Specific features of the formation of Pt(Cu) catalysts by galvanic displacement with carbon nanowalls used as support

Article abstract of DOI:10.1016/j.electacta.2012.04.124

Microamounts of Cu are applied by the methods of electrodeposition (Cu _ed) and magnetron sputtering (Cu_spr) on a new carbon material, carbon nanowalls (CNW). The galvanic displacement (GD) of Cu _ed and Cu_spr in a PtCl₄^2- solution (with 0.5 M H₂SO₄ as the supporting electrolyte) produces Pt(Cu)/CNW catalysts. The possibility of using open-circuit potential transients recorded in the course of GD for monitoring the surface layer composition is considered. The stable Pt(Cu)_st samples are characterized by several methods (SEM, TEM, XPS, voltammetry, etc.). It is shown that Pt(Cu)_st has structure of the core(Pt, Cu)-shell(Pt) type with the average atomic ratio Pt:Cu (%) ～ 57:43 for Cu_ed and ～80:20 for Cu_spr. The formation of the dense Pt shell is also confirmed by the data on the electrocatalytic activity of synthesized samples in the methanol oxidation reaction. The reasons for deviations in the properties of Pt(Cu) _st/CNW samples formed from Cu_ed and Cu_spr are discussed. The high specific surfaces of the Pt(Cu)_st/CNW catalyst obtained from Cu_ed (>40 m²/g Pt) with the simultaneous decrease in the Pt content makes this material promising for using in the platinum-catalyzed processes (particularly, in fuel cells).

Full text of DOI:10.1016/j.electacta.2012.04.124

Electrochimica Acta 76 (2012) 137–144
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Specific features of the formation of Pt(Cu) catalysts by galvanic displacement
with carbon nanowalls used as support
B.I. Podlovchenkoa,∗,1, V.A. Krivchenkob, Yu.M. Maksimova, T.D. Gladyshevaa, L.V. Yashinaa,
S.A. Evlashinb, A.A. Pilevskyb
a Department of Chemistry, Moscow State University, Moscow 119992, Leninskie Gory, Russia
b Institute of Nuclear Physics, Moscow State University, Moscow 119991, Leninskie Gory, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
Microamounts of Cu are applied by the methods of electrodeposition (Cued) and magnetron sputter-
ing (Cuspr) on a new carbon material, carbon nanowalls (CNW). The galvanic displacement (GD) of Cued
and Cuspr in a PtCl4 solution (with 0.5 M H2SO4 as the supporting electrolyte) produces Pt(Cu)/CNW
Received 18 November 2011
Received in revised form 17 April 2012
Accepted 18 April 2012
2−
catalysts. The possibility of using open-circuit potential transients recorded in the course of GD for mon-
itoring the surface layer composition is considered. The stable Pt(Cu)st samples are characterized by
several methods (SEM, TEM, XPS, voltammetry, etc.). It is shown that Pt(Cu)st has structure of the core(Pt,
Cu)–shell(Pt) type with the average atomic ratio Pt:Cu (%) ∼ 57:43 for Cued and ∼80:20 for Cuspr. The for-
mation of the dense Pt shell is also confirmed by the data on the electrocatalytic activity of synthesized
samples in the methanol oxidation reaction. The reasons for deviations in the properties of Pt(Cu)st/CNW
samples formed from Cued and Cuspr are discussed. The high specific surfaces of the Pt(Cu)st/CNW catalyst
obtained from Cued (>40 m2/g Pt) with the simultaneous decrease in the Pt content makes this material
promising for using in the platinum-catalyzed processes (particularly, in fuel cells).
Available online 14 May 2012
Keywords:
Carbon nanowalls (CNWs)
Copper
Platinum
Galvanic displacement
Core–shell structure
Electrocatalytic properties
© 2012 Elsevier Ltd. All rights reserved.
1. Introduction
with the lower platinum content to be obtained. Particularly, the
core(Pt, M1)–shell(Pt) particles were formed [7–10,13,14]. In acidic
The keen attention paid to the fuel cell problem led to the
development of many methods for synthesizing nanostructured
electrocatalytsts based on platinum-group metals [1–3]. However,
due to the high costs and the shortage of noble metals, the active
quest for the ways of lowering down their content in catalysts is
still continued. In doing this, the preparation of catalyst structures
by means of contact displacement (replacement) of a less noble
(less electropositive) metal M1 by a noble (more electropositive)
metal M2 under the open circuit conditions seems attractive [4–14].
The method of galvanic displacement (GD) is based on the reac-
solutions, copper was used most often as metal M1, because the
potential of the Cu/Cu2+ pair is sufficiently low but at the same
time more positive that the potential of the standard hydrogen
electrode; and, hence, the displacement of copper is uncompli-
cated by the parallel reaction of hydrogen evolution. The specific
features of the interaction of carbon-supported copper nanolayers
2−
or nanoparticles with PtCl6 anions were studied [8–10]. As the
supports, glassy carbon (GC) [8,9] and Vulcan XC-72 soot layers
applied on GC [10] were used. The gradual formation of the sta-
ble core(Pt, Cu)–shell(Pt) structures with the average Pt content
of 70–75 at.% was observed. However, the electrochemically active
specific surfaces of Pt(Cu) deposits were relatively small, 5–15 m2/g
Pt.
z+
0
tion: M1 + (n/z)M2 → M1n+ + (n/z)M2 as a result of which metal
M1 (from submonolayer to nanosized layer) is displaced partially
or completely by metal M2.
