Nickel and molybdenum containing mesoporous catalysts for ethylene oligomerization and metathesis

Article abstract of DOI:10.1039/c5nj02586a

An original Ni-AlSiO₂ mesoporous catalyst (Si/Al = 6.5, 2.0 wt% Ni) was prepared using a commercial mesoporous non-ordered silica as a support. It exhibited notable catalytic behaviour for the oligomerization of ethylene in both batch and flow mode. Thus, at 150 °C and 3.5 MPa, the catalyst was highly active (up to 130 g of oligomers per gram of the catalyst per hour), selective (C4, C6, C8, and C10 olefins, no cracking products) and stable (high conversion for 9 h on-stream). However, the catalytic performances were slightly lower compared to those obtained with the catalyst prepared using a mesostructured SBA-15 silica as a model support. When nickel-catalyzed oligomerization was coupled with a metathesis reaction catalyzed by MoO_x/(Al)SiO₂ or MoO_x/(Al)SBA-15, ethylene was efficiently converted into propylene. To determine the physico-chemical properties of catalysts, various techniques were used, such as powder X-ray diffraction, N₂ sorption, ²⁷Al MAS NMR, Raman spectroscopy and X-ray photoelectron spectroscopy.

Full text of DOI:10.1039/c5nj02586a

NJC
View Article Online
View Journal
PAPER
Nickel and molybdenum containing mesoporous
catalysts for ethylene oligomerization
and metathesis
Cite this:DOI: 10.1039/c5nj02586a
ab
b
a
a
Radu Dorin Andrei, Marcel Ionel Popa, Claudia Cammarano and Vasile Hulea*
2
An original Ni-AlSiO mesoporous catalyst (Si/Al = 6.5, 2.0 wt% Ni) was prepared using a commercial
mesoporous non-ordered silica as a support. It exhibited notable catalytic behaviour for the oligo-
merization of ethylene in both batch and flow mode. Thus, at 150 1C and 3.5 MPa, the catalyst was
highly active (up to 130 g of oligomers per gram of the catalyst per hour), selective (C4, C6, C8, and
C10 olefins, no cracking products) and stable (high conversion for 9 h on-stream). However, the
catalytic performances were slightly lower compared to those obtained with the catalyst prepared using
a mesostructured SBA-15 silica as a model support. When nickel-catalyzed oligomerization was coupled
Received (in Montpellier, France)
2
3rd September 2015,
Accepted 12th November 2015
x 2 x
with a metathesis reaction catalyzed by MoO /(Al)SiO or MoO /(Al)SBA-15, ethylene was efficiently
converted into propylene. To determine the physico-chemical properties of catalysts, various techniques
DOI: 10.1039/c5nj02586a
27
were used, such as powder X-ray diffraction, N
2
sorption, Al MAS NMR, Raman spectroscopy and X-ray
www.rsc.org/njc
photoelectron spectroscopy.
solid catalysts ethylene oligomerization leads selectively to
C4–C10 olefins with an even number of carbon atoms. In order
Introduction
Ethylene is one of the most important basic organic chemicals. to enlarge the product range, we developed some catalytic cascade
It is involved in a large number of industrial applications and processes, in which nickel-catalyzed ethylene oligomerization
among them, the oligomerization reaction is of considerable was assisted by a second catalyzed reaction. For example, we
interest. Indeed, ethylene oligomers (i.e. olefins C2n, n = 2–10) are proposed an original procedure in which ethylene oligomerization
valuable starting materials for a variety of chemicals, including (over Ni-AlMCM-41) was combined with the acid-catalyzed
lubricants, surfactants, alcohols, amines, and acids. Ethylene co-oligomerization of the resulting C4–C10 olefins (over proto-
12
oligomerization is typically catalyzed by transition metals con- nated AlMCM-41) to obtain hydrocarbons in the distillate range.
tained in both homogeneous and heterogeneous catalysts. Of On the other hand, we reported in a recent paper that ethylene
1
particular note are Ni-based complexes and Ni-exchanged can be directly converted into propylene by one-pot catalytic
2
13
porous materials, which exhibited very high activity and cascade reactions, using two heterogeneous catalysts. In a single
selectivity in this reaction. Various heterogeneous catalysts, flow reactor ethylene was first selectively dimerized/isomerized
3
including Ni-amorphous silica-alumina, Ni–(Al)-mesostructured over the Ni-AlSBA-15 catalyst to form 2-butenes, which reacted
4–8
9–11
materials,
and Ni-zeolites,
were capable of oligomerizing then with the excess of ethylene in a metathesis reaction catalyzed
ethylene, even under mild conditions. For these catalysts it was by MoO –SiO –Al O , to produce propylene. It is worth noticing
3
2
2 3
shown that the pore size is a crucial variable affecting their activity that a higher growth rate is in demand for propylene compared to
2
and stability. For example, the Ni-exchanged zeolites suffered ethylene during the last few decades and the need to develop
severe deactivation (due to the blocking of their micropores with sustainable sources of raw materials highly stimulated the interest
9
–11
heavy products),
while the catalysts with larger pores, i.e. for the conversion of ethylene to propylene through simpler
Ni–Al-mesoporous materials, exhibited very high productivity in procedures, without any addition of other hydrocarbons. The
4
–6
the ethylene conversion to oligomers. Typically, with Ni-based state of the art of these processes has been discussed in our
13
previous contribution.
