A high performance multi-walled carbon nanotube-supported palladium catalyst in selective hydrogenation of acetylene-ethylene mixtures

Article abstract of DOI:10.1016/j.apcata.2011.03.055

In this research, a palladium nanocatalyst was synthesised over a multi-walled carbon nanotube (MWCNT) support and then applied for selective hydrogenation of acetylene in an ethylene-richflow stream. This material displayed a very promising selectivity toward ethylene production with increasing temperature, and also suppressed oligomer formation during acetylene hydrogenation. New operating conditions for selective hydrogenation of acetylene in an ethylene-richflow were introduced. This nanocatalyst gave a considerably higher yield, as high as 93%, than that previously obtained for ethylene production. It was postulated that the governing mechanism for acetylene hydrogenation over 0.5 wt.% Pd/MWCNT was hydrogen transfer. Crown Copyright

Full text of DOI:10.1016/j.apcata.2011.03.055

Applied Catalysis A: General 399 (2011) 184–190
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
A high performance multi-walled carbon nanotube-supported palladium catalyst
in selective hydrogenation of acetylene-ethylene mixtures
H. Bazzazzadegan , M. Kazemeini , A.M. Rashidib
a
a,∗
a
Department of Chemical and Petroleum Engineering, Sharif University of Technology, Azadi Ave, Tehran 113659465, Islamic Republic of Iran
Nanotechnology Research Centre, Research Institute of Petroleum Industry (RIPI), Tehran, Islamic Republic of Iran
b
a r t i c l e i n f o
a b s t r a c t
Article history:
In this research, a palladium nanocatalyst was synthesised over a multi-walled carbon nanotube
Received 3 January 2011
Received in revised form 16 March 2011
Accepted 30 March 2011
Available online 7 April 2011
(
MWCNT) support and then applied for selective hydrogenation of acetylene in an ethylene-richflow
stream. This material displayed a very promising selectivity toward ethylene production with increasing
temperature, and also suppressed oligomer formation during acetylene hydrogenation. New operat-
ing conditions for selective hydrogenation of acetylene in an ethylene-richflow were introduced. This
nanocatalyst gave a considerably higher yield, as high as 93%, than that previously obtained for ethylene
production. It was postulated that the governing mechanism for acetylene hydrogenation over 0.5 wt.%
Pd/MWCNT was hydrogen transfer.
Keywords:
Selective hydrogenation
Ethylene yield
Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
Nanocatalyst
Multi-walled carbon nanotube
1
. Introduction
Ethylene is industrially produced by thermal or catalytic crack-
It is clearly known that the properties of fresh Pd catalysts are
strongly modified during their stabilisation in reactions of acety-
lene hydrogenation. It has been shown by various researchers
[7–13] that during conditioning of the fresh catalyst, by deposition
of carbonaceous species on the active surfaces of Pd, considerable
steric hindrance for absorbing ethylene molecules imposed. On the
other hand, a previous study has emphasised that by creating small
active sites within carbonaceous deposits (e.g., the special small
active sites for selective hydrogenation of acetylene), selectivity
for ethylene production could be improved [7]. Also, it has been
proposed by Al-Ammar and co-workers [14] that a dissociative
adsorbed species which might be removed easily through hydrogen
treatment at ambient temperature can act as the hydrogen transfer
medium. According to their suggestions, hydrogenation involves
hydrogen transfer between a dissociatively adsorbed species and
associatively adsorbed acetylene, rather than by direct addition of
hydrogen to adsorbed acetylene. In this regard, Podkolzin et al. [15]
confirmed this hydrogen transfer mechanism theoretically with
Density Functional Theory (DFT) calculations. Results from their
modelling suggested that there existed three main effects of hydro-
carbon coverage on selectivity in ethylene production. First, they
suggested that at high carbonaceous coverage, ethylene desorp-
tion became more competitive than further hydrogenation since
the presence of spectator species destabilised the reactive species.
