The roles of carbide and hydride in oxide-supported palladium nanoparticles for alkyne hydrogenation

Article abstract of DOI:10.1016/j.jcat.2011.06.025

Particle size affects the activity and selectivity to partial hydrogenation of 1-pentyne over oxide-supported palladium nanoparticles. Larger particles are intrinsically more selective because of the weaker bond strength of 1-pentene. In situ X-ray absorption near edge structure (XANES) at the Pd L₃ edge revealed the formation of a carbide-like phase as soon as the catalyst is exposed to alkyne, irrespective of particle size. The newly formed phase prevented hydride formation. Surface poisoning of the palladium carbide by alkyne is responsible for the constantly high selectivity, up to almost complete conversion. At almost 100% conversion, all catalysts show low selectivity. The lack of significant pentyne adsorption on the surface causes pentene to undergo consecutive reactions, such as isomerization and complete hydrogenation. The structure of the catalyst was that of carbide-like phase and did not change. Palladium hydride did not form under any of the conditions. Exposure of a carbided catalyst to pure hydrogen leads to partial reversal of the structure. Hydride is not essential for complete hydrogenation to occur.

Full text of DOI:10.1016/j.jcat.2011.06.025

Journal of Catalysis 283 (2011) 45–54
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
The roles of carbide and hydride in oxide-supported palladium nanoparticles
for alkyne hydrogenation
Min Wei Tew a, Markus Janousch b, Thomas Huthwelker b, Jeroen A. van Bokhoven a,b,
⇑
a ETH Zurich, Institute for Chemical and Bioengineering, 8093 Zurich, Switzerland
b Paul Scherrer Institute, 5232 Villigen, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Particle size affects the activity and selectivity to partial hydrogenation of 1-pentyne over oxide-sup-
ported palladium nanoparticles. Larger particles are intrinsically more selective because of the weaker
bond strength of 1-pentene. In situ X-ray absorption near edge structure (XANES) at the Pd L3 edge
revealed the formation of a carbide-like phase as soon as the catalyst is exposed to alkyne, irrespective
of particle size. The newly formed phase prevented hydride formation. Surface poisoning of the palladium
carbide by alkyne is responsible for the constantly high selectivity, up to almost complete conversion. At
almost 100% conversion, all catalysts show low selectivity. The lack of significant pentyne adsorption on
the surface causes pentene to undergo consecutive reactions, such as isomerization and complete hydro-
genation. The structure of the catalyst was that of carbide-like phase and did not change. Palladium
hydride did not form under any of the conditions. Exposure of a carbided catalyst to pure hydrogen leads
to partial reversal of the structure. Hydride is not essential for complete hydrogenation to occur.
Ó 2011 Elsevier Inc. All rights reserved.
Received 14 April 2011
Revised 28 June 2011
Accepted 30 June 2011
Available online 16 August 2011
Keywords:
Selective partial hydrogenation
1-Pentyne
Oxide-supported palladium nanoparticles
Palladium hydrides
Palladium carbide
Effect of particle size
Pd L3 edge in situ X-ray absorption near
edge structure
1. Introduction
[15,16]. In 1970s, Palczewska proposed the participation of hy-
drides in catalytic reactions and reviewed studies related to the
Chemoselective hydrogenation is a common reaction in the
chemical industry [1–4]. To achieve high yield, metal catalysts
are generally modified, for example by lead poisoning of palladium
to give a Lindlar catalyst [5,6], by adding dopants to Raney nickel
[7], and by partial poisoning of the catalyst surface. Partial hydro-
genation of alkynes is performed to remove traces of alkynes from
streams of alkenes to prevent poisoning of the catalyst during
polymerization and branching of poly-olefins [8,9]. Selective al-
kyne hydrogenation is characterized by the very strong adsorption
of the alkyne, which poisons the surface and prevents the alkene
from reacting further. The primary product from 1-alkyne is 1-al-
kene that converts into isomers and alkane in consecutive
reactions.
The structure of palladium catalysts changes readily as the con-
ditions to which the catalyst is exposed change. Palladium hy-
drides and carbides form easily, the latter of which forms after
adsorption of alkynes [10,11]. Hydride formation in palladium cat-
alysts depends on the size of the palladium particles [12–14]. Be-
cause the hydride is generally associated with unselective
hydrogenation, variation in catalytic performance is expected
catalytic reactivity of hydrogen on palladium and nickel hydride
phases [17]. She concluded that the catalytic activity of palladium
and nickel decreases when they transform into the respective hy-
dride phases. Results of the hydrogenation of acetylene over sup-
ported palladium catalysts, with or without excess ethane,
indicate that the formation of the b-hydride phase in large particles
enhanced the full hydrogenation of acetylene [4,18–23], while sur-
face hydrogen was much more selective to ethylene [24]. Similar
results were obtained for alkynes of higher molecular weight
[25–28]. Palladium particles smaller than 2 nm formed an
a phase,
which is more selective than the hydrogen-rich b phase [17,21,29–
32]. Under low pressure, hydrogenation of alkene did not occur
over a Pd(1 1 1) single crystal, because the hydrogen atoms dif-
fused so deeply into the bulk that they were inaccessible to the ad-
sorbed alkene. In contrast, alkene hydrogenation over palladium
nanoparticles under low pressure occurred, because the weak
binding of subsurface hydrogen meant that it was accessible to
the adsorbed alkene [15]. The authors also concluded that b-hy-
dride was more active and less selective to alkene than the
a phase.
The higher activity of the bulk-dissolved hydrogen than of the ad-
sorbed surface hydrogen was confirmed by theoretical calculations
[27,33,34].
During alkyne hydrogenation, adsorbed alkynes undergo hydro-
genation/dehydrogenation reactions, during which palladium car-
⇑
Corresponding author at: ETH Zurich, Institute for Chemical and Bioengineering,
HCI E 127, Wolfgang-Pauli-strasse 10, 8093 Zurich, Switzerland.
E-mail address: j.a.vanbokhoven@chem.ethz.ch (J.A. van Bokhoven).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.06.025

46
M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
bide may form [10,11]. The formation of palladium carbide during
acetylene hydrogenation at 100 °C was discussed as early as 1980s
[35]. The decomposition of acetylene on the palladium surface pro-
duced carbon atoms, which penetrated the lattice and formed a so-
lid solution of carbon in palladium. Two carbide phases, analogous
one), and X-ray absorption spectroscopy. All the catalysts had a
Pd weight loading of 1.9%, as determined by inductive couple plas-
ma mass spectrometry.