In some earlier studies, we showed that Ni-exchanged meso-
porous materials, such as Ni-AlMCM-41, Ni-AlMCM-48 and
Ni-AlSBA-15, with a carefully controlled texture, are ‘‘ideal’’ solid
catalysts for ethylene oligomerization. The aim of the present
work was to develop new oligomerization catalysts based on a
a
b
Institut Charles Gerhardt Montpellier, UMR 5253 CNRS/ENSCM/UM,
, rue de l’Ecole Normale, 34296 Montpellier Cedex 5, France.
8
E-mail: vasile.hulea@enscm.fr
4
–6
Faculty of Chemical Engineering and Environmental Protection, ‘‘Gheorghe
Asachi’’ Technical University of Iasi, 71 D. Mangeron Ave., 700050, Iasi, Romania
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
non-ordered inexpensive commercial silica. Note that this is plasma optical emission spectroscopy (ICP-OES) method, on a
the first time that this support is used for preparing Ni-based Perkin Elmer Optima 7000 DV apparatus. The XRD measure-
oligomerization catalysts. The results were compared with our ments were performed on a Bruker AXS D8 diffractometer,
6
earlier investigation using the Ni-AlSBA-15 catalyst. Additionally, using CuKa radiation and a Ni filter. The textural characteriza-
MoO /(Al)SiO2 and MoO /(Al)SBA-15 materials were prepared tion was achieved by using a conventional nitrogen adsorption/
x
x
and characterized. The Ni-based catalysts were coupled with desorption method, at À196 1C, on a Micrometrics ASAP 2020
MoO -containing metathesis catalysts in one-pot catalytic automatic analyzer. The structural and textural parameters were
x
cascade reactions in order to convert ethylene into propylene.
calculated from nitrogen isotherms and XRD data, as previously
described. Al MAS NMR spectra were recorded for the calcined
samples at room temperature using a Varian 600 MHz WB
6
27
Experimental
Materials and catalysts
Premium Shielded spectrometer, using 3.2 mm o.d. ZrO rotors
2
with a rotation speed of 20 kHz. A pulse width of 1 ms was used, with
a recycle delay of 1 s. The number of scan was 2600 and the
3 3
chemical shifts were referenced to an aqueous solution of Al(NO ) .
The Raman spectra were measured using a confocal microspectro-
meter (Labram HR, Jobin-Yvon) in air at 25 1C. The sample
volume was illuminated using an argon–krypton-ion laser beam
The commercial silica gel (Davicat (R) SI 1454; Lot # SP550-
10019ID35; a particulate size of 200 mm) was used as received
from Grace-Davidson. It is denoted in this study as SiO . SBA-15
2
silica was synthesized according to the conventional method,
using the (EO)20(PO)70(EO)20 triblock copolymer (Pluronic P123,
Aldrich), tetraethyl orthosilicate (TEOS, Aldrich) and HCl 2 M
(l = 647.1 nm) focused by an objective (*50LF, N.A. = 0.5, Olympus).
The scattered light was collected using the same objective and
(
Aldrich), at a molar ratio of 1TEOS : 0.016P123 : 4.9HCl : 40.5
À1
dispersed with a 1 cm spectral resolution using a grating of
1
4,15
H O.
The mixture was stirred for 24 h at 40 1C, and then it
À1
2
1
800 lines mm . X-ray photoelectron spectroscopy (XPS) measure-
was maintained for 48 h at 100 1C in a Teflon-lined autoclave
under static conditions. The solid product was filtered, washed
with water, dried in an oven at 80 1C overnight and calcined in
air flow at 550 1C for 8 h.
ments were performed on an ESCALAB 250 (Thermo Electron)
spectrometer equipped with an Al Ka source (1486.6 eV).
Catalytic experiments
AlSiO
and SBA-15 with sodium aluminate (54 Æ 1% Al
Erba), according to the procedure described by Andrei et al.
In a typical experiment, silica (4.0 g) was suspended at 25 1C for
2
and AlSBA-15 samples were obtained by grafting SiO
2
The catalytic oligomerization of ethylene was performed in
both semicontinuous (slurry batch reactor) and continuous
modes (fixed bed reactor). In the first case, the oligomerization
reaction was conducted in a 0.3 L well-mixed gas-slurry auto-
clave. Prior to each experiment, the catalyst (powder) was pre-
treated successively in a tubular furnace (550 1C, 8 h) and in an
oligomerization reactor (200 1C, 3 h) under nitrogen flow at
atmospheric pressure. The reaction was performed at 150 1C
and 3.5 MPa, using 0.5 g of the catalyst and 0.1 L of dry oxygen-
free n-heptane as a solvent, under constant stirring (1000 rpm).
