Experimental and computational studies of formic acid dehydrogenation over PdAu: Influence of ensemble and ligand effects on catalysis

Article abstract of DOI:10.1039/c6ta03654f

The critical role of the ligand effect and ensemble effect in enhancing formic acid (FA) dehydrogenation over PdAu catalysts was highlighted by both experimental and theoretical studies. FA dehydrogenation energy was calculated by DFT on PdAu model catalysts of different surface atomic arrangements. The Pd₃Au₁ surface exhibited the lowest reaction energy and kinetic barrier for FA dehydrogenation among four different PdAu surfaces. The Pd trimer played a critical role in stabilizing reaction intermediates. The experimental FA dehydrogenation activity of three different PdAu catalysts supported the theoretical results. In addition, the electronic interaction between the surface and subsurface layers also proved to contribute to the improved catalytic activity of PdAu catalysts via modification of Pd d states.

Full text of DOI:10.1039/c6ta03654f

Journal of
Materials Chemistry A
View Article Online
View Journal
PAPER
Experimental and computational studies of formic
acid dehydrogenation over PdAu: influence of
ensemble and ligand effects on catalysis†
Cite this: DOI: 10.1039/c6ta03654f
a
a
a
a
a
Jin Hee Lee,‡ Jinwon Cho,‡ Mina Jeon, Muhammad Ridwan, Hyun S. Park,
a
a
a
a
ab
Sun Hee Choi, Suk Woo Nam, Jonghee Han, Tae-Hoon Lim, Hyung Chul Ham*
abc
and Chang Won Yoon*
The critical role of the ligand effect and ensemble effect in enhancing formic acid (FA) dehydrogenation
over PdAu catalysts was highlighted by both experimental and theoretical studies. FA dehydrogenation
energy was calculated by DFT on PdAu model catalysts of different surface atomic arrangements. The
3 1
Pd Au surface exhibited the lowest reaction energy and kinetic barrier for FA dehydrogenation among
Received 2nd May 2016
Accepted 3rd August 2016
four different PdAu surfaces. The Pd trimer played a critical role in stabilizing reaction intermediates. The
experimental FA dehydrogenation activity of three different PdAu catalysts supported the theoretical
results. In addition, the electronic interaction between the surface and subsurface layers also proved to
contribute to the improved catalytic activity of PdAu catalysts via modification of Pd d states.
DOI: 10.1039/c6ta03654f
www.rsc.org/MaterialsA
transported using the existing infrastructure at a low cost. FA
can also release hydrogen in the presence of appropriate cata-
Introduction
Owing to the negative impacts of carbon-based fuels on the lysts at room temperature without CO formation which is toxic
environment, the need for benign and sustainable energy to polymer electrolyte membrane fuel cells (PEMFCs).
sources has drastically increased. Hydrogen has long been
recognized as an alternative because it has a high gravimetric efficient catalysts that enable the generation of H
For FA to be employed as a viable hydrogen energy carrier,
from FA with
energy density of 120 MJ kg , which is three times greater than high activity and selectivity are needed. In this context, several
2
ꢀ
1
ꢀ
1
7–10
11–21
that of gasoline (44 MJ kg ) with essentially no carbon dioxide homogeneous
and heterogeneous
catalysts have been
1
emission. Additionally, hydrogen can serve as a promising designed for hydrogen production from FA over the last few
medium for large scale energy storage, as it solves the ‘inter- years. In particular, Pd-based catalysts are the most active
2
–4
mittency problem’ of renewable energy sources. However, its heterogeneous catalysts for FA dehydrogenation. For example,
1
1–14,20,21
15
direct use is limited due to the lack of safe and reversible Pd-based alloys
hydrogen storage technologies.
modications exhibit enhanced catalytic activity, and among
To address the issue, the development of potentially revers- them, PdAu alloys showed excellent activity and selectivity for
or core–shells produced by structural
11,13,14,16,22
ible hydrogen storage systems with high hydrogen storage the desired reaction.
