Effect of 2-butene Cis/Trans isomers in the metathesis of ethylene and 2-butene over WO3/SiO2 catalysts

Article abstract of DOI:10.1007/s10562-014-1230-8

Pure 2 % trans-2-butene and the mixture of 1 % cis- and 1 % trans-2-butene were used as feed to react with ethylene in metathesis reaction in order to produce propylene to study the effect of feed on WO₃/SiO₂ catalysts with various tungsten oxide loadings (5, 7, 9, and 11 wt%), which were prepared by incipient wetness impregnation. These catalysts were characterized by X-ray diffraction, nitrogen adsorption (BET), scanning electron microscopy, energy dispersive X-ray spectroscopy, ion-exchange titration, FT-Raman, ammonia temperature programmed desorption (NH₃-TPD) and reactant-TPD. At high tungsten oxide loading, when the mixture of cis- and trans-2- butene was used as the reaction feed the conversion of 2-butene was lower than when pure trans-2-butene was used. However, the effect of cis-2-butene isomer on the conversion of 2-butene was less pronounced at low tungsten oxide loading. The dispersion of tungsten was a key factor to improve the activity of the tungsten catalysts when the mixture of cis- and trans-2-butene was used as a feed for metathesis reaction. Springer Science+Business Media New York 2014.

Full text of DOI:10.1007/s10562-014-1230-8

Catal Lett (2014) 144:920–927
DOI 10.1007/s10562-014-1230-8
Effect of 2-Butene Cis/Trans Isomers in the Metathesis of Ethylene
and 2-Butene Over WO3/SiO2 Catalysts
•
Narongrat Poovarawan Kongkiat Suriye
•
•
Sirachaya Kunjara Na Ayudhya Joongjai Punpranot
•
•
Francisco Jose Cadete Santos Aires Piyasan Praserthdam
´
Received: 20 November 2013 / Accepted: 26 February 2014 / Published online: 19 March 2014
Ó Springer Science+Business Media New York 2014
Abstract Pure 2 % trans-2-butene and the mixture of
1 % cis- and 1 % trans-2-butene were used as feed to react
with ethylene in metathesis reaction in order to produce
propylene to study the effect of feed on WO3/SiO2 catalysts
with various tungsten oxide loadings (5, 7, 9, and 11 wt%),
which were prepared by incipient wetness impregnation.
These catalysts were characterized by X-ray diffraction,
nitrogen adsorption (BET), scanning electron microscopy,
energy dispersive X-ray spectroscopy, ion-exchange titra-
tion, FT-Raman, ammonia temperature programmed
desorption (NH3-TPD) and reactant-TPD. At high tung-
sten oxide loading, when the mixture of cis- and trans-2-
butene was used as the reaction feed the conversion of
2-butene was lower than when pure trans-2-butene was
used. However, the effect of cis-2-butene isomer on the
conversion of 2-butene was less pronounced at low tung-
sten oxide loading. The dispersion of tungsten was a key
factor to improve the activity of the tungsten catalysts
when the mixture of cis- and trans-2-butene was used as a
feed for metathesis reaction.
Keywords Metathesis ꢀ Cis-2-butene ꢀ Trans-2-butene ꢀ
Propylene ꢀ WO3/SiO2
1 Introduction
The demand for propylene is growing rapidly in the
chemical industry. Propylene is mainly produced from
naphtha steam crackers [1], fluid catalytic crackers [2],
propane dehydrogenation [3], and Fischer–Tropsch chem-
istry [4]. It is used for making polypropylene, acrylonitrile,
alcohol, acrylic acid, and other petrochemical products.
Recently, metathesis reaction over WO3/SiO2 catalyst has
received considerable attention by the industry to produce
propylene [5]. The greatest advantage of WO3/SiO2 cata-
lyst when compared to Re- and Mo-based metathesis cat-
alysts, is that it is more robust [6–9], Westhoff and Moulijn
[10], Zhao et al. [11], and Spamer et al. [12] investigated
the influence of tungsten loading on the activity and
selectivity of the WO3/SiO2 catalysts in the metathesis of
ethylene and 2-butene using pure trans-2-butene feed.
They found that the conversion and the selectivity reach a
maximum at 8 % and higher tungsten oxide loadings and
the catalysts with loadings lower than 8 % tungsten oxide
loading have a faster deactivation.
