Aromatization of methanol and methylation of benzene over Mo2C/ZSM-5 catalysts

Article abstract of DOI:10.1016/j.jcat.2007.02.017

The adsorption and reaction pathways of methanol were investigated on various Mo₂C-containing catalysts characterized by XPS and surface acidity measurements. FTIR spectroscopy indicated the formation of a strongly bonded methoxy species on Mo₂C/ZSM-5(80) at 300 K that which was converted into adsorbed dimethyl ether at 373-473 K. TPD experiments following the adsorption of methanol on both ZSM-5 and Mo₂C/ZSM-5 at 300 K showed desorption profiles corresponding to unreacted methanol and decomposition products (H₂, CH₂O, CH₃CHO, CH₃{single bond}O{single bond}CH₃, and C₂H₄). Unsupported Mo₂C catalyzes only the decomposition of methanol. The same feature was observed for silica-supported Mo₂C. But a completely different picture was obtained when ZSM-5 was used as a support, which is known to be an active material in converting methanol into ethylene. The aromatization of methanol also occurred on this zeolite, but to only a limited extent. The deposition of Mo₂C on ZSM-5 markedly enhanced the formation of aromatics (benzene, toluene, xylene and C₉₊), however. The highest yield of the formation of aromatics was measured for 5% Mo₂C/ZSM-5 (SiO₂/Al₂O₃ = 80) at 773 K. It is assumed that Mo₂C opens a new route for the activation of methanol and also for the reactions of ethylene thus formed. Further experiments showed that Mo₂C/ZSM-5(80) was able to catalyze the methylation of benzene with methanol, which explains the formation of toluene, xylenes, and C₉₊ aromatics in the reaction of methanol alone.

Full text of DOI:10.1016/j.jcat.2007.02.017

Journal of Catalysis 247 (2007) 368–378
www.elsevier.com/locate/jcat
Aromatization of methanol and methylation of benzene
over Mo C/ZSM-5 catalysts
2
∗
Róbert Barthos, Tamás Bánsági, Tímea Süli Zakar, Frigyes Solymosi
Institute of Solid State and Radiochemistry, University of Szeged and Reaction Kinetics Research Group of the Hungarian Academy of Sciences, P.O. Box 168,
H-6701 Szeged, Hungary
Received 30 October 2006; revised 19 February 2007; accepted 21 February 2007
Abstract
The adsorption and reaction pathways of methanol were investigated on various Mo C-containing catalysts characterized by XPS and surface
2
acidity measurements. FTIR spectroscopy indicated the formation of a strongly bonded methoxy species on Mo C/ZSM-5(80) at 300 K that
2
which was converted into adsorbed dimethyl ether at 373–473 K. TPD experiments following the adsorption of methanol on both ZSM-5 and
Mo C/ZSM-5 at 300 K showed desorption profiles corresponding to unreacted methanol and decomposition products (H , CH O, CH CHO,
2
2
2
3
CH –O–CH , and C H ). Unsupported Mo C catalyzes only the decomposition of methanol. The same feature was observed for silica-supported
3
3
2
4
2
Mo C. But a completely different picture was obtained when ZSM-5 was used as a support, which is known to be an active material in converting
2
methanol into ethylene. The aromatization of methanol also occurred on this zeolite, but to only a limited extent. The deposition of Mo C on
2
ZSM-5 markedly enhanced the formation of aromatics (benzene, toluene, xylene and C ), however. The highest yield of the formation of
9+
aromatics was measured for 5% Mo C/ZSM-5 (SiO /Al O = 80) at 773 K. It is assumed that Mo C opens a new route for the activation
2
2
2
3
2
of methanol and also for the reactions of ethylene thus formed. Further experiments showed that Mo C/ZSM-5(80) was able to catalyze the
2
methylation of benzene with methanol, which explains the formation of toluene, xylenes, and C aromatics in the reaction of methanol alone.
9+
©
2007 Elsevier Inc. All rights reserved.
Keywords: Aromatization of methanol; Decomposition of methanol; Methylation of benzene Mo C; Mo C/ZSM-5; Mo C/SiO
2
2
2
2
1
. Introduction
convert methane into benzene with 80% selectivity at 10–15%
conversion, which ZSM-5 alone or promoted with Zn and Ga
cannot do [26–32]. Based on this finding, it seemed worthwhile
to examine the influence of Mo2C on the reaction and aromati-
zation of CH3OH on ZSM-5 and other oxidic supports.
The direct conversion of alkanes into aromatics is an impor-
tant process for industry, and thus extensive research on this
reaction has been performed. The results obtained on various
catalysts and the possible mechanism of this complex process
are well documented in several excellent reviews [1–6]. Pre-
vious studies have demonstrated that depending on the reac-
tion conditions and on the catalyst systems, methanol also can
be transferred into olefins and aromatics on ZSM-5 zeolites
2. Experimental
2
.1. Methods
[
7–16]. The latter process is promoted by Ga and, more par-
Catalytic reaction was carried out at 1 atm of pressure
in a fixed-bed, continuous-flow reactor comprising a quartz
tube connected to a capillary tube [23,24]. The flow rate was
ticularly, by Zn additive [12–16]. Recently it has been found
that Mo C is also an effective promoter in the aromatization of
several hydrocarbons and ethanol occurring on ZSM-5 [17–25].
Moreover, this catalyst exhibits a unique behavior in that it can
2
4
0 mL/min, the carrier gas was Ar, and the methanol content of
the carrier gas was 10%. A ∼0.3 g catalyst sample was loosely
compressed and broken into small (1–2 mm) fragments. Re-
action products were analyzed by gas chromatography using
a Hewlett-Packard 6890 gas chromatograph equipped with a
*
Corresponding author. Fax: +36 62 420 678.
