Cu/Al2O3catalysts modified with Pd for selective acetylene hydrogenation

Article abstract of DOI:10.1016/j.jcat.2014.08.016

A Cu/alumina catalyst has been modified by addition of various quantities of Pd. Characterisation suggests formation of a bimetallic with the surface dominated by Cu. Optimisation of the Cu:Pd ratio (50:1) resulted in a catalyst which combined the properties of Cu and Pd (high ethylene selectivity and activity at low temperature). Using a 3:1 H₂:acetylene feed, >99% acetylene conversion and >70% ethylene selectivity was attained at 373 K, representing a considerable reduction in the temperature necessary to achieve equivalent activity/selectivity over monometallic Cu. Increasing the H2:acetylene ratio to 10 resulted in >99% acetylene conversion at 363 K as well as enhanced ethylene selectivity (>80%). Enhanced activity of Cu at low temperature is attributed to H₂dissociation on Pd with spillover onto Cu sites where the reaction takes place. These bimetallic CuPd catalysts offer sufficient activity at low temperature and could represent a replacement to the current industrial PdAg catalysts.

Full text of DOI:10.1016/j.jcat.2014.08.016

Journal of Catalysis 319 (2014) 127–135
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Cu/Al O catalysts modified with Pd for selective acetylene
2
3
hydrogenation
Alan J. McCue a, , Callum J. McRitchie , Ashley M. Shepherd , James A. Anderson
⇑
a
b
a,c,
⇑
a
Surface Chemistry and Catalysis Group, Department of Chemistry, University of Aberdeen, Aberdeen AB24 3UE, UK
Chemistry Research Laboratory, University of Oxford, Oxford OX1 3TA, UK
Materials and Chemical Engineering Group, School of Engineering, University of Aberdeen, Aberdeen AB24 3UE, UK
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
A Cu/alumina catalyst has been modified by addition of various quantities of Pd. Characterisation
suggests formation of a bimetallic with the surface dominated by Cu. Optimisation of the Cu:Pd ratio
Received 21 May 2014
Revised 28 July 2014
Accepted 22 August 2014
(
50:1) resulted in a catalyst which combined the properties of Cu and Pd (high ethylene selectivity
and activity at low temperature). Using a 3:1 H :acetylene feed, >99% acetylene conversion and >70%
ethylene selectivity was attained at 373 K, representing a considerable reduction in the temperature
necessary to achieve equivalent activity/selectivity over monometallic Cu. Increasing the H :acetylene
ratio to 10 resulted in >99% acetylene conversion at 363 K as well as enhanced ethylene selectivity
>80%). Enhanced activity of Cu at low temperature is attributed to H dissociation on Pd with spillover
2
2
Keywords:
Copper
Palladium
Catalysts
Acetylene
Selective hydrogenation
Hydrogen spillover
(
2
onto Cu sites where the reaction takes place. These bimetallic CuPd catalysts offer sufficient activity at
low temperature and could represent a replacement to the current industrial PdAg catalysts.
Ó 2014 Elsevier Inc. All rights reserved.
1
. Introduction
phosphorous containing molecules such as diphenyl sulfide
[
18–20] and triphenylphosphine [21] with selectivity varying
Ethylene produced from naphtha crackers can be used for poly-
depending on the ligand structure suggesting the effect has yet
to be fully optimised [22].
ethylene production so long as it undergoes a rigorous purification
step to remove acetylene impurities. The preferred industrial
method involves selective hydrogenation using a Pd catalyst
modified with Ag or Au. Achieving high selectivity (i.e. avoiding
over-hydrogenation to ethane) is non-trivial and often relies upon
the addition of CO as a transient poison since it competes with eth-
ylene but not acetylene for adsorption sites [1,2]. However, such a
methodology is far from ideal since the quantity of CO added must
be monitored and adjusted to compensate during catalyst deacti-
vation. Traditionally, a selective Pd catalyst was associated with
small Pd ensembles which favour acetylene adsorption but limit
ethylene adsorption [2]. More recent studies indicate that the
interplay between Pd carbides [3–5] and hydrides [6,7] play impor-
tant roles in governing selective hydrogenation [8–15]. New
approaches to improving selectivity in Pd based catalysts include
alloying with Ga [16,17] or the adsorption of molecules which
act as a selectivity modifier. Examples include sulphur and
Whilst industrial catalysts almost exclusively use Pd as the
active metal, a number of other metals such as Cu [23–27], Ni
[28–30], Au [31–35], and Ag [36,37] show enhanced alkene selec-
tivity, notably in the absence of CO. Results of density functional
theory (DFT) calculations suggest that enhanced selectivity over
Au, Ag and Cu is related to the unfavourable adsorption energy of
the alkene species [34,38]. Catalysis over a Ni surface may be more
complex and involve hydride and/or carbide type species akin to
Pd [38]. Despite enhanced selectivity, the use of Cu, Au and Ag is
limited since hydrogen dissociation is an activated process and
requires the reaction to take place at temperatures approaching
473 K which far exceeds the temperatures required with Pd
2
(298–373 K). For example, the activation energy for H dissociation
on a Cu (111) surface has been calculated to be 0.46 eV [39] which
is far higher than on Pd (0.20 eV) [40]. When the higher reduction
temperature of the Pd free catalysts is also taken into account, it
means that these catalysts do not represent ‘drop in’ replacements
with existing industrial reactors and some authors acknowledge
these limitations [41]. An additional complication is that metals
such as Ni and Cu produce greater quantities of acetylene based
oligomers (most notably at low temperature) which can again be
attributed to a limited supply of hydrogen adatoms and may lead
⇑
Corresponding authors. Address: Surface Chemistry and Catalysis Group,
Department of Chemistry, University of Aberdeen, Aberdeen AB24 3UE, UK. Fax:
44 1224 272901.
E-mail addresses: a.mccue@abdn.ac.uk (A.J. McCue), j.anderson@abdn.ac.uk
J.A. Anderson).
+
(
http://dx.doi.org/10.1016/j.jcat.2014.08.016
021-9517/Ó 2014 Elsevier Inc. All rights reserved.
0

1
28
A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
to more rapid catalyst deactivation. Interestingly, Bridier et al. [27]
demonstrated that the temporary addition of CO to a Cu-hydrotal-
cite catalyst resulted in a decrease in the formation of oligomers in
favour of enhanced alkene selectivity. The observed effect was
attributed to a decrease in the size of the Cu ensembles through
a combination of surface restructuring and an ensemble dilution
effect due to carbon deposition.
To circumvent these problems, a possible solution could involve
combining the catalytic properties of Pd with one of the aforemen-
tioned selective metals. This would in principle create a catalyst
with well dispersed Pd atoms capable of hydrogen dissociation in
intimate contact with a second metal upon which the reaction
could take place via a hydrogen spillover mechanism. A natural
choice for the second metal is Cu since it readily forms alloyed par-
ticles with Pd [42,43] and some recent reports suggest that the
preparation of such a system may be viable for well-defined Cu
impregnation step followed by drying at 393 K and calcination at
673 K for 1 h. A 1.67% Pd/Al catalyst was prepared from
Pd(NO using the same method and a 3 h calcination at 673 K
for comparison purpose. Bimetallic catalysts are denoted as
‘X-CuPd’ where ‘X’ refers to the atomic ratio of Cu:Pd.
