Amorphous silica-alumina composite with regulated acidity for efficient production of hydrogen via steam reforming of dimethyl ether

Article abstract of DOI:10.1016/j.cattod.2019.01.056

In this work, we demonstrate that the amorphous silica-alumina composite (ASA) is an efficient solid acid for the steam reforming of dimethyl ether (SRD). Thus, a series of ASA with different Si/Al molar ratios was synthesized via the modified hydrolytic

Full text of DOI:10.1016/j.cattod.2019.01.056

Catalysis Today xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Amorphous silica-alumina composite with regulated acidity for efficient
production of hydrogen via steam reforming of dimethyl ether
a
a
b
a
a
Shan-Shan Wang , Yong-Hong Song , Yong-Hua Zhao , Chang Liu , Yong-Shan Xiao ,
b
a
a,⁎
Qijian Zhang , Zhao-Tie Liu , Zhong-Wen Liu
Key Laboratory of Applied Surface and Colloid Chemistry (MOE), School of Chemistry & Chemical Engineering, Shaanxi Normal University, Xi’an, 710119, China
a
b
School of Chemistry & Environmental Engineering, Liaoning University of Technology, Jinzhou, 121001, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Hydrogen production
Dimethyl ether
Steam reforming
Amorphous silica-alumina
Bifunctional catalyst
In this work, we demonstrate that the amorphous silica-alumina composite (ASA) is an efficient solid acid for the
steam reforming of dimethyl ether (SRD). Thus, a series of ASA with different Si/Al molar ratios was synthesized
via the modified hydrolytic method by using Al(NO ) ·9H O and/or NaAlO as Al precursor and tetra-
3 3 2 2
ethoxysilane as Si precursor, respectively. Irrespective of the synthesis conditions, all of ASA composites were in
an amorphous nature. However, both the acidity and the surface area were significantly affected by the synthesis
conditions, the extent of which depends on the specific precursor and Si/Al molar ratios. The ASA physically
mixed with a commercial Cu/ZnO/Al O was investigated as a bifunctional catalyst for SRD under the conditions
2 3
−
1
of P=0.1 MPa and gas hourly space velocity of 4000 h . Results indicate that both the conversion of dimethyl
ether (DME) and the hydrogen yield were strongly dependent on the acidic property of ASA and the ASA to Cu/
2 3
ZnO/Al O ratios. Following the two-step consecutive mechanism of SRD, the catalytic results were well ex-
plained based on the acidic property of ASA, and the synergetic effect between the two catalytic functions was
found to play a crucial role in determining the activity and stability of the bifunctional catalyst. Moreover, the
DME hydrolysis was revealed to be the rate-limiting step for SRD reactions in the case of ASA as an acidic
catalyst. By optimizing the synthesis conditions of ASA and the composition of the bifunctional catalyst, greater
than 99% of DME conversion and hydrogen yield was achieved, and was kept for a time on stream of 66 h
without an observable decrease. Thus, ASA is a very efficient solid acid for SRD.
1. Introduction
the steam reforming of the intermediate product of methanol (Eq. (2),
SRM) [2]. Moreover, undesirable side reactions (Eqs. (4)–(9)) may also
In recent years, dimethyl ether (DME), which characterizes in the
occur, the extent of which depends on the catalyst and reaction con-
ditions.
gas-like property and liquid-storage density, low cost, high hydrogen
content (13 wt.%), and low or no toxicity, is recognized as a promising
candidate to on-site supply hydrogen in mobile applications such as the
fuel cell vehicles. Moreover, in comparison with other hydrogen-pro-
duction routes from DME such as the catalytic partial oxidation, steam
reforming of DME (SRD) not only produces high-quality reformate but
also can be performed at lower temperatures of 473–673 K [1,2]. Thus,
after the pioneering works of Sobyanin at the beginning of the 2000s
3,4], investigations on SRD in the aspect of thermodynamics analyses
5,6], catalyst screening [7–13], on-board dynamic analysis [14], and
reaction engineering including types of reactors [15–17] and kinetics
18–20], are increasingly reported.
−
1
CH
CH
CH
CO
3
3
3
2
OCH
3
+ 3H
+ H O = 2CH
O = CO + 3H
= CO + H O Reverse water-gas shift (RWGS)
= CH + H O Methanation
→ CH + CO +H DME decomposition
2CO = C + CO Boudouard reaction
CH OCH → hydrocarbons DME to hydrocarbons(DTH)
2
O = 2CO
2
+ 6H
2
ΔH° = 135 kJ mol
SRD
(1)
−1
OCH
3
2
3
OH ΔH° = 37 kJ mol
DME hydrolysis (2)
−
1
OH + H
2
2
2
ΔH° = 49 kJ mol
SRM
(3)
(4)
(5)
(6)
(7)
(8)
+ H
2
2
CO + 3H
CH OCH
2
4
2
[
[
3
3
4
2
[
2
It is commonly accepted that SRD (Eq. (1)) occurs via two-step
series reactions, i.e., DME hydrolysis to methanol (Eq. (2)) followed by
3
3
⁎
Corresponding author at: No. 620, West Chang’an Avenue, School of Chemistry & Chemical Engineering, Shaanxi Normal University, Xi’an, 710119, China.
