Study of ethylene/2-butene cross-metathesis over W-H/Al2O 3 for propylene production: Effect of the temperature and reactant ratios on the productivity and deactivation

Article abstract of DOI:10.1016/j.jcat.2013.01.016

A highly active and selective catalyst based on supported tungsten hydride for the cross-metathesis between ethylene and 2-butenes yielding propylene has been investigated at low pressure with various temperatures and feed ratios. At low temperature (120 °C), the catalyst deactivates notably with time on stream. This phenomenon was extensively examined by DRIFTS, TGA, DSC, and solid-state NMR techniques. It was found that a large amount of carbonaceous species were formed due to a side-reaction such as olefin polymerization which took place simultaneously with the metathesis reaction. However, at 150 °C, the catalyst was stable with time and thereby gave a high productivity in propylene. Importantly, the slight increase in temperature clearly disfavored the side reaction. The ratio of ethylene to trans-2-butene was also studied, and surprisingly, the W-H/Al₂O₃ catalyst is stable and highly selective to propylene even at substoichiometric ethylene ratios.

Full text of DOI:10.1016/j.jcat.2013.01.016

Journal of Catalysis 301 (2013) 1–7
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Study of ethylene/2-butene cross-metathesis over W-H/Al2O3 for propylene
production: Effect of the temperature and reactant ratios on the productivity
and deactivation
Etienne Mazoyer a, Kai C. Szeto a, Nicolas Merle a, Sebastien Norsic a, Olivier Boyron a,
Jean-Marie Basset a,1, , Mostafa Taoufik a, , Christopher P. Nicholas b,
⇑
⇑
⇑
a Université Lyon 1, C2P2, CNRS UMR 5265, ICL, CPE Lyon, 43 Bd du 11 Novembre 1918, 69616 Villeurbanne, France
b Exploratory Catalysis Research, UOP LLC, a Honeywell Company, 25 East Algonquin Road, Des Plaines, IL 60017, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A highly active and selective catalyst based on supported tungsten hydride for the cross-metathesis
between ethylene and 2-butenes yielding propylene has been investigated at low pressure with various
temperatures and feed ratios. At low temperature (120 °C), the catalyst deactivates notably with time on
stream. This phenomenon was extensively examined by DRIFTS, TGA, DSC, and solid-state NMR tech-
niques. It was found that a large amount of carbonaceous species were formed due to a side-reaction such
as olefin polymerization which took place simultaneously with the metathesis reaction. However, at
150 °C, the catalyst was stable with time and thereby gave a high productivity in propylene. Importantly,
the slight increase in temperature clearly disfavored the side reaction. The ratio of ethylene to trans-
2-butene was also studied, and surprisingly, the W-H/Al2O3 catalyst is stable and highly selective to
propylene even at substoichiometric ethylene ratios.
Received 2 November 2012
Revised 17 January 2013
Accepted 17 January 2013
Keywords:
Heterogeneous catalysis
Alumina
Ethene
Propene
Ó 2013 Elsevier Inc. All rights reserved.
Tungsten hydrides
Surface organometallic chemistry
Cross-metathesis
1. Introduction
serve to accelerate the increase in propylene demand and its value
in the marketplace. The increasing demand for propylene therefore
Propylene supply has been a concern in recent years, mostly
owing to strong demand growth for polypropylene, propylene
oxide, and acrylonitrile [1–4]. Traditionally, propylene is produced
as a coproduct of steam or catalytic cracking. Much of the current
propylene production is therefore not ‘‘on purpose,’’ but as a by-
product of ethylene and gasoline production. This leads to difficul-
ties in coupling propylene production capacity with its demand in
the marketplace. Moreover, much of the additional steam cracking
capacity built in the near future will be based on using ethane as a
feedstock, which typically produces only ethylene as a final prod-
uct [5]. Although some hydrocarbons heavier than ethylene are
present, they are generally not produced in quantities sufficient
to allow for their recovery in an economical manner. In view of
the current high growth rate of propylene demand, this reduced
quantity of co-produced propylene from steam cracking will only
requires new ways for its production. The petrochemical industry
has reacted and technology developers have seen this potential
supply gap as an opportunity and have developed several ‘‘on-pur-
pose’’ propylene production technologies. These approaches in-
clude propane dehydrogenation [6–8], olefin metathesis [9–11],
and methanol to olefins [12–14]. All of these routes have been or
will soon be exploited commercially to varying degrees. Further-
more, there is an emphasis on the development of routes to
‘‘green’’ or ‘‘bio-propylene’’. The challenge of developing alterna-
tive technologies has become difficult because commercial tech-
nologies such as steam cracking and catalytic dehydrogenation
(UOP’s C3 Oleflex™ technology) are well established and continu-
ously subject to incremental improvements.