During the experiment ethylene (quality N25 from Air Liquide)
was continuously fed so that the total pressure was maintained
constant in the reactor. After reaction, the autoclave was cooled
to À20 1C and the products were collected, weighted and
analyzed by GC on a Varian 3900 chromatograph (FID, DB-1,
2 3
O
, Carlo
6
1
1
5 h, under stirring in 400 mL of aqueous solution containing
.1 g sodium aluminate, corresponding to a Si/Al ratio of 5.
The samples which resulted in sodium form (Na-AlSiO
Na-AlSBA-15) were subjected to successive ion exchange with
NH NO (99+%, Acros Organics) and Ni(NO O (98%, Alfa
Á6H
Aesar). Typically, 2 g of Na-AlSiO were contacted three times,
for 2 h at 25 1C, under constant agitation, with 100 cm of the
.5 M aqueous solution of NH NO to obtain the ammonium
2
and
4
3
3
)
2
2
2
3
0
4
3
form, NH -AlSiO . The sample in ammonium form was subjected
4
2
to successive nickel-ion exchanges with a 0.5 M aqueous solution
of nickel nitrate, following the same procedure as above. The
exchanged samples were calcined under air for 5 h at 550 1C
to achieve the oligomerization catalysts denoted in this study
6
0 m  0.32 mm, 3 mm film thickness).
In the flow mode, ethylene oligomerization was performed
Ni-AlSiO
2
and Ni-AlSBA-15.
in a stainless steel fixed bed reactor (i.d. 5 mm) using 0.5 g of
the Ni-based catalyst (0.15–0.25 mm sieve fraction). The pres-
sure was regulated via a back-pressure regulator. Before each
test, the catalyst was activated in the reactor at 550 1C under
nitrogen flow for 8 h. The reaction was conducted at 150 1C and
Supported MoO catalysts, denoted the MoO /support (support =
x
x
SiO , AlSiO , SBA-15 and AlSBA-15), were prepared using the
2
2
wet impregnation method, with an ammonium molybdate
aqueous solution ((NH Mo O, Aldrich-Sigma). Prior
4
)
6
7
O
24Á4H
2
to impregnation the supports were dried under vacuum. After
impregnation, the slurry was stirred at room temperature for
3.0 MPa of ethylene, without inert carrier gas. Oligomerization-
metathesis reactions were performed at 80 1C and 3.0 MPa, in a
similar mode, but using two consecutive catalyst beds consisting of
the Ni-based catalyst and the Mo-based catalyst, respectively. The
complete reactor effluent was analyzed online by gas chromato-
graphy (Varian CP-3800, FID), using a CP-PoraPLOT Q capillary
1
h and then the solvent was removed at 80 1C. The catalysts
thus obtained, having a Mo loading of 10%, were calcined in air
at 550 1C for 8 h.
Material characterization
column (25 m, 0.53 mm, 20 mm) and an auto-sampling valve.
À1
The composition of materials (dissolved in solution) was deter- Deactivation rates (in h ) were estimated from the parameter
mined by elemental analysis using the inductively coupled ((a
0
À a
t
)/(a
0
 t)) (where a
0 t
is the initial conversion and a the
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
conversion after t h on stream). For the quantification of the
products captured, the spent catalyst (20 mg) was analyzed by
thermogravimetry (TGA) using a NETZSCH TG 209C apparatus.
The temperature program started with an isothermal period of
5
1
min at 50 1C followed by a temperature ramp up to 900 1C at
0 1C min in synthetic air (20 mL min ).
À1
À1
Results and discussion
Material characteristics
Fig. 1 Nitrogen physisorption of (a) SiO
Al MAS NMR of (a) AlSiO2 and (b) Ni-AlSiO2 (right).
2
, (b) AlSiO
2 2
and (c) Ni-AlSiO (left).
27
Ni-exchanged oligomerization catalysts. The chemical, structural
and textural characteristics of SBA-15, AlSBA-15 and Ni-AlSBA-15
6
materials were widely examined in our previous study. They
However, no significant changes in the isotherm shape (results
are typical of the SBA-15 topology, with very well-ordered arrays
of mesoporous channels. The main properties of these materi-
als are summarized in Table 1. Fig. 1 shows the nitrogen
not shown) were observed for both SiO
of materials.
2
and SBA-15 types
DRUV-Vis and Raman spectroscopy were used as methods for
identifying the MoO species presents in catalysts. The UV-Vis
spectra of both MoO /SBA-15 and MoO /AlSBA-15 (Fig. 2) feature
2 2 2
sorption isotherms of SiO , AlSiO and Ni-AlSiO materials.
x
All samples show a type IV isotherm with a H1 type hysteresis
loop, typical for the mesoporous, but non-ordered materials.