N. Poovarawan ꢀ J. Punpranot ꢀ P. Praserthdam (&)
Center of Excellence on Catalysis and Catalytic Reaction
Engineering, Department of Chemical Engineering, Faculty of
Engineering, Chulalongkorn University, Bangkok 10330,
Thailand
The composition of reaction feed containing ethylene
and n-butene used in the metathesis reaction for commer-
cial-scale production of propylene has to be precisely
controlled in order to obtain both propylene yield and long
catalyst cycle time. The main problem resides in the fact
that the composition of 2-butene in terms of cis- and trans-
ratio is generally poorly controlled. Engelhardt and Zsinka
[13] found a higher rate of cross-metathesis between
1-butene and 2-butene in the feed containing trans-2-
butene reaction feed than that of the one containing cis-2-
e-mail: piyasan.p@chula.ac.th
N. Poovarawan ꢀ F. J. C. S. Aires
Institut de Recherches sur la Catalyse et l’Environnement de
´
Lyon, UMR 5256, CNRS/Universite Lyon1, 2 Avenue Albert
Einstein, 69626 Villeurbanne Cedex, France
K. Suriye ꢀ S. K. N. Ayudhya
SCG Chemicals Co., Ltd., 1 Siam Cement Road, Bangsue,
Bangkok 10800, Thailand
123

Effect of 2-Butene Cis/Trans Isomers
921
butene. This was attributed to the fact that cis-2-butene
converts easier to trans-2-butene than trans-2-butene con-
verts to cis-2-butene; so trans-2-butene gives a higher rate
of cross metathesis. However, the effect of cis- and trans-
isomer ratio in the feed has not been understood yet.
Therefore, an influence of cis- and trans-isomer in the feed
mixture of 2-butene over the WO3/SiO2 catalyst having
different tungsten oxide loading would be investigated in
this study.
equipped with a LaB6 gun and a Pentafet LinK-Isis EDX
spectrometer) was also used as a local probe to visualize
the morphology and distribution [14–18] of the tungsten
particles on the silica support.
Molecular structure of the tungsten oxide was deter-
mined by using Raman spectroscopy. The Raman spectra
were collected by projecting a continuous wave YAG laser
of Nd (810 nm) through the samples at room temperature
and were recorded using a PerkinElmer Spectrum GX
spectrometer.
Acidity of catalyst was determined by ammonia tem-
perature programmed desorption (NH3-TPD) and ion-
exchange titration. For NH3-TPD, 0.1 g of the catalyst was
treated at the temperature not greater than 873 K in a flow
of helium then cooled down to room temperature and sat-
urated with 15 % NH3/He mixture. After purging with
helium at room temperature for 1 h in order to remove
weakly physisorbed NH3, the sample was heated from
room temperature to 773 K under helium flow. NH3-TPD
was performed by using a Micrometrics Chemisorbs 2750
automated system equipped with ChemiSoft TPx software.
Ammonia effluent was finally kept in a boric acid solution
(20 g/dm3) and titrated with a 0.01 N aqueous HCl solu-
tion. The number of Brønsted acid sites was also estimated
by using a method involving an aqueous ion-exchange step
of catalyst’s H? ions with Na? ions, followed by titration
of the resulting solution [19].
2 Experimental
2.1 Preparation of WO3/SiO2 Catalysts
Silica gel, Davisil grade 646 was used as the support
(supplied by Aldrich). Catalysts with different tungsten
oxide loadings (5, 7, 9, and 11 wt%) were prepared by
incipient wetness impregnation of an aqueous solution of
ammonium metatungstate hydrate (Aldrich, 99.9 %) over
silica gel. The catalysts were dried at room temperature for
2 h and then 383 K for overnight. The calcination was
performed under air at temperature sufficiently removing
the residue during the synthesis in a tubular furnace.
2.2 Reaction Studies
Fixed bed reactor with 3 g of catalyst was used for per-
forming the catalytic test. The catalysts were activated at
the temperature not greater than 873 K for 2 h under
nitrogen then cooled down to reaction temperature at
673 K. Two types of reactant feed which are (i) 2 % trans-
2-butene with excess ethylene (balancing with nitrogen)
and (ii) 1 % trans-2-butene and 1 % cis-2-butene with
excess ethylene (balancing with nitrogen) were used in this
study in order to investigate the effect of cis- and trans-
isomer on the catalytic performance of metathesis reaction.