The production of propylene via olefin metathesis reaction is an
established method, having been in used in production for tens of
years, with a WO3/SiO2 catalyst used most frequently [11]. In re-
cent years, different attempts to upgrade C2 or C4 alkenes re-
sources to propylene via metathesis and to improve the existing
catalytic systems have been carried out [15–17]. Our laboratory
has been working extensively to develop alternative methods to
produce propylene by using alkene metathesis reactions [18–22].
In this context, we recently discovered a new catalytic reaction
⇑
Corresponding authors. Fax: +33 472431795 (M. Taoufik), +1 8473911287 (C.P.
Nicholas).
E-mail addresses: mostafa.taoufik@univ-lyon1.fr (M. Taoufik), Christopher.
Nicholas@uop.com (C.P. Nicholas).
1
Present address: KAUST Catalysis Center, King Abdullah University of Science and
Technology, Thuwal 23955–6900, Kingdom of Saudi Arabia.
0021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.01.016

2
E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
that transforms ethylene directly to propylene at low temperature
and pressure (150 °C, 1 bar) with a selectivity higher than 95% [23].
However, the tungsten hydride on alumina catalyst used here suf-
fers from catalyst deactivation with time on stream, leading to a
low yield of propylene if run without continuous regeneration. Re-
cent studies reveal that this same WH/Al2O3-(500) system is a stable
and highly active catalyst in the formation of propylene from linear
butenes at the same conditions (150 °C, 1 bar), but the selectivity
to propylene is moderate (55%) [24,25]. As the WH/Al2O3-(500) cat-
alyst system is active for both ethylene to propylene and butenes
to propylene and cross-metathesis of ethylene and 2-butenes is a
well-established reaction, we desired to investigate it in cross-
metathesis and compare this catalyst system to other known cata-
lysts (Scheme 1).
Scheme 2. Side reactions in ethylene and 2-butenes cross-metathesis.
500 °C prior to the grafting reaction. Excess molecular complex
was washed off with dry pentane, and the solid was dried under
high vacuum. The second step comprised a treatment of the latter
solid under H2 at 150 °C. The catalyst was then stored at ꢀ25 °C in
the glovebox. Characterization of the catalyst is consistent with
that described previously [36].
Multiple supported transition metal catalysts for alkene
metathesis have been reported and the most successful are based
on W, Mo, and Re. In particular, systems composed of highly loaded
Re2O7 supported on alumina or silica–alumina show promising ini-
tial activity and selectivity even at room temperature [26,27].
These systems suffer from unknown deactivation processes. Con-
versely, supported tungsten oxide catalysts are less active for
metathesis and generally require higher reaction temperatures
and pressure (>350 °C and 10 bar) [28], but are notably less expen-
sive and more robust than their rhenium and molybdenum coun-
terparts [29,30]. The formation of propylene from ethylene and
2-butenes is generally accompanied by several by-products (pent-
enes, hexenes, heptenes) due to disproportionation reactions of C4
olefins: (a) 2-butenes isomerization to 1-butene; (b) 1-butene self-
metathesis to form ethylene and 3-hexenes; (c) 1-butene and 2-
butenes cross-metathesis to form propylene and 2-pentenes
(Scheme 2). Optimal conditions to maximize the yield of propylene
are achieved by adjusting the temperatures, pressures, and ethyl-
ene/2-butenes ratio to minimize the side reactions [31–33].
In general, these supported transition metal oxide catalysts have
drawbacks such as unknown activation processes, significant deac-
tivation processes (metal reduction, coking), or require severe reac-
tion conditions (temperature and pressure) [34,35]. In an attempt
to overcome the disadvantages of the classical olefin metathesis
catalysts and to gain better molecular control over the catalyst ac-
tive site, we prepared the WH3/Al2O3-(500) system by the Surface
Organometallic Chemistry (SOMC) method. Herein, we study the
cross-metathesis between ethylene and trans-2-butene over
WH3/Al2O3-(500) to produce propylene at low temperature and pres-
sure as observed in the selective direct conversion of ethylene or
linear butenes to propylene. The performances of this catalyst have
been evaluated in a continuous flow reactor at 1 bar using different
temperatures and different reactant ratios in order to determine the
optimal propylene productivity, affording (i) the best conversion
rate, (ii) the highest selectivity in propylene, and (iii) the lowest
deactivation rate seen for this catalyst system.
2.2. Characterization
Infrared spectra were recorded on a Nicolet FTIR 6700 spectro-
photometer in diffuse reflectance mode, equipped with a MCT
detector. Typically, 64 scans with a resolution of 4 cmꢀ1 were ap-
plied for each spectrum. An air-tight cell with CaF2 window was
employed.