The isotherms point out a significant change of the textural
x
x
a strong absorption band at about 250 nm, which is assigned to
1
7,18
the highly dispersed tetrahedral MoO
x
species.
x
MoO /SBA-15
2 2
features of AlSiO and Ni-AlSiO materials compared to the
features an additional band at approximately 330 nm, which can
parent silica. These changes are quantified in Table 1 and
reveal a decrease of the total pore volume and of the BET
surface area. The modifications, similar to those observed in
the case of the SBA-15 type materials, can be attributed to a loss
of (ultra)microporosity and a decrease of the surface roughness
by a dissolution-redeposition of silica upon treatment in sodium
1
7,18
be ascribed to the polymerized octahedral MoO
Fig. 2 shows also the UV-Vis spectra recorded for MoO
and MoO /AlSiO catalysts. Both samples display large bands in
3
species.
/SiO
x 2
x
2
the range of 200 and 400 nm. The intense band at wavelength
higher than 300 nm indicates that the Mo species are mainly
octahedral MoO and even bulk a-MoO in both MoO /SiO and
6,16
x
3
x
2
aluminate solution.
Using the ionic exchange procedure
MoO
x
/AlSiO
2
samples.
described in the experimental section, the amount of nickel incor-
x x
The Raman spectra recorded for MoO /SBA-15 and MoO /
AlSBA-15 are reported in Fig. 3. Both samples display a band
porated in Ni-AlSiO and Ni-AlSBA-15 samples was 2.0 and 2.5 wt%,
2
respectively. Insertion of aluminium atoms into the silica network
À1
at 960 cm , which is assigned to the highly dispersed MoO
3
27
was confirmed by Al MAS NMR. The NMR spectrum of AlSiO and
2
19–21
monomers.
x
The Raman spectrum of the MoO /SBA-15 sample
2
Ni-AlSiO exhibited an intense signal centred at 55 ppm, which is
À1
displays three additional bands at 995, 819 and 367 cm , which
proves that this catalyst also contains polymerized octahedral
MoO and small crystallites of bulk a-MoO ,
characteristic of tetrahedrally coordinated aluminium species and
only a small signal at about 0 ppm (which is associated with
octahedrally coordinated aluminium species) (Fig. 1). As previously
1
9,22,23
which is in
x
3
agreement with the UV-Vis results.
6
reported for SBA-15 materials, this result proves that the treatment
The Raman spectra of MoO
shown) exhibited a band at 960 cm together with intense
bands at 995, 819 and 367 cm . The nature of the MoO
x
/SiO
2
À1
and MoO
x
/AlSiO (not
2
with sodium aluminate is a very efficient method for the post-
synthesis alumination of the silica.
MoOx supported metathesis catalysts. The properties of
metathesis catalysts (samples with a loading of 10 wt% Mo) are
summarized in Table 2. As expected, compared to the unmodified
À1
x
species in these catalysts was also checked by X-ray diffraction
at diffraction angles from 5 to 701 (Fig. 3). For both catalysts,
the diffraction lines of the crystalline phase of MoO (e.g. at
3
x
supports (see Table 1), the catalysts containing MoO species
showed a decrease of the pore volume and surface area.
1
9,24
23.3, 25.7, 27.3 and 33.781) are visible,
confirming the UV
results. However, the low intensity of these signals suggests
that only a small part of Mo is present as the crystalline phase.
a
Table 1 Properties of parent and modified SiO
2
and SBA-15 samples
2
À1
À1
À1
a
Sample
SiO
D, nm
S
BET, m g
V, mL g
V
mes, mL g
Table 2 Main properties of Mo-based metathesis catalysts
b
2
6.0
400
330
320
740
440
460
0.81
0.65
0.67
1.05
0.73
0.78
0.69
0.55
0.57
0.81
0.59
0.63
D,
nm
S
,
V,
mL g
V
,
Mo,
atoms nm
c
b
BET
mes
AlSiO
2
6.0
2
À1
À1
À1
À2
Sample
m g
mL g
d
b
Ni-AlSiO
2
6.0
e
b
SBA-15
8.4
7.9
7.9
MoO
MoO
MoO
MoO
x
x
x
x
/SiO
/AlSiO
/SBA-15
2
5.9
5.9
7.9
7.7
180
200
540
300
0.62
0.49
0.90
0.55
0.56
0.43
0.74
0.45
3.5
3.1
1.2
2.1
e
b
AlSBA-15
Ni-AlSBA-15
2
f
/AlSBA-15
a
Pore diameter (D), specific surface area (SBET), total pore volume (V),
mesoporous volume (Vmes). Average pore diameter. Si/Al = 6.5
mol/mol). 2.0 wt% Ni. Si/Al = 7.0 (mol/mol). 2.6 wt% Ni.
b
c
a
Pore diameter (D), specific surface area (SBET), total pore volume (V),
d
e
f
b
(
mesoporous volume (Vmes). Average pore diameter.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
Fig. 2 DRUV-vis spectra of MoO
x
/SBA-15 and MoO
x
/AlSBA-15 (left).
