Aromatization of n-butane and 1-butene over supported Mo2C catalyst

Article abstract of DOI:10.1016/j.jcat.2004.01.006

The reaction pathways of n-butane were investigated on Mo₂C deposited on ZSM-5 and SiO₂. Particular attention was paid to the effects of the composition of ZSM-5, to the influence of the Mo₂C loading, and to the nature of supports. ZSM-5 itself catalyzed the reaction of n-butane well above 800 K. Its efficiency sensitively depended on the composition of zeolite. Whereas the conversion of butane was ～90% on ZSM-5 with SiO₂/Al₂O₃=30 at 823 K, this value was only ～24% on the sample with SiO₂/Al₂O ₃=280. The dominant reaction was the cracking of butane yielding several C₁-C₃ compounds. Deposition of Mo₂C markedly changed the catalytic performance of ZSM-5, and the dehydrogenation and the aromatization processes came into prominence. This is particularly true in the case of less effective ZSM-5 (SiO₂/Al₂O ₃=280). From the extrapolation of selectivities to zero conversion we obtained that methane, ethane, ethylene, propylene, butene, and hydrogen are the primary products on pure and Mo₂C-containing ZSM-5. Aromatics are formed in a secondary process, in the oligomerization and aromatization of butenes. The favorable effect of Mo₂C is well exhibited in the case of SiO₂, which was practically inactive. For 2% Mo ₂C/SiO₂ at 823 K the selectivity of aromatics was 16-17% at a butane conversion of 26%. On this sample the main reaction was the dehydrogenation process. As the starting compound in the formation of aromatics is very likely butene, detailed measurements were performed on its reaction on the previously studied catalysts. 1-Butene exhibited a very high reactivity on pure ZSM-5 samples even at 723 K. The presence of 2% Mo₂C on the zeolites resulted only in a slight change in the conversion and product selectivities. The possible mechanism of the reactions and the role of Mo ₂C are discussed, taking into account our surface science studies on the reaction of butyl species on a Mo₂C/Mo(100) surface.

Full text of DOI:10.1016/j.jcat.2004.01.006

Journal of Catalysis 223 (2004) 221–231
www.elsevier.com/locate/jcat
Aromatization of n-butane and 1-butene over supported Mo C catalyst
2
∗
Frigyes Solymosi and Aleksandar Széchenyi
Institute of Solid State and Radiochemistry, University of Szeged and Reaction Kinetics Research Group of the Hungarian Academy of Sciences,
1
PO Box 168, H-6701 Szeged, Hungary
Received 4 December 2003; revised 10 January 2004; accepted 12 January 2004
Abstract
The reaction pathways of n-butane were investigated on Mo C deposited on ZSM-5 and SiO . Particular attention was paid to the effects
2
2
of the composition of ZSM-5, to the influence of the Mo C loading, and to the nature of supports. ZSM-5 itself catalyzed the reaction
2
of n-butane well above 800 K. Its efficiency sensitively depended on the composition of zeolite. Whereas the conversion of butane was
∼
90% on ZSM-5 with SiO /Al O = 30 at 823 K, this value was only ∼ 24% on the sample with SiO /Al O = 280. The dominant
2
2
3
2
2 3
reaction was the cracking of butane yielding several C –C compounds. Deposition of Mo C markedly changed the catalytic performance
1
3
2
of ZSM-5, and the dehydrogenation and the aromatization processes came into prominence. This is particularly true in the case of less
effective ZSM-5 (SiO /Al O = 280). From the extrapolation of selectivities to zero conversion we obtained that methane, ethane, ethylene,
2
2 3
propylene, butene, and hydrogen are the primary products on pure and Mo C-containing ZSM-5. Aromatics are formed in a secondary
2
process, in the oligomerization and aromatization of butenes. The favorable effect of Mo C is well exhibited in the case of SiO , which was
2
2
practically inactive. For 2% Mo C/SiO at 823 K the selectivity of aromatics was 16–17% at a butane conversion of 26%. On this sample
2
2
the main reaction was the dehydrogenation process. As the starting compound in the formation of aromatics is very likely butene, detailed
measurements were performed on its reaction on the previously studied catalysts. 1-Butene exhibited a very high reactivity on pure ZSM-5
samples even at 723 K. The presence of 2% Mo C on the zeolites resulted only in a slight change in the conversion and product selectivities.
2
The possible mechanism of the reactions and the role of Mo C are discussed, taking into account our surface science studies on the reaction
2
of butyl species on a Mo C/Mo(100) surface.
2

2004 Elsevier Inc. All rights reserved.
Keywords: Aromatization of n-butane; Reaction of 1-butene; Mo C/ZSM-5; Mo C/SiO
2
2
2
1
. Introduction
The upgrading of lower alkanes is an important subject
voted to the formation, structure, and reactivity of Mo2C on
ZSM-5 [10–15].
In our laboratory we continued our work in two direc-
tions: (i) elaborating the effect of Mo2C on the aromatization
of other hydrocarbons, ethane [16,17], ethylene [18], and
propane [19], and (ii) studying the chemistry of hydrocar-
bon fragments, CxHy, the primary products of the activation
of the above compounds, on Mo2C/Mo(100) in UHV by sev-
eral spectroscopic methods [20–24].
As a continuation of this research program the reac-
tion of n-butane and 1-butene was investigated on the same
Mo2C/ZSM-5 catalyst used in our previous works. Atten-
tion is paid to the low-temperature interaction of butane with
the catalyst, to the effects of the composition of ZSM-5 on
the catalytic performance of Mo2C, and to the influence of
preparation and pretreatments of the catalysts. Detailed mea-
surements are also performed on Mo2C deposited on inac-
tive oxidic supports. In order to obtain a deeper insight into
the mechanism of the formation of aromatics, the reaction of
of heterogeneous catalysis [1]. One of the great challenges
in this area is to transform methane into higher hydrocar-
bons with high conversionand selectivities, which represents
one route of converting the cheap raw materials into more
valuable compounds. The discovery of Wang et al. [2] and
Xu et al. [3] that methane can be converted into benzene
on MoO3/ZSM-5 opened a new route for the utilization of
methane. It turned out, however, that Mo2C not MoO3 is
the key component for the activation of methane, which is
formed from MoO3 during the induction period of the re-
action [4–9]. In subsequent works great attention was de-
*
Corresponding author.
E-mail address: fsolym@chem.u-szeged.hu (F. Solymosi).
This laboratory is a part of the Center for Catalysis, Surface and Ma-
1
terial Science at the University of Szeged.
0021-9517/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2004.01.006