TGA experiments were performed in a Mettler Toledo TGA/
DSC1 thermobalance. About 10 mg of the sample was placed in
alumina crucible. The samples were heated from 35 °C to 800 °C
at 20 °C minꢀ1. Air at a flow rate of 30 ml minꢀ1 was used as the
carrier gas. DSC analysis was performed in a Mettler Toledo DSC
1 instrument, equipped with an auto-sampler. The temperature
and the heat flow of the equipment were calibrated with an indium
standard. The sample was accurately weighed (around 10 mg) and
heated from 40 °C to 160 °C at 10 °C minꢀ1 with an empty alumi-
num pan as reference. Two successive heating and cooling cycles
were performed, and only the second run was considered. Dry
nitrogen with a flow rate set at 30 mL minꢀ1 was used as the purge
gas. The melting temperature (Tm) was measured at the top of the
endothermic peak. The STARe thermal analysis software was used
for the calculation.
13C CP-MAS solid-state NMR spectra were collected on a Bruker
NMR AVANCE 300 spectrometer. The sample was filled in a zirco-
nia impeller of 4 mm and then transferred into the probe with a
rotation speed of 10 kHz.
Elemental analysis was performed at the Central Analysis Ser-
vice of the CNRS (Solaize, France) to determine the tungsten load-
ing (found: 5.5 wt.% W).
2.3. Catalyst evaluation
Catalytic performance of trans-2-butene (99% provided by
Scott) and ethylene (99.95% purchased from Air Liquide) conver-
sion was carried out in a stainless steel continuous flow reactor
(PC4H8+C2H4 = 1 bar, T = 120–200 °C, total flow rate = 20 mL minꢀ1
or VHSV = 5200 hꢀ1). The gases were purified with a column of
molecular sieve and activated Cu2O/Al2O3 and controlled by Brooks
mass flow controllers. The catalyst was charged in the glovebox. A
4-way valve allowed isolation of the charged catalyst in the reactor
from the environment and extensive purging of the tubes. The
products were determined by an online HP 6890 GC equipped with
50 m KCl/Al2O3 column and FID.
2. Materials and methods
2.1. Catalyst preparation
The catalyst (WH3/Al2O3-(500)) was prepared according to pub-
lished procedure reported elsewhere [36] and consisted of two
steps. The first step involved grafting W(„CC(CH3)3)(CH2C(CH3)3)3
[37] on c-alumina at 66 °C under argon to obtain a well-defined
Al-O-W(„CC(CH3)3)(CH2C(CH3)3)2 fragment on the surface. The
c
3. Results and discussion
-alumina (Johnson Matthey, 200 m2 gꢀ1) was dehydroxylated at
We first studied the cross-metathesis of ethylene/trans-2-
butene (ratio 1:1) at different temperatures. At 120 °C, the reaction
undergoes a steep maximal conversion rate of 6:6 molC4H8þC2H4
molꢀW1 minꢀ1 at the start of reaction (1 h) before reaching a pseudo
Scheme 1. Ethylene/trans-2-butene cross-metathesis.

E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
3
Fig. 1. (a) Total carbon conversion ( ) and cumulated TON ( ) vs. time on stream at 120 °C. (b) Product distribution.
plateau of 1:9 molC4¼þC2¼ molꢀ1 minꢀ1, yielding an overall turnover
nals centered at 38, 20, and 14 ppm are detected, indicating the
presence of a small degree of branched paraffin which is most likely
due to incorporation of 2-butenes in the polymer [34,40]. The TGA
profile of the used catalyst is presented in Fig. 3. It clearly indicates
that carbonaceous species are formed at the surface of the catalyst
during reaction. It is noteworthy that the amount of carbonaceous
deposit is 35% of catalyst weight after the 48 h on stream.
DSC fusion profile (Fig. S2a) reveals that the species formed
melt mostly at 129 °C which is similar to the fusion temperature
of linear HDPE [41]. The crystallization profile (Fig. S2b) confirms
the presence of mainly one type of carbonaceous deposit.