E-mail address: fsolym@chem.u-szeged.hu (F. Solymosi).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.02.017

R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
369
thermal conductivity detector for the analysis of H2, CO, CO2,
H2O; a flame ionization detector for the analysis of organic
compounds; and PORAPAK Q + PORAPAQ S columns. The
calibration of all compounds was carried out separately. We
note that the values determined for individual compounds were
not altered by the presence of any other major products. The
conversion of methanol was calculated taking into account the
amount consumed. The selectivity for reaction products, S, was
defined as
To remove the excess carbon deposited on the Mo2C during
preparation and the Mo–O species produced during passivation,
the Mo2C-containing catalyst was reduced before the catalytic
measurements in situ in a H2 stream for 60 min at 873 K.
The MoO3/support samples were prepared by impregnating the
support with a basic solution of ammonium heptamolybdate to
yield 2, 5, and 10 wt% of MoO3. The suspension was dried and
calcined at 863 K for 5 h. The following materials were used
as supports: ZSM-5 with SiO2/Al2O3 ratios of 30, 80, and 280
2
and BET areas of 400, 425, and 400 m /g, and SiO2 (Aerosil,
xini
S = ꢀ
,
2
i
380 m /g). CH OH was a product of SHARLAU with a purity
3
xini
i
of 99.9%.
where xi is the fraction of product i, and ni is the number
of carbon atoms in each molecule of gaseous products. FTIR
spectra of adsorbed gases were recorded with a Biorad (Digi-
lab. Div. FTS 155) instrument with a wave number accuracy of
3
. Results
3
.1. Characterization of the catalyst
−1
±
4 cm . All the spectra presented are difference spectra. The
background was the spectrum of the pretreated sample before
the adsorption of reactants. Temperature-programmed desorp-
tion (TPD) curves were obtained by quadrupole mass spectrom-
etry.
All of the catalysts used in this study have been previ-
ously characterized by XPS measurements [23,24]. The BEs
for Mo(3d5/2) and Mo(3d3/2) showed some slight variation
with different samples, but dropped to 227.8–228.2 and 230.7–
The adsorption of methanol (10% in helium carrier gas) was
flown through the catalyst for 15 min at 300–304 K. The cat-
alyst was then purged with pure He to remove any weakly
adsorbed methanol and lined for 90 min. The gas flow was
introduced to the UHV system via a quartz capillary tube con-
taining a leak UHV valve. A constant operating pressure of
2
31.1 eV, respectively; that for C(1s) dropped to 283.8 eV.
These values are consistent with those attributed to Mo2C [34–
6]. In terms of the acidity of the catalysts, we refer to our pre-
3
vious work, which used the same samples [24]. We found that
during the preparation of Mo/ZSM-5 catalyst, Mo compounds
reacted with OH groups of ZSM-5. As a result, the Brönsted
sites decreased markedly in number, but were not completely
eliminated. The greater the carbide content, the greater the de-
crease. At the same time, the Lewis acidity was somewhat en-
hanced by Mo2C deposition.
−6
8
× 10 Torr was maintained in the system during all TPD
runs. TPD experiments were performed using a He flow of
4
0 mL/min and a heating rate of 10 K/min from 300 to 873 K.
The acidic properties of catalysts have been determined by
NH -TPD and FTIR methods using the adsorption of pyridine
3
and CO, as described previously [24]. The XPS measurements
were performed in a Kratos XSAM 800 instrument at a base
3
.2. FTIR and TPD measurements
−8
pressure of 10 Torr using MgKα primary radiation (14 kV,
0 mA). To compensate for possible charging effects, bind-
ing energies (BEs) were normalized to the Fermi level for the
The primary aim of the IR study was to determine the effect
1
of Mo2C on the adsorption of methanol on ZSM-5 and it surface
processes. Fig. 1A shows the IR spectra of ZSM-5(80) after the
adsorption of methanol at 300 K for 15 min, and subsequent
heating of the sample to higher temperatures under continu-
ous evacuation. Table 1 presents the absorption bands observed
at 300 K, along contains the characteristic IR features of vari-
ous compounds detected after the surface reaction of adsorbed
methanol [37–47]. The heat treatment of adsorbed layer caused
a gradual attenuation of absorption bands without any signifi-
cant shift or appearance of new spectral features. We found the
Mo C. The amount of coke deposited on the catalyst during
2
the reaction was determined by temperature-programmed reac-
tion (TPR). The catalyst was cooled in flowing argon and then
heated in a H stream with a rate of 5 K/min; the hydrocarbons
2
thus formed were then measured.
2
.2. Materials
Mo2C-containing catalysts were the same as used our other
same features in the presence of 2 or 10% Mo C (Fig. 1B);
2
studies [23,24]. Unsupported Mo2C was prepared by the car-
burization of MoO3 (product of ALFA AESAR) by C2H6/H2
however, after annealing the adsorbed layer, weak absorption
−
1
features appeared at 2971 and 2925 cm
.
[
33]. The oxide was heated under a 10% v/v C2H6/H2 gas
mixture from room temperature to 900 K at a heating rate of
.8 K/min. Afterward the sample was cooled to room tem-
perature under argon. The carbide was passivated in flowing
TPD spectra for various products are presented in Fig. 2.
In the case of pure ZSM-5(80), the following compounds
were formed: methanol (T = 352 K), methane (T = 354 K),
0
p
p
formaldehyde (T = 354 and 470 K), H O (T = 370 and
p
2
p
2
1
% O /He at 300 K. The surface area of Mo C is 20 m /g.
