Aromatization of Ethylene – Main Intermediate for MDA?

Article abstract of DOI:10.1002/cctc.201901655

Methane dehydroaromatization (MDA) over Mo/HZSM-5 has been hypothesized in literature to proceed via a two-step mechanism: methane is first converted to ethylene on the molybdenum (Mo) functionality and then ethylene is oligomerized, cyclized and dehydrogenated on the Br?nsted acid sites (BAS) of the HZSM-5 support. This hypothesis is tested by studying the conversion of ethylene at the same conditions as used for MDA, namely 700 °C, atmospheric pressure, and by co-feeding experiments with H₂ and CH₄. Our results suggest that ethylene is not the main intermediate for MDA, because the aromatic selectivities obtained from methane conversion are higher than selectivities measured during ethylene conversion. Furthermore, carbonaceous deposits formed during MDA have a lower density, are more hydrogenated and more active than the ones formed during ethylene aromatization (EDA). Similarly as for MDA, an activation period in which Mo carburizes to its active phase and an induction period, in which aromatics formation rates increase to their maximum are observed for ethylene conversion. The induction period, which was explained by the buildup of a hydrocarbon pool (HCP) is much faster with methane than with ethylene. This period, is attributed to a slow buildup of hydrocarbons, strongly adsorbed on Mo sites, because it is only observed with catalysts containing Mo. Hydrogen co-feeding with ethylene leads to the formation of more reactive coke species and a significantly prolonged lifetime of the catalyst, but not to a faster buildup of the HCP.

Full text of DOI:10.1002/cctc.201901655

Accepted Article
Title: Aromatization of ethylene – main intermediate for MDA?
Authors: Ina Vollmer, Edy Abou-Hamad, Jorge Gascon, and Freek
Kapteijn
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201901655
Link to VoR: http://dx.doi.org/10.1002/cctc.201901655
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
Aromatization of ethylene – main intermediate for MDA?
Ina Vollmer[a],*, Edy Abou-Hamad , Jorge Gascon[c],[a], Freek Kapteijn[a]
[b]
Abstract: Methane dehydroaromatization (MDA) over Mo/HZSM-5
has been hypothesized in literature to proceed via a two-step
mechanism: methane is first converted to ethylene on the
molybdenum (Mo) functionality and then ethylene is oligomerized,
cyclized and dehydrogenated on the Brønsted acid sites (BAS) of
the HZSM-5 support. This hypothesis is tested by studying the
conversion of ethylene at the same conditions as used for MDA,
namely 700 ºC, atmospheric pressure, and by co-feeding
Mo/HZSM-5 catalyst,
a
two-step mechanism has been
proposed. Methane first converts to C2 hydrocarbons on the Mo
site of the catalyst, which then get further dehydrogenated and
aromatized on the Brønsted acid sites (BASs) of the zeolite
support. Both ethylene and acetylene have been proposed as
likely intermediates for this mechanism.[3] Mériaudeau et al.
investigated whether ethylene or acetylene is a more likely
intermediate, comparing their reactivity at MDA conditions of
700 ºC and atmospheric pressure with methane conversion.[
They found that acetylene led to higher benzene formation rates,
likely due to its higher reactivity.
4]
2 4
experiments with H and CH . Our results suggest that ethylene is
not the main intermediate for MDA, because the aromatic
selectivities obtained from methane conversion are higher than
selectivities measured during ethylene conversion. Furthermore,
carbonaceous deposits formed during MDA have a lower in density,
are more hydrogenated and more active than the ones formed
during ethylene aromatization (EDA). Similarly as for MDA, an
activation period in which Mo carburizes to its active phase and an
induction period, in which aromatics formation rates increase to their
maximum are observed for ethylene conversion. The induction
period, which was explained by the buildup of a hydrocarbon pool
While methane cannot be activated directly on BASs, ethylene is
easily activated on the BASs of the zeolite even at temperatures
[
5]
as low as 200 ºC , but aromatics are formed at higher
[
6]
temperatures. At these high temperatures, however, coking
becomes significant leading to a fast drop-off in activity. This can
be mitigated to some extent by introducing a metal like Ga to the
zeolite. This not only decreases overall acidity because the
metal cation replaces the acidic proton to anchor to the zeolite
framework, it also enhances aromatics selectivity through
(HCP) is much faster with methane than with ethylene. This period,
[
4b,
is attributed to a slow buildup of hydrocarbons, strongly adsorbed on
Mo sites, because it is only observed with catalysts containing Mo.