E-mail address: zwliu@snnu.edu.cn (Z.-W. Liu).
https://doi.org/10.1016/j.cattod.2019.01.056
Received 30 September 2018; Received in revised form 18 December 2018; Accepted 25 January 2019
0920-5861/ © 2019 Elsevier B.V. All rights reserved.
Please cite this article as: Wang, S.-S., Catalysis Today, https://doi.org/10.1016/j.cattod.2019.01.056

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
CH
3
OH → hydrocarbons Methanol to hydrocarbons(MTH)
(9)
precursor. By using Al(NO
/3 as the alumina precursor, ASA-SN, where S and N refer to the
precursors of Al(NO and NaAlO , was synthesized. The ASA-SN with
3 3 2 2
) ·9H O and NaAlO at a fixed molar ratio of
1
According to the mechanism, a bifunctional catalyst with active
3
)
3
2
sites for catalyzing both the hydrolysis of DME and SRM are required to
complete SRD. Moreover, high hydrogen yields can be reasonably ex-
pected when the reaction rates of the DME hydrolysis and SRM are
sufficiently high and well matched [2]. Thus, the screening of the two
types of active components is extensively performed to improve the
SRD performance of the bifunctional catalyst. For the steam reforming
of the intermediate product of methanol, the Cu-based catalyst is con-
centrated [7–9,11–13,21] as a result of its sufficiently high activity and
low price although other catalysts such as Ga O [22], Mo C [10],
2 3 2
noble metals of Au [23], Rh [24,25], Pt [25], Pd [14,26], and the
transitional metal of Ni [27] are also reported.
However, great challenges are given to the DME hydrolysis as a
result of the lower kinetics rate at lower reaction temperatures and the
inhibiting effect of water [14,21,28]. Although much effort has been
paid on the development of alumina or zeolites as acid catalysts for SRD
a Si/Al molar ratio of 1/1, 1/4, 2/3, 3/2, 4/1, and 5/1 was abbreviated
as ASA-SN11, ASA-SN14, ASA-SN23, ASA-SN32, ASA-SN41, and ASA-
SN51, respectively.
2.2. Preparation of bifunctional catalysts
The bifunctional catalyst for SRD was a physical mixture of the ASA
and a commercial Cu/ZnO/Al
Co., Ltd.). To prepare the hybrid catalyst, the desired amount of ASA
and Cu/ZnO/Al were grinded into fine powders. Then, it was
pressed, crushed, and sieved into 40–60 mesh particles. The weight
ratio of Cu/ZnO/Al to ASA in the bifunctional catalyst was kept at
/1 if it was not specified.
2 3
O (C207-G, Changshu Kaituo Catalyst
2 3
O
2 3
O
1
[
2,21,29–33], the lower SRD activity of alumina due to a smaller
amount of acidic sites and easily coking of zeolites originated from the
stronger acidic sites lead either lower hydrogen yield or quick decrease
of the DME conversion [2,14,29–33]. In fact, amorphous silica-alumina
composite (ASA) has long been practiced as a solid acid catalyst for
many reactions in the petrochemical and other industries such as hy-
drocracking and isomerization [34]. Indeed, the ASA shows a moderate
acidity in comparison with alumina and zeolites although the origin of
the Brønsted acidity has not been unequivocally established [35,36].
However, it is conclusive that the acidity of ASA can be adjusted, to a
certain extent, by varying the silica to alumina ratios, the calcination
temperatures, and the preparation methods of the material [36–38].
Thus, quite a few works on the preparation and catalytic applications of
ASA have been reported in recent years [34,37,38]. In a recent work,
2
.3. Catalytic experimental procedure and product analyses
The SRD reaction was carried out in a quartz tubular reactor
(
i.d. = 8.0 mm). After loading the hybrid catalyst (0.6 g, 40–60 mesh),
it was reduced in a diluted H (10 vol.% H in N ) with a temperature
programmed procedure. Then, the water, DME, and N at a fixed molar
ratio of DME/H O/N = 1/4/5 mixed in a gas mixer with a constant
2
2
2
2
2
2
temperature of 150 °C were switched, and the SRD reaction was started
at the desired temperature, 0.1 MPa, and the gas hourly space velocity
−1
(
GHSV) of 4000 h . The deionized water was injected with an HPLC
syringe pump and was evaporated at 150 °C. The flow rate of DME and
the diluent N were controlled by the mass flow controller. The effluent
products were on-line analyzed by a gas chromatograph (GC-9560,
Huaai chromatographic analysis Co., Ltd.) equipped with a thermal
conductivity detector (TCD) and a flame ionization detector (FID). The
detailed GC analyses and the method for calculating the DME conver-
2
mesoporous Cu-Al
induced self-assembled method was comparatively investigated as a
2 3
bifunctional catalyst for SRD [39]. The mesoporous Cu-SiO -Al O
2 3 2 2 3
O and Cu-SiO -Al O prepared by the evaporation
2
2 2 3
shows clearly higher DME conversion and H yield than Cu-Al O al-
sion, the H
were described in our previous works [2,14].
