The enhancing effect of Bronsted acidity of supported MoO x species on their activity and selectivity in ethylene/trans-2- butene metathesis

Article abstract of DOI:10.1002/cctc.201400040

Supported catalysts with a nominal Mo surface density of 0.15 and 1.5 Mo atoms nm^-2 were synthesized by impregnation of alumina, silica, and alumina-silica supports with silica content between 1 and 70 wt %. They were tested for their activity and selectivity in the metathesis of ethylene and trans-2-butene to propene between 343 and 603 K at 125 kPa. The catalysts were characterized by UV/Vis, Raman, and IR spectroscopy, XRD and H₂ temperature-programmed reduction for elucidating the distribution, degree of polymerization, reducibility, and acidity of MoO_x species. We established that Bronsted acidity of highly dispersed tetrahedral and polymerized octahedral MoO_x species is required to ensure high metathesis activity. The acidic character of these species is influenced by their structure and support. Tetrahedral MoO_x species with Bronsted acidic character are only formed on supports possessing such acidity, whereas Bronsted acidic octahedral MoO_x is also created on supports without such acidic sites. Acidity meets structure: The activity and selectivity of MoO_x species on different supports in the metathesis of ethylene and trans-2-butene to propene are systematically elucidated and correlated to their structural and acidic properties. Tetrahedral and octahedral MoO_x species exhibit enhanced performance with increasing Bronsted acidity.

Full text of DOI:10.1002/cctc.201400040

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201400040
The Enhancing Effect of Brønsted Acidity of Supported
MoOx Species on their Activity and Selectivity in Ethylene/
trans-2-Butene Metathesis
Tobias Hahn,[a] Ursula Bentrup,[a] Marc Armbrꢀster,[b] Evgenii V. Kondratenko,*[a] and
David Linke*[a]
Supported catalysts with a nominal Mo surface density of 0.15
and 1.5 Mo atomsnmꢀ2 were synthesized by impregnation of
alumina, silica, and alumina–silica supports with silica content
between 1 and 70 wt%. They were tested for their activity and
selectivity in the metathesis of ethylene and trans-2-butene to
propene between 343 and 603 K at 125 kPa. The catalysts
were characterized by UV/Vis, Raman, and IR spectroscopy,
XRD and H2 temperature-programmed reduction for elucidat-
ing the distribution, degree of polymerization, reducibility, and
acidity of MoOx species. We established that Brønsted acidity
of highly dispersed tetrahedral and polymerized octahedral
MoOx species is required to ensure high metathesis activity.
The acidic character of these species is influenced by their
structure and support. Tetrahedral MoOx species with Brønsted
acidic character are only formed on supports possessing such
acidity, whereas Brønsted acidic octahedral MoOx is also creat-
ed on supports without such acidic sites.
Introduction
Propene metathesis to ethylene and 2-butenes over heteroge-
neous catalysts has been discovered in the middle of the last
century by Banks and Baily[1] and commercialized by Phillips
Petroleum Co.[2] Currently, this reaction is industrially per-
formed in the reverse direction, that is, metathesis of ethylene
and 2-butenes to propene, owing to the continuously increas-
ing demand for propene.[3] ABB Lummus offers such technolo-
gy using a silica-supported tungsten oxide catalyst operating
above 523 K.[4]
OH groups adjacent to metal oxides, for example, MoOx, WOx,
or ReOx, may be required for the generation of metathesis
active carbene species.[9] Very recently, Amakawa et al.[10] per-
formed a thorough mechanistic analysis of propene metathesis
on silica-supported (SBA-15) highly dispersed MoOx species.
These authors concluded that their Brønsted acidity is an es-
sential requirement for the in situ formation of metathesis
active metal carbene sites.[11] However, the SBA-15 support
does not possess any acidic sites. Therefore, it is unclear if and
how differently acidic supports influence the formation of
Brønsted acidic MoOx species. Further, the derived conclusion
for propene metathesis cannot be simply transferred to the
metathesis of ethylene and 2-butenes to propene, because 2-
to 1-butene isomerization can additionally occur on acidic sites
both of bare support and MoOx species and, thus, influence
their activity and selectivity.[12]
From an energetic viewpoint, supported catalysts with mo-
lybdenum oxide are promising candidates, because they oper-
ate at lower temperatures than tungsten-based catalysts.[3a,5]
Their activity is influenced by the degree of polymerization of
surface MoOx species. Isolated,[3a,6] dimeric[6c,d] MoOx species
and dispersed polymolybdates[6b,d] catalyze the target reaction
whereas crystalline MoO3 is catalytically inactive.[6d,e,7]
In addition to the nature of MoOx species, acidic properties
of catalysts also affect the metathesis activity. For example,
similarly structured MoOx species supported either on alumina
or silica exhibit lower activity than those on stronger acidic
supports consisting of alumina and silica.[8] Some previous
studies on propene metathesis claimed that Brønsted acidic
Against this background, the aim of the present work was
to systematically study the effect of Brønsted and Lewis acidity
of bare supports on the kind of formed MoOx species as well
as the influence of 1) the structure of MoOx species and 2) the
Brønsted and Lewis acidity of MoOx and supports on the cata-
lytic performance in metathesis of ethylene and trans-2-butene
to propene and in 2- to 1-butene isomerization. For this pur-
pose, we used commercially available alumina–silica supports
(Siralꢁ) with silica content varying from 1 to 70 wt% as well as
pure alumina and silica. According to previous studies,[7c,13] the
Mo loading was not higher than 1.5 Monmꢀ2 to minimize the
formation of catalytically inactive MoO3. Structural characteris-
tics of supported MoOx species were elucidated by UV/Vis and
Raman spectroscopy, XRD, and hydrogen temperature-pro-
grammed reduction (H2–TPR). Brønsted and Lewis acidic sites
both of bare supports and catalysts were characterized by IR
[a] T. Hahn, Dr. U. Bentrup, Dr. E. V. Kondratenko, Dr. D. Linke
Leibniz-Institut fꢀr Katalyse e.V. an der Universitꢁt Rostock (LIKAT)
Albert-Einstein-Str. 29a, 18059 Rostock (Germany)
E-mail: Evgenii.Kondratenko@Catalysis.de
David.Linke@Catalysis.de
[b] Dr. M. Armbrꢀster
Max-Planck-Institut fꢀr Chemische Physik fester Stoffe
Nçthnitzer Straße 40, 01187 Dresden (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201400040.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
spectroscopy using pyridine as basic probe molecule. Metathe-
sis of ethylene and trans-2-butene was investigated at a pres-
sure of 125 kPa under continuous-flow conditions in the tem-
perature range from 343 to 603 K.
spectively, as determined by integrating the corresponding IR
bands at 423 K. For better comparability, the values were nor-
malized to the SBET. Brønsted acidic sites were only detected on
Siral40 and Siral70. As such sites are created by charge equali-
zation in the alumina–silica network by protons,[8a,15a,17] an opti-
mal ratio between Al and Si is obviously required to generate
Brønsted acidity.[15b,16] With decreasing alumina content, the
Lewis acidity diminishes and approaches zero for pure silica as
previously reported for other alumina–silica materials.[14a,15,18]
This effect can be explained by a decreasing concentration of
coordinative unsaturated Al sites, which are responsible for the
Lewis acidity.[15,19]
Results
Surface properties
The specific surface areas (SBET) of bare supports and catalytic
materials are listed in Table 1 and Table 2, respectively. There is
only a slight decrease in SBET of the supports after Mo deposi-
tion up to 1.5 Monmꢀ2, that is, the support morphology has
persisted.