capacities is critical. Formic acid (FA) has garnered signicant
Despite the recent enhancement in activity and selectivity for
attention as a chemical hydrogen storage material because FA dehydrogenation on the PdAu nanocatalyst, the role of Au in
it is a stable, nontoxic liquid which has a high volumetric the hydrogen production rate and selectivity has not been
ꢀ
1
11,14,23
hydrogen storage density of 53 g L , much higher than profoundly understood; only a few studies
were conducted
), and can be to elucidate the inuence of PdAu alloys on FA dehydrogenation
2
ꢀ
1
compressed hydrogen gas (40 g L for 700 bar H
5,6
readily produced from biomass. Moreover, FA can easily be with mechanistic consideration. The ensemble effect, the
geometric effect induced by distinct surface atomic arrange-
ments, and the ligand effect, the electronic effect induced by
a
bonding interaction between metals which results in the
Fuel Cell Research Center, Korea Institute of Science and Technology (KIST), Seoul
1
36-791, Republic of Korea. E-mail: cwyoon@kist.re.kr; hchahm@kist.re.kr
modication of the electronic structure, are important factors
b
Clean Energy and Chemical Engineering, Korea University of Science and Technology, that affect the performance of bimetallic catalysts. The effect of
Daejeon 305-333, Republic of Korea
the PdAu alloy on FA dehydrogenation, however, has been
c
Department of Converging Science and Technology, KHU-KIST Graduate School,
Kyunghee University, Seoul, Korea
23
explained only by the ensemble effect. For example, Yuan et al.
reported the effects of a PdAu supported Au(111) monolayer
†
1
‡
Electronic supplementary information (ESI) available. See DOI:
0.1039/c6ta03654f
structure on FA dehydrogenation by density functional theory
11
(
DFT) calculations. Furthermore, Mullins et al. veried that
These authors contributed equally.
This journal is © The Royal Society of Chemistry 2016
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
distinct PdAu surface alloy ensembles on the Au(111) subsur- using BELCAT-M (Bel Japan Inc.). Five samples were evaluated:
face could affect both the reactivity and selectivity for FA Al O (for the blank), PdAu/Al O –CO (1), PdAu/Al O –H (2),
2
3
2
3
2
3
2
dehydrogenation. Despite these model catalysts demonstrating PdAu/Al O –N (3), and Pd/Al O (4). A catalyst was placed in
2
3
2
2 3
ꢁ
the importance of PdAu ensembles in a highly systematic a measurement tube and purged with He for 1 h at 25 C to
manner, the inuence of PdAu sublayers in the PdAu alloy on stabilize the background signal. CO was adsorbed on the cata-
catalytic activity (ligand effect) has not been well studied. Since lysts by supplying 7% CO diluted with He into the sample tube
the ligand effect could potentially play a pivotal role in H
2
-
for 30 min. The temperature of the tube was then increased to
ꢁ
ꢁ
ꢀ1
release from FA as well, it is necessary to scrutinize the effect. 400 C with a heating rate of 5 C min under the ow of He. To
Elucidation of both ensemble and ligand effects on catalytic obtain the amounts of adsorbed CO purely on the metallic
activity under practical catalytic conditions is also essential to surface, all the data were calibrated by subtracting relevant
minimize the gap between actual catalytic reactions and model CO-TPD data obtained from Al O as the blank sample.
2
3
24
reactions conducted in an ultra-high vacuum system.
Surface ensembles of PdAu catalysts were further monitored
We report here on experimental and theoretical evidence by CO vibrational frequencies using diffuse reectance infrared
that accounts for the enhanced activity for FA dehydrogenation Fourier transform (DRIFT) spectroscopy (Nicolet iS10, Thermo
over PdAu with a distinct surface structure under practical Scientic) equipped with a mercury cadmium telluride (MCT)
reaction conditions, demonstrating that both ensemble and detector. In a typical experiment, a catalyst was loaded in the
ligand effects played synergistic roles in facilitating FA DRIFT cell in powder form and purged with N
dehydrogenation.
2
to remove
moisture and oxygen inside the IR chamber. A background
spectrum was initially recorded under a N atmosphere. Then,
2
1
% CO gas diluted with N was supplied for 30 min at room
2
Experimental
temperature to adsorb CO onto the catalyst surface. The desired
Synthesis of catalysts
CO absorption spectra were obtained aer the removal of
2
gaseous CO residues by purging the cell with N gas for 20 min.
