Conversion of methanol to aromatics in fluidized bed reactor

Article abstract of DOI:10.1016/j.cattod.2014.02.007

Conversion of methanol to aromatics (MTA) was conducted over a fresh or a spent Zn/ZSM-5 catalyst in single stage fluidized bed (SSFB) and two stage fluidized bed (TSFB). Sampling at different stages of TSFB and operation in the temperature range of 250, 275, 300, 325, 350, 380 and 475 °C revealed the consecutive reaction mode from methanol to DME, C₁-C₄ hydrocarbons, C₅₊ nonaromatics to aromatics finally. High weight ratio of para-xylene (PX) in xylene in wide temperature range suggested that PX was the primary product of MTA, and other xylenes are produced by the iosmerization of PX. Other aromatics such as benzene, toluene and trimethylbenzene are finally produced by the dealkylation, alkylation or disproportionation of xylene. The adoption of TSFB reactor was effective to increase the yield of aromatics, compared to that using SSFB, due to the inhibition of backmixing of gases. Combination of TSFB and highly active catalyst with strong acids was effective to get high yield of aromatics under high space velocity of methanol.

Full text of DOI:10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Conversion of methanol to aromatics in fluidized bed reactor
a,b
b
b
a,∗
b
b
Tong Wang , Xiaoping Tang , Xiaofan Huang , Weizhong Qian , Yu Cui , Xinyu Hui ,
a
a
Wei Yang , Fei Wei
Department of Chemical Engineering, Tsinghua University, Beijing 100084, China
a
b
Coal Chemical Industry Division, Huadian Coal Industry Group Co. Ltd., Beijing 100031, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Conversion of methanol to aromatics (MTA) was conducted over a fresh or a spent Zn/ZSM-5 catalyst
in single stage fluidized bed (SSFB) and two stage fluidized bed (TSFB). Sampling at different stages of
TSFB and operation in the temperature range of 250, 275, 300, 325, 350, 380 and 475 C revealed the
Received 14 October 2013
Received in revised form 6 February 2014
Accepted 9 February 2014
Available online xxx
◦
consecutive reaction mode from methanol to DME, C1–C4 hydrocarbons, C5+ nonaromatics to aromatics
finally. High weight ratio of para-xylene (PX) in xylene in wide temperature range suggested that PX was
the primary product of MTA, and other xylenes are produced by the iosmerization of PX. Other aromatics
such as benzene, toluene and trimethylbenzene are finally produced by the dealkylation, alkylation or
disproportionation of xylene. The adoption of TSFB reactor was effective to increase the yield of aromatics,
compared to that using SSFB, due to the inhibition of backmixing of gases. Combination of TSFB and highly
active catalyst with strong acids was effective to get high yield of aromatics under high space velocity of
methanol.
Keywords:
ZSM-5
Methanol to aromatics: Fluidized bed
reactor
Yield
Selectivity
©
2014 Elsevier B.V. All rights reserved.
1
. Introduction
the conversion of gas reactants, but also increased the probability
of the undesirable side reaction of intermediates over the cata-
lyst [8–10]. Actually, over a metal–zeolite catalyst with Lewis and
Brosten acids, the MTA process included many reactions such as the
dehydration of methanol, hydrogen-transfer reaction from olefins
to paraffins, dehydrogenation of olefins or paraffins to aromatics in
presence of metal (such as Zn, Ag and Ga etc.), disproportionation,
isomerization, dealkylation or alkylation of aromatics etc. [1,2,11].
It remained unknown whether the operation in fluidized bed influ-
enced the product profiles of aromatics or not. Even worse, the
product evolution in MTA process over metal–zeolite catalyst, dif-
ferent from those in methanol to olefins (MTO) and methanol to
hydrocarbons (MTH) [11–14], was unclear yet, which increased the
difficulty for the controlled production of aromatics.