222
F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
1
-butene, assumedly the starting compound in the formation
Table 1
Characteristic data for ZSM-5 samples
of benzene, is also examined on Mo2C-containing catalysts.
In a short paper we already gave a preliminary account of
the reaction of n-butane on Mo2C/ZSM-5 (SiO2/Al2O3 =
2
SiO /Al O ratio
BET area (m /g)
Total acid sites/unit cell
2
2 3
30
400
425
425
400
9.79
6.94
3.22
1.32
8
0) [25]. The interaction of n-butane with MoO3 on ZSM-5
50
80
50
and the aromatization of n-butane on Mo2C/ZSM-5 were
also studied by Derouane’s school [26,27]. Both laboratories
showed that while n-butane is mainly cracked on ZSM-5,
the production of aromatics is markedly promoted when
Mo2C is deposited on ZSM-5. In addition, Derouane and co-
workers [26,27] also developed a method for the preparation
of Mo2C by a C4H10/H2 mixture. The large amount of Mo2C
2
and dwell time of 300 ms were used. Typically 10 scans
were accumulated for each spectrum. Fitting and deconvo-
lution of the spectra were made using the VISION software
(
Kratos). The pretreatments of the samples were performed
(
∼ 15%) used allowed them to perform XRD measurements
in the preparation chambers attached to the UHV system.
and to determine the structure of Mo2C formed under differ-
ent experimental conditions. They stated that the α-Mo2C
is a better catalyst than β-Mo2C, and the final temperature
of the carburization of MoO3/ZSM-5 effects both the con-
version and the selectivity to aromatics. This observation is
important in developing a more efficient Mo2C/ZSM-5.
2
.2. Materials
The gases used were of commercial purity (Linde). Four
ZSM-5 samples with different compositions were used. The
starting materials were NH4-ZSM-5 (Zeolite International),
which was calcined to produce H-ZSM-5 in air at 863 K for
5
h. Silica used were the products of Aerosil and Degussa.
2
. Experimental
The surface areas of the samples obtained are shown in Ta-
ble 1. The acidity of the zeolites are also listed in Table 1.
MoO3-containing catalysts were prepared by impregnat-
ing H-ZSM-5 or SiO2 with a basic solution of ammonium
heptamolybdate to yield different wt% of MoO3. The sus-
pension was dried at 373 K and calcined at 863 K for 5 h.
Supported Mo2C catalysts were prepared by the carburiza-
tion of calcined MoO3/ZSM-5 or MoO3/SiO2 by ethane
following the description of Green and co-workers [29]:
the MoO3-containing sample was heated under 10% v/v
C2H6/H2, from room temperature to 900 K at a heating rate
of 0.8 K min . After preparation the catalysts were cooled
down to room temperature under argon. The carbides were
passivated in flowing 1% O2/Ar at 300 K. The same cata-
lysts were made by carburization of supported MoO3 by a
CH4/H2 mixture [30]. Before most of the catalytic experi-
ments, the samples were treated with H2 at 873 K for 60 min
to remove the excess carbon. Unsupported Mo2C was also
2
.1. Methods
Catalytic reaction was carried out at 1 atm of pressure in
a fixed-bed, continuous flow reactor consisting of a quartz
tube (7 mm i.d.) connected to a capillary tube [3,5,6]. The
flow rate was in most cases 12 ml/min. The carrier gas
was Ar. The hydrocarbon content was 12.5%. Generally
0
.3 g of loosely compressed catalyst sample was used. Re-
action products were analyzed by gas chromatography using
a Hewlett-Packard 4890 gas chromatograph with an HP-Plot
Al2O3 and with a Poropak QS + S columns. The conversion
of n-C4H10 was calculated taking into account its amount
consumed, and also on the hydrogen and carbon basis. The
first two gave practically similar values. The selectivity for
reaction products, Si , was defined as
−1
xini
Si = ꢀ
,
xini
i
prepared by carburization of MoO with ethane.
3
where xi is the fraction of product i, and ni is the number of
carbon atoms in each molecule of gaseous products.
The amount of coke deposited on the catalyst during the
reaction was determined by temperature-programmed reac-
tion (TPR). The catalyst was cooled down in flowing argon
and then heated in a H2 stream with a rate of 5 K/min
and the hydrocarbons formed were measured. The acidity
of ZSM-5 samples was determined by the method described
before [28], which consists of the adsorption and desorption
of NH3.
Some characteristic XPS spectra for supported Mo2C are
presented in Fig. 1. The binding energies for Mo(3d5/2) and
Mo(3d3/2) were 227.8–228.2 and 230.7–231.1 eV, and for
C(1s) 283.8 eV. These values are consistent with those at-
tributed to Mo C [6,9,31].
2
3. Results
3.1. Interaction of n-butane with supported Mo2C
The XPS measurements were performed in a Kratos
−8
XSAM 800 instrument at a base pressure of 10 Torr using
Mg-Kα primary radiation (14 kV, 10 mA). To compensate
for possible charging effects, binding energies (BE) were
normalized to the Fermi level for the Mo2C. The pass en-
ergy was set at 40 eV, and an energy step width of 50 meV
Butane adsorbed weakly and nondissociatively on
Mo2C/ZSM-5 at 140–250 K. By means of FTIR spec-
troscopy we detected the characteristic absorption bands of
n-butane. Above 300 K new bands appeared at 2960, 2873,
−1
1381, and 1330 cm , which remained on the spectra even

