Characterization of bismuth adatom-modified palladium electrodes. The electrocatalytic oxidation of aliphatic aldehydes in alkaline solutions

Article abstract of DOI:10.1016/j.electacta.2006.05.048

Palladium-bismuth films supported on glassy carbon substrates were prepared by cathodic electrodeposition of palladium ad subsequent adsorption of bismuth species on the electrode surface. A mechanism regarding the irreversible adorption process of bismuth species on the palladium surface, was described. The resulting active film was characterized, and in particular, the electrooxidation of some aliphatic aldehydes was investigated in moderately alkaline solutions. The bismuth adlayer increases the overall rate of the oxidative processes through a combination of a third body mechanism and a true catalyst. The oxidation currents of propionaldehyde, evaluated in cyclic voltammetry increased linearly on increasing analyte concentration up to 8 mM. Under chronoamperometric conditions at the constant applied potential of -0.2 V versus SCE, the resulting calibration plot for propionaldehyde analysis was linear over the range of 1.5-8 mM with a correlation coefficient better than 0.97. Throughout the XPS investigation Pd 3d and Bi 4f detailed regions an average chemical composition of the surface composite film of Pd_96±2Bi_4±2 was evaluated.

Full text of DOI:10.1016/j.electacta.2006.05.048

Electrochimica Acta 52 (2006) 649–657
Characterization of bismuth adatom-modified palladium electrodes
The electrocatalytic oxidation of aliphatic aldehydes
in alkaline solutions
Innocenzo G. Casella∗, Michela Contursi
Dipartimento di Chimica dell’Universita`, Via N. Sauro 85, 85100 Potenza, Italy
Received 11 January 2006; received in revised form 26 May 2006; accepted 26 May 2006
Available online 10 July 2006
Abstract
Palladium–bismuth films supported on glassy carbon substrates were prepared by cathodic electrodeposition of palladium ad subsequent adsorp-
tion of bismuth species on the electrode surface. A mechanism regarding the irreversible adorption process of bismuth species on the palladium
surface, was described. The resulting active film was characterized, and in particular, the electrooxidation of some aliphatic aldehydes was investi-
gated in moderately alkaline solutions. The bismuth adlayer increases the overall rate of the oxidative processes through a combination of a third
body mechanism and a true catalyst. The oxidation currents of propionaldehyde, evaluated in cyclic voltammetry increased linearly on increasing
analyte concentration up to 8 mM. Under chronoamperometric conditions at the constant applied potential of −0.2 V versus SCE, the resulting
calibration plot for propionaldehyde analysis was linear over the range of 1.5–8 mM with a correlation coefficient better than 0.97. Throughout the
XPS investigation Pd 3d and Bi 4f detailed regions an average chemical composition of the surface composite film of Pd96 2Bi4 was evaluated.
2
© 2006 Elsevier Ltd. All rights reserved.
Keywords: Palladium; Bismuth; Oxides; Aldehydes oxidation; XPS
1. Introduction
have been widely studied and proposed in order to improve the
electrocatalytic activity of traditional noble metals for the oxida-
tion of organic molecules [1,2]. These metal adatoms can affect
the initial activity of the noble electrode substrate through dif-
ferent mechanism such as: (i) blocking the active sites of the
electrode in order to avoid the adsorption of poisoning species;
(ii) providing a favourable and distinct catalytic interaction with
the reactants and reaction intermediates (i.e., oxygen species
necessary for the surface electrode reactions); (iii) modifying
the electronic properties of the electrode surface in a such a
way that the poisoning species cannot be adsorbed on the active
sites. In this way, electrodeposition or irreversible adsorption of
As, Se, Sb, Sn, Pb, Hg, Cd, Bi, etc. on noble metal substrates
have been proposed as electrodes in order to obtain significant
enhancements in the electrocatalytic activity towards the elec-
trooxidation of small organic electroactive molecules [3–11].
Because of its strong adsorption and subsequent formation
of stable adlayer, bismuth adatom has been widely charac-
terized and used as modifier species in order to enhance the
electrocatalytic properties of platinum or palladium surfaces
[12–17]. It has been suggested that bismuth enhances the cat-
The electrochemical oxidation of organic species in aque-
ous solutions constitutes an interesting topic in the field of
the electrosynthesis, electroanalysis and hydrocarbon–air fuel
cells. In general, the efficiency of electrode reactions, depend
strictly upon the chemical and physical nature of the electrode
substrates. Noble metals such as gold and platinum are ideal
electrode substrates for the electrochemical oxidation of several
electroactive molecules because they show a high catalytic activ-
ity and a good corrosion resistance. Unfortunately, a serious con-
sequence of strong adsorption of organic reactants, intermediate
and/or products species can be the reason fouling of these elec-
trodes when are used as anodic materials. Hence, the scarcely
electrocatalytic efficiency of these electrodes is the result, in
many instances, of strong adsorption of organic molecules with
subsequent poisoning of the substrate electrodes. Metal adatoms
∗
Corresponding author.
E-mail address: casella@unibas.it (I.G. Casella).