Although aromatics, very important basic chemicals, were dom-
inantly produced from oil-based route to this date, the gradual
depletion of oil reserve resulted in the sustainably tight supply and
high cost of aromatics. In the past two decades, the route of convert-
ing methanol into aromatics (MTA) received intensive attention
considering methanol was easily available from wide resources
such coal, natural gas, shale gas and biomasses [1,2]. Technically,
the single pass conversion of methanol, over Zn or Cu or Ga sup-
ported on ZSM-5 catalyst at 470–500 C, gave the aromatics yield
of 50–70% (carbon base), the yield (25–40%) of gaseous hydrocar-
bon (C –C5) and 3–4% yield for coke [1–7]. To this date, most of
these works were carried out in lab-scale fixed bed reactor. Rare
report was found on using the fluidized bed reactor, which, how-
ever, may be much suitable for this process compared to the fixed
bed, since the high temperature operation always result in the rapid
deactivation of the catalyst, which called for the necessary con-
tinuous reaction and catalyst regeneration [4]. On the other hand,
backmixing of gases was widely observed in the single stage flu-
idized bed (SSFB), which not only decreased the driving force of
◦
1
In the present work, the conversion of methanol was studied not
only in SSFB, but also in two stage fluidized bed (TSFB). Actually,
sampling product in different stages of TSFB and the operation in
TSFB varying reaction temperature provided sufficient information
for the evolution of C –C5 hydrocarbon and aromatics products.
1
It allowed us to decouple this process, since most reactions above
were not only acidic sensitive, but also temperature sensitive or
even catalyst structure sensitive [1,11–18]. As result, we found the
product evolution route of MTA is different from that of MTO and
MTH [11–14]. The presence of metal over zeolite, independent of
the acid amount of the catalyst, was crucial to convert liquefied
petroleum gas (LPG) and C5+ alkanes and alkenes into aromatics
∗ Corresponding author. Tel.: +86 10 62794133; fax: +86 10 62772051.
E-mail addresses: qianwz@tsinghua.edu.cn, qianwz@mail.tsinghua.edu.cn
(W. Qian).
http://dx.doi.org/10.1016/j.cattod.2014.02.007
920-5861/© 2014 Elsevier B.V. All rights reserved.
0
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
T. Wang et al. / Catalysis Today xxx (2014) xxx–xxx
2
with high yield. The MTA process mainly followed a consecutive
reaction route from methanol to dimethyl ether (DME), C –C5
1
hydrocarbons and finally to aromatics. Moreover, para-xylene (PX)
was confirmed as the primary product of aromatics, while other
isomers of xylene (X) were produced by the isomerization of PX
subsequently, and B, T and trimethylbenzene (TriMB) was produced
by the dealkylation and alkylation of X finally. These knowledge
help us to understand the different MTA products in SSFB and TSFB.
High yield of aromatics in TSFB, compared to the SSFB, was due
to the inhibition of backmixing of gaseous reactants and interme-
diates [8–10]. These results were not only useful for the further
understanding of the catalytic issue on the product profiles, but
also provided engineering solutions for the scale up.
2
. Experimental
2.1. Catalyst preparation
The model catalyst used here was a Zn/ZSM-5 based catalyst
with the loading of Zn of 3%. Zn was firstly doped on the ZSM-5 with
Si/Al molar ratio of 38 by soaking the microsized ZSM-5 powder in
−1
◦
Fig. 1. Illustration of single stage fluidized bed (SSFB) and two stage fluidized bed
TSFB).
the solution of 0.5 mol L Zn(NO3)2 at 100 C for 24 h. Then the
powder was filtrated and washed by de-ion water for 5 times, then
(
◦
◦
was dried firstly at 105 C for 48 h and calcinated at 650 C for 6 h.