542 K), C H (T = 531 K), and H (T = 358 K) (Fig. 2A).
2
2
2
4
p
2
p
In certain cases, Mo2C was prepared in situ in the catalytic
reactor when the last step was missing. Supported Mo2C cat-
alysts were prepared in a similar manner by the carburiza-
tion of MoO3-containing supports with C2H6/H2 gas mixture.
A small amount of dimethyl ether (T ∼ 520 K) was also de-
p
tected in the desorbing products. Calculation showed that the
chemicals desorbed in the same peaks as methanol are mostly
its fragments. When 2% Mo2C was deposited on ZSM-5(80),

370
R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
only slight variations were experienced in the peak temper-
atures and in the amounts of desorbed compounds. At 10%
Mo2C content, however, new high-temperature peaks devel-
oped for H2 (Tp = 574 K), CO (Tp = 591 K), and CH4 (Tp =
3.3. Reaction of methanol
3.3.1. Pure Mo2C catalyst
Mo2C prepared by a C2H6/H2 gas mixture exhibited high ac-
tivity. The decomposition of methanol was observed at very low
temperature (473 K). At 623 K, the conversion already reached
5
72 K) (Fig. 2B). The characteristics of the desorption of other
products remained practically unaltered.
∼
90%. Besides hydrogen, CO, CO , and methane were the
2
main carbon-containing products. Aromatic compounds were
detected only in trace amounts even at 773 K. The formation of
dimethyl ether also occurred; however, its selectivity remained
below ∼6% in the temperature range of 573–773 K.
3
.3.2. Supported Mo2C catalyst
By depositing 2 or 10% Mo2C in highly dispersed state on
silica, we obtained a somewhat more effective catalyst. The
main C-containing products at 573 K were methane (S = 35%),
CO (S = 11%), and dimethyl ether (S = 32%). Whereas the se-
lectivity of the first two compounds increased with increasing
temperature, that of dimethyl ether decreased and almost disap-
peared at 773–873 K. Interestingly, ethylene was formed with
low selectivity, a maximum with 2.5% at 573–873 K. There was
no sign of formation of any aromatics.
Better results were obtained using zeolites. In harmony with
the previous studies [4,7–16], the ZSM-5 alone was an active
material in converting methanol into other compounds. The
reaction on ZSM-5 samples was observed at very low temper-
atures (423–473 K), yielding mostly dimethyl ether. The con-
version reached 99% even at 573 K. At this temperature, the
formation of several olefins and aromatics (benzene, toluene,
xylenes, and C9+) was seen. The production of aromatics de-
pended on the Si/Al ratio; it was the highest on ZSM-5(30) and
the lowest on ZSM-5(280) (Table 2). A general feature of ZSM-
Fig. 1. FTIR spectra of ZSM-5(80) (A) and 2% Mo C/ZSM-5(B) following the
2
adsorption of methanol (1.5 Torr) at 300 K, and after heating the samples to
different temperatures during continuous degassing.
5
samples is that the selectivity of ethylene and propylene grad-
Fig. 2. TPD spectra following CH OH adsorption on pure ZSM-5(80) (A), 10% Mo C/ZSM-5(80) (B) at 300 K.
3
2

R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
371
Table 1
Characteristic absorption bands following the adsorption of methanol and its decomposition products at 300 K
Mode
Gas phase
CH OH
CH OH
CH OH
3
CH OH
CH –O–CH
3
C H
2 4
3
3
3
3
CH OH
3
Pd/SiO
Mo C/Mo(100)
ZSM-5
Mo C/ZSM-5
ZSM-5
Mo C/ZSM-5
2
2
2
2
[44]
[47]
Present study
Present study
Present study
Present study
γ
3681
3000
OH
νasCH
2956
2846
2936
(2993) 2958
2857
(2996) 2958
2856
2972
2947
2842
2920
2852
3
νas(CH )
2
νsCH
2844
3
νaCH
δasCH
1441
1150
2
1477
1455
1474
1454
1468
1405
1472, 1465
1404
1457
1427
1477
1441
1379
3
δsCH
3
γ aCH
3
CH -wag (s)
2
νCC
νCO
1021
1033
Table 2
a
Characteristic data for the reaction of methanol on various catalysts
Catalyst
Tempera-
ture (K)
Selectivity (%)
CO Methane
0.6
ꢀ
Ethylene
Ethane
Propylene
Benzene
Toluene
Xylenes
C
aromatics
9
+
ZSM-5(30)
773
6.3
15.5
17.7
2.4
2.0
12.5
11.1
2.6
3.0
8.2
9.4
12.7
14.7
6.7
9.4
30.2
36.5
823
1.2
13.3
2
% Mo C/ZSM-5(30)
773
0.7
2.6
7.7
19.0
13.7
13.8
1.8
1.6
10.9
7.2
3.6
5.6
10.1
13.4
14.5
16.6
8.7
9.2
36.9
44.8
2
823
5
% Mo C/ZSM-5(30)
773
773
773
5.0
28.3
47.0
9.9
8.2
1.5
2.2
3.1
1.7
3.6
1.6
12.2
4.6
17.4
4.9
13.4
5.0
46.2
16.0
2
10% Mo C/ZSM-5(30)
17.8
2
ZSM-5(80)
0.0
0.6
2.4
7.9
19.8
18.9
0.9
1.2
24.1
20.4
1.6
2.1
5.6
7.2
9.3
12.9
3.7
6.2
20.3
28.3
823
2
5
1
% Mo C/ZSM-5(80)
773
0.4
2.1
4.9
17.0
16.4
13.9
0.7
0.9
21.4
12.1
0.4
0.2
6.9
8.9
12.4
17.7
5.4
12.3
25.1
39.0
2
823
% Mo C/ZSM-5(80)
