Electrodeposition of Cu+PdO and (Cu-Pd) +PdO composites

Article abstract of DOI:10.1149/1.3310021

The codeposition of PdO particles with a Cu matrix has been investigated with the aim of extending the composite deposition procedure to the preparation of materials potentially capable of catalyzing the low temperature combustion of methane. Cu+PdO composites with a low dispersed phase content were obtained from either basic pyrophosphate or acid sulfate baths. Codeposition of PdO particles with a Cu-Pd alloy matrix was effectively achieved by electrolyzing suspensions of PdO in solutions containing both Cu²⁺ and Pd ²⁺ ions, with large [Cu²⁺] / [Pd²⁺] ratios. (Cu-Pd) +PdO deposits with 20-25 Pd atom % were obtained in which only 1-2 atom % of Pd were alloyed in the matrix. These (Cu-Pd) +PdO composites were mechanically stable and highly porous throughout their thickness (their pore volume being ca. 60% of the total volume). Their true surface area was some thousands of times higher than their geometric area.

Full text of DOI:10.1149/1.3310021

D248
Journal of The Electrochemical Society, 157 ͑4͒ D248-D255 ͑2010͒
0
013-4651/2010/157͑4͒/D248/8/$28.00 © The Electrochemical Society
Electrodeposition of Cu + PdO and „Cu–Pd… + PdO
Composites
a
a,
a
b,
Enrico Verlato, Sandro Cattarin, * Nicola Comisso, Rosalba Gerbasi, *
Paolo Guerriero, Marco Musiani, * and Lourdes Vázquez-Gómez
b
a, ,z
a
a
b
Istituto per l’Energetica e le Interfasi, and Istituto di Chimica Inorganica e delle Superfici, Consiglio
Nazionale delle Ricerche, 35127 Padova, Italy
The codeposition of PdO particles with a Cu matrix has been investigated with the aim of extending the composite deposition
procedure to the preparation of materials potentially capable of catalyzing the low temperature combustion of methane. Cu
+
PdO composites with a low dispersed phase content were obtained from either basic pyrophosphate or acid sulfate baths.
Codeposition of PdO particles with a Cu–Pd alloy matrix was effectively achieved by electrolyzing suspensions of PdO in
2
+
2+
2+
2+
solutions containing both Cu and Pd ions, with large ͓Cu ͔/͓Pd ͔ ratios. ͑Cu–Pd͒ + PdO deposits with 20–25 Pd atom %
were obtained in which only 1–2 atom % of Pd were alloyed in the matrix. These ͑Cu–Pd͒ + PdO composites were mechanically
stable and highly porous throughout their thickness ͑their pore volume being ca. 60% of the total volume͒. Their true surface area
was some thousands of times higher than their geometric area.
©
2010 The Electrochemical Society. ͓DOI: 10.1149/1.3310021͔ All rights reserved.
Manuscript submitted November 12, 2009; revised manuscript received January 13, 2010. Published March 11, 2010. This was
Paper 2728 presented at the Vienna, Austria, Meeting of the Society, October 4–9, 2009.
The low temperature complete oxidation of methane and low
hydrocarbons to CO is a process of great fundamental and applied
interest because the catalytic oxidation process, when employed in
power generation, has unique advantages over the conventional ther-
mal combustion process in terms of environmental friendliness. In-
deed, it is characterized by very low emissions of residual unreacted
methane, which is a much more potent greenhouse gas than CO2,
In our exploratory work, the choice of matrix and dispersed
phase catalytic materials had to take into account the requirements
of the electrodeposition process ͑i.e., the matrix had to be an elec-
tronic conductor that could be easily and effectively deposited͒ and
those of the catalytic combustion ͑i.e., all materials had to be stable
at relatively high temperatures͒. On the basis of the information in
2
1
,2
1
,2
the literature, PdO was selected as the dispersed phase. Nickel and
copper were considered as possible matrix materials because ͑i͒ re-
and is an intrinsically ultralow NO process because nitrogen oxide
x
1
9
liable methods for their electrodeposition are available, ͑ii͒ their
melting points ͑1485 and 1083°C, respectively͒ are sufficiently
high, and ͑iii͒ their oxides possess themselves some catalytic activ-
production is negligible at the typical catalytic combustion tempera-
tures ͑e.g., Յ600°C͒. Many catalytic systems have been investi-
gated with the aim of optimizing their activity and robustness. Most
of them consist of composite materials in which particles of the
active component are supported on a matrix that may have some
direct catalytic activity itself.
1
,2,20,21
ity toward methane oxidation
Pd-based catalysts.
and were used as supports for
2
2
Our group has been interested, for over a decade, in the elec-
trodeposition of composite materials for electrocatalysis and has
produced, using this approach, both anodes for oxygen evolution
Experimental
Chemicals and materials.— PdO·H O ͑99.9% purity͒ was pur-
2
3
chased from Alfa Aesar and was used as received. The typical size
of most PdO particles, estimated by scanning electron microscopy
reaction ͑OER͒ and cathodes for hydrogen evolution reaction
4
͑
HER͒. The former electrodes consisted of oxide-matrix compos-
͑
SEM͒, was in the range of 100–300 nm.
ites, deposited at the anode by electrolyzing suspensions of OER
The cell used in the voltammetric and electrodeposition experi-
catalysts in electrolytes containing either Pb or Tl+ ions, which
2
+
ments consisted of a main central cylindrical compartment, hosting
the working electrode ͑generally rotating at 2000 rpm͒, connected
through glass frits to two lateral compartments, which hosted two Pt
were converted to PbO2 5 and Tl O , respectively. The latter
,6
7,8
2
3
were metal-matrix composites of a very large surface area, produced
9
,10
according to a variant of the pioneering work of Iwakura et al.,
2
wire counter electrodes ͑overall area 2 cm ͒ in the deposition ex-
who extended the composite deposition procedures from the tradi-
tional field of surface coatings with enhanced hardness, wear resis-
periments, or a Pt wire counter electrode and a saturated calomel
electrode ͑SCE͒ in voltammetric experiments; the interelectrode gap
1
1-17
tance, and self-lubricating properties
to the production of thin-
was ca. 3.5 cm. All compartments contained the same electrolyte
1
8
film materials with a much wider variety of functions.
