Hydrogen production by dimethyl ether steam reforming over copper alumina catalysts prepared using the sol-gel method

Article abstract of DOI:10.1016/j.apcata.2015.09.027

Several copper alumina catalysts prepared using the sol-gel method were examined for the production of hydrogen from dimethyl ether (DME). Each of the catalysts was used individually for DME steam reforming, without mixing of DME hydrolysis catalysts such

Full text of DOI:10.1016/j.apcata.2015.09.027

Applied Catalysis A: General 510 (2016) 20–26
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Hydrogen production by dimethyl ether steam reforming over copper
alumina catalysts prepared using the sol–gel method
∗
Kaoru Takeishi , Yoshimi Akaike
Department of Engineering, Graduate School of Integrated Science and Technology, Shizuoka University, 3-5-1 Jouhoku, Hamamatsu-shi, Shizuoka-ken,
432-8561, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 June 2015
Received in revised form 5 September 2015
Accepted 21 September 2015
Available online 29 October 2015
Several copper alumina catalysts prepared using the sol–gel method were examined for the production
of hydrogen from dimethyl ether (DME). Each of the catalysts was used individually for DME steam
reforming, without mixing of DME hydrolysis catalysts such as alumina and zeolite. In order to obtain
hydrogen by DME steam reforming, mixed catalysts with DME hydrolysis catalysts and methanol steam
reforming catalysts are usually used based on the reaction mechanism. However, the rate of hydrogen
production in the single use of copper alumina catalysts prepared using the sol–gel method is faster
than those of mixed catalysts. SEM-EDS analysis confirmed that alumina and copper coexist and are
well dispersed on the surfaces of sol–gel copper alumina catalysts. Therefore, two reactions of DME
hydrolysis over alumina sites and methanol steam reforming over copper sites occur more sequentially
in the case of sol–gel catalysts than in the case of mixed catalysts. Several experiments also suggest
that Cu–Zn(29–1 wt.%)/Al2O3 catalyst prepared using the sol–gel method is the best catalyst for more
Keywords:
Dimethyl ether (DME)
Steam reforming
Hydrogen
Sol–gel
Copper
◦
hydrogen production and less CO production from DME. DME steam reforming over the catalyst at 275 C
showed the following data; DME conversion: 95%, yield of H2: 95%, and CO concentration: 0.8 mol.%. We
have developed some outstanding new catalysts that produce hydrogen from DME at the low reforming
temperature.
©
2015 Elsevier B.V. All rights reserved.
1
. Introduction
tial fuel for hydrogen production to be used in fuel cells. DME is an
excellent hydrogen source, storage, and carrier [3]. 13 wt.% of DME
molecule (CH OCH ; MW 46.07) is H , and if one DME molecule
The fuel cell is expected as a method of restraining global warm-
3
3
2
ing because of its excellent energy efficiency. Steam reforming of
fossil fuels is being actively researched and developed as a hydro-
gen supply method for fuel cells. Dimethyl ether (DME) is expected
as a clean fuel for the 21st century [1], because DME burns with the
emission of no particulate matter (PM), no sulfur oxides (SOx), and
less nitrogen oxides (NOx). The cetane number of DME is 55–60,
and DME is attractive as a clean fuel for diesel engines. Its physi-
cal properties are similar to those of liquefied petroleum gas (LPG).
will be treated with H O six H2 molecules will be produced.
2
We have been studying the steam reforming of DME for hydro-
gen production. New types of catalysts have been developed using
the sol–gel [4,5]. Hydrogen production by steam reforming of DME
(SRDME) consists of two steps: the first is DME hydrolysis, and the
second is the steam reforming of methanol (SRM). The chemical
equation of SRDME (Eq. (1)) is separated for that of DME hydrolysis
(Eq. (2)) and that of SRM (Eq. (3)), as follows:
◦
DME is easily liquefied at −25 C and atmospheric pressure, and at
0.5 MPa and ambient temperature. DME is easily handled like LPG
CH OCH + 3H O → 6H + 2CO
3
3
2
2
2
(
(
1)
2)
and used for domestic fuels. DME is produced from coal, natural
gas, biomass, and so forth. DME is a multi-source and multi-use
fuel and energy. Global annual DME production capacity is approx-
imately 10 million metric tons and actual market use is about 3
million metric tons [2]. Recently, DME has also become a poten-
(
steam reforming of DME) ,
CH OCH + H O → 2CH OH
3
3
2
3
(
hydrolysis of DME) ,
2
CH OH + 2H O → 6H + 2CO
3
2
2
2
∗ Corresponding author. Fax : +81 53 478 1159.
E-mail address: takeishi.kaoru@shizuoka.ac.jp (K. Takeishi).
(3)
(
steam reforming of methanol)
http://dx.doi.org/10.1016/j.apcata.2015.09.027
0
926-860X/© 2015 Elsevier B.V. All rights reserved.