Generally, the deposition of MoOx on different supports
changes the surface acidity (Table 2). In the case of
0.15 Monmꢀ2 catalysts, 0.15MoS40 and 0.15MoS70 feature
Brønsted acidity higher than their corresponding bare sup-
ports. Upon increasing Mo loading to 1.5 Monmꢀ2, additional
Brønsted acidic sites were generated in the catalysts based on
supports with silica content of 10, 40, 70, and 100 wt%
(Table 2). The Mo loading also influences the amount of Lewis
acidic sites. Compared to unloaded alumina, Siral1 and Siral10,
The acidic properties of the pure supports and the support-
ed catalysts were characterized by IR spectroscopy of adsorbed
pyridine. The typical band of pyridine fixed to Brønsted acid
sites is observed at approximately 1540 cmꢀ1,[14] and a band at
approximately 1450 cmꢀ1 results from pyridine interacting with
Lewis acid sites.[14b,15] Representative spectra of pyridine ad-
sorbed on 1.5 Monmꢀ2 catalysts are shown in Figure SI-1 in
the Supporting Information. Table 1 and Table 2 contain the
calculated amounts of pyridine adsorbed on Brønsted and
Lewis acidic sites of bare supports and supported catalysts, re-
the corresponding catalysts with
a
Mo loading of
0.15 Monmꢀ2, exhibited a decrease in Lewis acidity. Contrarily,
an increase in Lewis acidity was observed for the catalysts
Table 1. Selected physicochemical properties and the suppliers of bare supports.
[a]
Sample
Nominal silica
content
[wt%]
SBET
Overall surface
acidity[b]
Brønsted
acidity[b]
Lewis
Supplier
Lot No.
acidity[b]
[m2 gꢀ1
]
[10ꢀ3
]
[10ꢀ3
]
[10ꢀ3
]
Alumina (Al)
Siral1 (S1)
Siral10 (S10)
Siral40 (S40)
Siral70 (S70)
Silica (Si)
0
1
10
40
70
100
155
260
350
475
360
315
26.13
22.58
16.74
15.09
9.22
0.00
0.00
0.00
2.82
3.92
0.00
26.13
22.58
16.74
12.27
5.31
Chempur
Sasol
Sasol
Sasol
Sasol
000178
S2497
S2849
S13581
S15053
07421 LZ
0.00
0.00
Aldrich
[a] Measured by BET method after drying and calcining at 773 K for 8 h in static air. [b] The acidity values were calculated from IR spectra of adsorbed
pyridine at 423 K.
Table 2. Selected physicochemical and catalytic properties of materials tested.
Sample
wMo
SBET
Overall surface
acidity[a]
Brønsted
acidity[a]
Lewis
S[b]
X[b]
Modified
contact time
[sgmLꢀ1
]
acidity[a]
[Mo atomsnmꢀ2
]
[m2 gꢀ1
]
[10ꢀ3
]
[10ꢀ3
]
[10ꢀ3
]
0.15MoAl
0.15MoS1
0.15MoS10
0.15MoS40
0.15MoS70
0.15 MoSi
1.5 MoAl
1.5 MoS1
1.5 MoS10
1.5 MoS40
1.5 MoS70
1.5 MoSi
0.20
0.18
0.17
0.21
0.13
0.16
1.48
1.33
1.37
1.35
1.61
1.47
150
260
340
390
320
305
140
255
315
365
275
240
18.13
16.92
16.24
16.05
17.41
2.03
35.21
28.90
23.14
25.64
17.71
17.17
0.00
0.00
0.00
3.49
6.69
0.00
0.00
0.00
3.94
9.40
8.07
7.08
18.13
16.92
16.24
12.56
10.72
2.03
35.21
28.90
19.21
16.25
9.64
0.00
0.07
0.72
0.71
0.70
0.97[c]
0.97
0.97
0.92
0.87
0.78
0.50[c]
0.11
1.37
1.08
0.22
0.66
0.68
1.09
0.82
0.27
0.11
0.07
0.14
0.55
0.06
0.01
0.39
0.39
0.21[c]
0.08
0.10
0.41
0.42
0.39
0.34[c]
10.08
[a] The acidity values were calculated from IR spectra of adsorbed pyridine at 423 K. [b] Propene selectivity (S) and 2-butene conversion (X) obtained at a re-
spective modified contact time at 423 K. [c] These values were obtained at 573 K.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
based on Siral40, Siral70 and silica. All catalysts with a Mo load-
ing of 1.5 Monmꢀ2 displayed higher Lewis acidity than the
corresponding bare supports.
Structural properties of supported MoOx species
In Figure 1, the UV/Vis spectra of differently loaded catalysts
based on Siral1 are exemplarily shown as well as those of pure
ammonium heptamolybdate (AHM) and MoO3, with the two
Figure 2. XRD patterns of a) 1.5MoAl, b) 1.5MoS1, c) 1.5MoS10,
d) 1.5MoS40, e) 1.5MoS70, f) 1.5MoSi, and g) MoO3.
Figure 1. UV/Vis spectra of a) MoO3, b) AHM, c) 1.5 MoS1, and d) 0.15MoS1
at 423 K after pretreatment in air at 773 K.
latter representing polymerized octahedral MoOx species and
crystalline MoO3, respectively. The spectra of other identically
loaded catalysts are relatively similar to each other (Figure SI-2
and SI-3). In detail, the spectra of 0.15 Monmꢀ2 catalysts fea-
ture a strong absorption below 300 nm. This absorption band
can be assigned to the ligand-to-metal charge transfer (LMCT)
bands in highly dispersed and small linear-chained tetrahedral
MoOx species, which are characterized by bands at approxi-
mately 250 nm and approximately 290 nm, respectively.[13c,d,20]
Increasing the Mo loading to 1.5 Monmꢀ2 gave rise to a new
band at approximately 330 nm, which can be ascribed to
LMCT in polymerized octahedral MoOx.[13c,d,20] However, the
presence of small amounts of crystalline MoO3 may not be ex-
cluded in 1.5 Monmꢀ2 loaded catalysts, considering the pres-
ence of a weak absorption band at approximately 400 nm.[20c]
Therefore, we also investigated these catalysts by powder XRD
analysis.