3
͑
8 cm in each compartment͒, but the PdO powders were suspended
Our group has recently become interested in exploring the poten-
tial of composite electrodeposition in the preparation of materials
able to catalyze the low temperature oxidation of methane. There are
some affinities between the catalysis of gas-phase reactions and that
of electrolytic processes. In both cases, ͑i͒ having catalysts with an
extended surface area is very important because these reactions are
heterogeneous and ͑ii͒ noble metals are among the most active cata-
lytic materials, but because of their high cost, there is a need to limit
their amount to a minimum. There are also significant differences;
e.g., thermal stability is important for combustion catalysts and
much less for electrocatalysts, whereas high electronic conductivity
is essential for electrocatalyst matrices and irrelevant for combustion
catalysts.
only in the central one; stirring of the suspensions was provided by
a magnetic bar rotating at ca. 1000 rpm. Combining electrode rota-
tion and vigorous stirring of the suspensions was shown to be effec-
tive in producing deposits with a homogeneous composition along
the disk radius. Unless differently specified, the electrolyte was in
contact with the atmosphere, and its temperature was maintained at
5
25°C with a thermostat.
2
Cu and Ni disk electrodes ͓0.2 or 0.3 cm geometric area, with a
poly͑tetrafluoroethylene͒ sheath of 1.0 cm diameter͔, polished with
emery paper 1000, rinsed with water, and dried in air, were used as
working electrodes. To avoid errors in the determination of the
Cu/Pd ratio, which might be caused by sampling the underlying
electrode material, alloy and composite samples submitted to the
energy-dispersive X-ray ͑EDX͒ analyses were always deposited on
Ni. The electrochemical equipment consisted of an EG&G
potentiostat/galvanostat model 273A, used in voltammetric and elec-
trodeposition experiments, and of an Autolab PGSTAT 100, used in
*
Electrochemical Society Active Member.
E-mail: m.musiani@ieni.cnr.it
z
Downloaded on 2015-06-24 to IP 128.210.126.199 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, 157 ͑4͒ D248-D255 ͑2010͒
D249
electrochemical impedance spectroscopy ͑EIS͒ experiments. The
(A) 0.185 M NaCl, 1.126 M Na
2
SO
B) Watt's Bath: 0.185 M NiCl , 1.126 M NiSO , 0.485 M H BO
3
4
, 0.485 M H BO
3
3
(
(
(
SEM analyses were performed with an FEI Quanta 200 FEG ESEM
instrument, equipped with a field-emission gun, operating in a high
vacuum condition at an accelerating voltage variable from 5 to 30
keV, depending on the observation needs. The EDX analyses were
obtained by using an EDAX Genesis EDX spectrometer at an accel-
erating voltage of 25 keV. The X-ray diffraction ͑XRD͒ patterns
were obtained by using a Philips X-PERT PW3710 diffractometer
with a Bragg–Brentano geometry, employing a Cu K␣ source ͑40
kV, 30 mA͒.
0.05
0.00
0.05
0.10
2
4
3
-
1
C) = (A) + 10 g L PdO
-1
D) = (B) + 10 g L PdO
A
C
-
-
B
D
Procedures.— The electrolytes used in the experiments aimed at
depositing Cu–PdO composites were obtained by suspending PdO
particles in either of the two Cu electrodeposition baths: ͑i͒ an acid
sulfate bath containing 0.88 M CuSO and 0.55 M H SO and ͑ii͒ a
-0.15
-
1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
4
2
4
mildly basic pyrophosphate bath containing 0.83 M Na P O and
E vs. SCE / V
4
2
7
0
.17 M Cu P O ͑pH 8.4͒. The former is the most common medium
2 2 7
19
for Cu deposition; the latter is a diluted version of another medium
recommended by Graham’s electroplating handbook. The acid sul-
Figure 1. CVs recorded with a Cu RDE ͑2000 rpm͒ in ͑curve A͒ 0.185 M
NaCl, 1.126 M Na SO , and 0.485 M H BO ; ͑curve B͒ 0.185 M
1
9
2
4
3
3
NiCl ·6H O, 1.126 M NiSO ·6H O, and 0.485 M H BO ; ͑curve C͒ 0.185 M
2
2
4
2
3
3
fate bath, with the addition of variable concentrations of PdSO , was
4
−1
NaCl, 1.126 M Na SO , 0.485 M H BO , and 10 g L PdO; ͑curve D͒
2
4
3
3
used in the deposition of Cu–Pd alloys and, with further addition of
PdO, in the deposition of ͑Cu–Pd͒ + PdO composites. The voltam-
metric curves were not corrected for the ohmic drop, which was
significant due to the cell geometry and possibly to some contact
resistance in the rotating disk electrode ͑RDE͒ assembly.
0
1
1
.185 M NiCl ·6H O, 1.126 M NiSO ·6H O, 0.485 M H BO , and
2 2 4 2 3 3
−1
0 g L PdO. All solutions had pH 4 and were kept at 55°C; sweep rate of
−1
00 mV s
.
Electrolyses were carried out under galvanostatic control, most
toward more negative potentials in a mildly acid NaCl, Na SO , and
−
2
2
4
often at Ϫ50 to −100 mA cm . The typical deposition charge was
H BO solution was HER, which became evident at E Ͻ −1.0 V.
−
2
3
3
1
00 C cm , which, in the deposition of pure Cu, would corre-
Curve B in the same figure shows that when Na salts were replaced
by the corresponding Ni salts, so as to obtain a classical Watt’s
spond to a deposit thickness of ca. 37 ␮m if the current efficiency
were 100%. The true current efficiency of the deposition processes
was estimated from the deposition charge and from the increase in
the mass of the electrodes.
The composition of the deposits containing Cu and Pd ͑formed
onto Ni RDEs͒ was obtained by EDX analyses. Areas of at least
19
bath, a current due to the Ni deposition flowed at E Ͻ −0.7 V.
