ru elec-
trodes and partially at a Cu electrode indicating relatively lower
The linearity of the iL–ω1/2 plot shown in Fig. 6 for three
different solutions confirms that the limiting current is under dif-
fusion control. The additional increase of the limiting current in
the presence of CoSO4 and FeSO4 in solution, i.e. CoFe plating

I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
2487
Fig. 6. The iL–ω1/2 plots obtained for three different solutions at pH 2.0. (a)
0
.28 mM RhCl3 in 0.3 M NH4Cl solution at −300 mV vs. SCE, (b) 0.3 M
NH4Cl solution + saccharin at −700 mV vs. SCE, (c) 0.3 M NH4Cl + 0.4 M
H3BO3 + 0.1 M FeSO4 + 0.04 M CoSO4 + 3 mM saccharin at −700 mV vs.
SCE.
solution, indicated the occurrence of Co2+ and Fe2+ reduction in
the potential range of the limiting current plateau. The results of
controlled potential electrodeposition described in Part 2 of this
work [21] show the formation of CoFeRh alloys in this potential
range.
3
.2. Chronoamperometry
The nucleation processes of Rh on GCE, Ru and Cu elec-
Fig. 7. Current–time transient curves for Rh deposition in 2.8 mM RhCl /0.3 M
3
2
2
trodes were investigated by chronoamperometry in 2.8 mM
RhCl3-0.3 M NH4Cl solution. A set of current transients
obtained at different overpotentials are shown in Fig. 7. These
transients exhibit a typical shape for a nucleation on foreign sub-
strate. The current–time profiles consist of an initial capacitive
current spike due to the charging of the double layer followed by
a rise of current due to the growth of Rh nuclei and an increase
of their number. As the nuclei grow, the overlap of neighboring
diffusion zones gives rise to a current maximum (imax) and falls
due to the overlapping of diffusion zones around neighboring
nuclei.
NH4Cl solution at pH 2.0. (a) GCE (area 0.07 cm ), (b) Ru (area 1 cm ), (c) Cu
2
(
area 1 cm ).
3.3. Nucleation and growth
Theoretical models have been developed and successfully
applied to study the mechanism of nucleation and growth
[22–25]. The model for electrodeposition on a foreign substrate
generally assumes that nucleation of metal nuclei initiates at
certain specific energetic sites including steps, kinks and other
surface defects. The density of nuclei as a function of time, N(t),
is usually described in terms of a nucleation rate constant, A
[23].
The analysis of the decaying portion of the current–time
curves show that the i–t1 plots are linear which indicates that
the reduction of Rh(III) species is under diffusion controlled
regime as described by the Cottrell equation:
/2
N = No[1 − exp(−At)]
(5)
1
/2
where No is the final nucleus density which may depend on
applied potential and solution conditions. For At → ∞ the nucle-
ation is instantaneous and all nuclei are formed at the same time.
This means that all grains have practically the same size and are
distributed homogeneously on the surface. As At → 0, the nucle-
ation is progressive and new nuclei are formed during the growth
nFD Co
i =
(4)
1
/2 1/2
π
t
where i is current density, n is the number of electrons involved,
F is the Faraday’s constant, D is the diffusion coefficient, Co, is
the bulk concentration, and t is the time.

2
488
I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
of existing nuclei. The result of progressive nucleation is a low
density of different-sized grains on the surface. As At → 0 Eq.
