Mechanistic insight into oxide-promoted palladium catalysts for the electro-oxidation of ethanol

Article abstract of DOI:10.1002/cssc.201402062

Recent advancements in the development of alternatives to proton exchange membrane fuel cells utilizing less-expensive catalysts and renewable liquid fuels, such as alcohols, has been observed for alkaline fuel cell systems. Alcohol fuels present the advantage of not facing the challenge of storage and transportation encountered with hydrogen fuel. Oxidation of alcohols has been improved by the promotion of alloyed or secondary phases. Nevertheless, currently, there is no experimental understanding of the difference between an intrinsic and a synergistic promotion effect in high-pH environments. This report shows evidence of different types of promotion effects on palladium electrocatalysts obtained from the presence of an oxide phase for the oxidation of ethanol. The correlation of mechanistic in situ IR spectroscopic studies with electrochemical voltammetry studies on two similar electrocatalytic systems allow the role of either an alloyed or a secondary phase on the mechanism of oxidation of ethanol to be elucidated. Evidence is presented for the difference between an intrinsic effect obtained from an alloyed system and a synergistic effect produced by the presence of an oxide phase.

Full text of DOI:10.1002/cssc.201402062

CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201402062
Mechanistic Insight into Oxide-Promoted Palladium
Catalysts for the Electro-Oxidation of Ethanol
Ulises Martinez, Alexey Serov, Monica Padilla, and Plamen Atanassov*[a]
Recent advancements in the development of alternatives to
proton exchange membrane fuel cells utilizing less-expensive
catalysts and renewable liquid fuels, such as alcohols, has been
observed for alkaline fuel cell systems. Alcohol fuels present
the advantage of not facing the challenge of storage and
transportation encountered with hydrogen fuel. Oxidation of
alcohols has been improved by the promotion of alloyed or
secondary phases. Nevertheless, currently, there is no experi-
mental understanding of the difference between an intrinsic
and a synergistic promotion effect in high-pH environments.
This report shows evidence of different types of promotion ef-
fects on palladium electrocatalysts obtained from the presence
of an oxide phase for the oxidation of ethanol. The correlation
of mechanistic in situ IR spectroscopic studies with electro-
chemical voltammetry studies on two similar electrocatalytic
systems allow the role of either an alloyed or a secondary
phase on the mechanism of oxidation of ethanol to be eluci-
dated. Evidence is presented for the difference between an in-
trinsic effect obtained from an alloyed system and a synergistic
effect produced by the presence of an oxide phase.
Introduction
The necessity for clean and renewable alternative fuel sources
to reduce our dependence on petroleum-based fuels has in-
creased interest in the use of ethanol as a fuel. Ethanol can be
obtained through the fermentation of renewable biomass re-
sources, such as sugar cane and corn. Furthermore, ethanol
has low toxicity, a high boiling point, and, more importantly,
an energy density similar to that of gasoline (8 kWhkgꢀ1).
Direct ethanol fuel cells (DEFCs) are a promising emerging
technology that can transform the chemical energy stored in
ethanol into electrical energy to power devices ranging from
portable applications to automobiles.[1] Nevertheless, the com-
mercialization of DEFCs has been limited by the use of expen-
sive and scarce platinum catalysts generally used for the oxida-
tion of small molecules. Recent studies have shown that palla-
dium catalysts, which are more abundant and less costly than
platinum, offer better performances than platinum when used
in an alkaline environment.[2–5] Furthermore, alkaline electro-
lytes offer the advantage of providing an environment in
which transition metals and rare-earth metals can be used as
co-catalysts to promote the catalytic activity and functionality
of palladium-based systems.
Developing better catalysts based on rational design re-
quires a fundamental understanding of the role that each com-
ponent plays in the reaction mechanism. Fundamental studies
on the ethanol oxidation reaction (EOR) on platinum-based
catalysts in acidic solutions have been widely reported in the
literature;[6–11] however, there are only a limited number of
mechanistic studies reported for alkaline environments. The
EOR is a very complex reaction with multiple pathways. In an
alkaline electrolyte, ethanol oxidation can result in 2-, 4-, or 12-
electron transfer [Eqs. (1)–(3)]. Unquestionably, the ideal cata-
lyst will convert ethanol into CO2 in a 12-electron transfer reac-
tion to obtain the maximum energy density from its oxidation.