In electrodes of low-temperature fuel cells, the catalytic layer is
often represented by Pt nanoparticles deposited on a porous car-
bon support. The formation of nanolayers or nanoparticles of metal
It is known that the support material can substantially affect
the properties of synthesized electrocatalysts (catalysts) such as
the average diameter of metal particles, their size distribution,
composition, stability, etc. Carbon nanotubes (CNT) are a suffi-
ciently new electrode material [15–20] that has already found its
use as the supports in the synthesis of different catalysts. How-
ever, the CNT may include traces of metals introduced by catalysts
(Ni, Fe, Co, etc.) and also nanographite particles [18,20]. This makes
their use difficult, especially, in the model systems employed in
solving fundamental problems associated with the synthesis of
M1 on this support and their subsequent displacement by platinum
(in the course of GD) allowed highly active bicomponent catalysts
∗
Corresponding author. Tel.: +7 495 939 4027; fax: +7 495 932 8846.
E-mail address: podlov@elch.chem.msu.ru (B.I. Podlovchenko).
ISE member.
1
0013-4686/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.electacta.2012.04.124

138
B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
catalysts (kinetics and mechanism of processes, structure optimiza-
tion, enhancing the activity, etc.). The recently synthesized carbon
nanowalls (CNW) prepared in the absence of catalysts by means of
the plasma enhanced chemical vapor deposition (PECVD) [21–25]
are very promising. The PECVD grown carbon nanowalls exhibited
the high electric conductivity and the large specific surfaces.
In the present study, CNW were used as the supports in syn-
thesizing the Pt(Cu)/C catalyst by the galvanic displacement of
preliminarily deposited copper. Taking into account the peculiar-
ities of the CNW structure (the presence of plate-like crystals)
[21–25], it seemed interesting to elucidate how the way of cop-
per deposition, namely, by electrodeposition as individual particles
platinum ions was broken and the cell was washed with supporting
electrolyte solution in the Ar flow. The stationarity criterion corre-
sponded to the potential variation by less than 1 mV/min. The real
surface of platinum was estimated based on the desorption currents
of upd copper [27–29]. As was shown in [28], for small Pt loads on
the carbon support with a high intrinsic surface, the determination
of the real surface of Pt based on Had becomes less reliable as com-
pared with its determination based on Cuad. This is associated with
the fact that the high intrinsic capacitance of the carbon support
can substantially change with the Pt deposition, which introduces
an error into the estimate of the Pt real surface. To determine the Pt
surface based on Cuad, one should subtract an I vs. E curve for Pt/C
(and not for C) from an I vs. E curve for CuadPt/C, i.e., the change
in the capacitance of the carbon part of the surface is taken into
account.
In experiments with Cuspr samples, the sample was placed in
deaerated supporting electrolyte solution, the electrode potential
was stabilized at 0.2 V, then the circuit was opened and the chloro-
platinite solution was added. The amount of Cuspr was determined
in a separate experiment for a sample taken from the same lot of
Cuspr/CNW electrodes used in our experiments on the galvanic dis-
placement of Cuspr. After the stabilization of its potential at 0.2 V,
an anodic potentiodynamic curve was recorded, and the amount of
deposited copper was estimated from the copper desorption peak.
To obtain electrodeposited (e.d.) Pt (or Pted) on CNW supports,
the same solution 0.5 M H2SO4 + 2 × 10−3 M H2PtCl4 was used. After
the electrode was pretreated by the same procedure as that used in
electroplating of Cu on CNW, the 0.5 M H2SO4 solution was changed
for the platinite solution and Pt was deposited with the cathodic
current of 0.1 mA.
or by vacuum splaying of an even layer, affects the formation and
2−
properties of Pt(Cu) particles. Anions PtCl4
were chosen as the
displacement agent, because in their presence copper was dis-
2−
placed in the ratio 1:1 (for PtCl6 Cu:Pt = 2:1) and their reduction
can proceed only to Pt (for PtCl62−, the formation of a sufficiently
2−
stable intermediate, namely, PtCl4 is possible).
2. Experimental
Films composed of dense array of CNW were used as the sup-
ports for platinum catalysts. CNW were grown on glassy carbon
substrates in plasma of dc glow discharge in a mixture of hydrogen
and methane. Details of experiments are described elsewhere [23].
The total geometric area of each GC plates was equal to 1 cm2.
After the synthesis of a CNW film on a GC plate, one side of
the plate was covered with polystyrene, i.e., the working (accessi-
ble to polarization) surface was 0.5 cm2. Platinum was applied on
the CNW layer by the galvanic displacement of either preliminar-
ily electrodeposited copper (Cued) or a copper layer sprayed on the
support (Cuspr) by the magnetron sputtering method. The solution
containing Pt ions was 2 × 10−3 M K2PtCl4 + 0.5 M H2SO4. The salt
K2PtCl4 was synthesized by the reduction of K2PtCl6 with hydrazine
sulfate [26]. In all experiments, 0.5 M H2SO4 served as the support-
ing electrolyte (H2SO4 was of the “e.p. (Across)” grade, water was
cleaned in the Milli-Q system). The measurements were carried out
at 19 1 ◦C. All potentials were related to the reversible hydrogen
electrode in the same solution (RHE).
Prior to electroplating Cu or Pt, CNW were subjected to the
electrochemical treatment. First of all, the electrode was polar-
ized anodically in the 0.5 M H2SO4 supporting electrolyte solution
at 1.1 V in the Ar flow for 10 min; then, after the changeover of
solution, the electrode potential was brought to 0 by applying a
small cathodic current. Then, the potential of 60 mV was estab-
lished (the background current Ibg did not exceed 10 ␮A) and the
anodic I vs. E curve was measured with the potential scan rate
ꢀ = 6.6 mV/s (the same ꢀ was used in recording the other potentio-
dynamic curves). Cu electrodeposits (Cued) were applied from the
5 × 10−3 M CuSO4 + 0.5 M H2SO4 solution in the galvanostatic mode.