They then proposed the effect related to changes in activity due to
unequal destabilisation of reactive hydrocarbon species. This later
view led to certain reaction pathways becoming unfavourable due
to steric hindrance. Finally, the last effect of such coverage was the
ability to exchange hydrogen with reactive species which provided
ing of higher hydrocarbons. After the separation of products,
acetylene impurities usually exist in the ethylene-richflow which
could poison traditional polymerisation catalysts. Catalytically
selective hydrogenation of acetylene is an important industrial pro-
cess to remove traces of acetylene in the ethylene-richflow for
the production of polymeric grade ethylene. Commercially avail-
able Pd (palladium)-based catalysts have problems of their own in
this regard. In other words, they result in poor ethylene selectiv-
ity at high acetylene conversion. In addition, they cause undesired
oligomer formation during acetylene hydrogenation. Furthermore,
catalyst deactivation by means of coke deposition requires frequent
regeneration processes of the catalyst utilised in the hydrogena-
tion reactor. Ultimately, these Pd-based catalysts are drastically
sensitive to temperature. Due to the temperature rise in the hydro-
genation reactor, such commercial catalysts lose further selectivity
to ethylene, thus, more ethylene is consumed, hence the temper-
ature is increased again. A typical behaviour of current catalysts
leads to declining ethylene selectivity as acetylene conversion or
temperature increases. Therefore, in practice for complete elimina-
tion of acetylene impurities in the ethylene feed stream, referred
to as tail-end processes, two reactors with inter-cooling operations
are usually applied [1–6].
∗ Corresponding author. Tel.: +98 21 66165425; fax: +98 21 6602 2853.
E-mail address: kazemini@sharif.edu (M. Kazemeini).
0
926-860X/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.03.055

H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
185
lower energy reaction pathways. Thus, it was expected that the
role of this mechanism would become more pronounced with the
enhancement of temperature and a reduction in the H /C ratio [7]
2
2
which led to more deposition of carbonaceous coverage.
Amongst different types of supports used in heterogeneous
catalysis, carbon materials have attracted growing interest due
to their specific characteristics, including (i) resistance to acidic
or basic media, (ii) the possibility to control porosity and surface
chemistry, up to certain limits and (iii) easy recovery of precious
metals by burning up the support associated with a low environ-
mental impact [16]. Also, carbon-supported catalysts have specific
properties which in some reactions might display thoroughly dif-
ferent performance compared with other catalyst supports such as
alumina, silica, etc. Accordingly, the inertness of the carbon sup-
port is a property that leads to much lower coking propensity
in comparison with alumina or silica supports [17]. For exam-
ple, in a hydro treating reaction such as hydrodesulphurisation, it
was observed [18–21] that the carbon-supported catalysts showed
more activity and lower coking propensity compared with conven-
tional alumina-supported catalysts. This was originally attributed
to the almost inert character of the carbon support [17]. In this
venue, it was clearly observed in recent research that, in compar-
ison to alumina or activated carbon, the carbon nano-structured
material (CNS)-supported Pd catalyst showed less green oil forma-
tion and no detectable coke deposition in selective hydrogenation
of acetylene [22]. Results of this paper indicated that the synthe-
sized, nano-structured Pd/CNS catalyst gave better and sustained
ethylene selectivity and activity in comparison to the reference cat-
alysts. Ultimately, it was observed that the 1 wt.% Pd/CNS could
Fig. 1. Raman spectra of synthesised MWCNT.
◦
display 98% ethylene selectivity after 4 h at 70 C, 1.0 atm, and feed
molar ratio of C H :H = 1:1 [22].
2
2
2
The current research presents the synthesis, characterisation
and evaluation of a selective hydrogenation nanocatalyst (i.e.;
Fig. 2. SEM micrographs of synthesised MWCNT.
0.5 wt.% Pd/MWCNT) showing very promising behaviour in acety-
lene hydrogenation, including suppression of oligomer formation
as well as greater stability toward deactivation. Furthermore,
with increasing temperature, this material’s selectivity to ethylene
production and acetylene conversion were both enhanced. Conse-
quently, when the Pd/MWCNT catalyst was utilised, high ethylene
yields in selective hydrogenation of acetylene were realised.
Pd (II) chloride monohydrate, [Pd(NH3)4]Cl2·H2O (obtained from
Aldrich Chem. Co.), as a Pd precursor. The concentration of Pd in
the solution was set to attain 0.5 wt.% on the support. The synthe-
sised MWCNT support was stirred in as the solution was added for
1 h. After that, the support was rinsed with distilled water, dried ini-
tially at ambient temperature overnight and then further for 14 h
◦
at 110 C. In order to prevent serious sintering of the Pd particles,
2
. Experimental
instead of calcinations at high temperature, the prepared catalyst
◦
was subjected to H2 treatment at 150 C for 2 h.
2.1. MWCNT synthesis
2.3. Characterisation techniques
In this research, high purity MWCNTs (>90%) were utilised as a
support for the Pd catalyst. This material was previously synthe-
sised by these researchers through a special CVD (chemical vapour
deposition) method decomposing methane on a Co–Mo/MgO cat-
The metal loading content of the nanocatalyst was determined
by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-
AES), Perkin Elmer Optima 3000DV. The amount of Pd loading on
the MWCNT was determined to be 0.5 wt.% of the support.