CH3CH2OH þ 2 OHꢀ ! CH3CHO þ 2 H2O þ 2 eꢀ
CH3CH2OH þ 4 OHꢀ ! CH3COOH þ 3 H2O þ 4 eꢀ
CH3CH2OH þ 12 OHꢀ ! 2 CO2 þ 9 H2O þ 12 eꢀ
ð1Þ
ð2Þ
ð3Þ
Improved stability, reactivity, and functionality obtained from
alloyed and supported catalysts have been attributed to two
different effects: promotion or intrinsic effects. Xu et al. dem-
onstrated a promotion effect on Pd electrocatalysts from oxide
phases (CeO2, NiO, Co3O4, and Mn3O4) for the oxidation of dif-
ferent alcohols.[3,12] Additionally, oxophilic metals, such as
ruthenium, tin, and nickel, improve the intrinsic catalytic activi-
ty of platinum and palladium electrocatalysts for the oxidation
of alcohols.[13–17] However, there few studies report the im-
proved catalytic activity of palladium for the oxidation of etha-
nol in alkaline environments. Du et al. reported the benefit of
alloying tin with palladium on carbon-supported catalysts.[14]
The study reports an increase in current density of two times
over commercial Pd/C catalysts, with an optimal Sn content of
14 at%. However, reaction studies suggest that ethanol is only
[a] Dr. U. Martinez,+ Dr. A. Serov, M. Padilla, Dr. P. Atanassov
Center for Emerging Energy Technologies
The University of New Mexico
1 University of New Mexico—MSC01 1120
Albuquerque, NM 87131 (USA)
E-mail: plamen@unm.edu
[+] Present Address:
Materials Physics and Applications
Los Alamos National Laboratory
Los Alamos, NM 87545 (USA)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201402062.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2351

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
partially oxidized to produce acetic acid as the main product.
Similarly, He et al. reported the promoting effect of Sn on
Pd2.5Sn/C catalysts, which resulted in increased current densi-
ties in 0.25m KOH.[18] To date, reports on using Sn as a co-cata-
lyst have been limited to PdꢀSn alloys with no reports on the
promotion effect of a SnO2 phase and, more importantly, there
are no reports that offer further understanding of the mecha-
nism of the promotion of alloyed/oxide phases on Pd electro-
catalysts for the oxidation of ethanol in alkaline environments.
This is the first report on the synthesis and characterization
of PdꢀSnO2 unsupported electrocatalysts with current densities
two times higher than that of pure Pd for the oxidation of eth-
anol in an alkaline electrolyte. Electrochemical activity studies
have been complemented with an insight into the mechanism
of oxidation based on in situ FTIR electrochemical spectroscop-
ic studies. Reaction mechanisms are presented for two differ-
ent electrolyte concentrations, 0.1 and 1m KOH, with an em-
phasis on the pH effect in relation to the electrochemical per-
formance and selectivity of the
The presence of both Pd and SnO2 phases was confirmed by
XRD. Figure 2 displays XRD spectra for both PdꢀSnO2(SSM) and
PdꢀSnO2(SP). Diffraction peaks at 2q angles of 40.0, 46.3, 67.7,
and 81.58 confirm the face-centered cubic (fcc) palladium
planes (111), (200), (220), and (311), respectively. Similarly,
peaks at angles of 26.6, 33.8, and 51.88 confirm diffraction
planes of (110), (101), and (211), respectively, for the SnO2
phase. The crystallite size of the Pd nanoparticles was calculat-
ed from the Pd (111) peak by using the Scherrer equation.
Crystallite sizes of the Pd nanoparticles are similar for both cat-
alysts: 6.5 nm for PdꢀSnO2(SP) and 6.9 nm for PdꢀSnO2(SSM).
On the other hand, the crystallite size for SnO2 differed de-
pending on the synthetic method. SnO2(SP) has a crystallite
size of 3.3 nm, whereas SnO2(SSM) resulted in a higher crystal-
lite size of 4.6 nm based on the (110) peak for SnO2.
The electrochemical activity of the PdꢀSnO2 electrocatalysts
relative to a Pd-only electrocatalyst is displayed in Figure 3.
Identical behavior is observed for PdꢀSnO2(SSM) and Pd when
EOR.
Results and Discussion
Morphological differences be-
tween two synthesized PdꢀSnO2
electrocatalysts, PdꢀSnO2(SSM)
(SSM=sacrificial
support
method) and PdꢀSnO2(SP), are
presented in Figure 1. Two differ-
ent methods were utilized for
the synthesis of the SnO2 phase,
which was then followed by the
chemical reduction of Pd onto
the oxide phase. Spray pyrolyzed
(SP) SnO2 consisted of submi-
cron-sized spherical particles
composed of agglomerated
nanoparticles (Figure 1, right,
and Figures S1 and S3 in the
Supporting Information). Con-
versely, tin oxide synthesized by
the sacrificial support method,
SnO2(SSM), consisted of large
micron-sized bulk particles (also
formed from small agglomerated
nanoparticles; Figure S2 in the
Supporting Information) with
porosity in the tens of nanome-
ters (Figure 1, left, and Figure S2
in the Supporting Information).
A more thorough analysis of the
dispersion of Pd and SnO2
phases by using SEM, energy-
dispersive spectroscopy (EDS),
and high-resolution (HR) TEM is
presented in Figures S1–S3 in
the Supporting information.