(
5) therefore becomes
N ∼ ANot
(6)
In order to distinguish between instantaneous and progres-
sive nucleations, the transients in Fig. 7 can be analyzed in
reduced form in terms of the maximum current, imax, and time at
which the maximum current is observed [24]. For instantaneous
nucleation
ꢂ
ꢃ
ꢄꢅ
2
ꢀ
ꢁ
2
i
tmax
−1.2564t
=
1.9542
1 − exp
(7)
2
i
t
tmax
max
and for progressive nucleation
ꢂ
ꢃ
ꢄꢅ2
2
ꢀ
ꢁ
2
i
tmax
−2.3367t
=
1.2254
1 − exp
(8)
2
2
t
i
t
max
max
In Fig. 8 we present the transients as reduced-variable plots to
enable comparison with a standard model for three-dimensional
nucleation followed by diffusion-limited growth. These plots are
obtained from Fig. 7a–c for three different electrodes by normal-
izing the two variables in terms of imax and tmax. Fig. 8 shows the
experimental data obtained on three electrodes (GCE, Ru,Cu),
which are in agreement with the model for three-dimensional
progressive nucleation, although deviations from the model
in the region controlled by diffusion, were observed. Similar
results, interpreted in terms of the theory of progressive nucle-
ation and growth had been obtained earlier in the cases of Rh
electrodeposition on GCE [5], polycrystalline gold electrode [9]
and pyrolitic graphite electrode [11].
The product ANo that defines the steady state nucleation rate
usually characterizes the kinetics of progressive nucleation. The
nucleation number density was calculated from Eqs. (9) and (10)
based on the progressive nucleation model [24,26].
(
nFCo)2
ANo = 0.3864 2
(9)
3
i
t
k
max t max
ꢃ
ꢄ
Fig. 8. Comparison of the experimental i–t transients obtained in 2.8 mM
RhCl3/0.3 M NH4Cl solution with the theoretical working curves for progressive
and instantaneous nucleation.
1/2
4
3
8πCoM
k =
(10)
ρ
3
where ρ is the density of Rh (12.42 g/cm ), Co is the bulk con-
−
6
3
centration of RhCl3 (2.8 × 10 mol/cm ), n = 3, and F is the
Faraday’s constant. From the ANo product, the value of the sat-
uration nuclear number density, Ns, can be calculated according
to Eq. (11) [23].
Table 1
Current–time transient data for reduction of Rh(III) in 0.3 M NH4Cl solution
−
6
Electrode
Potential
mV vs. SCE)
tmax
(s)
imax
ANo
(cm
Ns × 10
(cm
−
−2 −1
−2
)
(
(A cm 2)
s
)
ꢃ
ꢄ
1
/2
ANo
GCE
−290
−300
2.65
2.10
0.62
0.000340
0.000443
0.000741
3.61
4.27
58.7
6.17
5.62
20.9
Ns =
(11)
(12)
2
kD
−
325
2
i
tmax
max
Ru
Cu
−290
2.63
2.07
1.19
0.000871
0.000986
0.001370
0.62
0.90
2.45
2.68
3.40
5.61
D =
0
.2598(nFCo)2
−300
−325
The results of these calculations, obtained by using the D-
values derived from Eq. (12) [24], are represented in Table 1.
The results show increase in ANo and Ns with the increase of
cathodic potential similar to those reported in literature for three-
dimensional progressive nucleation and growth process [23].
−290
4.03
2.58
1.85
1.21
0.000660
0.000776
0.000903
0.001096
0.272
0.751
1.50
2.15
3.58
5.07
7.89
−
−
−
300
310
325
3.65

I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
2489
Fig. 9. SEM images of 0.5 ␮m thick Rh film deposited from NH4Cl solution:
a) Magnification = 1K× and (b) magnification = 20K×.
(
Fig. 10. XPS spectra of Rh film obtained by electrodeposition from NH4Cl
solution. (a) O 1s and (b) Rh 3d.
NotethatbothANo andNs valuesforeachpotentialareelectrode-
dependent and follow the order: GCE > Ru∼Cu. This effect can
be explained by the lower density of nucleation sites at Ru and
Cu electrodes compared to GCE.
◦
tilted and analysis was performed at an take-off angle of 75 to
analyze material not damaged by the ion beam. The standard
binding energies of different chemical states of each element
were analyzed according to literature data [27,28].