From this experiment, we see that propylene can be selectively
formed over the WH3/Al2O3 catalyst system (98% selectivity) with
high productivity by cross-metathesis of ethylene and trans-2-
butene at 120 °C. The initiation and propagation steps of this
reaction were elucidated by identifying the products formed while
heating to the reaction temperature of 120 °C at 1 bar and 1:1 ratio
ethylene/trans-2-butene. It is reasonable to postulate that the ini-
tiation of the catalytic process occurs with a similar mechanism to
that previously described for the direct conversion of ethylene to
propylene [23]: in the beginning of the reaction, we noted that
mainly ethane was released. It can be assumed that three ethylene
molecules insert into the tungsten tris-hydride precursor, affording
a tungsten tris(ethyl) species, [W][(CH2CH5)3]. This selectivity can
be explained by rapid ethylene insertion in the W-H bond, which is
energetically more favorable rather than insertion of sterically hin-
dered disubstituted olefins [42,43] (2-butenes), as described by the
W
number, TON, of 9000 after 48 h (Fig. 1a). A very high selectivity in
propylene of 98.7% is observed. The other products of the reaction
are pentenes (1.1%) and hexenes (0.2%) (Fig. 1b). The propylene
production rate for this reaction is thus 118:0 mmolC3¼ gꢀ1 hꢀ1
cata
at maximum and 32:5 mmolC3¼ gꢀ1 hꢀ1 after 48 h on stream. In
cata
contrast to the results obtained in the direct transformation of eth-
ylene to propylene (TON = 500 after 48 h) where the deactivation
rate is quite high (7.5% hꢀ1) [23], the catalyst in the beginning
(until 10 h) undergoes a fast deactivation with a rate of 1.7% hꢀ1
(Fig. 1a). At the end of the catalytic reaction, the deactivation is lin-
ear (0.23% hꢀ1) and the recovered catalyst is agglomerated rather
than as fine powder. Therefore, the catalyst used in the experiment
has been studied after 48 h on stream by DRIFT, solid-state NMR,
TGA, and DSC.
The DRIFT spectrum (Fig. 2) of the fresh catalyst exhibits bands at
3500–3800 cmꢀ1, 2750–3000 cmꢀ1, 1360–1465 cmꢀ1 corresponding
to m(OH) from alumina, m(CH) and d(CH) from residual hydrocarbon,
respectively. Finally, the m(WH) band is presented at 1913 cmꢀ1 [36].
The DRIFT spectrum of the used catalyst after 48 h on stream
shows a complete disappearance of the m(WH) mode, indicating that
all the W-H have been inserted by alkenes. Simultaneously, three
intense bands centered at 2923, 2852 and 1465 cmꢀ1, correspond-
ing to typical m(CH) and d(CH) absorptions for long-chain alkyl groups
were observed. This spectrum is similar to polyethylene obtained
by polymerization of ethylene on supported metal hydrides [38].
The formation of long-chain alkyls is further confirmed by the so-
lid-state NMR study (Fig. S1). A major peak at 32 ppm along with
a shoulder around 30 ppm is observed, assigned respectively to
amorphous and crystallized polyethylene [39]. Moreover, weak sig-
Cosse-Arlman mechanism [44,45]. Ethane formation is due to an
abstraction mechanism giving an ethyl-ethylidene species,
[W](@CHCH3)((CH2CH3) (Scheme 3).
a-
H
Fig. 2. DRIFT spectra of the fresh catalyst and used catalyst after 48 h on stream at 120 °C.

4
E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
after 48 h (Fig. 4a). In contrast to the results obtained at 120 °C
(TON = 9000 after 48 h), the conversion profile in the beginning is
quite linear with a negligible deactivation rate of only 0.03% hꢀ1
.
Similar selectivity to propylene (99%) is observed at 120 °C.
Trace amounts of pentenes and hexenes are also detected
(Fig. S3). The propylene production rate for this reaction is thus
133:4 mmolC3¼ gꢀ1 hꢀ1 at maximum and 97:9 mmolC3¼ gꢀ1 hꢀ1
cata
cata
after 48 h on stream. The used catalyst was also analyzed by TGA
where no significant weight loss was observed (Fig. 3) confirming
that deactivation by polymerization at 150 °C is negligible in accor-
dance with the slow deactivation rate with respect to the same
experiment carried out at 120 °C. This result shows that the poly-
merization by insertion of ethylene into the metal alkyl bond of the
active tungsten alkyl-alkylidene species is disfavored at higher
temperatures, as documented in the literature [40,47].
Fig. 3. TGA profile of the used catalyst after 48 h on stream at 120 °C, 150 °C, and
200 °C.
When the reaction is performed at 200 °C, the initial conversion
rate ð10:2 molC4¼þC2¼ molꢀ1 minꢀ1Þ is higher than that observed at
W
150 °C (Fig. 4) with a similar selectivity in propylene (99%, see
Fig. S4) giving
a high initial productivity of 180:9 mmolC3¼
hꢀ1. However, the conversion profile undergoes a linear deac-
ꢀ1
g
cata
tivation rate of 0.30% hꢀ1 which is 10-fold greater than the 150 °C
experiment giving an overall TON of 22000 with a propylene pro-
ductivity of 98:5 mmolC3¼ gꢀ1 hꢀ1 after 48 h on stream. Note that
the selectivity to propylene at the different temperatures studied
cata
Scheme 3. Initialization step by ethylene insertion.
(120, 150, and 200 °C) is fairly consistent around 99%.