◦
alyst at a temperature of 1000 C [23]. The high purity MWCNTs
were obtained through two stages involving (i) an acid treatment
to dissolve impurities such as metal catalysts and (ii) an oxidation
step. Initially, MWCNTs immersed in 18% HCl solution were stirred
The carbon materials were characterised by SEM (scanning elec-
tron microscope, Jeol JSM-7500F) and Raman spectroscopy, using
a Horiba Jobin-Yvon T64000 Raman spectrograph with 514.5 nm
excitation wavelength (see Figs. 1 and 2). As shown in Fig. 1, two
at ambient temperature for 16 h. Then, this material was treated in
◦
◦
6
M HNO3 at 40 C for 6 h, washed with distilled deionised water
−1
characteristic peaks at 1350 and 1582 Raman shift (cm ) are
and dried at 100 C. Finally, the resulting MWCNTs were oxidised
attributed to the MWCNT structure.
◦
at 400 C for 20 min to burn out all amorphous carbons.
XRD (X-ray diffraction) patterns were recorded on a Philips
Analytical Diffractometer using Cu Ka radiation. It is reiterated,
however, that due to low Pd loading on the MWCNT (i.e., 0.5 wt.%),
the XRD displayed no Pd peak, hence not shown here.
Through surface modification of the MWCNTs, high purity mate-
rials were functionalised in a boiling aqueous solution of 35% H O2
2
(
obtained from Fluka Chemicals) for 2 h.
Textural properties of the nanocatalyst were obtained utilis-
ing a Micromeritics ASAP-2010 nitrogen sorption system. Through
H2 treatment Prior to determination of the adsorption isotherm,
2
.2. Catalyst preparation
◦
Pd deposition on the functionalised MWCNTs was carried out via
the sample was out gassed and heated under vacuum at 300 C
the incipient wetness impregnation technique using tetra-amino
for 12 h. Ultimately, the BET (i.e. Brunauer–Emmett–Teller) equa-

1
86
H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
rates. Reactant and product streams were analysed using an online
gas chromatograph (GC, Agilent GC6890 N) fitted with an HP-PONA
capillary (50 m, 0.2 mm) connected to FID (flame ionisation detec-
tor) for the analyses of hydrocarbons and Poropak-Q (provided by
the Supelco) connected to a TCD (thermal conductivity detector)
for the analysis of permanent gases.
Prior to the catalyst being evaluated, as mentioned above, the
catalyst was reduced by passing a gaseous mixture of H /N volume
2
2
ratio of 1:3, at a total flow rate of 200 ml/min for 2 h at a temperature
◦
of 150 C.
The following formulae were applied to calculate the acetylene
conversion (%), reaction selectivity (%) for ethylene, ethane and
total butene production and ethylene production yield (%), respec-
tively:
Moles C H (In) − Moles C H (Out)
2
2
2
2
C H conversion (%) =
× 100
× 100
× 100
×100
2
2
Moles C H (In)
2
2
Moles C H (Out) − Moles C H (In)
2
2
2
2
C H selectivity (%) =
2
2
Moles C H (In) − Moles C H (Out)
2
2
2
2
Fig. 3. TEM image of the 0.5% Pd/MWCNT catalyst.
Moles C H (Out) − 0
2
2
C H selectivity (%) =
2
6
Moles C H (In) − Moles C H (Out)
2
2
2
2
tion was used to determine the specific surface area and total pore
2
3
volume. Values obtained were 132.3 m /g and 0.37 cm /g for the
Moles C H (Out) − 0
2
8
2
3
C H selectivity (%)=2×
MWCNT material and 110.3 m /g and 0.35 cm /g for the synthe-
sized nanocatalyst.
2
8
Moles C H (In) − Moles C H (Out)
2
2
2
2
The morphology of the Pd/MWCNT catalyst was characterised
by a TEM (transmission electron microscope). TEM investigations
were carried out using Zeiss 912 AB (120 kV) equipment. The sam-
ples for TEM studies were prepared by ultrasonic dispersion of
the catalyst in 2-propanol for 30 min, and the suspensions were
dropped onto a copper micro-grid covered by a perforated carbon
film. Fig. 3 clearly showed the multi-walled carbon nanotube struc-
ture, dispersion of metallic palladium clusters over MWCNTs and
cluster size distribution.
C H yield (%) = C H Conversion × C H Selectivity
2
4
2
2
2
2
Moles C2H4(Out) − Moles C2H4(In)
=
× 100
Moles C2H2(In)
It was noted that, as a result of the total hydrocarbon materials
mass balance, the above three selectivity formulae should sum up
to unity if no side reactions such as oligomer formation occurred.