3
.4. Electrodeposition and characterization of Rh films
obtained from NH4Cl solution
The “bulk” concentration of elements obtained after sput-
tering and calculated in weight percentage showed the
following composition: Rh (92.94 wt.%), O (6.95 wt.%) and Cl
3
.4.1. Morphology and surface roughness
Fig. 9 shows SEM images of a 0.5 ␮m thick Rh film at two
different magnifications. At lower magnification (Fig. 9a) the
cracks in the deposit were observed. This can be caused by
high tensile stress due to hydrogen embrittlement during elec-
trodeposition. Fig. 9b shows the compact Rh film consisting of
overlapped small nucleus. The appearance of the Rh films is
shiny with relatively low surface roughness.
(0.11 wt.%). Notably, a high content of oxygen, i.e. 6.95 wt.% or
31.8 at.%, was found in Rh film electrodeposited from RhCl3-
NH4Cl solution. Fig. 10a shows a convoluted high-resolution
XPS spectrum of the oxygen 1s core level. The O 1s spec-
trum shows an intense peak at 530.69 eV along with a peak
at 532.13 eV. The lower binding energy peak could be attributed
2−
to O type of species associated with rhodium oxide, close to
the binding energy reported for Rh2O3, i.e. 530.4 eV [27]. The
high intensity type of species at 532.17 eV could be assigned
3
.4.2. Chemical composition
The chemical compositions of Rh films were analyzed using
−
the XPS technique. The samples were sputtered for 12 min using
1
ing removed about 300–400 A of material. The samples were
to OH type species, and a weak peak at 533.9 eV indicates
+
kV Ar beam and 5 min using 500 eV beam. The sputter-
the presence of H2O in the deposit [29]. Fig. 10b shows a con-
voluted high-resolution spectrum for Rh 3d peaks. The peaks
˚

2
490
I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
Fig. 12. X-ray diffraction data from 0.8 ␮m thick CoFeRh film obtained from
CoFe plating solution.
3
.5.2. Crystal structure
Fig. 11. Effect of CoFeRh film thickness obtained from CoFe solution on surface
roughness.
The symmetric 2Θ–Θ XRD scans and 2Θ (Semann-Bohling
◦
method, ω = 1 ) scans were performed to determine the crystal
structure and texture of Rh films. The latter were used to avoid
the substrate diffraction peak input. Fig. 12 shows XRD patterns
of a 0.8 ␮m thick film. The XRD patterns of electrodeposited
CoFeRh film reveal the existence of face-centered cubic (fcc)
phase with (1 1 1), (2 0 0), (2 2 0), and (3 1 1) peaks. The aver-
age grain size was estimated from the deconvoluted fcc (1 1 1)
and (2 0 0) peaks by using the Scherrer formula. The correction
for instrumental broadening and for internal stress was used in
the evaluating grain size from XRD data. The electrodeposited
CoFeRh film shows a very small average grain size, i.e. 8.5 nm
for (1 1 1) and 6.5 nm for (2 0 0) texture. The measured lattice
located at 307.98 and 312.72 eV correspond to Rh 3d5/2 and Rh
3
d3/2 in metallic state, respectively. The peak at 309.33 eV could
be attributed to Rh2O3 compound possibly present at the grain
boundaries.
3
.5. Electrodeposition and characterization of CoFeRh
films obtained from CoFe solution
3
.5.1. Morphology and surface roughness
SEM image of electrodeposited CoFeRh film shows very
fine-grained films without cracks. The surface roughness was
imaged using the AFM technique. Fig. 11 shows three-
dimensional AFM images of Rh films taken at the same vertical
scale (50 nm/div) and for the scanning surface 5 ␮m × 5 ␮m.
The surface roughness, taken as an average of three scanning
surfaces, of CoFeRh films produced in a CoFe plating solution
is generally smaller than that of Rh films produced from NH4Cl
solution. The presence of saccharin as well as the lower pH of
the CoFe solution contributes to the improvement of roughness
and morphology.
˚
˚
parameter a = 3.803 A (Ref. [30], a = 3.80 A) suggests metallic
nature of Rh deposit.
