Electrochemical formation of bi-metal (copper-palladium) electrocatalyst supported on poly-3,4-ethylenedioxythiophene

Article abstract of DOI:10.1016/j.electacta.2006.06.015

Bi-metal modification of poly-3,4-ethylenedioxythiophene (PEDOT) layers by combining palladium and copper deposition is studied. Initial reference measurements are performed on electrodriven deposition and electroless precipitation of palladium, showing the existence of Pd nanocrystals on the PEDOT surface. It is established that by using the electrodeposition sequence - Cu (first step) - Pd (second step) only palladium remains available in the PEDOT layer. By reversing the deposition sequence - Pd (first step) - Cu (second step) a bi-metal modification of PEDOT is achieved. Depending on the parameters potential and time of the copper electroreduction the Pd nanocrystals coexist with either Cu-PEDOT-stabilized species or both Cu crystals and Cu-PEDOT-stabilized species. The voltammetric response of the modified PEDOT electrodes is tested for nitrate electroreduction in neutral solution. It is found that the bi-metal modified PEDOT shows a marked electroactivity for this reaction.

Full text of DOI:10.1016/j.electacta.2006.06.015

Electrochimica Acta 52 (2006) 816–824
Electrochemical formation of bi-metal (copper–palladium)
electrocatalyst supported on poly-3,4-ethylenedioxythiophene
M. Ilievaa, V. Tsakovaa,∗, W. Erfurthb
a Institute of Physical Chemistry, Bulgarian Academy of Sciences, Sofia, Bulgaria
b Max-Planck-Institut fu¨r Mikrostrukturphysik, Halle, Germany
Received 8 March 2006; received in revised form 6 June 2006; accepted 20 June 2006
Available online 24 July 2006
Abstract
Bi-metal modification of poly-3,4-ethylenedioxythiophene (PEDOT) layers by combining palladium and copper deposition is studied. Initial ref-
erence measurements are performed on electrodriven deposition and electroless precipitation of palladium, showing the existence of Pd nanocrystals
on the PEDOT surface. It is established that by using the electrodeposition sequence – Cu (first step) – Pd (second step) only palladium remains
available in the PEDOT layer. By reversing the deposition sequence – Pd (first step) – Cu (second step) a bi-metal modification of PEDOT is
achieved. Depending on the parameters potential and time of the copper electroreduction the Pd nanocrystals coexist with either Cu–PEDOT-
stabilized species or both Cu crystals and Cu–PEDOT-stabilized species. The voltammetric response of the modified PEDOT electrodes is tested
for nitrate electroreduction in neutral solution. It is found that the bi-metal modified PEDOT shows a marked electroactivity for this reaction.
© 2006 Elsevier Ltd. All rights reserved.
Keywords: PEDOT; Palladium; Copper; Electrocatalysis; Nitrate reduction
1. Introduction
Although metal particles deposition on CP layers is a field
of research started in parallel with the investigations on elec-
trochemical synthesis of CP layers, metal particles electrodepo-
sition in PEDOT is still scarcely studied [3–7]. One of the first
investigations [3] in this area addressed the palladium electrode-
position in PEDOT layers. This study shows some peculiarities
in the metal particles nucleation process involving large over-
potentials, necessary for initiation of the process, and the exis-
tence of marked induction periods before the onset of the metal
deposition. It was demonstrated that the palladium deposit is
electroactive for hydrogen sorption.
The electroreduction of copper ions was studied in a series
of recent papers [4–6]. It was established that depending on the
overpotential used for the electroreduction two different effects
couldbeobserved:(i)atlowoverpotentialsnocrystallinespecies
are formed but copper (after partial electroreduction) becomes
chemically stabilized by the polymer layer; (ii) at high over-
potentials the usual copper crystallization process is observed.
The detailed study of the copper stabilization process [4,5] has
shown that copper becomes captured in the bulk of the PEDOT
layer and that the specifically bonded copper atoms disturb the
PEDOT paramagnetic system [5]. It was also established that
the presence of stabilized copper species is at the origin of an
In the last years poly-(3.4-ethylenedioxythiophene) is one
of the most intensively studied conducting polymers (CP), the
studies being mainly directed to applications based on PEDOT’s
optical and semiconductor properties. Little attention is so far
paid to the possible involvement of PEDOT in electrocatalytic
applications by using it either as electrocatalyst itself or as
a media for dispersing catalytically active centers. Bearing in
mind the high conductivity of PEDOT in a large potential win-
dow as well as the lack of pH dependence of its conductivity, it
could be expected that this CP material may offer some interest-
ing prospects in the electrocatalytic aspect. The electrocatalytic
activity of PEDOT layers was recently demonstrated [1,2] for
the case of selective determination of dopamine and uric acid
in presence of excess ascorbic acid. In our work we address the
possibility for using metal particles modified PEDOT layers for
electrocatalytic purposes.