0013-4686/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2006.05.048

650
I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
alytic activity of platinum substrate electrodes toward formic
acid through an electronic effect [7], however bismuth may also
act as a third body, decreasing the ensemble size for poison for-
mation [16]. On the other hand, theoretical model calculation
has suggested that bismuth acts as a catalyst in the electroox-
idation processes (bifunctional mechanism) [7]. Recently, we
have compared the electrocatalytic activity of polycrystalline
gold and platinum electrodes modified with adsorbed bismuth
oxide adlayers toward the electrooxidation of several sugar acid
molecules in moderately alkaline medium [18]. In this study,
the bismuth adsorbed species acts as a true catalyst on the direct
electrooxidation pathway, providing the oxygen necessary for
the oxidation of organic molecules adsorbed on the platinum or
gold active sites.
that no features arose from the copper tape and from the sample
rod. Wide and detailed spectra were collected in fixed analyser
transmission (FAT) mode with a pass energy of 50 eV and a
channel width of 1.0 and 0.1 eV, respectively. The vacuum in
the analysis chamber was always better than 5 × 10−9 mbar.
The kinetic energy axis origin in all spectra was not cor-
rected for surface charging, but peak positions (BE) in the text
are corrected by referring to the C 1s peak after setting its BE to
284.6 eV. Satellites and a non-linear Shirley background were
substracted from the spectrum before curve fitting analysis. Ele-
mental surface stoichiometries were obtained from peak area
ratios corrected by appropriate sensitivity. Gaussian/Lorenzian
sum functions were used to fit Pd 3d, Bi 4f, O 1s and C 1s peaks
line-shapes.
In the present work, we report a procedure for the electro-
chemical deposition of palladium film onto a traditional glassy
carbon electrode substrate (GC/Pd) and its surface modifica-
tion with bismuth adlayer (GC/Pd/Bi). The resulting modified
films were characterized by electrochemical and X-ray photo-
electron spectroscopy (XPS) techniques. The electrocatalytic
properties of the GC/Pd and GC/Pd/Bi electrodes were inves-
tigated in 50 mM NaOH solutions towards the electrocatalytic
oxidation of some saturated aliphatic aldehydes.
2.3. Electrode modification
The palladium electrodeposition on the glassy carbon elec-
trodes was performed by voltage cycling (50 mV s−1) between
0.0 and −0.4 V versus SCE for 50 cycles in non-deaerated
solution containing 0. 5 mM Na2PdCl6, 50 mM HClO4 and
25 mM HCl. The bismuth-modified palladium electrodes were
prepared by immersion of the GC/Pd electrode in non-deaerated
50 mM HClO4 solution containing 200 ␮M Bi(NO3)3 at open
circuit for a given time. Occasionally, bismuth film was elec-
trodeposited on the glassy carbon surface at −0.6 V versus
SCE for 10 min in solution 50 mM HClO4 containing 200 ␮M
Bi(NO3)3. The resulting GC/Pd/Bi electrodes were then rinsed
with pure water and transferred to an electrochemical cell con-
taining 50 mM NaOH solution. Traces of palladium–bismuth
oxides were removed from the glassy carbon substrate by pol-
ishing the surface electrode with 0.05 ␮m ␣-alumina powder on
a polished micro-cloth and washing it with twice distilled water.
Quantitative analysis for evaluating the film composition was
mainly based on the relative areas of the XP signal of the Pd 3d
and Bi 4f corrected for the relevant sensitivity factors. Moreover,
thesurfacecoverageofBi(Γ Bi)andthesurfacefilmcomposition
of the investigated composite electrodes were also evaluated
by electrochemical analysis considering the charge under the
Ia1 and Ic1 waves associated at the oxidation and reduction of
bismuth and palladium species, respectively. In this respect, the
charge under the Ia1 wave was referred to Bi0 → BiIII redox
process while the Ic1 wave was referred to PdII → Pd0 redox
process.
2. Experimental
2.1. Reagents
Solutions were prepared from analytical-reagent grade chem-
icals without further purification and by using doubly distilled
deionized water. NaOH pellets (99%), HClO4 (70%), Na2PdCl6,
Bi(NO3)3·5H2O and aliphatic aldehydes are purchased from
Aldrich Chemical Co. (St. Louis, MO).
2.2. Apparatus
The voltammetric experiments were performed with an Auto-
lab PGSTAT 30 Potentiostat/Galvanostat (Eco Chemie, Utrecht,
The Netherlands) and the data were acquired using an Autolab
GPES software package Version 4.8. Cyclic voltammetry (CV),
was done in a three-electrode cell using a glassy carbon electrode
modified with palladium film and bismuth adlayer as a working
electrode, a SCE reference electrode and a platinum foil counter-
electrode. The electrode substrate (3 mm diameter) used in CV
was purchased from Amel (Milan, Italy). Current density values
(mA cm−2) are quoted with respect to apparent geometric area
of electrode substrate. Unless otherwise specified, experiments
were carried out at an ambient temperature (21 ◦C 2). When
necessary, the solutions were deoxygenated by bubbling with
nitrogen prior to the electrochemical experiments.
3. Results and discussion
3.1. Preparation and electrochemical behaviour of the
GC/Pd and GC/Pd/Bi electrodes
X-ray photoelectron spectra were collected using a Leybold
LH X1 spectrometer using unmonochromatized Al K␣ radiation
(1486.6 eV). The source was operated at 15 kV and 20 mA. The
binding energy (BE) scale was calibrated with respect to the Cu
2p3/2 (932.7 eV, with a full-width at half-maximum (FWHM)
of 1.75 eV) and Au 4f7/2 (84.0 eV with a FWHM of 1.20 eV)
signals. Spectra were recorded only after the wide scan showed
In order to minimize the hydrogen gas evolution from the
electrode surface and to improve the adherence and mechanical
film properties, the electrodeposition of palladium film was per-
formed by voltage cycling between 0.0 and −0.4 V versus SCE
for50cyclesindeaeratedsolutioncontaining0.5 mMNa2PdCl6,
50 mM HClO4 and 25 mM HCl. Fig. 1 shows a typical growth of
palladium film on the glassy carbon electrode during continuous

I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
651
Fig. 1. Consecutive cyclic voltammograms (recorded after each fifth scan) at
50 mV s−1 of palladium film growth on a glassy carbon electrode in a solution
containing 0. 5 mM Na2PdCl6, 50 mM HClO4 and 25 mM HCl.