Curve C, relevant to a sodium salt solution in which PdO particles
were suspended, shows a negative current at E Ͻ −0.65 V, thus
suggesting that PdO was electroactive because no other species
could be reduced. The current measured in the reverse sweep ͑to-
ward positive E͒ was more negative than that measured in the for-
ward sweep and was still negative at Ϫ0.50 V. Such a behavior is
typical of the nucleation of a new solid phase on the electrode sur-
face. After prolonged polarization of the electrode in the potential
range between Ϫ0.7 and Ϫ1.2 V, very thin, very poorly adherent,
gray deposits were seen. Curve D shows that the onset of a reduc-
tion current was shifted to a less negative potential when PdO par-
ticles were suspended in the Watt’s bath; at all E Ͻ −0.60 V, the
reduction current was higher in the presence than in the absence of
PdO. In the potential range between Ϫ1.0 and Ϫ0.45 V, the current
was more negative during the reverse than during the forward
2
00 ϫ 200 ␮m were sampled in these analyses; because the sam-
pling depth was estimated to be a few micrometers, the analyzed
7
deposit volume was ca. 10 times the typical particle volume. There-
fore, it could be assumed to contain a large number of PdO particles
and to be representative of the whole layers. The composition of the
deposits, either Cu–Pd alloys or PdO-containing composites, is ex-
pressed throughout this paper in terms of Pd atomic percent
NPd
Pd atom % = 100
͓1͔
NPd + NCu
where NPd and NCu represent atomic concentrations. Other elements,
e.g., oxygen, were neglected in the calculations.
Some voltammetric experiments aimed at assessing the feasibil-
ity of the codeposition of the PdO particles with a Ni matrix were
carried out using a conventional Watt’s bath: 1.126 M NiSO ·6H O,
0.2
4
2
1
9
0
.185 M NiCl ·6H O, and 0.485 M boric acid, pH 4, kept at 55°C.
The EIS and voltammetric tests aimed at characterizing the com-
2 2
0
.0
posite roughness and composition were carried out in a two-
compartment cell, with working and counter electrodes in the main
compartment and an SCE, used as reference, connected through a
Luggin capillary. In the EIS experiments, the working electrode was
polarized at its open-circuit potential; the frequency range of 100
mHz to 100 kHz was covered with eight points per decade and the
amplitude of the modulated potential was 10 mV rms. Some volta-
-0.2
A
C
-0.4
(
A) 0.88 M Na SO , 0.55 M H SO
2 4 2 4
-0.6
(B) 0.88 M CuSO , 0.55 M H SO
4
2
4
mmetric experiments were carried out with Pd sheet electrodes
-1
(C) = (A) + 10 g L PdO
2
B
D
-1
͑
0.4 cm area͒ after their electrochemical or thermal ͑2 h at 450°C͒
(D) = (B) + 10 g L PdO
-0.8
oxidation.
-
1.2 -1.0 -0.8 -0.6 -0.4 -0.2
E vs. SCE / V
0.0
0.2
0.4
Results and Discussion
Voltammetric study of PdO codeposition with Ni or Cu.— A pre-
liminary study aimed at assessing whether PdO particles could be
deposited with either a Ni or a Cu matrix was carried out by cyclic
voltammetry. The results are shown in Fig. 1 and 2, respectively.
Curve A in Fig. 1 shows that the only reduction process observed
by sweeping the potential of a Cu RDE from the open-circuit value
Figure 2. CVs recorded with a Ni RDE ͑2000 rpm͒ in ͑curve A͒ 0.88 M
Na SO and 0.55 M H SO ; ͑curve B͒ 0.88 M CuSO and 0.55 M H SO ;
curve C͒ 0.88 M Na SO , 0.55 M H SO , and 10 g L PdO; and ͑curve D͒
2
4
2
4
4
2
4
−1
͑
2 4 2 4
−
1
0.88 M CuSO , 0.55 M H SO , and 10 g L PdO; sweep rate of
4
2
4
−
1
100 mV s
.
Downloaded on 2015-06-24 to IP 128.210.126.199 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D250
Journal of The Electrochemical Society, 157 ͑4͒ D248-D255 ͑2010͒
(
a)
Figure 3. SEM image of a Cu + PdO deposit obtained by cathodic deposi-
−
1
tion from 0.83 M Na P O , 0.17 M Cu P O , and PdO 10 g L ͑pH 8.4͒
with a −100 mA cm deposition current density.
4
2
7
2
2
7
−
2
(b)
sweep. Also in this case, the deposits formed on the electrode were
very thin and adhered to the electrode very poorly.
A similar set of experiments, relevant to Cu deposition onto a Ni
RDE, is shown in Fig. 2. Curve A shows that HER took place at
E Ͻ −0.45 V in a Na SO and H SO solution. Copper deposition
2
4
2
4
from a CuSO and H SO solution occurred at a much more positive
4
2
4
potential ͑E Ͻ 0.07 V͒, as shown by curve B. Curve C shows that
the PdO particles were electroactive when suspended in the Na SO
2
4
and H SO solution; the fact that the current was more negative
2
4
during the reverse than during the forward sweep suggests that a
deposit was formed. However, the electroactivity of PdO was ob-
served only at potentials more negative than those necessary to Cu
deposition ͑comparing curves B and C͒, at variance with the behav-
ior observed with Ni ͑Fig. 1͒. When the PdO particles were sus-
pended in CuSO and H SO solution, no significant potential shift
4
2
4
was observed in the onset of Cu deposition; the only significant
effect of the PdO particles was enhancing the reduction current
Figure 4. SEM images of deposits obtained by cathodic deposition from
0.88 M CuSO4, 0.55 M H2SO4, and 10 g L PdO, with a −50 mA cm
deposition current density. Deposit ͑a͒ was obtained from a freshly prepared
suspension; deposit ͑b͒ was obtained from a suspension stirred during ca. 3
−
1
−2
͑
curve D͒.
The comparison of Fig. 1 and 2 suggests that the electroactivity
h.
of the PdO particles ͑which may be due both to their reduction when
they are brought in contact with the electrode surface and to the
catalysis of HER by Pd formed at the PdO surface and deposited on
the electrode͒ prevents the effective deposition of Ni. Instead, Cu
may be deposited in a potential range where PdO is inactive, at less
negative potentials than Ni, and so Cu and Cu matrix composites
can be grown. Indeed, prolonged electrolyses carried out in the
CuSO and H SO solution with suspended PdO particles led to the
Cu P O , and 10 g L PdO ͑pH 8.4͒ with a −100 mA cm−2 cur-
−
1
2
2
7
rent density. Under these and similar conditions, thick and porous
deposits were obtained, with a current efficiency close to 100%, but
this system had several drawbacks: The Pd content in the deposits
was limited to Յ2 atom %, the deposits were rather brittle and
their adhesion to the substrate was rather poor. Furthermore, prob-
ably due to local pH changes, precipitation of some salt, possibly a
basic Cu pyrophosphate, occurred simultaneously to deposit forma-
tion. For all these reasons, this route was abandoned.