The PdAu/Al
conventional incipient wetness impregnation.
nating solution containing Pd and Au was prepared by dissolv-
2
O
3
catalyst (5 wt%, Pd : Au ¼ 1 : 1) was prepared by
25,26
An impreg-
Catalytic activity test for FA dehydrogenation
ꢀ1
ing HAuCl $xH O (17 mg mL
, STREM Chemical) and PdCl2
4
2
ꢀ
H
2
O
1
An aqueous FA solution (ca. 1.0 M) containing distilled water
(9.7 mL) and FA (370 mL, 10 mmol) was added to a reaction tube
with a catalyst (50 mg, 1, 2, 3, or 4) and stirred at 300 rpm with
a magnetic stirrer. The quantities of gaseous products during
(
7.7 mg mLH O , Sigma Aldrich) in distilled water with gentle
2
2
ꢀ1
heating. g-Al
2
O
3
powder (Alfa Aesar, BET ¼ 220 m g ) was then
ꢀ1
added to the brownish aqueous solution (0.4 mLsolution
g
g-Al O3 ).
2
ꢁ
The obtained solids were ground and dried at 100 C to remove
ꢁ
dehydrogenation at 50 C were measured using a gas burette
residual H
2
O. The resulting powders were nally calcined at 400
ꢁ
connected to a real time recording system. The amount of H
2
produced aer 5 min was calculated as follows.
C for 4 h to ensure the formation of the PdAu alloy and then
reduced under 10% H gas at 250 C for 4 h to give PdAu/Al O –
ꢁ
2
2 3
H (2). Modication of surface PdAu arrangements was achieved
PV
2
Moles of H
2
produced in 5 min ðnÞ ¼
ꢁ
by a heat treatment of 2 at 250 C for 4 h either under N
2
to give
RT
ꢀ
ꢁ
1
ꢂVgas in 5 min
1
2
PdAu/Al O –N
2 3 2
2 3
(3) or 7% CO to afford PdAu/Al O –CO (1).
¼
ꢂ
0
:08206 ꢂ temperature
Characterization of catalysts
Theoretical details
The particle size distribution of the PdAu catalysts was analyzed by The calculations reported herein were performed on the basis of
scanning transmission electron microscopy (STEM) using a FEI spin polarized density functional theory (DFT) within the
27
Tecnai F20 at 200 kV. The surface distributions of Pd and Au in the generalized gradient approximation (GGA-PW91), as imple-
28
PdAu nanoparticles were measured by energy dispersive spec- mented in the Vienna Ab-initio Simulation Package (VASP). The
29
troscopy (EDS) at 200 kV. Due to the dri of the small sized- projector augmented wave (PAW) method with a planewave
nanoparticles under a high energy electron beam, we rst basis set was employed to describe the interaction between core
prepared larger nanoparticles with a size of 10 nm by the calci- and valence electrons. An energy cutoff of 350 eV was applied for
ꢁ
nation of the original PdAu materials at 800 C. The enlarged the planewave expansion of the electronic eigen functions. For
ꢁ
30
nanoparticles were then heat-treated using H
2
, CO or N
2
at 250 C. the Brillouin zone integration, we used a 5 ꢂ 5 ꢂ 1 Monkhorst–
To monitor the formation of PdAu alloys, X-ray photoelectron Pack mesh of k points to calculate geometries and total energies.
spectroscopy (XPS) was recorded on a PHI 50000 Versa Probe from Reaction pathways and barriers were determined using the
ꢀ8
31
Al Ka-ray (1486.6 eV) at 6.7 ꢂ 10 Pa. The enlarged PdAu nano- climbing image nudged elastic band method (CI-NEBM) for
particles were further characterized using high-angle annular dark each elementary step. For the Pd Au (111) model surfaces, as
eld (HAADF) microscopy (Talos F200X) coupled with scanning shown in Fig. S1,† we constructed a ve atomic-layer slab, each of
2
2

transmission electron microscopy (HAADF-STEM) and energy which contains four atoms. The slab is separated from its peri-
dispersive spectroscopy (EDS) with elemental mapping at 200 kV. odic images in the vertical direction by a vacuum space corre-
Relative quantities of Pd on the PdAu surface were deter- sponding to seven atomic layers. While the bottom two layers of
mined by CO-temperature programmed desorption (CO-TPD) the ve-layered slab are xed at corresponding bulk positions,
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
Journal of Materials Chemistry A
the upper three layers are fully relaxed using the conjugate energy barrier for the D1 step on the surface of II was 0.27 eV
gradient method until residual forces on all the constituent and 0.80 eV lower than those of III and IV, respectively, because
ꢀ2
ꢀ1
atoms become smaller than 5 ꢂ 10 eV A . The lattice constant three contiguous Pd atoms favorably stabilize the transition
˚
for bulk Pd is predicted to be 3.95 A which is virtually identical to state. In D2, the involvement of three metal atoms resulted in
the previous calculations and also in good agreement with the the lowest transition state energy on II. Although the differences
3
2
˚
experimental value of 3.89 A.
in the energy barriers were not as signicant as those in the rst
transition state due to the high energies of the starting inter-
mediates on the surfaces of III and IV, II showed the lowest
barrier among all catalysts (Fig. 1 and Table S1†).