Fig. 2. Comparison of the voltammogram profiles of a GC/Pd/Bi and GC/Pd
electrodes in 50 mM NaOH solution. The bismuth oxide adlayers were deposited
on the palladium film by direct immersion of the GC/Pd electrode in solution
containing 200 ␮M Bi(NO3)3 at open circuit for 10 min. Other experimental
conditions as in Fig. 1.
potential cycling. After 50 potential scans, an uniform and well
adherent deposit film appeared on the glassy carbon surface. It
is well known that in the negative potential range, hydrogen is
strongly adsorbed on the Pd surface and can be absorbed to a
large extent into the Pd lattice [19,20]. In general, two super-
ficially forms (i.e., ␣-PdH and ␤-PdH) of hydrogen species on
the Pd surface are reported [21] and the complex redox waves
observed, can be attributed to the ␤-PdH formation/desorption
processes of hydrogen from the Pd lattice to the surface and from
there into the electrolyte [22]. In fact, depending on the potential
and time, the diffusion of hydrogen atoms into bulk palladium
film (absorption processes) can result in the formation of either
␣ or the ␤ phase [23,24].
The bismuth adlayers on the palladium surface were prepared
by direct immersion of the GC/Pd electrode in solution con-
taining 200 ␮M Bi(NO3)3 at open circuit for 10 min. Typical
voltammogram profiles are shown in Fig. 2. The electrochem-
ical processes occurring at the GC/Pd electrode are relatively
complexes because several palladium oxide species are involved
(Ic1) corresponding to the reduction of oxide species and ref-
ormation of a clean surface of Pd0 layers. At potential values
more negative than −0.6 V there is a broad wave corresponding
to the hydrogen adsorption process. The voltammogram pro-
file of the GC/Pd/Bi electrode, shows some distinct features if
is compared with that of GC/Pd electrode. At sufficiently high
coverage, the adsorbed bismuth species suppresses completely
the adsorption/desorption processes of hydrogen on Pd surface.
In addition, the adsorbed bismuth induces a partial inhibition of
the formation of palladium oxides and oxygen evolution. It is
interesting to observe that during the positive sweep, there is a
wave (Ia1) that occurs at about 0. 1 V versus SCE and can be
related to the formation of a Bi(III) oxide film [29], according
to the following reaction:
2Biads + 6OH− ⇔ (Bi2O3)ads + 3H2O + 6e−
(1)
during the redox transition [19,25–27] such as PdO, Pd(OH)ads
,
The reduction of Bi2O3 and/or other bismuth oxide species such
as Bi(OH)3, BiOOH, Bi(OH)2+, Bi6(OH)126+, etc., is usually
observed as a single cathodic peak occurring prior to hydrogen
region. This reduction process can be accompanied by a partial
dissolution of bismuth oxide into the ionic species BiO2 and
formation of BiOads, (Bi(OH)2)ads and Biads species.
The experimental data reported here established that when
a palladium surface is exposed to an acidic solution contain-
ing Bi3+ ions, a spontaneous adsorption process of bismuth
species occurs. This behaviour is in agreement with previous
studies [30,31], where an irreversible adsorption process of bis-
muth species on platinized or platinum electrodes substrates
is observed. Taking into account our experimental data, and
in agreement with the literature [29,30], we believe that the
adsorption process of the bismuth species, proceeds through the
ionization of adsorbed hydrogen species with subsequent forma-
tion of hydroxyl and bismuth adsorbed species. Thus, a general
scheme regarding the irreversible adsorption process of bismuth
Pd(OH)2, PdO2, PdO3, etc. These electrochemical surface reac-
tions are assumed to occur in a stepwise fashion, superimposed
by a significant metal dissolution process strongly dependent
on the experimental conditions, such as applied potentials, pH
and chemical composition of the electrolytic solutions, etc. The
GC/Pd electrode shows during the anodic sweep, a broad peak
centred in the potential range between −0.7 and −0.5 V. This
peak is likely related to the desorption process of hydrogen
species from the palladium surface. Upon further anodic scan-
ning, the major redox process includes a large anodic peak at
about 0.0 V corresponding to the formation of Pd(OH)2 and/or
PdO monolayers [27,28]. Moreover, a simultaneous formation
ofhydrousoxideabove0.0 VviaformationofPd(OH)ads species
can be postulated. Oxygen evolution, observed at potentials
>0.5 V, obscures the overall processes regarding the formation
of higher valence oxides of palladium such as PdO2 and PdO3
species. During the cathodic sweep, there is a peak at −0.4 V
−

652
I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
potentials comprised between −0.4 and 0.1 V. The currents
increased linearly on increasing analyte concentration up to
8 mM and the relevant peak potential shifted in the positive
direction on increasing concentration of analyte. The inset
in Fig. 3A, shows the relevant calibration graph for the
analysed propionaldehyde. Similar current–potential profiles
and linear ranges of the concentrations were observed for the
other aliphatic aldehydes investigated such as formaldehyde,
acetaldehyde, butyraldehyde, valeraldehyde and hexanale. In
all cases, no significant electrooxidation currents attributable
to aliphatic aldehydes could be observed at glassy carbon
electrodes modified with electrodeposited bismuth film, in the
same alkaline medium and within the same potential window.