Hydrogen co-feeding with ethylene leads to the formation of more
reactive coke species and a significantly prolonged lifetime of the
catalyst, but not to a faster buildup of the HCP.
dehydrogenation of paraffins, as observed for Ga, Ag and Mo.
7
]
Another studied phenomenon is the delay in the onset of
aromatics production. For the MDA catalyst Mo/HZSM-5, it is
well known that there is one delay associated with the
carburization of Mo to its active phase called ‘activation period’
and a second delay due to the so-called ‘induction period’.[8] The
induction period has recently been associated with the buildup of
a hydrocarbon pool (HCP) analogous to what is observed for the
methanol to olefins process (MTO).[9] For the HCP, C2
hydrocarbon species have also been speculated to serve as the
intermediate, which react on the BAS to build up the HCP.
Introduction
Aromatics, such as benzene, toluene and xylene (BTX) are in
demand, because they enhance the octane number of fuels and
are used as feedstock for the fine-chemicals industry,
commodity goods, plastics and medicine. Alternative ways to
obtain BTX are desired, because the current source for BTX,
naphtha crackers, are revamped to gas crackers.[1] This is the
case especially in the United States due to the availability of
cheap shale gas.[2] There is also great interest in the direct
conversion of methane to aromatics, because of the high
availability of cheap methane.[2]
The present authors have recently found that the carburization
of Mo is achieved also with a treatment in carbon monoxide,
which creates
a catalyst immediately active but without
containing aromatic deposits that usually form during activation
in methane.[10] This allows for the clear separation of the
activation, the induction period and the formation of deactivating
coke species, which usually overlap in time.
In this study, we study ethylene as a possible intermediate by
comparing the catalytic behavior and selectivities of ethylene
aromatization (EDA) with the characteristics of MDA. Next
to comparing MDA and EDA, in this study the effect of co-
feeding methane and hydrogen and the role of Mo are
investigated. The coke formed during ethylene conversion
with different co-feeds and over different catalysts is
characterized and compared with the coke formed during
MDA.
For the methane dehydroaromatization (MDA) over the
[
a]
Dr. I. Vollmer, Prof. J. Gascon, Prof. F. Kapteijn
Catalysis Engineering, Chemical Engineering
Delft University of Technology
Van der Maasweg 9, 2629HZ Delft, The Netherlands
E-mail: i.vollmer@tudelft.nl
[
[
b]
c]
Dr. E. Abou-Hamad
Core Labs
King Abdullah University of Science and Technology
Thuwal 23955, Saudi Arabia
Prof. J. Gascon
We test whether Mo mainly acts in activating the C-H bond
in methane or if it can also aide aromatization. We also
investigate the proposed mechanism for the hydrocarbon
pool (HCP) formation and the possible role of ethylene as
KAUST Catalysis Center, Advanced Catalytic Materials
King Abdullah University of Science and Technology
Thuwal 23955, Saudi Arabia
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
a precursor for this process occurring during the induction period
of the catalyst. Finally, the activation of Mo/HZSM-5 in ethylene,
methane, in an ethylene/H mixture and in CO are compared to
2
separately study the activation and induction period and to
prevent coke formation during the activation period.
micropores of the zeolite.[11] When Mo is present on the catalyst,
two peaks are observed in the DTGA for both MDA and EDA.
The peak at lower temperature is associated with the Mo, while
the high temperature peak is associated with coke in the
micropores.[11] Both types of coke oxidize at lower temperatures
after MDA than after ethylene feeding over MoHZ as well as
over HZ. In addition, the coking rate is higher during ethylene
aromatization resulting in more coke deposits formed for the
same contact time (Error! Reference source not found. 5), which
1. Comparison of aromatics selectivity
2
also leads to higher pore filling as determined by N adsorption
measurements (Table S3).