The TGA analysis performed on the used catalyst (Fig. 3) does
not show a significant amount of carbonaceous species suggesting
that the observed deactivation does not arise from an ethylene
polymerization process as in the 120 °C experiment. We have pre-
viously shown that the highly oxophilic-supported metal species
are unstable at high temperature (P 150 °C) and consequently
change their coordination spheres either by opening MAOAM
bridges (M = Si, Al) [48] or by formation of unsaturated hydrocar-
bons bonded to the metal (cyclopentadienyl tuck-in complexes,
carbyne) to give inactive species in the conversion of alkanes and
olefins [49–51].
Our current experiments show that while propylene selectivity
is relatively unaffected by reaction temperature, catalyst stability
and propylene productivity are significantly affected by the reac-
tion temperature. At 120 and 200 °C, the catalyst undergoes a nota-
ble deactivation though we have shown that
2 different
mechanisms are likely responsible for the deactivation. On the
other hand, at 150 °C, the WH3/Al2O3 catalyst system is extremely
stable with time on stream. The overall turnover number achieved
after 48 HOS increases as the reaction temperature increases, pri-
marily due to the increased conversion level obtained in the
metathesis portion of the reaction. This suggests that overall yield
of propylene can be optimized through selection of reaction
parameters as the TON at 150 °C will eventually be greater than
that at 200 °C due to the lower deactivation rate.
Most importantly, excellent selectivity and productivity toward
propylene can readily be obtained over WH3/Al2O3 as catalyst even
at low temperature and 1:1 ratio of ethylene and trans-2-butene.
Typically, an excess of ethylene is utilized [52] in addition to high
temperature and elevated pressure in order to achieve high selec-
tivity to propylene for classical supported catalysts based on tung-
sten oxide. Hence, the same experiment (PC4H8+C2H4 = 1 bar,
T = 150 °C, total flow rate = 20 mL minꢀ1 or VHSV = 5200 hꢀ1) was
performed by using the classical heterogeneous catalyst [53,54]
based on tungsten oxide supported on silica (WO3/SiO2) synthe-
sized and activated as shown by Lokhat, et al. No significant activ-
ity was observed. The present result shows that the WH/Al2O3-(500)
catalyst is clearly more active and selective (99% in propylene) for
ethylene/2-butene cross-metathesis than the classical catalyst
WO3/SiO2 at low temperature and pressure. However, our working
Scheme 4. Intermediates for the production of propylene from ethylene and trans-
2-butene.
This newly formed carbene [W](@CHCH3)(CH2CH3) then under-
goes metathesis reaction with ethylene to yield propylene and a
new carbene, tungsten-methylidene-ethyl species [W](@CH2)(CH2-
CH3), following the classical Chauvin mechanism (A) [46]. The
reaction of trans-2-butene with the tungsten methylidene-ethyl
species gives a metallacyclobutene intermediate (A0) that by
decomposition liberates propylene (Scheme 4).
Alternatively, the highly electrophilic tungsten alkyl-alkylidene
species can undergo multiple insertions of ethylene at the tungsten
r
-alkyl bond to finally result in polyethylene probably responsible
for deactivation of the catalyst as described in the literature [34]. It
is generally known that higher temperatures decrease catalyst
activity in ethylene polymerization [40,47]. Hence, we studied
the effect of the temperature on the stability of the catalyst during
the cross-metathesis reaction between ethylene and trans-
2-butene.
The reaction of WH3/Al2O3-(500) with ethylene and trans-2-
butene at 150 °C in the ratio 1:1 undergoes a steep maximal con-
version rate of 7:4 molC4H8þC2H4 molꢀ1 minꢀ1 at the start of reaction
W
(1 h) before reaching
a pseudo plateau of 5:4 molC4¼þC2¼
molꢀW1 minꢀ1, yielding an overall turnover number, TON, of 16000

E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
5
Fig. 4. Total carbon conversion ( ) and cumulated TON ( ) vs. time on stream for ethylene/trans-2-butene metathesis at (a) 150 °C and (b) 200 °C.
conditions (low temperature and pressure) are very different from
the industrial conditions (350 °C, 35 bar).