Figure 3. Raman spectra of a) AHM, b) 1.5 MoAl, c) 1.5MoS1, d) 1.5 MoS10,
e) 1.5MoS40, f) 1.5MoS70, g) 1.5MoSi, and h) pure MoO3.
modes at approximately 950, approximately 860, and approxi-
mately 320 cmꢀ1, the spectra of 1.5MoAl and 1.5MoS10 are
characterized by a Raman shift at approximately 210 cmꢀ1
,
which is typical for the MoꢀOꢀMo deformation mode. Similar
bands are also present in the spectrum of AHM. Thus, the sur-
face of 1.5MoAl, 1.5MoS1, and 1.5 MoS10 is populated by
polymerized octahedral MoOx species. Compared to these
three sample spectra and in agreement with our UV/Vis and
XRD results, the Raman spectra of 1.5MoS40, 1.5MoS70, and
1.5MoSi proved that these catalysts contain polymerized octa-
hedral MoOx and MoO3 as concluded from the sharp Raman
bands at 992, 817, ꢁ663, 375, 334, ꢁ284, and
In agreement with the results of our UV/Vis study, no charac-
teristic MoO3 reflections are present in the XRD patterns of
0.15 Monmꢀ2 catalysts (Figure SI-4). For catalysts with
1.5 Monmꢀ2, XRD analysis proved the presence of small
amounts of nanosized (35–50 nm) MoO3 only in 1.5MoS40,
1.5MoS70, and 1.5MoSi (XRD reflections at 23.3, 25.7, 27.3,
and 33.78 in Figure 2).
To obtain information about the possible presence of XRD
amorphous MoO3 nanoparticles and surface MoOx species,
Raman spectroscopy was applied. In Figure 3, Raman spectra
of 1.5 Monmꢀ2 catalysts are shown together with the reference
spectra of AHM and MoO3. In addition to the Mo=O terminal
ꢁ237 cmꢀ1 [13d,20a,21]
.
The Raman spectra of 0.15 Monmꢀ2 cata-
lysts, excluding the one of 0.15MoSi, are mainly characterized
by a broad band at approximately 950 cmꢀ1 accompanied by
a shoulder at approximately 860 cmꢀ1 and a weak band at ap-
proximately 320 cmꢀ1 (Figure SI-5). According to previous stud-
ies,[20a,22] these bands were assigned to symmetric and antisym-
metric Mo=O terminal stretches and to Mo=O bending vibra-
tions, respectively. The absence of characteristic bands be-
tween 250 and 200 cmꢀ1, which are typical for MoꢀOꢀMo vi-
brations, and the congruent Raman shifts compared to the
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
2ꢀ[20a,22a,22b]
spectra of isolated tetrahedral coordinated MoO4
in-
1.5 MoSi with one significant exception that the reduction of
MoO3 overlaps with the reduction of polymerized MoOx spe-
cies with Tmax at approximately 880 K.
dicate the presence of highly dispersed tetrahedral MoOx spe-
cies. The Raman spectrum of 0.15MoSi features sharp bands at
875, 844, 790, 388, and 320 cmꢀ1, and a weak band at approxi-
mately 965 cmꢀ1. All these bands can neither be assigned to
MoO42ꢀ, Mo7O246ꢀ, silicomolybdic acid[23], nor MoO3.[13d,20a,21]
Nevertheless, the bands at 844 and 790 cmꢀ1 likely belong to
antisymmetric MoꢀOꢀMo stretches and point to bridged
MoOx. Hence, we assume that polymerized octahedral MoOx
already exists in 0.15MoSi.
The H2-TPR profiles of 0.15 and 1.5 Monmꢀ2 catalysts are
shown in Figure 4 and Figure SI-6, respectively. Reduction of
MoOx species in the 0.15 Monmꢀ2 catalysts containing alumina
in the support is characterized by a broad event with the tem-
In summary, the catalysts with 0.15 Monmꢀ2 are free of
MoO3, but possess either highly dispersed tetrahedral MoOx
species or linear chains of them or polymerized octahedral
MoOx, which all have been claimed to be active for metathesis
reactions of C2ꢀC4 olefins.[3a,6] The degree of polymerization of
surface MoOx species in these catalysts increases with an in-
crease in silica content in the support. It also increases with
Mo loading. Predominantly polymerized octahedral MoOx spe-
cies prevail on the surface of 1.5 Monmꢀ2 catalysts. Small
amounts of MoO3 are also present in the catalysts containing
40, 70, and 100 wt% silica in the support. To establish relation-
ships between the structure of MoOx species, their acidic, and
catalytic properties, we investigated metathesis of ethylene
and trans-2-butenes to propene.
Metathesis performance
Propene, 1-butene, cis-2-butene, and trace amounts of C5-ole-
fins, summarized as pentene, were observed over all catalysts.
For comparison, the unloaded supports were also tested under
the same conditions. They exhibited a similar activity for 2- to
1-butene isomerization, but did not produce propene (Table SI-
1). Therefore, the discussion below is dealing with the per-
formance of supported MoOx species.
In Figure 5(A), the rate of propene formation per surface
area [rBET calculated from Eq. (2)] at 423 K is shown, and the
propene selectivity values are listed in Table 2. Note that cis-2-
butene is not considered in the selectivity calculation [Eq. (4)].
The activity values of 0.15MoSi and 1.5MoSi were extrapolated
from those determined between 463 and 603 K by using Ar-
rhenius law (all measured rBET data are given in Table SI-2). Irre-
spective of the Mo loading, rBET passes a maximum with rising
silica content in the support. However, the maximum position
depends on the loading; for 0.15 and 1.5 Monmꢀ2 catalysts,
the highest activity was achieved over 0.15MoS70 and
1.5MoS40, respectively. To compare the intrinsic activity of dif-
ferently loaded catalysts, we calculated the rate of propene
formation per Mo atom [rMo calculated from Eq. (3)]. The results
are shown in Figure 5(B). It is obvious that the 1.5 Monmꢀ2
catalysts were in general more active than the lower loaded
counterparts with the exception of 1.5MoS70. The reasons are
examined in the discussion section.