773
1.9
6.6
19.7
32.6
6.8
5.9
1.0
1.1
3.8
1.7
2.5
2.2
9.0
6.4
22.3
12.1
29.1
28.3
62.8
49.1
2
823
0% Mo C/ZSM-5(80)
773
6.0
2.2
35.0
21.4
9.7
12.0
1.5
1.0
3.6
9.0
2.3
1.7
8.2
7.0
16.9
19.7
11.2
14.9
38.6
43.2
2
823
ZSM-5(280)
773
0.0
0.0
0.9
1.7
18.6
22.3
0.2
0.3
44.6
45.9
0.3
0.1
2.6
3.1
3.3
3.2
0.8
0.6
7.0
7.0
823
2
5
1
% Mo C/ZSM-5(280)
773
0.0
0.6
2.1
5.4
16.0
20.1
0.3
0.4
41.0
39.0
0.3
0.1
4.3
6.5
5.8
6.9
1.8
1.4
12.2
15.0
2
823
% Mo C/ZSM-5(280)
773
0.0
1.0
2.9
9.2
12.1
17.9
0.2
0.6
30.1
28.8
1.1
3.9
4.8
10.9
10.5
12.6
4.4
2.7
20.7
30.0
2
823
0% Mo C/ZSM-5(280)
2
773
0.7
2.7
4.7
17.8
15.3
16.2
0.4
0.9
28.3
19.2
2.0
3.5
9.6
11.4
11.0
14.7
3.2
4.8
25.8
34.4
823
a
The conversion of methanol at 773–823 K was ∼100%. Data were taken at 75 min of the reaction.
ually increased and that for butene decreased with increasing
temperature. The selectivities of aromatics, toluene, xylene, and
benzene, were affected only slightly by the changing tempera-
ture. Interestingly, methane, which was the main hydrocarbon
product on Mo2C/SiO2 catalysts, was formed only above 773 K
with low selectivity. The effect of temperature on the produc-
tion of several compounds on ZSM-5(80) is shown in Fig. 3A.
Note that the conversion and the product distribution on this
sample remained the same even after 10 h of reaction at 773 K.
Depositing Mo2C on ZSM-5 slightly enhanced the con-
version at low temperature and appreciably influenced the
product distribution and its temperature dependence. This is
illustrated for 5% Mo2C/ZSM-5(80) in Fig. 3B. Whereas the
selectivity of ethylene varied between 5 and 8% in the tem-
perature range of 573–873 K, that for propylene increased
from 3.5% at 573 K to 15% at 723 K, then fell to ∼7% at
773 K. The selectivity of methane was very low, 1.5–3%,
at 573–673 K, but increased dramatically at high tempera-
ture, to 20% at 773 K. On this sample, the selectivity of
aromatics grew gradually with increasing temperature, with
the highest value obtained at 773 K. Following the reac-
tion in time on stream at 773 K, we experienced an ini-

372
R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
Fig. 3. Reaction of methanol over ZSM-5(80) (A) and 5% Mo C/ZSM-5(80) (B) at different temperatures. Data were obtained during gradual heating. At every
2
temperature the sample was kept for 75 min.
Fig. 4. Changes in the selectivity of various products on 5% Mo C/ZSM-5(80) at 773 K in time on stream (A). Effects of Mo C content of ZSM-5(80) on the
2
2
selectivities of various products formed at 773 K (B). Data were taken at 75 min of the reaction.
tial decline in the selectivity of toluene and C9+ aromatics
timum temperature (773 K), the selectivity of total aromatics
reached 63%, twice than that found on Mo2C-free ZSM-5(80)
(Table 2). This is due mainly to the dramatically enhanced for-
mation of C9+ aromatics. Further increases in Mo2C content
produced somewhat less selective catalysts.
(Fig. 4A).
The effect of Mo2C content on the representative data is
shown in Fig. 4B. In terms of aromatization, a more marked
promoting effect was experienced using 5% Mo2C. At the op-

R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
373
Fig. 5. Reaction of methanol over ZSM-5(280) (A) and 5% Mo C/ZSM-5(280) (B) at different temperatures. Data were obtained during gradual heating. At every
2
temperature the sample was kept for 75 min.
The temperature dependence of the reaction for the ZSM-
(280) sample is presented in Fig. 5A. Adding Mo2C to this
ZSM-5 also enhanced its performance in terms of the formation
of aromatics (Fig. 5B). The highest selectivity value (30%) for
in argon flow. Some TPR curves are displayed in Fig. 7. The
hydrogenation of carbon started at ∼700 K yielding methane
(T = 873 K) and small amounts of ethylene (T = 848 K),
5
p
p
ethane, and propylene (T = 823 K). The total amount of car-
p
the total aromatics on 5% Mo C/ZSM-5(280) was attained at
C cat
bon was 0.36 mg /g . The peak temperatures agree well with
2
8
23 K. Important data for the reaction of methanol on ZSM-5
those determined after the reaction of several hydrocarbons
(
280)-based catalysts are also collected in Table 2. Note that
on Mo C/ZSM-5(80) prepared in the same manner [22–24];
2
repeating the measurements on Mo C-containing samples, we
found some scattering in selectivity, but the general features of
catalytic performance remained the same.
however, much less coke was formed in the aromatization of
methanol.