The GC/Pd electrode shows, a very low elctrocatalytic activity
for potentials higher than −0.2 V, where the electrooxidation of
the aldehyde is simultaneously accompanied with the formation
of higher oxidation state of the palladium. The anodic currents
are not linearly proportional to analyte concentrations, but
approach a limiting value with increasing concentrations
(i.e., up to 1.5–2.0 mM). Moreover, the comparison between
the electrochemical behaviour of the GC/Pd/Bi and GC/Pd
electrodes (see Fig. 3A and B) reveals, in presence of analyte,
a diminution of the charge under the Ic1 cathodic peak of
about 20% and 75%, respectively. Similarly, in presence
of hexanale (3 mM), was observed a diminution of the Ic1
peak of about 25% and 91% for the GC/Pd/Bi and GC/Pd
electrodes, respectively. The differences in the electrochem-
ical behaviour of the GC/Pd/Bi and GC/Pd electrodes are
particularly evidenced for aliphatic aldehydes with longer
aliphatic chains (i.e., valeraldehyde and hexanale). Thus the
influence of the bismuth adlayer on the catalytic activity
should increase in the order: formaldehyde < acethaldehyde
< propionaldehyde < butyraldehyde < valeraldehyde < hexanale.
It is likely, that bismuth adlayer on the palladium film induces
a partial inhibition of the adsorption processes of organic
reactants and/or intermediate reaction species on the catalytic
active sites. Thus, bismuth adatom acts as a third body type
mechanism in decreasing the reaction ensemble size. Similarly,
on platinum electrodes modified with bismuth adlayers, a
third body mechanism during the formic acid electrooxidation
process has been hypothesized [32]. On the other hand, the
potential needed for the incipient electrooxidation process of the
aldehydic molecules on the GC/Pd/Bi electrode is significantly
less positive than that observed in absence of bismuth adlayer
(see Fig. 3A and B). The total shift of the anodic polarization
curve (towards the more negative potentials) reaches approx-
imately 0.25 − 0.40 V, which is an interesting catalytic effect
because the oxidation overpotential is significantly reduced.
As comparison, Fig. 4 shows the relevant voltammograms
obtained for the electrooxidation of 3.2 mM hexanale in 50 mM
NaOH solution at a GC/Pd/Bi and GC/Pd electrodes. As can
be seen, with bismuth adatoms a significantly enhancement of
the catalytic currents, during the anodic sweep, is observed.
In addition, on the GC/Pd/Bi electrode the electrooxidation
process is shifted towards negative potentials by more than
0.4 V. Although it appears that bismuth acts as third body in
decreasing the reaction ensemble size, it may be contributing
Fig. 3. (A) Voltammograms at a GC/Pd/Bi in 50 mM NaOH (dashed curve) and
in solution containing increasing concentration of propionaldehyde (in 1.6 mM
steps). Inset: calibration graph of peak currents versus concentration of the propi-
onaldehyde. In the calibration graph, the peak currents are not corrected for the
background contribute. (B). Voltammograms at a GC/Pd electrode in 50 mM
NaOH (dashed curve) and in solution containing increasing concentration of
propionaldehyde. Other experimental conditions as in (A).
species on the palladium surface, may be the following:
Pd + H2O ⇔ Pd(OH)ads + Pd(H)ads
Bi3+ + Pd(H)ads ⇔ Pd(Bi∗)ads + H+
(2)
(3)
where (Bi*)ads indicates the bismuth ions and/or its oxides
species adsorbed on the palladium surface. Thus, the proposed
mechanism of bismuth adsorption, combining hydroxyl species
formation with bismuth adsorption, would lead to the simulta-
neous accumulation at the surface of oxygenated and bismuth
species (vide infra). Nevertheless, more experimental informa-
tion is necessary in order to confirm and/or define a better
detailed adsorption mechanism.
Typical cyclic voltammograms on GC/Pd/Bi and GC/Pd
electrodes in 50 mM NaOH solution containing increasing
concentrations of propionaldehyde are shown in Fig. 3A
and B, respectively. When propionaldehyde was added to
alkaline solution, the GC/Pd/Bi electrode shows, during the
positive sweep, a complex wave that appears in the range of

I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
653
Table 2
Effect of the applied potential on the GC/Pd electrode in non-deaerated 50 mM
HClO4 solution containing 200 ␮M Bi(NO3)3 on the bismuth surface con-
centration (Γ Bi) and charge associated to the oxidation process of 3 mM
propionaldehyde
Applied potential (V)
Γ Bi (nmol cm−2
)
PdxBiy
Charge (mC)
−0.8
−0.4
0.0
0.4
0.8
2.36
1.74
1.41
1.29
–
Pd78Bi22
Pd85Bi15
Pd88Bi12
Pd86Bi14
Pd
0.236
0.249
0.285
0.273
0.095
The potential was applied for 10 min under stirred solution with a magnetic bar
at about 300 rpm.
Other experimental conditions as in Table 1.
(i.e., 25 min of electrode immersion at open circuit), while the
catalytic efficiency towards the electrooxidation of propionalde-
hyde reached about the 90% of the maximum observed value.
Thus, the catalytic efficiency for the electrooxidation of alde-
hydes increases linearly with the bismuth loading up to values
of about 1 nmol cm−2 and increases slowly for higher surface
concentration of bismuth species.
Fig. 4. Comparison of the voltammograms relevant to the GC/Pd/Bi (solid
curve) and GC/Pd (dashed curve) electrodes in presence of 3.2 mM hexanale.