2
yield, and the selectivity of carbon-containing products
though it is easily deactivated.
Based on these understandings, in this work, we demonstrate that
ASA is a highly efficient and very promising solid acid for SRD. Thus, a
series of ASA was synthesized, and its synthesis parameters and Si/Al
molar ratio were optimized. According to the two-step mechanism of
SRD, the bifunctional catalyst was made by physically mixing ASA with
2
.4. Characterization methods
Powder X-ray diffraction (XRD) patterns were obtained at the am-
a commercial Cu/ZnO/Al
and stability for methanol synthesis. With the optimal ASA, the DME
conversion and H yield of > 99% were achieved over the bifunctional
2 3
O considering its reasonably high activity
bient temperature on a D8 advance X-Ray diffractometer with a
monochromatised Cu/K radiation (40 kV, 40 mA). The samples were
scanned from 5 to 70° (2θ) at a speed of 8° per minute.
The N adsorption-desorption were performed on a Micromeritics
ASAP 2020 instrument at −196 °C. Before the measurement, each
sample of about 100 mg was degassed at 300 °C for 12 h. The surface
area was calculated by the Brunaure-Emmett-Teller (BET) method, and
α
2
catalyst, and were kept for a time on stream (TOS) of at least 66 h
without any observable decrease.
2
2. Experimental section
the pore volume was determined at a relative pressure (P/P
The acidic property of ASA was evaluated by the temperature-pro-
grammed desorption of ammonia (NH -TPD) method, which is carried
0
) of 0.99.
2.1. Synthesis of ASA
3
The ASA was synthesized via the modified hydrolytic method as
out on a Micromeritics Auto 2920 instrument. For each test, about
0.05 g sample was loaded. Firstly, the sample was flushed with a He
flow at 500 °C for 1 h. Then, the temperature was decreased to 120 °C
reported in the reference [40]. To regulate the porous property of ASA,
cetyltrimethylammonium bromide (CTAB) was added as a template
during the synthesis. For a typical synthesis, 2.0 g CTAB was dissolved
in 50 ml deionized water at the ambient temperature. To the solution,
−1
under a He flow, and the diluted NH
was switched. After completing the adsorption of NH
30 min, it was flushed with a He flow for 2 h to remove the physically
adsorbed NH . Finally, NH -TPD experiments were conducted from 120
to 500 °C at a temperature ramp of 10 °C⋅ min under a He flow of 25
3
with a flow rate of 20 mL⋅ min
3
at 120 °C for
0
2
.03 mol NaAlO was added under vigorous stirring. Then, the absolute
ethanol solution of 0.03 mol tetraethoxysilane (TEOS) was slowly
added. After aging at 80 °C for 24 h, it was centrifugated and washed
thoroughly with deionized water. The solid was dried at 80 °C for 6 h,
and calcined in air at 600 °C for 8 h. The thus obtained ASA with a Si/Al
3
3
−
1
−
1
mL⋅ min
.
The used bifunctional catalysts were characterized by the thermo-
gravimetric differential scanning calorimetry (TG-DSC) method on a
Q1000DSC + LNCS + FACS Q600SDT thermal analyzer. The experi-
ments were performed from the room temperature to 1000 °C at a ramp
molar ratio of 1 was abbreviated as ASA-S1, where S stands for NaAlO
as the alumina precursor and 1 means the Si/Al molar ratio. Following
the same procedure by simply substituting Al(NO ·9H O for NaAlO
ASA-N1 was obtained, where N stands for Al(NO as the alumina
2
3
)
3
2
2
,
−
1
3
)
3
of 10 °C min
under the air atmosphere.
2

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 1. XRD patterns of different ASA composites.