The electrode potential was preliminarily stabilized in the support-
ing electrolyte solution at ∼0.4 V (Ibg < 5 ␮A); then, the supporting
electrolyte solution was changed in the Ar flow for a solution con-
taining copper ions and the cathodic current of 0.2 mA was applied.
The amounts of deposited copper estimated based on either the
cathodic currents of Cu2+ deposition or the anodic currents of
deposit dissolution virtually coincided ( 3%). After the deposition
of Cued, the cell was washed with deaerated supporting electrolyte
solution 7–8 times in the Ar flow while the potential was poten-
tiostatically maintained at E ∼ 0.2 V. Then, the circuit was opened
and the supporting electrolyte solution in the cell working part was
changed for the preliminarily deaerated 0.5 M H2SO4 + 2 × 10−3 M
H2PtCl4 solution. The time variations of the electrode potential
were recorded up to the establishment of its stationary value
Est. Upon reaching Est, the contact with the solution containing
The electrocatalytic activity of the resulting Pt(Cu) and Pt
deposits was tested in the reaction of methanol electrooxida-
tion by measuring stationary polarization curves in the 0.5 M
CH3OH + 0.5 M H2SO4 solution. The stationarity criterion was the
current change by less than 1%/min at a fixed E.
The morphology and the structural properties of samples were
studied by means of scanning electron microscopy (SEM) and trans-
mission electron microscopy (TEM). SEM images were obtained on
a scanning electron microscope JEOL JSM-6490 LV with accelerat-
ing voltage of 30 kV. Transmission electron microscope JEOL JEM
2100F with accelerating voltage of 200 kV was used in order to
obtain TEM images. The high resolution X-ray photoelectron spec-
troscopy (XPS) and the X-ray fluorescence (XRF) analysis were used
for studying the chemical composition of catalysts after the depo-
sition of platinum. XRF analysis was carried out on Inca Energy
350XT system. X-ray photoelectron spectra were recorded by a
spectrometer Axis Ultra DLD (Kratos) in monochromatic Al K␣ radi-
ation with the analyzer energy transmission of 20 eV. The analyzed
region was 300 × 700 ␮m. The preliminary energy scale calibration
corresponded to the following peaks of reference samples (ion-
bombardment-cleaned metal surface): Au 4f5/2 – 83.96 eV, Cu 2p3/2
– 932.62 eV, Ag 3d5/2 – 368.21 eV, Pt 4f7/2 – 90.99 eV with the accu-
racy within to 0.07 eV.
3. Results and discussion
3.1. Morphology of CNW and CNW-supported copper deposits
According to Fig. 1a, the original support material repre-
sents dens array of nanowalls preferentially oriented normally to
towards the surface. The average height of the nanowalls is approx-
imately ∼1 ␮m and their width is 0.5–1.0 ␮m. TEM study (insert in
Fig. 1) reveals that the average nanowall thickness is about 4–6 nm.
According to gravimetric data, the specific mass of the CNW film is
about 5 × 10−6 g/cm2.

B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
139
Fig. 1. SEM images of samples (a and b) as-grown CNW (inset – TEM image of the CNW); (c and d) Cued/CNW (48 ␮g/cm2 geom. surf.); (e and f) Cuspr/CNW (54 ␮g/cm2 geom.
surf.).
Copper electrodeposits (e.d.) on CNW (Fig. 1c and d) represent
crystals with a rather wide scatter in the size, from ∼20 nm to
∼200 nm. Copper was preferentially deposited on the edges of CNW
and in much smaller amounts on their sides (faces); moreover, finer
crystals were formed on the faces.
Fig. 1e and f shows the surface images of CNW covered with
magnetron-sputtered copper. It is seen that copper is distributed
evenly over both edges and faces and only strong magnification
revealed the presence of cracks (<5 nm) on certain faces.
The further run of curves allows the following three segments
to be singled out: a, b and c. The slow potential shift in segment a
is apparently associated with the increase in the Cu2+ concentra-
tion with the displacement of Cu according to reaction (1), where
Cu2+ + 2e ↔ Cu remains the potential-determining stage. In seg-
ment c, the potential reaches its stationary value Est (>0.8 V) which
corresponds to potentials of the beginning of oxygen adsorption
on platinum. This fact alone allows one to assume the absence of
considerable amounts of copper on the surface of Pt(Cu)st particles
formed at the stationary potential. At these potentials, even the
copper adatoms (Cuad) most strongly bound to the surface desorbed
[30,31]. The potential increase observed at E > 0.7 V was mainly
determined by the increase in the total charge (Q [32,33]) of the
platinum surface and unoccupied support areas by the reaction
3.2. Transients of open-circuit potential
Fig.
2 compares the open-circuit potential transients for
Cued/CNW and Cuspr/CNW electrodes recorded upon their contact
2−
with the solution containing PtCl4
anions (curves 1 and 2). In
PtCl42− + 2e → Pt0 + 4Cl−.
(2)
the first moment after the contact, a small but very quick potential
surge was observed, which was associated with the sharp increase
in the Cu2+ concentration in the near-electrode layer as the result
of the reaction
Thus, after the complete displacement of Cu and Cuad (E > 0.7 V),
under open circuit conditions, the electrons for reaction (2) are
supplied by the free charge [30,31] and the partial surface oxida-
tion of Pt(Cu) particles and/or support regions unoccupied by metal
particles.
Cu + PtCl42− → Cu2+ + Pt0 + 4Cl−.
(1)
However, on reaching sufficiently high potential values, the oxi-
2−
2−
dation of PtCl4 to PtCl6 starts
2−
PtCl42− + 2Cl− − 2e ↔ PtCl6
.