4
2
4
formation of thick deposits, as described below. The current en-
hancement due to the PdO particles ͑comparing curves C and D in
Fig. 2͒ probably originated from different phenomena: simultaneous
2
+
Cu and PdO reduction, enhanced HER, and deposition of rougher
deposits, with a larger effective area ͑see below͒. Results similar to
those in Fig. 2 were obtained with the Cu pyrophosphate deposition
bath.
Deposition of Cu ϩ PdO and ͑Cu–Pd͒ ϩ PdO composites from
acid copper sulfate solutions.— Several deposition experiments
−
2
were carried out in succession by imposing a −50 mA cm depo-
sition current density to Ni cathodes rotating at 2000 rpm in the
Deposition of Cu ϩ PdO composites from basic copper pyro-
phosphate solutions.— According to the Pourbaix diagrams, PdO is
quite unstable toward dissolution at pH Ͻ 1. Therefore, although
the most common Cu deposition bath is a sulfate bath with pH ca.
−
1
same 0.88 M CuSO , 0.55 M H SO , and 10 g L PdO suspen-
4
2
4
−
2
sion. The deposition charge was 100 C cm , and so the duration of
each electrolysis was 2000 s. Deposits were formed in all electroly-
ses with a current efficiency close to 100%, i.e., with negligible gas
evolution side reactions. A comparison of the SEM images of de-
posits obtained from ͑a͒ a freshly prepared suspension and from ͑b͒
a suspension stirred during ca. 3 h is shown in Fig. 4. A marked
1
9
0,
to operate with solutions in which PdO was stable, we initially
attempted the deposition of Cu + PdO composites using basic cop-
per pyrophosphate solutions. Figure 3 shows an SEM image of a
Cu + PdO deposit obtained from 0.83 M Na P O , 0.17 M
4
2
7
Downloaded on 2015-06-24 to IP 128.210.126.199 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, 157 ͑4͒ D248-D255 ͑2010͒
D251
3
2
2
0
5
0
2
1
1
0
5
0
5
0
No PdO
15
-
1
1
2
0 g L- PdO
1
0 g L PdO
1
0
5
0
0
3
6
9
12
15
18
0
5
10
15
20
25
3
-
1
[
PdSO ] 10 / M
PdO in Suspension / g L
4
Figure 5. Dependence of the composition of Cu–Pd and ͑Cu–Pd͒ + PdO
Figure 6. Dependence of the composition of ͑Cu–Pd͒ + PdO deposits on the
deposits on the concentration of PdSO added to 0.88 M CuSO and 0.55 M
concentration of PdO suspended in 0.88 M CuSO , 0.55 M H SO , and 0.006
4
4
4
2
4
−
1
−1
H SO solutions containing no ͑ᮀ͒, 10 g L ͑b͒, or 20 g L ͑᭡͒ PdO in
suspension. Deposition current density and deposition charge were
50 mA cm and 100 C cm , respectively. The lines are just guides for
M PdSO solutions. Deposition current density and deposition charge were
2
4
4
−
2
−2
−50 mA cm
calculated
and 100 C cm
according to
,
respectively. The continuous curve,
−
2
−2
−
Pddeposit = 34 ϫ ͓1.25PdOsuspension/͑1
the eyes.
+ 1.25PdOsuspension͔͒, is a guide for the eyes.
increase in the deposit roughness from ͑a͒ to ͑b͒ is evident. EDX
analyses of the samples shown in Fig. 4 showed that the Pd contents
were 1.9 and 8.7 atom %, respectively.
formed in which the PdO particles were dispersed in a Cu–Pd alloy
matrix; henceforth, these materials will be denoted as ͑Cu–Pd͒
+ PdO. The difference in the Pd content between ͑Cu–Pd͒ + PdO
This evolution in the composition of the deposits can be inter-
2
3
preted by taking into account that the equilibrium condition for the
reaction
2
+
composites and Cu–Pd alloys increased as ͓Pd ͔ increased to be-
+
2+
come essentially independent of it. The Pd content in the
PdO + 2H = Pd + H O
͓2͔
2
2+
͑
3
Cu–Pd͒ + PdO composites obtained with ͓Pd ͔ in the range of
is
−3
–18 ϫ 10 M was much higher than the sum of the Pd content in
2
+
the alloys ͑obtained without suspended PdO͒ and the Pd content in
log͓Pd ͔ = −2.35 − 2pH
͓3͔
2
+
the Cu + PdO composites ͑obtained in the absence of Pd in solu-
tion͒, thus suggesting a catalytic effect. In principle, two alternative
and therefore PdO should dissolve in a strongly acidic solution, to
−
3
reach an equilibrium concentration of 4.5 ϫ 10 M in a pH 0 so-
explanations may be proposed for the phenomena shown in Fig. 5:
2
+
2+
lution. Simultaneous reduction of Cu and Pd might then lead to
2+
͑
i͒ The presence of Pd in the deposition bath enhanced the depo-
24-26
the formation of a Cu–Pd alloy,
with a consequent increase in
the Pd content of the deposit. According to Simonet et al., Pd
spontaneously deposits on Cu from acid sulfate solutions of Pd ,
and this might be another possible route for the entry of Pd in the
deposit.
The dissolution kinetics of PdO ͑10 g L ͒ in a 0.88 M Na SO
and 0.55 M H SO solution was monitored by recording the UV/
sition rate of PdO, which was codeposited more effectively with a
Cu–Pd alloy than with pure Cu, or ͑ii͒ the presence of PdO in the
suspension enhanced the atomic fraction of Pd in the Cu–Pd alloys
formed at the cathode. This issue is discussed below on the basis of
the XRD data.