3. Results and discussion
3.1. Structural and textural properties
The XRD patterns of the ASA are shown in Fig. 1. For all of the
samples, only a broad peak centered at 2θ of about 25° was clearly
observed, which is easily assigned to the characteristic diffraction of the
amorphous silica. Moreover, there was still no any diffraction from
aluminum species even for the sample with the highest content of
alumina (ASA-SN14), suggesting that aluminum species are well dis-
tributed in the matrix of silica. When the broad peak was concerned, its
intensity and span were slightly varied depending on the synthesis
parameters, i.e., the Al precursor and Si/Al ratio. Thus, irrespective of
the Al precursor and Si/Al ratio, the amorphous ASA composite with
the uniformly distributed aluminum in silica was obtained.
2
The N adsorption-desorption isotherms of the ASA are shown in
Fig. 2. Irrespective of the samples, the type IV isotherms with H2
hysteresis loops were exclusively observed according to the IUPAC
classification [41]. However, significant difference in the maximum
Fig. 2. N2 adsorption-desorption isotherms of various ASA composites.
Table 1
Summary of the textural properties of the ASA.
2
capacity of the N adsorbed, the size, shape and the relative pressures of
the hysteresis loop was apparently observed (Fig. 2). This indicates the
complex pore structure of the samples with the pore blocking in a wide
range of pore necks and the filling of micropores as discussed in re-
ference [41]. As reflected from the textural properties summarized in
Table 1, the increase in the Si/Al ratio of ASA generally led to an in-
creased BET surface area and a continuously decreased total pore vo-
lume, and ASA-SN51 having the highest Si/Al ratio of 5 showed the
highest surface area and smallest total pore volume. As a result, a re-
versed changing pattern for the average pore size was expectedly no-
ticed. As reflected from the isotherms (Fig. 2), the changing pattern of
these textural properties with increasing the Si/Al ratio of ASA can be
easily explained as the decreased amount of mesopores and the in-
creased amount of micropores. When the aluminum precursor for
synthesizing ASA was concerned, ASA-N1 showed much higher surface
area than ASA-S1 or ASA-SN11. This indicates that higher surface-area
Sample
BET surface area
Total pore volume
/cm ⋅ g
Average pore diameter
/nm
2
−1
3
−1
/m ⋅ g
ASA-SN14
ASA-SN23
ASA-SN11
ASA-SN32
ASA-SN41
ASA-SN51
ASA-S1
276
225
240
421
568
741
213
519
1.78
1.02
0.94
0.94
0.59
0.55
0.96
0.63
25.8
18.2
15.6
6.6
5.6
3.0
18.0
3.9
ASA-N1
significant impact on the amount and strength of acidic sites over ASA.
Moreover, weaker acidic sites were predominant irrespective of the
synthesis parameters of ASA composites, which is agreeable with re-
ference results [34–36]. When the aluminum precursor was taken into
account (Fig. 3A), the ASA-N1 showed the highest amount of weaker
acidic sites. In contrast, the highest amounts of total and stronger acidic
sites were obtained over ASA-SN11. In the cases of changing the Si/Al
ratios (Fig. 3B), the peak temperature for the desorption of NH
slightly increased with increasing the Si/Al ratio from 1/4 (˜165 °C) to
/1 (˜210 °C). However, an insignificant decrease of the peak tem-
perature was observed if the Si/Al ratio was further increased to be 5/1
˜200 °C). If the amount of acidic sites were concerned, it was not
3 3 2
ASA can be obtained by using Al(NO ) ·9H O as the precursor, which
should be closely related to its hydrolytic behavior. Thus, both the
aluminum precursor and the Si/Al ratio had strong impacts on the BET
surface area, pore volume, and pore size of ASA composites.
3
was
4
3.2. Acidic features
(
3
Generally, the desorption of NH at the peak temperature of about
monotonically changed with the Si/Al ratio of ASA. The amount of
acidic sites was decreased in the order of ASA-SN41 > ASA-SN14 ≈
ASA-SN32 ≈ ASA-SN51 > ASA-SN11 > ASA-SN23. Thus, the changing
of aluminum precursor and Si/Al ratio led the varied amount and
2
00 and 350 °C can be assigned to weaker and stronger acidic sites over
3
solid acids such as zeolites, respectively [2]. As indicated from the NH -
TPD patterns of different ASA composites (Fig. 3), the synthesis para-
meters, i.e., the aluminum precursor and the Si/Al ratio, had a
3

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 4. The time-on-stream DME conversion and H
of carbon-containing products at TOS of 3 h (B) over the hybrid catalyst of 0.3 g
ASA + 0.3 g Cu/ZnO/Al under the conditions of T = 300 °C, P = 0.1 MPa,
2
yield (A) and the selectivity
2 3
O
−
1
and GHSV = 4000 h
.
3
Fig. 3. NH -TPD patterns of ASA composites synthesized under different con-
ditions.
To investigate the effect of Si/Al ratios of ASA as a solid acid on the
SRD performance of the hybrid catalyst, ASA with largely varied Si/Al
ratios from 5/1 to 1/4 were synthesized by using Al
strength of acidic sites over ASA.