The conversion of methane is around ~10 mol% at the applied
conditions and would lead to an ethylene concentration of
~
5 mol%. Therefore, EDA is performed feeding 5 mol% of
ethylene diluted in N to the bare zeolite (HZ) and compared to
feeding 95 mol% methane in to Mo/HZSM-5, MoHZ
Figure 1). It was found that these conditions lead to similar
2
N
2
(
rates of aromatics formation (Figure S1a). The selectivities
observed in these two cases, however, develop differently with
time on stream (TOS) (Figure 1). While benzene selectivity is
similar in the beginning of the reaction, it drops much faster
during ethylene feeding over HZ. In addition, the naphthalene
selectivity is much lower for ethylene feeding. While benzene
selectivity decreases, ethylene conversion stays high
(
Figure S1b) indicating a high coking rate for ethylene feeding
over HZ (Figure S2). This leads to a full deactivation after about
00 min TOS, whereas the MoHZ catalyst under MDA
3
conditions is still active. Coke selectivities, shown in Figure S2,
increase up 50 mol% for EDA, while they stay below 10 mol%
for MDA. Thus, coke amounts formed are much higher during
ethylene conversion on HZ as well, and the coke species formed
are denser (Table S3).
Figure 2. TGA (top) and DTGA (bottom) profiles after reaction, feeding
5 ml/min 95 mol%CH /5 mol%N to MoHZ, 5 mol%C /95 mol%N to HZ
and MoHZ and 5 mol%C /10 mol%H /85 mol%N to MoHZ. TGA was
performed in air with a heating rate of 10 °C/min. Total catalyst weight (mcat
was the same for all experiments, 500 mg. Coke amounts can be calculated
by (1-w/w
1
4
2
2
H
4
2
2
H
4
2
2
)
o
)*mcat.
Although virtually no differences are observed in the 13C NMR
spectra taken on the spent catalysts (Figure S4), the density of
the carbonaceous deposit formed from ethylene is higher than
Figure 1. Comparison of selectivity to aromatics when feeding 15 ml/min
9
5
5 mol%CH
mol%C
4
/5 mol%N
/95 mol%N
2
to
MoHZ
(blue
symbols)
or
feeding
2
from methane, as determined by N adsorption measurements
(Table S3) in line with a lower reactivity. To analyze whether this
2
H
4
2
to HZ (grey symbols).
is due to a lower H/C ratio in the feed for EDA, a co-feeding
experiment with
H
2
is performed with
H
2
in an amount
2
. Characterization of carbonaceous deposits
resembling that when 10 mol% methane is fully converted. The
resulting coke has a lower density and oxidizes at a lower
temperature.
The more reactive nature of carbonaceous deposited during
MDA is demonstrated by thermogravimetric analysis (TGA) in
Figure 2.
For EDA on the bare HZ, only one coke oxidation peak is
observed at high temperature (~650 °C), which is associated
with diffusion-limited combustion of coke deposited in the
During H
identical as during an experiment solely with ethylene feed up to
00 min on stream, but with co-feeding the benzene yield stays
stable for the duration of the experiment and does not drop as in
2
co-feeding, the profiles of aromatic yields look almost
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
EDA (Figure 3). In addition, the final coke amount deposited is
lower, especially the more refractory coke (Figure 2 and
Table S2). The oxidation of the Mo-associated coke formed
during the co-feeding experiment shifts to lower temperatures
and thus becomes more similar to the species observed after
benzene selectivity exceeds thermodynamic predictions. In
contrast, for MDA the product distribution resembles what is
found by thermodynamic considerations.
MDA, while the coke species associated with BAS stay the 4. The function of Mo in ethylene
same as during EDA (Figure 2). Thus, the higher H/C ratio of
methane leads to the formation of less coke which is slightly
aromatization
more hydrogenated and reactive. Also the ratio of the reactive
and refractory coke is higher.
The question arises whether aromatization can be achieved on
Mo as it was observed earlier that the Mo functionality cannot be
separated from the BAS of the bifunctional Mo/HZSM-5 catalyst.
In addition, it was demonstrated that Mo supported on
silicalite-1, which has the same topology as ZSM-5 but lacks the
BAS, was able to produce aromatics during MDA.[12] After
introducing Mo to the zeolite, the acid site density and overall
acidity is lowered as Mo cations replace some of the acidic
protons of the zeolite. This is clear from Py-IR measurements
In conclusion, the lower density and higher reactivity of
carbonaceous deposits formed after MDA over MoHZ is mostly
attributed to the formation of more hydrogenated coke species.