Figure 1. SEM micrographs of PdꢀSnO2 electrocatalysts: left: PdꢀSnO2 (SSM), right: PdꢀSnO2 (SP); top: 10k mag-
nification, middle: 100k magnification, bottom: contrast image of 100k magnification for the ease of identification
of Pd and SnO2 phases.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2352

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Pd and PdꢀSnO2(SSM) must have the same active sites. Never-
theless, the presence of SnO2 shifted the oxidation peak to
lower potentials (Figure 3b). This shift could be caused by the
increase in OHꢀ at the reaction surface by the SnO2 phase,
which increased the rate of reaction and produced higher cur-
rents between ꢀ0.40 and ꢀ0.05 V.
The same promotion effect observed for the HOR region of
PdꢀSnO2(SP) electrocatalysts exists for the oxidation of ethanol.
A significantly lower onset potential was observed for Pdꢀ
SnO2(SP): ꢀ0.90 V (ꢀ0.75 V for Pd and ꢀ0.70 for PdꢀSnO2
(SSM)). Additionally, the peak current density for PdꢀSnO2(SP)
is two times higher than those of PdꢀSnO2(SSM) and Pd (Fig-
ure 3b). The lower onset potential observed must be caused
by an intrinsic effect of Pd(Sn) alloying at the surface of the
spherical SnO2 particles. This phenomenon is further discussed
in the Supporting Information through Rietveld refinement of
the XRD data along with HRTEM (Figures S2 and S3 in the Sup-
porting Information). Hence, the PdꢀSnO2(SP) electrocatalyst is
truly a ternary catalyst: Pd(Sn)ꢀSnO2. Peak current densities are
observed because of the passivation process of the palladium
electrocatalysts. Because PdꢀOH is oxidized into PdO, further
adsorption of ethanol is inhibited, which results in a decrease
in the current density.
Figure 2. XRD spectra for the PdꢀSnO2 electrocatalysts.
The main goal in the design of novel electrocatalysts is to
obtain high electrochemical activity with the required selectivi-
ty. Therefore, these reports should be complemented with an
insight into the reaction mechanism to obtain the desired opti-
mization.
Reports of mechanistic studies based on in situ experiments
for the EOR in alkaline environments are limited. Comprehen-
sive studies on platinum electrodes in alkaline electrolytes
have been reported by Lai et al.[19] and Christensen et al.[20] In
situ FTIR studies on palladium electrodes have been published
by Zhou et al.[21] and Fang et al.[22] Zhou et al. provided infor-
mation about intermediates and products of ethanol electro-
oxidation in a solution of 0.1m NaOH and 0.1m ethanol. Their
studies reported acetate as the main product of the reaction
with very low selectivity to CO2. Moreover, they suggested that
the depletion of NaOH from the interface layer might have af-
fected the selectivity of e ethanol oxidation, but indicated the
need for further experiments for confirmation. Fang et al. re-
ported ethanol electro-oxidation at different pH values (0.01 to
5m NaOH) with analysis of products and intermediates by
using in situ FTIR spectroelectrochemistry. They observed an
optimal electrochemical performance in a 1m NaOH electro-
lyte. Product analysis found that acetate was the main product
for concentrations higher than 0.5m NaOH. Evidence of CꢀC
cleavage was observed at pH values lower than 13, but they
did not provide a possible mechanism for the reaction leading
to the production of CO2. Both of these reports provide in-
sights into the role of hydroxide ions in the selectivity and
overall performance of the electro-oxidation of ethanol. How-
ever, no clear understanding of the mechanism of reaction to
achieve CꢀC cleavage has been provided and is therefore still
required. Herein, we aim to provide a better insight into the
mechanism of the reaction at two different pH values to
obtain a more comprehensive understanding of the EOR in al-
Figure 3. Cyclic voltammograms of PdꢀSnO2 electrocatalysts in a) 1m KOH
and b) 1m ethanol/1m KOH at 228C, 20 mVsꢀ1; inset: onset potentials for
the EOR.
no fuel is present (Figure 3a); this indicates that there is no
effect on the H and OH adsorption properties of Pd by the ad-
dition of SnO2. Conversely, the onset potential and peak cur-
rent density potential for the hydrogen oxidation reaction
(HOR) region of the PdꢀSnO2(SP) catalyst are shifted to lower
potentials by about 100 mV; this indicates a probable intrinsic
promotion effect of the SnO2 on Pd. No effect on the onset po-
tential for the oxidation of ethanol is observed for PdꢀSnO2-
(SSM) when compared with that for Pd; this means that both
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2353

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
kaline environments by using palladium electrocatalysts pro-
moted by an oxide phase.
for ethanol (n˜ =1045 cmꢀ1) and acetate (n˜ =1410 cmꢀ1) as
a function of the potential applied. The overlay of these peak
areas together with the FSV provides important information
about the mechanism of reaction for the oxidation of ethanol.
The voltammogram in Figure 4b displays two onset poten-
tials at ꢀ0.94 and ꢀ0.61 V. These two potentials represent two
different electron-transfer reactions. At ꢀ0.94 V, ethanol spe-
cies are starting to be adsorbed at the surface of the catalyst.