27
2
+
−
1
2
4
The data in Fig. 5 allows us to interpret the dependence of the
deposit composition on the number of successive electrolyses per-
formed with the same suspension ͑see above discussion of Fig. 4͒:
As the duration of the interaction between the PdO particles and the
2
4
visible spectra of solution samples after various interaction times.
2
+
During the first 6 h, the Pd concentration increased, essentially
linearly, by 1.6 ϫ 10−3 M/h, to yield a Pd concentration in excess
of that given by Eq. 3, which neglects the effect of sulfate anions.
2+
2
+
acid solution increased, the Pd concentration in the solution built
up, the matrix material deposited at the cathode changed from es-
sentially pure Cu to Cu–Pd alloys, and composites with a high con-
tent of either alloyed Pd͑0͒ or PdO were formed.
Cu–Pd alloys were deposited from 0.88 M CuSO and 0.55 M
4
H SO solutions containing variable concentrations of PdSO , at a
2
4
4
−
2
−
50 mA cm current density, and were submitted to the EDX
Figure 5 also shows that the overall Pd content was higher for
samples prepared with the higher amount of PdO in suspension ͓͑᭡͒
vs ͑b͔͒. This issue was further investigated by varying the amount
of PdO suspended in 0.88 M CuSO , 0.55 M H SO , and 0.006 M
2
+
analysis. Figure 5 ͑ᮀ͒ shows that for Pd concentrations in the
range of 0–18 ϫ 10−3 M, the Pd content in the alloy varied roughly
linearly and never exceeded 5 atom %. The Pd content in the alloys
increased quite sharply as the Pd2+ concentration was further in-
creased ͑data not shown in Fig. 5͒. The dependence of the compo-
sition of deposits obtained from 0.88 M CuSO and 0.55 M H SO
4
2
4
PdSO solutions ͑again, freshly prepared suspensions were used in
4
each experiment͒. The results are shown in Fig. 6, where the experi-
4
2
4
mental data are roughly interpolated by a curve that phenomenologi-
solutions containing 10 or 20 g L−1 PdO on Pd concentration was
then investigated. Data are shown in Fig. 5 ͓͑b͒ and ͑᭡͒, respec-
tively͔. In these experiments, a freshly prepared suspension was
used in each electrolysis, and thus, PdO dissolution might cause an
increase in the Pd2+ concentration by 9 ϫ 10 M at most. Figure 5
shows that the Pd content in these samples was much higher than
2+
28
cally reproduces the behavior predicted by the Guglielmi model.
This curve shows that in spite of some data scattering ͑somewhat
larger than the error expected in the EDX analyses͒, the composition
of ͑Cu–Pd͒ + PdO composites depends on the PdO concentration in
a way commonly observed for many composite deposition systems:
When the dispersed phase concentration in the suspension is pro-
gressively increased, its concentration in the deposit increases mark-
edly at first and then more weakly up to a limiting value.
−4
2
+
2+
that in the alloys obtained with the same ͓Cu ͔/͓Pd ͔ ratio without
suspended PdO. Therefore, it can be concluded that composites were
Downloaded on 2015-06-24 to IP 128.210.126.199 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D252
Journal of The Electrochemical Society, 157 ͑4͒ D248-D255 ͑2010͒
trode was measured at its open-circuit potential ͑Ϫ0.025, Ϫ0.04,
0.015, and Ϫ0.02 V, respectively͒. None of the electrodes had an
ideal capacitive behavior, which agrees with Härtinger and
+
(a)
3
1
Doblhofer, who reported that deviations in the impedance of Cu
electrodes immersed in neutral media from an ideally capacitive
behavior were caused by OH− electrosorption. However, a fairly
wide frequency range ͑delimited in Fig. 8a by a couple of vertical
dotted lines͒ in which the slope of the curves approached Ϫ1 could
be identified in each plot. For Cu and Cu–Pd alloy electrodes, this
range was found at medium-high frequency, whereas for the
͑
Cu–Pd͒ + PdO composite a capacitive behavior was best ap-
proached at much lower frequencies ͑between 0.1 and 10 Hz͒. The
latter behavior is typical of porous electrodes, the capacity of which
is equal to that of a flat electrode of the same area as the developed
pore surface, and can be measured only at frequencies low enough
3
2
that the ac signal penetration depth is as large as the pore depth. At
higher frequencies ͑e.g., 50–2000 Hz͒, the plot relevant to the
͑
Cu–Pd͒ + PdO composite shows a linear branch with a slope close
to Ϫ1/4; in a Nyquist plot of the impedance of the ͑Cu–Pd͒
PdO composite ͑Fig. 8b͒ the high frequency linear portion forms
with the real axis an angle close to 22.5°, much smaller than 45°, the
+
3
3
value expected for cylindrical pores. Similar results were previ-
ously reported by Gourbeyre et al., who ascribed them to the pres-
(
b)
34
ence of branched pores.
The surface roughness of each electrodeposit ͑defined as the ratio
between its true surface area and its geometric area͒ was obtained by
dividing its capacity by the capacity of the polished Cu electrode
͑
which is equivalent to assuming the latter to be ideally flat͒. Be-
cause the impedance response of the electrodes was never ideally
capacitive, the linear portions of the impedance plots ͑Fig. 8a͒ were
analyzed, assuming a series resistance-constant phase element ͑R-
CPE͒ combination as a model, with the CPE impedance given by
1
ZCPE
=
͓4͔
␣
Q͑j␻͒
The parameters Q and ␣ were evaluated by the graphical method
3
5
proposed by Orazem et al., and the effective double-layer capacity
3
6
was computed according to the formula proposed by Brug et al.
for blocking electrodes
Ceff = Q1/␣R͑1−␣͒/␣
͓5͔
Figure 7. SEM images of deposits obtained from 0.88 M CuSO , 0.55 M
4
−
1
H SO , and 0.003 M PdSO solutions in which ͑a͒ no PdO or ͑b͒ 10 g L
2
4
4
PdO were suspended. In both cases, deposition current density and deposi-
The surface roughness of the Cu electrodeposit was only marginally
higher than that of polished Cu. The Cu–Pd alloy had a markedly
higher surface roughness ͑5.8͒. The surface roughness of the com-
posite was as high as 7300, ca. 3 orders of magnitude higher than
that of the alloy. In spite of the coarse approximations ͑i.e., assum-
ing the double-layer capacity of the unit electrode surface to be
independent of the electrode composition and the polished Cu elec-
trode to be ideally flat͒, the results in Fig. 8 prove the major effect of
codeposition of PdO particles on the true surface area of the depos-
its.
tion charge were −50 mA cm− and 100 C cm , respectively.