Fig. 13 shows a TEM bright field image and a THEED. The
image at high resolution reveals the high crystallinity with grain
size around 6–8 nm, which is consistent with XRD results. The
electrodeposited CoFeRh from CoFe is polyscrystalline, metal-
lic film with “impurities” at the grain boundaries. Selected area
diffraction from the region of about 300 nm diameter, gave rings
patterns consistent with fcc structure and a lattice parameter of
˚
a = 3.79 A. The first four reflections in the pattern correspond to
Fig. 13. TEM bright field images and THEED diffraction pattern of electrodeposited CoFeRh from CoFe plating solution.

I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
2491
Table 2
by electrodeposition from CoFe plating solution. In addition,
SIMS analysis (not shown here) revealed the presence of H, C,
and N elements in Rh films.
Chemical analysis of elements in Rh films
Analysis
method
Rh
(wt.%)
O
S
Fe
Co
(wt.%)
Cl
(wt.%)
(wt.%) (wt.%) (wt.%)
Fig. 14 shows a high-resolution XPS spectra for CoFeRh film
electrodeposited from CoFe solution. The Rh 3d peaks located at
307 and 312 eV correspond to 3d5/2 and 3d3/2 peaks in metallic
state. The 2p peaks for Fe located at 707 and 720 eV corre-
sponding to its metallic state are not well-resolved from other
peaks in spectrum. Similarly, 2p peaks for Co located at 778
and 793 eV characteristic for the metallic state of Co were not
detected in XPS spectra. Both elements (Co and Fe) are proba-
bly present in the form of oxides, hydroxides and sulfides at the
grain boundaries. The convoluted XPS spectra for S 2p peaks
indicates the presence of FeS (161.55 eV) and CoS (162.73 eV).
The reported values for FeS (161.6 ± 0.1 eV) [32] and for CoS
RBS
EDS
AES
85.45
76.1
89.1
5.81
11.06
3.1
4.28
8.52
6.3
2.72
2.49
–
1.73
1.77
–
–
–
1.5
(
1 1 1), (2 0 0), (2 2 0), and(3 1 1)texturewhichareinaccordance
with the XRD results.
3
.5.3. Chemical composition
The “bulk” concentration of elements in electrodeposited
˚
CoFeRh films obtained after removal of 300–400 A from the
surface of CoFeRh film by sputtering were analyzed by RBS,
EDS, and AES methods. The quantitative analysis of elements,
represented in weight percentage, is dependent on the sensi-
tivity of the particular method. The results shown in Table 2
proved the presence of light elements in CoFeRh films obtained
(
162.6 ± 0.2 eV) [33] are in agreement with binding energies
obtained. The peak around 165 eV can be attributed to the pres-
ence of sulfur in the form of sulfate. The O 1s spectrum shows
peaks typical of metal hydroxides (MxOHy) at 532 eV and metal
oxides (MxOy) at 530 eV, respectively.
Fig. 14. High-resolution XPS spectra for electrodeposited CoFeRh films from CoFe plating solution.

2
492
I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
The C 1s spectrum in Fig. 14 shows two peaks at 283
4. Conclusions
and 284.5 eV. HPLC analysis of the solution obtained after
dissolving of electroplated CoFe film in the presence of sac-
charin, revealed the presence of benzamide, saccharin and
o-toluenamide in the deposit [31]. The benzene ring itself gives
a C 1s peak at 284.7 eV and, therefore the peak at 284.5 eV could
be attributed to the benzene ring present in all three compounds
detected in the CoFe deposit. The peak around 283 eV is often
interpreted as a carbide in the form of a metal carbide (MxCy).
However, the electrochemical reduction of organic compounds
to carbides during electrodeposition seems unlikely. If the peak
at 283 eV is due to metal carbide it could be produced by the
sputtering process before the XPS analysis [34].
It has been demonstrated that the electrochemistry of Rh(III)
species is influenced by each of the compounds present in CoFe
plating solution, i.e. NH Cl, H BO , CoSO , FeSO , and sac-
4
3
3
4
4
charin.
The results support progressive nucleation of Rh films,
obtained from NH Cl solution, on all three-electrode surfaces.