Therefore, the electron-transfer reaction occurring at this po-
tential must be ethanol dehydrogenation and adsorption, as
described by Equations (4)–(6):
Detailed electrochemical analysis was performed by in situ
IR reflection absorbance spectroscopy (IRRAS) to provide
a better understanding of the mechanism of the reaction of
the highly active PdꢀSnO2(SP) electrocatalyst. Figure 4a dis-
plays the IR spectra obtained for the EOR in a 1m KOH and 1m
ethanol electrolyte. The adsorption of ethanol at the surface of
the catalyst and its oxidation is identified by characteristic IR
peaks for ethanol at n˜ =1086 and 1045 cmꢀ1, which corre-
spond to the CO stretching in CCO (Table 1).[21] In conjunction
with the IR spectra, Figure 4b displays the FSV for the corre-
sponding oxidation of ethanol at 1 mVsꢀ1; plotted on the op-
posite y axis are the calculated areas for characteristic peaks
CH3CH2OH ! CH3CH2OHads
CH3CH2OHads ! CH3CHOHads þ Hads
Hads þ OHꢀ ! H2O þ 1 eꢀ
ð4Þ
ð5Þ
ð6Þ
According to DFT calculations by Du et al.,[14] ethanol ad-
sorption is exothermic and does not involve any charge trans-
fer. Therefore, the current increase observed following ethanol
adsorption in Figure 4b must be due to the dehydrogenation
of the a-carbon [Eq. (5)][14,23] and further oxidation of adsorbed
hydrogen [Eq. (6)]. As the potential increases, the current
reaches a limit at about ꢀ0.7 V, which coincides with the maxi-
mum reached by the ethanol adsorbed species. At this poten-
tial, all active sites must be covered by adsorbed ethanol spe-
cies, CH3CHOHads. No more adsorbed hydrogen is oxidized,
which results in a decrease in current. Furthermore, the onset
potential for the oxidation of the CH3CHOHads species can be
observed at ꢀ0.61 V. At this onset potential, acetate species
begin to appear (Figure 4a and b). Acetate formation can be
observed by means of the characteristic IR bands appearing at
n˜ =1549 and 1410 cmꢀ1, which correspond to the OꢀCꢀO out-
of-phase and in-phase stretching, respectively.[20–22] For the
CH3CHOHads species to convert into acetate species, these ad-
sorbed species must undergo a series of dehydrogenation
steps and further oxidation to produce the acetate ion. To
better understand this mechanism of reaction and to gain an
insight into the role of the SnO2 phase in promoting the cata-
lytic performance of Pd, further in situ IRRAS studies were per-
formed by using SnO2(SP) particles under the same conditions
as those used for the oxidation of ethanol on PdꢀSnO2(SP)
(Figure S6 in the Supporting Information).
Figure 4. a) IRRAS spectra of the ethanol electro-oxidation reaction by Pdꢀ
SnO2(SP) catalyst in a 1m KOH electrolyte. b) Forward sweep voltammogram
(FSV) of the oxidation reaction. Conditions: 1m ethanol, 1 mVsꢀ1, 228C,
0 rpm.
Correlating the results obtained from the analysis of SnO2
(Figure S6 in the Supporting Information) with the analysis
from PdꢀSnO2(SP) indicates that the role of SnO2 in the selec-
tivity and performance of PdꢀSnO2(SP) is to provide hydroxide
species to the reaction surface, which cause an increase in the
turnover frequency for the formation of acetate ions, but in-
hibit ethanol from further reacting to completely oxidize to
CO2. Therefore, the mechanism of the reaction to convert ad-
sorbed ethanol species into acetate in a 1m electrolyte solu-
tion will involve the further dehydrogenation of CH3CHOHads to
form the acetate ion [Eqs. (7)–(9)]:
Table 1. IRRAS band assignments for the electro-oxidation of ethanol.
Wavenumber [cmꢀ1
]
Assignment
3200–3400
2340
1600
OH stretch
CO2 asymmetric stretch[10, 41]
HOH bend
1549, 1410
1270
1086, 1045
OCO stretch (in acetate)[20–22]
CO stretch (in ꢀCOOH)[7,9]
CO stretch (in ethanol)[21]
CH3CHOHads þ 2 OHꢀ ! CH3COads þ 2 H2O þ 2 eꢀ
ð7Þ
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2354

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
CH3COads þ OHads ! CH3COOHads þ H2O
CH3COOHads þ OHꢀ ! CH3COOꢀ þ H2O
ð8Þ
ð9Þ
These findings are corroborated by the presence of acetate
species appearing exactly at the onset potential of the ethanol
oxidation, as observed in Figure 4b. The amount of acetate
produced increases as the potential is increased and reaches
a maximum at the peak current density occurring at ꢀ0.10 V.