2
−2
Characterization of ͑Cu–Pd͒ ϩ PdO deposits.— Figure 7 shows
the SEM images of a Cu–Pd alloy obtained from a ͑a͒ 0.88 M
CuSO , 0.55 M H SO , and 0.003 M PdSO solution and of a ͑b͒
4
2
4
4
͑
Cu–Pd͒ + PdO composite obtained from an identical solution in
−
1
which 10 g L PdO was suspended. The EDX analyses showed
that these samples contained 0.6 and 6.7 Pd atom %, respectively.
Clearly, besides increasing the Pd content, codeposition of PdO par-
ticles induced a marked increase in the deposit roughness. Such a
behavior is commonly observed when particles of sufficiently high
electronic conductivity are used as dispersed phases in composite
Figure 9 shows a fracture section of a ͑Cu–Pd͒ + PdO deposit
grown to a thickness of ca. 200 ␮m. Both magnifications show that
the deposit is highly porous throughout its whole depth and that
pores with sizes variable from a few micrometers down to 100 nm
or less are present. The thickness of the sample shown in Fig. 9 was
quite constant along its radius, and thus its total volume could be
easily calculated with fairly good precision. From the deposition
charge, the volume of deposited Cu was estimated ͑neglecting other
components of the deposits͒. By comparing these two volumes, it
was estimated that the pore volume was ca. 60% of the total volume.
Some Cu–Pd alloys with up to 25 Pd atom % were deposited from
2
9
deposition and has been exploited in the preparation of electro-
catalysts with a large effective area.
3,5-8,18,30
To obtain a quantitative estimate of the effective area of the alloy
and composite deposits, their impedance was measured in an inert
electrolyte ͑0.5 M Na SO ͒, so as to determine the double-layer
2
4
capacity of the electrodes, which, in the first approximation, may be
assumed to be proportional to the true area, irrespective of the
sample composition. Figure 8a shows the plots of the imaginary part
of the impedance vs frequency, in logarithmic–logarithmic coordi-
nates, for a Cu electrode polished with Al O and for Cu, Cu–Pd,
2
+
2+
2+
2+
Cu and Pd acid sulfate solutions with appropriate ͓Cu ͔/͓Pd ͔
concentration ratios. SEM images and impedance measurements,
not reported here, showed that their surface roughness was signifi-
cantly higher than that of the sample shown in Fig. 7a. Although the
Cu–Pd alloys were not investigated in detail, it may be speculated
2
3
and ͑Cu–Pd͒ + PdO electrodeposits; the impedance of each elec-
Downloaded on 2015-06-24 to IP 128.210.126.199 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, 157 ͑4͒ D248-D255 ͑2010͒
D253
8
(
b)
4
3
2
1
0
1
7
6
5
4
3
2
1
0
1
1
1
1
1
0
0
0
0
(
a)
(
Cu-Pd)+PdO composite
Cu-Pd Alloy
0.75 Hz
Electrodeposited Cu
Polished Cu
Figure 8. ͑a͒ Frequency dependence of
the imaginary part of the impedance of Cu
͓
͑᭺͒ polished; ͑b͒ electrodeposited͔,
Cu–Pd ͑᭢͒, and ͑Cu–Pd͒ + PdO ͑᭿͒ elec-
trodes, measured in 0.5 M Na2SO4, at the
respective open-circuit potentials. ͑b͒ Ny-
quist plot of the impedance of the
0
͑
Cu–Pd͒ + PdO electrode.
-
10
4
2.2 Hz
2
-
1
0
1
2
3
4
5
1
0
10
10
10
10
10
10
0
1
3
4
5
frequency / Hz
2
Re(Z) / Ω cm
that they might provide a route to the deposition of porous Pd-
containing materials onto substrates of complicated geometry for
which composite deposition might be ineffective.
atomic fraction of the alloy components, corresponds to a Cu Pd
80 20
composition; such a composition agrees well with the EDX analysis
of the same sample, considering the precision of the analyses and
Further work was aimed at establishing the chemical status of the
Pd atoms contained in the deposits. The cyclic voltammogram ͑CV͒
of a ͑Cu–Pd͒ + PdO deposit was recorded in a 0.5 M Na SO solu-
2
4
tion ͑in which, due to the absence of electroactive species, only
redox processes involving the electrode material or the solvent are
possible͒ and compared to those of oxidized Pd electrodes and of a
Cu electrodeposit. The results are shown in Fig. 10. The potential of
the thermally oxidized Pd electrode was swept at first toward a
negative potential ͑curve A͒: A reduction peak was observed at
3
7
Ϫ0.54 V, which may be attributed to the PdO reduction, followed
at more negative potentials by water reduction ͑with hydrogen ad-
sorption, absorption, and evolution͒. In the reverse sweep, peaks at
0
.42 and 0.82 V may be ascribed to the reoxidation of hydrogen and
Pd. For electrochemically oxidized Pd ͑curve B͒, the potential of a
polished Pd electrode was swept at first toward positive potentials
͑
up to 1.7 V, not shown in Fig. 10͒, causing Pd oxidation; a reduc-
tion peak was observed at Ϫ0.13 V, suggesting that the species
produced by electrochemical oxidation, presumably more strongly
hydrated, was easier to reduce than the thermal oxide. A weak re-
duction peak around Ϫ0.12 V was observed also for the Cu elec-
trodeposit ͑curve C͒; it is ascribed to the reduction of CuO formed
by atmospheric oxidation and upon immersion of the electrode in
the Na SO solution. Finally, the voltammogram of ͑Cu–Pd͒
2
4
+
PdO showed reduction peaks at Ϫ0.28 and Ϫ0.43 V during the
initial sweep toward negative potentials ͑curve D͒. These peaks are
compatible with the reduction of both CuO and incorporated PdO
particles. The Cu and ͑Cu–Pd͒ + PdO deposits were obtained with
the same deposition charge; the much larger reduction current mea-
sured in the latter case was probably a result of the much larger
surface area of the composite deposit.