Other experimental conditions as in Fig. 3A.
to the enhancement of the kinetics of the electrooxidation
processes through a bifunctional mechanism [33,34] in which
Pd and Bi species promote specific interactions with reactants
and/or reaction intermediate.
The effect of applied potential during the deposition pro-
cedure on the efficiency of bismuth adsorption and catalytic
performance are also evaluated. Table 2 shows the relevant
results. As can be seen, increasing the applied potentials in the
range comprised between −0.8 and 0.4 V during the deposition
procedure, the surface concentration of bismuth species (Γ Bi
nmol cm−2) exhibit a decrease, while the resulting composite
films show an electrocatalytic activity almost independent from
the deposition conditions and film composition. On the con-
trary, for higher applied potentials, a sensible decrease of the
bismuth coverage with subsequent attenuation of the catalytic
performance was observed. This result suggests that, at high
applied potentials a partial transformation of bismuth oxide into
In order to elucidate some aspects of the kinetics of the
bismuth adsorption, various time of immersion of the GC/Pd
electrode in non-deaerated 200 ␮M Bi(NO3)3 solution at open
circuit and under various applied potentials were investigated.
Assuming that the bismuth oxidation and palladium reduction
processes (i.e., Bi0/Bi3+; Pd2+/Pd) are 100% efficient on the
voltammetric time scale, and assuming an almost homogeneous
surface distribution of the Pd–Bi species, the film composi-
tion can be tentatively evaluated by electrochemical technique
measuring the charge associated under the wave Ia1 and Ic1
related to the oxidation and reduction of bismuth and palladium
species, respectively. Table 1 summarizes the relevant results.
As can be seen, after 5 min of time immersion, the bismuth
coverage reached about 30% of the maximum coverage value
the cationic species such as Bi(OH)2+, Bi6(OH)12 or nega-
6+
(2m+n−3)−
tively charged BiOm(OH)n
species may facilitate the
desorption process of the bismuth adlayer from the electrode
surface. Alternatively, the formation of higher valence oxides of
palladium such as PdO2 and PdO3 species can induce a partial
inhibition of the adsorption process of bismuth species on the
electrode surface.
Table 1
Effect of the immersion time of the GC/Pd electrode in non-deaerated 50 mM
HClO4 solution containing 200 ␮M Bi(NO3)3 on the bismuth surface con-
centration (Γ Bi) and charge associated to the oxidation process of 3 mM
propionaldehyde
Fig. 5 shows some of the chronoamperometric curves (I–t) for
the GC/Pd electrode recorded at a constant potential of −0.2 V
versusSCEin50 mMNaOHsolutioninpresenceof2 mMpropi-
onaldehyde. After the addition of 50 ␮M Bi3+ in solutions (curve
A), immediately the electrooxidation current increases without
undesired electrode poisoning effects. In contrast, in absence
of Bi3+ ions in solution (curve B), the GC/Pd electrode shows
a rapid and sensible attenuation of the catalytic performance
and the relevant electrooxidation current decreases rapidly. It is
interesting to observe that, Bi3+ ions absorb efficiently on the
palladium surface also in presence of adsorbed analytes and/or
their intermediates. Thus, bismuth adatoms are able to prevent
the poison formation or to remove the formed organic adsorbate
molecules on the electrode surface through a third-body type
mechanism or through an electronic effect [35]. Fig. 6 com-
Immersion time (min)
Γ Bi (nmol cm−2
)
PdxBiy
Charge (mC)
0.0
2.0
5.0
10.0
15.0
25.0
–
Pd
0.087
0.223
0.255
0.261
0.282
0.290
0.32
0.54
1.01
1.18
1.94
Pd98Bi2
Pd96Bi4
Pd91Bi9
Pd92Bi8
Pd88Bi12
The charge (mC) was evaluated by integration of the oxidation wave during the
anodic sweep between −0.5 and 0.15 V at 50 mV s−1 in 50 mM NaOH solution
containing 3 mM propionaldehyde.
The bismuth surface concentration (Γ Bi) and the film composition (PdxBiy) were
evaluated by integration of the Ia1 and Ic1 peaks considering the Bi0 → BIII and
PdII → Pd0 redox process, respectively. The GC/Pd electrode was obtained by
cycling the potentials between 0.0 and −0.4 V vs. SCE for 50 cycles in non-
deaerated solution containing 0. 5 mM Na2PdCl6, 50 mM HClO4 and 25 mM
HCl.

654
I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
Fig. 5. Chronoamperometric measurements obtained at a GC/Pd electrode in
50 mM NaOH solution. (a) Addition of 2 mM propionaldehyde; after 200 s were
added 50 ␮M Bi3+ ions (curve A). The curve B is obtained in absence of Bi3+
ions. The solution was stirred with a magnetic bar at about 500 rpm. Electrode
potential: −0.2 V vs. SCE.
Fig. 7. XP detailed Pd 3d spectra of the composite Pd–Bi film obtained as in
Fig. 2.
pares the chronoamperometric responses of the GC/Pd/Bi and
GC/Pd electrodes, obtained at a constant potential of −0.2 V
versus SCE in 50 mM NaOH solution in presence of increasing
concentration of propionaldehyde. As a consequence of poi-
soning effects, the chronoamperometric profile of the GC/Pd
electrode appears rather insensitive to the propionaldehyde addi-
tion. On the contrary, the GC/Pd/Bi electrode responds rapidly
to millimolar concentration changes and the resulting calibra-
tion plot for propionaldehyde analysis is linear over the range
of 1.5–8 mM with a correlation coefficient better than 0.97.
investigation Pd 3d, Bi 4f, O 1s and C 1s regions were collected
and analysed for each specimen. Typical XP spectra of Pd 3d
and Bi 4f of a GC/Pd/Bi electrode are shown in Figs. 7 and 8,
respectively. As can be seen, the energy spectrum of the Pd 3d
shows two asymmetrical signals corresponding to the Pd 3d5/2
and Pd 3d3/2 with a spin-orbit splitting, in agreement with the
literature [36], of about 6.0 eV. In this study, the Pd 3d5/2 peak
gives four contributions at 335.2 0.1, 336.0 0.2, 337.1 0.3
and 338.1 0.5 eV, which on the basis of their BEs can be ten-
tatively assigned to Pd metal, PdOads, PdO and PdO2 species,
respectively [36–40].