2.2. Gas-phase hydrogenation of 1-pentyne
to the
a and b hydride phases, were observed when the palladium
catalysts were exposed to acetylene at high temperature [10,36–
38]. In situ XPS has revealed the formation of a surface carbide in
bulk palladium during the selective hydrogenation of 1-pentyne
[25]. Full hydrogenation occurred over a surface with bulk-dis-
solved hydride, indicating the instability of the carbide phase over
which selective hydrogenation occurred. The formation of palla-
dium carbide was also observed during acetoxylation of ethylene
[39], selective hydrogenation of butadiene [40], and the selective
hydrogenation of propyne [27,27] and 1-pentyne [11,25,26,41].
These studies showed that the structure of a catalyst is affected
by the composition of the gas phase. Theoretical calculations
showed that the formation of carbide is energetically feasible
[27,42–44] and that carbides probably form at the steps-like sites
at near-surface positions and diffuse toward the center of the pal-
ladium particles. Penetration to the bulk of the particle and forma-
tion of dense carbide structures are energetically unfeasible [27]. It
was suggested that the carbide formed mainly in the near-surface
region of the catalyst and that it completely suppressed the
absorption of hydrogen [11,25,27] and, hence, increased selectiv-
ity. Palladium carbide has also been correlated with the loss of
activity, stabilization of supported Pd catalysts, and high selectivity
to alkene.
Despite results from different groups suggesting the formation
of the palladium carbide during reactions [11,25], only very few
characterization data of this species have been obtained under
reaction condition, especially in case of nano-sized particles. In this
work, we performed gas-phase hydrogenation of 1-pentyne over
supported palladium nanoparticles and determined the structure
of the palladium particles in situ by means of X-ray absorption
near edge structure (XANES) at the Pd L3 edge during selective
and non-selective hydrogenation. The L3 edge XANES probes the
empty d density of states; thus, palladium in different oxidation
states can be distinguished easily. Furthermore, hydrides can be
identified in the near edge when there is a Pd–H antibonding state
[14,45,46]. Because of the depth of penetration of high energy X-
rays, XAS can be implemented under actual catalytic conditions
[47]. The energy of the X-rays at the Pd L3 edge (3175 eV) is mod-
erate, which therefore required the adaptation of an in situ cell
[48]. The measurements were taken at the PHOENIX I beamline
of the Swiss Light Source. The lowest measurable palladium con-
centration in fluorescence mode was about 0.4 wt.%. The in situ
spectroscopic data obtained enabled us to determine directly the
role of hydride and carbide during selective and non-selective al-
kyne hydrogenation. We found that, irrespective of the palladium
particle size, palladium carbide forms during the selective reaction
and that, over time, the particles are completely carbided. Com-
plete and selective hydrogenation occurred over this surface.
The gas-phase hydrogenation of 1-pentyne was performed in a
continuous-flow, fixed-bed tubular quartz reactor. The tempera-
ture difference of the oven was kept constant within 1 °C. The tem-
perature at the bottom of the catalyst bed was measured with a
chromel–alumel thermocouple. All the temperatures mentioned
in the text refer to the measured temperature of the catalyst
bed; these values at most varied by a few degrees. The catalyst
was diluted with alumina or silica to give different dilution factors,
i.e., the ratio between the amount in weight of pure catalyst and
that of the respective diluents ranged from 100 to 400 to achieve
different conversion levels. Sieved fractions (45 < x < 60 mesh) of
the diluted catalyst, 0.05 g, were used in each experiment. A fresh
batch of catalyst was used in each experiment. Prior to the catalytic
reaction, the catalyst was reduced in situ at 2°/min to 150 °C and
left for 1 h in a flow of pure hydrogen (30 mL/min). The catalyst
was then cooled in pure hydrogen to the desired reaction temper-
ature. 1-Pentyne (Fluka, 99% purity) in a stream of helium was
introduced into the system and passed through a saturator at
0 °C. The 1-pentyne/helium flow was mixed with pure hydrogen
before reaching the catalyst bed and the H2/1-pentyne ratio ad-
justed. The reaction products were analyzed with an Agilent
6890A gas chromatograph (GC) equipped with an FID detector
and an HP-5 (50 m ꢀ 0.32 mm) column. Under differential condi-
tions, the reactions were performed at a constant reaction temper-
ature of 80 °C. The sample dilutions were adjusted to give a
conversion of around 10%. Conversion and selectivity were also
determined as a function of hydrogen to 1-pentyne ratio and of
increasing gas flow at 90 °C. When varying the hydrogen to 1-pen-
tyne ratio, the flow rates of hydrogen were adjusted while the flow
rates of helium, which contained the 1-pentyne, remained con-
stant. When varying the gas flow, the total flow rates of helium-
1-pentyne and of hydrogen were adjusted to maintain a H2/1-pen-
tyne ratio of 4. We did not find that the selectivity and conversion
depend on the flow rate per min, but on the space time. The reac-
tions were performed at a constant reaction temperature of 45, 75,
and 100 °C to achieve different conversion levels. To identify the
effect of increasing hydrogen partial pressure, an additional exper-
iment was performed at 45 °C at a H2/1-pentyne ratio of 10.
The TOF of the products was calculated using the following
equation:
moleculeproduct=t
TOF ðsꢁ1Þ ¼
;
ðiÞ
nSurf:Pd
where nSurf.Pd is the number of exposed palladium atoms as deter-
mined by CO chemisorption [49].
The selectivity (S) of the products was calculated using the fol-
lowing equation:
2. Experimental
Aproduct A ꢂ Wproduct
A
Sproduct
¼
ꢀ 100%;
ðiiÞ
A
molall
2.1. Synthesis and characterization of silica-and alumina-supported
palladium nanoparticles
products
where Ai is the GC peak area and Wi the calibrated mol coefficient.