Figure 4. Profiles of consumed H2 during H2–TPR of 0.15 Monmꢀ2 loaded
catalysts: a) 0.15MoSi, b) 0.15MoS70, c) 0.15MoS40, d) 0.15MoS10,
e) 0.15MoS1, and f) 0.15 MoAl.
perature of maximal hydrogen consumption (Tmax) at approxi-
mately 860 K. An additional reduction event starts above
950 K. Its Tmax depends on the alumina content; the higher the
content, the higher the temperature. This huge shift in Tmax is
owing to the weaker interaction of MoOx with silica-rich than
with alumina-rich supports.[24] The H2–TPR profile of 0.15MoSi
is more complex and consists of four less resolved maxima of
hydrogen consumption. This means that, in contrast to other
0.15 Monmꢀ2 catalysts, the surface of 0.15MoSi is populated
by MoOx species of several different types. Together with our
UV/Vis and Raman studies, we put forward that this catalyst
additionally possesses polymerized octahedral MoOx species,
whereas other 0.15 Monmꢀ2 catalysts contain a large fraction
of highly dispersed tetrahedral MoOx species and linear chains
of them. Considering previous studies,[7e,13e,f,25] the Tmax values
at approximately 750 and above 1000 K in the TPR profiles of
1.5MoAl, 1.5MoS1, and 1.5MoS10 indicate that the surface of
these catalysts is populated by polymerized octahedral MoOx
and probably by Al2(MoO4)3. Contrarily, 1.5MoS40 possesses
MoO3, which is characterized by a single reduction event at ap-
proximately 850 K.[13f] The latter is the major species in
1.5MoSi. The H2–TPR profile of 1.5MoS70 is similar to that of
The support and the Mo loading also significantly influence
the propene selectivity. However, there is no general effect
valid for all catalyst studied. As shown in Table 2, the selectivity
over the 0.15 Monmꢀ2 catalysts increases with a decrease in
the content of alumina in the support. Contrarily, it decreases
over the 1.5 Monmꢀ2 catalysts.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
tain Mo loading increases with decreasing Lewis acidity in the
support. Moreover, Lewis acidity in supported catalysts de-
pends on the degree of polymerization of surface MoOx spe-
cies and the support. This can be explained as follows. As long
as the amount of Mo in the impregnation solution is lower
than the amount of acidic sites on support surface, highly dis-
persed tetrahedral MoOx species will be formed as found in
0.15MoAl, 0.15MoS1 and 0.15MoS10. As the Lewis acidity in
these samples is lower than in the bare supports, it can be
suggested that highly dispersed tetrahedral MoOx species
block the Lewis acidic sites of the support.[15b] Furthermore, we
suppose that the probability of interaction of MoOx species
with non-Lewis acidic sites increases with a decrease in sup-
port Lewis acidity in, for example, Siral40, Siral70, and silica,
with a constant amount of MoOx in the impregnation solution.
Consequently, higher polymerized MoOx species, for example,
chained MoOx, were stabilized on the surfaces of 0.15MoS40,
0.15MoS70, and 0.15 MoSi than on the surfaces of 0.15MoAl,
0.15MoS1 and 0.15MoS10. Such species are responsible for
the creation of additional Lewis acidic sites.[13d] The high Lewis
acidity of 1.5 Monmꢀ2 catalysts obviously results from a high
concentration of polymerized MoOx species. Similar to the
0.15 Monmꢀ2 catalysts, the lower the Lewis acidity of the sup-
ports, the higher the polymerization degree of surface MoOx
species in the 1.5 Monmꢀ2 catalysts.
In summary, we demonstrated that the support and the Mo
loading influence the Brønsted and Lewis acidity of the surface
MoOx species. The support Brønsted acidity is decisive for the
formation of MoOx species with Brønsted acidic character. On
the other hand, the Lewis acidity of the support affects the dis-
persion and the polymerization degree of MoOx. Highly dis-
persed MoOx species are preferentially formed on strong Lewis
acidic supports.
Figure 5. Effect of the silica content in the Al2O3–SiO2 supports on A) rBET and
&
B) rMo of propene formation over catalysts with a Mo loading of ( )
0.15 Monmꢀ2 and ( ) 1.5 Monmꢀ2. Reaction conditions: 423 K, 125 kPa, and
*
ethylene/trans-2-butene/nitrogen=5:5:1.
Discussion
Metathesis performance
Surface properties and MoOx species formed
Differently from previous studies on metathesis reactions,[3a,6]
we can only partly support the conclusion on the effect of the
structure of surface MoOx species on the catalyst activity. For
example, all 0.15 Monmꢀ2 tested catalysts are free of catalyti-
cally inactive MoO3 and possess only highly dispersed tetrahe-
dral or polymerized octahedral MoOx species, which should be
active for the target reaction. However, the only active cata-
lysts are 0.15MoS40 and 0.15MoS70 (Figure 5). Similarly,
1.5MoAl, 1.5MoS1, and 1.5MoS10 possessing only polymer-
ized octahedral MoOx species strongly differ in their activity,
and the latter is up to approximately 20 times more active
than the former two. Surprisingly, 1.5MoS40 is even more
active than 1.5MoS10 despite the fact that it possesses both
polymerized octahedral MoOx species and catalytically inactive
nanosized MoO3 (Figure 5). The amount of MoO3 further in-
creases in 1.5MoS70 and 1.5MoSi resulting in a decrease in
The complementary characterization techniques revealed that
Mo loading and kind of the support are two key factors gov-
erning the acidic and structural properties of surface MoOx
species. We demonstrated that Brønsted acidity of supported
catalysts can be tuned by selecting appropriate supports and
Mo loadings. Irrespective of the Mo loading (0.15–
1.5 Monmꢀ2), the support Brønsted acidity favors the forma-
tion of additional Brønsted acidic sites after deposition of
MoOx species. As this effect depends on the Mo loading, it is
obvious that the new Brønsted acidic sites were generated by
MoOx species. This conclusion is supported by the fact that
1.5MoS10 and 1.5MoSi also feature Brønsted acidic sites al-
though their respective supports did not possess such sites.