3.2. XPS characterization of GC/Pd and GC/Pd/Bi
electrodes in alkaline solutions
The high-resolution spectrum of the Bi 4f shows two asym-
metrical main signals with a spin-orbit splitting of about
5.4 eV, corresponding to Bi 4f7/2 and Bi 4f5/2 contributes
[36]. The deconvolution of the Bi 4f7/2 signal produces gen-
In order to investigate the chemical nature and surface com-
position of the catalysts we first analyse the detailed spectra of
the GG/Pd/Bi composite electrodes and then compare the results
with those obtained from the GC/Pd electrodes. Throughout this
Fig. 6. Chronoamperometric measurements obtained at a GC/Pd/Bi and GC/Pd
electrode in 50 mM NaOH solution containing increasing concentration in
2.0 mM steps of propionaldehyde. Other experimental conditions as in Fig. 5.
Fig. 8. XP detailed Bi 4f spectra of the composite Pd–Bi film obtained as in
Fig. 2.

I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
655
erally three contributes centred at 158.1 0.3, 158.9 0.4 and
159.8 0.4 eV with a full-width at half-maximum of about
1.4 0.1 eV. The significant dispersions of the binding ener-
gies can be explained by the fact that the bismuth forms a
mixture of complexes species on the electrode surface and not
pure metal oxides species [40]. In addition, a shift to higher BE
values compared with those obtained with bismuth metal sam-
ples was previously observed [15], and can be understood as a
consequence of an electronic charge transfer to the Pd oxides.
Thus, in agreement with the literature data [15,40–42], the sig-
nal at 158.1 0.3 eV can be attributed to Bi0, while the signals
at 158.9 0.4 and 159.8 0.4 eV can be related to BiIII oxide
species. Tentatively, the peak at 158.9 0.4 eV can be attributed
(2m+n−3)−
to BiOOH, Bi(OH)3 and/or the BiOm(OH)n
species,
while the peak at 159.8 0.4 eV can be attributed to Bi2O3
oxides. In this study the surface composition of the GC/Pd/Bi
electrode was mainly based on the measured areas of the XP sig-
nal for the Pd 3d and Bi 4f peaks. Thus the chemical composition
of the film, was monitored as a function of the immersion time of
the GC/Pd electrode in non-deaerated 200 ␮M Bi(NO3)3 solu-
tion at open circuit. In addition, the chemical composition of the
filmsevaluatedbyXPStechniquewascomparedwiththesurface
concentration of adsorbed Bi species (Γ Bi) calculated by elec-
trochemical technique. The relevant results are summarized in
Table3. Ascanbeseen, after5 minoftheimmersiontimeinsolu-
tion containing 200 ␮M Bi(NO3)3, the average composition of
the film was Pd95Bi5. These results are in agreement with those
obtained in the previous section, where a fast and irreversible
adsorption process is established between palladium and bis-
muth complexes species. The further examination of Table 3
reveals that the shape and the BEs of the various contributions
Fig. 9. XP detailed O 1s spectra of the composite Pd–Bi film obtained as in
Fig. 2 and polarized at −0.4 V vs. SCE for 5 min in 50 mM NaOH solution.
of the Pd 3d signal relevant to GC/Pd/Bi electrode are identical
to those obtained for the GC/Pd electrode. Thus, although the
Pd–Bi film shows interesting electrocatalytic activity towards
the electrooxidation of aliphatic aldehydes, the present XPS
study does not indicate any chemical interaction between the
palladium and bismuth active species.
It is interesting to observe that the film composition eval-
uated here, is in good accordance with those calculated by
electrochemical technique. The small differences in the metal
percent composition (i.e., PdxBiy) of the composite films can be
explained by considering, in first approximation, the different
depth-profile sampling between the electrochemical and XPS
technique. Thus the surface composition of the GC/Pd/Bi elec-
trodes estimated by electrochemical and XPS techniques are in
good agreement, indicating that the quantitative analysis proce-
dures adopted here are sufficiently reliable.
The deconvolution of the O 1s signal exhibits generally four
well-resolved contributions with BEs ranging between 530 and
536 eV. A typical O 1s curve fitting of a GC/Pd/Bi electrode
polarized at −0.4 V versus SCE is shown in Fig. 9. Unfor-
tunately, the O 1s signal do not contribute to an exhaustive
elucidation of the chemical surface composition of the film,
because the specific assignment of the O 1s peaks is more com-
plex and rather ambiguous owing to partial overlapping of the
various oxygen species such as OH− ions, chemisorbed water
and oxygen-containing organic compounds derived by oxidized
carbonaceous species (i.e., glassy carbon substrate, adventitious
carbon contaminants derived by diffusion pump oil in the resid-
ual gas, etc.). In addition, the contemporary presence of palla-
dium and bismuth oxides induces further complications and the
analysis of the O 1s peak is further complicated by the interfer-
ence with the neighbouring Pd 3p3/2 signal [37] (BE = 531.4 eV).