2 3
O
3 2
)
Temperature programmed reduction (TPR) experiments were
carried out using a TPDRO 1100 instrument with a TCD detector.
Profiles were collected in a temperature range of 313–873 K using
À1
a heating rate of 5 K min . TPR results are presented per gram of
sample and hydrogen consumption quantified based on a response
factor determined using a CuO standard.
Powder X-ray diffraction (XRD) patterns of the calcined samples
were collected using an X’Pert Powder diffractometer supplied by
PANalytical. Data were collected in the range 20–60° 2h with a step
size of 0.03° with a sample acquisition time of ca. 10 min.
FTIR of adsorbed CO was performed using a PE Spectrum 100
FTIR spectrometer with sample presented as a 16 mm diameter
self-supporting disc. The disc was suspended in a quartz holder
and held in a conventional vacuum line which permitted in situ
(
111) single crystals [44,45]. The original authors identified iso-
lated Pd atoms as the active site for dissociative H adsorption;
2
however, a recent DFT study suggested that incorporation of Pd
atoms into both the surface and subsurface layers reduces the bar-
rier for dissociation [39]. In this study, we report a simple prepara-
evacuation and gas manipulation. Sample was first heated and
À1
reduced in 100 ml min of H
2
for 1 h (323 K for 1.67% Pd/Al
2
O
3
tion of powdered Cu/Al
2
O
3
catalysts promoted with a range of Pd
and 523 K for all other samples). The sample cell was cooled and
À5
loadings and identify the Cu:Pd ratio which offers optimum activ-
ity at low temperature (363 K) whilst offering high selectivity.
Some recent reports have demonstrated that CuPd systems can
offer enhanced selectivity when Cu is added as a structural modi-
fier to bulk [46] and supported CuPd particles [47,48]. In these
cases, the improved selectivity is attributed to a reduction of the
Pd ensemble size (in the same manner as Ag acts with Pd) as
opposed to the reaction taking place on a Cu surface. It should be
noted that catalytic tests described in this report have been
performed under conditions selected to attain high selectivity to
ethylene whilst removing as much acetylene as possible (i.e. close
to 100% conversion). As such extraction of kinetic parameters
evacuated to a pressure of ca. 4 Â 10 mbar. An initial spectrum
À1
(25 scans, 4 cm resolution) was collected prior to stepwise intro-
duction of increasing CO overpressures (0.1–80 Torr). Results are
presented as difference spectra relative to the initial scan collected
prior to exposure to CO. In one case, a 50:50 w/w mixture of 10%
2 3 2 3
Cu/Al O and 1.67% Pd/Al O (effective metal loading – 5% Cu
and 0.84% Pd) was prepared in order to examine how the spectra
would appear when the two metals were present in the same disc
but as separate metals.
XPS data were collected using a VG Escalab II spectrometer
using Al K
a radiation (1486.6 eV) and a hemispherical analyser
for detection of electrons. The resulting spectra were analysed
using CasaXPS peak fitting software and sample charging corrected
using C 1s as a reference at 285.0 eV.
(
reaction rates, TOFs, activation energies) are of limited value given
that in some cases rates may not reflect a maximum attainable due
to reagent availability. However, such conditions do represent an
excellent choice in order to access the propensity of a catalyst to
over-hydrogenate acetylene to form ethane.
All catalyst testing was performed in a Microactivity reference
reactor (PID Eng
& Tech, supplied by Micromeritics) using
100 mg sample diluted with 400 mg SiC supported in a 9 mm
stainless steel reactor. Prior to reaction, samples were reduced in
3
2 2 2 3
0% H /N for 1 h (323 K for 1.67% Pd/Al O and 523 K for all other
2
. Experimental
samples). Reactions were performed at 1 bar pressure with a
mixture of 0.6% acetylene/1.8% hydrogen/balance N
2
to give a
À1
1
0% Cu/Al
O
2 3
was prepared by wet impregnation. The Al
2
O
3
2
H :acetylene ratio of 3:1 and a space velocity of ca. 24,000 h
.
2
À1
3
À1
(
Aeroxide-Alu C, Evonik, 100 m g
,
q
= 2.9 cm g , particle size
m) was pre-wetted with distilled water and
O dissolved in distilled
The reactor temperature was varied in intervals from 323 to
423 K with 5 h time on stream (TOS) at each temperature. Analysis
of the effluent gas was performed using a PE Clarus 580 GC fitted
with an FID and a 30 m  0.53 mm elite alumina capillary column.
Acetylene conversion was calculated as the amount reacted
divided by the amount introduced. Selectivity to ethylene and eth-
ane was calculated as the amount formed divided by the amount of
acetylene reacted. Ethylene yield has been calculated as acetylene
estimated as 250
the appropriate amount of Cu(NO
water was added dropwise under vigorous stirring. After complete
addition, the slurry was stirred for 1 h followed by drying at 393 K
and calcination at 673 K for 3 h in a flow of 100 ml min . A series
l
3
)
2
Á3H
2
À1
of catalysts with different Cu:Pd ratios (Table 1) were prepared by
adding the appropriate amount of Pd(NO
3
)
2
in a second wetness
Table 1
Nominal metal loadings and Cu:Pd atomic ratio.
Catalyst
Cu (wt%)
Pd (wt%)
Cu:Pd atomic ratio
Estimated CuO crystallite size (nm)a
b
1
1
7
5
2
1
1
0% Cu/Al
00-CuPd
5-CuPd
0-CuPd
5-CuPd
0-CuPd
2
O
3
10
10
10
10
10
10
–
–
–
23 ( 24)
0.17
0.22
0.33
0.67
1.67
1.67
100
75
50
25
10
–
23
23
23
23
22
–
2 3
.67% Pd/Al O
a
Estimated from XRD pattern of calcined sample using the Scherrer equation.
Sample after second calcination.
b

A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
129
conversion multiplied by ethylene selectivity. Selectivity to oligo-
mers was calculated based on a carbon balance, although analysis
allowed for C4 and C6 oligomer fractions be analysed in more
detail. These conditions were purposefully chosen in order to
achieve high or full acetylene conversion and to assess ethylene
selectivity under such conditions. The advantage of operating
under such conditions is that it represents a challenge to achieve
high ethylene selectivity. Industrially, it is also important to reduce
the acetylene concentration down to low ppm level whilst
maintaining high selectivity – thus, mirroring our experimental
conditions. It is noted, however, that at such high conversion, the
reaction rate may be limited by mass transfer, as accessed by
Weisz–Prater [49] and Mears criterion [50,51], and thus, extraction
of kinetic and mechanistic information is not appropriate [2].
(a)
(b)
300
400
500
600
700
800
900
3
. Results
Temperature / K
XPS analysis was performed on 10-CuPd and 50-CuPd samples
2 3 2 3
Fig. 1. TPR profiles for (a) 10% Cu/Al O and (b) 1.67% Pd/Al O .
in order to calculate the Cu:Pd metal ratios (Table 2). Both samples
exhibited well-defined Cu peaks with a Cu 2p3/2 binding energy of
9
32.5 and 932.7 eV for 10-CuPd and 50-CuPd, respectively. The dif-
feature up to ca. 460 K is associated with reduction of Pd species
ference between the Cu 2p3/2 signal for Cu metal, Cu O and CuO is
2
which are harder to reduce due to strong interactions with the
minimal; however, a strong satellite peak observed at ca. 943 eV
confirms that the predominant species is CuO [52]. Much weaker
signals were observed for binding energies associated with Pd
Al
2
O
3
support [54].