DRUV-vis spectra of MoO
x
/SiO
2
and MoO
x
/AlSiO (right).
2
Fig. 4 Mo 3d XP spectrum (normalized with respect to the Si 2p emission)
x
of a MoO /SBA-15 sample.
Table 3 XPS binding energy for metathesis catalysts
Binding energy
Sample
a-MoO
Mo 3d5/2
Mo 3d3/2
3
232.80
233.02
233.31
233.13
233.12
235.90
236.06
236.31
236.28
236.17
Fig. 3 Raman spectra of MoO
x
/SBA-15 and MoO
/AlSiO (right).
x
/AlSBA-15 (left). X-ray MoO
x
x
x
x
/SiO
/AlSiO
2
2
patterns of MoO /SiO and MoO
x
2
x
2
MoO
MoO
MoO
/SBA-15
/AlSBA-15
x
Summarizing, various MoO species coexist in all samples:
monomeric, polymeric, and crystalline Mo oxide. Highly
dispersed (monomeric) species are preferentially produced on
average catalytic activity and the oligomer distribution with
those previously obtained with Ni-AlSBA-15. It is important
to note that the results obtained with Ni-AlSBA-15 were the
highest ever reported in the literature with Ni-based microporous
and mesoporous catalysts. The result obtained with the Ni-
6
the AlSBA-15 type support. This behavior can be related to the
À2
surface density of the MoO species on supports: 2 Mo nm for
x
À2
AlSBA-15 and 3–3.5 Mo nm for SiO /AlSiO (Table 2). On
2
2
the other hand, for each class of materials, the aluminated
supports (material exhibiting acidic properties) allowed us to
AlSiO catalyst is inferior in terms of average productivity (per
2
weight of the catalyst) compared to the Ni-AlSBA-15 catalyst.
Also the turnover frequency (TOF, expressed as the number of
molecules of ethylene transformed by the nickel site per hour)
prepare catalysts with MoO
x
species better dispersed than the
purely silica supports. This is in good agreement with previous
1
8,22
studies,
x
which claimed that highly dispersed MoO species
2
was slightly lower with Ni-AlSiO compared to Ni-AlSBA-15,
are preferentially formed on acidic supports.
though still remarkably high. To estimate the TOF values, each
Ni ion introduced to the silica-alumina support was considered
as a single active site. Note that the high TOF values exhibited by
the Ni-exchanged mesoporous materials were often comparable
The oxidation state of molybdenum in the MoO species was
x
investigated by XPS. The spectrum of the MoO /SBA-15 catalyst
x
(considered as example for this study) in the Mo 3d region is
shown in Fig. 4. Similar spectra, consisting of a single couple of
two well-resolved bands (3d3/2 and 3d5/2), were obtained for all
metathesis catalysts. The values of the binding energies (BE) of
2+
to those obtained in ethylene oligomerization using Ni -based
27–29
complexes as homogeneous catalysts.
2
For both Ni-AlSiO and Ni-AlSBA-15 catalysts, the oligomers
catalysts are slightly higher than those of the bulk a-MoO
Table 3). These results suggest that in all samples the Mo
atoms are essentially in a single oxidation state, most probably
3
consisted of C4, C6, C8 and C10 olefins, with small amounts of
C12 olefins. Only traces of alkanes and odd carbon number
alkenes were identified, indicating that the hydrogen transfer
or the acid catalyzed cracking reactions are not occurring to a
significant level. C4 and C6 olefins were the main products:
(
6
+ 25
6+
Mo . It is important to note that the species Mo are considered
as precursors for generating the active site in metathesis reac-
26
23
tions. On the other hand, according to Thielmann et al., the BE
shifted to higher values compared to bulk a-MoO indicating the
3
presence of isolated and only small (not bulk) clusters of dispersed Table 4 Representative catalytic behaviour of Ni-based catalysts for
ethylene oligomerization
molybdenum oxide species.
Selectivity to, wt%
a
Si/Al,
mol mol
Ni, A1,
wt% g g
TOF,
s
Catalytic results
Ethylene oligomerization in the batch mode. Ni-AlSiO was
used as a catalyst in ethylene oligomerization tests performed Ni-AlSBA-15 7.0
at 150 1C, under 3.5 MPa of ethylene, using n-heptane as a
solvent. The reaction was run for 60 min. Table 4 compares the 3.5 MPa, reaction time 60 min, batch mode.