Deactivation is faster with ethylene than with methane due to a
higher selectivity to carbonaceous deposits, indicated as coke.
(
Figure S9), where the vibrations for pyridine adsorbed on BAS
are less intense for MoHZ than for HZ. Further, the conversion
of ethylene over a bare zeolite catalyst with a higher Si/Al ratio
(
HZ-40, Si/Al = 40) than HZ and thus lower acidity is compared
with MoHZ. After introduction of Mo HZ has the same Brønsted
acidity as HZ-40 as is shown by Py-IR (Figure S9).
Nevertheless, aromatic yields observed over MoHZ are much
higher than over HZ-40 (Figure 4). Since coke selectivities are
lower over MoHZ, it is concluded that Mo does not only speed
up the activation of ethylene but also plays an important role in
the aromatization of ethylene likely during the dehydrogenation
step.
a)
Figure 3. Yield to aromatics when feeding 15 ml/min 5 mol%C
2 4 2
H /95 mol%N
or 5 mol%C /10 mol%H /85 mol%N to MoHZ (closed symbols).
2
H
4
2
2
3. Thermodynamic considerations
Comparison of thermodynamic predictions with the experimental
observations are used to determine whether the aromatization of
the two reactions is under thermodynamic or under kinetic
control. At the reaction conditions applied, naphthalene and
benzene (predicted selectivities 54 mol% and 41 mol%) are the
most stable products for MDA, while methane also becomes a
b)
preferred
product
for
ethylene
conversion
(
naphthalene: 56 mol%, methane: 33 mol%, benzene: 11 mol%).
Thermodynamically, much lower benzene selectivity is predicted
during ethylene conversion (Figure S6). However, in the initial
phase of the reaction, benzene selectivity is much higher while
naphthalene and methane selectivities (Figure S6) are much
lower than what is predicted by thermodynamics. At the same
time, experimentally observed selectivities are very close to
thermodynamic predictions for MDA over MoHZ. It has to be
noted that if coke was considered in the thermodynamic
calculations it would be the sole predicted product. Due to the
shape-selectivity of the zeolite and kinetics however the next
most preferred products naphthalene and benzene are formed.
It appears that benzene formation from ethylene is kinetically
favored over formation of the bigger molecule naphthalene as
Figure 4. a) Ethylene conversion and b) yield to benzene and naphthalene
with time on stream when feeding 15 ml/min of 5 mol%C H /95 mol%N to
2 4 2
either MoHZ (black) or HZ-40 (red).
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
5. Activation and induction period
Both the activation and the induction period are much shorter
during MDA. To investigate this further, a co-feeding experiment
was performed where the inert N
2 4
was partially replaced by CH
Aromatics are instantaneously formed over HZ and HZ-40, while
over MoHZ there is a delay in the formation. The delay in the
onset of aromatics production and the slow increase in
aromatics production to its maximum is only observed when Mo
is present on the catalyst. A similar behavior was observed for
methane aromatization[8] and was coined 'activation period' for
the delay and 'induction period' for the increase afterwards.[9a]
(
Error! Reference source not found. 5b). Benzene forms after
less than 14 min (14 min being the time resolution of the GC
measurement) during MDA. CO is produced until benzene is
observed (Figure S11), demonstrating the carburization of Mo
oxide to Mo (oxy-)carbide. During EDA and co-feeding
experiments, CO evolution is observed for almost 28 min, thus
carburization takes at least twice as long with ethylene than with
methane.
For MDA, it was found that Mo carburizes before being active to
convert methane to aromatics. To test if Mo carbide is also the
active phase in EDA, a CO pretreatment was applied to MoHZ
prior to contact with ethylene (Figure 5a). This pretreatment
effectively eliminated the delay in the onset of aromatics
formation and the CO formation, due to Mo oxide reduction,
during this period (Figure S10). The increase of aromatic yields
to their maximum, however, remained. The two phases can thus
be clearly assigned to the activation of Mo through carburization
and the buildup of hydrocarbons inside the pores of the zeolite,
respectively. [9a]
The faster activation period observed can be explained by the
higher carburization activity of methane, leading to about
6
mol% initial methane conversion during this period in the co-
feeding experiment (Figure S12). We discuss possible
explanations for the shortened induction period. One could be
that methane is just much faster to react than ethylene thus
leading to the buildup of a critical level of hydrocarbon filling of
the pores and thus a faster evolution of products from the pores.