The TPR profiles for the bimetallic samples are presented in
Fig. 2 and are more complex than the two monometallic samples.
The first important observation is that irrespective of the Cu:Pd
ratio, there is no evidence of the feature indicative of palladium
hydride formation/decomposition. Instead, the positive features
observed at 347 and 366 K for 10-CuPd and 25-CuPd are attributed
to reduction of PdO since the integrated band area approximately
follows the Pd metal loading. Samples with Cu:Pd ratios of >10:1
show an additional peak in the temperature range 375–396 K
which is thought to be associated with reduction of a mixed oxide
phase. At higher temperatures, at least two features are apparent
with the most prominent being attributed to the two-step reduc-
tion of CuO (as observed for the monometallic Cu sample). 10-CuPd
sample exhibits these peaks at 432 and 460 K which represents a
marked decrease in reduction temperature compared with mono-
metallic Cu. This, therefore, serves as evidence of the close proxim-
ity of the metals on the surface and also that hydrogen spillover
from Pd to Cu occurs. The remainder of the Cu:Pd ratios show Cu
reduction features at temperatures similar to monometallic Cu.
3d5/2 and 3d3/2. For both samples, the binding energy (ca. 337 eV)
would suggest the presence of PdO. The Cu:Pd ratio for 10-CuPd
was calculated from the integrated Cu 2p3/2 and Pd 3d5/2 signals
to be 11.1 which is remarkably similar to the nominal ratio
(Table 1). The equivalent calculation for 50-CuPd gives a ratio of
1
3.8:1, although this value should not be considered as definitive
but merely as a guideline since the Pd signal was exceptionally
weak. Evaluation of XRD data indicates that the average CuO
crystallite size for all samples is in the range 22–24 nm (Table 1).
Clearly, the relative surface compositions as well as metal oxida-
tion states and crystallite size would be expected to change follow-
ing the reduction process.
The H
tion features at 464 and 507 K which are attributed to the two-step
reduction of Cu (i.e. CuO ? Cu O ? Cu). Integration of the
2 2 3
-TPR profile for 10% Cu/Al O (Fig. 1) shows two reduc-
2
combined peak area suggests complete reduction, although the
intensity ratio for the two peaks is not even, suggesting that
+
2+
the sample resides in a mixture of Cu and Cu states following
2 3
FTIR spectra collected on a 10% Cu/Al O sample reduced at
+
2+
calcination. Indeed, the relative proportion of Cu and Cu as well
as the temperature at which these reduction stages take place has
previously been shown to depend on both Cu dispersion and
calcination temperature [53]. The TPR profile does not suggest
the formation of a copper aluminate phase to any significant
extend which is consistent with previous reports which highlight
that calcination temperatures in excess of 873 K are necessary to
form such a phase [26]. The general reduction temperature used
throughout this study for samples containing Cu is 523 K which
523 K and exposed to CO at room temperature (Fig. 3a) show
two bands at 2113 and 2100 cm which are assigned to CO line-
arly adsorbed on Cu and Cu , respectively [55]. Equivalent spectra
of 1.67% Pd/Al were recorded after in situ reduction at 323 K
(Fig. 3b). Whilst the absorbance band intensity is less intense than
observed with monometallic Cu as many as 6 absorbance bands
can be observed – 3 high-frequency bands (2150–2000 cm
and 3 low-frequency bands (2000–1700 cm ). The band at
2117 cm is assigned to on-top adsorption on Pd cations [56],
although the intensity is distorted at higher overpressures as the
gas phase CO doublet centred at 2143 cm becomes more prom-
inent. Bands at 2095 and 2077 cm are assigned to linear adsorp-
tion steps/edges and terraces, respectively. Features at 1973 and
À1
+
0
2 3
O
À1
)
À1
À1
0
may not result in complete reduction but a mixture of Cu and
+
À1
Cu states (see FTIR results later). The H
2
-TPR profile for 1.67%
(Fig. 1) shows a negative feature at 338 K, due to the
decomposition of palladium hydride which readily forms upon
À1
Pd/Al
2
O
3
À1
exposure to H
2
at ambient temperature [7]. A broad reduction
1930 cm are assigned to
l
2
bridge-bonded CO, most likely on
Table 2
XPS binding energy for Cu and Pd signals and the calculated Cu:Pd ratio.
Catalyst Binding energy of Cu (eV)
Binding energy of Pd (eV)
3d5/2
Cu:Pd ratio
2
p
3/2
2p1/2
3d3/2
1
5
0-CuPd
0-CuPd
932.5
932.7
952.3
952.6
337.4
336.8
342.9
342.4
11.1
13.8

1
30
A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
0.16
0.14
0.12
0.10
0.08
4
2
1
5
2
1
0
0
1
0
0
(
e)
.7
.3 torr
(
(
d)
c)
0.06
0.04
0
.02
.00
(
(
b)
0
a)
2300
2200
2100
2000
1900
-1
1800
1700
300
400
500
600
700
800
900
Wavenumber / cm
Temperature / K
Fig. 4. FTIR spectra of reduced 50:50 w/w mixture of 10% Cu/Al
Al after exposure to CO at 298 K at increasing pressures.
2
O
3
and 1.67% Pd/
Fig. 2. TPR profiles for (a) 100-CuPd, (b) 75-CuPd, (c) 50-CuPd, (d) 25-CuPd and (e)
0-CuPd.
2 3
O
1
À1
steps/edges and terraces. The shoulder at 1878 cm (most appar-
ent at higher pressure) is attributed to
three-fold sites.
is consistent with expectation based on the lower enthalpy of sub-
limation of Cu [43]. There is no evidence of pronounced bands
which can be assigned to Pd, although the 10-CuPd material does
3
l hollow-bound CO in
À1
Determining the composition of the surface layer of the bime-
tallic samples is of the upmost importance with regards to catalytic
performance. In order to identify if one metal predominated on the
surface or if there were monometallic Pd particles present, FTIR
spectra were collected after reduction of the CuPd samples at
exhibit an unusually broad tail in the region 2070–2050 cm per-
haps indicative of a degree of linearly adsorbed CO (see catalytic
data later). Importantly, there is no evidence of multiply bound
CO (comparison of Figs. 4 and 5) which suggests that monometallic
Pd nanoparticles are absent, and that there are no pronounced
ensembles of Pd on the surface of the bimetallic particles.
5
23 K and exposure to CO. Firstly, an experiment was performed
with a physical mixture of the two monometallic catalysts (compo-
sition: 5 wt% Cu and 0.84 wt% Pd) to identify which spectral
features should be observed when the metals are present as mono-
metallic particles (Fig. 4). The spectra essentially appear as a
composite of those presented in Fig. 3. Bands present at 2113
Catalyst testing was conducted using powdered catalyst
(R ꢀ 125
lm). Using such fine powders, it is unexpected that the
reaction rate should be limited by internal mass transfer. This
was confirmed by application of the Weisz–Prater [49] criterion
which was met assuming that gas phase diffusivity fell within
À1
À7
À7
2
and 2100 cm are assigned to CO adsorption on Cu (as before)
the range, 1.83 Â 10 to 2.57 Â 10 m s or greater. These values
are lower than the effective pore diffusivities measured by Asplund
[57] for acetylene hydrogenation catalysts in the form of pellets
and crushed pellets of larger particle size than those employed
here. Furthermore, a 5-point Arrhenius plot (323–373 K) used to
calculate apparent activation energy for acetylene conversion up
to 99.5% (R = 0.986) give a value of 47.3 kJ mol (for 50-CuPd)
similar to the value (46.8 kJ mol ) reported by McGowan et al.