(3)
Assume that the stationary potential (Est) of Pt(Cu)st particles is
mainly determined by the equilibrium in reaction (3). In the first
approximation which ignores the activity coefficients of compo-
nents of reaction (3), we can write its equilibrium potential (Eeq) as
follows:
cPtCl
cPtCl4 cC2l
2.3RT
2F
6
Eeq = E0
+
log
(4)
where E0 is the standard potential equal to 0.672 V [34]; cPtCl , cPtCl
6
4
and cCl are the solution concentrations of PtCl62−, PtCl4 and Cl−,
respectively; 2.3RT/2F ≈ 0.028 V at 293 K. It is seen that Eeq of reac-
tion (3) depends strongly on the concentration of chloride anions.
We assume that Cl− anions are mainly formed in reaction (1) and
the degree of copper displacement is ∼70% [8–10]. Then, for the
initial copper content of ∼0.03 mg and the cell volume of 10 ml,
the chloride concentration is assessed as cCl ≈ 6 × 10−4 M. Accord-
2−
Fig. 2. Transients of open-circuit potential recorded upon bringing (1 and 2)
Cu/CNW- and (3) Pt/CNW electrodes into contact with 2 × 10−3 M H2PtCl4 solution:
ing to Eq. (4) for cPtCl = 2 × 10−4 M and Eeq = Est ≈ 0.9 V, we find
(1) Cued, (2) Cuspr
.
4

140
B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
Fig. 3. SEM images of Pt(Cu)st/CNW: (a and b) after the displacement of Cued and (c and d) after the displacement of Cuspr
.
cPtCl ≈ 1.4 × 10−5 M. As seen, the oxidation of even a small part
the relatively fast potential variation in this region was mainly due
to the increase in the total surface charge Q as a result of reaction
(2) in the absence of considerable oxygen adsorption [32,33].
6
2−
2−
of PtCl4 anions to PtCl6 is in principle sufficient for potentials
corresponding to the onset of oxygen adsorption on platinum (its
oxidation) to be established on the Pt(Cu)/CNW electrode. Indeed,
this is a rough estimate due to the very approximate nature of sev-
eral assumptions made and, particularly, because the formation of
Cl− anions by reaction (2) was neglected. The calculations only con-
3.3. Morphology and composition of mixed Pt(Cu)st deposits.
Fig. 3 shows the surface images of Pt(Cu)st/CNW samples
synthesized upon the displacement of either e.d. Cu (a and b)
or a layer of sprayed copper (c and d). The comparison of
Fig. 1(c,d) and Fig. 3(a,b) demonstrate that the metal particles are
still preferentially accumulated on CNW edges but, on the whole,
the number of particle increases and their size decreases. The dis-
integration of original Cu particles seems reasonable if for no other
reason than due to the increase in the mass of metal particles,
because the higher Pt atomic mass as compared with Cu is not com-
pensated by the high specific weight of Pt (for 100% displacement
of copper by platinum, the volume should increase by ∼30%). Quite
a lot of fine particles appear on the faces, coarser particles remain
on edges. Due to the high conductivity of CNW, new particles on the
faces can form due to the Pt deposition (reaction (2)) on the carbon
surface areas substantially remote from the copper ionization sites
(Cu − 2e → Cu2+). However, this does not allow us to assume a pri-
ory that such particles should contain no copper at all. The copper
firmed that the role of reaction (3) in the formation of the stationary
2−
potential of Pt(Cu)st particles in the presence of PtCl4
anions
may be substantial. On the whole, this is mixed a potential the
establishment of which depends also on such factors as the begin-
ning of the irreversible oxygen adsorption on Pt, the possibility of
the formation of carbon oxides, the presence of microimpurities,
e.g., dissolved oxygen traces that can be found even in thoroughly
deaerated solutions.
Curve 3 in Fig. 2 is a transient of the open-circuit potential
recorded in 0.5 M H2SO4 + 2 × 10−3 M H2PtCl4 solution on a CNW-
supported e.d. Pt (48 ␮g/cm2 of geometric surface) in the absence
of copper on the surface. In this case, the potential increase was
associated only with the increase in the total surface charge Q by to
reaction (2). The gradually establishing Est also lies in the potential
region of the beginning of oxygen adsorption on Pt. This confirms
the assumption that Est on Pt(Cu)st particles (curves 1 and 2) was
determined by the processes that occur of the platinum surface.
The fact that Est on e.d. Pt/CNW is even lower than on
Pt(Cu)st/CNW electrodes is rather unexpected. This is probably due
to the fact that in the absence of copper in solution, the Cl− anions
were formed only by reaction (2). Their lower concentration in solu-
tion and, correspondingly, the lower adsorption values should favor
the earlier adsorption of oxygen on platinum [32,35]. In particular,
this can lead to the earlier inhibition of reaction (2) as compared
with the process of formation of Pt(Cu)st particles by the displace-
ment of copper.
inclusion can occur due to the formation of copper adatoms (Cuad
)
on Pt surfaces formed at potentials <0.7 V [30,31]. This mechanism
of preparation of mixed electrodeposits was considered in [36,37]
for Pd–Cu and Pt–Cu deposits.
The displacement of sprayed copper produces a result at first
glance unexpected. Upon reaching the stationary state in the course
of displacement of copper uniformly distributed over the CNW sur-
faces, the Pt(Cu)st deposit finds itself mostly on the CNW edges
and in their vicinity (Fig. 3c and d). The metal particles on “faces”
are present in insignificant amount. It is seen that the deposit rep-
resents coarse conglomerates of fine particles. The formation of
such particles on CWN “faces”, which is followed by their travel
to edges, appears improbable taking into account the relatively
In segment b of E vs. ꢁ curves, which corresponds to the displace-
ment of copper (curves 1 and 2, Fig. 2), the reaction Cu0 ↔ Cu2+ + 2e
has already stopped to be the potential-determining. Apparently,

B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
141
Fig. 4. TEM images of Pt(Cu)st/CNW corresponding to the displacement of Cued for (a) particles with the core–shell structure, (b) particles with the homogenous structure
(some particles are marked by arrows for clarity, magnified images see in Supporting information).