Results and discussion
To investigate the effects of bimetallic PdAu structure on FA
Numerous reports suggested that for bimetallic surfaces,
dehydrogenation activity, spin-polarized density functional their surface structures change upon gas adsorption due to
6,33–37
theory (DFT) calculations were performed. First, we scrutinized metal segregation on the surface.
Since CO preferentially
how the FA dehydrogenation kinetics were affected by surface binds to Pd over Au, the Pd segregation energies in the alloy
arrangements (ensembles) of Pd atoms on four different were evaluated as a function of the Pd coverage. We assessed
PdAu(111) model alloys (I–IV) which had varying surfaces and the energetically preferred composition of Pd in the rst
subsurface PdAu arrangements (Fig. 1a). The rst surface layers surface layer by calculating the total energy variation (DE) as
of model PdAu catalysts consisted of pure Pd (I), triangular a function of surface Pd concentration. Under a CO environ-
shaped Pd (II), linear shaped Pd (III), and mono Pd (IV), ment, the surface composition of 75 atom% Pd [Pd Au /Pd Au /
3
1
1
3
respectively. Fig. 1b shows changes in the total energy and Pd Au (111), II] was the most stable structure, followed by 50
2
2
activation barriers for FA dehydrogenation based on two atom% Pd [Pd Au /Pd Au /Pd Au (111), III] and 25 atom% Pd
2
2
2
2
2
2
sequential reactions: (i) HCOOH* / HCOO* + H* (D1) and [Pd Au /Pd Au /Pd Au (111), IV] (Table 1). This indicated the
1
3
3
1
2
2
*
2
(
ii) HCOO*/CO þ H* (D2). Among all surface models, II higher availability of triangular shaped Pd trimers (three-fold
showed the most favorable reaction energies (DE ¼ ꢀ0.18 eV for hollow sites associated with three Pd atoms) on the PdAu
D1; ꢀ0.70 eV for D2) and lowest kinetic barriers (E ¼ 0.60 eV for surface than linear shaped Pd atoms (two-fold bridge sites
a
D1; 0.87 eV for D2) (Table S1†), suggesting that the triangular associated with two Pd atoms) and Pd monomers (one-fold top
shaped Pd atoms on the PdAu alloy surface played an important sites of a single Pd atom which is completely surrounded by Au
role in the catalysis, which originates from the strong attractive atoms) under a CO atmosphere.
force between Pd and the reaction intermediates in the transi-
We prepared several PdAu catalysts possessing different
tion states. For instance, FA requires three adjacent metal surface atomic concentrations according to the theoretical
25–29
atoms for the coordination of two oxygens and one bridging calculations and the previous reports.
The PdAu alloy
hydrogen in the rst transition state (TS1, Fig. S1†). The kinetic nanoparticles were impregnated on g-alumina according to the
2
5,26
reported procedure
and the as-prepared catalysts were ther-
, and
ꢁ
mally treated at 250 C with three distinct gases of CO, H
2
N to alter the surface atomic arrangements of the PdAu cata-
2
lysts. The resulting catalysts were denoted as PdAu/Al O –CO (1),
2
3
PdAu/Al O –H (2), and PdAu/Al O –N (3), respectively. We also
2
3
2
2
3
2
prepared pure Pd/Al
PdAu alloys was veried by the shi of the Pd 3d5/2 peak from
35.2 eV (pure Pd) to 335.4 eV (PdAu alloys) in the XPS spectra
2 3
O (4) for comparison. The formation of
3
22
(Fig. S2 and Table S2†). STEM revealed the formation of PdAu
nanoparticles with the average sizes of 1.2–1.3 nm (Fig. S3†). The
particle sizes were found to be nearly identical aer heat treat-
ment. To compare the surface concentrations of Pd and Au in
the prepared materials, we measured EDS line proles of 1–3.