[58]. Values of 40 kJ mol [59] and between 40 and 50 kJ mol
À1
and bands at 2077, 1986, 1930 and 1878 cm assigned to adsorp-
0
tion on Pd (as before). Interestingly, the intensity ratio of the Cu to
+
Cu bands is higher in the physical mixture compared with mono-
metallic Cu. This would suggest that Cu is more reduced and may
imply that Pd promoted reduction can take place even without
bimetallic particle formation. Importantly, this experiment high-
lights that with a 6 fold excess of Cu relative to Pd, absorbance
bands associated with Pd particles can clearly be observed. Results
of the experiments performed with the range of CuPd ratios are
shown in Fig. 5. For all samples, the spectra appear similar and
2
À1
À1
À1
À1
[60] have been reported for acetylene hydrogenation to ethylene
under conditions which have been tested and confirmed to be rep-
resentative of reaction in the absence of transfer resistances. This is
not the case for data obtained at acetylene conversion levels closer
0
+
are dominated by the characteristic Cu /Cu doublet suggesting
that the samples predominantly expose Cu on the surface which
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.10
(a)
(b)
39.4
1
9
9.5
.4
0.08
6
32.4
5.5
4.5
2.1
1.2
0.5
1
7
4
2
6.3
.9
.3
.0
0.06
0.2 torr
0.9
0
.04
.02
0.4
0.2 torr
0
0.00
2300
2200
2100
2000
1900
1800
2300
2200
2100
2000
1900
1800
1700
-1
-1
Wavenumber / cm
Wavenumber / cm
2 3 2 3
Fig. 3. FTIR spectra of reduced monometallic catalysts after exposure to CO at 298 K at increasing pressures. (a) 10% Cu/Al O and (b) 1.67% Pd/Al O .

A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
131
0
0
0
0
0
0
0
0
0
.8
.7
.6
.5
.4
.3
.2
.1
.0
(
a)
0
.14 (b)
67
40
0.12
65
30
3
1
8
4
2
0
2
6
1
8
4
2
1
0
0
6
0.10
0.08
.8
.6
.3 torr
0.1 torr
0.06
0.04
0.02
0.00
2
300
2200
2100
2000
1900
1800
2300
2200
2100
2000
1900
1800
-1
-1
Wavenumber / cm
Wavenumber / cm
0
.6
.5
.4
.3
.2
.1
.0
0.6
0.5
0.4
0.3
0.2
(c)
(d)
0
0
0
0
0
0
8
40
20
10
5
2
1
0.4
0
6
6
3
1
8
4
2
1
0
0
5
6
.5
.2 torr
0.1 torr
0
.1
.0
0
2300
2200
2100
2000
1900
1800
2300
2200
2100
2000
1900
1800
-1
-1
Wavenumber / cm
Wavenumber / cm
Fig. 5. FTIR spectra of reduced CuPd catalysts after exposure to CO at 298 K at increasing pressures. (a) 10-CuPd, (b) 25-CuPd, (c) 50-CuPd and (d) 100-CuPd.
to 100% where reaction rates become limited by reagent availabil-
ity and consequently rates (and TOF values) will not be reported.
1
00
1
00 (a)
(b)
Results indicate (Table 3) that at low temperature (323 K) the
intrinsic ability of Cu to convert acetylene is low and temperatures
approaching 423 K are necessary to achieve full conversion. At all
temperatures investigated, ethane is produced only in minor
quantities verifying that Cu is intrinsically selective in terms of
C2 products produced. Inspite of this, ethylene selectivity is limited
to between 53% and 72% due to oligomers formed in large quanti-
ties. These results are consistent with previous literature reports
which highlight that high temperatures are necessary for Cu to
be both active and selective for alkyne hydrogenation [23–27].
9
0
80
8
6
4
2
0
0
0
0
0
7
0
0
6
50
40
3
0
20
10
0
1
2 3
.67% Pd/Al O displayed much higher conversion (Table 3) at
low temperature (>85% at 323 K) but almost exclusively formed
ethane with only a small amount of oligomers. Comparison of
the Cu and Pd reaction data highlights the activity of Pd at low
temperatures and the selectivity of Cu at high temperatures and
correlates well with similar tests reported using propyne [38].
The different CuPd catalysts were screened for activity/selectiv-
ity at 323, 373 and 423 K (Figs. 6–8). 10-CuPd shows enhanced activ-
ity at 323 K relative to monometallic Cu. However, the product
distribution favours the formation of ethane suggesting that the
reaction is taking place, at least in part, on Pd sites, or sites strongly
Fig. 6. (a) Acetylene conversion and (b) product selectivity for CuPd catalysts at
23 K after 5 h time on stream. Acetylene conversion (grey), ethylene selectivity
yellow), ethane selectivity (cyan) and oligomer selectivity (pink). (For interpreta-
tion of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
3
(
Table 3
2 3 2 3
Acetylene conversion and ethylene/ethane/oligomer selectivity over 10% Cu/Al O and 1.67% Pd/Al O .
Catalyst
Temperature (K)
323
Acetylene conversion (%)
Ethylene selectivity (%)
Ethane selectivity (%)
Oligomer selectivity (%)
1
0% Cu/Al
2
O
3
9.59
60.71
99.20
53.2
60.5
71.5
9.9
8.6
8.0
37.0
30.8
20.5
3
4
73
23
1
.67% Pd/Al
2
O
3
323
85.17
0.0
93.1
6.9

1
32
A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
(
a)
100
Table 4
100
(b)
Ethylene yield at 323, 373 and 423 K for CuPd catalysts.
90
80
70
Catalyst
Ethylene yielda (%)
8
6
4
2
0
0
0
0
0
323 K
373 K
423 K
1
7
00-CuPd
5-CuPd
3.3
1.4
7.1
17.1
14.3
23.6
19.5
71.1
66.3
6.6
56.9
76.1
80.3
68.0
9.7
60
50
50-CuPd
2
1
5-CuPd
0-CuPd
40
30
a
Ethylene yield = acetylene conversion  ethylene selectivity.
20
10
Based on the success of the screening tests, additional reactions
were performed to identify the lowest possible temperature at
which 50-CuPd offers sufficient activity and selectivity. Fig. 9 pre-
sents the time on stream data at a range of temperatures in the
range 323–423 K. At 323 and 333 K there was variation with activ-
ity/selectivity as the catalyst approached steady state conditions.
At higher temperatures there was little sign of deactivation or
marked changes in activity/selectivity beyond experimental scat-
ter. 50-CuPd produces close to 70% conversion at 363 K and
0
Fig. 7. (a) Acetylene conversion and (b) product selectivity for CuPd catalysts at
73 K after 5 h time on stream. Acetylene conversion (grey), ethylene selectivity
yellow), ethane selectivity (cyan) and oligomer selectivity (pink). (For interpreta-
tion of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
3
(
9
9.6% conversion at 373 K. The biggest difference in selectivity
between these two temperatures is the selectivity to oligomers
which is almost 35% at 363 K which is far from desirable. In order
to increase the rate of hydrogen dissociation there are two options
100
100
–
2
either increase temperature or increase the H :acetylene ratio.