2
Next, we examined the effect of space velocity on the forma-
tion of various compounds. Previous studies have shown that
the product distribution sensitively depends on the contact time
of the methanol [9,11,12]. To decrease the conversion, we used
only 0.1 g of catalyst. The conversion of methanol and the se-
lectivity of aromatics and olefins decreased with increasing flow
3
.4. Study of the reaction between methanol and benzene
It has been reported previously that the methylation of ben-
zene to C –C aromatics can be easily realized by methanol
7
9+
over acidic ZSM-5 catalysts at 673 K [13–16]. Because these
aromatic compounds are formed with increased selectivities
and yields on Mo2C-doped ZSM-5, it appears likely that they
are produced in secondary processes also promoted by Mo2C.
As shown in Fig. 8, flowing benzene + methanol gas mix-
ture through Mo2C/ZSM-5 catalyst at 573–773 K, a significant
enhancement in the rate of formation of C7–C9+ aromatics
occurred compared with the case in which benzene was not
present. Note that benzene underwent very little decomposi-
tion in this temperature range, so its products make a negligible
contribution to the methylation of benzene. Repeating this ex-
periment with Mo2C-free zeolite, we also found enhanced for-
mation of methylated benzene, but with much lower yields than
on the Mo2C/ZSM-5 catalyst under the same experimental con-
ditions.
rate on both the pure and Mo C-containing ZSM-5(80). This
2
tendency was compensated for by the increased selectivity of
dimethyl ether, which was hardly detectable at low space veloc-
ity. Whereas on pure ZSM-5(80), this change was not extensive,
on 5% Mo C/ZSM-5, all of the CH products disappeared at
2
x
higher space velocities, and only dimethyl ether was formed.
The results obtained for pure and 5% Mo C/ZSM-5(80) are dis-
2
played in Fig. 6B.
3
.3.3. TPR measurements for used catalyst
The amount and the reactivity of coke deposited on the cat-
alysts during the reaction was determined based on its TPR
with hydrogen. The samples were treated with the reacting gas
mixture for 12 h at 773 K, then cooled to room temperature

374
R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
Fig. 6. Effects of space velocity on the product selectivity over ZSM-5(80) (A) and 5% Mo C/ZSM-5(80) (B) at 773 K.
2
Fig. 7. Formation of hydrocarbons in the TPR measurements following methanol reaction on 5% Mo C/ZSM-5(80) at 773 K for 12 h.
2
4
. Discussion
frequency range indicates that surface OH groups are consumed
in the reaction with methanol. The absorption bands detected
(Fig. 1 and Table 1) suggest that most of the adsorbed methanol
dissociates on ZSM-5(80) at 300 K,
4
.1. Adsorption and dissociation of methanol
IR spectroscopic studies showed that methanol strongly in-
teracts with pure ZSM-5(80). The negative feature in the OH
CH OH = CH O + H(a)
(1)
3
(a)
3
(a)

R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
375
Fig. 8. Rate of formation of C7–C
function of reaction temperature. Data were obtained during gradual heating. At every temperature the sample was kept for 75 min.
aromatics from methanol and methanol + benzene mixture (1:1) on pure ZSM-5(80) and on 5% Mo C/ZSM-5(80) as a
9+
2
and
The appearance of formaldehyde (Tp = 470 K) in the desorb-
ing products indicates the occurrence of the dehydrogenation
of methoxy species,
H(a) + OH(a) = H2O(g),
(2)
to give methoxy species. This conclusion is supported by the
positions of the ν (CH ) and ν (CH ) vibration at 2958 and
CH3O(a) = CH2O(g) + H(a).
We expected to find CO as a result of the decomposition of
(5)
as
3
s
3
−1
2
857 cm , respectively. Treatment of silica discs with CH3OH
in an autoclave at 300–523 K for 20 h produced methylated sil-
formaldehyde,
−
1
icas, which gave IR bands at 2958 and 2858 cm assigned to
Si–OCH3 compound [41]. Heating the adsorbed layer caused
no significant changes in the IR spectra. Adding Mo2C to
ZSM-5 had little effect on the spectral features determined for
pure ZSM-5 at 300 K (Fig. 1); the species adsorbed on ZSM-5
determine the characteristics of the spectra. After annealing the
adsorbed layer at higher temperature, a weak absorption band
CH2O(a) = CO(g) + H2(g),
but there was no sign of CO desorption from pure ZSM-5. We
discuss the possible reaction pathway of the formation of ethyl-
ene in the following section.
(6)
The characteristics of TPD on 10% Mo2C-containing ZSM-
5
remained qualitatively the same. A new feature is the ap-
pearance of the high-temperature peaks for H2, CO, and CH4
Fig. 2B), which suggests the presence of a strongly bonded
CxHy species on the Mo2C.
−
1
developed at 2971 and 2925 cm . Taking into account the
characteristic spectral features of various compounds possibly
formed in the reaction of adsorbed methanol (Table 1), the band
(
−
1
at 2971 cm can be attributed to the vibration of dimethyl
−
1
ether, and the band at 2925 cm can be ascribed to the vi-
bration of ethylene. The formation of dimethyl ether suggests
the coupling of two adsorbed methanols,
4
.2. Aromatization of methanol
As in the case of the reaction of C1–C5 hydrocarbons, Mo2C
2
CH3OH(a) = CH3–O–CH3(a) + H2O(g).
(3)
alone cannot convert methanol into aromatics; it only catalyzes
its decomposition to CO, H2, and CH4. The situation remained
the same when Mo2C was deposited on high-surface area SiO2.
A completely different picture was found on ZSM-5-based
catalysts, however. ZSM-5 alone catalyzes the conversion of
methanol to olefins, and to aromatics to a lesser extent [7–16,
Careful analysis of the IR spectra of annealed layers to iden-
tify other species did not yield positive results.