Gas product obtained from the reaction was analyzed by
using FID-gas chromatography (Shimadzu 2014).
Reactant-TPD was used to determine the adsorption
behavior of the reactant. Catalyst sample was treated at the
temperature not greater than 873 K in a flow of nitrogen for
1 h, then cooled down to room temperature and saturated
with the reactant. After purging with nitrogen at room
temperature for 1 h to remove weakly physisorbed reac-
tant, the sample was heated from room temperature
to 673 K under a flow of nitrogen (30 cm3/min). The
resulting gaseous desorption was sampled every 2 min and
analyzed by using a FID-Shimadzu 2014.
3 Results and Discussion
2.3 Catalyst Characterization
3.1 Characteristics of WO3/SiO2 with Various
Tungsten Oxide Loadings
Crystallite phase and size were investigated by X-ray dif-
fraction, XRD (Siemens D5000) using Ni filter Cu Ka
radiation. Specific surface area (SBET) was determined
from nitrogen adsorption studies by using Micrometrics
Chemisorbs 2750.
XRD diffractograms of the SiO2 support and the WO3/SiO2
catalysts containing various amounts of tungsten are shown
in Fig. 1. Only a broad peak at 2h = 22.5° was observed
for the SiO2 support indicating the amorphous phase of
silica [20]. The characteristic peaks at 2h = 23.12°, 23.60°,
and 24.38° which corresponds to crystalline tungsten oxide
(orthorhombic WO3) [21] were observed for all WO3/SiO2
catalyst. The XRD peak intensity of the crystalline phase of
tungsten increased with increasing tungsten oxide loading
The catalyst morphology and tungsten-distribution over
the silica support were investigated by scanning electron
microscopy (SEM) together with energy dispersive X-ray
(EDX) spectroscopy (JEOL JSM-5800LV SEM equipped
with a Pentafet LinK-Isis series 300 EDX spectrometer).
Transmission electron microscopy (JEOL JEM 2010 TEM
123

922
N. Poovarawan et al.
11 wt%). Similar result was reported by Li and colleagues
[22] that the XRD peaks of WO3 crystalline phase in WO3/
SiO2 catalysts could be detected only when tungsten oxide
loading was higher than 5 wt%.
TEM observation of the catalyst with 9 wt% (Fig. 2)
clearly shows that the tungsten is dispersed on the surface
of SiO2 support as large agglomerates. A closer inspection
by high resolution TEM of these agglomerates (Fig. 2c)
reveals that their structure is an orthorhombic WO3 which
is consistent with the XRD results.
Figure 3 shows the typical SEM micrographs with the
EDXmappingoftungstenfor thevariousWO3/SiO2 catalysts.
The EDX mapsclearly show that tungsten is welldispersed all
over the catalyst grains surfaces for all the samples. The dis-
persion of tungsten decreases with increasing tungsten oxide
loading as determined by SEM–EDX.
Fig. 1 XRD diffractograms of the support and the WO3/SiO2
catalysts having various % tungsten oxide loadings: (a) SiO2 support,
(b) 5 wt% WO3/SiO2, (c) 7 wt% WO3/SiO2, (d) 9 wt% WO3/SiO2
and (e) 11 wt% WO3/SiO2
indirectly indicating that the dispersion of tungsten specie
decreased with increasing tungsten oxide loading. The
optimum tungsten loading and calcination temperature for
catalyst preparation were determined to be 9 wt% and 773
K, respectively [12]. The results suggested that tungsten
crystal species were very small crystal and well dispersed
on the surface of silica support, which were invisible by
XRD, for the catalyst sample with low tungsten oxide
loading (5 wt%) [11] and the dispersion of tungsten
decreased with increasing tungsten oxide loading (7, 9, and
The average amounts of tungsten on the catalyst grains
as determined by SEM–EDX are shown in Table 1 and
were found to be in good agreement with the targeted
amount of tungsten oxide loading varying from 5 to 11
wt% ( 1 wt%). The surface areas of the SiO2 support and
the WO3/SiO2 containing various tungsten oxide loadings
are given in Table 1. The surface areas of the WO3/SiO2
catalysts (*180–182 m2/g) were not significantly different
comparing to the SiO2 support (*185 m2/g).