This option seems unlikely because of the low methane
conversion (0-2 mol%, Figure S12) observed. An alternative
explanation is that the aromatics formed from methane are more
hydrogenated because of a higher H/C ratio in CH
and thus more reactive than the ones forming from ethylene.
This is however refuted by the fact that when co-feeding H with
4 2 4
than in C H
a)
2
ethylene in a mixture leading to the same H/C ratio as during
MDA, the induction period remains the same as during pure
EDA. Instead it could be that the aromatics formed initially
strongly interact with Mo and are only freed once a certain pore
filling and chain length is achieved. Methane could act to
destabilize the strong interaction of the aromatics with Mo and
effectively free them from the pores.
That these species strongly interact with the catalyst is
demonstrated by a flushing experiment. The catalyst was
flushed with He at the reaction temperature of 700 ºC after
feeding ethylene until the induction period was completed and
the benzene yield fully developed, at 100 min on stream. The
evolution of all products immediately stops once the flow is
switched from ethylene to pure He (Error! Reference source not
found.6 and Figure S13). Only naphthalene continues evolving
for another 100 min. This is expected, because the naphthalene
yield is limited by a sublimation equilibrium in the downstream
flow lines from the reactor to the GC which are kept at 150 ºC.
Higher line tracing temperatures were not used, to avoid
ethylene reacting in the lines.[5] The fact that the benzene yield
immediately achieves the same level as prior to flushing and
does not experience an induction period indicates that the HCP
species are retained on the catalyst.
b)
Figure 5. a) Yield to benzene and naphthalene when feeding 15 ml/min
5
mol%C
pretreatment (closed symbols). For the CO-treatment a 30 ml/min flow of
.5 mol%CO/97.5 mol%He is passed over the catalyst while heating to 780 ºC
2 4 2
H /95 mol%N to MoHZ with (open symbols) and without a CO
2
with 10 ºC/min and keeping it at this temperature for 30 min before cooling to
the reaction temperature of 700 °C. b) Yield to benzene and naphthalene
when feeding 15 ml/min 5 mol%C
2
H
4
/95 mol%N
2
(open symbols) or 15 ml/min
of 5 mol%C /90 mol%CH /5 mol%N
2
H
4
4
2
(closed symbols) to MoHZ.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
Conclusions
The two-step mechanism as proposed earlier for MDA does not
cover the full picture, ethylene likely not being the main
intermediate of the reaction.
The hydrocarbon pool (HCP) forming during MDA is made up by
less dense and more hydrogenated species than the ones
forming from ethylene.
The carbonaceous deposits formed from methane are more
reactive than the ones formed from ethylene. Co-feeding
hydrogen with ethylene leads to reactive and hydrogenated coke
species and prolongs the lifetime of the catalyst.
Mo has to be carburized first before being active for ethylene
aromatization. This carburization can be achieved by CO,
similarly to what has been observed for MDA. Furthermore,
methane is a better carburizing agent than ethylene.
Figure 6. Observed yield to benzene and naphthalene when the reactor is
flushed with a 15 ml/min He flow for 130 min after 100 min on stream feeding
15 ml/min 5 mol%C2H4/95 mol%N2 to MoHZ.
Experimental Section
Since 100 mol% conversion of ethylene is achieved under the
applied conditions it can be assumed that the concentration of
ethylene throughout MDA at any given point is rather low, if
ethylene were to be the main intermediate. Our experiments with
Catalyst synthesis: Catalytic tests are performed on the H-form the
commercial HZSM-5 zeolites with Si/Al = 24 (PQCorporation, CBV
5
(
020E) denoted HZ and on another commercial catalyst with Si/Al = 40
Zeolyst, CBV 8014) denoted HZ-40. Mo is introduced to HZ through IWI
by dissolving appropriate amounts of ammonium heptamolybdate
(NH Mo 24, AHM) in a volume of water needed to fill the pores of the
5
mol% of ethylene in the flow are however representative,
because the two-step mechanism dictates that any product has
to be formed from the intermediate ethylene. And thus the
ethylene flow is chosen so that similar aromatics formation rates
are achieved (Figure S1a).