F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
223
Fig. 1. XPS spectra of Mo C/ZSM-5 and Mo C/SiO catalysts:
2
2
2
(
A) MoO /ZSM-5; (B) Mo C (Aldrich); (C) Mo C/SiO ; (D) Mo C/
3 2 2 2 2
ZSM-5.
Fig. 2. Effects of SiO /Al O ratio on the reaction of n-butane at 823 K.
2
2 3
Data were taken at 40 min on stream.
after degassing the sample at 300–400 K. These bands can
be attributed to the vibrations of butylidyne formed in the
strong interaction of n-butane with Mo2C [32]. No such
spectral features were observed on pure ZSM-5 under simi-
lar conditions.
Table 2
Distribution of butenes formed in the reaction of n-butane on different cat-
alysts at 823 K
Catalyst
Si/Al
Selectivity (%)
ratio
trans-2- 1-Butene Isobutene
cis-2-
Butene
Butene
3
3
.2. Reaction of n-butane
H/ZSM-5
(30)
(30)
(50)
(50)
(80)
(80)
(280)
0.36
0.85
0.94
1.26
1.52
2.38
3.59
8.93
0.18
0.74
0.74
1.03
1.18
1.95
2.54
7.40
0.51
1.34
1.59
2.09
2.55
3.79
5.75
14.42
0.25
0.66
0.72
0.97
1.16
1.83
2.67
6.92
2
% Mo C/ZSM-5
2
.2.1. Effects of the composition of ZSM-5
First we examined the catalytic behavior of four ZSM-5
H/ZSM-5
% Mo C/ZSM-5
2
2
samples with different compositions. Characteristic data for
the effect of the composition of ZSM-5 are plotted in Fig. 2.
As regards the conversion of n-butane, the ZSM-5 with
SiO2/Al2O3 = 30 (further Si/Al = 30) was found to be the
most active catalyst. At 823 K the conversion was above
H/ZSM-5
2% Mo C/ZSM-5
2
H/ZSM-5
2
% Mo C/ZSM-5 (280)
2
Data were taken at 75 min of time on stream.
9
0%. Another feature is the high stability of the catalyst,
which is characteristic for all ZSM-5 samples used. The
highest selectivity was measured for propane (26–28%),
xylene (18–20%), ethane (14.6%), methane (14.0%), ben-
zene (11–12%), ethylene (6.7%), and propene (4.9%). It is
interesting that butenes were formed with low selectivity,
of the olefins increased. This is particularly true for the for-
mation of butenes, the selectivity of which was more than an
order of magnitude higher on ZSM-5 (Si/Al = 280) com-
pared to the value measured for zeolite (Si/Al = 30). We
note here that toluene was formed on all zeolites, but its se-
lectivity remained at a low level (0.12–0.47%).
Detailed analysis of butenes showed that it consists of
1-butene, trans-2-butene, isobutene and cis-2-butene. Their
distribution is shown in Table 2. It appears that isobutene is
produced with higher selectivities on all ZSM-5 samples.
∼
1.0%.
The efficiency of the ZSM-5 gradually decreased with the
increase of the SiO2/Al2O3 ratio, and at the value of 280, the
conversion of butane was only 23–24%. At the same time the
selectivity of benzene and xylene diminished, whereas that

224
F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
3
.2.2. Effects of Mo2C
From the study of the effect of the preparation of Mo2C
the selectivity of the formation of butenes occurred and, at
the same time, a significant reduction in the selectivity of
propane.
We experienced similar phenomena in the case of less ef-
fective ZSM-5 samples. The addition of 2% Mo2C to ZSM-5
we observed that the preparation of Mo2C by carburization
of unsupported or ZSM-5-supported MoO3 by ethane al-
ways produced a better catalyst compared to that made by
methane. This feature appeared in both the conversion and
the selectivity to formation of aromatics. The possible reason
is the higher surface area or dispersion of Mo2C. Therefore
in the subsequent measurements the Mo2C samples were
made using ethane. As regards the catalytic performance
of unsupported Mo2C (0.5 g) we found a well measurable
reaction of n-butane at 823 K. The initial conversion was
(
Si/Al = 80) increased the selectivity to aromatics from 12
to 27% and markedly diminished that to propane. On in-
creasing the amount of Mo2C we observed a decrease in
the conversion and in the selectivities of aromatics, methane,
ethane, and propane, and a marked increase in that of butenes
(
Fig. 4A). The results obtained for ZSM-5 (Si/Al = 280) are
plotted in Fig. 4B. While this zeolite exhibited a very little
aromatizing tendency, the presence of Mo2C promoted the
formation of both benzene and xylene and slightly increased
the conversion too. The selectivity of aromatics rose from 6
to 13–15%. A significant enhancement was observed in the
selectivity of butenes, whereas the formation of lower hy-
drocarbons, particularly that of propane, has been reduced.
Mo2C exerted only a slight influence on the formation of
toluene measured on the ZSM-5 samples.
1
3
0–15%, which decayed to a constant value, 4–5% after 20–
0 min of reaction. The main process was the dehydrogena-
tion yielding various butenes and hydrogen. The selectivity
to butenes was 75–80%, but the products of cracking re-
actions were also observed. Benzene and xylene were also
formed, but their total selectivity was below 0.5%.
Next, the effects of different amounts of Mo2C were ex-
amined on all ZSM-5 samples. As ZSM-5 (Si/Al = 30) it-
self exhibited high activity (see results in Fig. 2), a great
improvement of its catalytic performance with Mo2C addi-
tion was not expected. The deposition of 2% Mo2C onto
this ZSM-5 led to a decay in the conversion of butane
of a few percent, and somewhat improved the aromatiza-
tion capability of ZSM-5, as indicated by the higher se-
lectivity of aromatics. As shown in Fig. 3, the selectivi-
ties of various products changed very little with time on
stream. A further increase in the amount of Mo2C pro-
duced a less effective catalyst, and was also unfavorable
for the formation of aromatics. A dramatic enhancement in
When the temperature was raised to 873 K, the selectivi-
ties of all aromatic compounds, benzene, xylene, and toluene
were considerably increased, mainly at the expense of the
formation of propane, and resulted in higher yields of aro-
matics. This is well illustrated in the Table 3, where the
yields of various products are collected.
As regards the distribution of butenes on Mo2C/ZSM-5
samples we found the same features as for pure ZSM-5.
Isobutene formed with higher selectivities on all Mo2C-
containing catalyst (Table 2).
Fig. 3. Selectivities for various products formed in the reaction of n-butane on ZSM-5 (Si/Al = 30) (A) and 2% Mo C/ZSM-5 (Si/Al = 30) (B) at 823 K in
2
time on stream.