Based on previous studies, we suggest that potential Brønsted
acidic sites over low-loaded catalysts are generated on boun-
daries between MoOx species and the support (MoꢀOHꢀ the metathesis activity. Thus, in agreement with previous stud-
Al(Si))[9d,26] and for higher-loaded catalysts additionally between
Mo atoms (MoꢀOHꢀMo).[26a]
ies,[3a,6] we also proved that crystalline MoO3 is catalytically in-
active in the metathesis of ethylene and trans-2-butene. How-
ever, the activity of highly dispersed tetrahedral or polymerized
octahedral MoOx species appears to be a function of the used
In good agreement with previous studies,[13c,d,15b] we also es-
tablished that the degree of polymerization of MoOx at a cer-
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
5
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
support. In contrast to previous studies highlighting the posi-
tive effect of overall catalyst acidity (Brønsted plus Lewis) on
metathesis activity,[8a,c] we could not establish any direct corre-
lation between the metathesis activity and overall acidity in
bare supports (Table 1) or catalysts (Table 2). For example, de-
spite the similar overall acidity of 0.15MoS70 and 0.15MoS10,
the rate of propene formation over the former catalyst is ap-
proximately seven times higher than over the latter. Such con-
tradiction was also found for 1.5MoS1 and 1.5MoS10. What
are the true factors governing catalytic activity of similarly
structured surface MoOx species?
As demonstrated above, deposition of MoOx species increas-
es the Brønsted acidity of bare supports. This is a clear indica-
tion that the acidity increases as a result of the presence of
MoOx species, that is, Brønsted acidic OH groups should be
connected to or located in the vicinity of these species. Impor-
tantly, the catalysts possessing such sites also exhibit high ac-
tivity for propene formation. To illustrate this correlation, we
plotted the rMo values for all tested catalysts at 423 K against
the Brønsted acidity also measured at this temperature. From
Figure 6(A) it is clear that the rate of propene formation
strongly increases with rising Brønsted acidity, while the cata-
lysts without detectable Brønsted acidic character feature very
low metathesis activity. This is valid for the catalysts possessing
highly dispersed tetrahedral or polymerized octahedral MoOx
species. However, the activity to propene formation decreases
with increasing amount of MoO3 in the catalyst irrespective of
their Brønsted acidic character, for example, 1.5MoS70. The
positive effect of the Brønsted acidity on the activity of highly
dispersed tetrahedral or polymerized octahedral MoOx species
can be explained as follows.
Mol and co-workers[9a,b,27] have previously reported on the
positive role of Brønsted acidity in Re2O7-catalyzed metathesis
of propene and methyl oleate. They suggested that reduced
rhenium species and neighboring Brønsted acid sites partici-
pate in the generation of metal carbene species. Very recently,
Amakawa et al.[10] postulated that Brønsted acidity also plays
an important role in propene metathesis over highly dispersed
MoOx species on SBA-15. Brønsted OH groups are required to
form MoVI–isopropoxide species upon propene adsorption.
Such species are further oxidized by lattice oxygen of MoOx to
gas-phase acetone yielding reduced MoIV. This species reacts
with gas-phase propene to form the metathesis active MoVI–al-
kylidene species. Assuming that Brønsted acidic MoOx sites in
our catalysts are also required to in situ generate metathesis
active metal carbene species, the correlation in Figure 6(A) is
established. The role of Brønsted acidity for the generation of
metathesis sites is indirectly supported by the fact that rMo of
2-butene to 1-butene isomerization is less dependent on the
acidity than rMo of propene formation.
Figure 6. Effect of the Brønsted acidity of 0.15 Monmꢀ2 catalysts (full sym-
bols) and 1.5 Monmꢀ2 catalysts (empty symbols) on the rate (rMo) of the for-
~
,
!
); and C) pentene ( , 3).
&
*
^
mation of A) propene ( , ), B) 1-butene
Reaction conditions: 423 K, 125 kPa, and ethylene/trans-2-butene/nitro-
gen=5:5:1.
ports. Merely for a certain Mo loading, the activity for 2- to 1-
butene isomerization rises along with the Brønsted acidity, but
there is a significant difference in activity between 0.15 and
1.5 Monmꢀ2 catalysts (Figure 6(B)). As shown in Figure 6(C),
differently loaded catalysts also strongly differ in the formation
rate of pentene, which is actually formed through metathesis
pathways, for example, 1-butene plus 2-butene. Based on the
results of catalyst characterization, we suggest that polymer-
ized octahedral MoOx species are intrinsically more selective to
propene formation than highly dispersed tetrahedral MoOx
species.
From a selectivity viewpoint, Brønsted acidity is not the only
governing factor. If propene selectivities are compared over
the most active catalysts, that is, 0.15MoS40, 0.15MoS70,
1.5MoS10, 1.5MoS40, and 1.5 MoS70 (Table 2), the
1.5 Monmꢀ2 catalysts are more selective than the 0.15 Monmꢀ2
ones. This large difference in the selectivity cannot be only ex-
plained by the different acidity of the catalysts or bare sup-
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
6
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
In summary, we have demonstrated that Brønsted acidity is
required for supported highly dispersed tetrahedral or poly-
merized octahedral MoOx species to become active and selec-
tive for propene formation from ethylene and trans-2-butene.
Catalysts possessing high amounts of crystalline MoO3 cannot
be improved by the Brønsted acidity. The acidic character of
catalytically active species can be tuned by the support and
the Mo loading. In general, polymerized octahedral MoOx spe-
cies perform better than the highly dispersed ones, particularly,
in terms of their selectivity to propene. To understand the
effect of the nature of MoOx species on propene selectivity, we
are currently investigating individual reaction pathways in the
metathesis of ethylene and trans-2-butene over catalysts pos-
sessing either highly dispersed tetrahedral or polymerized oc-
tahedral MoOx species.
Support and catalyst characterization
Nitrogen physisorption experiments were performed at 77 K on
a Belsorp-MiniII (Rubotherm). Prior to measuring, the samples were
degassed at 523 K under vacuum (2 Pa). The specific surface area
(SBET) was determined according to the BET method.[28]
The weight concentration of Mo in the calcined catalysts was de-
termined by inductively coupled plasma optical emission spec-
trometry by using a Varian 715 emission spectrometer. Each
sample (10 mg) was dissolved in a mixture of aqua regia (6 mL)
and hydrofluoric acid (2 mL). This solution was treated by micro-
wave application at 473 K and 600 kPa (Multiwave, Anton Paar/Per-
kinElmer) followed by dilution with distilled water up to 100 mL.
The resulting solution was analyzed afterwards. Using the Mo
weight concentration and the SBET values, we calculated apparent
surface densities summarized in Table 2.
For discriminating between Brønsted and Lewis acidity, we per-
formed IR measurements of adsorbed pyridine by using a Tensor
27 spectrometer (Bruker) equipped with a heatable and evacuable
in-house-developed reaction cell with CaF2 windows connected to
a gas dosing and evacuation system. The powdery samples were
pressed into self-supporting wafers with a diameter of 20 mm and
a weight of 50 mg. Prior to pyridine adsorption, the samples were
pretreated at 663 K for 10 min in synthetic air (50 mLminꢀ1) fol-
lowed by cooling to RT and evacuation. Pyridine was adsorbed at
298 K until saturation. Then the reaction cell was evacuated for re-
moving physisorbed pyridine. After heating the sample in vacuum
up to 423 K, the IR spectra of adsorbed pyridine were recorded.