◦
heated to 160 C by the hot wires outside to avoid the condensation
of water, hydrocarbons with carbon number larger than 4. Yield of
aromatics was calculated by the carbon base, without considering
the amount of water. The analysis using GC gave a resolution of
The powder was grinded to an average size of 100 ␮m. In addition
◦
to the fresh catalyst, the catalyst was also treated in 700 C in 100%
steam for 6 h. We called it here as spent catalyst with weak acids
[
4] used for the comparison study.
1
0 ppm for the component of product.
Acidity of the catalyst was analyzed by NH –TPD in a Quanta-
3
Since the catalyst had relatively long life time and the activ-
chrome automated chemisorption analyzer from room tempera-
◦
◦
−1
ity of the catalyst can be well regenerated by burning coke, the
experimental data used in the present work remained very good
repeatability.
ture to 850 C with a ramp of 10 C min . Brunauer–Emmett–Teller
BET) surface area of catalyst are recorded in a Quantachrome auto-
(
mated surface area and porosity analyzer with N as the adsorption
2
gas.
3
. Result and discussions
Effect of Zn was to stabilize the acids of ZSM-5 and to increase
the dehydrogenation ability of the catalyst [1,2]. The Zn loading of
the fresh catalyst and spent catalyst was measured by the titra-
3.1. Characterization of the fresh and spent catalysts
−
1
tion method. Firstly, catalyst was solved by using 0.5 mol L HCl at
◦
BET surface area of fresh and spent catalyst was 136.4 and
5
0 C for 30 h. Secondly, the soluble Zn was titrated by using EDTA.
2
−1
1
37.3 m g , respectively, indicating its stable pore structure. The
Indictor was Xylenol Orange disodium salt and the masking agent
was ammonium fluoride, which was used to decrease the influ-
ence of Al in the solution. End point of titration was set when the
color of solution changed from pink to light yellow. The Zn loading
in spent catalyst is 98–99% of that in fresh catalyst, indicating the
nearly unchanged Zn loading after the reaction.
doping of Zn on HZSM-5 mainly resulted in the formation of Lewis
acids in large amount, which was favorable to increase the dehydro-
genation ability of the catalyst [1,2]. In addition, the modification
of zeolite by Zn also stabilized the middle strong acids [1]. Fig. 2
showed the NH –TPD result of fresh and spent catalyst. The treat-
ment by steam under high temperature resulted in the significant
3
+
3
◦
decrease in the amount of weak acidity in 200–350 C and strong
2.2. MTA experiment in fluidized bed reactor
For MTA reaction in SSFB (Fig. 1 left), 600 g catalyst was directly
added into the reactor. For the operation in TSFB, 400 g catalyst
was added in the lower stage of TSFB (Fig. 1 right), which was
close to the bottom distributor, 200 g catalyst was added in the
upper stage of TSFB. A perforate plate with pore size of 1 mm was
placed between the lower and upper stage in order to avoid the
mixing of catalysts in different stages. Thus the information of
gases or liquid product and the catalysts in different stages can
Fresh
Spent
o
735
C
o
8
00 C
◦
be obtained separately. Temperature range was 250–475 C, pres-
sure was 0.4–0.5 MPa (absolute pressure) and gross space velocity
−1
−1
of methanol was 0.2–0.6 g g cata h .
The product flowing out of the reactor was directly sent to the
GC (Shimadzu, GC2014, Japan, equipped with two flame ionization
detectors (FID)). One FID was used to detect hydrogen, CO, CO and
200
400
600
800
2
gaseous hydrocarbons with carbon number of 1–6. Another was
used to analyze the methanol, dimethyl ether (DME), hydrocarbons
with carbon number of 5–14. The pipe between reactor and GC was
o
Temperature ( C)
Fig. 2. NH3–TPD of fresh and spent catalyst of Zn/ZSM-5.