Nevertheless, in this study an analytical approach regarding the
comparisonoftherelativeareasofthePd3d, Bi4fandO1speaks
Table 3
XPS data (BEs, eV) of the GC/Pd/Bi electrodes and chemical composition of
the active film in terms of PdxBiy as a function of the immersion time of the
GC/Pd electrode in non-deaerated 200 ␮M Bi(NO3)3 solution at open circuit
Immersion
time (min)
Pd 3d5/2
Bi 4f7/2
(Γ Bi nmol cm−2
)
PdxBiy
0.0
5.0
335.1
336.0
336.9
338.0
–
–
–
–
–
335.2
336.2
337.4
338.6
158.0
159.2
160.2
1.56
1.66
1.81
Pd95Bi5
Pd93Bi7
Pd92Bi8
10.0
20.0
335.3
336.2
337.3
338.6
158.2
159.3
160.2
335.2
336.0
337.1
338.1
158.2
159.2
160.1
The chemical film composition (PdxBiy) was evaluated by measuring the areas
of the XP signal of the Pd 3d and Bi 4f corrected for the relevant sensitivity
factors.
The GC/Pd electrode was obtained as reported in Table 1.

656
I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
Table 4
4. Conclusions
XPS data (BEs, eV) of the GC/Pd/Bi and GC/Pd electrodes polarized for 5 min
at various potentials in 50 mM NaOH solution
Surface modification of palladium sites electrodeposited
on the glassy carbon electrodes were readily obtained in
non-deaerated 200 ␮M Bi(NO3)3 solution under open circuit
conditions for a given time. The spontaneous adsorption
of bismuth species induces the simultaneous formation of
hydroxyl-adsorbed species on the palladium film. The catalytic
activity of the GC/Pd/Bi electrode towards the electrooxidation
of some aliphatic aldehydes was evaluated by cyclic voltamme-
try and chronoamperometric techniques in moderately alkaline
solutions. The surface coverage of bismuth (Γ Bi) was evaluated
by electrochemical measurements and a direct correlation
with the catalytic reactivity towards the electrooxidation of
aliphatic aldehydes was verified. The surface composition of
the active Pd–Bi composite film was evaluated by XPS analysis
throughout the investigation of the detailed Pd 3d, Bi 4f and O
1s regions. After various electrochemical treatments, an average
Sample
Potential
(V)
Pd 3d5/2 Bi 4f7/2 O 1s
O 1s/M PdxBiy
GC/Pd/Bi −0.4
335.1
335.8
336.8
337.9
157.8
158.6
159.5
531.0 24.8
Pd95Bi5
Pd96Bi4
Pd98Bi2
532.5
534.0
535.6
GC/Pd
−0.4
335.2
336.0
336.6
337.6
530.5
532.4
534.1
536.3
2.1
5.2
2.1
7.4
3.1
GC/Pd/Bi
GC/Pd
0.15
335.1
336.0
337.0
337.8
157.9
158.5
159.5
530.6
532.1
533.7
535.5
0.15
0.8
335.2
355.9
336.7
337.7
530.4
532.1
533.9
536.2
chemical composition of the composite film of Pd96 2Bi4
2
was obtained. An unexpected high value of the O 1s/M atomic
ratio (i.e., O 1s/M = 24.8) in the double region of potentials
was also observed, indicating that the Bi adsorbs efficiently and
irreversibly oxygen and/or hydroxyl species on the electrode
surface.
GC/Pd/Bi
GC/Pd
335.2
336.0
337.2
338.3
158.5
159.4
530.6
532.2
533.9
535.9
0.8
335.3
336.2
337.1
337.8
530.2
531.9
533.8
535.8
Acknowledgment
This work was supported by Ministero dell’Universita’ e
della Ricerca Scientifica e Tecnologica (MURST, COFIN 2004).
The GC/Pd film was obtained as reported in Table 1.
The composite Pd–Bi film was obtained by immersion of the GC/Pd electrode
in non-deaerated 200 ␮M Bi(NO3)3 solution at open circuit for 10 min.
O 1s/M represents the atomic ratios taken between the areas of O 1s signal and
the sum of the Pd 3d + Bi 4f signals, corrected for the relevant sensitivity factors.
References
[1] R.R. Azdic, in: H. Gerischer (Ed.), Advances in Electrochemistry and Elec-
trochemical Engineering, vol. 13, Wiley, New York, 1984.
[2] N. Xonoglou, G.J. Kokkinidis, Electroanal. Chem. 12 (1984) 485.
[3] F. Kadirgan, B. Beden, C. Lamy, J. Electroanal. Chem. 136 (1982) 119.
[4] J.M. Feliu, A. Ferna`ndez-Vega, A. Aldaz, J. Clavilier, J. Electroanal. Chem.
256 (1988) 149.
[5] H. Druliolle, K.B. Kokoh, B. Beden, J. Electroanal. Chem. 385 (1995) 77.
[6] H.-W. Lei, H. Hattori, H. Kita, Electrochim. Acta 41 (1996) 1619.
[7] E. Leiva, T. Iwasita, E. Herrero, J.M. Feliu, Langmuir 13 (1997) 6287.
[8] V. Climent, E. Herrero, J.M. Feliu, Electrochim. Acta 44 (1998) 1403.
[9] Y. Morimoto, E.B. Yeager, J. Electroanal. Chem. 444 (1998) 95.
[10] F. Matsumoto, M. Harada, N. Koura, S. Uesugi, Electrochem. Commun. 5
(2003) 42.