Wi was 1 for 1-pentene, trans-2-pentene, cis-2-pentene, and
pentane and 0.5 for the oligomers, assuming that the oligomers
consist mainly of hydrocarbons with 10 carbon atoms (C10), as sug-
gested by the retention time in the GC. The carbon balance was
closed within the accuracy of the GC. Only at the initial stage of
the reaction did alkyne adsorb irreversibly onto the catalyst.
The catalysts were prepared by incipient-wetness impregnation
with either alumina or silica, as described elsewhere [14]. The cat-
alysts are Pd/SiO2-10.5, Pd/Al2O3-3.6, Pd/SiO2-2.8, and Pd/Al2O3-
1.3; the numbers at the end refer to the average particle size in
nanometer, determined from scanning transmission electron
microscopy, CO chemisorption (assuming a CO to Pd ratio of

M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
47
2.3. Pd L3 edge XANES
Fig. S1 (supporting information) shows the conversion of 1-pen-
tyne on Pd/SiO2-10.5 at different H2/1-pentyne ratios at constant
temperature of 90 °C. Starting from a H2/1-pentyne ratio of 8, the
conversion gradually decreased from 45% to 15% at decreasing
H2/1-pentyne ratio to 2. The corresponding selectivity to pentenes
was 67% and 98%, while the selectivity to pentane was 20% and 2%
respectively. The amount of oligomers remained below 5% regard-
less of the H2/1-pentyne ratio used. Increasing the H2/1-pentyne
ratio back to 4 and then 6 led to conversion and selectivity trend
which were similar to those values obtained during decreasing ra-
tio. Our data confirmed that hydrogenation of alkyne usually
shows a reaction order of between 0.5 and 1.5 in hydrogen and
zero or even negative in alkyne [1,52–54]. Hydrogen adsorbs
weakly compared to the alkyne, and the rate determining step is
normally the addition of hydrogen to an adsorbed intermediate
[55,56]. The hydrogen coverage is always very low and virtually
the complete conversion range is the surface covered with alkyne.
The structure of Pd/SiO2-10.5 and Pd/SiO2-2.8 during selective
and non-selective hydrogenation of 1-pentyne was determined
by means of Pd L3 edge XANES. The measurements were taken at
the X07MB (PHOENIX I) beamline of the Swiss Light Source with
a double crystal Si(1 1 1) monochromator to select the energy of
the incoming beam. High-order harmonics and thermal load of
the monochromator crystals were reduced with a set of two plane
rotatable mirrors. The fluorescent X-ray was detected by means of
a silicon drift detector with an energy resolution of ꢃ130 eV,
placed at 90° to the incoming beam. The sample made an angle
of 45° to both, beam and fluorescence detector. A very thin poly-
mer window was attached to the detector to enable detection of
the fluorescence of elements down to the carbon K edge [50].
The determination of each XANES spectrum took about 15 min.
Two mirrors in a Kirkpatrick–Baez configuration provided a focal
spot size of about 10
incoming intensity of the beam (I0) was monitored by a thin poly-
l
m ꢀ 5
l
m
(horizontal ꢀ vertical). The
3.2. Effect of temperature and conversion on activity and selectivity
ethylene foil covered by a 30-nm-thick nickel layer.
An aluminum plug-flow reactor, which can be heated to 400 °C,
was used [48]. The reactor was placed in a stainless steel chamber
which was vacuumed and refilled multiple times with nitrogen
and helium to minimize the disturbance of air which could
strongly affect the quality of the data. The chamber was eventually
filled with 800 mbar of helium. The inlet and outlet of the reactor
were fed through the chamber by Swagelok connections. The cata-
lyst was diluted with silica (dilution factors from 2 to 5) to control
the conversion. Sieved fractions (45 < x < 60 mesh) of the catalyst
were packed in the plug-flow reactor and covered by a kapton film,
0.008 mm thick. Prior to the reaction, 4 mg catalyst was reduced
in situ at 150 °C at 2°/min, left for 1 h, and cooled to the desired
reaction temperature in pure hydrogen (30 mL/min). Spectra of
the catalyst in pure hydrogen were collected at 150 °C and then
at 40 and 100 °C respectively. 1-Pentyne (Fluka, 99% purity) was
introduced into the system in a stream of helium, which passes
through a saturator at 0 °C. The 1-pentyne/helium flow was mixed
with pure hydrogen before reaching the catalyst bed, and the H2/1-
pentyne ratio was adjusted to 4. With the GC, conversion and
selectivity were determined as functions of increasing gas flow in
mL/min. XANES spectra of each flow were recorded. The reactions
were performed at a constant reaction temperature of 40 or 100 °C.
Measurements taken at the middle and at the bottom of the reactor
yielded identical results. The XAS data were processed according to
standard procedures using the XDAP software package (Version
2.3) [51].
Fig. 2a–d show the conversion and selectivity to pentenes, pen-
tane, and oligomers over Pd/SiO2-10.5 at 100 °C. Different levels of
conversion were obtained by varying the flow and the dilution fac-
tor. The dilution factor means the ratio between the amounts in
weight of pure catalyst and that of the respective diluents. At
10–30% conversion (Fig. 2a), the system showed about 92% selec-
tivity to pentenes, 5% to oligomers, and 3% to pentane. The pent-
enes consisted of only 1-pentene (Fig. 2b). At almost 100%
conversion, there was consecutive 1-pentene isomerization to
cis- and trans-2-pentenes as well as full hydrogenation. The selec-
tivity to 1-pentene eventually decreased to 0% (Fig. 2c and d).
Fig. 2e and f give the data recorded at 45 °C. The conversion was
below 5%, and the reaction was 100% selective to 1-pentene. The
very low conversion made it difficult to detect small amount of
products. Oligomer formation was lower than 5% in all cases.
Fig. 3a and b give the reaction data under the same conditions as
in Fig. 2e and f, with the exception that the H2/1-pentyne ratio
was 10. Increasing the hydrogen pressure led to twice as much
conversion, but the selectivity did not change.
Fig. 4a–d gives the conversion and product selectivity of Pd/
Al2O3-1.3 as a function of increasing gas flow at 100 °C. At conver-
sion between 15 and 84% (Fig. 4a and b), the system showed 60 to
80% selectivity to pentenes, consisted mainly of 1-pentene. At 98%
conversion, the selectivity to pentenes dropped to 48%; at >99%
conversion, full hydrogenation occurred. At these high conversion
levels, 1-pentene also converted to trans-2-pentene (23%) and
cis-2-pentene (9%). At 100% conversion (Fig. 4c and d), full hydro-
genation dominated (93% selectivity to pentane, 6% to oligomers,
and 1% to pentenes). Fig. 4e and f gives the data of the reaction
at 45 °C. The conversion was below 5%, and selectivity was about
82% to 1-pentene, 7% to pentane, and 11% to oligomers.