short times of the mixed deposit formation. Apparently, the mech-
anism of spatially separated reactions PtCl42− + 2e → Pt0 + 4Cl− and
Cu0 − 2e → Cu2+ should be preferred. It should be assumed that
upon the formation of Pt(Cu) particles on edges and in their vicin-
ity, the further deposition of Pt mainly proceeds on these particles
due to the electrons donated by copper atoms from the copper lay-
ers on the faces. The uniform distribution of copper, on the one
hand, provides the high total currents of its dissolution and, on the
other hand, hinders the formation of Pt(Cu)st particles on the faces,
because this requires the destruction of the copper layer. Although
in the proposed mechanism of sprayed copper displacement the
incorporation of copper into the platinum deposit through cop-
per adatoms is, as was noted above, possible, one, however, should
expect the lower copper content as compared with the formation
of Pt(Cu)st particles from e.d. Cu on CNW. Indeed, this assumption
confirms the results of the XRF analysis of the average contents
of Pt and Cu in Pt(Cu)st/CNW deposits shown in Table 1 (the total
content of Pt and Cu atoms was assumed to be 100%). Among the
other factors that can explain the preferential deposition of Pt on
CNW edges at the Cuspr displacement, mention should be made of
the different electric field gradients on the edges and side walls of
observed in 5–10 min after the beginning of the corrosion process
was associated [38] with the formation of a dense platinum layer.
The TEM examination of Cued/CNW and Pt(Cu)st/CNW(from
Cued) samples was carried out. On the whole, these images con-
firmed the SEM data (Figs. 1d and 3b). For Cued/CNW, no new
information was obtained (these TEM images are omitted). Fig. 4
shows the TEM images for Pt(Cu)st/CNW, which are the most inter-
esting from our point of view. Fig. 4a demonstrates a particle
(marked with a cross) with the core–shell structure (the core with
the diameter of ∼10 nm and the shell with the thickness of ∼5 nm).
A conglomerate of such particles can be seen to the left of the former
particle.
Fig. 4b demonstrates TEM image of CNW decorated with par-
ticles with the linear size of 1–4 nm. These particles have a
homogeneous bulk structure with the interplanar spacing of about
0.21 nm. This is close to the interplanar spacing for Pt hkl (1 1 1) and
(2 0 0). Based on the XPS and electrochemical data (see below) it can
be concluded that these are Pt particles. Thus, the TEM data show
that Pt(Cu)st particles prepared by the GD method differ not only
in the size but also in the composition. In this connection, it can
be said that the application of the term “core–shell” to Pt(Cu)st/C
systems prepared by galvanic displacement is to a certain extent
conditional.
2−
CNW and also of the difference in the discharge kinetics of PtCl4
anions on Pt and Cu.
It is interesting that Pt(Cu)st particles synthesized by the GD
method from e.d. Cu on CNW contained a noticeably larger amount
of copper as compared with the case where GC [8,9] and carbon
black [10] were used as the support. The reason for this may con-
sist in the accumulation of the main mass of the original copper
deposit on CNW edges. The displacement of copper in the upper
layers of such deposit and the agglomeration of Pt(Cu) particles
formed strongly hinder the dissolution of copper located in the bulk
of coarse conglomerates. This assumption is also confirmed by the
results on the dissolution of Pt–Cu alloys in sulfuric acid solution
saturated with air [38]. The sharp inhibition of copper dissolution
The specific surfaces of Pt(Cu)st/CNW (from Cued) samples
related to 1 g Pt in the deposit were 46( 15) m2/g, whereas for
the directly deposited Pt, the surface was 2–3 times smaller (see
Table 1). Earlier, the much lower S values were obtained for the
galvanic displacement of Cued from GC [8,9] and Vulcan XC-72 soot
[10] (see Section 1). Apparently, this is associated with the higher
dispersivity of Cued on CNW. In line with this, the displacement
of Cuspr led to the formation of Pt(Cu)st with the lower specific
surfaces (Table 1).
3.4. The composition of the surface layer of Pt(Cu)st particles (XPS
analysis)
Table 1
Characteristics of Pt(Cu)st/CNW electrodes: mCu, the original mass of Cu; Pt:Cu, the
composition of the Pt(Cu)st deposit; mPt, the amount of Pt in Pt(Cu)st deposit; S,
specific surface.
The XPS results point to the formation of a continuous sur-
face layer of metal platinum with the thickness of 2–3 nm (the
atomic ratio Cu/Pt ≈ 1/99). Analytical Cu 2p peak on the general X-
ray photoelectron spectrum (not shown) is virtually absent. Fig. 5
demonstrates Cu 2p and Pt 4f spectra. The continuity of surface layer
was attested by the fact that copper was detected in its unoxidized
state. The Cu 2p peak is described by a single doublet with the bind-
ing energy BE(Cu 2p3/2) = 931.76 eV. The formation of core–shell
structures can be concluded based on the comparison of atomic
mCu (␮g) (deposit type)
Pt:Cu (at.%)
mPt (␮g)
SPt (m2/g)
24.5 (e.d.)
24.3 (e.d.)
24.0 (e.d.)
14.0 (spr)
27.6 (spr)
Pt (e.d. without Cu)
56:44
54:46
60:40
82:18
78:22
43.3
39.0
44.2
43.0
66.1
40.0
61.2
45.6
31.0
17.1
10.9
20.1

142
B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
Fig. 5. X-ray photoelectron spectra of Pt(Cu)st/CNW: (a) Cu 2p; (b) Pt 4f. Points – data obtained experimentally, line – fitting with the Gaussian/Lorentzian convolution.
ratios of metals found by a volume method (XPF) and a method for
analyzing the surface (XPS). These results suggest that platinum is
largely accumulated in the surface layer.