4
The nucleation kinetic parameters were found to vary with
potential and were electrode-dependent.
The CoFeRh films obtained from CoFe solution are generally
smoother than Rh films obtained from NH Cl solution. The elec-
4
trodeposited CoFeRh is polyscrystalline metallic fcc film with
small grain size, i.e. 8.5 nm for (1 1 1) and 6.5 nm for (2 0 0)
texture.
3
.6. Origin of foreign elements in Rh and CoFeRh films
The chemical composition of “bulk” Rh-film obtained from
NH4Cl solution showed a high content of oxygen (6.95 wt.%)
accompanied with 0.11 wt.% of chlorine. The CoFeRh films
obtained from CoFe solution showed even higher level of light
elements present in deposit, i.e. O, S, C, Cl.
The overall reduction of complexes [RhCln(H2O)6 ](
3−n)+
−n
with n = 3 can be formalized by the following equation:
−
[
RhCl3(H2O)3] + 3e → Rh + 3Cl + 3H2O
(13)
The increase of content of light elements in Rh and CoFeRh
films is associated to the 3-electron reduction of octahedral
The [RhCln(H2O)6 ](
3−n)+
complexes were proposed to
−n
(
3−n)+
[
RhCln(H2O)
]
complexes which can result with pre-
be in equilibrium and interchangeable on the time scale of
the experiment, which explains their reduction along a sin-
gle voltammetric wave. However, the reduction mechanism is
rather complex and may involve numerous reaction pathways
involving a preceding chemical reactions such as interconver-
6−n
cipitation of the adsorbed 1- or 2-electron reduction product
into Rh deposit as an “impurities” at the grain boundaries.
(
3−n)+
sion of complexes [RhCln(H2O)6
]
, where n equals 0 to
, in homogeneous solution before or after electron transfer and
following chemical reaction through release of leaving groups
−n
References
6
[
[
[
[
1] G.A. Somorjai, Introduction to Surface Chemistry and Catalysis, Wiley
Interscience, New York, 1994.
2] J.M. Thomas, W.J. Thomas, Principles and Practice of Heterogeneous
Catalysis, VCH, Weinheim, 1997.
3] M.W. Lane, C.E. Murray, F.R. McFeely, P.M. Vereecken, R. Rosenberg,
Appl. Phys. Lett. 83 (2003) 2330.
4] I. Goswami, R. Laxman, Semiconduct. Int. 27 (2004) 49.
(
Cl−, H2O)afterelectrontransfer. Therearemanypossiblepath-
ways leading to neutral compounds after the first or second
electron transfer to octahedral complexes. The neutral com-
pounds can precipitate in the deposit, presumably at the grain
boundaries. For example, the neutral product, [RhCl2(H2O)4],
could be formed through ligand interconversion after first elec-
tron transfer. The chemical analysis (Fig. 10, Table 2) of Rh and
CoFeRh films indicates the presence of H2O and chlorine in the
deposit and supports such a hypothesis.
[5] D. Pletcher, R.I. Urbina, J. Electroanal. Chem. 421 (1997) 137.
[
[
[
[
6] D. Pletcher, R.I. Urbina, J. Electroanal. Chem. 421 (1997) 145.
7] R. Gomez, J.M. Felin, Electrochim. Acta 44 (1998) 1191.
8] L.A. Kibler, M. Kleinert, D.M. Kolb, J. Electroanal. Chem. 467 (1999) 240.
9] M. Arbib, B. Zhang, V. Lazarov, D. Stoychev, A. Milchev, C. Buess-
Herman, J. Electroanal. Chem. 510 (2001) 67.
The presence of sulfur in the form of metal sulfides is due to
the reductive transformation of saccharin during CoFeRh elec-
trodeposition to SO2 [31]. Rh catalyzes the reductive conversion
of SO2 to H2S, which reacts with metal salts. This explains
the relatively high concentration of sulfur in deposit. The pres-
ence of carbon and nitrogen are due to the included organic
molecules in the Rh deposit such as saccharin, benzamide and
o-toluenebenzamide.