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
T. Wang et al. / Catalysis Today xxx (2014) xxx–xxx
3
3
3
2
2
1
1
5
0
5
0
5
0
5
0
In this case, when the operation temperature between two stages
remained the same, it did not change such probability. Second, the
decreased yield of X was due to the transformation of X to other
aromatics since it was the primary product of MTA. Actually, In
this case, the disproportionation of X contributed to both the yield
increase of T and TriMB. Thus, the yield increase of B and excess
T should be mainly due to the dealkylation of X. Quantitatively,
the consumption of 1.96% X gave 0.98% TriMB and 0.98% T via the
disproportionation route. And the increase of 1.65% B and 5.59% T
should decrease the yield of X by 7.24% and increase the yield of
methane by 8.89% simultaneously. However, the practical yield of
methane was just increased by 0.58% practically, far lower than the
calculated 8.89%. And the yield of X decreased by only 5.6%, not
9.2% totally above. These mismatch in the yield of X and methane
suggested the complex transformation nature between different
hydrocarbons. The dealkylation of X was not directly associated to
the production of methane. In addition, the disproportionation of
X is difficult to conduct than the dealkylation of X and gave more B
and T than TriMB.
Detailed profiles of C1–C5 hydrocarbon (Fig. 3) suggested the
gross yield of ethane and propane was slightly increased, but the
yield of C2–C4 olefins and that of butane and C5+ nonaromat-
ics were significantly decreased in the further transformation in
upper stage of TSFB. Apparent, the increased yield of aromatics
was from the transformation of these olefins and active paraffins.
And the increased yield of ethane and propane was probably due
to the hydrogen-transfer reaction from ethylene and propylene,
respectively. Totally, the transformation of olefins and paraffins
contributed to the formation of aromatics in larger percentage and
lower stage
upper stage
C1
C2 C2= C3 C3= C4 C4= C5+
B
T
X
C9+
Product profile
Fig. 3. Product profiles sampled at different stages of TSFB (fresh catalyst, 0.3 h−1,
.4 MPa).
0
◦
acidity in 735 C. However, the ratio of very strong acidity centered
at 800 C was increased to some degree. It suggested that the seri-
ous dealumination occurred in the steam treatment, in agreement
with the previous report [4].
◦
3
.2. Evolution of MTA product in different stages of TSFB
Products were sampled at exit of different stages of TSFB (Fig. 3).
Basically, the products in the exit of lower stage represented the
immediate product, but that in the exit of upper stage represented
the final product. The yield of C –C5 hydrocarbon and aromatics
1
the inert C –C2 paraffins in smaller percentage. These results sug-
over the fresh catalyst was 46% and 54%, respectively, in the exit of
lower stage of TSFB (Fig. 4). Further conversion of these products
in the upper stage of TSFB decreased the yield of gaseous hydrocar-
bon to 41.7% and increased the yield of aromatics to 58.3%. Since
the conversion of methanol in exit of lower stage of TSFB was com-
pleted, the increased yield of aromatics was directly attributed to
the further conversion of C –C5 hydrocarbons.
In detail, the yield of B, T and TriMB was increased by 1.65%,
.57% and 0.98%, respectively, while the yield of X was decreased
1
gested the excellent performance of this catalyst, since methane
with the lowest cost was high undesirable in this process.
3.3. Effect of temperature on the product profile of MTA
In order to further study the product evolution in MTA pro-
1
cess, we used the spent catalyst with weak acids and carried out
the experiment in relatively low temperature (Fig. 4), since these
effects were both effective to suppress the dealkylation or dispro-
6
by 5.6%. When methanol was totally converted, the increased yield
of TriMB was not due to the alkylation of X, but due to the dispro-
portionation of X. Weight ratio of X in aromatics decreased from
◦
portionation of X. Actually, the conversion of methanol in 250 C
did not produce aromatics and C5+ nonaromatics at all, but to pro-
duce DME via dehydration. Off gas (CH , C H and C H ) and LPG
6
2.02% in the exit of lower stage to 49.22% in the exit of upper
4
2
6
2
4
(
C3–C hydrocarbons) was the main product. Apparent, the fact is
stage. Two factors may contribute to the decreasing trend of the
yield of X with the reaction time. First, the further conversion of
LPG and C5+ alkanes and alkenes (which was called as nonaromatics
below) contributed to more B and T, but less X probably. However,
X was the most thermodynamically favorable than T and B [19].