It should be noted that the Cu 2p3/2 peak is shifted by ca. 1 eV
to the lower binding energies, which is considerably lower as com-
pared with elemental copper. This may indicate that the latter is
accumulated in the “core” in its intermetallic state.
The analysis of Pt 4f spectrum revealed that the found param-
eters are completely identical to those obtained for pure platinum
foil; the peak corresponds to the energy of 70.9 eV and is charac-
terized by the asymmetry parameter ˛ = 0.22, i.e., pertains to the
volume state. It is well known from the literature [39,40] that the
decrease in the size of platinum clusters from 5 to 1 nm can shift the
Pt 4f7/2 peak to the higher binding energies by ∼1 eV and decreases
asymmetry. The absence of these effects suggests that in the mate-
rial under study, platinum particles with sizes fitting this range are
preferentially in the electric contact with one another; hence, one
can speak of their “conglomeration”.
Fig. 6. Anodic potentiodynamic curves of (1) Pted/CNW and (2 and 3) Pt(Cu)st/CNW
3.5. Voltammetric characteristics of Pt(Cu)st/CNW electrodes.
electrodes: (1) e.d. Pt (80 ␮g/cm2 of geometric surface); (2 and 3) prepared from: (2)
Cued (48 ␮g/cm2 of geometric surface), (3) Cuspr (54 ␮g/cm2 of geometric surface).
Fig. 6 shows that the potentiodynamic curves of Pt(Cu)st/CNW
Supporting electrolyte 0.5 M H2SO4. Insert: parts of anodic potentiodynamic curves
electrodes obtained by the galvanic displacement of either e.d. Cu
(curve 2) or sprayed Cu (curve 3) are very close to the similar curve
for e.d. Pt on CNW (curve 1). Both curves demonstrate well pro-
nounced hydrogen regions; their double-layer regions contain no
peaks or waves of ionization of bulk and/or upd copper. Thus, the
electrochemical results together with the XPS data point to the
formation of the shell(Pt)–core(Pt, Cu) structures and the virtually
total absence of copper in the surface layer of particles.
for Pt(Cu)st/CNW(from Cued): (1) background, (2) after the electrode exposure to
0.5 M H2SO4 solution at 0.3 V for 3 h (see explanations in the text).
The anodic potentiodynamic curves of the desorption of the
monolayer of copper adatoms from two types of Pt(Cu)st/CNW elec-
trodes and also from the Pted/CNW electrode are also sufficiently
close (Fig. 7).
To assess the degree of permeability of the Pt shell for copper
atoms in the core, we considered the peculiarities of the behavior
of copper adatoms on a Pt electrode. As it was shown in [28,30], the
Cuad monolayer (Cuad ML) accumulated the e.d. Pt remains on the
surface even after the repeated washings of both the electrode and
the cell working part with the thoroughly deaerated supporting
electrolyte (sulfuric acid) solution. Hence, were the shell perme-
able, the stabilization the Pt(Cu)st/CNW electrode at the potential
of the formation of Cuad ML would result in the exit of copper to
the surface in the form of adatoms. The anodic potentiodynamic
curve of the Pt(Cu)st/CNW electrode (with originally e.d. copper)
measured after the exposure of the electrode to 0.5 M H2SO4 at
E = 300 mV for 3 h (insert in Fig. 6) demonstrated the appearance of
Fig. 7. Anodic potentiodynamic curves for Pted/CNW (1 and 4) and Pt(Cu)st/CNW (2
and 3) in the presence (1–3) and in the absence (4) of Cuad on the surface: (2) Cued
(3) Cuspr
,
.

B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
143
conclusion [10] that the presence of copper in the cores of Pt(Cu)st
particles has no activating effect on the stationary process of CH3OH
electrooxidation.
From our point of view, the results of polarization measure-
ments in CH3OH solutions agree with the other results of the
present study and confirm the presence of a sufficiently dense Pt
layer on the surface of Pt(Cu)st particles. The fact that the CH3OH
electrooxidation currents on the Pt(Cu)st deposit proved to be
somewhat lower than on e.d. Pt are best explained by the structural
factors, particularly, the size effect. For Pt particles incorporated
into polymer matrices, the “negative” size effect (a decrease in the
process rate on small particles) was observed [43] in the reac-
tion of methanol oxidation. The comparison of SEM images of
Pt(Cu)st/CNW (Fig. 3) and Pted/CNW (Fig. 9) surfaces made it possi-
ble to conclude that Pt electroplates contained virtually no crystals
measuring less than 10 nm, whereas Pt(Cu)st deposits contained a
lot of such particles. According to SEM and TEM data (Figs. 3 and 4),
the sizes of certain Pt(Cu)st particles synthesized of GD did not
exceed 2–3 nm. The increase in activity of Pt particles in the reaction
of methanol oxidation (MOR) as their size increased from ∼2.5 to
∼6.5 nm was mentioned in [44]. The same paper cites many studies
in which the “negative” size effect was also observed for MOR on
platinum catalysts. Usually, two main factor that determine this
effect are distinguished [43,44]: (i) the large particles are more
“convenient” for multisite adsorption of the CH3OH molecule, (ii)
on small Pt particles, the O and/or OH adsorption is stronger, which
inhibits their interaction with methanol chemisorptions products.
However, the different MOR rates on Pt(Cu)st/C and Pted/C can
be associated, as well, with factors other than the size effect, for
instance, the specifics of the methanol/CO2 transport on the “stuck
together” agglomerates of Pt(Cu)st particles and/or the changes
in the ratio of stable products of CH3OH oxidation (CO2, HCOOH,
H2CO).