The sample was also studied by TPO (temperature programmed
oxidation) utilising a Micromeritics TPD-2900 apparatus connected
to a Pfeiffer Vacuum-300 mass spectrometer. In this experiment,
3
. Results and discussion
3.1. Stabilisation of the Pd/MWCNT catalyst
1
0 mg of used Pd/MWCNT catalyst (i.e., the sample was aged for 50 h
by passing a gaseous mixture composed of 1.2 vol% C H , 45 vol%
2
2
Fig. 4 indicates the acetylene conversion and selectivity to
C2H and 2.4 vol% H2 being diluted by N2 at a total flow rate of
4
ethane, ethylene and butene species during the stabilisation period
of the synthesised nanocatalyst. It was observed that the selective
◦
1
35 ml/min and temperature of 120 C),was maintained in a stream
of 5 vol% oxygen and 95%heliumwith a flow rate of 60 cm /min,
3
◦
hydrogenation reaction over Pd/MWCNT at a temperature of 120 C
◦
◦
and heated from 50 up to 550 C at 10 C/min. Due to burning of
the MWCNT support during this experiment, for comparison and
analysis of the coke burn-off of the used catalyst, the fresh catalyst
was also tested. It should be noted that samples tested were not
out-gassed before the experiment in order to detect all adsorbing
species.
reached steady state approximately 160 min after initiation. How-
ever, the catalyst was aged for another 50 h to investigate formation
possibilities of undesired oligomers as side reaction products. Dur-
ing this time of aging, the selectivity toward ethane; ethylene and
butene species as well as, acetylene conversion were not shown
any change and kept constant. In addition, in regards to the total
carbonic mass balance, these three selectivity’s values summed
up to unity. These results illustrated that no detectable oligomers
2.4. Catalytic experiments
(e.g., oligomer formation was considered as a side reaction) were
The performance of the Pd/MWCNT catalyst was studied using a
formed by the nanocatalyst under the studied reaction conditions.
Furthermore, persistent activity of the nanocatalyst (e.g., acetylene
conversion has not decreased), displayed that no sensitive deacti-
vation of the nanocatalyst occurred in this period.
During stabilisation of the catalyst, ethylene selectivity was
sharply improved at approximately 20 min after initiation. The data
shown in Fig. 4 demonstrated that the decline in selectivity to
butene species was mainly responsible for the drastic improvement
in selectivity to ethylene during this period. Ethane selectivity, on
the other hand, decreased slightly during this time, which in turn
affected the ethylene selectivity as well. GC analysis indicated that,
at the beginning of the catalyst conditioning period, nearly 70 wt.%
of the inlet acetylene was converted into butene species. However,
fixed-bed differential stainless-steel micro-reactor (with an inter-
nal diameter of 12 mm) operated to eliminate heat and mass
transfer gradients. The catalyst (∼0.3 g) was placed in the mid-
dle of the micro-reactor. Glass beads were placed below and above
the catalyst bed to maintain its position and to insure proper pre-
heating of the feed. Inlet gas mixtures were composed of cylinder
gases of ethylene (99.95% pure, provided by the Matheson Co.),
hydrogen (99.999% pure, provided by the Roham Gas Co.), acety-
lene (99.96% pure, provided by the Matheson Co.) and nitrogen
(
99.999% pure, provided by the Roham Gas Co.). Four mass flow
controllers (Brooks, Model 5850E) equipped with a control panel
Brooks 0154) were utilised to automatically adjust the inlet flow
(

H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
187
Fig. 4. Variation of conversion and selectivity over Pd/MWCNT catalyst as a function of time for acetylene hydrogenation by passing a gaseous mixture composed of 1.2 vol%
◦
C2H2, 45 vol% C2H4 and 2.4 vol% H2 being diluted by N2 at a total flow rate of 135 ml/min and temperature of 120 C.
during the catalyst stabilisation time, this concentration declined
to ca. 35 wt.% of the inlet value. In addition, it should be highlighted
that no butadiene or other diene species were detected during
any of the experiments. Other researchers have suggested that
butadiene and other diene species act as precursors for oligomer
formation and more carbonaceous deposition on the catalyst sup-
port [24,25]. It seemed that the nanocatalyst developed in this work
had no tendency to produce oligomers. In other words, instead of
the formation of long chain compounds, it led toward the forma-
tion of relatively short chain compounds such as butene species
on its surface. This was emphasised by the detection of butene
species formed through the addition of H2 to butadiene [8] over
the Pd/MWCNT catalyst for which side reactions of oligomers and
the formation of long chain carbonaceous species (i.e., deposits)
were suppressed. This behavioural pattern of Pd/MWCNT might be
considered similar to what was observed by Kim et al. [26] in which
the addition of the TiO2 to a Pd catalyst decreased the chain length
of the oligomers.