4
not in consistent with the hydrocarbon-pool mechanism [11,12].
The aromatic-intermediates, if existed, will firstly release from the
pores of ZSM-5 zeolite, which size is comparable to the molecular
size of B, T and PX, and larger than those of ethylene or propyl-
◦
ene. In 275–325 C, DME and methanol were quickly converted into
◦
off gas, LPG and aromatics [13,14]. Upon to 380 C or above, the
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
conversion of DME to aromatics was complete and it is difficult to
detect DME in product at 475 C. But it is unable to exclude its exist-
aromatics
nonaromatics
offgas
◦
ence as an adsorbed state over the catalyst yet. In total temperature
LPG
◦
range of 250–475 C, the yield of aromatics sustainably increased
DME
methanol
with the temperature, while the yield of off gas was insensitive to
◦
the increase of temperature. Finally, the high temperature (475 C)
resulted in the drastic conversion of nanoaromatics (C5+) to aromat-
ics. In comparison, increasing the pressure from 0.4 MPa to 0.5 MPa
◦
at 350 C did not increase the aromatics significantly (Fig. 7), but
drastically increased the conversion of LPG to C5+ nonaromatics.
Apparent, aromatization of nanoaromatics calls a much higher tem-
perature to overcome the energy barrier than the aromatization of
methanol or DME.
As to detailed profiles of aromatics, the weight ratio of X in aro-
matics (X/a) is 53.5–55% over the spent catalyst in the exit of upper
stage in different temperatures (Fig. 5), higher than that (49.22%)
as using the fresh catalyst in the same exit. It confirmed that the
250
300
350
400
o
450
500
Temperature ( C)
Fig. 4. Temperature dependent yield of different products from the conversion of
methanol in TSFB (spent catalyst, 0.3 h , 0.4 MPa).
−1
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
4
T. Wang et al. / Catalysis Today xxx (2014) xxx–xxx
8
6
4
2
0
0
0
0
30
C5+
X/a
C4H10
X
25
20
15
10
5
0
C4H8
C3H8
C3H6
C2H6
C2H4
CH4
PX/X
0
2
50
300
350
400
450
500
250
300
350
400
o
450
500
o
Temperature ( C)
Temperature ( C)
Fig. 5. Temperature dependent yield of aromatics, selectivity of PX in X and selec-
tivity of X in TSFB (spent catalyst, 0.3 h , 0.4 MPa)
Fig. 6. Profile of offgas and LPG under different temperatures in TSFB (spent catalyst,
0.3 h , 0.4 MPa).
−1
−1
dealkylation or disproportionation of X into other aromatics was
suppressed by using a catalyst with weak acids. The result further
confirmed that X is the primary product of aromatics, independent
4
4
3
5
0
5
0
.4MPa
0.5 MPa
◦
of temperature in the range of 250–475 C. Moreover, the weight
ratio of PX in X is sustainably higher than 65% in the temperature
◦
30
25
range of 275–325 C. Apparent, the shape selectivity of ZSM-5 cat-
alyst made PX as the primary product with high selectivity in X in
lower temperature, because the diffusion coefficient of PX was 1000
folds that of MX or OX inside the pores of the ZSM-5, which pore
size exhibited a shape selectivity on X product [20]. Note here that,
although the low temperature suppressed the isomerization of PX
to MX and OX significantly, the isomerization of PX still occurred
over the present catalyst with relatively stronger acidity, compared
to that to prepare PX with high selectivity up to 90% [13,14]. Actu-
20
15
10
5
0
ally, the serious isomerization between X at the temperature above
C1 C2 C2= C3 C3= C4 C4=C5+ B
T EB PX MX OX C9+
◦
3
50 C resulted in the establishment of the equilibrium between
Product profile
PX, MX and OX, which gave the lowest selectivity (23–24%) of PX
in X. Apparent, the isomerization of X [17,18] was much tempera-
ture sensitive and much acidic sensitive compared to the alkylation
or dealkylation, or disproportionation of X. Note that, the dispro-
Fig. 7. Pressure effect on the product profiles from the conversion of methanol in
TSFB (spent catalyst, 380 C, 0.3 h ).