À1
À1 À1
À1
Catalyst
h
C4 C6 C8 C10+
2
Ni-AlSiO2
6.5
2.0 130
2.6 175
3.7
3.9
56 31 10
41 37 15
3
7
a
catalystÀ1
À1
Average productivity goligomers
g
h
. Conditions: 150 1C,
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
2
7% with Ni-AlSiO and 78% with Ni-AlSBA-15. C4 fraction was
Paper
8
exclusively linear (no isobutene detected) and 2-butene predo-
minated as compared with 1-butene among C4 isomers. The
linear hexenes (75%) also predominated as compared with
other ethylene trimers.
Ethylene oligomerization in the flow mode. The catalytic
2
behavior of the Ni-AlSiO catalyst was also evaluated for ethylene
oligomerization carried out in the flow mode, using a fixed-bed
dynamic reactor. The reaction was conducted over 9 h on stream,
at 150 1C, 3.0 MPa of ethylene, and a weight hourly space velocity
À1
(
WHSV) of 10 h . The initial conversion of ethylene was
2
Fig. 6 TG and DTG curves of the spent Ni-AlSiO catalyst.
about 90% and it declined smoothly, with a deactivation rate
À3 À1
of 19.3 Â 10
h
(Fig. 5). The product distribution (47% C4,
3
4% C6, 16% C8 and 3% C10+) did not change during the of the products, particularly the bulkier ones, and results in a
catalytic test. The average composition of butenes consisted of lower deactivation rate of these catalysts.
8% 1-C4, 38% trans-2-C4 and 44% cis-2-C4.
Ethylene oligomerization and metathesis. Another objective
1
On the Ni-AlSBA-15 catalyst, the initial conversion of ethylene was of this study was to investigate the direct conversion of ethylene
À3 À1
about 93% and the deactivation rate was 1.6 Â 10
h
(12 times to propylene in a single flow reactor, using two heterogeneous
6
lower than that of the Ni-AlSiO catalyst). The product distribution catalysts, through cascade dimerization, isomerization, and
2
was similar for Ni-AlSiO and Ni-AlSBA-15 catalysts.
metathesis reactions (Scheme 1).
2
The efficiency of the regeneration process was evaluated on
Two consecutive catalyst beds consisting of the Ni-based
the spent Ni-AlSiO
2
catalyst. It consisted of calcination at 550 1C catalyst and the Mo/support catalyst, respectively, were placed in
for 8 h with a mixture air-nitrogen (25/75, v/v) in the reactor. the reactor. Based on previous results obtained in oligomeriza-
1
3
The regenerated catalyst exhibited initial catalytic properties tion and metathesis reactions and preliminary experiments,
similar to those of the original catalyst, but the deactivation the following parameters were chosen for conducting the
À3 À1
rate slightly increased (23 Â 10
h
). Additional TG measure- oligomerization-metathesis reactions: T = 80 1C, p = 3.0 MPa
À1
ments under oxidizing conditions were carried out on the spent and the flow rate = 33 mL min of pure ethylene. WHSV was
À1
À1
catalysts in order to evaluate the amount of products confined 16.5 h for the oligomerization catalyst (0.15 g) and 5 h for
into the pores. As can be seen in Fig. 6, the total weight loss the metathesis catalyst (0.50 g). A moderate conversion of
was about 30%, consisting of two fractions: 22% in the range of ethylene is favorable, because it limits the formation of higher
5
3
0–300 1C (light and middle oligomers) and 8% in the range of oligomers in the oligomerization step, and increases the conver-
00–750 1C (heavy products). The TGA weight loss for the spent sion in the metathesis step. Different catalytic systems were
Ni-AlSBA-15 catalyst was only 10%. This major difference in tested and the results are summarized in Table 5. Under the
the weight loss exhibited by these two catalysts may explain why present reaction conditions the major olefins were propylene
Ni-AlSiO (having narrower and non-ordered pores) showed a and butenes (480%). Small amounts of C5 and C6+ olefins were
2
higher deactivation rate. On the other hand, the high stability also formed. The presence of C5 and C7 olefins indicates that the
of the Ni-AlSBA-15 catalyst can be related to the SBA-15 topology Mo-based catalysts were also able to catalyze metathesis reactions
where the large interconnected mesopores facilitate the diffusion involving olefins higher than C4. The best catalytic system in
terms of selectivity to propylene was the couple Ni-AlSBA-15 and
MoO /AlSBA-15. Moreover, the metathesis catalysts based on SBA-
x
15 type supports are better than those prepared on commercial
silica supports. MoO /AlSiO and MoO /SiO show low selectivity
x
2
x
2
to C3, suggesting that they are less active metathesis catalysts.
These results can be related to the properties of the metathesis
catalysts discussed above. Indeed, MoO
x
/AlSBA-15 contains highly
1
9,24
dispersed MoO (which are very active metathesis species),
x
Fig. 5 Ethylene conversion and product spectrum vs. time on stream on
Ni-AlSiO
2
. (B ) % ethylene conversion, (&) % C4, (n) % C6, (x) % C8 (Â| )
Scheme 1 Reaction sequence: ethylene dimerization, butene isomeriza-
tion and cross-metathesis.