9
8
7
6
5
0
0
0
0
0
Increasing temperature beyond 373 K is undesirable since the
aim of this work is to produce a Cu-based catalyst which can be
used at industrially applicable temperatures. Instead, a reaction
80
60
40
20
0
was performed with 50-CuPd but using a 10:1 H
2
:acetylene ratio
(Fig. 10). Under these conditions, activity at 343 K is somewhat
unstable with steady state conversion not apparent after 5 h TOS;
however, at 353 and 363 K conversion in excess of 99% was
40
30
observed. Selectivity at these temperatures show
decrease in oligomer formation (3:1 H :acetylene = 35% vs 10:1 H2-
acetylene = 10%). The step changed in selectivity at 353 K is cur-
rently not well understood.
a marked
20
2
:
10
0
4
. Discussion
.1. Monometallic Cu and Pd catalysts
Prior to discussing the merits of bimetallic CuPd catalysts, it is
4
Fig. 8. (a) Acetylene conversion and (b) product selectivity for CuPd catalysts at
23 K after 5 h time on stream. Acetylene conversion (grey), ethylene selectivity
yellow), ethane selectivity (cyan) and oligomer selectivity (pink). (For interpreta-
tion of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
4
(
important to consider the catalytic properties of the monometallic
catalysts. Numerous literature studies indicate that Cu is not espe-
5h
100
influenced by the presence of Pd (compare Fig. 6 and Table 3). The
actual ethane selectivity is 53% which is lower than observed with
pure Pd (93%) which would suggest that Cu acts as a surface diluent
decreasing the size of the Pd ensembles and enhancing selectivity as
previously reported [47]. Cu:Pd ratios of 75:1 and higher result in
activity which is either equivalent to or less than monometallic Cu
and suggests that insufficient Pd is present to enhance activity. 25-
CuPd and 50-CuPd catalysts demonstrate enhanced activity at
90
8
0
0
7
60
50
3
23 K but also selectivity very similar to 10% Cu/Al
increasing the temperature to 373 K, the conversion increases to
99% and with an ethylene selectivity of >65%. Neither catalyst pro-
2
O
3
. Upon
40
3
0
0
>
2
duces ethane to any significant extent which suggests that the reac-
tion is predominantly taking place on Cu with Pd playing a
secondary role in activating the surface towards hydrogen dissocia-
tion. As temperature is increased to 423 K, 50-CuPd stands out as
being the most selective with >80% selectivity to ethylene and less
than 10% selectivity to ethane at >99.8% conversion. Comparison of
the ethylene yield (Table 4) further highlights that 50-CuPd is the
most selective catalyst with an ethylene yield of 71.1% and 80.3%
achieved at 373 and 423 K, respectively.
10
0
323 K 333 K 343 K 353 K 363 K 373 K 398 K 423 K
Fig. 9. Data for 50-CuPd for 5 h time on stream at temperatures between 323 and
23 K. Acetylene conversion (grey square), ethylene selectivity (yellow circles),
ethane selectivity (cyan triangles) and oligomer selectivity (pink diamonds). (For
interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
4

A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
133
1
00
is clear that selectivity over monometallic Pd is a complex
problem, it is important to recognise that industrial catalysts are
generally based on PdAAg alloys which offer enhanced alkene
selectivity. The role of Ag is likely twofold: firstly, it acts as a
surface diluent reducing Pd ensemble size which is thought to
increase selectivity [2]. Secondly, Ag should also limit the forma-
tion of sub-surface hydrogen [6].
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
5
h
4.2. CuAPd interactions
Previous reports on a Cu (111) single crystal modified with iso-
lated palladium atoms demonstrated that hydrogen dissociation
and spillover onto Cu happened readily [44,45]. The same group
have more recently demonstrated that this strategy may be
extended to powdered catalysts but required the use of galvanic
reduction to control Pd deposition onto a Cu surface for use in
phenylacetylene hydrogenation [64]. Encouraged by these results,
we prepared a series of powder catalysts with different Cu:Pd
ratios with the aim of preparing a catalyst with the following
characteristics:
3
43 K
353 K
363 K
Fig. 10. Data for 50-CuPd for 5 h time on stream at temperatures between 343 and
63 K using a 10:1 H :acetylene ratio. Acetylene conversion (grey square), ethylene
selectivity (yellow circles), ethane selectivity (cyan triangles) and oligomer
selectivity (pink diamonds). (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
3
2
1
. Bimetallic CuPd particles with no monometallic Pd particles.
cially active for alkyne hydrogenation at low temperature and
tends to produce significant quantities of oligomers, leading to
deactivation [23–26,51]. It is therefore necessary to operate at
higher temperatures (423–523 K) where oligomers are formed to
a lesser extent. Even at high temperature, significant oligomer
formation can take place with acetylene [27] being more prone
to self-coupling than propyne [27] or butyne [61]. Importantly,
copper is inherently selective in terms of the C2 products produced
with little or no alkane formed indicating that it represents an
interesting choice of metal for alkyne hydrogenation [23–27,56].
The selective nature of Cu has been explored by DFT calculations
and is generally attributed to the difference in adsorption energies
for acetylene and ethylene [38]. The activity and selectivity of 10%
2. A surface dominated by Cu without segregated Pd islands.
3. Catalytic properties dominated by Cu but enhanced by Pd.
Previous literature studies indicate that Cu and Pd readily form
alloy particles across a wide range of compositions [42,43]. FTIR
spectra of the bimetallic catalysts after reduction (Fig. 5) show fea-
0
+
tures which are attributed to CO adsorbed on Cu and Cu , strongly
suggesting that the surface is dominated by Cu atoms. The lack of
absorption bands on the bimetallic catalysts in the range 2080–
À1
1700 cm suggests that there is little or no contribution from Pd
(compare Fig. 3b and Fig. 5). More importantly, the absence of
bands due to adsorption of CO in bridging and threefold hollow
modes suggests that there are no monometallic Pd particles pres-
ent or particles containing larger Pd only facets. It is noted that
2 3
Cu/Al O used in this study, correlates well with literature reports
both in terms of activity and selectivity, with the best ethylene
selectivity of 71.5% achieved at 423 K (Table 3). Both the lack of
activity at low temperature and the tendency to form oligomers
stem from the difficulty of dissociating H on a Cu surface. Theoret-
2
ical studies have calculated the activation energy barrier for disso-
ciation for a Cu (111) surface to be in the range 0.45–0.50 eV
10-CuPd sample exhibits a possible shoulder at approximately
À1
2060 cm
which may be related to an increased Pd surface
concentration. Comparison of the spectra collected for bimetallic
samples (Fig. 5) with the physical mixture of the two monometallic
samples (Fig. 4) highlights the expected spectral features for a sam-
ple containing separate individual particles of Cu and Pd. Given
that the spectra for bimetallic samples more closely resemble
those for monometallic Cu samples, reinforces the conclusion that
the majority, if not all of the Pd exists as an alloyed phase.