∗
Corresponding author. Tel.: +359 2 8719307; fax: +359 2 971 2688.
E-mail address: tsakova@ipchp.ipc.bas.bg (V. Tsakova).
0013-4686/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2006.06.015

M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
817
enhancement of the copper electrocrystallization process, result-
ing in increase of the crystal numbers [6].
voltage used for the SEM imaging is 25 keV. All images are
obtained by using secondary electron detector. Spot EDX anal-
ysis with beam diameter of several nanometers is performed
locally at specific objects – high contrast structures or locations
between visible metal crystals.
Electropolymerization of PEDOT layers is performed using
an aqueous microemulsion [23–25] consisting of 0.068 M 3,4-
ethylenedioxythiophene (Bayer), 0.04 M polyoxyethylene-10-
laurylether (Sigma) and 0.5 M LiClO4 (Aldrich). Polymeriza-
tion of EDOT occurs at constant anodic potential, Ea = 1.03 V
for different times. The polymerization charge passed during
synthesis of the PEDOT layers is used for estimating the thick-
ness, d of the CP layers on a 240 mC cm−2 per 1 ␮m basis [26].
The difference in thickness of PEDOT layers denoted with one
and the same value of d does not exceed 5%.
An alternative of the electrodriven metal particles deposition
on conducting polymer layers is the electroless metal precipita-
tion, which occurs by coupling spontaneous polymer oxidation
and metal ions reduction. This process was initially suggested as
a way of recovery of precious metals from acid solutions [8–10]
based on a self-sustained mechanism [11,12]. In few cases elec-
troless precipitation was exploited as a mean of producing metal
deposit by a two step procedure involving initial reduction of the
polymer layers in absence of metal ions and following dipping
of the reduced polymer in the metal ions containing solution.
In this way were obtained palladium [13] and silver [14] parti-
cles in polyaniline. To our knowledge electroless precipitation
of metal particles by using initially reduced PEDOT has not been
explored.
The aim of the present study is to investigate the possibili-
ties for coupling palladium and copper deposition and obtaining
bi-metal modified PEDOT layers. There are two reasons for the
interest in combining these two metals. The first one is based on
the specific electrochemistry of Cu and Pd, offering a number
of ways for approaching the synthesis of the bi-metal compos-
ite, includingelectrolessandelectrodrivendepositions, chemical
stabilization and crystallization processes. Although copper sta-
bilization seems to be a general property of the thiophene-based
CPs [15–18], this process has never been involved in further
metal modification of these polymer layers. Thus there is a basic
interest in the understanding of the role of this thiophene-related
process for metal particles modification.
Copper electroreduction is performed in an aqueous solu-
tion of 0.033 M CuSO4 and 0.5 M H2SO4 at different constant
potentials. The equilibrium potential of copper in this solution
2+
is E0Cu/Cu = 0.28 V. As was shown before [5] at E = 0.25 V
instead of copper crystallization partial reduction and further
stabilization of the copper species takes place. The presence of
stabilized copper becomes evident by the appearance of a new
redox couple in the voltammetric curve measured in supporting
electrolyte (0.5 M H2SO4).
Palladium electrodriven or electroless deposition occurs in an
aqueoussolutionof0.002 MPdSO4 and0.5 MH2SO4. Theequi-
2+
librium potential of Pd in this solution is E0Pd/Pd = 0.855 V.
In the case of electroless deposition the PEDOT layer is ini-
tially subjected to electrochemical reduction by keeping the
PEDOT coated electrode at E = −0.05 V for 1200 s in support-
ing electrolyte (0.5 M H2SO4). After completing this procedure
the electrode is transferred to the palladium containing solution.
The electrocatalytic performance of the metal modified
PEDOT layers is tested in aqueous solution of 0.02 M NaNO3
and 0.1 M NaClO4.
The second reason for the interest in copper and combined
copper–palladium modification of PEDOT relates to possible
electrocatalytic applications. In fact copper alone (in the crys-
talline form) is a catalytic active material for electrooxidation
of a number of organic compounds, e.g. organic acids and poly-
hydric compounds. On the other hand Cu and especially the
Cu–Pd couple is of specific interest for nitrate ions electrore-
duction [19–22]. Thus the second aim of this work is to test the
behaviour of copper and copper–palladium modified PEDOT
layers with respect to the nitrate ions electroreduction reaction.