a
b
d
(1 1 2 )
~65
(0 2 0 )
orthorhombic WO3
[2 0 -1]
c
d(0 2 0) 0.38 nm
Fig. 2 TEM images of the 9 wt% WO3/SiO2 catalyst: a general
clearly detected, and d 2D-FFT of particle in b revealing d(0
interplane distances (white circle) as well as very faint d(1
interplane distances planes (grey circle) indicating that the particle is
oriented in the [2 0 -1] direction
2
0)
2)
overview of the WO3 particles on the SiO2 support, b one single-
crystalline particle, c high resolution detail of a particle in b where
only showing d(0 2 0) interplane distances of orthorhombic WO3 are
1
123

Effect of 2-Butene Cis/Trans Isomers
923
Fig. 3 SEM images (a–d) and
corresponding EDX W-M line
mapping (e–h) over the WO3/
SiO2 catalysts having various %
tungsten oxide loadings: a, e
5 wt% WO3/SiO2, b, f 7 wt%
WO3/SiO2, c, g 9 wt% WO3/
SiO2, and d, h 11 wt% WO3/
SiO2
The structures of tungsten species on the surface of
catalyst samples were investigated by FT-Raman as shown
in Fig. 4. Raman bands at 707–720 and 803–808 cm-1
were assigned to be O–W–O bending modes and W–O
stretching modes, respectively [23]. These Raman bands
represented the tungsten octahedral of the bulk crystalline
WO3 [23]. The band at 960–980 cm-1 was assigned to be
terminal W=O bonds of the isolated surface tetrahedral
tungsten oxide species (i.e., at the crystal surface) [21, 23].
Increasing of tungsten oxide loadings lead to the increase
123

924
N. Poovarawan et al.
Table 1 The specific surface area, relative intensities ratio of the
Raman band between tetrahedral tungsten species and octahedral
tungsten species (I970/I805), and the amount of tungsten on the cata-
lyst’s surface of the various WO3/SiO2 catalysts and the SiO2 support
Samples
Support 5 wt% 7 wt% 9 wt% 11 wt%
WO3/
SiO2
WO3/
SiO2
WO3/
SiO2
WO3/
SiO2
The surface area
(BET, m2/g)
184.6
na
182.2
0.38
4.03
181.0
0.35
5.77
180.0
0.34
7.54
179.7
0.34
I970/I805 (Raman
spectroscopy)
Fig. 5 NH3-TPD profile of WO3/SiO2 catalyst having various %
tungsten oxide loadings: (a) 5 wt% WO3/SiO2, (b) 7 wt% WO3/SiO2,
(c) 9 wt% WO3/SiO2, and (d) 11 wt% WO3/SiO2
The amount of
tungsten on the
surface of
na
10.96
catalyst (SEM–
EDX, %)
Table 2 The amount of acidity estimated from NH3-TPD and ion-
exchange over the various WO3/SiO2 catalysts
Samples
5 wt%
WO3/SiO2 WO3/SiO2 WO3/SiO2 WO3/SiO2
7 wt%
9 wt%
11 wt%
805 (cm-1
)
Acidity from NH3-TPDa
970(cm-1
)
e
d
c
b
a
lmol NH3/g catalyst
250
235
3.9
177
2.2
157
1.6
mmol NH3/g tungsten 6.3
Acidity from NH3-TPDb
lmol NH3/g catalyst
230
215
3.6
164
2.1
143
1.4
mmol NH3/g tungsten 5.8
Acidity from ion exchange
lmol H?/g catalyst
mmol H?/g tungsten
Pseudo Lewis aciditya
lmol H?/g catalyst
mmol H?/g tungsten
Pseudo Lewis acidityb
lmol H?/g catalyst
mmol H?/g tungsten
a
65
80
95
98
1.6
1.3
1.2
1.0
1100
1000
900
800
700
600
500
400
Raman Shift (cm-1)
185
4.6
155
2.6
82
60
1.0
0.6
Fig. 4 FT-Raman spectra of the SiO2 support and the WO3/SiO2
catalyst having various % tungsten oxide loadings: (a) SiO2 support,
(b) 5 wt% WO3/SiO2, (c) 7 wt% WO3/SiO2, (d) 9 wt% WO3/SiO2
and (e) 11 wt% WO3/SiO2
165
4.1
135
2.3
69
46
0.9
0.5
Acidity from area of NH3 TPD
b
of the intensities of peak at 804 and 710 cm-1, obviously.