Conclusions
Catalysts testing in combination with their thorough character-
ization by complementary methods provided novel fundamen-
tal insights into the effect of the dispersion, degree of poly-
merization, and acidity of MoOx species on the activity and se-
lectivity in the metathesis of ethylene and trans-2-butene to
propene. The propene production over highly dispersed tetra-
hedral and polymerized octahedral MoOx species is significant-
ly enhanced by the induced Brønsted acidity. In addition to
the acidity, the structure of MoOx species influences the selec-
tivity to propene; octahedral species perform superior to the
tetrahedral ones.
UV/Vis measurements of supported catalysts were performed by
using an AVASPEC fiber optical spectrometer (Avantes) equipped
with a DH-2000 deuterium–halogen light source and a CCD array
detector. BaSO4 was used as a white reference material. A high-
temperature reflection UV/Vis probe consisting of six radiating op-
tical fibers and one reading fiber was threaded through the fur-
nace to face the wall of the quartz reactor at the position where
catalysts were held.[29] UV/Vis spectra were collected in the range
from 220 to 700 nm at 423 K. The recorded reflectance (R) was
converted into the Kubelka–Munk function (F) according to Equa-
tion (1). Prior to the measurements, the catalysts were heated to
773 K (temperature ramp of 5 Kminꢀ1) in nitrogen (5.0, Air Liquide,
With respect to catalyst development, catalytic materials
should possess as high as possible concentration of surface
MoOx species with Brønsted acidic character and be free of cat-
alytically inactive MoO3. The acidity can be tuned by the Mo
loading or/and by the support Brønsted acidity; the higher the
Mo loading or/and the acidity, the higher the Brønsted acidity
of the MoOx species. For a certain Mo loading, the Lewis acidi-
ty of the support determines the dispersion and the degree of
polymerization of MoOx species. As a consequence, the result-
ing structure of supported MoOx species and their Brønsted
acidic character can be tuned by a proper choice of support.
20 mLminꢀ1
) and then oxidized in a flow of synthetic air
(25 mLminꢀ1/reactor as a mixture of 20 mLminꢀ1 nitrogen and
5 mLminꢀ1 oxygen (4.8, Air Liquide) for 3 h. The UV/Vis spectrum
of AHM was measured at 298 K in nitrogen.
2
ð1 ꢀ RÞ
ð1Þ
FðRÞ ¼
2 ꢂ R
Experimental Section
XRD measurements were performed on a Theta/Theta diffractome-
ter X’Pert Pro (Panalytical) with CuKa radiation (a=1.5418 ꢄ, 40 kV,
40 mA) and an X’Celerator RTMS Detector. The alignment was
made according to a silicon standard. The data of calcined catalysts
were collected at RT in the 2 q range from 5 to 708. The phase
composition of the samples was determined by using the program
suite WinXPOW (Stoe & Cie) with inclusion of the powder diffrac-
tion file PDF2 of the international center of diffraction data. The
Scherrer equation was used to calculate the size of MoO3
crystallites.[30]
Catalyst preparation
The catalysts were prepared by incipient wetness impregnation of
commercial oxide supports with different Al/Si ratios. In Table 1,
their abbreviations, selected physicochemical properties, and the
suppliers are listed. The supports were initially dried at 383 K for
24 h followed by calcination at 773 K for 8 h in static air. Aqueous
solution of AHM (NH4)6Mo7O24 ꢃ4H2O (Riedel–de Haen) was drop-
wise added to the dried supports to yield catalyst precursors with
atomic loadings of 0.15 or 1.5 Monmꢀ2. Afterwards the catalyst
precursors were treated at 383 K in static air under agitation for
10 h followed by calcination at 773 K in static air for 8 h. The final
catalysts are denoted XMoY, whereby “X” stands for the nominal
Mo loading (Monmꢀ2) and “Y” for the support.
Raman spectra of MoO3, AHM, and supported catalysts were ac-
quired in the range from 1100 to 200 cmꢀ1 by using a Labram010
Raman microscope (Horiba Jobin Yvon) with a 1808 back scattering
system using a 532 nm laser (100 mW) and with the 50ꢃ ocular.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
7
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Acquisition time was between 1 and 1000 s depending on catalyst.
20 scans were recorded and averaged for each sample. The experi-
ments were performed at RT in air without any pretreatment.
activities of differently loaded catalysts, we also calculated the Mo-
related reaction rate (rMo) using Equation (3). The highest relative
standard deviation of rBET was 7.28% for 1.5MoS1 at 423 K. The
product selectivity (S) was calculated according to Equation (4).
Temperature-programmed reduction of supported catalysts with
hydrogen (H2-TPR) was performed in a setup containing eight indi-
vidually heated plug-flow fixed-bed quartz reactors. The catalyst
amount was calculated according to D. A. Monti and A. Baiker.[31]
Prior to H2–TPR, the catalysts were oxidized in a 5 vol% O2/Ar flow
(Air Liquide, 10 mLminꢀ1/reactor) at 773 K for 3 h. Afterwards they
were cooled down to 323 K in argon (5.0, Linde Gas) and flushed
for 3 h. Hereafter, the argon flow was replaced by a 5 vol% H2/Ar
(Air Liquide) flow of 10 mLminꢀ1/reactor and the temperature of
a selected reactor was increased to 1173 K with a heating rate of
10 Kminꢀ1. Other catalysts were sequentially tested in a similar
way. Hydrogen consumption and water formation were monitored
by a quadrupole mass spectrometer (Pfeiffer Vacuum OmniStar
200). The following atomic mass units were analyzed: 40 (Ar), 18
(H2O), and 2 (H2). The concentration of hydrogen was determined
from the respective atomic mass unit by using the sensitivity
factor determined by analyzing calibration gas mixtures.
ꢀ
ꢁ
_
molpropene
m2catalyst ꢂ s
xpropene ꢂ VF
ð2Þ
ð3Þ
ð4Þ
rBET
¼
SBET ꢂ mcatalyst ꢂ VM
ꢀ
ꢁ
_
moleculepropene
Mo ꢀ atom ꢂ s
xpropene ꢂ VF ꢂ NA
rMo
¼
SBET ꢂ mcatalyst ꢂ VM ꢂ wMo
bC;i ꢂ xi
Si ¼
ð3 ꢂ xpropene þ 4 ꢂ x1ꢀbutene þ 5 ꢂ xpentene
Þ
˙
For which xi, bC,i, VF, mcatalyst, VM, NA and wMo stand for mole fraction
of component i, number of carbon atoms in component i, feed
flow rate, catalyst mass, molar volume, Avogadro number, and
weight concentration of Mo, respectively.