3. Results and discussion
The experiments on metal modification of the PEDOT lay-
ers are performed in three series (Table 1), the first one (Series
I) consisting of measurements on palladium deposition alone
using electrodriven and electroless deposition. These experi-
ments served as reference base for further comparison with the
casesofbi-metaldepositiononPEDOTlayers. SeriesII(Table1)
consists in Cu electroreduction as first step followed by Pd depo-
sition (second step). The experimental sequence in Series III is
reversed – Pd deposition is the first step and Cu electroreduction
– the second.
2. Experimental
All electrochemical experiments are performed in three
electrode set up using platinum working electrodes, mer-
cury/mercury sulfate reference electrode and platinum plate as
counter electrode. The working electrode used in most of the
experiments is a platinum single crystal bead, sealed in glass,
with surface area, S = 0.02 cm2. Platinum plate working elec-
trodes with S = 2 cm2 are used for preparing specimens for SEM
and EDX analysis. All potentials in the text are referred to the
SHE. The electrolyte solutions are de-aerated before measure-
ments by bubbling of argon.
3.1. Deposition of Pd
The electrochemical experimentation is performed by means
of a computer driven potentiostat/galvanostat (PAR 263 A).
SEM imaging and EDX elemental analysis are accomplished at
a computer driven JSM 6340F (Jeol) scanning electron micro-
scope, equipped with ROENTEC EDX system. The acceleration
3.1.1. Electrodriven deposition of Pd
Previous experiments on electrodriven palladium deposition
in PEDOT [3] were performed under potentiostatic conditions
using rather thin (about 300 nm) PEDOT layers. In the present
investigations we study the deposition process at much thicker

818
M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
Table 1
Series
Experiment
First step
Second step
Available metal
Effect
Pded
Pdes
Pd electrodriven deposition, E = 0.28 V
Pd electroless, precipitation
–
–
Pd
Pd
Pd nanocrystals, 20–50 nm
Pd nanocrystals, 20–50 nm
I
Cust–Pded
Cu electroreduction, E = 0.25 V
Pd electrodriven deposition
Pd (no Cu)
Cu is exchanged for Pd, enhanced
Pd deposition
II
Cust–Pdes
Pded–Cu
Cu electroreduction, E = 0.25 V
Pd electroless precipitation
Pd (no Cu)
Pd and Cu
Cu is exchanged for Pd
Pd electrodriven deposition, E = 0.28 V
Cu electroreduction, E = 0.25 V
Cu available in both stabilized and
crystalline form
III
Pdes–Cu
Pd electroless precipitation
Pd electroless precipitation
Cu electroreduction, E = 0.25 V
Cu electroreduction, E = 0.28 V
Pd and Cu
Pd and Cu
Cu available in both stabilized and
crystalline form
Cu available in the stabilized form
Pdes–Cust
(>1 ␮m) PEDOT layers. This is done because both electroless
palladium deposition and copper stabilization are favored by a
higher redox charge, i.e. higher thickness of the PEDOT layer.
This was shown before for the amount of stabilized copper
species in PEDOT [4]. The same qualitative observation was
made here for the amount of electroless deposited palladium.
Fig. 1 shows potentiostatic current transients of palladium
deposition at three different thicknesses of the PEDOT layers.
Large overpotentials (0.475 and 0.575 V, correspondingly) and
long times are needed for driving the metal deposition reaction.
Asisshownbythetwocurrenttransients, registeredatE = 0.38 V
for the 1.1 ␮m layers it is difficult to obtain reproducible results
at these thicknesses. Obviously, the current reflects a compli-
cated situation involving nucleation and growth of palladium
crystals in the course of diffusion of Pd2+ ions to the polymer
surface but also inside the polymer layer [3]. SEM observation
of the PEDOT surface after potentiostatic palladium deposition
reveals the existence of palladium nanocrystals with diameters
between 20 and 50 nm. Electrochemical evidence for the avail-
able palladium deposit is obtained by cycling the Pded/PEDOT
layer in supporting electrolyte in the hydrogen sorption region
(Fig. 2, dashed line).
3.1.2. Electroless precipitation of Pd
Electroless palladium precipitation was performed by dip-
ping the initially reduced PEDOT layer in the palladium con-
taining solution. The open circuit potential transient measured
in the course of the deposition process (Fig. 3) shows that oxi-
dation of the PEDOT layer occurs rapidly in the first 10 min
followed by a gradual potential increase up to a steady state
value, which is about 140 mV below the palladium equilib-
rium potential. Experiments performed at different precipitation
times using layers with constant thickness have shown that the
amount of deposited palladium does not vary when increasing
the precipitation time beyond 10 min. Therefore, this precipita-
tion time was fixed in further experiments performed at PEDOT
layers with different thickness. As already demonstrated for the
case of electroless precipitation of silver in polyaniline [14], in
this case, too, it is found that the amount of precipitated metal
increases with increasing polymer layer thickness. SEM obser-
vation of the Pdes/PEDOT layer (Fig. 4) reveals the existence of a
largeamountofsmall(20–50 nm)structureswithbrightcontrast.