(
3 3 2 2
NO ) ·9H O + NaAlO as the aluminum precursor considering that the
3
.3. SRD performance
ASA-SN11 based bifunctional catalyst showed the highest SRD perfor-
mance. The TOS SRD results under the same conditions are given in
Fig. 5. As indicated in Fig. 5A, stable DME conversions over different
catalysts were between ˜40 - 55%, which is dependent on the Si/Al ratio
of the ASA. Specifically, the highest DME conversion was obtained over
ASA-SN41 (˜55%), which is slightly higher than that over ASA- SN32
2 3
The ASA physically mixed with Cu/ZnO/Al O was evaluated as a
bifunctional catalyst for SRD under the conditions of T = 300 °C,
P = 0.1 MPa, the inlet molar ratio of DME/H O/N = 1/4/5, and
GHSV = 4000 h . The time-on-stream DME conversion, H yield, and
the selectivity of carbon-containing products are shown in Fig. 4. The
DME conversion over the bifunctional catalyst was decreased in the
order of ASA-SN11 > ASA-S1 > ASA-N1 (Fig. 4A). Moreover, the same
changing pattern was observed in the case of H
selectivity of methanol and hydrocarbons as by-products was decreased
in a reversed order (Fig. 4B). Noteworthy, CH and CO were not de-
tected under these conditions, indicating that the RWGS (Eq. (4)), the
methanation (Eq. (5)), the DME decomposition (Eq. (6)), and Bou-
douard reactions (Eq. (7)) were greatly suppressed. As a result, almost
the same DME conversion and H
case of ASA-SN11-based bifunctional catalyst (Fig. 4A). Thus, ASA is an
effective solid acid in decreasing side reactions during SRD. If the NH
TPD results were correlated with the catalytic results, the changing
order for the total amount of acidic sites over ASA was coincident with
2
2
−1
2
(
˜50%). Moreover, ASA-SN14, ASA-SN23, and ASA-SN11 showed al-
most the same DME conversion (˜40-45%), which is very similar to that
over ASA-SN51. When the H yield was examined (Fig. 5B), the same
changing pattern was observed. Again, the H yield (˜54%) over ASA-
SN41 based catalyst was very similar to that of the DME conversion
˜55%), indicating the negligible side reactions. This was more clearly
2
2
yield. As a result, the
2
4
(
reflected from the selectivity of carbon-containing products (Fig. 5C),
i.e., less than 1% selectivity of methanol plus hydrocarbons. Note-
worthy, the highest selectivity of methanol plus hydrocarbons over all
of the hybrid catalysts was less than 4%, which represents the merit of
ASA as a solid acid for SRD in comparison with zeolites or alumina
2
yield was observed, especially in the
3
-
[
3
14,29–33]. When the acidic property of ASA characterized by NH -
TPD was compared with the SRD result over different catalysts, the
total amount of acidic sites was again revealed to the crucial factor in
determining the SRD performance, and the highest performance of
ASA-SN41 based catalyst was consistent with its highest amount of
acidic sites. The question may be raised from the contribution of tex-
tural properties of the ASA on the SRD performance of the hybrid
2
those for the DME conversion and the reverse H yield, indicating that
the total amount of acidic sites over ASA is the key parameter in de-
termining the SRD performance of the hybrid catalyst. Among the
aluminum precursors used, the ASA-SN11 synthesized with Al
(
NO
3 3 2 2
) ·9H O + NaAlO showed the highest SRD performance since that
catalyst. However, if the changing order of the DME conversion or H
2
the Si/Al ratio was the same for ASA-S1, ASA-N1, and ASA-SN11.
4

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 6. Effect of the weight ratio of ASA-SN41 to Cu/ZnO/Al
stream DME conversion and H yield (A) and the selectivity of carbon-con-
taining products at TOS of 3 h (B) under the conditions of T = 300 °C,
2 3
O on the time-on-
2
−
1
P = 0.1 MPa, and GHSV = 4000 h
.
3.4. Synergetic effect and long-term stability
Considering the two-step series nature of the SRD reaction, hybrid
catalysts with varied weight ratios of ASA-SN41 to Cu/ZnO/Al O were
2 3
evaluated under the same reaction conditions. From the results shown
in Fig. 6, the most obvious observation was that the DME conversion,
H
2
yield, and the stability of the hybrid catalyst were strongly depen-
dent on the composition of the hybrid catalyst (Fig. 6A). With in-
creasing the weight ratio of ASA-SN41 to Cu/ZnO/Al from 1/2 to 1/
, the initial DME conversion and the H yield were almost unchanged.