Acknowledgements
Catalytic tests
Financial support by Deutsche Forschungsgemeinschaft (DFG,
project KO 2261/3-1, Li 2047/1-1) is gratefully acknowledged.
Metathesis of ethylene (3.5, Air Liquide) and trans-2-butene (2.0,
Linde Gas) to propene was investigated in a parallel operating mul-
tichannel fixed-bed reactor system. Up to 15 samples could be an-
alyzed under identical reaction conditions. This setup was fully au-
tomated and enabled controlling gas flows, three temperature
zones (pre-, reactor, and post-heating), reactor selection, and prod-
uct sampling. To exclude the effect of internal mass transport on
catalytic results, we used catalysts in the form of particles of 250–
355 mm. The catalysts (0.02–0.60 g) were loaded into quartz tube
reactors (4 mm inner diameter) and sandwiched between ꢁ20 mg
quartz wool (Roth) and ꢁ1.2 g silicon carbide (F 30, ESK-SiC), with
the former located downstream below the catalyst bed. Initially,
Keywords: acidity · metathesis · molybdenum · structure
elucidation · supported catalysts
[1] R. L. Banks, G. C. Bailey, Ind. Eng. Chem. Prod. Res. Dev. 1964, 3, 170–
173.
[2] Hydrocarbon Processing, Phillips Petroleum Company, 46, 1967.
[3] a) K. J. Ivin, J. C. Mol in Olefin Metathesis and Metathesis Polymerization
(2), Academic Press, London, 1997, pp. 397–410; b) J. C. Mol, J. Mol.
Catal. A 2004, 213, 39–45.
[4] G. Parkinson, Chem. Eng. 2001, 108, 27–35.
[5] J. C. Mol, J. A. Moulijn in Adv. Catal. Vol. 24 (Eds.: H. P. D. D. Eley, B. W.
Paul), Academic Press, 1975, pp. 131–171.
the catalysts were heated to 773 K (temperature ramp of 5 Kminꢀ1
in followed by feeding air
nitrogen flow (20 mLminꢀ1
)
a
)
(25 mLminꢀ1/reactor as a mixture of 20 mLminꢀ1 nitrogen and
5 mLminꢀ1 oxygen) for 3 h and then cooled down to reaction tem-
perature in the nitrogen flow and flushed for additional 90 min.
The experiments were performed at 125 kPa in a temperature
range from 343 to 603 K by using a feed containing ethylene,
trans-2-butene, and nitrogen in a ratio of 5:5:1 at a total flow rate
between 11 and 22 mLminꢀ1/reactor. All gases were purified over
molecular sieve 3 ꢄ filters (Roth). Nitrogen was further purified
with an additional AlO-750-2 filter (Pure Gas Products).
[6] a) M. Anpo, M. Kondo, Y. Kubokawa, C. Louis, M. Che, J. Chem. Soc. Fara-
´
day Trans. 1 1988, 84, 2771–2782; b) P. Topka, H. Balcar, J. Rathousky, N.
ˇ
ˇ
zilkovꢅ, F. Verpoort, J. Cejka, Microporous Mesoporous Mater. 2006, 96,
44–54; c) J. Handzlik, Surf. Sci. 2007, 601, 2054–2065; d) D. P. Debecker,
B. Schimmoeller, M. Stoyanova, C. Poleunis, P. Bertrand, U. Rodemerck,
E. M. Gaigneaux, J. Catal. 2011, 277, 154–163; e) T. Ono, M. Anpo, Y. Ku-
bokawa, J. Phys. Chem. 1986, 90, 4780–4784; f) B. Zhang, Y. Li, Q. Lin, D.
Jin, J. Mol. Catal. 1988, 46, 229–241.
[7] a) D. P. Debecker, M. Stoyanova, U. Rodemerck, E. M. Gaigneaux, J. Mol.
Catal. A 2011, 340, 65–76; b) D. P. Debecker, M. Stoyanova, U. Rode-
merck, A. Lꢆonard, B.-L. Su, E. M. Gaigneaux, Catal. Today 2011, 169,
60–68; c) J. Handzlik, J. Ogonowski, J. Stoch, M. Mikołajczyk, Appl.
Catal. A 2004, 273, 99–104; d) B. Zhang, N. Liu, Q. Lin, D. Jin, J. Mol.
Catal. 1991, 65, 15–28; e) R. Thomas, J. A. Moulijn, V. H. J. De Beer, J.
Medema, J. Mol. Catal. 1980, 8, 161–174.
[8] a) D. P. Debecker, D. Hauwaert, M. Stoyanova, A. Barkschat, U. Rode-
merck, E. M. Gaigneaux, Appl. Catal. A 2011, 391, 78–85; b) S. Liu, X. Li,
W. Xin, S. Xie, P. Zeng, L. Zhang, L. Xu, J. Nat. Gas Chem. 2010, 19, 482–
486; c) J. Handzlik, J. Ogonowski, J. Stoch, M. Mikołajczyk, P. Michorczyk,
Appl. Catal. A 2006, 312, 213–219; d) J. Handzlik, J. Ogonowski, J.
Stoch, M. Mikolajczyk, Catal. Lett. 2005, 101, 65–69.
The feed components, that is, ethylene and trans-2-butene, and re-
action products, that is, propene, 1-butene and pentene (as the
sum of all pentene isomers), were quantified by an online gas
chromatograph (6890N, Agilent). Analyzes were taken every
8.2 min. The separation of hydrocarbons was performed on a CP-
Al2O3/KCl column (25 m nominal length, 320 mm nominal diameter,
5 mm nominal film thickness). A flame ionization detector was used
for quantifying feed components and reaction products. Nitrogen
was analyzed on a second channel using a HP-Molsieve column
(30 m nominal length, 530 mm nominal diameter, 25 mm nominal
film thickness) and a thermal conductivity detector. The initial
metathesis performance (activity and selectivity) was determined
after approximately 0.5 h on stream. To ensure good accuracy, cat-
alytic tests were repeated several times and the results were aver-
[9] a) X. Xiaoding, C. Boelhouwer, D. Vonk, J. I. Benecke, J. C. Mol, J. Mol.
Catal. 1986, 36, 47–66; b) A. Andreini, X. Xiaoding, J. C. Mol, Appl. Catal.
1986, 27, 31–40; c) E. A. Lombardo, M. Lo Jacono, W. K. Hall, J. Catal.
1978, 51, 243–255; d) D. T. Laverty, J. J. Rooney, A. Stewart, J. Catal.