TPD measurements revealed that a fraction of the adsorbed
methanol on ZSM-5 desorbed and another fraction decomposed
to water, hydrogen, methane, and ethylene. Very small amounts
of formaldehyde and dimethyl ether also were released. We as-
sume that the desorption of methanol is mostly the result of the
associative reaction of adsorbed methoxy and hydrogen,
4
8–50]. Its catalytic performance depends on the Si/Al ratio
or, in other words, on the number of Brönsted sites, which de-
creases with increasing silica content. This is reflected in the
production of aromatics; the highest value was found on ZSM-
5(30), and the lowest was found on ZSM-5(280) (Table 2).
CH3O(a) + H(a) = CH3OH(g).
(4)

376
R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
Adding Mo2C to the ZSM-5 samples significantly enhanced
the formation of aromatics on all zeolite samples (Table 2). The
best results were obtained at 5% Mo2C content; any further in-
crease in Mo2C loading led to a decrease in the total selectivity
of aromatics. This feature can be attributed to the lowering of
acidic centers by Mo2C. An exception is ZSM-5 (280), which
contains fewer Brönsted sites, in which the catalytic effect of
Mo2C plays a more dominant role.
The study of the reaction pathways of CH3 on Mo2C/Mo
(100) surface by several spectroscopic methods has disclosed
that instead of its coupling into ethane, it releases hydrogen,
and the recombination of CH to ethylene is the favored reac-
2
tion [52].
Although a huge amount of IR spectroscopic data support
the cleavage of the O–H bond (i.e., the formation of methoxy
species in the dissociation of methanol on solid surfaces), the
rupture of the C–O bond on metals as another step is also
considered. Based on XPS, SIMS, and TPD results, Winograd
et al. [53] suggested a bimolecular mechanism in which two
neighboring adsorbed methanol molecules produce a methoxy,
a methyl group, and water on the surface. This assumption
sparked an extensive debate. Yates et al. [54] found no indi-
cation of C–O bond scission by isotopic mixing studies. In
harmony with this observation, we found that CH3 produced by
the photodissociation of CH3I on Pd(100) surface is fully dehy-
drogenated to C around 300 K with no spectroscopic evidence
for the existence of CH2 and CH species above this temperature
4
.2.1. Reaction of methanol on pure ZSM-5
Before interpreting the foregoing results, it should be men-
tioned that because of its great technological importance, the
methanol-to-olefins (MTO) process on various zeolites has
been the subject of extensive research [7–16,48–50]. The cen-
tral problem is the formation of C–C bonds, which is also a
key step in the production of aromatics. Chang and Silvestri [8]
proposed that carbene-like intermediate is produced by a con-
certed α-elimination mechanism involving both Brönsted-acid
sites and basic sites.
Another view is that methanol is activated on the acidic sites
[
55,56]. In contrast, Rebholz and Kruse [57], using field ioniza-
+
of the catalysts to form carbenium ion, CH [10], which reacts
3
tion spectroscopy, reported that CH3 fragments formed at 200 K
are stable up to 500 K. Chen et al. claimed [58] that methyl
fragments are stable only up to 400 K; above this temperature,
they dehydrogenate to CH2 and CH species. Recently, Morkel
et al. [59] suggested that there are two competing decomposi-
tion pathways of adsorbed methanol on Pd(111): dehydrogena-
tion to CO and H2 and methanolic C–O bond scission. The
time-dependent evolution of CO/CHxO and of carbonaceous
deposits CHx (x = 0–3) was monitored; however, the existence
of the latter has not been established by vibration spectroscopy.
In the light of this debate, we may count on the cleavage of
the methanolic C–O bond on Mo2C, which involves a bimole-
cular mechanism in which two neighboring adsorbed methanol
molecules produce a methoxy and methyl species, the reaction
of which leads to further products. The active sites for this step
could be the carbon-deficient site on the Mo2C surface, which
may have a high affinity toward oxygen. The reactivity of this
site has been exhibited in the promotion of CO dissociation at
with methanol or dimethyl ether to give a compound containing
a C–C bond,
.
(7)
Several variations of these mechanisms appear in the literature.
At present, most authors accept the reaction pathway, which in-
volves the transient formation of carbenium ion. In terms of the
production of aromatics, the ethylene formed should be acti-
vated, which proceeds on the Brönsted sites of ZSM-5 by the
successive deprotonation of ethylene and hydride transfer to
carbonium ions [4,51].
4
.2.2. Effects of Mo2C
When explaining the promoting effect of Mo2C, we must
take into account the fact that the deposition of Mo2C on ZSM-
markedly reduces the number of the acid sites of the zeolite
24]. In light of this feature, we may consider that Mo2C opens
3
00 K [60] and also in the scission of C–I bond in the adsorbed
5
[
alkyl iodides [61]. The release of strongly bonded O in the form
of CO occurred only above ∼960 K [60], whereas that of ad-
sorbed I occurred above 1050 K [61].
However, the role of the Mo2C is not solely to open a new
route for the activation of methanol to ethylene—it also cat-
alyzes its aromatization. It is very likely that Mo2C provides
dehydrogenation centres for adsorbed ethylene resulting in dif-
ferent products, which are converted to aromatics on the acidic
sites of ZSM-5 [62,63].
a new route for the activation of methanol. Along with O–H
bond scission proceeding on ZSM-5, cleavage of the C–O bond
of adsorbed CH3OH on Mo2C also occurs to yield a CH3 radi-
cal,
CH3OH(a) = CH3(a) + OH(a)
and/or
(8)
CH3O(a) = CH3(a) + O(a),
(9)
To further refine the picture, we compared the rate of aro-
matics formation from methanol and ethylene on Mo2C/ZSM-5
which further decomposes to CH2, the recombination of which
leads to the formation of ethylene,
(
80) under exactly the same experimental conditions. Data pre-
CH3(a) = CH2(a) + H(a),
(10)
(11)
sented in Fig. 9 show that the aromatization process begins
at lower temperatures and occurs much more rapidly from
methanol than from ethylene. This suggests that the allylic and
other CxHy species generated in the activation of methanol have
a higher tendency and reactivity for transformation into aro-
matics compared with the “stable” ethylene molecule. Another
2
CH2(a) = C2H4(a).