Fig. 5. Selectivity for ethylene production versus reaction temperature for ethylene-
rich feeds containing H2/C2H2 molar ratios of 1, 1.2 and 1.3.
3.2. Catalytic performance of Pd/MWCNT
Catalytic tests to study activity and selectivity performance
were carried out on the stabilised sample aged for 50 h at the reac-
tion conditions presented in Section 3.1. In these tests, gaseous
mixtures of 93 vol% ethylene and H /C H molar ratios of 1, 1.2
2
2
2
and 1.3 with a total flow rate of 135 ml/min as reactor feeds were
passed over the catalyst bed and the product gases were analysed
by GC. The reaction temperature was set to vary in the range of
◦
ca. 90–180 C. In addition, at each test point, the temperature was
maintained for 15 min to reach a steady state before the product
was analysed. The results shown in Figs. 5 and 6 were comple-
mentary perspectives of one another. These figures showed that for
the Pd/MWCNT catalyst, ethylene selectivity increased with both
temperature and acetylene conversion. In other words, as the tem-
perature rose, acetylene conversion and ethylene selectivity were
enhanced.
These results were on one hand, in contrast with the previously
observed selective behaviour of many hydrogenation catalysts
reported earlier in the literature [26–32], while on the other hand,
they supported a few rather specific previous studies [33–36].
Molero et al. [34] worked on a Pd foil as an acetylene hydrogena-
Fig. 6. Selectivity for ethylene production versus acetylene conversion for ethylene-
rich feeds containing H2/C2H2 molar ratios of 1, 1.2 and 1.3.

1
88
H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
Fig. 8. Ethylene production yield versus reaction temperature for ethylene-rich
feeds containing H2/C2H2 molar ratios of 1, 1.2 and 1.3 for GHSV of 5000 and
−
1
36,000 h
.
taking place over the Pd/MWCNT catalyst. This will be discussed in
the following section.
In the current investigation, it was observed that, with increased
Fig. 7. The TPO profile of fresh and used nanocatalyst.
hydrogen partial pressure in the feed mixture in which the H /C2
2
molar ratio was increased to ca. 1.4 and beyond, the catalyst
completely lost its selectivity toward ethylene. However, the
hydrogenation reaction continued to consume almost all of the
hydrogen in the feed. Under these conditions, the Pd/MWCNT
material was known to act as an overactive hydrogenation cata-
lyst which had no selectivity towards ethylene. This phenomenon
tion catalyst. They determined that ethylene selectivity increased
drastically with temperature from 30% at 300 K to 94% at 470 K
when 100 Torr of acetylene and hydrogen were utilised [34]. Subse-
quently, Mei et al. [35] applied the Monte Carlo simulation method
for this case and under the same experimental conditions employed
by Moleroand co-workers [34]. These conditions included a feed
H /C molar ratio of 1/1 and a temperature range of 300–500 K.
suggested that, at a critical H /C2 molar ratio, a change in hydro-
2
genation mechanisms might occur over this catalyst. Similar effects
2
2
The simulated results of Mei et al. emphasised the enhancement of
ethylene selectivity with increased temperature. They rationalised
that, although the rates of acetylene and ethylene hydrogenation
both increased with temperature, the desorption rate of ethylene
increased faster than that of ethylene hydrogenation with temper-
ature. Increased surface coverage of acetylene with temperature
was detected as still another complementary reason for increased
ethylene selectivity with temperature. It should be noted that their
simulated results were obtained assuming no side reactions such as
oligomerisation (e.g., which would have led to diene intermediates)
occurred. Therefore, it was concluded that, if in practice, during
acetylene hydrogenation over a catalyst, side reactions (e.g., mainly
oligomer formation) did not occur or were suppressed by the cat-
alytic performance, the ethylene selectivity would improve with
temperature [35,36]. This behaviour was fairly similar to what was
observed over the Pd/MWCNT catalyst investigated in the present
study.
The temperature-programmed-oxidation (TPO) of the used
nanocatalyst investigated in this research illustrated that very little
coke was deposited on this material (see Fig. 7). Ultimately, it was
confirmed that no side reactions for oligomer formation or coke
deposition occurred on the nanocatalyst studied in this work. In
Fig. 7, a small amount of CO2 detected at low temperatures might
refer to adsorption of carbonaceous species over the catalyst.