Such a result was similar to that of Li and colleagues [22].
The FT-Raman results were in accordance to the XRD
results that more crystalline WO3 was formed at higher
tungsten oxide loading. According to Huang et al. [21], the
tetrahedral tungsten oxide species are the active sites for
the metathesis reaction of ethylene and 2-butene to produce
propylene. The ratios of the relative intensities of the FT-
Raman bands at 970 and 805 cm-1 (I970/I805) reflect the
relative contents of metathesis-active sites on the WO3/
SiO2 catalyst [24]. It was found that the ratios of the rel-
ative intensities of FT-Raman band between 970 and
805 cm-1 for all the catalysts were essentially the same. In
other words, although the tungsten active sites increased
with increasing tungsten oxide loading, the crystallite size
of tungsten oxide being formed also increased as suggested
by the XRD results, resulting in no changes in the relative
content of active sites.
Acidity from titration of H3BO3 with HCl
The catalysts acidity was determined by the NH3-TPD
and the ion-exchange titration methods. The total acidity
was based on the NH3-TPD results. The Brønsted acidity,
which is typically attributed to the doubly bonded oxygen
atom (W=O) on the surface of the crystalline WO3 [25],
was determined by ion-exchange titration. The Lewis
acidity, which is generally attributed to two tungsten atoms
bonded to an oxygen atom (W–O–W) within the crystalline
WO3 [26], was determined from the difference between the
acidity deduced from NH3-TPD and the acidity deduced
from ion-exchange titration method. The NH3-TPD profiles
are shown in Fig. 5. All the WO3/SiO2 catalysts exhibited
only one peak in the NH3-TPD profiles within the tem-
perature range of 298–773 K. 11 wt% WO3/SiO2 catalyst
showed a stronger acidity than the other WO3/SiO2 cata-
lysts as shown by a shift of the desorption peak towards
123

Effect of 2-Butene Cis/Trans Isomers
925
Fig. 6 Correlation between 2-butene conversion and acidity from ion
exchange over the WO3/SiO2 catalyst: filled squares conversion for
the pure 2 % trans-2-butene with excess ethylene feed and filled
triangles conversion for the mixture of 1 % cis- and 1 % trans-2-
butene with excess ethylene feed
Fig. 8 2-Butene conversion and propylene selectivity over the WO3/
SiO2 catalyst having various % tungsten oxide loadings: filled
triangles with continuous line conversion, filled triangles with dotted
line selectivity for the pure 2 % trans-2-butene with excess ethylene
feed and filled squares with continuous line conversion, filled squares
with dotted line selectivity for the mixture of 1 % cis- and 1 % trans-
2-butene with excess ethylene feed
3.2 Effect of 2-Butene Cis–Trans Isomers
in the Metathesis of Ethylene and 2-Butene Over
WO3/SiO2 Catalysts
A gas phase metathesis reaction of ethylene and 2-butene
was carried out in a fixed bed continuous flow reactor at
673 K under atmospheric pressure. The n-butene reactants
used were either pure 2 % trans-2-butene (balancing with
N2) or the mixture of 1 % cis- and 1 % trans-2-butene
(balancing with N2). Conversions of 2-butene and propylene
selectivity as a function of % tungsten oxide loading over the
various WO3/SiO2 catalysts after 10 h time-on-stream are
shown in Fig. 8. At higher % tungsten oxide loading from 5
to 11 wt%, the conversion of 2-butene increased from 12.50
to 70.00 % for the metathesis testing with pure 2 % trans-2-
butene and excess ethylene feed whereas it increased from
12.00 to 40.00 % for those testing with the mixture of 1 %
cis- and 1 % trans-2-butene with excess ethylene. Both cases
show that there were no significant improvements in term of
2-butene conversion when % tungsten oxide loading was
increased from 9 to 11 wt%. A similar trend was reported by
Zhao et al. [11] in which the conversion of 2-butene reached
an optimum point at 9 wt% tungsten oxide loading. This is
attributed to the presence of a pseudo-amorphous surface
phase, which would have a minor contribution compared to
the crystalline WO3 phase, for loadings higher than 9 wt%.