F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
225
Fig. 4. Effects of Mo C content of ZSM-5 on the reaction of n-butane at 823 K: (A) Si/Al = 80; (B) Si/Al = 280. Data were taken at 40 min on stream.
2
Table 3
Some characteristic data for reaction of n-butane on various ZSM-5 and Mo C/ZSM-5 catalysts at 823 and 873 K
2
Catalyst
Si/Al
Conversion
(%)
Yields
ratio
Propane
Butene
Benzene
Xylene
Aromatics
ZSM-5
(30)
(30)
(30)
(50)
(50)
(50)
(80)
(80)
(80)
96.74
86.40
99.18
82.91
85.18
96.79
66.39
52.03
65.43
70.49
24.54
24.89
30.96
14.69
36.95
32.80
35.15
33.24
27.05
8.59
2.08
25.70
9.41
4.51
18.23
6.71
3.85
7.43
3.44
0.28
1.02
0.05
0.25
0.17
0.18
0.09
1.27
2.78
0.50
3.27
2.37
1.29
4.06
4.91
4.26
3.17
2.65
9.36
5.03
7.90
15.74
11.40
19.46
20.34
10.12
14.79
28.00
3.81
12.09
23.53
1.97
5.38
8.34
8.07
0.38
1.14
1.88
0.97
4.07
4.20
2.54
2.14
19.13
17.13
21.14
11.28
16.81
18.34
5.90
8.62
9.86
12.2
0.99
1.29
2.22
0.79
3.00
3.21
2.58
2.04
29.71
32.30
49.14
15.41
28.90
41.86
8.09
14.17
18.20
30.27
1.41
2.43
4.11
1.47
7.07
7.42
5.12
4.18
2% Mo C/ZSM-5
2
a
a
2% Mo C/ZSM-5
2
ZSM-5
2% Mo C/ZSM-5
2
2% Mo C/ZSM-5
2
ZSM-5
2
2
2
% Mo C/ZSM-5
% Mo C/ZSM-5
% Mo C/ZSM-5
2
a
2
b
(80)
2
ZSM-5
(280)
(280)
(280)
(Aerosil)
(Aerosil)
(Aerosil)
(Aerosil)
(Cabosil)
2
2
2
2
2
1
1
% Mo C/ZSM-5
% Mo C/ZSM-5
% Mo C/SiO
% Mo C/SiO
% Mo C/SiO
0% Mo C/SiO
0% Mo C/SiO
2
2
a
2
2
2
2
2
b
a
2
2
2
2
2
a
Data were taken at 873 K.
Data were taken at a flow rate of 7.5 ml/min.
b
3
.2.3. TPR measurement for used catalysts
In order to establish the amount and the reactivity of
ing methane (Tp = 853 K), ethane (Tp = 773 K), ethylene
(T = 753 K), and propane (T = 713 K). In the absence of
p
p
carbonaceous species formed on the catalyst surface dur-
ing the reaction, TPR measurements were performed with
hydrogen. When the excess carbon formed on Mo2C dur-
ing carburization was removed with H2 treatment at 873 K,
the subsequent TPR measurements gave only few carbon-
containing compounds. As the data presented in Fig. 5 show
a small amount of carbon (1.76 mg/g catalyst) was deposited
on 2% Mo2C/ZSM-5 (Si/Al = 80) during the reaction of
n-butane, which reacted with hydrogen above 500 K yield-
2 p
Mo C, the T values appeared at somewhat higher temper-
atures, but the amount of carbon calculated from the hydro-
carbons formed was more than an order of magnitude lower
(0.105 mg/g).
3.2.4. Effects of space velocity
Results showing the influence of space velocity on the
conversion and selectivities for various products are pre-
sented in Fig. 6. As expected the conversion is decreased

226
F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
Fig. 5. TPR spectra of 2% Mo C/ZSM-5 (Si/Al = 80) after reaction of n-butane at 823 K for 120 min.
2
Fig. 6. Effects of space velocity on the reaction of n-butane on ZSM-5 (Si/Al = 80) (A) and on 2% Mo C/ZSM-5 (Si/Al = 80) (B) at 823 K.
2
with the rise of the space velocity. The selectivities of aro-
matics also decreased, whereas an increase was observed
in these values of all unsaturated compounds, like ethene,
propene, and butene.
for the dehydrogenation and decomposition of n-butane. At
823 K we measured 8–12% of conversion. As regards the
formation of aromatics Mo C/SiO exhibited the higher ac-
tivity; therefore, more detailed measurements were carried
out only with this catalyst.
2
2
3
.2.5. Effects of supports
Mo2C deposited on other oxidic supports (alumina, tita-
The pure silica showed very little activity toward the
reaction of n-butane. Even at 873 K, the conversion was
below 2%. It was interesting to see first how the start-
nia, magnesia, and silica) was found to be an active catalyst