3. Results
3.1. Activity and selectivity under different conditions
Based on the data in Figs. 2–4, selectivity versus conversion
plots were constructed. Fig. 5a–d give the selectivity versus con-
version plots of the other catalysts. Selectivity to 1-pentene,
trans-2-pentene, and cis-2-pentene is referred to as selectivity to
pentenes. Except for conversion below about 5%, all the catalysts
showed constant selectivity to pentenes and pentane with near
100% conversion. Only at 100% conversion did full hydrogenation
dominate and the selectivity to pentenes decreased to 0%. More-
over, under these conditions, 1-pentene isomerized. The selectivity
to pentenes over Pd/SiO2-10.5 was 90% compared to 80% over Pd/
Al2O-1.3. An increase of 10% in selectivity is extensive at such high
level of selectivity. These plots indicate that selectivity is deter-
mined by the level of conversion and tends to be independent of
temperature and that particle size affects the intrinsic selectivity
to partial hydrogenation.
Fig. 1a shows the conversion under constant temperature of
80 °C. Nearly a constant conversion of 12% was obtained for Pd/
SiO2-10.5 and Pd/SiO2-2.8 throughout the reaction. Both alumina-
supported catalysts, particularly Pd/Al2O3-3.6, showed high activ-
ity at the beginning of the reaction and became deactivated over
time. The reactivity of each catalyst was compared by means of
TOF based on the number of adsorbed CO molecules (Fig. 1b). Pd/
SiO2-10.5, with the largest particle size, converted six times more
pentyne than Pd/Al2O3-1.3 with the smallest particles and twice
as much as Pd/SiO2-2.8 per site per second. Due to the strong deac-
tivation of Pd/Al2O3-3.6, it is not discussed. Fig. 1c–e show the cor-
responding selectivity of each catalyst to pentenes, pentane, and
oligomers. Pd/SiO2-10.5 showed a constantly higher selectivity to
pentenes, while Pd/Al2O3-1.3 showed the lowest.

48
M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
20
15
10
5
60
50
40
30
20
10
0
(a)
(b)
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
time, mins
time, mins
100
98
96
94
92
90
88
10
8
(c)
(d)
6
4
2
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
time, mins
time, mins
10
8
(e)
6
4
2
0
0
50
100
150
200
250
300
time, mins
Fig. 1. (a) Conversion of 1-pentyne over ( ) Pd/SiO2-10.5, ( ) Pd/Al2O3-3.6, ( ) Pd/SiO2-2.8, and ( ) Pd/Al2O3-1.3 under differential conditions at 80 °C. (b) The
corresponding TOF. Corresponding selectivity to (c) pentenes, to (d) pentane, and to (e) oligomers.
3.3. Pd L3 edge XANES of reference samples
low and selectivity to pentenes was high. The collected spectra,
therefore, correspond to the catalyst structure in the selective re-
gion. For the experiments at 100 °C, both catalysts show high con-
version and high selectivity to full hydrogenation (pentane). The
spectra, therefore, correspond to the catalyst structure in the
non-selective region (vide infra).
Figs. 7 and 8 show the spectra of Pd/SiO2-2.8 and Pd/SiO2-10.5
collected at 40 °C. The formation of hydride of both catalysts in
H2 was clearly indicated by the Pd–H antibonding state in the spec-
tra. Upon exposure to the reactant mixture, the Pd–H antibonding
state of Pd/SiO2-10.5 and Pd/SiO2-2.8 rapidly disappeared, indicat-
ing complete consumption of bulk hydrogen (Figs. 7 and 8, blue).
At the same time, broadening of the whiteline occurred, indicating
formation of a carbide-like phase. When the gas flow was increased
to 57.1 mL/min, the spectra of both catalysts (Figs. 7 and 8, green)
resembled that of the carbide-like phase (Fig. 6, pink): the white-
line became broader and intensity increased. These experiments
indicate that formation of a carbide-like phase occurs in the selec-
tive region for both catalysts.
Fig. 6a shows the Pd L3 edge XANES spectra of palladium
nanoparticles, palladium hydride, and palladium carbide. The
Pd L3 edge XANES spectrum of the bare particles of Pd/SiO2-
10.5 was measured at 300 °C in He (Fig. 6, black [14]). It is char-
acterized by an absorption edge at around 3175 eV together with
an intense feature on the absorption edge, known as the white-
line. The typical hydride spectrum was obtained by exposing
Pd/SiO2-10.5 to pure hydrogen at 40 °C (Fig. 6, red). The edge
shifts to higher energy, the intensity of the whiteline decreases,
a new Pd–H antibonding state forms at around 3180 eV, and a
change in the intensity at above 3185 eV occurs [14,45,46]. The
spectrum of palladium carbide was measured under pure 1-pen-
tyne, which leads to the formation of carbide (Fig. 6, pink) [10].
The edge shifts to higher energy and is very close to that of the
hydride. Furthermore, the whiteline is significantly broader, and
the intensity is between that of the bare particle and the hydride.
Unlike the spectrum of hydride, no clear antibonding state was
observed.
Figs. 9 and 10 show the spectra of Pd/SiO2-10.5 and of Pd/SiO2-
2.8 at 100 °C. The formation of hydride of both catalysts was also
found at 100 °C in H2 (Figs. 9 and 10, red), although to lesser extent
than at 40 °C. For both catalysts, the Pd–H antibonding state disap-
peared as soon as the system was exposed to the reactant mixture
at 24.5 mL/min (Figs. 9 and 10, blue). Again for both catalysts,
3.3.1. In situ Pd L3 edge XANES
Table 1 lists the conversion and selectivity during the Pd L3 edge
experiments at 40 and 100 °C. The trends were similar to the pre-
vious catalytic data. For the experiments at 40 °C, conversion was

M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
49
100
80
1.0
0.8
0.6
0.4
0.2
0.0
100
80
60
40
20
0
20
15
10
60
(a)
(b)
40
5
20
0
0
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
100
80
60
40
20
0
30
25
20
15
10
5
(c)
(d)
0
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
100
80
60
40
20
0
100
80
60
40
20
0
5
4
(e)
3
(f)
2
1
0
23 28 33 38 43 48 53 58 63 68
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
Flow, mL/min
Fig. 2. Conversion (N) and selectivity of pentenes ( ), pentane ( ), oligomers ( ), 1-pentene ( ), t-2-pentene ( ), and c-2-pentene ( ) as a function of total flow of Pd/SiO2-
10.5 at 100 °C and dilution factors of 100 (a and b) and 10 (c and d), and at 45 °C and a dilution factor of 100 (e and f).