The strong difference in the morphology of e.d. Cu (Fig. 1c and
d) and e.d. Pt (Fig. 9) was quite unexpected. Obviously, in con-
trast to Cu, platinum is sufficiently uniformly distributed over the
CNW surface without any preferential deposition on the edges and
demonstrates no substantial scatter in the particle size for the aver-
age size of ∼35 nm. The calculation of the specific surface according
to the model of spheres gave ∼8 m2/g. This is approximately 2 times
lower than S of Pted on CNW, determined based on upd copper
desorption, although at the high magnification, one can see that
sphere-like particles with the mentioned average sizes represent
conglomerates formed by bounded together crystals with linear
size of several nm. Thus, there is a considerable differences in the
surface structures of Pt(Cu)st particles and e.d. Pt particles formed
on CNW.
Fig. 8. Steady-state polarization curves of methanol electrooxidation (0.5 M
CH3OH + 0.5 M H2SO4) on (1) Pted/CNW and (2 and 3) Pt(Cu)st/CNW: (2) Cued, (3)
Cuspr. Specific current densities (j) are given per cm2 of the true surface.
a small wave at the potentials of ionization of copper adatoms. As
regards the charge consumed in the ionization of Cuad, the average
rate of copper exit was assessed as ∼2 nA/cm2 of the real surface
area. Indeed, such a slow rate of the copper exit cannot have any
noticeable effect on the cyclic voltammograms even recorded with
the slow potential scan rate used here.
A certain difference in the adsorption of oxygen is observed
in the initial region of the oxygen potential range for three elec-
trodes under consideration (Fig. 6). This can be associated with
the different sizes and different distributions of catalyst parti-
cles over the carbon support surface. In these potential regions,
the CNW surface can also be somewhat oxidized, which is initi-
ated by the presence of catalytically active Pt particles. It is seen
that the oxide (this can also be OH species) formation [41,42]
on Pted/CNW electrodes in 0.5 M H2SO4 occurs a little earlier (at
less positive potentials) as compared with the Pt(Cu)/CNW elec-
trodes. This agrees with the assumption that the difference in Est
for Pt(Cu)st and Pt particles on CNW (Fig. 2) can be due to the differ-
ence in the oxygen adsorption at least in the range of low surface
coverages.
3.6. Electrooxidation of methanol on Pt(Cu)st/CNW electrodes
According to result shown in Fig. 8, the specific rates of methanol
electrooxidation (calculated per cm2 of the electrode surface deter-
mined based on the upd copper desorption) on the Pt(Cu)st/CNW
electrodes prepared by the displacement of Cued and Cuspr, are
close to one another. They do not exceed (and are even somewhat
lower) the specific rate of CH3OH oxidation on the Pted/CNW elec-
trode with the close amounts of deposited Pt. This agrees with the
The problem of the reason for such a different distribution of Cu
and Pt electroplates on CNW is very interesting on its own right but
Fig. 9. SEM images of e.d. Pt (80 g/cm2 geom. surf.) on CNWs.

144
B.I. Podlovchenko et al. / Electrochimica Acta 76 (2012) 137–144
goes beyond the frames of the present study. It can also be men-
tioned that the strong difference in the charge and the structure
[3] C. Conntanceau, S. Brimand, C. Lamy, J.-M. Leger, L. Dubau, S. Rousseau, F. Vigier,
Electrochimica Acta 53 (2008) 6865.
[4] S.R. Brankovic, J.X. Wang, R.R. Adzic, Surface Science 477 (2001) L173.
[5] M.B. Vukmirovic, J. Zhang, K. Sasaki, A.U. Nilekar, F. Uribe, M. Mavrikakis, R.R.
Adzic, Electrochimica Acta 52 (2007) 2257.
[6] Y. Yu, Y. Hu, X. Liu, W. Deng, X. Wang, Electrochimica Acta 54 (2009) 3092.
[7] M. Van Brussel, G. Kokkinidis, I. Vandendael, C. Buess-Herman, Electrochem-
istry Communications 4 (2002) 808.
of Cu2+ and PtCl4 ions allows one to expect a substantial differ-
2−
ence between their discharge mechanisms and, hence, between the
nucleation sites for Pt and Cu crystals.
[8] S. Papadimitriou, A. Tegou, E. Pavlidou, S. Armyanov, E. Valova, G. Kokkinidis,
S. Sotiropoulos, Electrochimica Acta 53 (2008) 6559.
4. Conclusions
[9] S. Papadimitriou, S. Armyanov, E. Valova, A. Hubin, O. Steenhaut, E. Pavlidou, G.
Kokkinidis, S. Sotiropoulos, Journal of Physical Chemistry C 114 (2010) 5217.
[10] B.I. Podlovchenko, T.D. Gladysheva, A.Yu. Filatov, L.V. Yashina, Russian Journal
of Electrochemistry 46 (2010) 1189.
[11] D. Gokcen, S.-E. Bae, S.R. Brankovic, Electrochimica Acta 56 (2011) 5545.
[12] D.-Y. Park, H.S. Jung, Y. Rheem, C.M. Hangarter, Y.-I. Lee, J.M. Ko, Y.-H. Choa,
N.V. Myung, Journal of Physical Chemistry C 114 (2010) 5217.
[13] N. Kristian, Y. Yu, J.-M. Lee, X. Liu, X. Wang, Electrochimica Acta 56 (2010) 1000.
[14] P. Wang, H. Li, H. Feng, H. Wang, L. Lei, Journal of Power Sources 195 (2010)
1099.
[15] W. Gang, X. Bo-Qing, Journal of Power Sources 174 (2007) 148.
[16] J. Prabhuram, T.S. Zhao, Z.H. Liang, R. Chen, Electrochimica Acta 52 (2007) 2649.
[17] Z. Liu, X.Y. Ling, L. Hong, J.Y. Lee, Journal of Power Sources 167 (2007) 272.
[18] J. Li, Y. Liang, G. Bing, X. Zhu, X. Tian, Electrochimica Acta 54 (2010) 1277.