Moreover, at this tungsten oxide loading, tetrahedral tung-
sten oxide species (i.e., at the crystal surface) reached their
maximum level.
Fig. 7 Correlation between propylene selectivity and acidity from
ion exchange over the WO3/SiO2 catalyst: filled squares conversion
for the pure 2 % trans-2-butene with excess ethylene feed and filled
triangles conversion for the mixture of 1 % cis- and 1 % trans-2-
butene with excess ethylene feed
higher temperature. Total acidity was calculated based on
the integrated peak areas of the NH3-TPD curves for each
catalyst. In addition, titration with 0.01 M aqueous HCl
solution was employed to confirm the acidity measured for
each WO3/SiO2 catalyst. As summarized in Table 2, total
Lewis acidity was found to decrease with increasing
tungsten oxide loading whereas the total Brønsted acidity
increased with increasing tungsten loading, according with
the linear correlation between metathesis conversion and
acidity sites as shown in Fig. 6. The slope of correlation of
catalysts from pure 2 % trans-2-butene feed was steeper,
indicating that conversion of ones from pure 2 % trans-2-
butene has been influenced by acidity sites more than that
of catalysts from mixture feed. In addition, the similar
trend of linear correlation between propylene selectivity as
shown in Fig. 7 was also observed for all catalysts.
The 2-butene conversion and propylene yield of 5 wt%
WO3/SiO2 catalysts were rather similar over both of feeds,
which are pure 2 % trans-2-butene with excess ethylene
123

926
N. Poovarawan et al.
Table 3 2-Butene conversion, propylene selectivity, yield, and % difference over the various WO3/SiO2 catalysts tested with pure 2 % trans-2-
butene with excess ethylene feed and the mixture of 1 % cis- and 1 % trans-2-butene with excess ethylene feed
Reactants
Trans-2-butene:cis-2-butene (2:0)
Trans-2-butene:cis-2-butene (1:1)
A % difference
of conversione
A % difference
of yieldf
Conversiona
Selectivity
Yieldc
Conversionb
Selectivity
Yieldd
5 wt% WO3/SiO2
7 wt% WO3/SiO2
9 wt% WO3/SiO2
11 wt% WO3/SiO2
12.50
45.00
65.00
70.00
69.00
50.00
49.00
51.00
8.63
22.50
31.85
35.70
12.00
34.00
40.00
40.00
71.70
66.00
62.00
61.00
8.60
22.44
24.80
24.40
4.00
24.44
38.46
42.86
0.24
0.27
22.14
31.65
amount of butene in feedꢁamount of butene remained in products
Conversion ¼
Selectivity ¼
ꢂ 100
ꢂ 100
amount of butene in feed
amount of propylene
amountof butene in feedꢁamount of butene remained in products
amount of propylene
amount of butene in feed
Yield ¼
ꢂ 100
e
aꢁb
% Difference of conversion ¼
ꢂ 100
a
f
cꢁd
% Difference of yield ¼
ꢂ 100
c
Table 4 Reactant-TPD result summarizing the amount of ethylene, cis-2-butene and trans-2-butene adsorbed over various WO3/SiO2 catalysts
dosed with pure 2 % trans-2-butene with excess ethylene feed and the mixture of 1 % cis- and 1 % trans-2-butene with excess ethylene feed
Trans-2-butene:cis-2-butene (2:0)
Trans-2-butene:cis-2-butene (1:1)
% Difference
of reactant
desorption
Ethylenea
Trans-2-buteneb
(lmol/g cat)
Ethylenec
Trans-2-butened
(lmol/g cat)
Cis-2-butenee
(lmol/g cat)
(lmol/g cat)
(lmol/g cat)
5 wt% WO3/SiO2
7 wt% WO3/SiO2
9 wt% WO3/SiO2
11 wt% WO3/SiO2
9.16
26.70
39.54
47.45
110.06
175.04
229.21
275.05
8.45
19.40
27.63
34.40
73.33
78.40
82.60
85.92
37.00
44.20
48.00
50.30
0.37
29.61
41.13
47.10
% Difference of reactant desorption ¼ ðaþbÞꢁðcþdþeÞ ꢂ 100
ðaþbÞ
feed and mixture of 1 % cis-, 1 % trans-2-butene with
excess ethylene feed. Differences in 2-butene conversion
and propylene yield were seen when tungsten oxide load-
ing increased to 7 wt% and higher loading. For tung-
sten oxide loading C7 wt%, the use of the pure 2 % trans-
2-butene with excess ethylene feed resulted in higher
2-butene conversion than those of use of the mixture of
1 % cis- and 1 % trans-2-butene with excess ethylene feed.