The EDS line proles with these enlarged nanoparticles
conrmed that the “surface Pd” concentration of 1 (CO treated)
was found to be higher than 2 (H
Fig. S3†). The results again supported that the CO treatment can
alter the surface PdAu atomic ratio.
2 2
treated) and 3 (N treated)
(
The surface structure of Pd was further characterized by
temperature programmed desorption (TPD) and diffuse reec-
tance infrared Fourier transform (DRIFT) spectroscopy using
CO gas. As depicted in Fig. 2, the Pd coverage on the surface was
signicantly altered upon calcination. The sample with pure Pd
showed the highest CO adsorption capacity as it has only Pd on
the surface. The CO desorption temperatures and adsorbed CO
quantities on the PdAu catalyst surfaces increased in the order
Fig. 1 (a) The PdAu model catalysts used for DFT calculations and (b)
the calculated potential energy diagram for the FA dehydrogenation
on I–IV. Yellow and gray balls represent the Au and Pd atoms,
respectively.
This journal is © The Royal Society of Chemistry 2016
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Table 1 Variation in the total energy (DE for CO adsorption) as a function of the surface Pd composition. Yellow and gray balls represent the Au
and Pd atoms, respectively
Model (1st layer/2nd layer/substrate)
Surface Pd composition (%)
Top view of the Au
2
Pd
2
slab
DE (eV)
0.28
Pd
4
3
/Au
4
/Pd
2
Au
2
(I)
100
75
Pd
Au
1
2
/Pd
1
2
Au
3
2
/Pd
2
2
Au
2
2
(II)
ꢀ0.05
Pd
2
1
Au
/Pd
Au
/Pd
Au
(III)
50
25
0
Pd
Au
3
/Pd
3
Au
1
/Pd
2
Au
2
(IV)
0.17
40
of 1 > 2 $ 3, indicating that 1 possessed a Pd-enriched surface (Table S3†). The CO intensities attributed to Pd atop sites
with numerous CO binding sites, i.e., the increased CO decreased as the amount of surface Pd increased, which indi-
adsorption area of 1 originated from the increased Pd surface cated that the atomic surface arrangements were altered by the
coverage because only negligible amounts of CO can adsorb on heat treatment under different atmospheric conditions. Despite
38
Au atoms at room temperature. In addition, the increased CO 4 having the highest number of hollow sites as it consists of
desorption temperature of 1 indicated the presence of specic a pure Pd surface, 4 exhibited a lower relative ratio for hollow
Pd sites that interacted strongly with CO, which is likely sup- sites than 1. It implies that CO adsorption on the bridge sites is
ported by the DFT results (vide supra). In DRIFT spectra, CO favorable over the hollow sites on the pure Pd surface at high
41,42
species adsorbed on Pd surfaces are known to exhibit CO coverage.
three characteristic vibrational frequencies depending on the
The modied bimetallic PdAu catalysts likely provided
38–41
ꢀ1
binding modes at distinct sites:
atop site), (ii) 1900–2050 cm
(i) ca. 2080 cm (linear, different dehydrogenation activities owing to their distinct
(2-fold, bridge site), and surface ensembles. To verify the inuence of atomic ensembles
ꢀ
1
ꢀ
1
(
iii) 1800–1900 cm (3-fold, hollow site). Given the Pd surface at the PdAu surfaces, we performed dehydrogenation reactions
ꢁ
coverage of the as-prepared catalysts, the number of Pd hollow using a 1.0 M aqueous FA solution at 50 C. The catalytic activity
sites at the surface likely increased in the order of 1 > 2 > 3. increased as the surface Pd/Au ratio increased (1 > 2 > 3; Fig. 4a),
Indeed, DRIFT studies supported the TPD results; the intensity which was closely correlated with the relative quantity of 3-fold
of the CO vibration of 1 increased at the Pd hollow sites (1800– hollow sites. The amount of H
2
produced for 5 min catalyzed by
ꢀ
1
1900 cm ), while 3 exhibited a negligible intensity in the same 1 (0.13 mmol) was found to be much greater than those of 2
region (Fig. 3). The population of hollow sites with respect to
total Pd sites for 1 and 2, calculated using tted peak areas,
was 4.3 and 2.8 times greater than that of 3, respectively
Fig. 3 The DRIFT spectra of the vibrations of CO adsorbed on different
Fig. 2 The CO-TPD profiles of the as-prepared catalysts.