This maximum is observed as a result of the passivation of the
interface layer Pd catalyst by PdO, which starts to form at
around ꢀ0.15 V;[24,25] hence, at this potential, Pd loses its elec-
trochemical activity and causes a decrease in current. No other
products or intermediates are eminent in the IR spectra, al-
though acetaldehyde formation, which has been reported as
a possible intermediate of acetate formation, could be masked
by the predominant OꢀH vibration of the electrolyte occurring
at n˜ ꢁ1630 cmꢀ1.
Electrolyte concentration effects for the EOR have been sug-
gested by several researchers. Lai et al. studied the electro-oxi-
dation of ethanol on platinum and gold electrodes, and report-
ed a relationship between the nature of the electrolyte and
the selectivity and electrochemical activity of the reaction.[19]
Their results showed that higher electrochemical activity was
obtained at higher pH values when compared in different
buffer solutions; furthermore, CꢀC bond breaking was ob-
served in the absence of strongly adsorbed anions. Fang et al.
published a report on in situ FTIR studies performed on palla-
dium electrodes in electrolytes with pH values ranging from
0.1 to 5m NaOH.[22] They found that CꢀC bond cleavage only
occurred at electrolyte concentrations lower than 0.5m. How-
ever, no mechanism was suggested by the authors for the
pathway leading to CꢀC bond cleavage, and thus, the need for
a further understanding of the mechanism of oxidation leading
to the complete oxidation of ethanol in alkaline solutions re-
mains. The following section reports studies performed to cor-
roborate these findings and we propose a mechanism for oxi-
dation leading to CO2/CO3ꢀ formation in 0.1m hydroxide by
using the PdꢀSnO2(SP) electrocatalyst.
Figure 5. a) IRRAS spectra of ethanol oxidation by PdꢀSnO2(SP) in a 0.1m
KOH electrolyte. b) FSV of the oxidation reaction. Conditions: 1m ethanol,
1 mVsꢀ1, 228C, 0 rpm.
nism is the adsorption of ethanol onto the electrocatalyst sur-
face [Eqs. (4)–(6)]. Following the adsorption of ethanol, the
onset potential for the EOR occurs at ꢀ0.45 V, which is 160 mV
higher than the onset potential observed in 1m KOH. Thus,
the further dehydrogenation of ethanol to form an adsorbed
acetyl intermediate must occur from PdꢀOHads species that
form at higher potentials[24] with hydroxide ions provided by
SnO2, as observed in a 1m KOH electrolyte [Eq. (10)].
The in situ IRRAS spectra of the EOR in a 0.1m KOH electro-
lyte are reported in Figure 5a. Intermediates and products ap-
pearing in Figure 5b differ from those present in a 1m KOH
electrolyte solution. These differences include the absence of
predominant peaks for acetate products at n˜ =1549 and
1410 cmꢀ1, as seen in a 1m electrolyte, as well as the appear-
ance of new species at n˜ =1270 and 2340 cmꢀ1. Although
there are no predominant peaks for acetate, as observed in
Figure 4a, the peak at n˜ =1408 cmꢀ1 together with the broad-
ening of the peak centered at n˜ ꢁ1626 cmꢀ1 might indicate
the formation of a CH3COO product or intermediate participat-
ing in a reaction mechanism similar to that outlined by Equa-
tions (7) and (8).
CH3CHOHads þ 2 OHads ! CH3COads þ 2 H2O
ð10Þ
Formation of acetic acid species (n˜ =1270 cmꢀ1) appear at
the onset potential (Figure 5b). Additionally, a distinctive
sudden increase in current occurring at ꢀ0.16 V correlates with
the appearance of CO2 species (n˜ =2340 cmꢀ1) and is therefore
the potential at which CꢀC cleavage occurs. Although it is ex-
pected that the complete oxidation of ethanol in an alkaline
environment would result in carbonate formation, the combi-
nation of the consumption of limited available hydroxide spe-
cies to partially dehydrogenate ethanol [Eq. (10)] as well as the
dissociative adsorption of water on Pd, which occurs at these
intermediate potentials [Eq. (11)], generates a low-pH environ-
ment near the reaction surface in which CO2 and acetic acid
Figure 5b provides an overlay of the forward sweep of the
electro-oxidation reaction along with the integrated peak areas
for the species of interest: ethanol (n˜ =1045 cmꢀ1), acetic acid
(n˜ =1270 cmꢀ1), and CO2 (n˜ =2340 cmꢀ1). Analogous to electro-
oxidation in a 1m KOH electrolyte, the first step of the mecha-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2355

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
could be formed.
situ IR studies at different electrolyte concentrations can be ap-
plied to the rational design of next-generation electrocatalysts
for the oxidation of many fuels. For ethanol oxidation, im-
proved results were obtained from a ternary Pd(Sn)ꢀSnO2 cata-
lytic system. These catalysts would require alloyed electron-
dense metals that promote the electronic activity of Pd, while
retaining its active sites, as well as surface oxides to assist with
the increase in surface OH ions.