◦
−1
◦
portionation of X only became significant above 325 C, while the
dealkylation of X to B and T is always significant in all temperature
ranges. It is probably due to the latter can be done inside pores,
while the former reaction just occurred on the outer surface of the
catalyst. In order to confirm this point, the conversion of ethanol
to aromatics was also carried out over the same catalyst (data not
shown here). Surprisingly, no ethylbenzene in large amount was
produced, suggesting the surface alkylation of B, T with ethyl groups
to ethylbenzene was of very low probability. In contrast, B and T in
large mount were produced by the de-alkylation of X and to give
a similar aromatic distribution to that using methanol as the feed-
stock. Furthermore, feedstock of methanol with B and T was used
in MTA (data not shown here). The presence of B and T retards
the formation of B and T, and increased the yield of X and TriMB.
The result suggested that there existed an equilibrium between the
production of X and their alkylation to produce B and T and further
validated that X was the primary product of MTA.
from firstly olefins to then aromatics. During the evolution of C –C3
2
olefins and aromatics olefins with the temperature, the yield of C5+
◦
nonaromatics approached to the maximal value at 380 C. Thus, we
proposed that the consecutive reaction route from methanol to aro-
matics was comprised of the firstly produced DME, off gas, LPG, C5+
nanoaromatics, and then PX (Fig. 8). Isomerization of PX produced
MX and OX [17,18] subsequently, and B, T and TriMB were finally
produced from the transformation of X. Apparent, wide profiles of
aromatics was due to the transformation of X over a catalyst with
strong acids in high temperature. Note here that, based on pure
ZSM-5 zeolite in MTH process, the formation of aromatics and C5+
nonaromatics followed the competitive route [12–14]. In contrast,
the doping of Zn on ZSM-5 increased the dehydrogenation ability
of the catalyst and was effective to convert C5+ nonaromatics into
aromatics in MTA process. Thus the conversion of methanol to aro-
matics mostly followed a consecutive route, under the assistance
of high temperature.
In addition, the hydrogen-transfer reaction was less signifi-
◦
cant below 300 C and to give high priority propylene, but became
◦
serious above 350 C to give higher yield of propane than propyl-
◦
ene (Fig. 6). Further increasing temperature to 475 C resulted in
the very low yield of propylene. In addition, the increase of tem-
perature resulted in the gradually lower yield of olefins and the
sustainably increasing yield of aromatics [21]. Apparent, C –C
2
3
olefins and aromatics was not the paralleled product proposed
by hydrocarbon-pool mechanism [11,12], but consecutive product
Fig. 8. Consecutive reaction mode of the production of aromatics from methanol.
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
T. Wang et al. / Catalysis Today xxx (2014) xxx–xxx
5
3
3
2
2
1
1
5
0
5
0
5
0
SSFB
TSFB
SSFB
5
4
3
2
1
0
0
0
0
0
0
TSFB
5
0
COCO2H2 C1 C2 C2=C3 C3= C4 C4=C5+ B
Product profile
T
EB PX MX OXC9+
off gas
LPG
C5+
aromatics
Product profile
Fig. 9. Reactor effect on the product profiles from the conversion of methanol (spent
catalyst, 475 C, 0.25 h , 0.4 MPa).
◦
−1
Fig. 11. Reactor effect on the product profiles from the conversion of methanol
(
◦ −1
fresh catalyst, 475 C, 0.6 h , 0.4 MPa).