À1
C10+; conditions: 150 1C, 3.0 MPa of ethylene, WHSV = 10 h
.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
Table 5 Catalytic behaviour in the ethylene oligomerization-metathesis Conclusion
reaction
We showed that a mesoporous non-ordered inexpensive Ni-
AlSiO catalyst can be successfully prepared by post-synthesis
After 1 h on stream After 3 h on stream
2
a
b
c
C2 conv., C3, C4, C2 conv., C3, C4,
alumination of commercial silica, followed by ion exchange
with nickel. This solid proved to be a versatile, robust and
efficient catalyst for the selective oligomerization of ethylene
under mild conditions. Thus, productivity up to 130 g of olefins
(C4–C10) per g of the catalyst per hour was obtained in the
batch mode, at 150 1C. Using a fixed-bed flow reactor, the
Catalyst couple
%
%
%
%
%
%
Ni-AlSBA-15 + MoO
Ni-AlSBA-15 + MoO
x
/SBA-15 41
/AlSBA-15 43
54 35 39
71 23 40
29 55
35 63
x
Ni-AlSiO
Ni-AlSiO
2
+ MoO
+ MoO
x
/SiO
/AlSiO
2
40
41
18 63 38
38 46 39
9
79
2
x
2
27 53
a
b
Ethylene conversion. Concentration of propylene in products. catalyst also showed good activity and stability. However, these
c
Concentration of butenes in products. Conditions: 80 1C, 3.0 MPa.
catalytic behaviours are to some extent inferior to those exhib-
ited by the expensive mesoporous Ni-AlSBA-15 catalyst. This
difference is due to the narrow pore size of the silica-based
catalyst, which suffered a more rapid deactivation with respect
to the mesoporous model material. These results clearly illus-
trate that the structural features of Ni catalysts strongly affect
the catalytic activity/stability in C2 conversion. By combining
while the other catalysts contain mainly polymerized octahedral
MoOx and even a small amount of bulk a-MoO , which are
3
known to be less active species in metathesis.
On the other hand, previous studies have reported on the
relationship between catalytic activity and acidity of the catalytic
These results suggested that similarly structured
MoO species exhibited higher activity when they were supported
18,30–33
materials.
the oligomerization catalysts with the MoO
x 2
/(Al)SiO or the
x
MoO /(Al)SBA-15 metathesis catalyst, in an one-pot cascade
x
on stronger acidic supports.
process, at 80 1C, ethylene was converted into propylene.
The effects of the time on stream on the C3 and C4 selectivity
profiles are given in Fig. 7 for the best catalytic couple. The initial
selectivity to propylene was 71%, which is very close to that
The metathesis behaviour was governed by the texture and
the acidity of the supports, as well as the nature of the MoO
x
species. Thus, the catalysts based on SBA-15 type supports
were again better than those prepared on commercial silica
obtained with the catalytic system Ni-AlSBA-15 + MoO
3
–SiO
in a previous study. In that case, the selectivity to
propylene decreased progressively from 70% to 46% at 5 h on
stream. In the present study, with Ni-AlSBA-15 + MoO /AlSBA-15
2
–
1
3
2 3
Al O
supports. The best catalyst was MoO /AlSBA-15, with highly
x
x
dispersed MoO species and surface acidity.
x
the selectivity to C3 decreased rapidly from 71% to 35% at 3 h on
stream, and 24% at 6 h on stream (Fig. 7). The selectivity to
butenes increased from 23% to 72%. These results indicate that
Acknowledgements
the activity of the metathesis catalyst diminished with time on We gratefully acknowledge François Fajula (ICGM) for fruitful
stream, while the oligomerization catalyst maintained its activity discussions. Eddy Petit (IEM, Montpellier) is acknowledged for
during the catalytic test. The low reaction temperature may performing Raman measurements. Philippe Gaveau (ICGM) is
contribute to the deactivation process by preventing fast product thanked for his assistance with the NMR spectroscopy.
desorption from the metathesis active sites.
In summary, the results reported here are promising in terms
of catalyst productivity and selectivity to C3–C4 olefins, but References
additional work is needed to optimize the metathesis catalyst
and the reaction conditions in order to improve the catalyst
lifetime and selectivity to propylene.
1
D. Vogt, in Applied Homogeneous Catalysis with Organo-
metallic Compounds, ed. B. Cornils and W. A. Herrmann,
Wiley-VCH, Weinheim, 2002, vol. 1, pp. 240–253.
2
3
A. Finiels, F. Fajula and V. Hulea, Catal. Sci. Technol., 2014,
4, 2412.
M. D. Heydenrych, C. P. Nicolaides and M. S. Scurrell,
J. Catal., 2001, 197, 49.
4
5
V. Hulea and F. Fajula, J. Catal., 2004, 225, 213.
M. Lallemand, A. Finiels, F. Fajula and V. Hulea, Stud. Surf.
Sci. Catal., 2007, 170, 1863.