However, a significant decrease in both DME conversion and the H
2 3
O
1
2
2
yield at a TOS of 6 h was observed over the hybrid catalyst with less
ASA-SN41 while they were still stable in the case with more ASA-SN41.
If the weight ratio of ASA-SN41 to Cu/ZnO/Al
from 1/1 to 2/1, a clear increase in the DME conversion and the H
yield was achieved. If it was further increased from 2/1 to 3/1, a clearly
lower H yield was obtained although the DME conversion was kept
almost the same. This was evidently revealed from the selectivity of
carbon-containing products (Fig. 6C) as a result of more pronounced
side reactions. According to the SRD mechanism with two-step series
reactions [2,4], DME is firstly hydrolyzed to methanol (Eq. (2)) cata-
lyzed by the acid, and the subsequent steam reforming of the produced
2 3
O was further increased
2
2
Fig. 5. The time-on-stream DME conversion (A), H yield (B) and the selectivity
of carbon-containing products at TOS of 3 h (C) over the hybrid catalysts of
.3 g ASA + 0.3 g Cu/ZnO/Al under the conditions of T = 300 °C,
2
0
2
O
3
P = 0.1 MPa, and GHSV = 4000 h 1.
−
yield over different hybrid catalysts was compared with that of the BET
surface area or total pore volume of ASA (Table 1), no simple correla-
tion was found. For example, although ASA-SN51 showed a sig-
2 2
methanol into CO and H over Cu-based catalyst (Eq. (3)). As the two
2
−1
reactions are closely interrelated, synergetic effects between the two
catalytic functions have been reported [2,8,14,21,30]. In our case, all of
the conditions were kept the same except the weight ratio of ASA-SN41
nificantly higher BET surface area (741 m ⋅ g ) than that of ASA-SN11
2
−1
(
240 m ⋅ g ), almost the same SRD performance was achieved over
these two catalysts. Thus, the textural property of ASA plays a less
important role in determining the SRD performance of the hybrid cat-
alyst if it has.
2 3
to Cu/ZnO/Al O . Thus, the synergetic effect between ASA and the Cu-
based catalyst, i.e., the two catalytic functions are reciprocally inter-
acted, could be realized to explain these observations because that ei-
ther a smaller (1/2) or higher (3/1) weight ratio of ASA-SN41 to Cu/
5

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
coke deposited over the spent catalyst, which is consistent to the cat-
alytic results given in Fig. 8. Thus, the long-term stability of the hybrid
catalyst for SRD can be reasonably expected, and the ASA composite is
proved a promising solid acid for the titled reaction.
4. Conclusions
In summary, a series of amorphous ASA composites was synthe-
sized, and the ASA physically mixed with a commercial Cu/ZnO/Al O
2 3
was studied as a bifunctional catalyst for SRD. Independent on the
synthesis conditions, all of the ASA composites were amorphous, and no
any crystallized aluminum species were formed even for the ASA with
the highest amount of aluminum (Si/Al molar ratio of 1/4). However,
2
−1
the BET surface area (213 - 741 m ⋅ g ), the total pore volume (0.55 -
3
−1
1
.78 cm ⋅ g ), and the average pore size (3.0–25.8 nm) of ASA were
significantly varied by changing the aluminum precursor or the Si/Al
ratios. Moreover, the amount and strength of acidic sites over ASA was
closely related to the synthesis conditions although weaker acidic sites
were predominant over all of the samples. In the cases of the catalytic
reactions, the DME conversion and the hydrogen yield over the bi-
functional catalyst were decreased in the order of ASA-SN11 > ASA-
S1 > ASA-N1. Regarding the effect of S/Al ratios, the DME conversion
and the hydrogen yield were decreased in the order of ASA-
SN41 > ASA-SN32 > ASA-SN11 ≈ ASA-SN14 ≈ ASA-SN23 ≈ ASA-
SN51. Following the two-step consecutive mechanism of SRD, the cat-
alytic results were well explained based on the acidic property of ASA,
and the total amount of acidic sites was the crucial factor in de-
termining the SRD performance of the hybrid catalyst. Moreover, the
DME hydrolysis was revealed to be the rate-limiting step for the titled
reaction. By optimizing the synthesis conditions of ASA and the com-
position of the bifunctional catalyst, an active, selective, and stable
bifunctional catalyst for SRD was achieved. Thus, ASA was concluded to
be an efficient and very promising solid acid for SRD.
2
Fig. 7. Effect of reaction temperature on the DME conversion and H yield over
2 3
the hybrid catalysts of 0.4 g ASA-SN41 + 0.2 g Cu/ZnO/Al O under the con-
−
1
ditions of P = 0.1 MPa and GHSV = 4000 h
.
Declarations of interest
None.