Local EDX analysis performed at these structures located at the
polymer surface shows that these are palladium crystals. A simi-
lar result (not shown here) is obtained in the case of electrodriven
deposition. In spite of this observation it could be expected that
Fig. 1. Potentiostatic current transients (black lines) of palladium deposition
measured at PEDOT layers at E = 0.38 V for d = 1.1 ␮m and E = 0.28 V for d = 1.4
and 1.9 ␮m. The grey lines denote measurements performed in Pd-free solution
at E = 0.38 V (dashed line) and E = 0.28 V (full line). The deposition potential
was stepped from the equilibrium potential of Pd in the palladium plating solu-
Fig. 2. Voltammetric scans measured in 0.5 M H2SO4 after palladium electro-
driven (0.28 V, 900 s) (dashed line) and electroless (black line) depositions at
PEDOT layers. The grey line denotes the voltammetric scan of the PEDOT layer
before palladium deposition, d = 1.5 ␮m and v = 100 mV/s.
tion EPd/Pd = 0.855 V.
2+

M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
819
ference in the shape of the hydrogen desorption peaks indicating
that irrespective of the deposition procedure (electrodriven or
electroless) the Pd crystals are in both cases electroactive. (The
coincidence in the hydrogen sorption charges of both curves in
Fig. 2 is occasional.)
The two types of palladium deposit are further used for exper-
iments involving copper electroreduction.
3.2. Deposition of Pd in presence of stabilized copper
In these experiments the first step was copper electroreduc-
tion at low overpotential in PEDOT and the subsequent step
– palladium deposition (Table 1 – IInd series of experiments).
Copper was electroreduced at E = 0.25 V for 2400 s. According
to our previous studies [5] in these conditions copper becomes
partially reduced and stabilized by the PEDOT layer without
building of copper crystals. After the copper electroreduction
step the electrode was disconnected electrically, transferred to
thepalladiumcontainingsolutionandconnectedagain. Theelec-
trode remained in the palladium containing solution for several
seconds before being polarized.
Fig. 3. Open circuit potential transient measured in the palladium contain-
ing solution after dipping reduced PEDOT layers with different thickness,
d = 0.9 ␮m (ꢀ, (ꢁ) and d = 1.7 ␮m (ꢂ). (ꢁ) are obtained at Cust (0.25 V, 2400 s),
PEDOT layer. Circles (᭹) denote data obtained in Pd-free solution at an initially
reduced PEDOT layer, d = 2.1 ␮m.
The comparison of the palladium electrodriven deposi-
tion in absence and presence of pre-stabilized copper species
(Fig. 5) shows enhanced palladium deposition in the presence
of Cu-stabilized species. The voltammetric measurements in the
hydrogen sorption region (Fig. 6) gives direct evidence for the
increased amount of palladium deposited in this case.
Palladium electroless precipitation was also tested in absence
and presence of pre-stabilized copper species. No influence of
the stabilized-Cu was observed on the amount of electroless
precipitated palladium.
Both types of Cu–Pd (Cust–Pded and Cust–Pdes) modified
PEDOT specimens were analysed by EDX and observed in
SEM. In both cases only palladium and no copper could be
detected after palladium deposition. This is most probably due
to the exchange of the less noble copper atoms by palladium
in the course of palladium electrodeposition. In order to check
this hypothesis in a separate experiment a Cust–PEDOT layer
the location of the Pd crystals differs depending on the deposi-
tion procedure. Thus preferred deposition at the polymer surface
should occur in the course of the PEDOT oxidation driven elec-
trolessprecipitationwhereasdefectsinthepolymerstructureand
possible nucleation directly at the underlying metal electrode
surface may be involved in the electrodriven deposition. The
last suggestion was made in a previous investigation [3] on pal-
ladium deposition in PEDOT performed in the same palladium
plating solution. Large induction periods in the potentiostatic
current transients were found before observing the typical cur-
rent behaviour corresponding to diffusion limited 3D growth of
metal crystals. Based on this observation and additional support-
ing XPS measurements the authors suggested that the palladium
nuclei appear at the metal/polymer interface and a filament-like
metal structure grows through the polymer layer [3]. The com-
parison of the voltammetric response in the hydrogen sorption
region of both types of palladium deposits (Fig. 2) shows no dif-
Fig. 5. Potentiostatic current transients of palladium deposition measured at
PEDOT layers with thickness d = 1.9 ␮m at E = 0.28 V in absence (grey line)
and presence (black line) of pre-stabilized copper species (0.25 V, 2400 s).