100
100
80
60
40
20
0
6
5
4
3
2
1
0
80
60
40
20
0
(a)
(b)
30
40
50
60
70
80
90
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
Flow, mL/min
Fig. 3. (a and b) Conversion (N) and selectivity of pentenes ( ), pentane ( ), oligomers ( ), 1-pentenes ( ), t-2-pentene ( ), and c-2-pentene ( ) as a function of total flow
of Pd/SiO2-10.5 at 45 °C and a dilution factor of 100 and a H2/pentyne ratio of 10.
broadening of the whiteline and an increase in intensity were
observed when the gas flow increased to 40.8 mL/min for Pd/
SiO2-2.8 or to 57.1 mL/min for Pd/SiO2-10.5 (Figs. 9 and 10, green),
which suggest the formation of a carbide-like phase. The formation
of such phase occurs at 100 °C even in the presence of excess
hydrogen at full conversion. These results were not affected by
where the spectra were recorded, i.e., at the middle or bottom of
the reactor. For Pd/SiO2-10.5, the spectra collected after exposure
to excess hydrogen at a hydrogen/1-pentyne ratio of 10 at 100 °C
(gas flow = 42.5 mL/min) and to only 1-pentyne at 100 °C caused
no significant difference in the spectra. Instead, the whiteline re-
mained shifted and broader (Fig. 9, orange and pink), along with
an increase in the intensity, indicating the stable presence of the
carbide-like phase. For both catalysts, reexposure to pure hydrogen

50
M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
100
80
60
40
20
0
100
(b)
80
(a)
60
40
20
0
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
100
80
60
40
20
0
30
20
10
0
(c)
(d)
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
100
80
60
40
20
0
20
100
80
60
40
20
0
15
(e)
10
5
(f)
0
23 28 33 38 43 48 53 58 63 68
23 28 33 38 43 48 53 58 63 68
Flow, mL/min
Flow, mL/min
Fig. 4. Conversion (N) and selectivity of pentenes ( ), pentane ( ), oligomers ( ), 1-pentenes ( ), t-2-pentene ( ), and c-2-pentene ( ) as a function of total flow of Pd/
Al2O3-1.3 at 100 °C at dilution factors of 300 (a and b) and 50 (c and d), and at 45 °C at a dilution factor of 100 (e and f).
(Figs. 9 and 10, cyan) caused the weak reappearance of the Pd–H
antibonding feature, indicative of hydride. Its intensity remained
low, indicating that only some of the carbide-like phase was con-
verted into hydride.
small amount of pentenes, which did not desorb from the surface
but underwent direct consecutive hydrogenation.
4.2. Reaction mechanism and effect of particle size
During the hydrogenation of alkynes, they adsorb strongly on
the catalyst surface. The first formed olefin may desorb or undergo
consecutive reactions, such as hydrogenation and isomerization
(Fig. 11) [3]. Competition between desorption of the pentene and
the consecutive reactions to pentane determine the intrinsic selec-
tivity of the catalyst. High selectivity to pentene is expected when
k4, k5, and k6, which refer to the extent, to which pentane and olig-
omers form from 1-pentyne but without the formation of desorbed
pentene. When k1/kꢁ1 ꢄ k3/kꢁ3, strong and preferential adsorption
of 1-pentyne poisons the surface, effectively displacing the pen-
tene and inhibiting its re-adsorption. Only when the concentration
of 1-pentyne is low, can the pentene undergo consecutive reactions
such as isomerization (k6) and full hydrogenation (k5) (Figs. 2–5).
The intrinsic selectivity to partial hydrogenation in the selective
regime was affected by the particle size of palladium. Large parti-
cles (Pd/SiO2-10.5) were intrinsically more selective to partial
hydrogenation. Smaller particles were more selective to the forma-
tion of pentane and oligomers. DFT calculations gave the adsorp-
tion energy of acetylene and ethylene on various transition metal
surfaces; they uniquely correlate with the energy of the carbon-
4. Discussion
4.1. Selectivity versus conversion
Selectivity of hydrogenation of alkynes over palladium-based
catalysts is generally attributed to the thermodynamic effect: al-
kynes adsorb stronger than the corresponding alkenes [57–60].
The selectivity of the reaction is essentially determined by the rel-
ative surface coverage of adsorbed alkyne respectively alkene on
the catalyst. A sharp decrease in selectivity occurred only at close
to full conversion of the alkyne, because of the ability of alkene
to re-adsorb [1]. The low alkyne coverage enables the fast reaction
of alkene, such as in isomerization and full hydrogenation, which
are faster than alkyne hydrogenation [1,8,61]. During complete
hydrogenation over palladium black, the subsurface carbon under-
goes hydrogenation [11,25]. In contrast, the structure of the nano-
particles did not change during the transition from the selective to
the unselective regime. The constant and small amount of pentane
at all conversion levels probably originated from the presence of a

M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
51
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
(b)
pentenes
(filled)
(a)
pentenes
(filled)
pentane
(unfilled)
pentane
(unfilled)
0
20
40
60
80
100
0
20
40
60
80
100
Conv. (%)
conv. (%)
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
(c)
(d)
pentenes
(filled)
pentenes
(filled)
pentane
(unfilled
pentane
(unfilled)
0
20
40
60
80
100
0
20
40
60
80
100
conv (%)
conv. (%)
Fig. 5. Selectivity to pentenes and pentane versus conversion plot of: (a) Pd/SiO2-10.5, (b) Pd/Al2O3-3.6, (c) Pd/SiO2-2.8, and (d) Pd/Al2O3-1.3. Selectivity versus conversion
plots include data from experiments performed at a reaction temperature of 45 °C (d),75 °C (ꢀ), or 100 °C (N).