Pd þ H2O ! PdO þ 2 Hþ þ 2 eꢀ
ð11Þ
It is possible that PdO, which forms at around ꢀ0.25 V,[24]
might assist in the CꢀC cleavage of the adsorbed acetyl spe-
cies by creating an oxametallacyclic ethanol conformation
analogous to that hypothesized for the PtꢀRh system.[26]
Furthermore, different final products, CO2 and acetic acid,
could possibly be a result of different crystallographic planes,
as suggested by Wang and Liu.[27] Close-packed (111) planes
are limited to produce acetic acid as the final product
[Eq. (12)], following the oxidation of the adsorbed acetyl spe-
cies from Equation (10).
Experimental Section
Synthesis: Two different methods were utilized to synthesize SnO2:
spray pyrolysis (SP)[6,15,28–30] and the sacrificial support method
(SSM).[31–39] SnO2(SP) was synthesized by using a SnCl4 precursor
(Sigma–Aldrich) dissolved in deionized water to a final concentra-
tion of 5 wt%. The precursor solution was ultrasonically atomized
and pyrolyzed by using a quartz tube in a furnace operating at
5008C with air (0.5 Lminꢀ1) as the carrier gas. Pyrolyzed particles
were air dried and collected on a Teflon filter. Collected oxide pre-
cursor was heat treated in air at 3008C for 2 h. Alternately, SnO2
was also synthesized through a sacrificial support method. The sac-
rificial support method was based on the implementation of silica
materials instead of conventional carbon supports. First, silica
(Cab-O-Sil, EH-5, surface area: ꢁ400 m2 gꢀ1) was dispersed in water
by using an ultrasound bath. SnCl4 (Sigma–Aldrich) was then
added to the solution of silica, followed by drying at 858C over-
night. As-obtained, dry SnCl4/SiO2 composite material was calcined
at 3508C in air, resulting in the formation of SnO/SiO2. The total
loading of tin oxide on silica was calculated to be 20 wt%.
Pd(NO3)2 precursor was chemically reduced onto both oxides by
using an excess amount of NaBH4 added dropwise to the solution
of precursors/silica under constant ultrasonication. The resulting
black slurry was aged for 2 h. In case of SnO2(SSM), the silica sup-
port was etched by means of 7m KOH for 8 h. PdꢀSnO2 catalysts
were washed with deionized water until neutral reaction of water.
The Pd/Sn ratio (atomic) was selected to be 3:1. To compare the
catalytic activity of PdꢀSnO2 catalysts with Pd, unsupported Pd ma-
terial was synthesized by the sacrificial support method.
CH3COads þ OHads ! CH3COOH
ð12Þ
CꢀC cleavage has lower energy barriers on (100) surfaces fol-
lowing the dehydrogenation of the b-hydrogen [Eqs. (13)–
(15)].[23]
CH3COads þ OHads ! CH2COads þ H2O
CH2COads ! COads þ CH2ads
COads þ OHads ! CO2 þ Hþ þ eꢀ
ð13Þ
ð14Þ
ð15Þ
A combination of in situ IR spectroelectrochemical studies
and pseudo-steady-state sweep voltammograms has been em-
ployed to further understand the ethanol electro-oxidation re-
action in alkaline electrolytes by using palladium-based cata-
lysts. It was determined that, in high concentrations of OHꢀ,
the fate of the reaction was a 4-electron transfer with acetate
as the final product. As the pH is lowered, the limited concen-
tration of OHꢀ allows for the reaction to continue to complete
oxidation and produce CO2. It was also determined that the
role of an oxide phase was to increase the reaction rate by
providing OHꢀ to the electrochemical reaction surface. Further-
more, an intrinsic promotion effect was obtained from Pd(Sn)
alloying, which allowed for a lower onset potential.
Characterization: The phase composition and morphologies of the
synthesized catalysts were characterized by using XRD, SEM,
HRTEM, and EDS. Powder XRD spectra were recorded by using
a Scintag Pad V diffractometer with DataScan 4 software (MDI, Inc.)
for system automation and data collection. CuKa radiation (40 kV,
35 mA) was used with a Bicron Scintillation detector (with a pyrolit-
ic graphite curved crystal monochromator). Data sets were ana-
lyzed with Jade 9.5 Software (MDI, Inc.) by using the ICDD (Interna-
tional Center for Diffraction Data) PDF2 database (rev. 2004) for
phase identification.
SEM was performed on a Hitachi S-5200 instrument, with a resolu-
tion of 0.5 nm at 30 kV and 1.7 nm at 1 kV, equipped with a PGT
EDS system. TEM was performed on a JEOL 2010F FASTEM field-
emission gun scanning transmission electron microscope equipped
with an Oxford EDS system. The probe size was 1.0 nm and the ac-
celerating voltage was 200 kV.