Fig. 4. SEM obtained after electroless palladium precipitation on PEDOT
(d = 1.7 ␮m).

820
M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
Fig. 8. Copper electroreduction transients measured at E = 0.25 V at Pded
(0.28 V, 900 s) modified PEDOT (d = 1.0 ␮m) (dashed line) and at Pdes modified
PEDOT (d = 1.9 ␮m) (black solid and dotted lines). The grey line is obtained at
E = 0.28 V at Pdes modified PEDOT (d = 1.9 ␮m).
Fig. 6. Voltammetric scans measured in 0.5 M H2SO4 after palladium elec-
trodriven deposition (0.28 V, 900 s) performed in absence (dashed line) and
presence (black line) of pre-stabilized copper species (0.25 V, 2400 s). The grey
line denotes the voltammetric scan at PEDOT alone, d = 1.9 ␮m, v = 100 mV/s.
Thus the deposition of palladium performed after copper
electroreduction at low overpotential results finally in the for-
mation of palladium nanocrystalline deposit alone. In the initial
presence of copper stabilized species the potential electrodriven
Pd deposition is enhanced whereas the electroless palladium
precipitation remains unaffected.
was just dipped in the palladium containing solution without
initially reducing the PEDOT layer and without applying any
potential. After several minutes the electrode was transferred
to the supporting electrolyte. The voltammetric scan indicated
availabilityofpalladiumandcompleteabsenceofcopperspecies
giving direct evidence for the Cu–Pd displacement reaction. This
effect could be in the origin of the observed enhancement of the
palladium deposition in the potential electrodriven case. It could
be suggested that the sites where exchange of Cu for Pd takes
place could act as new active sites for palladium electrodeposi-
tion. This process would hardly affect the palladium electroless
precipitation process, which is driven by the oxidation capacity
of the PEDOT layer alone.
3.3. Electroreduction of copper in presence of deposited Pd
In these experiments the first step is palladium deposition and
the subsequent step – copper electroreduction (Table 1 – IIIrd
sequence of experiments).
Fig. 8 shows a set of potentiostatic current transients obtained
in copper containing solution at Pd-modified PEDOT layers. The
experiments (black lines) are performed at copper electroreduc-
tion potential, E = 0.25 V, which is known to result in copper
stabilization alone when applied to non-modified PEDOT lay-
ers [5]. In the latter case a falling current transient coming to a
steady state value is registered. This current behaviour does not
imply the formation and growth of copper crystals with expand-
ing surface. In the presence of palladium deposited crystals,
however, a rising current, as expected for the case of nucle-
ation and 3D growth of metal crystals is observed at the same
copper electroreduction potential (black solid and dashed lines
in Fig. 8), indicating the initiation of copper crystallization. A
falling transient (grey line in Fig. 8) could be obtained if only
shifting the copper electroreduction potential to 0. 28 V. In fact
this potential corresponds to the equilibrium potential of copper
in the electrolyte solution.
SEM observation (Fig. 7) of the Cu–Pd modified PEDOT
surface shows in both cases the availability of Pd crystals with
the same characteristics (i.e. size and number) as in the case of
direct Pd deposition in absence of pre-stabilized copper species.
The type of the copper deposit (obtained when using PEDOT
layers with pre-deposited palladium) is investigated by SEM.
Two Pdes–Cu–PEDOT specimens are prepared by depositing
palladium in one and the same way. For the first specimen the
copper deposition step is stopped in the region of the low steady
state current (Fig. 8, solid dots) of copper electroreduction. For
the second specimen the copper deposition step is prolonged so
as to obtain the late rising part of the current transient (Fig. 8,
Fig. 7. SEMofaCu–Pded–PEDOTspecimen, obtainedafterCuelectroreduction
(0.25 V, 3600 s) followed by Pd deposition (0.28 V, 1200 s), d = 1.5 ␮m.

M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
821
Fig. 9. SEM of (a) Pdes–Cu (0.25 V, 300 s) PEDOT and (b) Pdes–Cu (0.25 V, 600 s) PEDOT. EDX spectrum (c) obtained between the copper crystals observed in
SEM (b), d = 1.9 ␮m.
solid line). SEM observation shows that in the first case there
are no detectable Cu crystals on the PEDOT layer (Fig. 9a),
although EDX analysis shows the availability of both Cu and
Pd all over the surface. For the second specimen (advanced
stage of copper electroreduction) copper crystals are seen at
the Pdes–Cu–PEDOT surface (Fig. 9b). EDX analysis (Fig. 9c)
performed in the regions between the copper crystals indicates
availability of both Cu and Pd. Thus the type of the copper
deposit relates to falling or rising part of the copper deposition
transient, i.e. to the initial or advanced stage of the electroreduc-
tion process.