2.0
Table 1
Kinetic data for the Pd L3 edge experiments at 40 and 100 °C.
1.5
Pd/SiO2-10.5
Pd/SiO2-2.8
(A)
(B)
(C)
(D)
40 °C
100 °C
40 °C
100 °C
1.0
0.5
0.0
24.5 mL/min
40.8 mL/min
57.1 mL/min
Conversion (%)
Selectivity (%)
Pentane
Pentene
Oligomers
100
33
100
87
10
3
5
80
15
93
0
7
Conversion (%)
Selectivity (%)
Pentane
Pentene
Oligomers
12
95
100
0
100
0
64
31
5
80
13
7
3165 3170 3175 3180 3185 3190 3195
Energy, eV
Conversion (%)
Selectivity (%)
Pentane
Pentene
Oligomers
10
93
12
Fig. 6. Pd L3 edge spectra of Pd/SiO2-10.5, at 300 °C in He (black, [14]) and at 40 °C
in H2 (red) and in 1-pentyne (pink). (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)
0
100
0
56
38
6
0
83
17
surface bond, as determined from the adsorption energy of methyl
groups [57]. An energy range for metals and alloys with an optimal
combination of high activity and selectivity was proposed. Surfaces
in this energy range were predicted to show weak binding with
ethylene but strong binding with acetylene, thus enabling partial
hydrogenation. Both ethylene and acetylene showed weaker bond-
ing on palladium with subsurface carbon and hydrogen than on
clean palladium, and such catalysts are generally more selective.
Nevertheless, palladium with subsurface hydrogen had a signifi-
cantly weaker effect on the selectivity and reactivity on the palla-
dium surface than on the surface with subsurface carbon. Selective
hydrogenation occurred on a carbided surface, showing that the
strong pentyne bonding is responsible for the selective hydrogena-
tion. We found that the larger particles showed higher selectivity.
Larger particles may adsorb more weakly as long as the surface is
covered with alkynes; the weaker adsorption of alkenes has a
stronger effect and, thus a higher selectivity. Weaker adsorption
of poisonous species may enhance the rate. In addition to a differ-
ent adsorption strength, the selectivity to pentenes is influenced by
the relative amount of planes and edges and corner atoms and the
kind of exposed surface of the palladium particles. Big particles
have larger extended planes than small particles, which have more
edge and corner sites. Terminal alkyne can adsorb on the surface in
a parallel or perpendicular manner, while internal alkyne can only
bond in a parallel state, assuming facile removal of the alkynic
hydrogen [62]. Large particles thus favor 2-pentyne hydrogenation
while for 1-pentyne hydrogenation, the differences were less sig-
nificant. In terms of activity, our TOF results agree with those ob-
tained for the hydrogenation of 1-butyne over oxide-supported
palladium catalysts, where the reaction activity decreased as the
particle size decreased [63,64].

52
M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
2.0
1.5
1.0
0.5
0.0
2.5
2.0
1.5
1.0
0.5
0.0
3165 3170 3175 3180 3185 3190 3195
3165 3170 3175 3180 3185 3190 3195
Energy, eV
Energy, eV
Fig. 7. Pd L3 edge spectra of Pd/SiO2-10.5 at 40 °C in H2 (red), during reaction at
40.8 mL/min (blue) and 57.1 mL/min (green). The spectra correspond to kinetic data
in Table 1 (A). (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
Fig. 10. Pd L3 edge spectra of Pd/SiO2 -2.8 at 100 °C in H2 (red), during reaction at
24.5 mL/min (blue) and 40.8 mL/min (green). Reexposed to H2 at 100 °C (cyan). The
spectra correspond to kinetic data in Table
1 (B). (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of
this article.)
2.0
1.5
1.0
0.5
Oligomers (C10),
Isomerisation
PdCx
k6
H2
k4
H
2
k ,
2
H
5
k ,
2
1-pentyne (ads)
Pentenes (ads)
Pentane (ads)
k-2
k1
k-1
k3
Pentenes (g)
k-3
k7 k-7
1-pentyne (g)
Pentane (g)
0.0
3165 3170 3175 3180 3185 3190 3195
Fig. 11. Reaction mechanism during hydrogenation of alkynes [26].
Energy, eV
Fig. 8. Pd L3 edge spectra of Pd/SiO2-2.8 at 40 °C in H2 (red), during reaction at
24.5 mL/min (blue) and 57.1 mL/min (green). The spectra correspond to kinetic data
in Table 1 (C). (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
4.3. The structure of palladium during reaction
The results of this study indicated that the exposure of fully re-
duced catalysts to hydrogen always produced hydrides [14,45,46].
As soon as the catalyst was exposed to 1-pentyne, the hydrides
transformed into a disordered, carbide-like phase, even in the pres-
ence of excess hydrogen. We only observed the formation of the
carbide-like phase after the exposing the catalyst to pentyne, not
pentene. Scheme 1 describes the structures that form after consec-
utive exposure of the catalysts to different gas mixtures. Alkynes
bond very strongly to the palladium surface and readily undergo
hydrogenation and dehydrogenation, which gradually fully trans-
form the palladium particles into a palladium carbide-like phase,
probably starting at the (sub) surface. Reexposure of the carbide-
like phase to hydrogen for 1 h (Figs. 9 and 10) restores the hydride
structure only to a small extent, indicating the stability of the
nano-sized carbide-like phase, as opposed to the instability of sub-
surface carbide in massive palladium [11,25]. The presence of
hydrogen does not hinder the dehydrogenation of alkyne, which
enables the incorporation of carbon into the palladium structure;
under 1-pentyne and hydrogen, the carbide-like phase also forms.
Although it takes some time for the particles to be carbided, selec-
tive hydrogen is found on fully occurs over particles with the car-
bide-like phase. The structure of our nanoparticles did not change
when changing to the unselective regime. Even though the pres-
ence of hydride is often associated with complete hydrogenation,
it is not essential for complete hydrogenation to occur.