[19] A.A. Mikhaylova, E.K. Tusseeva, N.A. Mayorova, A.Yu. Rychagov, Yu.M.
Volfkovich, A.V. Krestinin, O.A. Khazova, Electrochimica Acta 56 (2011) 3656.
[20] C.L. Scott, M. Pumera, Electrochemistry Communications 13 (2011) 426.
[21] Y. Ando, X. Zhao, M. Ohkohchi, Carbon 35 (1997) 153.
[22] Y.H. Wu, P.W. Qiao, T.C. Chong, Z.X. Shen, Advanced Materials 14 (2002) 64.
[23] V.A. Krivchenko, V.V. Dvorkin, N.N. Dzbanovsky, M.A. Timofeyev, A.S. Stepanov,
A.T. Rakhimov, N.V. Suetin, O.Yu. Vilkov, L.V. Yashina, Carbon 50 (2012) 1477.
[24] M. Hiramatsu, M. Hori, Carbon Nanowalls: Synthesis and Emerging Applica-
tions, Springer, NY, 2010.
[25] B. Zheng, Y. Kehan, L. Ganhua, W. Pengxiang, M. Shun, Ch. Junhong, Carbon 49
(2011) 1858.
[26] N.N. Chernyaev (Ed.), Synthesis of Complex Compounds of Platinum Group
Metals, Nauka, Moscow, 1964.
[27] T.D. Gladysheva, B.I. Podlovchenko, Z.A. Zikrina, Electrochimiya 23 (1987) 1440.
[28] T.D. Gladysheva, B.I. Podlovchenko, Vestik Moskovskogo Uni- versiteeta Seriya
2 Khimiya 38 (1997) 3.
[29] C.L. Green, A. Kucernak, Journal of Physical Chemistry B 106 (2002) 1036.
[30] T.D. Gladysheva, B.I. Podlovchenko, Electrochimiya 35 (1999) 573.
[31] B.I. Podlovchenko, U.E. Zhumaev, Yu.M. Maksimov, Journal of Electroanalytical
Chemistry 651 (2011) 30.
It was shown that CNW can be successfully used as the supports
in the preparation of Pt(Cu) catalysts by the method of galvanic
displacement.
•
The possibility of using transients of open-circuit potential for
monitoring the surface layer composition in the course of GD
in systems in which the displaced non-noble metal represents
a phase deposit is analyzed.
•
It was found that the Pt(Cu)st/CNW system, which exhibits the
properties of stable structures of the shell(Pt)–core(Pt, Cu) type,
really represents a mixture of particles differing in not only their
size but also the composition.
•
The specific activity of Pt(Cu)st electrodes in the reaction of
methanol oxidation (calculated per cm2 of the real surface)
turned out to be some lower than the activity of the Pted/CNW
electrodes, which was explained by the size effect.
•
A comparison of characteristics of Pt(Cu)st/CNW synthesized
from Cued and Cuspr allowed us to reveal the important role
played by the spatially separated reactions of Cu dissolution and
Pt deposition in the GD.
•
The Pt(Cu)st/CNW catalyst synthesized based on Cued exhibited
the high specific surface and the smallest content of Pt as com-
pared with analogous catalysts described in the literature. This
might be of certain interest for practice.
[32] A.N. Frumkin, Potentsialy Nulevogo Zaryada (Zero-Charge Potentials), Nauka,
Moscow, 1981.
Acknowledgment
[33] B.I. Podlovchenko, E.A. Kolyadko, Journal of Electroanalytical Chemistry 506
(2001) 11.
[34] A.J. Bard, F. Sholz, M. Stratmann, Ch.J. Pickett (Eds.), Encyclopedia of Electro-
chemistry, Inorganic Chemistry, vol. 7, Textbooks-Barnes and Noble/Wiley-
VCH, 2007, p. 43.
The work was financially supported by the Russian Foundation
for Fundamental Research, Project Nos. 12-03-00998a and MK-
4994.2012.2.
[35] N. Li, J. Lipkowski, Journal of Electroanalytical Chemistry 491 (2000) 95.
[36] B.I. Podlovchenko, Ju.M. Maksimov, T.L. Azarchenko, E.N. Egorova, Elec-
trochimiya 30 (1994) 258.
Appendix A. Supplementary data
[37] D.-L. Lu, M. Ishihara, K. Tanaka, Electrochimica Acta 43 (1998) 2325.
[38] Y. Hosh, T. Yoshida, A. Nishikata, T. Tsuru, Electrochimica Acta 56 (2011) 5302.
[39] J.H. Tian, F.B. Wang, Zh.Q. Shan, R.J. Wang, J.Y. Zhang, Journal of Applied Elec-
trochemistry 34 (2004) 461.
[40] T. You, O. Niva, M. Tomita, S. Hirono, Analytical Chemistry 75 (2003) 2008.
[41] B.E. Conway, Progress in Surface Science 49 (1995) 331.
[42] G. Yerkiewicz, G. Vatankhah, J. Lessard, M.P. Soriaga, Y.-S. Park, Electrochimica
Acta 49 (2004) 1451.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.electacta.2012.04.124.
References
[43] Yu.M. Maksimov, B.I. Podlovchenko, T.L. Azarchenko, Electrochimica Acta 43
(1998) 1053.
[44] S.H. Joo, K. Kwon, D.J. You, Ch. Pak, H. Chang, J.M. Kim, Electrochimica Acta 54
(2009) 5746.
[1] T.R. Ralph, M.P. Hogarth, Platinum Metals Review 46 (3) (2002) 117.
[2] W. Vielstich, A. Lamm, H. Gasteiger (Eds.), Handbook of Fuel Cells, Fundamen-
tals, Technology, Applications (Hardcover), J. Willy and Sons, London, 2003.