Figure 11 shows the X-ray diffractograms of ͑A͒ the as-
purchased PdO powder used in the electrodeposition experiments,
͑
B͒ an electrodeposited Cu –Pd alloy, and ͓͑C͒ and ͑D͔͒ two elec-
77 23
trodeposited ͑Cu–Pd͒ + PdO composites. The PdO diffractogram
shows very broad and not very intense peaks, located at angles,
38
which agree with the literature, which show that the PdO powders
were almost amorphous ͑the crystal size estimated by the Scherrer
formula is in the range of 20–30 Å͒. The diffractogram of the Cu–Pd
alloy shows a set of peaks located at positions intermediate between
3
9
40
26
those of Cu and those of Pd ͑at variance with Milhano et al.,
who reported the Cu–Pd alloys deposited from perchlorate solutions
to be amorphous͒. Both Cu and Pd crystallize in the cubic system,
and their lattice parameters are 3.6150 and 3.8902 Å, respectively.
The lattice parameter of the Cu–Pd alloy was calculated to be 3.670,
which, assuming that the lattice parameter varies linearly with the
Figure 9. SEM images, at two different magnifications, of a fracture section
of a ͑Cu–Pd͒ + PdO composite.
Downloaded on 2015-06-24 to IP 128.210.126.199 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D254
Journal of The Electrochemical Society, 157 ͑4͒ D248-D255 ͑2010͒
0
0
0
.02
.01
.00
A: PdO
B: Cu-Pd alloy, 23 at% Pd (EDX)
C: (Cu-Pd)+PdO, 16 at% Pd (EDX)
D: (Cu-Pd)+PdO, 20 at% Pd (EDX)
6000
(
100)
-
-
-
-
-
-
0.01
0.02
0.03
0.04
0.05
0.06
C
5
4
3
2
1
000
000
000
000
000
0
A
(
200)
D
A: Thermal PdO
(
222)
(
220)
B: Electrochemical PdO
C: Cu deposit
(311)
B
D
D: (Cu-Pd)+PdO composite
-
1.2
-0.8
-0.4
0.0
0.4
0.8
E vs. SCE / V
C
Figure 10. CVs obtained with thermally or electrochemically oxidized Pd,
Cu, or ͑Cu–Pd͒ + PdO electrodes in 0.5 M Na SO ; sweep rate of
2
4
−
1
1
00 mV s .
B
the possible deviations from Vegard’s law. The diffractograms of the
Cu–Pd͒ + PdO composites, which contained 16 and 20 atom % Pd,
A
͑
respectively ͑determined by EDX͒, showed peaks at positions fairly
close to those of pure Cu. Calculations of the lattice parameters and,
hence, of the composition of the crystalline component of the de-
posits yielded Cu98.3Pd1.7 and Cu97.8Pd2.2 for their matrices. There-
fore, one can conclude that only a minor fraction of the Pd atoms
contained in the ͑Cu–Pd͒ + PdO composites were present as alloyed
Pd͑0͒ in the matrix, while the large majority was present as embed-
ded PdO, or possibly as Pd generated by PdO reduction. The fact
that the reflections typical of PdO ͑notably the most intense one
expected at 33.84°͒ were absent from the diffractograms ͓͑C͒ and
20 30 40 50 60 70 80 90 100
2
θ / Degrees
Figure 11. X-ray diffractograms of ͑A͒ as-purchased PdO powder, ͑B͒ elec-
trodeposited Cu–Pd alloy, and ͓͑C͒ and ͑D͔͒ electrodeposited ͑Cu–Pd͒
PdO composites. The overall Pd contents in the electrodeposits, deter-
mined by EDX, were ͑B͒ 23, ͑C͒ 16, and ͑D͒ 20 atom %. The diffractograms
are shifted along the intensity axis for the sake of clarity.
+
͑
D͔͒ is not surprising if the quasi-amorphous nature of PdO is con-
sidered.
The XRD data in Fig. 11 allow us to rule out the hypothesis that
ous variation in the composition of the deposition baths and there-
fore in the composition and properties of the materials deposited
therefrom.
the Cu–Pd alloys obtained in the presence of suspended PdO are
much richer in alloyed Pd than those obtained in its absence. In fact,
the Pd content of the matrices is comparable with those of alloys
obtained in the absence of suspended particles. Thus, one must con-
clude that the codeposition of PdO particles becomes much more
effective as soon as matrices consisting essentially of Cu contain as
little as 1–2 atom % Pd. To the best of our knowledge, this is an
uncommon phenomenon in composite electrodeposition, which can
2
+
5
. Unexpectedly, the presence of low concentrations of Pd ͑2
2
+
orders of magnitude below ͓Cu ͔͒ strongly favors the codeposition
of PdO. ͑Cu–Pd͒ + PdO composites with overall Pd content above
2
0 atom % are thus easily obtained; in these deposits, ca. 1/10 of the
Pd atoms are alloyed with Cu in the matrix, the others being present
as PdO particles, possibly partly reduced to Pd.
2
+
be tentatively explained by supposing that the reduction of Pd ions
adsorbed at the PdO surface greatly facilitates the PdO incorporation
The catalytic properties of the ͑Cu–Pd͒ + PdO composites will
be tested in a successive stage of this project.
11
in the matrix.
Acknowledgments
Conclusions
The authors acknowledge the financial support of the Italian
Ministry for Economic Development ͑MSE͒, MSE-CNR Agreement
on National Electrical System. The authors are indebted to the FILA
INDUSTRIA CHIMICA SPA, San Martino di Lupari, Padova, Italy,
owner of the Fei-ESem FEI Quanta 200 FEG instrument for allow-
ing its use for the research work described in this paper.
An exploratory study on the possibility of codepositing PdO par-
ticles with metal matrices, so as to obtain porous composite materi-
als potentially able to catalyze the low temperature combustion of
methane, has led to the following conclusions:
1
. Codeposition of PdO with a Ni matrix is precluded because at
2
+
the fairly negative potentials at which Ni is reduced, PdO is elec-
troactive; its cathodic reactions and, most probably, water reduction
at Pd centers prevent the effective deposition of a robust Ni matrix
able to incorporate Pd.