F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
227
Fig. 7. Reaction of n-butane on 2% MoO /SiO (A) and on 2% Mo C/SiO (B) at 873 K in time on stream.
3
2
2
2
ing material, 2% MoO3/SiO2 used for the preparation of
Mo2C/SiO2,behaves toward the reaction of n-butane. At the
beginning of the reaction, up to about 30 min, the forma-
tion of CO and H2O was observed, very likely as a result
of the reduction of MoO3. Other compounds were methane,
propane, and propene. Benzene and xylene were detected
even at 5 min of reaction, but only in minor amounts. With
the progress of the reduction of MoO3, and possibly with
the formation of carbide layer, the conversion and the selec-
tivities to aromatics significantly increased attaining max-
imum values at 20 min, when the production of CO and
H2O ceased. At the same time the selectivities to propene
and butene decreased. After 20–25 min opposite changes
were experienced. Similar features were registered at 823 K,
where the maximum occurred at 30 min. Data are shown in
Fig. 7A.
When Mo2C was prepared on SiO2, the formation of
CO and H2O was not observed and only slight changes
were experienced in the selectivities of products with time
on stream (Fig. 7B). The catalytic behavior of Mo2C/SiO2
was sensitive to the preparation of the sample, to the ori-
gin of the silica, and particularly to the amount of excess
carbon formed during the preparation of Mo2C. The 2%
Mo2C/SiO2 (Aerosil) exhibited a somewhat better catalytic
performance compared to Mo2C on Cabosil SiO2. In this
case the conversion of n-butane at 873 K approached 40%,
which gradually decreased with the time. Similarly as in the
case of Mo2C/ZSM-5, removing the excess carbon with H2
enhanced the activity of the catalyst; the extent of the en-
hancement depended on the amount of coke deposited on
the catalyst during the carburization process. At 2% Mo2C
content, the major products were various butenes, their total
selectivity attained a value of 40–45%. Interestingly, ben-
zene, xylene, and toluene also formed with a total selectivity
of 20–25%. Other products were in decreasing selectivities:
methane, propene, ethylene, propane, ethane, and isobutane.
Higher Mo C loading resulted in the enhanced selectivity of
2
butenes and in the slight diminution of that for the formation
of benzene. Some characteristic data obtained for various
samples are also listed in Table 3.
3
.3. Reaction of 1-butene on supported Mo2C
As the products of the dehydrogenation of n-butane
are butenes, it was important to examine the reaction of
-butene on the previously studied catalysts. The reactiv-
1
ity of 1-butene on ZSM-based catalysts is very high: the
conversion of 1-butene on pure ZSM-5 samples (Si/Al = 30
and 80) was more than 90% even at 723 K. In the calcula-
tion of the conversion, the isomerization of 1-butene, which
occurred to a great extent, was not taken into account. The
main products were propane followed by xylene, isobutane,
and benzene. The selectivities of aromatics were lower on
ZSM-5 (Si/Al = 80) compared to ZSM-5 (Si/Al = 30) and
their values slightly increased on the effect of 2% Mo2C.
The selectivities of the main products formed on pure and
Mo2C-containing ZSM-5 (Si/Al = 80) at 723 K are plot-
ted with time on stream in Fig. 8. At a lower temperature,
623 K, the selectivities of all the three aromatic compounds

228
F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
Fig. 8. Selectivities for various products in the reaction of 1-butene on ZSM-5 (Si/Al = 80) (A) and on 2% Mo C/ZSM-5 (Si/Al = 80) (B) at 773 K in time
2
on stream.
Table 4
Characteristic data for the reaction of 1-butene on ZSM-5 and 2% Mo C/ZSM-5 samples at 723 K
2
Catalyst
Si/Al
Conversion
(%)
Selectivity(%)
Yield of
ratio
Methane
Ethane
Ethylene
Propane
Propene
Isobutane
Benzene
Xylene
aromatics
H/ZSM-5
(30)
(30)
(80)
(80)
(280)
(280)
98.5
98.0
95.1
88.5
84.5
83.84
1.42
1.32
0.62
0.25
0.21
0.114
2.33
2.33
1.25
0.60
0.39
0.57
2.57
2.70
5.80
5.10
8.39
6.87
35.37
30.94
21.59
16.58
8.76
3.63
3.30
9.82
11.28
23.55
26.2
9.82
9.63
15.31
19.28
12.82
9.95
8.83
10.94
5.16
3.65
1.71
1.69
27.74
30.12
26.65
32.02
18.93
20.53
36.38
40.60
30.26
31.63
18.08
19.28
2
% Mo C/ZSM-5
2
H/ZSM-5
% Mo C/ZSM-5
2
2
H/ZSM-5
% Mo C/ZSM-5
2
6.28
2
Data were taken at 40 min of time on stream.
were reduced, whereas the formation of isobutane, n-butane,
and pentane greatly increased.
tivities for aromatics determined on both pure and Mo2C
containing ZSM-5 (Si/Al = 280) were only with a few per-
cent higher.
It was interesting to examine the reaction of 1-butene on
ZSM-5 (Si/Al = 280), too, which exhibited very little arom-
atizing behavior of n-butane. On pure ZSM-5 the conversion
of 1-butene was 84.5% at 723 K: the major process was
the cracking of butene yielding propene isobutane, propane,
ethylene, methane, and ethane in decreasing selectivities. In
addition, xylene (S = 18.7%), benzene (S = 1.7%), toluene
Mo2C/SiO2 was a much less active catalyst for the
1
-butene. The conversion was only 13% at 723 K, and se-
lectivities to aromatics were 3–5%. At 823 K the conversion
approached 20%, and the total selectivity of aromatics at-
tained a value of 7–8%.
(
S = 1.0%), and several C5 compounds were also formed.
4
4
4
. Discussion
On 2% Mo2C/ZSM-5 neither the conversion of 1-butene nor
the product distribution were altered to an appreciable ex-
tent. Most important data are collected in Table 4. Since the
reaction pathways of hydrocarbons, particularly the aroma-
tization of alkanes and alkenes, are strongly influenced by
the temperature, an attempt was made to measure the reac-
tion at 823 K, where the conversion of n-butane was studied.
The cracking processes of butene were very fast at this tem-
perature resulting in a larger deposition of coke, too. This
caused a more significant deactivation. Nevertheless, selec-
.1. Reaction of n-butane
.1.1. Pure ZSM-5 samples
It has been disclosed before that ZSM-5 is an efficient cat-
alyst for the reaction of n-butane [1]. The predominant prod-
uct on ZSM-5 (Si/Al = 56) at high conversion and at 773 K
was propane. The yield of aromatics was found to be very
small. Several attempts have been made to enhance the cat-
alytic activity of zeolites and to alter the product distribution