surfaces.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 5 Projected local density of states plots on the d states of surface
Pd atoms for the pure Pd(111) and Pd
3 1 1 3 2 2
Au /Pd Au /Pd Au (111). The
displayed vertical line at 0 eV stands for the Fermi level.
decomposition. For the Pd–Au system in particular, the
ensemble effect is a primary effect where the catalytically more
active component (Pd) is diluted by the less active component
43
(
Au). In earlier studies on late-transition metals such as Pd
and Au, the importance of the d state of the surface was
emphasized over the s and p states in dening chemisorption
44
and catalytic properties. In the presence of a Pd-dominated
Fig. 4 (a) H
2
production from catalytic FA dehydrogenation over surface supported by an Au dominated subsurface, such as in I
different catalysts. (b) Correlation between the amounts of H
produced in 5 min and the relative quantities of Pd hollow sites existed
at the catalysts.
2
and II, there is an increase in the d electron density of surface
Pd atoms owing to charge transfer from the subsurface Au
24,45
atoms.
Together with ensemble effects, the charge transfer
in II leads to a reduced binding energy of HCOO compared to
that in Pd(111), as there is no ensemble effect on the surface.
This decreased activity toward formate chemisorption is related
to the signicant modication of the d states in surface Pd
atoms owing to electronic interactions with subsurface Au
atoms. Fig. 5 shows the local density of states (LDOS) projected
on the d states of surface Pd atoms for II and pure Pd(111). A
noticeable reduction in the occupied LDOS near the Fermi level
(
0.081 mmol) and 3 (0.066 mmol). These results indicated that
a high Au surface concentration was detrimental because it
inhibited the formation of Pd hollow sites, as supported by the
results obtained by DFT methods. The importance of hollow
sites was highlighted again by the close correlation between
amounts of H production and ratios of hollow Pd sites on PdAu
2
catalysts (Fig. 4b).
Although ensemble effects can reasonably explain the cata-
lytic behavior of the as-prepared PdAu catalysts, ligand effects
could also play an important role in accelerating the FA dehy-
drogenation. The presence of the ligand effect can be supported
by the result of the lower H -release activity of 4 compared to 1,
despite the higher quantity of the 3-fold hollow Pd sites in 4
(
ꢀ0.5 < E ꢀ E
f
< 0) was found for II as compared to that of pure
Pd(111). This explained the lower activity of Pd atoms toward
HCOO chemisorption in the alloy as compared to pure Pd(111).
It is worth noting that the higher the LDOS near the Fermi level,
46
the higher the activity of surface atoms toward chemisorption.
2
The calculated d-band centers for II and Pd(111) were ꢀ1.97 eV
and ꢀ1.71 eV, respectively.
(Fig. 4a) i.e., the enhanced dehydrogenation activity for 1 orig-
inated additionally from the electronic perturbation induced by
Au-containing sublayers. To verify this hypothesis, we further
optimized HCOO adsorption on II as well as on pure Pd(111)
Conclusions
(
(
as a model surface for 4) in a bridging bidentate mode
Fig. S4†). A substantial reduction in the binding energy of
In summary, the possible roles of PdAu alloys in the rate of FA
dehydrogenation were demonstrated as independent functions
of (i) surface atomic ensembles (side by side interactions) and
HCOO on surface Pd atoms in II (Ebind ¼ ꢀ2.34 eV) was pre-
dicted as compared to pure Pd atoms (Ebind ¼ ꢀ2.42 eV), sug-
gesting a reduced adsorption energy of surface Pd atoms in II
owing to Pd–Au interactions, which likely contributed to the FA
(ii) surface charge polarization (interactions between top and
second layers). Several surface compositions were prepared by
varying the calcination conditions and the resulting catalytic
This journal is © The Royal Society of Chemistry 2016
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
activities were evaluated. Triangular shaped Pd sites were 16 J. H. Lee, J. Ryu, J. Y. Kim, S.-W. Nam, J. H. Han, T.-H. Lim,
proposed to play pivotal roles in facilitating FA dehydrogena-
tion by reducing the reaction energy and kinetic barrier. In
S. Gautam, K. H. Chae and C. W. Yoon, J. Mater. Chem. A,
2014, 2, 9490–9495.