Fig. 8. The long-term DME conversion and H
2
yield over the hybrid catalysts of
2 3
.4 g ASA-SN41 + 0.2 g Cu/ZnO/Al O under the conditions of T = 350 °C,
−
0
1
P = 0.1 MPa, GHSV = 4000 h
.
Acknowledgments
The financial supports by the National Natural Science Foundation
of China (21636006), Shaanxi Innovative Team of Key Science and
Technology (2013KCT-17), and the Fundamental Research Funds for
the Central Universities (GK201901001) are highly appreciated.
ZnO/Al
version and H
2
O
3
was unfavorable to the H
2
yield. Thus, a high DME con-
2
yield can be obtained only when the acid catalysis for
the DME hydrolysis is well matched with the Cu-based catalysis for the
steam reforming of the intermediate product of methanol. Moreover,
the hydrolysis of DME can be reasonably proposed to be the rate-de-
termining step for the titled reaction because that more acid catalyst
References
was required to achieve the highest H
ASA-SN41 to Cu/ZnO/Al of 2/1. This is clearly different from the
case where the bifunctional catalyst is composed of zeolite and Cu/
ZnO/Al , i.e., the lower rate of SRM than that of the DME hydrolysis
14].
To reveal the impact of the reaction temperature on the synergetic
effect, the hybrid catalyst of ASA-SN41 and Cu/ZnO/Al with the
optimal weight ratio of 2/1 was evaluated for SRD (Fig. 7). With in-
creasing the reaction temperature from 300 to 350 °C, the DME con-
version were increased to be about 100%, which is almost the same to
2
yield, i.e., the weight ratio for
[
[
1] K. Takeishi, H. Suzuki, Appl. Catal. A Gen. 260 (2004) 111–117.
2] X. Long, Q. Zhang, Z.-T. Liu, P. Qi, J. Lu, Z.-W. Liu, Appl. Catal. B Environ. 134-135
2 3
O
(2013) 381–388.
2 3
O
[3] V.A. Sobyanin, S. Cavallaro, S. Freni, Energy Fuel 14 (2000) 1139–1142.
[
[
4] V.V. Galvita, G.L. Semin, V.D. Belyaev, T.M. Yurieva, V.A. Sobyanin, Appl. Catal. A
Gen. 216 (2001) 85–90.
[
5] T.A. Semelsberger, R.L. Borup, J. Power Sources 155 (2006) 340–352.
2
O
3
[6] K. Faungnawakij, N. Viriya-empikul, W. Tanthapanichakoon, Int. J. Hydrogen
Energy 36 (2011) 5865–5874.
[
7] K. Faungnawakij, R. Kikuchi, N. Shimoda, T. Fukunaga, K. Eguchi, Angew. Chem.
Int. Ed. 47 (2008) 9314–9317.
[8] K. Faungnawakij, K. Eguchi, Catal. Surv. Asia 15 (2011) 12–24.
[9] M. Nilsson, K. Jansson, P. Jozsa, L.J. Pettersson, Appl. Catal. B Environ. 86 (2009)
the H
2
yield. This clearly indicates that the two catalytic functions for
1
8–26.
the DME hydrolysis and SRM are still matched very well. Thus, a stable
performance of the hybrid catalyst for SRD is reasonably expected. As
[
[
10] F. Solymosi, R. Barthos, A. Kecskemeti, Appl. Catal. A Gen. 350 (2008) 30–37.
11] P.V. Snytnikov, S.D. Badmaev, G.G. Volkova, D.I. Potemkin, M.M. Zyryanova,
V.D. Belyaev, V.A. Sobyanin, Int. J. Hydrogen Energy 37 (2012) 16388–16396.
12] L. Oar-Arteta, A.T. Aguayo, A. Remiro, J. Bilbao, A.G. Gayubo, Ind. Eng. Chem. Res.
2
shown in Fig. 8, ˜100% DME conversion and H yield were kept over
[
[
[
[
66 h without any observable decreases. To reveal the possible coke
5
4 (2015) 11285–11294.
deposition during SRD, the hybrid catalyst tested after SRD for 66 h was
applied to TG-DSC analysis, and the results are given in Fig. S1 (Sup-
plementary material). By comparing the TG-DSC patterns of the fresh
and used hybrid catalysts, it is conclusive that there was no detectable
13] C. Ledesma, E. López, T. Trifonov, Á. Rodríguez, J. Llorca, Catal. Today (2019),
https://doi.org/10.1016/j.cattod.2018.03.011.
14] J. Li, Q. Zhang, X. Long, P. Qi, Z.-T. Liu, Z.-W. Liu, Chem. Eng. J. 187 (2012)
2
99–305.
15] F.A. Elewuwa, Y.T. Makkawi, Int. J. Hydrogen Energy 40 (2015) 15865–15876.
6

S.-S. Wang et al.