(
4
)
6
7
O
zeolite powder (400 µl/g). The AHM solution is introduced to the zeolite
dried at 120 °C for 10 h und mixing. The impregnated samples are dried
overnight at 80 °C and calcined in static air at 550 °C for 5 h using a
heating rate of 2°C/min. Samples are prepared with a Mo/Al ratio of 0.6,
which is equivalent to 3.4 wt% and are denoted as MoHZ.
The presented results cast doubt on the proposal of ethylene as
the main intermediate of MDA. Firstly, because benzene and
naphthalene selectivities are quite different when feeding
ethylene or methane at levels encountered in MDA even when
Thermogravimetric analysis (TGA) is performed on a Mettler Toledo
TGA/SDTA851e using 10 to 20 mg of sample under flowing 100 ml/min
of air at a heating rate of 10 °C/min up to 800 °C. The amount of coke
formed during the reaction is calculated from the weight loss between
300 °C and 800 °C of the catalyst after 7 h of reaction.
2
co-feeding H . Secondly, thermodynamic limits are reached
during MDA, while EDA seems to be commandeered by kinetics.
Thirdly, the carbonaceous deposits and the HCP species formed
during EDA are much less reactive than the ones forming during
MDA. This is evident from the longer induction period
encountered with ethylene feeding and the higher oxidation
temperature during TGA. This cannot solely be attributed to a
Catalytic testing is performed in a quartz reactor tube with an inner
diameter of 6 mm, using 500 mg catalyst pelletized and sieved to 212 to
355 µm. A reaction temperature of 700 °C is applied in all tests and the
reactor is brought to reaction temperature under a 15 ml/min flow of He
with a heating rate of 10 ⁰C/min. The flow is switched to the reaction
mixture once the final temperature is reached and the first GC injection is
done after the catalyst is already subjected to the reactant flow for 5 min.
A weight hourly space velocity (WHSV) of 1.21 h-1 (based only on
4 2 2 4
higher H/C ratio in the feed for CH , as co-feeding H /C H does
not shorten the induction period and the TGA profile of coke
deposited during this experiment does not fully resemble the
TGA profile measured on a catalyst after MDA.
methane flow) is used when feeding methane corresponding to 15 ml/min
STP flow of 95 mol%CH
reactant, a 15 ml/min flow of 5 mol%C
would be the flow of C corresponding to 10 mol% conversion of CH
4
entirely to ethylene. In addition, co-feeding experiments are performed
with either methane or hydrogen where 15 ml/min flows of
4
/5 mol%N
2
. For experiments with ethylene as a
The higher coking rate observed during EDA compared to MDA
has implications for an industrial process aiming at aromatics. It
has been proposed to split up MDA in two steps.[13] In a first
2 4
H
/95 mol%N is used, as this
2
2
H
4
reactor methane would be coupled to produce ethylene, which
would then be aromatized in a second reactor. However, as
more and less reactive coke is formed during ethylene feeding, it
is more recommended to form aromatics directly from methane
in one step. When ethylene aromatization with a high selectivity
to BTX is desired, co-feeding methane and or hydrogen should
be beneficial enhancing aromatics selectivity and prolonging the
lifetime of the catalyst.
5
5
mol%C
2
H
4
/90 mol%CH
4
/5 mol%N
2
or
mol%C
2
H
4
/10 mol%H /85 mol%N
2
2
are used. For all experiments with
2
ethylene and/or methane, N serves as the internal standard.
Reactant conversion is calculated according to Equation 1, where
퐴푖,푖푛 (ꢀ = 퐶 퐻 , 퐶퐻 표푟 푁 ), the integrated peak area from the GC analysis
is determined from three initial GC calibration injections, measured while
sending the reaction mixture through the bypass. Product analysis is
achieved on an Interscience Trace GC with one TCD and two FIDs.