2.5
2.0
1.5
1.0
0.5
0.0
3165 3170 3175 3180 3185 3190 3195
Energy, eV
Fig. 9. Pd L3 edge spectra of Pd/SiO2-10.5 at 100 °C in H2 (red), during reaction at
24.5 mL/min (blue) and 57.1 mL/min (green). In excess hydrogen at a H2/-pentyne
ratio of 10 (orange) at 100 °C. In 1-pentyne at 100 °C (pink) and reexposed to H2 at
40 °C (cyan). The spectra correspond to kinetic data in Table 1 (D). (For interpre-
tation of the references to color in this figure legend, the reader is referred to the
web version of this article.)

M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
53
Palladium
Hydride
(a)
Surface hydrogen
Carbon
Support
High temperature, inert
1-pentyne
(c)
(b)
During reaction below 100% conversion (selective)
≤ 100 oC in hydrogen
(d)
During reaction at full conversion (non-selective)
Scheme 1. Structures that form after consecutive exposure of the catalysts to different gas mixtures: (a) bare palladium nanoparticles, (b) formation of surface hydrogen and
hydride under hydrogen, (c) formation of palladium carbide-like phase during the selective hydrogenation, and (d) arbide during non-selective hydrogenation.
5. Conclusions
Acknowledgments
During the hydrogenation of 1-pentyne, the alkyne poisons the
surface of palladium nanoparticles and effectively displaces the
pentene to desorb from the surface. The inhibition of the reaction
of pentenes leads to suppressed conversion and high selectivity,
the latter of which depends on particle size; the larger particles
are more selective and have a higher TOF. The intrinsic selectivity
depends on the relative bond strength of the alkynes and alkenes.
Pentene isomerization and full hydrogenation dominate when the
concentration of 1-pentyne is low enough for pentene to undergo
secondary reactions. This occurs at high conversion. Alkyne under-
goes hydrogenation and dehydrogenation, which gradually trans-
form the palladium particles into a palladium carbide-like phase.
Surface poisoning by alkyne of the palladium carbide-like phase
is responsible for the high selectivity below 95% conversion. The
nano-sized carbided palladium particles did not change in the
non-selective regime, and full hydrogenation occurred over the
carbided particles. Exposure of the carbided catalyst to pure hydro-
gen leads to subtle changes in the structure. Hydride was not de-
tected during reaction. Full hydrogenation can also occur in the
absence of hydride.
The authors acknowledge the Swiss National Science Founda-
tion (SNF) for financial support. The Pd L3 edge experiments were
performed on the X07MB (PHOENIX I) beamline at the Swiss Light
Source, Paul Scherrer Institute, Villigen, Switzerland. The authors
thank Dr. Jeffrey T. Miller of the Chemical Technology Division of
the Argonne National Laboratory, United States for the samples,
the workshop of the ETH Zurich for constructing the in situ plug-
flow reactor, and Erich DeBoni and Beat Meier of the Swiss Light
Source for optimizing the plug-flow reactor.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcat.2011.06.025.
References
[1] G.C. Bond, Metal-Catalysed Reactions of Hydrocarbons, Springer, New York,
2005.

54
M.W. Tew et al. / Journal of Catalysis 283 (2011) 45–54
[2] S.A. Nikolaev, L.N. Zanaveskin, V.V. Smirnov, V.A. Averyanov, K.L. Zanaveskin,
[35] A. Frackiewicz, A. Janko, Acta Crystallogr. Sect. A 34 (1978) S377.
[36] J. Stachurski, A. Frackiewicz, J. Less-Common Metals 108 (1985) 249.
[37] S.B. Ziemecki, G.A. Jones, J. Catal. 95 (1985) 621.
[38] J.A. Mccaulley, Phys. Rev. B: Condens. Matter Mater. Phys. 47 (1993) 4873.
[39] A.H. Zaidi, Appl. Catal. 30 (1987) 131.
Russ. Chem. Rev. 78 (2009) 231.
[3] Á. Molnár, A. Sárkány, M. Varga, J. Mol. Catal. A: Chem. 173 (2001) 185.
[4] A. Borodzin´ ski, G.C. Bond, Catal. Rev. Sci. Eng. 48 (2006) 91.
[5] H. Lindlar, Helv. Chim. Acta 35 (1952) 446.
[6] H. Lindlar, R. Dubuis, Org. Synth. 5 (1973) 880.
[7] L.W. Covert, R. Connor, H. Adkins, J. Am. Chem. Soc. 54 (1932) 1651.
[8] A.N.R. Bos, K.R. Westerterp, Chem. Eng. Process. 32 (1993) 1.
[40] T. Ouchaib, J. Massardier, A. Renouprez, J. Catal. 119 (1989) 517.
[41] D. Teschner, Z. Revay, J. Borsodi, M. Havecker, A. Knop-Gericke, R. Schlogl, D.
Milroy, S.D. Jackson, D. Torres, P. Sautet, Angew. Chem. Int. Ed. 47 (2008) 9274.
[42] N. Seriani, F. Mittendorfer, G. Kresse, J. Chem. Phys. 132 (2010) 024711.
[43] P. Sautet, F. Cinquini, Chemcatchem 2 (2010) 636.
[44] I.V. Yudanov, K.M. Neyman, N. Rosch, Phys. Chem. Chem. Phys. 6 (2004) 116.
[45] A.V. Soldatov, S. Dellalonga, A. Bianconi, Solid State Commun. 85 (1993) 863.
[46] T. Kubota, Y. Kitajima, K. Asakura, Y. Iwasawa, Bull. Chem. Soc. Jpn. 72 (1999)
673.
[47] J.A. van Bokhoven, T. Ressler, F.M.F. De Groot, G. Knop-Gericke, in: B.M.
Weckhuysen (Ed.), In situ Spectroscopy of Catalysts, American Scientific
Publishers, 2004, p. 123.
[48] N. Weiher, E. Bus, B. Gorzolnik, M. Moller, R. Prins, J.A. van Bokhoven, J.
Synchrotron Radiat. 12 (2005) 675.
´
[9] A. Borodzinski, Catal. Lett. 63 (1999) 35.
[10] S.B. Ziemecki, G.A. Jones, D.G. Swartzfager, R.L. Harlow, J. Am. Chem. Soc. 107
(1985) 4547.