The data shown in Fig. 5 displayed that, at each reaction tem-
of high H /C2 molar ratios had been observed in previous studies
2
[37–39]. It was proposed by Sarkany et al. [40] that this sudden
change in catalytic behaviour was caused by a drastic reduction in
carbonaceous deposits and clearing off of the catalyst surface from
coke coverage at high H /C2 molar ratios. Supported by the TPO
2
results (Fig. 7), the present authors suggest that a thin carbonaceous
layer was deposited on the catalyst during aging of the synthesised
Pd/MWCNT. This layer, consequently, played an essential role in
the occurrence of the observed selective behaviour of the catalyst
in the hydrogenation of acetylene. However, under the reaction
conditions, this proposed carbonaceous deposit on Pd/MWCNT to
some extent did not impose serious limitations on the internal dif-
fusion of acetylene molecules such as what was reported by Tracey
et al. over a Pt (pellatin) film [33] or had no effects on the facilita-
tion of hydrogen spillover to the support, since both these effects
implied enhancement of selectivity toward ethane production.
Fig. 8 displayed the ethylene production yield with respect
to the reaction temperature. It was clearly shown that the cat-
alytic performance improved with increased temperature. The
ethylene yield (i.e., obtained as a product of conversion with
selectivity) was almost the same, especially when the tempera-
◦
ture rose above 130 C for different molar ratios of H /C H . This
2
2
2
perhaps emphasised that the aforementioned governing hydro-
genation mechanism acting over the Pd/MWCNT catalyst became
overwhelmed compared with other hydrogenation routes as the
temperature increased. Therefore, the effect of the H /C H molar
perature with a decline in the H /C2 ratio, ethylene selectivity
2
2
2
2
increased. Also, differences in ethylene selectivity between runs
ratio upon catalytic performance at relatively higher temperatures
(e.g., above 130 C) was blurred.
◦
with feeds containing H /C2 molar ratios of 1 and 1.3 decreased
2
◦
from about 25% to 10% as the temperature increased from 89 C to
80 C. It was clear that with increased reaction temperature, the
In Fig. 8, the ethylene yield for a feed with total flow equal
◦
−1
1
to GHSV (gas hourly space velocity) of 5000 h
was displayed.
influence of the H /C2 molar ratio on ethylene selectivity as well
as acetylene conversion became less pronounced (see Fig. 6). This
behaviour was related to the governing hydrogenation mechanism
This revealed that, by regulating the total flow, one might
enhance acetylene conversion to nearly 100% and as a result,
achieve high yields of ethylene production. Indeed, experimen-
2

H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
189
Fig. 9. Ethylene selectivity versus acetylene conversion at a constant temperature for ethylene-rich feeds containing a H2/C2H2 molar ratio of 1 and GHSV of 5000 and
−
1
3
6,000 h
.
tal runs in the present research led to an ethylene yield of
3%.
Another consequence of the present work which resulted
with increased acetylene conversion [33]. In addition, as a result
of enhanced carbonaceous species deposition and the hydrogen
transfer mechanism prevailing over the nanocatalyst the increased
temperature weakened the progress of these series reactions.
9
through selectivity enhancement by temperature increase was that
the tail-end processes could possibly be operated in a single reac-
tor and inter-cooling no longer be necessary. This was due to the
fact that the temperature increase in the reactor (i.e., because of the
exothermic nature of the reaction) caused better catalytic perfor-
mance (i.e., higher selectivity and yield), hence no need to decline
the temperature (which meant employing an inter-cooling stage)
was felt. This is an improvement to the current industrial double
reactor setup utilised.
3.4. Effects of MWCNT support on catalytic behaviour
It has previously been claimed that catalyst deactivation and
coke deposition on the support increase with the support’s acidity
in hydrogenation reactions [24]. Accordingly, Asplund [24] showed
that oligomerisation occurred on the alumina support; moreover,
coke formation was about four to five times faster on Pd/␥-Al O3
2
compared with the ␣-Al O -supported catalyst. Based upon the
2
3
3
.3. Selective hydrogenation mechanisms on Pd/MWCNT
obtained results, it seemed that the complete inertness of the
MWCNT-supported catalyst utilised in this work, led to neither
oligomer nor long-chain hydrocarbon depositions formed on the
catalyst compared to alumina, silica or other oxide-supported Pd
catalysts. This was indeed a satisfactory advantage for the catalyst
developed in this investigation.