[
0
39,62], although experimentally determined values closer to
.4 eV have been reported [63]. As such, monometallic Cu does
not represent a viable industrial catalyst unless limited hydrogen
dissociation can be overcome allowing for use at lower
temperature.
Pd is the metal of choice for many low-moderate temperature
hydrogenation reactions since it can dissociatively adsorb hydro-
The key to producing a catalyst where selectivity is governed by
Cu and not Pd is optimising the Cu:Pd ratio. Previous literature
reports demonstrate that small Cu:Pd ratios in the range 0.3–1.5
produce a catalyst which is dominated by Pd active sites with Cu
acting to dilute the Pd surface and reduce ensemble size [47].
Results obtained in this study with a 10:1 Cu:Pd ratio (Fig. 6)
produces a catalyst which is both active and unselective at low
temperature (323 K) – both traits associated with Pd dominated
catalysis. It can be concluded that using this preparation route, a
Cu:Pd ratio of greater than 10 is necessary to produce a catalyst
which is not dominated by Pd-like behaviour. Similarly, Cu:Pd
ratios of 75:1 and 100:1 produce activity and selectivity which
appears very similar to pure Cu which infer too high a Cu:Pd ratio
(Fig. 6). It is interesting to note that the single crystal study of
Sykes and co-workers [44,45] observed enhanced hydrogen disso-
ciation at a Cu:Pd ratio of approximately 100, thus highlighting
differences between powder and single crystal studies. A slight
discrepancy observed within this study is that both 75-CuPd and
2 3
gen around room temperature. The 1.67% Pd/Al O catalyst used
in this study reaffirms this, with >85% acetylene conversion
observed at only 323 K (Table 3). However, a number of studies
highlight that monometallic palladium is not always selective
and is prone to over-hydrogenation forming the alkane in signifi-
cant quantities (>93% selectivity in this work) [19,21,38]. Selectiv-
ity over Pd is now thought to be governed by the interplay between
hydride [6,7] and carbide [3–5] phases both on the surface and in
the sub-surface region. Through the combined use of DFT calcula-
tions and new surface science techniques such as prompt gamma
activation analysis (PGAA) it is possible to probe the palladium
hydride phase like never before [3,8]. This has led to several groups
showing that the migration of dissolved hydrogen adatoms from
the bulk to the surface results in rapid C@C bond hydrogenation
[
8–12], although some studies highlight this may also occur for tri-
2 3
100-CuPd are less active than 10% Cu/Al O at 323 K. This effect
ple bonds [6]. It has also been shown that carbon plays a crucial
role in replenishment of sub-surface hydrogen [9–13]. Whilst it
is likely associated with a decrease in the active metal surface area
brought about during the second calcination step necessary using

1
34
A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
the sequential impregnation method. Comparison of the CuO crys-
tallite size determined from XRD suggests a slightly larger crystal-
lite size following a second calcination, although the difference
was minimal (Table 1). The impact on activity was verified by sub-
jecting 10% Cu/Al O to a second calcination which resulted in
2 3
decreased acetylene conversion at 323 K (5.4% vs 9.6%).
stability and ethylene selectivity of up to 75% but require opera-
tional temperatures approaching 473 K in order to achieve high
acetylene conversion [17]. The ternary CuNiFe catalyst reported by
Bridier and Perez-Ramírez [41] demonstrated perfect selectivity of
100% for propene and 80% selectivity for ethylene but required acti-
vation at 773 K and operation at 523 K. This does not therefore rep-
resent a direct replacement for PdAg catalysts (as acknowledged by
the authors) [41]. NiAZn is reported to be very selective, although
insufficient information was reported to make comment on the tem-
perature necessary to achieve high conversion [14]. The CuPd cata-
lysts described in this work offer ethylene selectivity of up to 80%
at only 363 K and therefore could be used industrially if longer term
stability was demonstrated. To the best of our knowledge, the only
system reported to date which could operate at lower temperature,
with higher selectivity than 50-CuPd is bimetallic AuANi. Nikolaev
and co-workers reported that 100% ethylene selectivity could be
Both 25:1 and 50:1 Cu:Pd ratios show enhanced activity at
3
23 K (Fig. 6) compared to monometallic Cu but also display selec-
tivity that suggest the reaction takes place on a Cu surface. These
two catalysts demonstrate that the concept of H activation on
2
Pd and spillover onto Cu can be transferred from single crystal
studies to powdered catalysts [59]. However, it should be noted
that any modification to the metal composition and its electronic
properties might also influence the relative rates of ethylene
desorption relative to further hydrogenation. Weaker adsorption
of ethylene on copper compared with palladium facilitates desorp-
tion with respect to hydrogenation of ethylene although this needs
to be balanced by sufficiently strong acetylene adsorption to avoid
significant loss in hydrogenation rate of the latter. This implies a
delicate balance between retention of strongly adsorbing Pd sites
and more weakly adsorbing Cu sites. The activity and selectivity
of 25:1 and 50:1 Cu:Pd even exceeds that observed with 10% Cu/
2 3
obtained at 357 K over 0.27% Au 0.09% Ni/c-Al O [35,60]. It is worth
noting that this study has been conducted using a feed stream which
does not contain ethylene, whereas the industrial process takes
place in a large excess of ethylene. To verify industrial applicability,
bimetallic catalysts should be tested under conditions more repre-
sentative of the industrial process. However, previous studies indi-
cate that selectivity trends observed in single reagent experiments
can give a good indication of what can be expected using mixed C2
feed streams [19]. Tests with CuPd catalysts in the presence of both
acetylene and ethylene are being undertaken but fall out with the
scope of this manuscript.
Al
an ethylene selectivity of 61% (Table 3) whereas 50-CuPd produces
99% conversion and 71% selectivity (Fig. 7). The difference in
2 3
O . At 373 K monometallic Cu produces 61% conversion and
>
selectivity arises primarily as a result of the formation of fewer
oligomers which suggests that hydrogen adatoms are more readily
available on the bimetallic surface, consistent with Pd enhancing
the ease with which molecular hydrogen dissociates. It is also help-
ful to consider the ethylene yield at 373 K. Monometallic Cu pro-
duces an ethylene yield of 36.7% whereas 50-CuPd produces a
considerably improved yield of 71.1% (Table 4).