F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
229
by adding different metals and oxides to ZSM-5 [33–41]. As
4.1.2. Effects of Mo2C
regards the aromatization of n-butane, Zn and Ga proved to
be the most active promoters [41–46].
The addition of Mo2C to ZSM-5 markedly influenced the
catalytic performance of all ZSM-5 samples, and this feature
was well exhibited on the less effective ZSM-5 samples. The
main effects of Mo2C are as follows:
By extrapolating the selectivities to zero percent conver-
sion, on pure ZSM-5 (Si/Al = 56) the following compounds
were found: methane, ethane, propene, ethylene, butenes,
and hydrogen, which were considered as the primary reac-
tion products [44]. This product distribution was described
by the following reactions,
(i) With the exception of ZSM-5 (Si/Al = 280), the total
conversion of n-butane is slightly decreased,
(ii) The selectivities and yields for aromatics are enhanced
at the expense of the cracking reactions. Higher Mo2C
loading leads to a less active and selective catalysts.
(iii) The production of hydrogen and butenes is markedly
increased.
CH4 + C3H7+
ꢀ
(1)
(2)
(3)
+
+
n-C4H10 + H → C2H6 + C2H
5
+
ꢁ
H2 + C4H9
,
assuming the transient formation of pentacoordinated carbo-
nium ions. The carbenium ions may release protons to give
lower alkenes. As the selectivities for the above compounds
decreased in the later stage of the process, the formation
of propane came into the fore, another mode of activation
of n-butane, namely the hydride transfer reaction was as-
sumed [44].
Our systematic study showed that the conversion of n-
butane on ZSM-5 samples sensitively depends on the com-
position of zeolites, although there is no significant differ-
ence in their surface area. The composition also exerted a
drastic influence on the reaction pathway of n-butane; at a
low Si/Al ratio the aromatics are formed with higher selec-
tivities and relatively higher yields, whereas at a high Si/Al
ratio (280) aromatic compounds were produced with low (2–
The latter feature indicates that Mo C promotes the dehy-
drogenation of n-butane by providing new dehydrogenation
centers. As a consequence the production of aromatics is
considerably increased very likely through dimerization and
oligomerization of butenes on the Brönsted sites of ZSM-5.
When the selectivities of various products are extrapolated to
2
zero conversion for 2% Mo C-containing ZSM-5 (Si/Al =
2
80), we obtained the data listed in Table 5. This shows that
although the product ratios are different for Mo C/ZSM-5
2
samples, the primary products are the same as on the pure
ZSM-5. The dehydrogenation of n-butane is more favored
on Mo C/ZSM-5. Another important result is that aromatics
2
are not primary products, as they are missing at zero con-
version even on Mo C-containing ZSM-5, too. This implies
2
again that both benzene and xylene are produced in the sec-
3
%) selectivities (Fig. 2). This change in the selectivities can
ondary processes.
be attributed to the variation of acidic sites, which decrease
with the increase of Si/Al ratio (Table 1).
From the extrapolation of selectivities determined for
ZSM-5 (Si/Al = 80) to zero percent conversion (Fig. 8),
we obtained the same primary products as above for zeolite
In previous studies it was established that zeolite acid-
ity is an important parameter not only for the oligomer-
ization of alkenes but also for the production of alkenes
from alkanes [1]. The number of acid sites and acid strength
are, however, considerably reduced by loading Mo C onto
2
(
Si/Al = 56) [44]. The values of the three reactions are in the
ZSM-5 [47], so we have to assume that this drawback is
overcompensated by the enhanced production of butenes in
the dehydrogenation process.
ratio 38:29:15 (Table 5). This suggests that the rupture of the
C–C bond on this ZSM-5 is more dominant than that of the
C–H bond. A change occurred in this ratio for ZSM-5 (280)
The significant effect of Mo C on the activation of n-
2
(
Table 5), indicating that on zeolites with decreased acidity
butane and on the subsequent reactions is well exhibited by
the cleavage of the C–H bond became more dominant.
As the total conversion increases, benzene and xylene
appear in the products. This suggests that the primary prod-
ucts, very likely the butenes, undergo secondary reactions
involving oligomerization, cyclization, and hydrogen trans-
fer before they leave the pore system of ZSM-5.
2 2
the results obtained on Mo C/SiO . In this case the situa-
tion is simpler as silica is practically inactive for the reaction
of n-butane. Loading the silica with Mo2C produced a rela-
tively active catalyst (Table 3). The main process is the de-
hydrogenation of n-butane, which occurs with a selectivity
of 50–70%. Cracking reactions may also proceed as indi-
Table 5
Selectivities in n-butane conversion at zero conversion (extrapolated) at 823 K
Catalyst
Si/Al
Selectivity (in moles produced per 100 mol butane converted)
ratio
Methane
Propene
Ethane
Ethylene
Hydrogen
Butene
Isobutane
Aromatics
H-ZSM-5
(80)
(80)
(280)
38
40
39
1
36
35
35
2
29
14
33
1
40
34
35.5
3
15
20
22
16
20
22
80
4
7
7
0
0
0
0
2
% Mo C/ZSM-5
2
H-ZSM-5
% Mo C/SiO
2
(Aerosil 380)
120
29
2
2
Data were taken from the measurements of the effect of space velocity.