the dominant reaction mechanism [24] but that ethylene coordi-
nates strongly and gives cross-metathesis products when it enters
the reaction cycle. Increasing the ethylene content to 18% slightly
improves overall conversion and again significantly increases the
propylene selectivity to 90%. With an ethylene content of 35%,
the overall conversion level is the highest observed. Given these re-
sults, as the concentration of ethylene in the feed exceeds above
10%, ethylene/2-butene cross-metathesis becomes the dominant
reaction. We are unable to rule out ethylene dimerization to 1-bu-
tene followed by isomerization to 2-butene, as in the ethylene to
propylene mechanism, as a reaction pathway here. But, given that
the overall conversion level and deactivation rate for cross-metath-
esis are consistent with the butenes to propylene data published
previously [24,25] rather than ethylene to propylene data [23],
and that the moles of ethylene converted (9.8 mmol hꢀ1) are con-
Given the differences in reactivity and deactivation rates seen in
conversion of ethylene to propylene [23] and 2-butene to propyl-
ene [24] as well as to obtain the maximum selectivity and produc-
tivity in propylene, the influence of the proportion of ethylene in
the feed during the cross-metathesis reaction between ethylene
and trans-2-butene catalyzed by WH3/Al2O3 was then studied to
optimize the reaction conditions and its influence on deactivation
rate. The study has been conducted at the same conditions
(PC4H8+C2H4 = 1 bar, T = 150 °C, total flow rate = 20 mL minꢀ1 or
VHSV = 5200 hꢀ1) by varying the proportion of ethylene from 0%,
10%, 18%, 35%, 45% to 70% mol in trans-2-butene feed.
The selectivity to propylene increases with the proportion of
ethylene in the feed, starting from 55% when no ethylene was in
the feed to 99% when 70% of ethylene was added (Fig. 5). The max-
imum propylene production rate during the cross-metathesis reac-
tion is obtained when the reactor is fed with 35% of ethylene in the
trans-2-butene, with 142:8 mmolC3¼ gcata hꢀ1 (Fig. 6). The maximal
sistent with the moles of propylene produced (19.5 mmol hꢀ1
)
according to the 2:1 mol ratio of the cross-metathesis reaction
(Scheme 1), the extent of dimerization (pathway for direct conver-
sion of ethylene to propylene) is limited under these conditions.
When the ethylene proportion in the feed exceeds 35%, the con-
version rate and the productivity to propylene decrease even
though the selectivity remains constant (>98%). At these higher
ethylene content levels, the productivity of propylene is also higher
than that of either the butenes to propylene [24,25] or ethylene to
propylene [23] reaction showing that cross-metathesis of ethylene
and 2-butenes is predominant. As previously noted, ethylene
dimerization is the slowest reaction occurring in the system at
150 °C [23]. Additionally, we have previously shown that the con-
centration of 1-butene in the 2-butene feed affects both selectivity
to propylene (as well as pentenes and hexenes) and productivity
conversion rates span from 5.0 to 9:2 molC4H8þC2H4 molꢀ1 minꢀ1
,
W
whereas the deactivation rate is quite constant for all ratios: ca.
0.11% hꢀ1. (Fig. 7). At low ethylene content in the feed, notable
amounts of pentenes are observed. These products originate from
2-butenes isomerization on tungsten-hydride moiety to 1-butene
followed by cross-metathesis as highlighted in Scheme 2 [24].
A detailed analysis of the latter results shows that the addition
of a small amount of ethylene (10%) to the trans-2-butene feed
does not increase the overall conversion (Fig. 7), but does signifi-
cantly improve the selectivity to propylene, increasing it from
55% to 79% (Fig. 5), thus implying that butenes to propylene is still
Fig. 6. Productivity in propylene at 20 h on stream for ethylene/trans-2-butene
metathesis on WH/Al2O3-(500) at 150 °C vs. ethylene molar percentage in the trans-
2-butene feed.
Fig. 5. Selectivity at 20 h on stream for ethylene/trans-2-butene metathesis on WH/
Al2O3-(500) at 150 °C vs. ethylene molar percentage in the trans-2-butene feed.

6
E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
Fig. 7. Total carbon conversion vs. time on stream for ethylene/trans-2-butene metathesis on WH/Al2O3-(500) at 150 °C for different feed compositions 0%, 10%, 18%, 35%, 45%,
and 70% mol. of ethylene in the 2-butene feed.
[24]; here, we do not observe a decreased selectivity to propylene
nor an increase in heavy olefins, suggesting that the dimerization
and isomerization reactions are limited.
References
[1] F. Cavani, N. Ballarini, A. Cericola, Catal. Today 127 (2007) 113.
[2] Z. Nawaz, F. Wei, J. Ind. Eng. Chem. 16 (2010) 774.
[3] J.S. Plotkin, Catal. Today 106 (2005) 10.
[4] A.H. Tullo, Chem. Eng. News 81 (2003) 15.
[5] D. Seddon, Petrochemical Economics: Technology Selection in
Constrained World, Imperial College Press, London, 2010.
[6] K. Fukudome, N.-o. Ikenaga, T. Miyake, T. Suzuki, Catal. Sci. Technol. 1 (2011)
987.
[7] M.S. Kumar, D. Chen, A. Holmen, J.C. Walmsley, Catal. Today 142 (2009) 17.
[8] J. Schaeferhans, S. Gomez-Quero, D.V. Andreeva, G. Rothenberg, Chem. Eur. J.
17 (2011) 12254.