4.3. Mechanism of enhanced activity
There are two questions which remain to be answered. Firstly,
can CuPd catalysts form a hydride phase? Recent work from
Gapipaud et al. [66] and Friedrich et al. [46] suggests that CuPd
mixtures do not form a hydride phase for Cu:Pd bulk ratios of 0.5
and above. Based on these reports, it is not expected that any of
the Cu:Pd ratios described in this report should be able to form a
hydride phase. Indeed, TPR measurements confirm the absence of
Pd-hydride (Fig. 2). Secondly, what is the nature of the key site
which promotes hydrogen dissociation and spillover? The original
report from Tierney et al. [44] suggested that the active site for dis-
sociation is isolated Pd atoms but highlighted that Pd populated
both surface and sub-surface sites when dosed on a Cu (111) sur-
face. A more recent study by Fu and Luo [39] calculated the varia-
tion of the barrier which limits hydrogen dissociation on a Cu
Inspite of a significant improvement in activity at lower tem-
perature, there still remains two challenges for CuPd-based cata-
lysts. Firstly, operation at lower temperature tends to lead to
more significant quantities of oligomers, therefore reducing alkene
yield. Secondly, extensive oligomer formation can lead to more
rapid deactivation. TOS data for 50-CuPd (Fig. 9) does not show
any signs of deactivation over a 5 h period despite considerable oli-
gomer formation and suggests that the bimetallic catalysts are
more stable than monometallic Cu. This may be in part associated
with the type of oligomeric species formed. Large oligomers will
result in deactivation whereas shorter oligomers such as C4 and
C6 species will desorb. Over 50-CuPd, a significant quantity of
the oligomers desorb and based on GC analysis, over 80% of those
which desorb are C4 compounds. Similar findings were reported
for a Pd surface where ensemble size was reduced by exposure
to diphenyl sulphide [20]. The formation of predominantly C4 olig-
omers suggests that fewer larger oligomers are forming meaning
less significant deactivation. Indeed, the formation of C4 com-
pounds such as 1,3-butadiene and butene are desirable from an
industrial point of view but would of course complicate down-
stream processing. Since the bimetallic catalysts prepared in this
work produce relatively little ethane, an option for reducing oligo-
(
111) surface modified with Pd. Their results show that the addi-
tion of a single Pd atom decreases this barrier from 0.46 eV to
.29 eV. However, when a surface Pd atom was bonded directly
0
to sub-surface Pd atoms, a more significant decrease in the disso-
ciation barrier was obtained (ꢁ0.05 eV). Although the characterisa-
tion presented in this report is limited, it is noted that XPS analysis
for 10-CuPd and 50-CuPd in their calcined states gave Cu:Pd ratios
of 11:1 and ca. 14:1 being determined, although FTIR of CO on the
reduced catalysts suggest that the proportion of surface Pd is
diminished between calcination and reduction states. Given that
the two catalysts which offer enhanced activity (25-CuPd and 50-
CuPd) appear to contain very little Pd in surface sites (Fig. 5b and
c), it appears entirely plausible that a non-trivial amount of Pd
occupies the near-surface layers. The enhanced activity is therefore
thought to be associated with small clusters of Pd atoms which
encompass both the surface and sub-surface region.
mer formation is to operate at higher H
2
:acetylene ratios. Further
TOS data at a 10:1 H :acetylene ratio (Fig. 10) shows that oligomer
2
formation can be reduced to approximately 10% at 363 K with eth-
ane selectivity remaining around 10%. It should be noted that the
option to operate at an increased H :acetylene ratio is not readily
2
available for Pd-based catalysts since this would promote ethane
formation. The step changed in selectivity at 353 K is currently
not well understood but may be related to carbon lay down.
The concept of using a bimetallic or trimetallic catalyst for alkyne
hydrogenation is not uncommon in the literature. Notable examples
offering high selectivity include PdGa [16,17], NiAZn [14], AuANi
5. Conclusions
A simple sequential impregnation methodology allowed bime-
tallic CuPd catalysts to be prepared with a Cu dominated surface.
[
35,65] and CuNiFe [41]. Intermetallic PdGa catalysts offer excellent

A.J. McCue et al. / Journal of Catalysis 319 (2014) 127–135
135
Both 25:1 and 50:1 Cu:Pd ratios result in a combination of the
catalytic properties of Cu and Pd for acetylene hydrogenation (i.e.
high ethylene selectivity and activity at low temperature). Use of
[27] B. Bridier, M.A.G. Hevia, N. López, J. Pérez-Ramírez, J. Catal. 278 (2011) 167–
72.
28] S. Abelló, D. Verboekend, B. Bridier, J. Pérez-Ramírez, J. Catal. 259 (2008) 85–
5.
[29] D.L. Trimm, I.O.Y. Liu, N.W. Cant, Appl. Catal. A: Gen. 374 (2010) 58–64.
1
[
9
5
0-CuPd resulted in >99% acetylene conversion and >70% ethylene
[
[
30] D.L. Trimm, N.W. Cant, I.O.Y. Liu, Catal. Today 178 (2011) 181–186.
31] J. Jia, K. Karaki, J.N. Kondo, K. Domen, K. Tamaru, J. Phys. Chem. B 104 (2000)
selectivity at 373 K, corresponding to at least a 50 K decrease in the
temperature needed to obtain the same degree of conversion over
monometallic Cu. The selectivity to ethylene can be further
11153–11156.
[32] T.V. Choudhary, C. Sivadinarayana, A.K. Datye, D. Kumar, D.W. Goodman, Catal.
Lett. 86 (2003) 1–8.
enhanced by increasing the H
2
:acetylene ratio (>99% conversion,
[
[
33] J.A. Lopez-Sanchez, D.L. Lennon, Appl. Catal. A: Gen. 291 (2005) 230–237.
34] Y. Segura, N. López, J. Pérez-Ramírez, J. Catal. 247 (2007) 383–386.
>80% selectivity to ethylene at 363 K). Enhanced activity is attrib-
uted to hydrogen dissociation on Pd with spillover to neighbouring
Cu sites where the reaction takes place. These results suggest that
CuPd catalysts could be used in the temperature range where Pd
catalysts are currently employed industrially.
[35] S.A. Nikolaev, V.V. Smirnov, Catal. Today 147S (2009) S336–S341.
[
[
36] A. Sárkány, Z. Révay, Appl. Catal. A: Gen. 243 (2003) 347–355.
37] G. Vilé, D. Baudouin, I.N. Remediakis, C. Copéret, N. López, J. Pérez-Ramírez,
ChemCatChem 5 (2013) 3750–3759.
[38] B. Bridier, N. López, J. Pérez-Ramírez, Dalton Trans. 39 (2010) 8412–8419.
[
[
39] Q. Fu, Y. Luo, ACS Catal. 3 (2013) 1245–1252.
40] N. Lopez, Z. Lodziana, F. Illas, M. Salmeron, Phys. Rev. Lett. 93 (2004) 146103-
References
1–146103-4.
[
[
41] B. Bridier, J. Perez-Ramírez, J. Am. Chem. Soc. 132 (2010) 4321–4327.
42] M. Fernández-García, J.A. Anderson, G.L. Haller, J. Phys. Chem. 100 (1996)
[
1] S.A. Nikolaev, I.L.N. Zanaveskin, V.V. Smirnov, V.A. Averyanov, K.L. Zanaveskin,
Russ. Chem. Rev. 78 (2009) 231–247.
2] A. Borodzinski, G.C. Bond, Catal. Rev. 50 (2008) 379–469.
3] D. Teschner, J. Borsodi, A. Wootsch, Z. Revay, M. Kaveckre, A. Knop-Gericke, S.D.
Jackson, R. Schlögl, Science 320 (2008) 86–89.
16247–16254.
[
[
[
[
[
43] M. Fernández-García, A. Martínez-Arias, C. Belver, J.A. Anderson, J.C. Conesa, J.
Soria, J. Catal. 190 (2000) 387–395.
44] H.L. Tierney, A.E. Baber, J.R. Kitchin, E.C.H. Sykes, Phys. Rev. Lett. 103 (2009)
[
[
4] H. Gabasch, K. Hayek, A. Knop-Gericke, R. Schlögl, J. Phys. Chem. B 110 (2006)
246102-1–246102-4.
4
947–4952.
5] M. García-Mota, B. Bridier, J. Pérez-Ramírez, N. López, J. Catal. 273 (2010)
2–102.