230
F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
Fig. 9. Product selectivity as a function of n-butane conversion over ZSM-5 (Si/Al = 80) (A), ZSM-5 (Si/Al = 280) (B), and 2% Mo C/ZSM-5 (Si/Al = 80)
2
(C).
cated by the formation of various lower hydrocarbons. An
interesting feature of this catalytic system is the formation
of aromatics, benzene and xylene.
1-butene was very limited on Mo2C/SiO2. This feature sug-
gests that the direct transformation of the primary prod-
uct of n-butane activation on Mo2C plays a crucial role
in the process leading to the formation of aromatics. In
other words, a fraction of the intermediate formed at the
Mo2C/ZSM-5 or Mo2C/SiO2 interface, or more precisely on
the highly dispersed Mo2C interacting with the acid sites of
these support, is effectively converted into aromatics before
they can be transformed into other molecules. This interme-
diate is very likely butyl species,
From the extrapolation of the selectivities to zero per-
cent conversion we determined the primary products of the
reaction on this sample, too (Fig. 9). These are basically
different compared to ZSM-5-based catalysts. The predom-
inant products were hydrogen and butenes, and methane,
ethane, propane, and ethylene appeared only in traces (Ta-
ble 5). This suggests that the dehydrogenation is the primary
process on Mo2C/SiO2, and all the other reactions, including
cracking and aromatization, occur subsequently. The forma-
tion of aromatics over this catalyst strongly supports the
idea that the role of Mo2C is not only the activation of n-
butane and the production of butenes, but somehow Mo2C
also participates in the steps leading to aromatics. As silica
contains no Brönsted sites we may speculate that during the
carburization of MoO3 or the deposition of Mo2C itself cre-
ates Brönsted sites on silica. Detailed infrared spectroscopic
measurements, however, did not prove this assumption [47].
Nevertheless, it was demonstrated that the Mo deposited on
SiO2 reacted with OH groups of the support and induced
Lewis acidity. Carburization resulted in further OH con-
sumption and the appearance of stronger Lewis acid sites.
Accordingly, we may assume that the oligomerization and
aromatization processes of butenes may proceed on Lewis
acidic sites of the Mo2C/SiO2 catalyst.
C4H10(a) → C4H9(a)
which is formed in the activation of n-butane on Mo2C. This
idea seems to be supported by the results obtained in the
study of the reactions of butyl species on Mo C/Mo(100) in
2
UHV system [48]. A C H fragment was prepared by the
4
9
thermal and/or photo dissociation of C H I. TPD measure-
4
9
ments revealed that beside the dehydrogenation and hydro-
genation processes the coupling of C H species also oc-
4
9
curred to a very limited extent.
5. Conclusion
(i) The conversion and reaction pathways of n-butane on
ZSM-5 sensitively depend on the composition of zeo-
lite. The main reaction is the cracking, and the aromati-
zation proceeds mainly on the sample with a low Si/Al
ratio.
The study of the reaction of 1-butene on the previously
applied catalysts shed more light on this complex system.
We obtained that Mo2C exerted only slight promotion on
the aromatization of 1-butene on ZSM-5 samples. In har-
mony with this, the formation of aromatic compounds from
(ii) Deposition of Mo C changed the catalytic performance
2
of the ZSM-5 and promoted the dehydrogenation and
aromatization processes.