Catalysis Today xxx (xxxx) xxx–xxx
[
[
16] C.-F. Yan, W. Ye, C.-Q. Guo, S.-L. Huang, W.-B. Li, W.-M. Luo, Int. J. Hydrogen
Energy 39 (2014) 18642–18649.
17] A.S. Fedotov, D.O. Antonov, V.I. Uvarov, M.V. Tsodikov, S.N. Khadzhiev, Pet.
Chem. 58 (2018) 62–67.
157–161.
[30] K. Takeishi, Y. Akaike, Appl. Catal. A Gen. 510 (2016) 20–26.
[31] A.G. Gayuboa, J. Vicente, J. Ereña, L. Oar-Arteta, M.J. Azkoiti, M. Olazara,
J. Bilbao, Appl. Catal. A Gen. 483 (2014) 76–84.
[
[
18] D. Feng, Y. Wang, D. Wang, J. Wang, Chem. Eng. J. 146 (2009) 477–485.
19] L. Oar-Arteta, A.T. Aguayo, A. Remiro, A. Arandia, J. Bilbao, A.G. Gayubo, Chem.
Eng. J. 306 (2016) 401–412.
20] F.A. Elewuwa, Y.T. Makkawi, Int. J. Hydrogen Energy 41 (2016) 19819–19828.
21] T.A. Semelsberger, K.C. Ott, R.L. Borup, H.L. Greene, Appl. Catal. B Environ. 65
[32] J. Vicente, A.G. Gayubo, J. Ereña, A.T. Aguayo, M. Olazar, J. Bilbao, Appl. Catal. B
Environ. 130-131 (2013) 73–83.
[33] H. Chu, Q. Zhang, Y. Chu, Q. Ding, Q. Wang, H. Xu, L. Zhang, Catal. Today (2019),
https://doi.org/10.1016/j.cattod.2018.03.027.
[34] K.A. Nadeina, O.V. Klimov, I.G. Danilova, V.Yu. Pereyma, E.Yu. Gerasimov,
I.P. Prosvirin, A.S. Noskov, Appl. Catal. B Environ. 223 (2018) 22–35.
[35] Z. Wang, Y. Jiang, O. Lafon, J. Trébosc, K. Kim, C. Stampfl, A. Baiker, J.-
P. Amoureux, J. Huang, Nat. Commun. (2016), https://doi.org/10.1038/
ncomms13820.
[
[
(
2006) 291–300.
22] S. Zhou, K. Ma, Y. Tian, M. Meng, T. Ding, Y. Zha, T. Zhang, X. Li, RSC Adv. 6
2016) 52411–52420.
[
(
[23] A. Gazsi, I. Ugrai, F. Solymosi, Appl. Catal. A Gen. 391 (2011) 360–366.
[24] G. Halasi, T. Sansagi, F. Solymosi, ChemCatChem 1 (2009) 311–317.
[25] L. Oar-Arteta, A. Remiro, F. Epron, N. Bion, A.T. Aguayo, J. Bilbao, A.G. Gayubo,
Ind. Eng. Chem. Res. 55 (2016) 3546–3555.
[36] V.S. Escribano, G. Garbarino, E. Finocchio, G. Busca, Top. Catal. 60 (2017)
1554–1564.
[37] M. Zabeti, J. Baltrusaitis, K. Seshan, Catal. Today 269 (2016) 156–165.
[38] K. Larmier, C. Chizallet, S. Maury, N. Cadran, J. Abboud, A.-F. Lamic-Humblot,
E. Marceau, H. Lauron-Pernot, Angew. Chem. Int. Ed. 56 (2017) 230–234.
[39] Y. Zang, X. Dong, C. Wang, Chem. Eng. J. 313 (2017) 1583–1592.
[40] L. Kou, Y. Wang, Y. Dong, D. Han, B. Ni, Z. Yan, Mater. Lett. 121 (2014) (2014)
212–214.
[
[
[
[
26] E. Ramos, L. Davin, I. Angurell, C. Ledesma, J. Llorca, ChemCatChem 7 (2015)
2
179–2187.
27] R. Gonzalez-Gil, C. Herrera, M.A. Larrubia, P. Kowalik, I.S. Pieta, L.J. Alemany, Int.
J. Hydrogen Energy 41 (2016) 19781–19788.
28] T. Kawabata, H. Matsuoka, T. Shishido, D. Li, Y. Tian, T. Sano, K. Takehira, Appl.
Catal. A Gen. 308 (2006) 82–90.
[41] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Olivier, F. Rodriguez-Reinoso,
J. Rouquerol, K.S.W. Sing, Pure Appl. Chem. 87 (2015) 1051–1069.
29] K. Faungnawakij, R. Kikuchi, T. Fukunaga, K. Eguchi, Catal. Today 138 (2008)
7