[10] S. Langerock, L. Heerman, J. Electrochem. Soc. 151 (2005) C155.
[
[
11] O. Brylev, L. Rone, D. Belmanger, J. Electroanal. Chem. 581 (2005) 22.
12] F. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, John Willey &
Sons, 1998, pp. 908–910.
[
[
13] I. Tabakovic, S. Riemer, V. Inturi, J. Electrochem Soc. 149 (2002) C18.
14] I. Tabakovic, S. Riemer, V. Vas’ko, V. Sapozhnikov, M. Kief, J. Elec-
trochem. Soc. 150 (2003) C635.
[
15] D. Pletcher, R. Greef, R. Pest, L.M. Peter, J. Robinson, Instrumental Meth-
ods in Electrochemistry, Horwood Publishing, Ltd., England, 1985, pp.
In addition, the foreign elements in the Rh and CoFeRh films
185–187.
(
O, H, Cl, Fe, Co) can be formed through reduction of metal
[
[
16] G.J. Hills, D.J. Schiffrin, J. Thompson, Electrochim. Acta 19 (1974) 657.
17] A.B. Soto, E.M. Arce, M. Palomar-Pardave, I. Gonzales, Electrochim. Acta
complexes, ML, presented in Eqs. (14) and (15):
4
1 (1996) 2647.
[18] S. Sobri, S. Roy, J. Electrochem. Soc. 152 (2005) C593.
19] G. Trejo, A.F. Gil, I. Gonzales, J. Electrochem. Soc. 142 (1995) 3404.
[20] J. Horkans, J. Electrochem. Soc. 126 (1979) 1861.
+
ML + e → MLadsorMLs
(14)
(15)
[
MLads + e → M + L−
[
21] I. Tabakovic, S. Riemer, V. Vas’ko, J.M. Qiu, M. Sun, M. Kief, Electrochim.
Acta, in preparation.
22] M. Palomar-Pardave, M.T. Ramirez, I. Gonzales, A. Serrya, B.R. Schar-
ifker, J. Electrochem. Soc. 143 (1996) 1551.
where M can be Co or Fe, and MLs is the included complex in
the deposit, L = OH , Cl , or HSO4 .
[
−
−
−

I. Tabakovic et al. / Electrochimica Acta 53 (2008) 2483–2493
2493
[
23] G. Gunawardena, G. Hills, I. Montenegro, B. Scharifker, J. Electrochem.
Soc. 138 (1982) 225.
[29] C.D. Wagner, D.A. Zatko, R.H. Raymond, Anal. Chem. 52 (1980)
1445.
[
[
[
24] B. Scharifker, G. Hills, Electrochim. Acta 28 (1983) 879.
25] B.R. Scharifker, J. Mostany, J. Electroanal. Chem. 177 (1984) 13.
26] J.-J. Lee, B. Miller, X. Shi, R. Kalish, K. Wheeler, J. Electrochem. Soc.
[30] C. Kittel, Introduction to Solid State Physics, John Willey & Sons, New
York, 1986.
[31] I. Tabakovic, S. Riemer, K. Tabakovic, M. Sun, M. Kief, J. Electrochem.
Soc. 153 (2006) C586.
1
48 (2001) C183.
[
[
27] J.F. Mondler, W.E. Stickle, P.E. Sobel, K.D. Bomben, Handbook of X-ray
Photoelectron Spectroscopy, Perkin-Elmer, Eden Praire, MN, 1992.
28] A. Briggs, M.P. Seah (Eds.), Practical Surface Analysis, vol. 1, John Willey
[32] P. Marcus, J.M. Grimal, Corros. Sci. 31 (1990) 377.
[33] C. Battistoni, L. Gastaldi, A. Lapecirella, G. Mattogno, C. Vitcoli, J. Phys.
Chem. 47 (1986) 899.
&
Sons, New York, 1990.
[34] G.N. Kashin, V.I. Makhnjuk, Surf. Interf. Anal. 21 (1994) 870.