2
4
4
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
Product yields are calculated according to Equation 2, where 퐹 [푚표푙/ℎ]
X. Guo, Catal Lett 1996, 40, 207-214; d) F. Solymosi, J. Cserényi, A.
Szöke, T. Bánsági, A. Oszkó, Journal of Catalysis 1997, 165, 150-161.
a) P. Mériaudeau, L. V. Tiep, V. T. T. Ha, C. Naccache, G. Szabo,
Journal of Molecular Catalysis A: Chemical 1999, 144, 469-471; b) P.
Mériaudeau, V. T. T. Ha, L. V. Tiep, Catal Lett 2000, 64, 49-51.
S. Bessell, D. Seddon, Journal of Catalysis 1987, 105, 270-275.
P. Qiu, J. H. Lunsford, M. P. Rosynek, Catal Lett 1998, 52, 37-42.
a) V. R. Choudhary, P. Devadas, S. Banerjee, A. K. Kinage,
Microporous and Mesoporous Materials 2001, 47, 253-267; b) M.-F.
Hsieh, Y. Zhou, H. Thirumalai, L. C. Grabow, J. D. Rimer,
ChemCatChem 2017, n/a-n/a.
푛
denotes the molar flow of molecule ꢁ, 푥 the number of carbons and 푦 the
number of hydrogens in a product molecule, while #퐶ꢂ푒푎푐푡푎푛푡 denotes the
number of carbons in the reactant molecule. The integral selectivity to
each gaseous product ꢀ over the whole time on stream (TOS) is obtained
using Equation 3, where the number of the GC injection is denoted by 푗.
Integral coke selectivity over the whole TOS calculated with Equation 4.
[4]
[5]
[6]
[
7]
ꢃꢄꢅꢆꢇꢈꢆꢉꢈ,ꢊꢉ
ꢃꢄꢅꢆꢇꢈꢆꢉꢈ,ꢍ푢ꢈ
ꢃꢋꢌ,ꢍ푢ꢈ
(
)−(
)
ꢃꢋꢌ,ꢊꢉ
ꢃ
푋ꢂ푒푎푐푡푎푛푡 =
∗ 100%, 푟ꢎꢏꢐꢑꢏꢁꢑ = 퐶2퐻4 표푟 퐶퐻4
ꢄꢅꢆꢇꢈꢆꢉꢈ,ꢊꢉ
ꢃꢋꢌ,ꢊꢉ
[8]
[9]
a) H. Jiang, L. Wang, W. Cui, Y. Xu, Catal Lett 1999, 57, 95-102; b) J.
Lu, Y. Xu, Y. Suzuki, Z.-G. Zhang, Catalysis Communications 2010, 12,
127-131; c) C. Sun, S. Yao, W. Shen, L. Lin, Catal Lett 2008, 122, 84-
Equation 1
ꢒ
ꢔ ꢖ
ꢓ ꢕ ,ꢍ푢ꢈ
푌
푖
[푚표푙%] = ꢒ
∗ 푥 ∗ 100%
,
ꢀ = 푝푟표푑ꢘꢐꢑ, 퐶ꢙ퐻ꢚ
ꢄꢅꢆꢇꢈꢆꢉꢈ,ꢊꢉ∗#ꢗꢄꢅꢆꢇꢈꢆꢉꢈ
90; d) D. Wang, J. Lunsford, M. Rosynek, Top Catal 1996, 3, 289-297.
Equation 2
∑
ꢝ ꢜꢊ
a) N. Kosinov, A. S. G. Wijpkema, E. Uslamin, R. Rohling, F. J. A. G.
Coumans, B. Mezari, A. Parastaev, A. S. Poryvaev, M. V. Fedin, E. A.
Pidko, E. J. M. Hensen, Angewandte Chemie 2018, 57, 1016-1020; b)
E. G. Derouane, J. B. Nagy, P. Dejaifve, J. H. C. van Hooff, B. P.
Spekman, J. C. Védrine, C. Naccache, Journal of Catalysis 1978, 53,
∑푇푂ꢛ 푆푖[푚표푙%] = ∑
∗ 100%, 푗 = 퐺퐶 ꢀꢁ푗ꢎꢐꢑꢀ표ꢁ# Equation 3
ꢝ ꢞꢄꢅꢆꢇꢈꢆꢉꢈ
∑푇푂ꢛ 푆푐ꢟ푘푒[푚표푙%] = ∑ 푋
ꢠ ∑ ∑
푇푂ꢛ
푆푖 Equation 4
ꢡ
ꢂ푒푎푐푡푎푛푡
푖
40-55; c) I. M. Dahl, S. Kolboe, Catal Lett 1993, 20, 329-336; d) I.