It was discussed in the previous sections that the formation
of a thin carbonaceous layer on the Pd/MWCNT catalyst seriously
affected the catalytic performance. On the other hand, evidences
revealed that the deposited layer determined the path of hydro-
genation over the nanocatalyst. Following is the summary of results
obtained in this study in order to rationalise that the govern-
ing mechanism in acetylene hydrogenation over the synthesised
Pd/MWCNT catalyst was a hydrogen transfer mechanism:
On the other hand, if the MWCNTs are used in catalysis,
their conductive properties make them clearly different compared
to activated carbon [16]. In accordance with the obtained TEM
image of the present catalyst, it seemed that the special electronic
properties of the functionalised MWCNT support (e.g., electronic
conductivity) during loading of metallic Pd particles are what
might be responsible for the production of relatively homogenised
and similar active sites. Furthermore, this conduction property of
the support made the active catalytic sites relatively homogenous
through electrons transferring between them. Consequently, the
observed behavioural patterns were displayed by this nanocatalyst
in the present research. Subscribing to this viewpoint, it might be
expected that the relatively conductive Pd/MWCNT hybrid should
behave similarly to a Pt film [33] or Pd foil [34] in the selective
hydrogenation of acetylene (e.g., see Figs. 5 and 6). Perhaps the
nanocatalyst utilised in the present study might even perform bet-
ter than them, in light of its properties of suppressing undesired
oligomer formation and better stability (i.e., toward deactivation).
By increasing the H /C2 molar ratio to a critical amount and
2
beyond, the catalyst completely lost its selectivity toward ethylene.
This meant that a significant change in the hydrogenation mech-
anism occurred over the Pd/MWCNT material which was due to
clearing off the catalyst surface of carbonaceous deposits.
By increasing the temperature or decreasing the H /C2 molar
2
ratio, the selectivity to ethylene improved (see Fig. 5). These
changes led to more carbonaceous species to be formed on the
catalyst.
Finally, acetylene conversion, as well as ethylene selectivity,
increased with temperature (see Fig. 6) which meant that when
undesired side reactions were suppressed over the Pd/MWCNT
catalyst, the hydrogen transfer mechanism prevailed. Therefore,
activity boosted when temperature increased.
Fig. 9 clearly confirmed the above results. It shows that ethy-
lene selectivity decreased while acetylene conversion increased
at constant temperatures. This figure highlighted the essential
role of the hydrogen transfer mechanism in acetylene hydrogena-
tion over the Pd/MWCNT catalyst. Furthermore, it emphasised
that this mechanism prevailed as the governing factor when tem-
perature increased. It is noteworthy that, in previous studies, it
was shown that due to the occurrence of hydrogenation reactions
in series, at constant temperature ethylene selectivity decreased
4. Conclusions
In this research, a selective hydrogenation catalyst as well as
new process conditions for the selective hydrogenation of acety-
lene in an ethylene-richflow were obtained and discussed. The
prepared nanocatalyst acted very satisfactorily at high tempera-
tures in such a way that the elimination of acetylene impurities
in ethylene feeds was possible in a single hydrogenation reactor.

1
90
H. Bazzazzadegan et al. / Applied Catalysis A: General 399 (2011) 184–190
In other words, through this route, industrially selective hydro-
genation in a tail-end process might be operated in a single reactor
as opposed to the current industrial double reactor setup. Hence,
inter-cooling is no longer necessary. As a further consequence of
this, only small amounts of hydrogen entering the feed will be
needed in a single reactor setup utilising this catalyst. Ultimately, a
very significant advantage of the present catalyst was its high yield
for ethylene production which reached a maximum amount of 93%.
High ethylene yields as well as suppression of oligomer forma-
tion and greater stability toward deactivation during the selective
hydrogenation of acetylene impurities in an ethylene-rich flow
makes the Pd/MWCNT catalyst a promising candidate for a success-
ful selective hydrogenation catalyst for future ethylene production
from acetylene-rich flow streams.
[7] A. Borodzinski, G.C. Bond, Catal. Rev. 48 (2006) 91–144.
[
[
8] A. Borodzinski, A. Cybulski, J. Appl. Catal. A 198 (2000) 51–66.
9] A. Borodzinski, J. Catal. Lett. 63 (1999) 35–42.
[
10] A. Borodzinski, J. Catal. Lett. 71 (2001) 169–175.
[11] A. Molnar, A. Sarkany, M. Varga, J. Mol. Catal. A: Chem. 173 (2001) 185–221.
[
[
12] D. Duca, F. Frusteri, A. Parmaliana, G. Deganello, J. Appl. Catal. 146 (1996)
69–273.
2
13] D. Duca, D. Arena, A. Parmaliana, G. Deganello, J. Appl. Catal. A 172 (1998)
207–211.
[14] A.S. Al-Ammar, G.J. Webb, J. Chem. Soc. 74 (1978) 657–664.
15] S.G. Podkolzin, R. Alcala, J.A. Dumesic, J. Mol. Catal. A: Chem. 218 (2004)
17–227.
[
2
[16] P. Serp, M. Corrias, P. Kalck, J. Appl. Catal. A 253 (2003) 337–358.
[17] F. Rodriguez-Reinoso, Carbon 36 (1998) 159–175.