Table 3 summarizes % difference of conversion between
the use of pure 2 % trans-2-butene with excess ethylene
feed and the use of the mixture of 1 % cis- and 1 % trans-
2-butene with excess ethylene feed. The % differences
increased with increasing tungsten oxide loading. In the
other words, for metathesis reaction, when the dispersion of
tungsten over the SiO2 support was fairly low, the presence
of the mixture between cis- and trans-2-butene isomers had
a strong effect on 2-butene conversion. Such effect was not
markedly observed when the catalyst contained low tung-
sten oxide loading (i.e., 5 wt% WO3/SiO2). Propylene
yield for the metathesis testing with pure 2 % trans-2-
butene with excess ethylene feed was higher than the
testing with the mixture of cis/trans-2-butene feed espe-
cially 9 and 11 wt% loading.
The temperature programmed desorption of the reactants
(reactant-TPD) was performed in order to elucidate the
adsorption/desorption behavior of pure 2 % trans-2-butene
and excess ethylene as well as the mixture of 1 % cis- and
1 % trans-2-butene with excess ethylene over the WO3/SiO2
catalyst. The amounts of ethylene, trans-2-butene, and cis-2-
butene adsorbed in the WO3/SiO2 catalyst that were esti-
mated by the means of reactant-TPD are summarized in
Table 4. It was found that the adsorption of the reactants
increased with increasing tungsten oxide loading. The linear
correlation between reactant desorption and acid sites is
shown in Fig. 9, agreed with correlation between metathesis
activity and acidity sites as mentioned above.
In the case of higher tungsten dispersion over SiO2
support (5 % WO3/SiO2 catalyst), there was no significant
difference of the total amount of the adsorbed reactants
between the pure 2 % trans-2-butene with excess ethylene
and the mixture of 1 % cis- and 1 % trans-2-butene with
excess ethylene. However, the difference was more
123

Effect of 2-Butene Cis/Trans Isomers
927
350
300
250
200
150
100
50
tungsten is low. This is in a good agreement with reactant-
TPD in the result of lower propylene yield over higher %
tungsten oxide loading of WO3/SiO2 catalyst.
Acknowledgments The financial supports from the Thailand
Research Fund (TRF) and the Office of Higher Education Commis-
sion are gratefully acknowledged. The authors would like to thank the
Royal Golden Jubilee PhD Scholarship from TRF and SCG Chemi-
cals for Narongrat Poovarawan.
0
60
70
80
90
100
References
Acidity from ion exchange (µmol/g-catalyst)
1. Grootjans J, Vanrysselberghe V, Vermeiren W (2005) Catal
Today 106:57–61
2. Ladwig PK, Asplin JE, Stuntz GF, Wachter WA, Henry BE, Fung
SC (2004) Patent Application Publication: US 6,803,494B1
3. Nawaz Z, Tang X, Zhang Q, Wang D, Fei W (2009) Catal
Commun 10:1925–1930
Fig. 9 Correlation between temperature programmed desorption of the
reactants (reactant-TPD) and acidity from ion exchange over the WO3/
SiO2 catalyst: filled squares reactant-TPD for the pure 2 % trans-2-
butenewith excess ethylene feed and filled trianglesreactant-TPDfor the
mixture of 1 % cis- and 1 % trans-2-butene with excess ethylene feed
¨
4. Botes FG, Bohringer W (2004) Appl Catal A 267:217–225
5. Mol JC (2004) J Mol Catal A 213:39–45
6. Spamer A, Dube TI, Moodley DJ, van Schalkwyk C, Botha JM
(2003) Appl Catal A 255:133–142
7. Ivin KJ, Mol JC (1997) Olefin metathesis and metathesis poly-
merization. Academic Press, San Diego
pronounced over the WO3/SiO2 catalyst sample having
higher tungsten oxide loadings which means lower tung-
sten dispersion over SiO2 support. This is in a good
agreement with the metathesis testing results.