Another fraction of CH3 species is hydrogenated into
methane,
CH3(a) + H(a) = CH4(g).
(12)

R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
377
Fig. 9. Comparison of the reaction of methanol (A) and ethylene (B) on 5% Mo C/ZSM-5(80) at different temperatures. Data were obtained during gradual heating.
2
At every temperature the sample was kept for 75 min. The selectivity scale refers only for total aromatics.
important feature is that greater amounts of C8–C9+ aromat-
ics are formed in the reaction of methanol compared with the
reaction of ethylene. In the latter case, C9+ aromatics were pro-
duced only in traces. A possible reason for the different product
distribution is that using methanol as a starting reagent also re-
sults in the methylation of benzene formed. This assumption
has been proven by the study of the reaction between methanol
and benzene. As shown in Fig. 8, the methylation of benzene
with methanol occurs on pure ZSM-5; however, adding Mo2C
to the zeolite greatly promotes this process, indicating the high
reactivity of hydrocarbon fragments formed in the activation of
methanol.
of methanol and for the further reactions of compounds
formed.
(iv) A study of the co-reaction between benzene and methanol
revealed that the methylation of benzene by methanol on
ZSM-5 is greatly promoted by Mo C, suggesting that we
2
can count on the occurrence of this process even in the
aromatization of methanol alone.
Acknowledgments
The work was supported by the Hungarian National Office
of Research and Technology (NKTH) and the Agency for Re-
search Fund Management and Research Exploitation (KPI) un-
der contract RET-07/2005.
5
. Conclusions
(
i) FTIR spectroscopic measurements showed that methanol
adsorbs dissociatively on ZSM-5(80) at 300 K, yielding
adsorbed methoxy species. This process is apparently in-
fluenced little by the presence of Mo2C. After annealing
the adsorbed layer on Mo2C/ZSM-5, the transitory forma-
tion of dimethyl ether is also detected.
References
[1] D. Seddon, Catal. Today 6 (1990) 351.
[2] Y. Ono, Catal. Rev. Sci. Eng. 34 (1992) 179.
[3] M. Guisnet, N.G. Gnep, F. Alario, Appl. Catal. A Gen. 89 (1992) 1.
[4] P. Meriaudeau, C. Naccache, Catal. Rev. Sci. Eng. 39 (1997) 5, and refer-
ences therein.
(
ii) Unsupported Mo2C is an effective catalyst for the decom-
position of methanol above 573 K, yielding hydrogen, CO,
CO2, methane, and dimethyl ether. Thus is also true for
silica-supported Mo2C.
[5] C.D. Chang, in: G. Ertl, H. Knoezinger, J. Weitkamp (Eds.), Handbook of
Heterogeneous Catalysis, VCH, Weinheim, 1997, pp. 1894–1908.
[6] G.A. Olah, Á. Molnár, Hydrocarbon Chemistry, Wiley–Interscience, New
York, 2003.
[7] C.D. Chang, J.C.W. Kou, W.H. Lang, S.M. Jackob, J.J. Wise, K.J. Sil-
vestri, Ind. Eng. Chem. Prod. Res. Dev. 171 (1978) 255.
(
iii) ZSM-5 zeolite also an active catalyst for the decomposi-
tion of methanol. Depending on its composition, it cat-
alyzes the formation of aromatics at high temperatures.
The deposition of Mo2C on ZSM-5 markedly enhances
the formation of aromatics (benzene, toluene, xylene, and
C9+), very likely by opening new routes for the activation
[8] C.D. Chang, A.J. Silvestri, J. Catal. 47 (1977) 269.
9] C.D. Chang, W.H. Lang, R.L. Smith, J. Catal. 56 (1979) 169.
10] Y. Ono, T. Mori, J. Chem. Soc. Faraday Trans. 1 77 (1981) 2209.
11] Y. Ono, E. Imai, T. Mori, Z. Phys. Chem. 115 (1979) 169.
12] Y. Ono, H. Adachi, Y. Sendoda, J. Chem. Soc. Faraday Trans. 1 84 (1988)
1091.
[
[
[
[

378
R. Barthos et al. / Journal of Catalysis 247 (2007) 368–378
[
[
[
[
[
[
[
13] C.D. Chang, Catal. Rev. Sci. Eng. 25 (1983) 1.
[40] C. Morterra, M.J.D. Low, J. Phys. Chem. 73 (1969) 321.
[41] N. Takezawa, H. Kobayashi, J. Catal. 25 (1972) 179.
[42] A.B. Morrow, J. Chem. Soc. Faraday Trans. 1 70 (1974) 1527.
[43] L. Kubelkova, P. Schürer, P. Jiru, Surf. Sci. 18 (1969) 245.
[44] J. Raskó, J. Bontovics, F. Solymosi, J. Catal. 146 (1994) 22.
[45] T.R. Forester, S.T. Wong, R.F. Howe, J. Chem. Soc. Chem. Commun.
(1986) 1611.
[46] T.R. Forester, R.F. Howe, J. Am. Chem. Soc. 109 (1987) 5076.