F. Solymosi, A. Széchenyi / Journal of Catalysis 223 (2004) 221–231
231
(
iii) Reaction products determined at zero conversion on
Mo2C/ZSM-5 samples suggested that aromatics are
formed in the secondary processes very likely in the
reactions of butenes on the acidic sites of ZSM-5.
iv) Mo2C catalyzed the dehydrogenation and aromatiza-
tion of n-butane even when it was deposited on an inac-
tive silica support, which suggested that the oligomer-
ization and aromatization processes of butenes might
proceed on the Lewis sites of the Mo2C/SiO2.
[14] Y.-Y. Shu, D. Ma, L.-L. Su, L.-Y. Xu, Y.-D. Xu, X.-H. Bao, Stud. Surf.
Sci. Catal. 136 (2001) 27.
[15] R. Ohnishi, L. Xu, K. Issoh, M. Ichikawa, Stud. Surf. Sci. Catal. 136
(2001) 393.
[
[
16] F. Solymosi, A. Szöke, Appl. Catal. A 166 (1998) 225.
17] F. Solymosi, R. Németh, Catal. Lett. 62 (1999) 197.
(
[18] F. Solymosi, A. Szöke, Stud. Surf. Sci. Catal. 119 (1998) 355.
[19] F. Solymosi, R. Németh, L. Óvári, L. Egri, J. Catal. 195 (2000) 316.
[20] F. Solymosi, L. Bugyi, A. Oszkó, Catal. Lett. 57 (1999) 103.
[21] L. Bugyi, F. Solymosi, A. Oszkó, I. Horváth, J. Catal. 185 (1999) 160.
[22] L. Bugyi, A. Oszkó, F. Solymosi, Surf. Sci. 516 (2002) 74.
[23] L. Bugyi, A. Oszkó, F. Solymosi, Surf. Sci. 519 (2002) 139.
(
v) As Mo2C exerted much less influence on the reaction
of 1-butene occurring on ZSM-5 and only slightly en-
hanced the formation of aromatics, it was inferred that
the butyl species, the primary product of the activation
of butane on Mo2C, is effectively converted to a com-
pound leading to aromatics.
[24] L. Bugyi, A. Oszkó, F. Solymosi, Surf. Sci. 539 (2003) 1.
[
[
25] F. Solymosi, R. Németh, A. Széchenyi, Catal. Lett. 82 (2002) 213.
26] S. Yuan, S.B. Derouane-Abd Hamid, Y. Li, P. Ying, Q. Xin, E.G. Der-
ouane, C. Li, J. Mol. Catal. A: Chem. 184 (2002) 257.
[
[
[
27] S. Yuan, S.B. Derouane-Abd Hamid, Y. Li, P. Ying, Q. Xin, E.G. Der-
ouane, C. Li, J. Mol. Catal. A: Chem. 180 (2002) 245.
28] J.R. Anderson, K. Foger, T. Mole, R.A. Rajadhyakska, J.V. Sanders,
J. Catal. 58 (1997) 114.
29] A.J. Brungs, A.P.E. York, J.B. Claridge, C. Márquez-Alvarez, M.L.H.
Green, Catal. Lett. 70 (2000) 117.
Acknowledgments
[
[
30] J.S. Lee, S.T. Oyama, M. Boudart, J. Catal. 106 (1987) 125.
31] L. Leclercq, M. Provost, H. Pastor, J. Grimblot, A.M. Hardy, L.
Gengembre, G. Leclercq, J. Catal. 117 (1989) 371.
The work was supported by the Hungarian Academy of
Science and by Grant OTKA T 38233 and TS 40877.
[
32] M.A. Chesters, C. De La Cruz, P. Gardner, E.M. McCash, P. Pudney,
G. Shadi, N. Sheppard, J. Chem. Soc., Faraday Trans. 86 (1990) 2757.
33] M. Guisnet, N.S. Gnep, F. Alario, Appl. Catal. A 89 (1992) 1.
34] N. Kumar, L.E. Lindfors, R. Byggningsbacka, Appl. Catal. A 139
[
[
References
(
1996) 189; Appl. Catal. A 147 (1996) 175.
[
1] Y. Ono, Catal. Rev.-Sci. Eng. 34 (1992) 179.
[
[
35] W.J.H. Dehertog, G.F. Fromen, Appl. Catal. A 189 (1999) 63.
36] I.M. Harris, J. Dwyer, A.A. Garforth, C.H. Mcateer, W.J. Ball, Stud.
Surf. Sci. Catal. 47 (1989) 271.
[2] L. Wang, L. Tao, M. Xie, G. Xu, J. Huang, Y. Xu, Catal. Lett. 21 (1993)
35.
[
[
[
[
3] Y. Xu, L. Wang, M. Xie, X. Gou, Catal. Lett. 30 (1995) 135.
4] F. Solymosi, A. Erdöhelyi, A. Szöke, Catal. Lett. 32 (1995) 43.
5] F. Solymosi, A. Szöke, J. Cserényi, Catal. Lett. 39 (1996) 157.
6] F. Solymosi, J. Cserényi, A. Szöke, T. Bánsági, A. Oszkó, J. Catal. 165
[
37] T. Yahsima, S. Ejiri, K. Kato, M.M. Ishaq, M. Tanigawa, T. Komatsu,
S. Namba, Stud. Surf. Sci. Catal. 447 (1996) 100.
[38] N. Kumar, R. Byggningsbacka, L.E. Lindfors, React. Kinet. Catal.
Lett. 61 (1997) 297.
(
1997) 150.
7] D.W. Wang, J.H. Lunsford, M.P. Rosynek, Top. Catal. 3 (4) (1996)
99.
8] J.H. Lunsford, M.P. Rosynek, D.W. Wang, Stud. Surf. Sci. Catal. 107
1997) 257.
9] D.W. Wang, J.H. Lunsford, M.P. Rosynek, J. Catal. 169 (1997) 347.
[
[
39] P.C. Doolan, P. Pujado, Hydrocarbon Process. 68 (1989) 72.
40] P. Fejes, J. Halász, I. Kiricsi, Z. Kele, G. Tasi, E. Derouane, K. Dooley,
J.N. Armor, Stud. Surf. Sci. Catal. 75 (1993) 421.
[
[
[
2
[41] T. Inui, Y. Ishihara, K. Kawasi, H. Matsuda, Stud. Surf. Sci. Catal. 49
(
(1989) 1183.
[42] G. Giannetto, A. Montes, N.S. Gnep, A. Fluorentino, P. Cartaud, M.
Guisnet, J. Catal. 145 (1994) 86.
[
10] R.W. Borry III, E.C. Lu, K. Young-ho, E. Iglesia, Stud. Surf. Sci.
Catal. 119 (1998) 403.
11] J.-Z. Zhang, M.A. Long, R.F. Howe, Catal. Today 44 (1998) 293.
12] L.S. Liu, L. Wang, R. Ohnishi, M.J. Ichikawa, J. Catal. 181 (1999)
[
[
[
[
[
[
43] Y.F. Chu, A.W. Chester, US patent 4392989, 1983.
[
[
44] Y. Ono, K. Kanae, J. Chem. Soc., Faraday Trans. 87 (1991) 663, 669.
45] R. Le Van Mao, J. Yao, B. Sjiariel, Catal. Lett. 6 (1990) 23.
46] R. Le Van Mao, L. Dufresne, J. Yao, Appl. Catal. 65 (1990) 143, 173.
47] F. Solymosi, L. Óvári, J. Mol. Catal. A: Chem. 207 (2004) 35.
48] L. Bugyi, A. Oszkó, F. Solymosi, Surf. Sci., submitted for publication.
175.
[
13] S.B. Derouane-Abd Hamid, J.R. Anderson, I. Schmidt, C. Bouchy,
C.J.H. Jacobsen, E.G. Derouane, Catal. Today 63 (2000) 461.