Additional evidence in this respect was obtained through
electrochemical measurements – by studying the voltammetric
response of the Pd–Cu-modified PEDOT in supporting elec-
trolyte. In the case of a rising copper deposition transient (dashed
line in Fig. 8) a large stripping peak is observed when first
scanning the potential in the positive direction (Fig. 10a), corre-
sponding to the dissolution of crystalline copper. In the second
scan (solid line in Fig. 10b) the redox couple characterizing
copper stabilized in PEDOT [4,5] is clearly seen together with
hydrogen adsorption/desorption peaks giving evidence for the
presence of Pd, too. In the case of Cu electroreduction at the
Cu equilibrium potential value (grey transient in Fig. 8), the
voltammetric curve (grey line in Fig. 10b), measured in sup-
porting electrolyte, indicates again the availability of palladium
and of stabilized copper species, alone.
The measurements presented so far point (i) to the existence
of both Pd and Cu in PEDOT layers, where palladium deposition
is the first- and copper electroreduction – the subsequent step and
(ii) to the initiation of Cu crystallization on Pd–PEDOT layers
at lower overpotential than on non-modified PEDOT layers.
The established reduction of the overpotential for copper
crystal formation in presence of pre-deposited Pd crystals
deserves special attention. This effect should be commented
by taking into account that copper underpotential deposition
takes place at palladium electrodes [27–29]. Thus it could be
suggested that the palladium crystals, acting as nanoelectrodes
become covered by a copper upd layer and provide new, energet-
ically favored, active sites for copper crystal formation. These
active sites could more easily promote the copper nucleation
process in comparison to the non-modified PEDOT surface. It
is worth noting that the visible number of Cu crystals observed
at the Pdes–Cu–PEDOT surface (4.107 cm−2) is low in compar-
ison to the number of electroless deposited palladium particles
(>109 cm−2). Thus it seems that not all of the existing palla-
dium nanocrystals are used for copper crystal formation. Indeed
EDX analysis performed at locations between the copper crys-
tals shows the availability of large amounts of both palladium

822
M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
Fig. 11. Voltammetric curves measured in 0.02 M NaNO3 + 0.1 M NaClO4 solu-
tion at PEDOT (grey line), Cust (0.25 V, 2400 s) PEDOT (dashed line) and
Cust (0.25 V, 2400 s) +Cucryst (0.21 V, 60 s) PEDOT (black line). The dotted
line is obtained at the Cust + Cucryst–PEDOT specimen in nitrate-free solution,
d = 1.8 ␮m, v = 20 mV/s .
electroreduction. For comparison the voltammetric curve mea-
sured in absence of nitrate ions (Fig. 11, dotted line) is shown
for this case, too. The cathodic current shoulder registered in the
nitrate containing solution overlaps the potential region where
nitrate reduction proceeds at bulk copper electrodes in neu-
tral solutions. Thus it becomes evident that stabilized copper
species are not catalytically active for nitrate reduction but have
a positive effect by impeding the hydrogen reduction reaction.
The latter effect is very probably connected to impeded diffu-
sion of the hydrogen ions through the polymer porous structure
where Cu-atoms are coordinated in a specific way by the EDOT
molecules. The copper crystals however are electroactive for the
nitrate electroreduction.
Fig. 10. First (a) and second (b) voltammetric scans measured in 0.5 M H2SO4
at Pded (0.28 V, 900 s) Cu (0.25 V, 180 s) PEDOT (black lines), Pdes–Cu (0.28 V,
2400 s) PEDOT (grey line), v = 5 mV/s, d = 1.0 ␮m.
The voltammetric response due to the presence of palla-
dium particles alone is also checked in the nitrate contain-
ing electrolyte (Fig. 12, dashed line). As should be expected
and copper. Thus the reduced copper species are consumed in
parallel for copper crystal formation and copper stabilization by
the PEDOT layer. This result is also in line with the electrochem-
ical measurement showing that copper in the stabilized form is
available in the Pded–Cu–PEDOT specimen.