1976, 45, 110–113.
aged. The activity was defined as surface-related reaction rate (rBET
and was calculated according to Equation (2). For comparing the
)
[10] K. Amakawa, S. Wrabetz, J. Krçhnert, G. Tzolova-Mꢀller, R. Schlçgl, A.
Trunschke, J. Am. Chem. Soc. 2012, 134, 11462–11473.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
8
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[11] Y. Chauvin, Angew. Chem. Int. Ed. 2006, 45, 3740–3747; Angew. Chem.
2006, 118, 3824–3831.
[21] J. Medema, C. van Stam, V. H. J. de Beer, A. J. A. Konings, D. C. Konings-
berger, J. Catal. 1978, 53, 386–400.
[12] J. Houzvicka, V. Ponec, Catal. Rev. Sci. Eng. 1997, 39, 319–344.
[13] a) E. Hillerovꢅ, H. Morishige, K. Inamura, M. Zdrazˇil, Appl. Catal. A 1997,
156, 1–17; b) H. Hu, I. E. Wachs, S. R. Bare, J. Phys. Chem. 1995, 99,
10897–10910; c) C. C. Williams, J. G. Ekerdt, J. M. Jehng, F. D. Hardcastle,
I. E. Wachs, J. Phys. Chem. 1991, 95, 8791–8797; d) C. C. Williams, J. G.
Ekerdt, J. M. Jehng, F. D. Hardcastle, A. M. Turek, I. E. Wachs, J. Phys.
Chem. 1991, 95, 8781–8791; e) R. Thomas, E. M. van Oers, V. H. J.
de Beer, J. A. Moulijn, J. Catal. 1983, 84, 275–287; f) R. Thomas, E. M. va-
n Oers, V. H. J. de Beer, J. Medema, J. A. Moulijn, J. Catal. 1982, 76, 241–
253.
[14] a) M. R. Basila, T. R. Kantner, K. H. Rhee, J. Phys. Chem. 1964, 68, 3197–
3207; b) F. E. Kiviat, L. Petrakis, J. Phys. Chem. 1973, 77, 1232–1239.
[15] a) P. O. Scokart, F. D. Declerck, R. E. Sempels, P. G. Rouxhet, J. Chem. Soc.
Faraday Trans. 1 1977, 73, 359–371; b) S. Rajagopal, J. A. Marzari, R. Mir-
anda, J. Catal. 1995, 151, 192–203.
[22] a) W. P. Griffith, P. J. B. Lesniak, J. Chem. Soc. A 1969, 0, 1066–1071; b) H.
Knoezinger, H. Jeziorowski, J. Phys. Chem. 1978, 82, 2002–2005; c) D. S.
Kim, K. Segawa, T. Soeya, I. E. Wachs, J. Catal. 1992, 136, 539–553; d) G.
Mestl, T. K. K. Srinivasan, Catal. Rev. Sci. Eng. 1998, 40, 451–570.
[23] a) Y. Barbaux, A. R. Elamrani, E. Payen, L. Gengembre, J. P. Bonnelle, B.
Grzybowska, Appl. Catal. 1988, 44, 117–132; b) S. Kasztelan, E. Payen,
J. B. Moffat, J. Catal. 1988, 112, 320–324; c) M. A. Banares, H. C. Hu, I. E.
Wachs, J. Catal. 1995, 155, 249–255.
[24] a) J. Brito, J. Laine, Polyhedron 1986, 5, 179–182; b) S. Rajagopal, H. J.
Marini, J. A. Marzari, R. Miranda, J. Catal. 1994, 147, 417–428; c) J. Re-
galbuto, J.-W. Ha, Catal. Lett. 1994, 29, 189–207.
[25] R. Thomas, J. A. Moulijn, J. Mol. Catal. 1982, 15, 157–172.
[26] a) T. Kitano, S. Okazaki, T. Shishido, K. Teramura, T. Tanaka, J. Mol. Catal.
A 2013, 371, 21–28; b) W. K. Hall in Chemistry and Uses of Molybdenum:
Proceedings of the Fourth International Conference (Eds.: H. F. Barry,
P. C. H. Mitchell), Climax Molybdenum Company, Golden, Colorado, USA
1982 (August 9–13), p. 224.
[16] J. A. Schwarz, B. G. Russell, H. F. Harnsberger, J. Catal. 1978, 54, 303–
317.
[17] a) R. A. van Santen, G. J. Kramer, Chem. Rev. 1995, 95, 637–660; b) C.
Chizallet, P. Raybaud, Angew. Chem. Int. Ed. 2009, 48, 2891–2893;
Angew. Chem. 2009, 121, 2935–2937.
[18] R. E. Sempels, P. G. Rouxhet, J. Colloid Interface Sci. 1976, 55, 263–273.
[19] J. A. R. Van Veen, P. A. J. M. Hendriks, E. J. G. M. Romers, R. R. Andrea, J.
Phys. Chem. 1990, 94, 5275–5282.
[20] a) H. Jeziorowski, H. Knoezinger, J. Phys. Chem. 1979, 83, 1166–1173;
b) O. Collart, P. Van Der Voort, E. F. Vansant, E. Gustin, A. Bouwen, D.
Schoemaker, R. Ramachandra Rao, B. M. Weckhuysen, R. A. Schoon-
heydt, Phys. Chem. Chem. Phys. 1999, 1, 4099–4104; c) E. L. Lee, I. E.
Wachs, J. Phys. Chem. C 2007, 111, 14410–14425.
[27] J. C. Mol, Catal. Today 1999, 51, 289–299.
[28] K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rou-
querol, T. Siemieniewska, Pure Appl. Chem. 1985, 57, 603–619.
[29] O. Ovsitser, M. Cherian, A. Brꢀckner, E. V. Kondratenko, J. Catal. 2009,
265, 8–18.
[30] P. Scherrer, Gçttinger Nachrichten Math. Phys. 1918, 2, 98–100.
[31] D. A. M. Monti, A. Baiker, J. Catal. 1983, 83, 323–335.
Received: January 15, 2014
Published online on && &&, 0000
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
9
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
T. Hahn, U. Bentrup, M. Armbrꢀster,
E. V. Kondratenko,* D. Linke*
Acidity meets structure: The activity
and selectivity of MoOx species on dif-
ferent supports in the metathesis of
ethylene and trans-2-butene to propene
are systematically elucidated and corre-
lated to their structural and acidic prop-
erties. Tetrahedral and octahedral MoOx
species exhibit enhanced performance
with increasing Brønsted acidity.
&& – &&
The Enhancing Effect of Brønsted
Acidity of Supported MoOx Species on
their Activity and Selectivity in
Ethylene/trans-2-Butene Metathesis
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&10
&
ÞÞ
These are not the final page numbers!