[9] P. Amigues, Y. Chauvin, D. Commereuc, C.T. Hong, C.C. Lai, Y.H. Liu, J. Mol. Catal.
65 (1991) 39.
[10] S. Liu, X. Li, W. Xin, S. Xie, P. Zeng, L. Zhang, L. Xu, J. Nat. Gas Chem. 19 (2010)
482.
[11] J.C. Mol, J. Mol. Catal. A 213 (2004) 39.
[12] A. Sassi, M.A. Wildman, H.J. Ahn, P. Prasad, J.B. Nicholas, J.F. Haw, J. Phys. Chem.
B 106 (2002) 2294.
[13] S. Svelle, U. Olsbye, F. Joensen, M. Bjørgen, J. Phys. Chem. C 111 (2007) 17981.
[14] W. Wu, W. Guo, W. Xiao, M. Luo, Chem. Eng. Sci. 66 (2011) 4722.
[15] X. Zhu, X. Li, S. Xie, S. Liu, G. Xu, W. Xin, S. Huang, L. Xu, Catal. Surv. Asia 13
(2009) 1.
The lowered propylene productivity at greater than 35% ethyl-
ene content in the feed is probably due to the stronger coordina-
tion of ethylene (smallest and most reactive of the alkenes) than
2-butenes on the active sites, thereby limiting the probability of
the catalyst to undergo a productive cross-metathesis reaction to
produce propylene. Degenerate metathesis pathways are not indi-
cated in Scheme 4, but with a high concentration of ethylene in the
feed, the probability of undergoing non-productive degenerate
metathesis to continuously yield an ethylidene species is high.
Even though the production of propylene from ethylene/2-butene
cross-metathesis is a stoichiometric reaction, having a lower than
stoichiometric ethylene to 2-butenes ratio allows for the highest
productivity of propylene over the W-H/Al2O3 catalyst system be-
cause the relative rates of productive and non-productive metath-
esis are best matched.
a Carbon-
[16] D.P. Debecker, K. Bouchmella, M. Stoyanova, U. Rodemerck, E.M. Gaigneaux,
P.H. Mutin, Catal. Sci. Technol. 2 (2012) 1157.
[17] D.P. Debecker, M. Stoyanova, F. Colbeau-Justin, U. Rodemerck, C. Boissiere,
E.M. Gaigneaux, C. Sanchez, Angew. Chem. Int. Ed. 51 (2012) 2129.
[18] J.M. Basset, J. Thivolle-Cazat, M. Taoufik, E. Le Roux, C. Coperet, Pat. Appl.
FR2872509, 2006.
[19] C. Coperet, J.M. Basset, J. Thivolle-Cazat, E. Le Roux, M. Taoufik, Pat.
Appl.WO2006013251, 2006.
[20] C.P. Nicholas, E. Mazoyer, M. Taoufik, J.M. Basset, P.T. Barger, Pat. Appl.
US20110245569, 2011.
[21] C.P. Nicholas, E. Mazoyer, M. Taoufik, J.M. Basset, P.T. Barger, J.E. Rekoske, Pat.
Appl. US20110245570, 2011.
[22] M. Taoufik, E. Mazoyer, C.P. Nicholas, J.M. Basset, Pat. Appl. US20110245560,
2011.
[23] M. Taoufik, E. Le Roux, J. Thivolle-Cazat, J.-M. Basset, Angew. Chem. Int. Ed. 46
(2007) 7202.
[24] E. Mazoyer, K.C. Szeto, J.-M. Basset, C.P. Nicholas, M. Taoufik, Chem. Commun.
48 (2012) 3611.
4. Conclusion
Propylene can be obtained selectively and in high productivity
by cross-metathesis between ethylene and 2-butenes catalyzed
by WH3/Al2O3 prepared by the SOMC method. In contrast to the
classical system (WO3/SiO2), the present catalyst yields exceptional
productivity of propylene at low temperature and pressure.
Unprecedentedly, the highest propylene productivity is obtained
at less than stoichiometric ratios of ethylene in the trans-2-butene
feed. The current system may have a remarkable economical and
energetic impact since a lower amount of ethylene is required.
The challenge we will pursue is to fully investigate the catalyst life-
time in an industrial context and the regeneration under hydrogen
to recover the initial WH3/Al2O3 catalyst.
[25] E. Mazoyer, K.C. Szeto, S. Norsic, A. Garron, J.-M. Basset, C.P. Nicholas, M.
Taoufik, ACS Catal. 1 (2011) 1643.
[26] J.C. Mol, Catal. Today 51 (1999) 289.
[27] M. Stoyanova, U. Rodemerck, U. Bentrup, U. Dingerdissen, D. Linke, R.W.
Mayer, H.G.J.L. Rotgerink, T. Tacke, Appl. Catal. A 340 (2008) 242.