45] G. Kyriakou, M.B. Boucher, A.D. Jewell, E.A. Lewis, T.J. Lawton, A.E. Baber, H.E.
Tierney, M. Flytzani-Stephanopoulos, E.C.H. Sykes, Science 335 (2012) 1209–
9
1212.
[
[
6] N.A. Khan, S. Shaikhutdinov, H.J. Freund, Catal. Lett. 108 (2005) 159–164.
7] J. Sá, G.D. Arteaga, R.D. Daley, J. Bernardi, J.A. Anderson, J. Phys. Chem. B 110
}
[
[
[
46] M. Friedrich, S.A. Villaseca, L. Szentmiklósi, D. Teschner, M. Armbruster,
Materials 6 (2013) 2958–2977.
47] S.K. Kim, J.H. Lee, I.Y. Ahn, W.J. Kim, S.H. Moon, Appl. Catal. A: Gen. 401 (2011)
(
2006) 17090–17095.
[
8] M. Armbr u} ster, M. Behrens, F. Cinquini, K. Föttinger, Y. Grin, A. Haghofer, B.
Klötzer, A. Knop-Gericke, H. Lorenz, A. Ota, S. Penner, J. Prinz, C. Rameshan, Z.
Révay, D. Rosenthal, G. Rupprechter, D. Teschner, D. Torres, R. Wagner, R.
Widmer, G. Wowsnick, ChemCatChem 4 (2012) 1048–1063.
12–19.
48] A. Cooper, B. Bachiller-Baeza, J.A. Anderson, I. Rodríguez-Ramos, A. Guerrero-
Ruiz, Catal. Sci. Technol. 4 (2014) 1446–1455.
49] P.B. Weisz, C.D. Prater, Adv. Catal. 6 (1954) 143.
50] D.E. Mears, Ind. Eng. Chem. Process. Des. Dev. 10 (1971) 541.
51] S.T. Oyama, X. Zhang, J. Lu, Y. Gu, T. Fujitani, J. Catal. 257 (2008) 1–4.
52] J. Batista, A. Pintar, D. Mandrino, M. Jenko, V. Matin, Appl. Catal. A: Gen. 206
[
[
[
[
[
9] S. Schauermann, N. Nilius, S. Shaikhutdinov, H.J. Freund, Acc. Chem. Res. 46
(
2013) 1673–1681.
10] W. Ludwig, A. Savara, K.-H. Dosert, S. Shauermann, J. Catal. 284 (2011) 148–
56.
11] W. Ludwig, A. Savara, R.J. Madix, S. Schauermann, H.J. Freund, J. Phys. Chem. C
16 (2012) 3539–3544.
12] M. Wilde, K. Fukutani, W. Ludwig, B. Brandt, J.-H. Fischer, S. Schauermann, H.J.
Freund, Angew. Chem. Int. Ed. 47 (2008) 9289–9293.
13] F. Viñes, C. Loschen, F. Illas, K.M. Neyman, J. Catal. 266 (2009) 59–63.
14] F. Studt, F. Abild-Pedersen, T. Bligaard, R.Z. Sørensen, C.H. Christensen, J.K.
Nørskov, Angew. Chem. Int. Ed. 47 (2008) 9299–9302.
[
[
[
1
(
2001) 113–124.
[
[
[
[
53] J.M. Dumas, C. Geron, A. Kribii, J. Barbier, Appl. Catal. 47 (1989) L9–L15.
54] V.H. Sandoval, C.E. Gigola, Appl. Catal. A: Gen. 148 (1996) 81–96.
55] P. Hollins, Surf. Sci. Rep. 16 (1992) 51–94.
56] J.A. Anderson, M. Fernandez-Garcia, A. Martinez Arias, in: J.A. Anderson, M.
Fernandez-Garcia (Eds.), Supported Metals in Catalysis, second ed., ICP,
London, 2004, p. 123.
57] S. Asplund, J. Catal. 158 (1996) 267–278.
58] W.T. McGowan, C. Kemball, D.A. Whan, M.S. Scurrell, J. Chem. Soc., Faraday
Trans. 1 (76) (1977) 632–647.
59] A.N.R. Bos, E.S. Bootsma, F. Foeth, H.W.J. Sleyster, K.R. Westerterp, Chem. Eng.
Process. 32 (1993) 53–63.
60] A. Pachulski, R. Schödel, P. Claus, Appl. Catal. A: Gen. 445–456 (2012) 107–120.
61] R.A. Koeppel, J.Y. Wehrli, M.S. Wainwright, D.L. Trimm, N.W. Cant, Appl. Catal.
A: Gen. 120 (1994) 163–177.
62] B. Hammer, J.K. Nørskov, Nature 376 (1995) 238–240.
63] T. Kammler, J. Küppers, J. Chem. Phys. 111 (1999) 8115–8123.
64] M.B. Boucher, B. Zugic, G. Cladaras, J. Kammert, M.D. Marchinkowski, T.J.
Lawton, E.C.H. Sykes, M. Flytzani-Stephanopoulous, Phys. Chem. Chem. Phys.
1
[
[
[
[
[
[
15] B. Yang, R. Burch, C. Hardacre, G. Beaddock, P. Hu, J. Catal. 305 (2013) 264–276.
16] J. Osswald, R. Giedigkeit, R.E. Jentoft, M. Armbr u} ster, F. Girgsdies, K. Kovnir, R.
Ressler, Y. Grin, R. Schlögl, J. Catal. 258 (2008) 210–218.
[
[
[
17] J. Osswald, K. Kovnir, M. Armbr u} ster, R. Giedigkeit, R.E. Jentoft, U. Wild, Y. Grin,
R. Schlögl, J. Catal. 258 (2008) 219–227.
[
[
18] F.M. McKenna, R.P.K. Wells, J.A. Anderson, Chem. Commun. 47 (2011) 2351–
2
353.
[
[
19] F.M. McKenna, J.A. Anderson, J. Catal. 281 (2011) 231–240.
20] F.M. McKenna, L. Mantarosie, R.P.K. Wells, C. Hardacre, J.A. Anderson, Catal. Sci.
Technol. 2 (2012) 632–638.
21] A.J. McCue, F.M. McKenna, J.A. Anderson, unpublished results.
22] A.J. McCue, J.A. Anderson, Catal. Sci. Technol. 4 (2014) 272–294.
23] J.T. Wehrli, D.J. Thomas, M.S. Wainwright, D.L. Trimm, Appl. Catal. 66 (1990)
99–208.
24] J.T. Wehrli, D.J. Thomas, M.S. Wainwright, D.L. Trimm, Appl. Catal. 70 (1991)
53–262.
[
[
[
[
[
[
1
5 (2013) 12187–12196.
65] S.A. Nikolaev, V.V. Smirnov, A. Yu Vasil’kov, V.L. Podshibikhin, Kinet. Catal. 51
2010) 375–379.
66] J. Gapipaud, M.H. Martin, L. Roué, D. Guay, J. Phys. Chem. C 117 (2013) 2688–
698.
[
[
1
(
[
2
2
[
[
25] N.J. Ossipoff, N.W. Cant, J. Catal. 148 (1994) 125–133.
26] B. Bridier, N. López, J. Pérez-Ramírez, J. Catal. 269 (2010) 80–92.