Conclusions
For the first time, detailed mechanistic studies on palladium-
based electrocatalysts were performed. The presented in situ
IR reflection absorbance spectroscopy (IRRAS) studies con-
firmed the effect of electrolyte concentration on the kinetics
and selectivity of the ethanol electro-oxidation reaction. At 1m
KOH, SnO2 acted as a co-catalyst to provide hydroxide ions to
the interface layer, increase the turnover rate, and limit the
final product to acetate. Conversely, at 0.1m KOH, a limited
concentration of hydroxide ions near the reaction surface were
consumed by the partial dehydrogenation of ethanol to limit
the formation of acetate and allow for adsorbed ethanol spe-
cies to proceed to complete oxidation and produce CO2 aided
by the formation of PdꢀOH at higher potentials.
Electrochemical
measurements:
An
aqueous
suspension
(5 mgmLꢀ1 in H2O/isopropanol, 4:1) of each catalyst was prepared
and sonicated to disperse the powder in solution. Aliquots (10 mL)
of the aqueous suspension were deposited onto a glassy carbon
rotating disc electrode (RDE) with an area of 0.196 cmꢀ2 and al-
Although the results reported herein are for a specific cata-
lytic system, knowledge gained from the combination of in
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2356

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
[9] G. A. Camara, T. Iwasita, J. Electroanal. Chem. 2005, 578, 315–321.
[10] F. Vigier, C. Coutanceau, F. Hahn, E. M. Belgsir, C. Lamy, J. Electroanal.
Chem. 2004, 563, 81–89.
[11] J. P. I. de Souza, S. L. Queiroz, K. Bergamaski, E. R. Gonzalez, F. C. Nart, J.
Phys. Chem. B 2002, 106, 9825–9830.
[12] C. Xu, Z. Tian, P. Shen, S. P. Jiang, Electrochim. Acta 2008, 53, 2610–
2618.
[13] R. M. Modibedi, T. Masombuka, M. K. Mathe, Int. J. Hydrogen Energy
2011, 36, 4664–4672.
[14] W. Du, K. E. Mackenzie, D. F. Milano, N. A. Deskins, D. Su, X. Teng, ACS
Catal. 2012, 2, 287–297.
[15] E. E. Switzer, T. S. Olson, A. K. Datye, P. Atanassov, M. R. Hibbs, C. J. Cor-
nelius, Electrochim. Acta 2009, 54, 989–995.
lowed to air dry. Electrochemical measurements were conducted in
0.1 and 1m KOH with N2 as the purging gas. Deposited electroca-
talyst was stabilized by repeating cyclic voltammograms at
50 mVsꢀ1 prior to the addition of ethanol. Ethanol (HPLC grade;
Sigma–Aldrich) was added to the electrolyte at a concentration of
1m. A Pt coil was used as a counter electrode and a Hg/HgO elec-
trode (XR440, Radiometer Analytical SAS) was used as a reference
electrode. Cyclic voltammograms were normalized to the Pd active
mass, m, calculated from m=MQnꢀ1Fꢀ1, in which M is the molar
mass, Q is the charge transferred, n is the number of electrons
transferred, and F is the Faraday constant.
[16] Z. Zhang, L. Xin, K. Sun, W. Li, Int. J. Hydrogen Energy 2011, 36, 12686–
12697.
[17] T. Maiyalagan, K. Scott, J. Power Sources 2010, 195, 5246–5251.
[18] Q. He, W. Chen, S. Mukerjee, S. Chen, F. Laufek, J. Power Sources 2009,
187, 298–304.
[19] S. C. S. Lai, S. E. F. Kleijn, F. T. Z. ꢁztꢂrk, V. C. van Rees Vellinga, J. Koning,
P. Rodriguez, M. T. M. Koper, Catal. Today 2010, 154, 92–104.
[20] P. A. Christensen, S. W. M. Jones, A. Hamnett, J. Phys. Chem. C 2012, 116,
24681–24689.
[21] Z.-Y. Zhou, Q. Wang, J.-L. Lin, N. Tian, S.-G. Sun, Electrochim. Acta 2010,
55, 7995–7999.
[22] X. Fang, L. Wang, P. K. Shen, G. Cui, C. Bianchini, J. Power Sources 2010,
195, 1375–1378.
[23] E. D. Wang, J. B. Xu, T. S. Zhao, J. Phys. Chem. C 2010, 114, 10489–
10497.
[24] Z. X. Liang, T. S. Zhao, J. B. Xu, L. D. Zhu, Electrochim. Acta 2009, 54,
2203–2208.
IR spectroscopy: IRRAS experiments were performed at room tem-
perature with a Nicolet 6700 FTIR spectrometer equipped with
a mercury cadmium telluride (MCT) detector cooled with liquid ni-
trogen. The experimental setup was described previously.[6] For
each spectrum, 128 interferograms were added together at a reso-
lution of 8 cmꢀ1 with unpolarized light. Absorbance units of the
spectra are defined as A=ꢀlog(R/R0), in which R and R0 represent
reflected IR intensities that corresponded to the sample and refer-
ence single beam spectrum, respectively. Thus, a positive peak in
the resulting spectrum indicated the production of a species,
whereas a negative peak indicated consumption or a decrease in
the concentration of a species relative to the reference spectrum.