3
.4. Comparison of MTA in SSFB and TSFB
although the yield of aromatics was increased and propane was sig-
nificantly decreased, the yield of H2 was not increased due to the
simultaneously increased yield of butane and C5+ nonaromatics.
As follows, the product profile at different space velocity of
Based on these results, we are able to understand the hydro-
dynamics of SSFB and TSFB on the product profiles of gaseous
hydrocarbon and aromatics (Fig. 9). Actually, compared to the TSFB,
the backmixing of gases and catalyst in SSFB prolonged the contact
time of gases with catalyst and lowered the driving force to achieve
aromatics. Under the same temperature and pressure, methanol
was totally converted in different reactors, but gave different prod-
−1
methanol (0.25 and 0.4 h ) in TSFB was tested. Compared to the
case of low space velocity, gas velocity under in the reactor was
increased high space velocity and the contact time of catalyst
with gaseous reactants or intermediates was further shortened. As
a result, the self decomposition of methanol to CO/CO2 and the
hydrogen-transfer reaction from propane to propylene and from
butane to butene were both significantly suppressed. Note that,
operation in high space velocity resulted in the yield increase of
C5+ nonaromatics, not aromatics (Fig. 10), further confirming that
the C5+ nonaromatics were intermediate products when LPG com-
ponents were converted into aromatics. However, if using fresh
catalyst, the C5+ nonaromatics can be further converted into aro-
uct profiles. Basically, weight ratio of CO/CO was 1.12% in SSFB, and
2
was further decreased to 0.86% in TSFB, indicating the self decom-
position of methanol was significantly suppressed in TSFB. The yield
of LPG, mainly the propane, was decreased by 5.09% in TSFB, as
compared to that in SSFB. It contributed to the yield increase of
aromatics by 4.43% and C5+ nonaromatics by 1.36% in the former
reactor. Apparent, the driving force of transforming LPG to aromat-
ics was significantly increased in TSFB, mainly due to inhibiting the
gases backmixing between different stages [8–10]. In detail, the
increased yield of 4.43% for aromatics in total contributed to the
−1
matics in TSFB, even under much higher space velocity of 0.6 h
(Fig. 11). Apparent, the combination of TSFB with the catalyst with
strong acidity is effective to achieve high yield of aromatics in high
space velocity.
1
.93% increase for B, 2.78% increase for T. The yield of X remained
unchanged, but the yield of TriMB was even decreased a little bit.
The increased yield of B and T, accompanied with the yield increase
of CH4 or C H , was due to the dealkylation of X in TSFB. Interest-
4. Conclusion
2
6
ingly, the backmixing of gases in SSFB diluted the concentration
of PX in outer surface, and consequently, suppressed the surface
isomerization of PX. As a result, the weight ratio of PX in X in
SSFB was 24.94%, higher than that (23.11%) in TSFB. In addition,
The study of the conversion of methanol in two-stage flu-
idized bed under wide temperatures over fresh and spent catalyst
with different acidities revealed the consecutive reaction mode of
methanol to DME, LPG, C5+ nonaromatics to PX, then to MX and
OX and finally to B, T and TriMB. The yield of aromatics can be
increased in TSFB compared to that in SSFB, due to the inhibition of
the backmixing of gaseous between different stages. High efficient
production of aromatics, e.g. under high space velocity of methanol,
called for the combination of advanced reactor technology (e.g.
TSFB) with highly active catalyst with strong acidities.
3
2
2
1
1
0
5
0
5
0
-
1
0
0
.25h
.4 h
-1
Acknowledgements
The authors thank the support of National 863 Program of China
2012AA051003), NSFC Program of 21376135 and NSFC Key Pro-
(
gram of 51236004.