18] J.C. Duchet, E.M. VanOers, V.H.J. De Beer, R. Prins, J. Catal. 80 (1983) 386–402.
19] P. Arnoldy, E.M. Van Oers, V.H.J. De Beer, J.A. Moulijn, R. Prins, J. Appl. Catal. 48
[
[
(
1989) 241–252.
[20] V.H.J. De Beer, F.J. Derbyshire, C.K. Groot, R. Prins, A.W. Scaroni, J.M. Solar, Fuel
3 (1984) 1095–1100.
6
In addition, based upon the results obtained in this work and
comparisons made with other data available in the literature, acety-
lene hydrogenation on the Pd/MWCNT catalyst was rationalised to
take place via a hydrogen transfer mechanism.
[
[
21] A.V. Scaroni, R.G. Jenkins, P.L. Walker, J. Appl. Catal. 14 (1985) 173–183.
22] W.G. Augustyn, R.I. MCrindle, N.J. Coville, J. Appl. Catal. A 388 (2010) 1–6.
[23] A.M. Rashidi, B. Amini, A. Mohajeri, Kh.J. Jozani, US Patent 20,080,274,277
2008).
(
[
[
24] S. Asplund, J. Catal. 158 (1996) 267–278.
25] M. Larsson, J. Jansson, S. Asplund, J. Catal. 178 (1998) 49–57.
Acknowledgements
[26] W.J. Kim, J.H. Kang, I.Y. Ahn, S.H. Moon, J. Catal. 226 (2004) 226–229.
[
27] P. Praserthdam, B. Ngamsom, N. Bogdanchikova, S. Phatanasri, M. Pramotthana,
J. Appl. Catal. A 230 (2002) 41–51.
The authors would like to express their thanks to the RIPI,
Tehran, Iran for their financial support of this project. In addition,
they gratefully acknowledge very helpful discussions with Profes-
sor Y. Mortazavi of the Nano Science Centre of Excellence at the
Tehran University.
[
28] Q. Zhang, J. Li, X. Liu, Q. Zhu, J. Appl. Catal. A 197 (2000) 221–228.
[29] J.H. Kang, E.W. Shin, W.J. Kim, J.D. Park, S.H. Moon, J. Catal. Today 63 (2000)
83–188.
1
[
[
30] J.H. Kang, E.W. Shin, W.J. Kim, J.D. Park, S.H. Moon, J. Catal. 208 (2002) 310–320.
31] E.W. Shin, J.H. Kang, W.J. Kim, J.D. Park, S.H. Moon, J. Appl. Catal. A223 (2002)
161–172.
[
32] N. Wongwaranon, O. Mekasuwandumrong, P. Praserthdam, J. Panpranot, J.
Catal. Today 131 (2008) 553–558.
References
[
[
33] S. Tracey, A. Palermo, J.P. Vazquez, R.M. Lambert, J. Catal. 179 (1998) 231–241.
34] H. Molero, B.F. Bartlett, W.T. Tysoe, J. Catal. 181 (1999) 49–56.
[
[
1] M.M. Johnson, D.W. Walker, G.P. Nowack, US Patent 4,484,015 (1984).
2] Y.J. Huang, C.F. Shun, L.G. Daniel, E.L. Mohundro, J.E. Hartgerink, US Patent
[35] D. Mei, P.A. Sheth, M. Neurock, C.M. Smith, J. Catal. 242 (2006) 1–15.
[36] J. Osmwald, K. Kovnir, M. Armbruster, R. Giedigkeit, R.E. Jentofta, U. Wilda, et al.,
J. Catal. 258 (2008) 219–227.
[37] G.C. Bond, P.B. Wells, J. Catal. 5 (1966) 65–73.
[38] G.C. Bond, P.B. Wells, J. Catal. 6 (1966) 397–410.
5
,332,705 (1994).
[
[
[
3] G.C. Bond, P.B. Wells, J. Catal. 5 (1965) 65–69.
4] A.N.R. Bos, K.R. Westerterp, J. Chem. Eng. Pro. 32 (1993) 1–6.
5] H. Arnold, F. Dobert, J. Gaube, G. Ertl, H. Knoerzinger, J. Weitkamp, Handbook
of Heterogeneous Catalysis, VCH, Weinheim, 1997, pp. 2165–2166.
6] K. Flick, C. Herion, H.M. Allman, US Patent 5,856,262 (1999).
[39] R.B. Moyes, D. WWalker, D.A. Wells, D.A. Whan, J. Appl. Catal. 55 (1989) L5–L8.
[40] A. Sarkany, A. Horvath, A. Beck, J. Appl. Catal. 229 (2002) 117–125.
[