8. Moffat AJ, Johnson MM, Clark A (1970) J Catal 18:345–348
9. Luckner RC, Wills GB (1973) J Catal 28:83–91
10. Westhoff R, Moulijn JA (1977) J Catal 46:414–416
11. Zhao Q, Chen SL, Gao J, Xu C (2009) Transit Met Chem
34:621–627
12. Spamer A, Dube TI, Moodley DJ, van Schalkwyk C, Botha JM
(2003) Appl Catal A 255:153–167
13. Engelhardt J, Zsinka I (1985) J Mol Catal 28:169–182
Both cis- and trans-2-butene adsorb on metathesis active
sites, however, the results in this study showed much lower
conversion of 2-butene when the mixture of cis/trans-2-
butene were used as the feed. The reactant-TPD records
that the presence of cis-2-butene led to lower amount of
reactant adsorb. At higher loading (tungsten oxide loading
more than 7 wt%), the lower dispersion led to agglomer-
ation of tungsten as seen from TEM in Fig. 2. According to
the lower tungsten dispersion, the active sites positions
were close to each other and this obstructed cis/trans-2-
butene absorption. Cis/trans-2-butene could access to the
active sites difficultly, leading to the lower conversion. The
effect of 2-butene cis/trans isomers became less pro-
nounced at lower tungsten oxide loading (better tungsten
dispersion) because the active sites positions were faraway.
Therefore, it was proposed here that tungsten dispersion
affected the reactant adsorption and catalyst activities.
´
14. Cadete Santos Aires FJ, Sautet Ph, Fuchs G, Rousset JL, Melinon
P (1993) Microsc Microanal Microstruct 4:441–452
15. Rousset JL, Cadete Santos Aires FJ, Bornette F, Cattenot M,
Pellarin M, Stievano L, Renouprez AJ (2000) Appl Surf Sci
164:163–168
16. Garcia Cervantes G, Cadete Santos Aires FJ, Bertolini JC (2003)
J Catal 214:26–32
´
17. Mahfouz R, Cadete Santos Aires FJ, Brenier A, Ehret E, Roumie
M, Nsouli B, Jacquier B, Bertolini JC (2010) J Nanopart Res
12:3123–3136
´
18. Pattamakomsan K, Ehret E, Morfin F, Gelin P, Jugnet Y, Prakash
S, Bertolini JC, Panpranot J, Cadete Santos Aires FJ (2011) Catal
Today 164:28–33
´
19. Lopez DE, Goodwin JG Jr, Bruce DA (2007)
245:381–391
20. Yu X, Wang M, Li H (2000) Appl Catal A 202:17–22
J Catal
4 Conclusions
21. Huang S, Chen F, Liu S, Zhu Q, Zhu X, Xin W, Feng Z, Li C,
Wang Q, Xu L (2007) J Mol Catal A 267:224–233
22. Lu G, Li X, Qu Z, Wang Y, Chen G (2008) Appl Surf Sci
255:3117–3120
23. Baserga A, Russo V, Di Fonzo F, Bailini A, Cattaneo D, Casari CS, Li
Bassi A, Bottani CE (2007) Thin Solid Films 515:6465–6469
24. Limsangkass W, Phatanasri S, Praserthdam P, Panpranot J, Ja-
reewatchara W, Kunjara Na Ayudhya S, Suriye K (2013) Catal
Lett 143:919–925
Effect of 2-butene cis/trans isomers on the metathesis
reaction of ethylene and 2-butene to propylene over WO3/
SiO2 catalysts having various % tungsten oxide loadings
was investigated in this study. With higher % tung-
sten oxide loading, the relative content of metathesis-active
sites on the WO3/SiO2 catalyst was still maintained while
the dispersion of tungsten was lower. It was discovered that
the cis/trans isomers could obstruct each other accessing to
the metathesis-active site, especially when the dispersion of
25. Sun C, Dong L, Yu W, Liu L, Li H, Gao F, Dong L, Chen Y
(2011) J Mol Catal A 346:29–38
´
26. Gora A, Brocławik E (2004) J Mol Catal A 215:187–193
123