[11] D. Teschner, E. Vass, M. Hävecker, S. Zafeiratos, P. Schnörch, H. Sauer, A. Knop-
Gericke, R. Schlögl, M. Chamam, A. Wootsch, A.S. Canning, J.J. Gamman, S.D.
Jackson, J. McGregor, L.F. Gladden, J. Catal. 242 (2006) 26.
[12] P.C. Aben, J. Catal. 10 (1968) 224.
[13] M. Boudart, Adv. Catal. 20 (1969) 153.
[14] M.W. Tew, J.T. Miller, J.A. van Bokhoven, J. Phys. Chem. C 113 (2009) 15140.
[15] A.M. Doyle, S.K. Shaikhutdinov, S.D. Jackson, H.J. Freund, Angew. Chem. Int. Ed.
42 (2003) 5240.
[16] W. Palczewska, in: Z. Paal, P.G. Menon (Eds.), Hydrogen Effects in Catalysis,
Marcel Dekker, New York, 1988, p. 373.
[17] W. Palczewska, Adv. Catal. 24 (1975) 245.
[18] A. Borodzin´ ski, R. Dus´, R. Frak, A. Janko, W. Palczewska, in: G.C. Bond, P.B.
Wells, F.C. Tompkins (Eds.), Proceedings of the 6th International Congress on
Catalysis, London, July 12–16, 1976, Chemical Society, London, 1977.
[49] A. Borodzin´ ski, Langmuir 13 (1997) 5613.
[50] A.M. Flank, G. Cauchon, P. Lagarde, S. Bac, M. Janousch, R. Wetter, J.M.
Dubuisson, F.M. Idir, T. Langlois, D. Moreno, D. Vantelon, Nucl. Instrum.
Methods Phys. Res. Sect. B 246 (2006) 269.
[51] D.C. Koningsberger, B.L. Mojet, G.E. van Dorssen, D.E. Ramaker, Top. Catal. 10
(2000) 143.
´
[19] A. Borodzinski, A. Janko, React. Kinet. Catal. Lett. 7 (1977) 163.
[52] C.A. Hamilton, S.D. Jackson, G.J. Kelly, R. Spence, D. de Bruin, Appl. Catal. A 237
(2002) 201.
[53] L. Brandaõ, D. Fritsch, A.M. Mendes, L.M. Madeira, Ind. Eng. Chem. Res. 46
(2007) 377.
[54] G.C. Bond, D.A. Dowden, N. Mackenzie, Trans. Faraday Soc. 54 (1958) 1537.
[55] G. Webb, in: C.H. Bamford, C.F.H. Tipper (Eds.), Comprehensive Chemical
Kinetics, vol. 21, Elsevier, Amsterdam, 1978, p. 1.
[56] S. Bailey, F. King, in: R.A. Sheldon, H. van Bekkum (Eds.), Fine Chemicals
through Heterogeneous Catalysis, Wiley, New York, 2001, p. 351.
[57] F. Studt, F. Abild-Pedersen, T. Bligaard, R. Sørensen, C. Christensen, J. Nørskov,
Angew. Chem. Int. Ed. 120 (2008) 9439.
[58] J.B. Conn, G.B. Kistiakowsky, E.A. Smith, J. Am. Chem. Soc. 61 (1939) 1868.
[59] G.F. Taylor, S.J. Thomson, G. Webb, J. Catal. 12 (1968) 150.
[60] X.C. Guo, R.J. Madix, J. Catal. 155 (1995) 336.
[61] A.N.R. Bos, E.S. Bootsma, F. Foeth, H.W.J. Sleyster, K.R. Westerterp, Chem. Eng.
Process. 32 (1993) 53.
[20] G. Carturan, G. Facchin, G. Cocco, S. Enzo, G. Navazio, J. Catal. 76 (1982) 405.
[21] C.E. Gigola, H.R. Aduriz, P. Bodnariuk, Appl. Catal. 27 (1986) 133.
[22] H.R. Aduriz, P. Bodnariuk, M. Dennehy, C.E. Gigola, Appl. Catal. 58 (1990) 227.
[23] Q.W. Zhang, J. Li, X.X. Liu, Q.M. Zhu, Appl. Catal. A 197 (2000) 221.
[24] N.A. Khan, S. Shaikhutdinov, H.J. Freund, Catal. Lett. 108 (2006) 159.
[25] D. Teschner, J. Borsodi, A. Wootsch, Z. Révay, M. Hävecker, A. Knop-Gericke,
S.D. Jackson, R. Schlögl, Science 320 (2008) 86.
[26] D. Teschner, J. Borsodi, Z. Kis, L. Szentmiklosi, Z. Revay, A. Knop-Gericke, R.
Schlogl, D. Torres, P. Sautet, J. Phys. Chem. C 114 (2010) 2293.
[27] M. Garcia-Mota, B. Bridier, J. Perez-Ramirez, N. Lopez, J. Catal. 273 (2010) 92.
[28] B. Bridier, N. Lopez, J. Perez-Ramirez, Dalton Trans. 39 (2010) 8412.
[29] G.C. Bond, P.B. Wells, The poisoning of a palladium catalyst by mercury vapor,
in: Proceedings of 2nd International Congress on Catalysis, Technip, Paris,
1960, p. 1159.
[30] A. Sarkany, A.H. Weiss, L. Guczi, J. Catal. 98 (1986) 550.
[31] H. Arnold, F. Dobert, J. Gaube, in: G. Ertl, H. Knözinger, J. Weitkamp (Eds.),
Handbook of Heterogenous Catalysis, vol. 5, VCH, Germany, 1997, p. 3266.
[32] W.J. Kim, J.H. Kang, I.Y. Ahn, S.H. Moon, J. Catal. 226 (2004) 226.
[33] A. Michaelides, P. Hu, A. Alavi, J. Chem. Phys. 111 (1999) 1343.
[34] V. Ledentu, W. Dong, P. Sautet, J. Am. Chem. Soc. 122 (2000) 1796.
[62] S.D. Jackson, C.A. Hamilton, G.J. Kelly, D. de Bruin, React. Kinet. Catal. Lett. 73
(2001) 77.
[63] J.P. Boitiaux, J. Cosyns, S. Vasudevan, Appl. Catal. 6 (1983) 41.
[64] J.P. Boitiaux, J. Cosyns, S. Vasudevan, Appl. Catal. 15 (1985) 317.