3.4. Electroreduction of nitrate ions at metal modified
PEDOT
To our knowledge there is no previous experience in studying
the nitrate electroreduction reaction at metal modified conduct-
ing polymer electrodes and there is no data on the behaviour of
PEDOT alone in this solution. Therefore, the response of the
polymer layer in the nitrate containing solution is first checked
(Fig. 11, grey line). It is obvious that there is no electrocatalytic
response for the nitrate reduction reaction and that hydrogen
reduction starts at about −0.7 V. The voltammetric curve mea-
sured at Cust–PEDOT (Fig. 11, dashed line) shows an inhibition
of the hydrogen reduction reaction but still no response for
nitrate electroreduction. The combination of stabilized and crys-
tallized copper species deposited in the PEDOT layer (Fig. 11,
solid line) results in the appearance of a cathodic current shoul-
derinthevoltammetriccurve, indicativefortheprocessofnitrate
Fig. 12. Voltammetric curves measured in 0.02 M NaNO3 + 0.1 M NaClO4 solu-
tion at PEDOT (grey line), Pdes–PEDOT (dashed line) and Pdes–Cu (0.25 V,
600 s) PEDOT (black line). The dotted line is obtained at the Pdes–Cu–PEDOT
specimen in nitrate-free solution, d = 1.8 ␮m, v = 20 mV/s .

M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
823
the Pd particles, deposited at the PEDOT layer, promote the
hydrogen reduction reaction and are of no direct use for the
studied electroreduction reaction. However, the use of the bi-
metal Pd–Cu–PEDOT-modified electrode (Fig. 12, solid line)
shows a well defined voltammetric wave of nitrate reduction.
For comparison the voltammetric response measured in absence
of nitrate ions (Fig. 12, dotted line) is shown in the figure, too.
The deviation of the voltammetric curve measured in nitrate-
containing solution from the one measured in nitrate-free solu-
tion begins exactly at the point where nitrate reduction starts
at bulk copper electrodes in neutral solution. Thus the bi-metal
modified PEDOT electrode combines inhibited hydrogen reduc-
tion due to the Cu-stabilized species and promoted nitrate reduc-
tion due very probably to the favorable co-existence Pd and
Cu crystals. It is interesting to note the marked decrease in
the PEDOT redox charge for the bi-metal modified electrode.
Bearing in mind that the copper electrodeposition transient was
in this case of the rising type and thus implying nucleation
and growth of copper crystals this feature is very probably
due to decreased active PEDOT surface blocked by the copper
crystals.
step)–Cu(second step) deposition sequence results in a bi-
metal modification of the PEDOT layer.
The use of copper stabilization for modification of the
PEDOT layers reveals the specific role of these species and
namely the possibility to block the diffusion through the porous
polymer structure of small ions, e.g. H+. The present results
support also our previous findings [6] on the important role
of the Cu-stabilized species for the metal electrocrystallization
process. For the cases of both copper [6] and palladium crys-
tallization the presence of copper-stabilized species results in
increasing the amount of subsequently electrodeposited metal.
This effect is due to the provision of new active sites for
nucleation, the latter being either the stabilized copper atoms
themselves or new sites at the polymer surface resulting from
cooperative effect of electronic redistribution along the chains
[6].
Finally, our investigation shows a first attempt to use metal
modified PEDOT layers for the electroreduction of nitrate ions
in neutral solutions. It is established that
1. The presence of Cu-stabilized species in the PEDOT layer
has a positive effect by impeding the hydrogen reduction
reaction.
2. The electrodeposited copper crystals are electroactive for the
nitrate reduction reaction.
3. The use of the bi-metal (Pd–Cu) modified PEDOT results in
a marked voltammetric response and seems to be of special
interest due to the simultaneous blocking of the interfering
hydrogen reduction reaction and promoting the nitrate elec-
troreduction.
4. Conclusions
The investigations presented in this study reveal various pos-
sibilities for palladium and copper–palladium modification of
PEDOT layers:
1. It is demonstrated that palladium electroless precipitation is
a suitable alternative to the conventional electrodriven depo-
sition and results in palladium nanocrystals with number
density >109 cm−2, evenly distributed over the polymer sur-
face.
2. The presence of stabilized copper species in the PEDOT layer
before the onset of palladium deposition enhances the palla-
dium electrocrystallization process due very probably to the
presence of new active sites (Pd atoms exchanged for Cu)
in the PEDOT layer. In contrast, the palladium electroless
process, driven by the PEDOT oxidation capacity, remains
unaffected by the presence of copper stabilized species. For
both electrodriven and electroless palladium deposition the
Cu (first step)–Pd (second step) deposition sequence results
in a final deposit consisting of palladium alone due to the
displacement of the less noble copper for palladium.