addition to Pd ensemble effects, the dehydrogenation activity 17 K. Koh, J.-E. Seo, J. H. Lee, A. Goswami, C. W. Yoon and
was also inuenced by the Pd electronic structure. Our study T. Asefa, J. Mater. Chem. A, 2014, 2, 20444–20449.
revealed the importance of both ensemble and ligand effects for 18 J. Cho, S. Lee, J. Han, S. P. Yoon, S. W. Nam, S. H. Choi,
dehydrogenation catalysis, which will provide valuable insight
into the development of highly capable catalysts not only for FA
K.-Y. Lee and H. C. Ham, J. Phys. Chem. C, 2014, 118,
22553–22560.
dehydrogenation, but also for other Pd based alloy materials for 19 A. Bulut, M. Yurderi, Y. Karatas, M. Zahmakiran, H. Kivrak,
a number of chemical transformations.
M. Gulcan and M. Kaya, Appl. Catal., B, 2015, 164, 324–333.
0 Y. Karatas, A. Bulut, M. Yurderi, I. E. Ertas, O. Alal,
M. Gulcan, M. Celebi, H. Kivrak, M. Kaya and
M. Zahmakiran, Appl. Catal., B, 2016, 180, 586–595.
2
Acknowledgements
This research was supported by the Technology Development 21 M. Yurderi, A. Bulut, N. Caner, M. Celebi, M. Kaya and
Program to Solve Climate Changes of the National Research M. Zahmakiran, Chem. Commun., 2015, 51, 11417–11420.
Foundation (NRF) funded by the Ministry of Science, ICT & 22 Z.-L. Wang, J.-M. Yan, H.-L. Wang, Y. Ping and Q. Jiang,
Future Planning (2015M1A2A2074688). The authors also J. Mater. Chem. A, 2013, 1, 12721–12725.
acknowledge the nancial support from the institutional 23 D. W. Yuan and Z. R. Liu, J. Power Sources, 2013, 224, 241–
program of Korea Institute of Science and Technology
249.
2E26590) as well as from the Korea Electric Power Corporation 24 F. Gao and D. W. Goodman, Chem. Soc. Rev., 2012, 41, 8009–
(
and Korea Southern Power Co., Ltd.
8020.
2
5 L. Kesavan, R. Tiruvalam, M. H. A. Rahim, M. I. bin Saiman,
D. I. Enache, R. L. Jenkins, N. Dimitratos, J. A. Lopez-
Sanchez, S. H. Taylor, D. W. Knight, C. J. Kiely and
G. J. Hutchings, Science, 2011, 331, 195–199.
Notes and references
1
2
3
4
5
S. Satyapal, J. Petrovic, C. Read, G. Thomas and G. Ordaz,
Catal. Today, 2007, 120, 246–256.
A. Gonz ´a lez, E. McKeogh and B. O. Gallachoir, Renewable
Energy, 2004, 29, 471–489.
26 D. A. Bulushev, S. Beloshapkin, P. E. Plyusnin, Y. V. Shubin,
V. I. Bukhtiyarov, S. V. Korenev and J. R. H. Ross, J. Catal.,
2013, 299, 171–180.
´
´
H. Chen, T. N. Cong, W. Yang, C. Tan, Y. Li and Y. Ding, Prog. 27 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
Nat. Sci., 2009, 19, 291–312. 1996, 77, 3865–3868.
D. Anderson and M. Leach, Energy Policy, 2004, 32, 1603– 28 J. F. G. Kresse, VASP the Guide, Vienna University of
614. Technology, Vienna, Austria, 2001.
R. K. Ahluwalia and J. K. Peng, Int. J. Hydrogen Energy, 2008, 29 P. E. Bl ¨o chl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994,
1
33, 4622–4633.
50, 17953–17979.
6
7
F. Jo ´o , ChemSusChem, 2008, 1, 805–808.
J. F. Hull, Y. Himeda, W.-H. Wang, B. Hashiguchi, R. Periana,
30 P. E. Bl ¨o chl, O. Jepsen and O. K. Andersen, Phys. Rev. B:
Condens. Matter Mater. Phys., 1994, 49, 16223–16233.
D. J. Szalda, J. T. Muckerman and E. Fujita, Nat. Chem., 2012, 31 G. Henkelman, B. P. Uberuaga and H. J ´o nsson, J. Chem.
, 383–388. Phys., 2000, 113, 9901–9904.