[47] H.H. Hwu, J.G. Chen, Surf. Sci. 536 (2003) 75.
[48] M. Stöcker, Microporous Mesoporous Mater. 29 (1999) 3.
[49] J.F. Haw, W. Song, D.M. Marcus, J.B. Nicholas, Acc. Chem. Res. 36
(2003) 317.
[50] W. Wang, A. Bucholz, M. Seiler, M. Hunger, J. Am. Chem. Soc. 125
(2003) 15260, and references therein.
[51] M.I. Poustma, in: J.A. Rabo (Ed.), Zeolite Chemistry and Catalysis, in:
ACS Monograph, vol. 171, Am. Chem. Soc., Washington, DC, 1976,
p. 437.
[52] F. Solymosi, L. Bugyi, A. Oszkó, Catal. Lett. 57 (1999) 103.
[53] R.J. Levis, J. Zhicheng, N. Winograd, J. Am. Chem. Soc. 111 (1989)
4605.
14] M.O. Adebajo, R.F. Howe, M.A. Long, Catal. Today 63 (2000) 471.
15] M.O. Adebajo, M.A. Long, Catal. Commun. 4 (2003) 71.
16] H.A. Zaidi, K.K. Pant, Catal. Today 96 (2004) 155.
17] F. Solymosi, A. Szoke, Appl. Catal. 166 (1998) 225.
18] F. Solymosi, R. Nemeth, L. Ovari, L. Egri, J. Catal. 195 (2000) 316.
19] 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.
[
[
[
[
[
[
[
20] F. Solymosi, R. Nemeth, A. Szechenyi, Catal. Lett. 82 (2002) 213.
21] F. Solymosi, A. Szechenyi, J. Catal. 223 (2004) 221.
22] A. Széchenyi, F. Solymosi, Appl. Catal. Gen. A 306 (2006) 149.
23] F. Solymosi, R. Barthos, Catal. Lett. 101 (2005) 235.
24] R. Barthos, F. Solymosi, J. Catal. 235 (2005) 60.
25] A. Széchenyi, R. Barthos, F. Solymosi, Catal. Lett. 110 (2006) 85.
26] L. Wang, L. Tao, M. Xie, G. Xu, J. Huang, Y. Xu, Catal. Lett. 21 (1993)
35.
[27] F. Solymosi, A. Erdohelyi, A. Szoke, Catal. Lett. 32 (1995) 43.
[28] F. Solymosi, A. Szoke, J. Cserenyi, Catal. Lett. 39 (1996) 157.
[29] F. Solymosi, J. Cserenyi, A. Szoke, T. Bansagi, A. Oszkó, J. Catal. 165
(1997) 150.
[
[
30] D.W. Wang, J.H. Lunsford, M.P. Rosynek, J. Catal. 169 (1997) 347.
31] D. Ma, Y.Y. Shu, M.J. Cheng, X.D. Xu, X.H. Bao, J. Catal. 194 (2000)
[54] X. Guo, L. Hanley, J.T. Yates Jr., J. Am. Chem. Soc. 111 (1989) 3155.
[55] F. Solymosi, K. Révész, J. Am. Chem. Soc. 113 (1991) 9145.
[56] F. Solymosi, A. Berkó, Z. Tóth, Surf. Sci. 285 (1993) 197.
[57] M. Rebholz, N. Kruse, J. Chem. Phys. 95 (1991) 7745.
[58] J.-J. Chen, Z.-C. Jiang, Y. Zhou, B.R. Chakraborty, N. Winograd, Surf.
Sci. 328 (1995) 248.
105, and references therein.
[
[
32] L.S. Liu, L. Wang, R. Ohnishi, M.J. Ichikawa, J. Catal. 181 (1999) 175.
33] A.J. Brungs, A.P.E. York, J.B. Claridge, C. Marquez-Alvarez, M.L.H.
Green, Catal. Lett. 70 (2000) 117.
[
34] L. Leclercq, M. Provost, H. Pastor, J. Grimblot, A.M. Hardy, L. Gengem-
[59] M. Morkel, V.V. Kaichev, G. Rupprechter, H.-J. Freund, I.P. Prosvirin,
V.I. Bukhtiyarov, J. Phys. Chem. B 108 (2004) 12955.
[60] F. Solymosi, L. Bugyi, A. Oszkó, Catal. Lett. 57 (1999) 103.
[61] A.P. Farkas, Á. Koós, L. Bugyi, F. Solymosi, Surf. Sci. 600 (2006) 2355.
[62] F. Solymosi, A. Sz o˝ ke, Stud. Surf. Sci. Catal. 119 (1998) 355.
[63] F. Solymosi, in: E.G. Derouane, et al. (Eds.), Sustinable Strategies for the
Upgrading of Natural Gas: Fundamentals, Challenges and Opportunities,
Springer, New York, 2005, p. 25.
bre, G. Leclercq, J. Catal. 117 (1989) 371.
[
[
[
[
[
35] Á. Koós, A. Oszkó, F. Solymosi, Appl. Surf. Sci. 253 (2007) 3022.
36] L. Ovari, F. Solymosi, J. Mol. Catal. A Chem. 207 (2004) 35.
37] A.N. McDonald, J. Phys. Chem. 62 (1958) 175.
38] L.D. Beliakova, A.V. Kiselev, Dokl. Akad. Nauk SSSR 119 (1958) 2.
39] E. Borello, A. Zecchina, C. Morterra, J. Phys. Chem. 71 (1967) 2938;
E. Borello, A. Zecchina, C. Morterra, J. Phys. Chem. 71 (1969) 321.