3. A bi-metal modification of PEDOT is achieved by perform-
ing in the first step palladium deposition and in the second
step – copper electroreduction. In the presence of palladium
nanocrystals the overpotential for initiating copper crystal-
lization is reduced due very probably to the involvement of
new active sites in the nucleation process. The new active
sites should be related to the Cu upd process taking place on
palladium. Depending on the duration of the second – cop-
per electroreduction step, the initially deposited palladium
nanocrystals coexist with either stabilized copper species
(falling copper electroreduction transient) or stabilized and
crystalline copper (rising part of the copper electroreduc-
tion transient observed at longer times). Thus the Pd (first
Acknowledgements
Financial support of the FP6 European project NANOPHEN
(INCO-CT-2005-016696) and of the project X-1405 of the Bul-
garian Ministry of Education and Science is gratefully acknowl-
edged.
References
[1] S.S. Kumar, J. Mathiyarasu, K.L. Phani, J. Electroanal. Chem. 578 (2005)
95.
[2] S.S. Kumar, J. Mathiyarasu, K.L.N. Phani, J.K. Jain, V. Yegnaraman, Elec-
troanalysis 17 (2005) 2281.
[3] V. Tsakova, S. Winkles, J.W. Schultze, J. Electroanal. Chem. 500 (2001)
574.
[4] M. Ilieva, V. Tsakova, Synth. Met. 141 (3) (2004) 281.
[5] M. Ilieva, V. Tsakova, Synth. Met. 141 (3) (2004) 287.
[6] M. Ilieva, V. Tsakova, Electrochim. Acta 50 (2005) 1669.
[7] S. Biallozar, A. Kupniewska, V. Jusalajtene, Bull. Electrochem. 20 (5)
(2004) 231.
[8] Y.P. Ting, K.G. Neoh, E.T. Kang, K.L. Tan, J. Chem. Tech. Biotechnol. 59
(1994) 31.
[9] E.T. Kang, Y.P. Ting, K.G. Neoh, K.L. Tan, Synth. Met. 69 (1995) 477.
[10] L.M. Abrantes, A.C. Cascalheira, M. Savic, Portug. Electrochim. Acta 17
(1999) 157.
[11] L.M. Abrantes, J.P. Correia, Mater. Sci. Forum 191 (1995) 235.
[12] J.A. Smith, M. Josowicz, J. Janata, J. Electrochem. Soc. 150 (2003) E384.

824
M. Ilieva et al. / Electrochimica Acta 52 (2006) 816–824
[13] A. Mourato, A.S. Viana, J.P. Correia, H. Siegenthaler, L.M. Abrantes, Elec-
trochim. Acta 49 (2004) 2249.
[21] J.F.E. Gootzen, P.G.J.M. Peeters, J.M.B. Dukers, L. Lefferts, W. Visscher,
J.A.R. van Veen, J. Electroanal. Chem. 434 (1997) 171.
[22] A.C.A. de Vooys, R.A. van Santen, J.A.R. van Veen, J. Mol. Catal. A:
Chem. 154 (2000) 203.
[23] V. Tsakova, S. Winkels, J.W. Schultze, Electrochim. Acta 46 (2000) 759.
[24] F.-H. Haegel, J. Schlu¨pen, J.W. Schultze, S. Winkels, Ch. Stromberg, Elec-
trochim. Acta 46 (2001) 3973.
[14] S. Ivanov, V. Tsakova, Electrochim. Acta 50 (2005) 5616.
[15] G. Tourillon, E. Dartyge, H. Dexpert, A. Fontaine, A. Jucha, P. Lagarde,
D.E. Sayers, J. Electroanal. Chem. 178 (1984) 357.
[16] G. Tourillon, E. Dartyge, A. Fontaine, A. Jucha, Phys. Rev. Lett. 57 (1986)
603.
[17] D. Guay, G. Tourillon, A. Fontaine, Fraday Discuss. Chem. 89 (1990)
41.
[25] C. Stromberg, V. Tsakova, J.W. Schultze, J. Electroanal. Chem. 547 (2003)
125.
[18] M.R. Guascito, P. Boffi, C. Malitesta, L. Sabbatini, P.G. Zambonin, Mater.
Chem. Phys. 44 (1996) 17.
[26] N. Sakmeche, J.J. Aaron, M. Fall, S. Aeiyach, M. Jouini, J.C. Lacroix, P.C.
Lacaze, Chem. Commun. (1996) 2723.
[19] X. Xing, D. Sherson, J. Electroanal. Chem. 270 (1989) 273.
[20] G.E. Dima, A.C.A. de Vooys, M.T.M. Koper, J. Electroanal. Chem.
554–555 (2003) 15.
[27] S. Szabo, J. Electranal. Chem. 77 (1977) 193.
[28] T. Chierchie, C. Mayer, Electrochim. Acta 33 (1988) 341.
[29] T. Solomun, J. Electranal. Chem. 261 (1989) 229.