5
References
0
COCO2H2 C1 C2 C2=C3 C3= C4 C4=C5+ B
T EB PX MX OXC9+
[
1] J.G. Zhang, W.Z. Qian, X.P. Tang, K. Shen, T. Wang, X.F. Huang, F. Wei, Acta Phys.
Chim. Sin. 29 (2013) 1281.
Product profile
[
2] Y. Inoue, K. Nakashiro, Y. Ono, Microporous Mater. 4 (1995) 379.
Fig. 10. Space velocity effect on the product profiles from the conversion of
methanol in TSFB (spent catalyst, 475 C, 0.4 MPa).
[3] J.A. Lopez-Sanchez, M. Conte, P. Landon, W. Zhou, J.K. Bartley, S.H. Taylor, A.F.
Carley, C.J. Kiely, K. Khalid, G.J. Hutchings, Catal. Lett. 142 (2012) 1049.
◦
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007

G Model
CATTOD-8898; No. of Pages6
ARTICLE IN PRESS
T. Wang et al. / Catalysis Today xxx (2014) xxx–xxx
6
[
[
[
[
4] T. Tian, W.Z. Qian, X.P. Tang, S. Yun, F. Wei, Acta Phys. Chim. Sin. 26 (2010) 3305.
5] D. Freeman, R.P.K. Wells, G.J. Hutchings, Chem. Commun. 18 (2001) 1754.
6] H.A. Zaidi, K.K. Pant, Catal. Today 96 (2004) 155.
7] U. Olsbye, S. Svelle, M. Bjorgen, P. Beato, T.V.W. Janssens, F. Joensen, S. Bordiga,
K.P. Lillerud, Angew. Chem. Int. Ed. 51 (2012) 5810.
8] S.G. Diao, W.Z. Qian, G.H. Luo, F. Wei, Y. Wang, Appl. Catal. A 286 (2005) 30.
9] W.Z. Qian, T. Liu, Z.W. Wang, F. Wei, Z.F. Li, G.H. Luo, Y.D. Li, Appl. Catal. A 260
[13] E.G. Derouane, J.B. Nagy, P. Dejaifve, J.H.C.V. Hooff, B.P. Spekman, J.C. Vedrine,
C. Naccache, J. Catal. 53 (1978) 40.
[14] M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino c, L. Palumbo, S.
Bordiga, U. Olsbye, J. Catal. 249 (2007) 195.
[15] J.H. Kim, T. Kunieda, M. Niwa, J. Catal. 173 (1998) 433.
[16] S.R. Zheng, H.R. Heydenrych, A. Jentys, J.A. Lercher, J. Phys. Chem. B 106 (2002)
9552.
[
[
(
2004) 223.
10] X.B. Wei, H.B. Shi, W.Z. Qian, G.H. Luo, Y. Jin, F. Wei, Ind. Eng. Chem. Res. 48
2009) 128.
[17] M. Guisnet, N.S. Gnep, S. Morin, Microporous Mesoporous Mater. 35 (2000)
47.
[18] L.L. Li, X.W. Nie, C.S. Song, X.W. Guo, Acta Phys. Chim. Sin. 29 (2013) 754.
[19] Y.V. Joshi, K.T. Thomson, J. Phys. Chem. C 112 (2008) 12825.
[20] R.L.V. Mao, V. Ragaini, G. Leofanti, R. Fois, J. Catal. 81 (1983) 418.
[21] A. Kecskemeti, R. Barthos, F. Solymosi, J. Catal. 258 (2008) 111.
[
[
[
(
11] D.M. McCann, D. Lesthaeghe, P.W. Kletnieks, D.R. Guenther, M.J. Hayman, V.V.
Speybroeck, M. Waroquier, J.F. Haw, Angew. Chem. Int. Ed. 47 (2008) 5179.
12] S. Ilias, A. Bhan, ACS Catal. 3 (2013) 18.
Please cite this article in press as: T. Wang, et al., Conversion of methanol to aromatics in fluidized bed reactor, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.02.007