[28] Y.D. Wang, Q.L. Chen, W.M. Yang, Z.K. Xie, X. Xu, D. Huang, Appl. Catal. A 250
(2003) 25.
[29] W. Grunert, R. Feldhaus, K. Anders, E.S. Shpiro, K.M. Minachev, J. Catal. 120
(1989) 444.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2013.01.016.

E. Mazoyer et al. / Journal of Catalysis 301 (2013) 1–7
7
[30] A. Montreux, F. Petit, Industrial Applications of Homogeneous Catalysis, Reidel,
Dordreicht, 1988.
[31] S. Huang, F. Chen, S. Liu, Q. Zhu, X. Zhu, W. Xin, Z. Feng, C. Li, Q. Wang, L. Xu, J.
Mol. Catal. A 267 (2007) 224.
[32] S.L. Liu, S.J. Huang, W.J. Xin, J. Bai, S.J. Xie, L.Y. Xu, Catal. Today 93–5 (2004)
471.
[42] H. Malani, S. Hayashi, H. Zhong, R. Sahnoun, H. Tsuboi, M. Koyama, N.
Hatakeyama, A. Endou, H. Takaba, M. Kubo, C.A. Del Carpio, A. Miyamoto, Appl.
Surf. Sci. 254 (2008) 7608.
[43] X. Li, J. Guan, A. Zheng, D. Zhou, X. Han, W. Zhang, X. Bao, J. Mol. Catal. A 330
(2010) 99.
[44] E.J. Arlman, P. Cossee, J. Catal. 3 (1964) 99.
[33] R. Thomas, J.A. Moulijn, J. Mol. Catal. 15 (1982) 157.
[34] X. Li, W. Zhang, X. Li, S. Liu, H. Huang, X. Han, L. Xu, X. Bao, J. Phys. Chem. C 113
(2009) 8228.
[45] P. Cossee, J. Catal. 3 (1964) 80.
[46] J.L. Herisson, Y. Chauvin, Makromol. Chem. 141 (1971) 161.
[47] S. Matsui, T. Fujita, Catal. Today 66 (2001) 63.
[35] A. Spamer, T.I. Dube, D.J. Moodley, C. van Schalkwyk, J.M. Botha, Appl. Catal. A
255 (2003) 133.
[48] G. Saggio, A. de Mallmann, B. Maunders, M. Taoufik, J. Thivolle-Cazat, J.M.
Basset, Organometallics 21 (2002) 5167.
[36] E. Le Roux, M. Taoufik, C. Coperet, A. de Mallmann, J. Thivolle-Cazat,
J.M. Basset, B.M. Maunders, G.J. Sunley, Angew. Chem. Int. Ed. 44 (2005)
6755.
[49] E. Le Roux, M. Chabanas, A. Baudouin, A. de Mallmann, C. Coperet, E.A.
Quadrelli, J. Thivolle-Cazat, J.M. Basset, W. Lukens, A. Lesage, L. Emsley, G.J.
Sunley, J. Am. Chem. Soc. 126 (2004) 13391.
[37] D.N. Clark, R.R. Schrock, J. Am. Chem. Soc. 100 (1978) 6774.
[38] G. Tosin, C.C. Santini, J.-M. Basset, Top. Catal. 52 (2009) 1203.
[39] G. Nowaczyk, S. Glowinkowski, S. Jurga, Solid State Nucl. Magn. Reson. 25
(2004) 194.
[40] C. Forte, M. Hayatifar, G. Pampaloni, A.M.R. Galletti, F. Renili, S. Zacchini, J.
Polym. Sci. A: Polym. Chem. 49 (2011) 3338.
[50] F. Rataboul, M. Chabanas, A. de Mallmann, C. Coperet, J. Thivolle-Cazat, J.M.
Basset, Chem. Eur. J. 9 (2003) 1426.
[51] S. Soignier, M. Taoufik, E. Le Roux, G. Saggio, C. Dablemont, A. Baudouin, F.
Lefebvre, A. de Mallmann, J. Thivolle-Cazat, J.M. Basset, G. Sunley, B.M.
Maunders, Organometallics 25 (2006) 1569.
[52] Q. Zhao, S.-L. Chen, J. Gao, C. Xu, Trans. Met. Chem. 34 (2009) 621.
[53] H. Liu, S. Huang, L. Zhang, S. Liu, W. Xin, L. Xu, Catal. Commun. 10 (2009) 544.
[54] D. Lokhat, M. Starzak, M. Stelmachowski, Appl. Catal. A 351 (2008) 137.
[41] M. Trujillo, M.L. Arnal, A.J. Mueller, E. Laredo, S. Bredeau, D. Bonduel, P. Dubois,
Macromolecules 40 (2007) 6268.