[11] K.-I. Tsceng, M.-C. Yang, J. Electrochem. Soc. 150 (2003) H156.
[12] S.A. Campbell, R. Parson, J. Chem. Soc. Faraday Trans. 88 (2003) 833.
[13] R. Carbo`, R. Albalat, J. Claret, J.M. Feliu, J. Electroanal. Chem. 446 (1998)
79.
[14] S.P.E. Smith, H.D. Abrun˜a, J. Phys. Chem. B 102 (1998) 3506.
[15] G. Wittstock, A. Stru¨bing, R. Szargan, G. Werner, J. Electroanal. Chem.
444 (1998) 61.
is adopted, as consequence the exact assignment of the relevant
contributions under the O 1s signal is not essential. Looking at
Table 4 it is interesting to observe that the atomic ratios O 1s/M,
expressed as:
O 1s
M
O 1s
=
Pd 3d + Bi 4f
is strictly correlated to the presence of adsorbed bismuth species
on the palladium electrode surface and the applied potential.
In particular, under prolonged polarization at −0.4 V versus
SCE the GC/Pd/Bi and GC/Pd electrodes shown an atomic
ratios O 1s/M of 24.8 and 2.1, respectively. The above results
clearly show that the adsorbed bismuth adlayer undergoes a sur-
face redox process that leads, particularly in the double layer
region, to the formation of adsorbed oxygen-containing com-
plexes species. The significant high value of the O 1s/M atomic
ratio observed in the double layer region, suggests, that the Bi
species (i.e., in particular the Bi at lower oxidation state) adsorb
preferentially and efficiently oxygen and/or hydroxyl ions on
the electrode surface [43]. Thus, in agreement with the results
obtained in the previous section, the bismuth adsorption process
leads to significant formation of hydroxyl-adsorbed species on
the modified palladium surface.
[16] S.P.E. Smith, H.D. Abrun˜a, J. Electroanal. Chem. 467 (1999) 43.
[17] M. Wenkin, P. Ruiz, B. Delmon, M. Devillers, J. Mol. Catal. A 180 (2002)
141.
[18] I.G. Casella, M. Gatta, M. Contursi, J. Electroanal. Chem. 561 (2004) 103.
[19] T. Chierchie, C. Mayer, W.J. Lorenz, J. Electroanal. Chem. 135 (1982) 211.
[20] M. Elam, B.E. Conway, J. Electrochem. Soc. 135 (1988) 1678.
[21] J.-P. Chevillot, J. Farcy, C. Hinnen, A. Rousseau, J. Electroanal. Chem. 64
(1975) 39.
[22] R.C. Salvarezza, M.C. Montemayor, E. Fatas, A.J. Arvia, J. Electroanal.
Chem. 313 (1991) 291.

I.G. Casella, M. Contursi / Electrochimica Acta 52 (2006) 649–657
657
[23] A.E. Bolza´n, J. Electroanal. Chem. 380 (1995) 127.
[24] C.-C. Hu, T.-C. Wen, J. Electrochem. Soc. 141 (1994) 2996.
[25] M.I. Manzanares, A.G. Pavese, V.M. Solis, J. Electroanal. Chem. 310
(1991) 159.
[26] A.E. Bolza`n, A.J. Arvia, J. Electroanal. Chem. 322 (1992) 247.
[27] M.-C. Jeong, C.H. Pyun, In-H. Yeo, J. Electrochem. Soc. 140 (1993) 1986.
[28] I.G. Casella, Electrochim. Acta 44 (1999) 353.
[29] Z. Grubacˇ, M. Metikos-Hukovic´, Electrochim. Acta 44 (1999) 4559.
[30] S. Szabo, F. Nagy, J. Electroanal. Chem. 87 (1978) 261.
[31] J. Clavilier, J.M. Feliu, A. Aldaz, J. Electroanal. Chem. 243 (1988) 419.
[32] S.P.E. Smith, K.F. Ben-Dor, H.D. Abrun˜a, Langmuir 16 (2000) 787.
[33] M. Watanabe, S. Motoo, J. Electroanal. Chem. 60 (1975) 259.
[34] M. Watanabe, S. Motoo, J. Electroanal. Chem. 69 (1976) 429.
[35] E. Herrero, A. Fernandez-Vega, J.M. Feliu, A. Aldaz, J. Electroanal. Chem.
350 (1993) 73.
[36] C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, G.E. Mouilenberg
(Eds.), Handbook of X-Ray Photoelectron Spectroscopy, Perkin-Elmer
Corporation, Eden Prairie, Minnesota, 1978.
[37] K.S. Kim, A.F. Gossmann, N. Winograd, Anal. Chem. 46 (1974) 197.
[38] J. Pollmann, R. Franke, J. Hormes, H. Bo¨nnemann, W. Brijoux, A. Schulze
Tilling, J. Electron Spectros. Rel. Phenom. 94 (1998) 219.
[39] M. Brun, A. Berchet, J.C. Bertolini, J. Electron Spectros. Rel. Phenom. 104
(1999) 55.
[40] M. Wenkin, P. Ruiz, B. Delmon, M. Devillers, J. Mol. Catal. A 180 (2002)
141.
[41] C. Hinnen, C. Nguyen Van Huong, P. Marcus, J. Electron Spectros. Rel.
Phenom. 73 (1995) 293.
[42] K. Uchida, A. Ayame, Surf. Sci. 357 (1996) 170.
[43] U.W. Hamm, D. Kramer, R.S. Zhai, D.M. Kolb, Electrochim. Acta 43
(1998) 2969.