Yarulina, A. D. Chowdhury, F. Meirer, B. M. Weckhuysen, J. Gascon,
Nature Catalysis 2018, 1, 398-411; e) K. Hemelsoet, J. Van der
Mynsbrugge, K. De Wispelaere, M. Waroquier, V. Van Speybroeck,
ChemPhysChem 2013, 14, 1526-1545.
Acknowledgements
Financial support from the SABIC-NWO CATC1CHEM CHIPP
project is gratefully acknowledged. Thanks go to Dr. Christoph
Dittrich (SABIC), Dr. Frank Mostert (SABIC) and Dr. T.
Alexander Nijhuis (SABIC) for helpful discussion.
[
[
10] I. Vollmer, B. van der Linden, S. Ould-Chikh, A. Aguilar-Tapia, I.
Yarulina, E. Abou-Hamad, Y. G. Sneider, A. I. Olivos Suarez, J.-L.
Hazemann, F. Kapteijn, J. Gascon, Chemical Science 2018, 9, 4801-
4807.
11] a) D. Ma, D. Wang, L. Su, Y. Shu, Y. Xu, X. Bao, Journal of Catalysis
2002, 208, 260-269; b) H. Zheng, D. Ma, X. Liu, W. Zhang, X. Han, Y.
Xu, X. Bao, Catal Lett 2006, 111, 111-114; c) D. Ma, Y. Lu, L. Su, Z.
Xu, Z. Tian, Y. Xu, L. Lin, X. Bao, The Journal of Physical Chemistry B
Keywords: methane dehydroaromatization • ethylene
aromatization • HZSM-5 • molybdenum • coking
2
002, 106, 8524-8530; d) R. Ohnishi, S. Liu, Q. Dong, L. Wang, M.
[
1]
M. Bender, in New Technologies and Alternative Feedstocks in
Petrochemistry and Refining Vol. 2, DGMK Conference, Dresden,
Germany, 2013.
Ichikawa, Journal of Catalysis 1999, 182, 92-103; e) N. Kosinov, E. A.
Uslamin, F. J. A. G. Coumans, A. S. G. Wijpkema, R. Y. Rohling, E. J.
M. Hensen, ACS Catalysis 2018, 8459-8467; f) A. Martínez, E. Peris,
M. Derewinski, A. Burkat-Dulak, Catalysis Today 2011, 169, 75-84.
12] N. Kosinov, F. J. A. G. Coumans, E. A. Uslamin, A. S. G. Wijpkema, B.
Mezari, E. J. M. Hensen, ACS Catalysis 2016, 520-529.
[
[
2]
3]
N. E. T. Laboratory, (Ed.: U. D. o. Energy), US Department of Energy,
US, 2011.
[
[
a) Y. Xu, S. Liu, X. Guo, L. Wang, M. Xie, Catal Lett 1994, 30, 135-149;
b) L. Y. Chen, L. W. Lin, Z. S. Xu, X. S. Li, T. Zhang, Journal of
Catalysis 1995, 157, 190-200; c) Y. Xu, W. Liu, S.-T. Wong, L. Wang,
13] C. Dittrich, A process for carrying out endothermic, heterogeneously
catalyzed reactions, 2015, WO2015082375A1.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201901655
FULL PAPER
FULL PAPER
Does it all revolve around ethylene?
According to our study this is very
unlikely. In contrast to what had
been hypothesized, the formation of
aromatics from methane does not
proceed solely via the formation of
ethylene in a first step. Likely other
intermediates and aromatics can
directly form on Mo, while
Ina Vollmer*, Edy Abou-Hamad, Jorge
Gascon, Freek Kapteijn
Page No. – Page No.
Aromatization of ethylene – main
intermediate for MDA
intermediate alkenes get aromatized
on Brønsted acid sites in the zeolite
pores of the catalyst.
This article is protected by copyright. All rights reserved.