References
1
2
. T. V. Choudhary, S. Banerjee, and V. R. Choudhary, Appl. Catal., A, 234, 1 ͑2002͒.
. P. Gélin and M. Primet, Appl. Catal., B, 39, 1 ͑2002͒.
3. S. Cattarin and M. Musiani, Electrochim. Acta, 52, 2796 ͑2007͒.
4
. L. Vázquez-Gómez, S. Cattarin, P. Guerriero, and M. Musiani, Electrochim. Acta,
2, 8055 ͑2007͒.
. M. Musiani, F. Furlanetto, and P. Guerriero, J. Electroanal. Chem., 440, 131
1997͒.
6. M. Musiani, F. Furlanetto, and R. Bertoncello, J. Electroanal. Chem., 465, 160
1999͒.
2
. Codeposition of PdO with a Cu matrix is possible, both from
basic pyrophosphate and acid sulfate baths.
. The deposits obtained from basic pyrophosphate baths are
5
5
3
͑
brittle, have scanty adherence to the substrates, and have low Pd
content ͑typically Ͻ2 atom %͒. Furthermore, these baths are diffi-
cult to handle.
͑
7
8
9
. M. Musiani and P. Guerriero, J. Electrochem. Soc., 145, 549 ͑1998͒.
. M. Musiani, F. Furlanetto, and P. Guerriero, J. Electrochem. Soc., 145, 555 ͑1998͒.
. C. Iwakura, N. Furukawa, and M. Tanaka, Electrochim. Acta, 37, 757 ͑1992͒.
4
. PdO is not fully stable in the acid sulfate bath as it undergoes
2
+
dissolution to give Pd . The dissolution process causes a continu-
10. C. Iwakura, M. Tanaka, S. Nakamatsu, H. Noue, M. Matsuoka, and N. Furukawa,
Downloaded on 2015-06-24 to IP 128.210.126.199 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, 157 ͑4͒ D248-D255 ͑2010͒
D255
Electrochim. Acta, 40, 977 ͑1995͒.
1. J. R. Roos, J. P. Celis, J. Fransaer, and C. Buelens, J. Met., 42, 60 ͑1990͒.
2. J. P. Celis, J. R. Roos, C. Buelens, and J. Fransaer, Trans. Inst. Met. Finish, 69, 133
1991͒.
3. C. Buelens, J. Fransaer, J. P. Celis, and J. R. Roos, Bull. Electrochem., 8, 371
1992͒.
25. D.-L. Lu, M. Ichihara, and K. Tanaka, Electrochim. Acta, 43, 2325 ͑1998͒.
1
1
26. C. Milhano and D. Pletcher, J. Electroanal. Chem., 614, 24 ͑2008͒.
27. J. Simonet, P. Poizot, and L. Laffont, J. Electroanal. Chem., 591, 19 ͑2006͒.
28. N. Guglielmi, J. Electrochem. Soc., 119, 1009 ͑1972͒.
͑
1
2
9. T. W. Andersen, C. H. Pitt, and L. S. Livingston, J. Appl. Electrochem., 13, 429
1983͒.
͑
͑
14. J. Fransaer, J. P. Celis, and J. R. Roos, Met. Finish., 91, 97 ͑1993͒.
15. A. Hovestad and L. J. J. Jansen, J. Appl. Electrochem., 25, 519 ͑1995͒.
16. C. T. J. Low, R. G. A. Wills, and F. C. Walsh, Surf. Coat. Technol., 201, 371
3
3
3
0. H. J. Miao and D. L. Piron, Electrochim. Acta, 38, 1079 ͑1993͒.
1. S. Härtinger and K. Doblhofer, J. Electroanal. Chem., 380, 185 ͑1995͒.
2. J.-P. Candy, P. Fouilloux, M. Keddam, and H. Takenouti, Electrochim. Acta, 26,
͑
2006͒.
1
029 ͑1981͒.
1
7. A. Hovestad and L. J. J. Jansen, in Modern Aspects of Electrochemistry, B. E.
Conway, C. G. Vayenas, R. E. White, and M. E. Gamboa-Adelco, Editors, Vol. 38,
p. 475, Kluwer, New York ͑2005͒.
3
3
3
3
3. R. de Levie, in Advances in Electrochemistry and Electrochemical Engineering, P.
Delahay, Editor, Vol. VI, p. 329, Interscience, New York ͑1967͒.
4. Y. Gourbeyre, B. Tribollet, C. Dagbert, and L. Hyspecka, J. Electrochem. Soc.,
1
1
8. M. Musiani, Electrochim. Acta, 45, 3397 ͑2000͒.
9. A. K. Graham, Electroplating Engineering Handbook, pp. 231–257, Van Nostrand
1
53, B162 ͑2006͒.
5. M. E. Orazem, N. Pébère, and B. Tribollet, J. Electrochem. Soc., 153, B129
2006͒.
Reinhold, New York ͑1971͒.
͑
2
2
0. M. C. Marion, E. Garbowski, and M. Primet, J. Chem. Soc., Faraday Trans., 86,
6. G. J. Brug, A. L. G. van den Eeden, M. Sluyters-Rehbach, and J. H. Sluyters, J.
3
027 ͑1990͒.
Electroanal. Chem. Interfacial Electrochem., 176, 275 ͑1984͒.
37. M. Grdeń, M. Łukaszewski, G. Jerkiewicz, and A. Czerwiński, Electrochim. Acta,
53, 7583 ͑2008͒.
38. ICDD card 41-1107 ͑1989͒.
39. ICDD card 04-0836 ͑1953͒.
1. J. G. McCarthy, Y.-F. Chang, V. L. Wong, and M. E. Johansson, Prepr., Pap. - Am.
Chem. Soc., Div. Fuel Chem., 42, 158 ͑1997͒.
2
2
2. T. Ishihara, H. Shigematsu, Y. Abe, and Y. Takita, Chem. Lett., 3, 407 ͑1993͒.
3. M. Pourbaix, Atlas d’Equilibres Electrochimiques, p. 358, Gauthier-Villars & Cie,
Paris, France ͑1963͒.
2
4. R. Lacmann, Electrochim. Acta, 34, 1225 ͑1989͒.
40. ICDD card 46-1043 ͑1993͒.
Downloaded on 2015-06-24 to IP 128.210.126.199 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).