6
7
8
9
R. D. Andrei, M. I. Popa, F. Fajula and V. Hulea, J. Catal.,
2015, 323, 76.
M. Tanaka, A. Itadani, Y. Kuroda and M. Iwamoto, J. Phys.
Chem. C, 2012, 116, 5664.
M. Lallemand, A. Finiels, F. Fajula and V. Hulea, J. Phys.
Chem. C, 2009, 113, 20360.
M. Lallemand, A. Finiels, F. Fajula and V. Hulea, Appl.
Catal., A, 2006, 301, 196.
Fig. 7 Selectivity to C3 (&), C4 (n), and conversion (B ) of C2 vs. time
on stream. Catalysts: Ni-AlSBA-15 + MoO
.0 MPa of ethylene.
x
/AlSBA-15; conditions: 80 1C,
3
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
1
1
1
1
0 M. Lallemand, O. A. Rusu, E. Dumitriu, A. Finiels, F. Fajula 22 C. C. Williams, J. G. Ekerdt, J. M. Jehng, F. D. Hardcastle,
and V. Hulea, Appl. Catal., A, 2008, 338, 37. A. M. Turek and I. E. Wachs, J. Phys. Chem., 1991, 95, 8781.
1 J. Heveling, A. Van Der Beek and M. De Pender, Appl. Catal., 23 J. P. Thielmann, G. Weinberg and C. Hess, ChemCatChem,
A, 1988, 42, 325. 2011, 3, 1814.
2 A. Lacarriere, J. Robin, D. Swierczynski, A. Finiels, F. Fajula, 24 D. P. Debecker, B. Schimmoeller, M. Stoyanova, C. Poleunis,
F. Luck and V. Hulea, ChemSusChem, 2012, 5, 1787.
P. Bertrand, U. Rodemerck and E. M. Gaigneaux, J. Catal.,
2011, 277, 154.
3 R. D. Andrei, M. I. Popa, F. Fajula, C. Cammarano, A. Al
Khudhair, K. Bouchmella, P. H. Mutin and V. Hulea, ACS 25 J. F. Moulder, W. F. Stickle, P. E. Sobol and K. Bomben, in
Catal., 2015, 5, 2774.
4 D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson,
B. F. Chmelka and G. D. Stucky, Science, 1998, 279, 548.
Handbook of X-ray Photoelectron Spectroscopy, ed. J. Chastain,
Perkin-Elmer Corporation (Physical Electronics), Eden Prairie,
MN, 2nd edn, 1992.
1
1
1
5 A. Galarneau, H. Cambon, F. Di Renzo, R. Ryoo, M. Choi 26 K. Amakawa, S. Wrabetz, J. Kroehnert, G. Tzolova-Muller,
and F. Fajula, New J. Chem., 2003, 73, 27.
6 A. Galarneau, M. Nader, F. Guenneau, F. Di Renzo and
A. Gedeon, J. Phys. Chem. C, 2007, 111, 8268.
R. Schloegel and A. Trunschke, J. Am. Chem. Soc., 2012,
134, 11462.
27 C. Carlini, M. Marchionna, A. M. Raspolli Galletti and
G. Sbrana, Appl. Catal., A, 2001, 206, 1.
17 E. L. Lee and I. E. Wachs, J. Phys. Chem. C, 2007, 111, 14410.
1
8 T. Hahn, U. Bentrup, M. Armbr u¨ ster, E. V. Kondratenko and 28 M. Helld o¨ rfer, W. Milius and H. G. Alt, J. Mol. Catal. A:
D. Linke, ChemCatChem, 2014, 6, 1664. Chem., 2003, 197, 1.
9 P. Topka, H. Balkar, J. Rathousky, N. Zilkova, F. Verpoort 29 F. Speiser, P. Braunstein and L. Saussine, Acc. Chem. Res.,
and J. Cejka, Microporous Mesoporous Mater., 2006, 96, 44. 2005, 38, 784.
0 S. Kim, K. Segawa, T. Soeya and I. E. Wachs, J. Catal., 1992, 30 B. Zhang, N. Liu, Q. Lin and D. Jin, J. Mol. Catal., 1991, 65, 15.
1
2
2
136, 539.
31 J. Handzlik, J. Ogonowski, J. Stoch and M. Mikołajczyk,
1 K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Appl. Catal., A, 2004, 273, 99.
Coordination Compounds, John Wiley & Sons, New York, 5th 32 D. P. Debecker, M. Stoyanova, U. Rodemerck and
edn, 1997; S. J. Palmer, R. L. Frost, G. Ayoko and T. Nguyen,
J. Raman Spectrosc., 2008, 39, 395.
E. M. Gaigneaux, J. Mol. Catal. A: Chem., 2011, 340, 65.
33 S. Lwin and I. E. Wachs, ACS Catal., 2014, 4, 2505.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.