The reference spectrum was collected at ꢀ1.04 V (vs. Hg/HgO). A
thin layer of ink was deposited onto a polished glassy carbon elec-
trode with a diameter of 5 mm. A ZnSe hemisphere was used as
the IR window and the working electrode was pressed against the
window to create a thin solution layer with a thickness of a few mi-
crometers. The incident angle of the IR radiation passing through
the ZnSe window was 368.[40] Argon was used to purge the electro-
lyte while dry air was used to purge the spectrometer and cham-
ber, reducing the spectral interference from ambient CO2 and
water vapor.
´
[25] M. Grden, M. Lukaszewski, G. Jerkiewicz, A. Czerwinski, Electrochim. Acta
2008, 53, 7583–7598.
[26] A. Kowal, M. Li, M. Shao, K. Sasaki, M. B. Vukmirovic, J. Zhang, N. S. Mar-
inkovic, P. Liu, A. I. Frenkel, R. R. Adzic, Nat. Mater. 2009, 8, 325–330.
[27] H.-F. Wang, Z.-P. Liu, J. Am. Chem. Soc. 2008, 130, 10996–11004.
[28] U. Martinez, K. Asazawa, B. Halevi, A. Falase, B. Kiefer, A. Serov, M. Padil-
la, T. Olson, A. Datye, H. Tanaka, P. Atanassov, Phys. Chem. Chem. Phys.
2012, 14, 5512.
[29] D. A. Konopka, S. Pylypenko, P. Atanassov, T. L. Ward, ACS Appl. Mater.
Interfaces 2010, 2, 86–95.
[30] E. E. Switzer, A. K. Datye, P. Atanassov, Top. Catal. 2007, 46, 334–338.
[31] A. Serov, U. Martinez, A. Falase, P. Atanassov, Electrochem. Commun.
2012, 22, 193–196.
Acknowledgements
This material is based upon work supported by the DOE-EPSCoR
grant no. DE-FG02–08ER46530. U.M. was funded by the National
Science Foundation Graduate Research Fellowship under grant
no. DGE-0903444.
[32] A. Serov, M. H. Robson, K. Artyushkova, P. Atanassov, Appl. Catal. B
2012, 127, 300–306.
[33] A. Serov, M. H. Robson, M. Smolnik, P. Atanassov, Electrochim. Acta
2012, 80, 213–218.
[34] A. Serov, M. H. Robson, B. Halevi, K. Artyushkova, P. Atanassov, Electro-
chem. Commun. 2012, 22, 53–56.
[35] M. H. Robson, A. Serov, K. Artyushkova, P. Atanassov, Electrochim. Acta
2013, 90, 656–665.
[36] A. Serov, A. Aziznia, P. H. Benhangi, K. Artyushkova, P. Atanassov, E.
Gyenge, J. Mater. Chem. A 2013, 1, 14384–14391.
Keywords: electrochemistry
spectroscopy · palladium
· ethanol · fuel cells · IR
[1] E. H. Yu, X. Wang, U. Krewer, L. Li, K. Scott, Energy Environ. Sci. 2012, 5,
5668–5680.
[37] A. Serov, U. Martinez, P. Atanassov, Electrochem. Commun. 2013, 34,
185–188.
[2] C. Xu, L. Cheng, P. Shen, Y. Liu, Electrochem. Commun. 2007, 9, 997–
1001.
[38] A. Serov, M. H. Robson, M. Smolnik, P. Atanassov, Electrochim. Acta
2013, 109, 433–439.
[3] C. Xu, P. k. Shen, Y. Liu, J. Power Sources 2007, 164, 527–531.
[4] L. Ma, D. Chu, R. Chen, Int. J. Hydrogen Energy 2012, 37, 11185–11194.
[5] E. Antolini, Energy Environ. Sci. 2009, 2, 915–931.
[39] S. Brocato, A. Serov, P. Atanassov, Electrochim. Acta 2013, 87, 361–365.
[40] P. W. Faguy, N. S. Marinkovic, Appl. Spectrosc. 1996, 50, 394–400.
[41] X. H. Xia, H. D. Liess, T. Iwasita, J. Electroanal. Chem. 1997, 437, 233–240.
[6] D. A. Konopka, M. Li, K. Artyushkova, N. Marinkovic, K. Sasaki, R. Adzic,
T. L. Ward, P. Atanassov, J. Phys. Chem. C 2011, 115, 3043–3056.
[7] Q. Wang, G. Q. Sun, L. H. Jiang, Q. Xin, S. G. Sun, Y. X. Jiang, S. P. Chen,
Z. Jusys, R. J. Behm, Phys. Chem. Chem. Phys. 2007, 9, 2686–2696.
[8] R. B. Kutz, B. Braunschweig, P. Mukherjee, R. L. Behrens, D. D. Dlott, A.
Wieckowski, J. Catal. 2011, 278, 181–188.
Received: February 14, 2014
Revised: April 3, 2014
Published online on July 1, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2351 – 2357 2357