A. Boddien, D. Mellmann, F. G ¨a rtner, R. Jackstell, H. Junge, 32 H. C. Ham, G. S. Hwang, J. Han, S. W. Nam and T. H. Lim,
4
8
9
P. J. Dyson, G. Laurenczy, R. Ludwig and M. Beller, Science,
011, 333, 1733–1736.
J. Phys. Chem. C, 2009, 113, 12943–12945.
33 S.-Y. Lee, N. Jung, J. Cho, H.-Y. Park, J. Ryu, I. Jang, H.-J. Kim,
E. Cho, Y.-H. Park, H. C. Ham, J. H. Jang and S. J. Yoo, ACS
Catal., 2014, 4, 2402–2408.
2
B. Loges, A. Boddien, H. Junge and M. Beller, Angew. Chem.,
Int. Ed., 2008, 47, 3962–3965.
1
1
1
1
1
0 C. Fellay, P. J. Dyson and G. Laurenczy, Angew. Chem., Int. 34 F. Gao, Y. Wang and D. W. Goodman, J. Phys. Chem. C, 2010,
Ed., 2008, 47, 3966–3968. 114, 4036–4043.
1 W.-Y. Yu, G. M. Mullen, D. W. Flaherty and C. B. Mullins, 35 S. M. Oxford, P. L. Lee, P. J. Chupas, K. W. Chapman,
J. Am. Chem. Soc., 2014, 136, 11070–11078.
M. C. Kung and H. H. Kung, J. Phys. Chem. C, 2010, 114,
¨
2 S. Zhang, O. Metin, D. Su and S. Sun, Angew. Chem., Int. Ed.,
013, 52, 3681–3684.
17085–17091.
2
36 F. Abild-Pedersen and M. P. Andersson, Surf. Sci., 2007, 601,
3 Z.-L. Wang, J.-M. Yan, Y. Ping, H.-L. Wang, W.-T. Zheng and
Q. Jiang, Angew. Chem., Int. Ed., 2013, 52, 4406–4409.
4 D. A. Bulushev, S. Beloshapkin, P. E. Plyusnin, Y. V. Shubin,
1747–1753.
37 R. Anton, H. Eggers and J. Veletas, Thin Solid Films, 1993,
226, 39–47.
V. I. Bukhtiyarov, S. V. Korenev and J. R. H. Ross, J. Catal., 38 L. Ouyang, G.-j. Da, P.-f. Tian, T.-y. Chen, G.-d. Liang, J. Xu
013, 299, 171–180. and Y.-F. Han, J. Catal., 2014, 311, 129–136.
5 K. Tedsree, T. Li, S. Jones, C. W. A. Chan, K. M. K. Yu, 39 S. K. Kim, C. Kim, J. H. Lee, J. Kim, H. Lee and S. H. Moon,
2
1
P. A. J. Bagot, E. A. Marquis, G. D. W. Smith and
J. Catal., 2013, 306, 146–154.
S. C. E. Tsang, Nat. Nanotechnol., 2011, 6, 302–307.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
Journal of Materials Chemistry A
40 S. D. Ebbesen, B. L. Mojet and L. Lefferts, Phys. Chem. Chem. 43 J. Zhang, H. Jin, M. B. Sullivan, F. C. H. Lim and P. Wu, Phys.
Phys., 2009, 11, 641–649. Chem. Chem. Phys., 2009, 11, 1441–1446.
4
1 J. B. Giorgi, T. Schroeder, M. B ¨a umer and H.-J. Freund, Surf. 44 J. K. Norskov, T. Bligaard, J. Rossmeisl and
Sci., 2002, 498, L71–L77. C. H. Christensen, Nat. Chem., 2009, 1, 37–46.
4
2 I. V. Yudanov, R. Sahnoun, K. M. Neyman, N. R ¨o sch, 45 Y.-S. Lee, Y. Jeon and S.-J. Oh, J. Korean Phys. Soc., 2000, 47,
J. Hoffmann, S. Schauermann, V. Joh ´a nek, H. Unterhalt,
451–455.
G. Rupprechter, J. Libuda and H.-J. Freund, J. Phys. Chem. 46 Y. Y. Tong, T. Yonezawa, N. Toshima and J. J. vanderKlink,
B, 2003, 107, 255–264. J. Phys. Chem., 1996, 100, 730–733.
This journal is © The Royal Society of Chemistry 2016
J. Mater. Chem. A