K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
21
Most researchers [6–21] have reported mixed catalysts with
DME hydrolysis catalysts and SRM catalysts, because of the above-
mentioned equations. However, we focused on a single use type
catalyst prepared using the sol–gel method in terms of catalyst
activity, catalyst life, and for ease of industrial processing that
eliminates the mechanical mixing procedure. We have already
reported that the copper alumina catalysts prepared using the
sol–gel method in single use produce more hydrogen than phys-
ically mixed catalysts by commercial catalysts of DME hydrolysis
and SRM [4]. As Volkova et al. mentioned [22], these catalysts pre-
pared using the sol–gel method are the first bifunctional catalysts
reported out. After that Volkova et al. and Snytnikov et al. developed
catalysts and the catalysts themselves, and lead to a deterioration
of activity. However, we consider that the sintering before reaction
is a smaller problem for a comparison of catalyst activity than sin-
tering while the reaction takes place. Instead of the duration test
of the catalysts, the catalysts after this intense pretreatment were
compared on the activity, selectivity and so forth. The alumina in
the catalyst was found to be mainly ␥-alumina by measured X-ray
diffraction analysis. The condition of copper is metallic form and
zinc is ZnO form, but for ease of calculation for preparation the
catalyst is called such as Cu–Zn(4.5–0.5 wt.%)/Al O3 catalyst.
2
Comparison catalysts of SRDME, Cu–Zn(9–1 wt.%)/SiO , were
2
prepared by almost same way of the above-mentioned sol–gel
method. Difference is the selection and change of AIP to tetraethyl
orthosilicate (TEOS; purity of (C H5O) Si: 95%; Wako). This
the bifunctional Cu–Ce/Al O3 catalysts [22,23]. In our previous
2
paper, we speculated that the Lewis acid sites on ␥-Al O for DME
2
3
2
4
hydrolysis and active sites for SRM such as copper are coexistent
and well dispersed on the surfaces of the catalysts. It is the reason
that the Cu/Al O catalysts prepared using the sol–gel method are
Cu–Zn(9–1 wt.%)/SiO2 catalyst prepared using the sol–gel method
is one of the excellent catalysts for H2 production by SRM. For
example, H2 yield by SRM over the Cu–Zn(9–1 wt.%)/SiO2 catalyst
2
3
◦
◦
effective for hydrogen production by DME steam reforming.
at 350 C and at 400 C under the atmospheric pressure are 99.3%
and 97.9%, respectively [24–26]. As mentioned above, SRDME (Eq.
(1)) consists of DME hydrolysis (Eq. (2)) and SRM (Eq. (3)), so excel-
lent SRM catalysts have excellent performance for H2 production
by SRDME. ␥-Al O was prepared by AIP hydrolysis without Cu and
In this paper, we analyzed the surface of Cu–Zn/Al O3 catalysts
2
prepared using the sol–gel method in order to confirm our spec-
ulation by SEM-EDS. We also investigated for the improvement of
further activities. In a previous paper [4], catalysts with 10 wt.%
metal loading were examined from the viewpoint of the sol–gel
method. However, the activities of SRM increases with the increase
2
3
Zn nitrates as mentioned above. ␥-Al O is one of acid catalysts, and
2
3
it is good catalyst for DME hydrolysis [4]. These catalysts were also
ground by using an agate mortar until the diameter of each grain
of powder was less than 150 ␮m, respectively. Each portion of the
of Cu loading percent in Cu/SiO2 and Cu/Al O3 prepared using the
2
sol–gel method, and the optimum percent was 30–50 wt.%Cu for
◦
SRM [24–28]. In this paper, for optimization of Cu–Zn–Al O3 ratio,
powder was also calcined at 500 C for 5 h. These two catalysts were
2
the influences of Cu loading, and amount of Zn addition were also
investigated for improvement of the hydrogen yield and the hydro-
gen production rate from DME.
used for SRDME in two mixing ways, details are mentioned in the
results session.
2
.2. Surface analysis of Cu–Zn/Al O catalyst prepared using the
2 3
sol–gel method
2
. Experimental
Scanning Electron Microscope-Energy Dispersive X-ray Spec-
trometer (SEM-EDS) analysis was performed for the surface
2.1. Catalyst preparation
characterization of Cu–Zn/Al O3 catalysts prepared using the
2
All catalysts used for this research were prepared using the
sol–gel method. The system used was JEOL JSM-6060LA, and the
acceleration voltage was 5 kV. For easy detection of Cu and Zn, much
metal loading catalyst, Cu–Zn(25–15 wt.%)/Al O catalyst prepared
sol–gel method. For example, Cu–Zn/Al O3 was obtained by the
2
hydrolysis of the mixed solution with aluminum isopropoxide
2
3
(
AIP), Cu(NO ) , Zn(NO ) , water, and small amount of ethylene
using the sol–gel method was used for this analysis. After the
aforesaid preparation method and the calcination, this catalyst was
reduced and evacuated under the same conditions as mentioned
above. The reduced catalyst was kept at He gas atmosphere in a
closed Pyrex glass tube. Before SEM-EDS analysis, the glass tube
was broken, and the catalyst was kept in the tube over night. This
procedure is a sort of mild oxidation. If this reduced catalyst were
placed on a sheet of paper from the broken tube, the oxidation heat
would burn the paper, and the heat of the paper burning would
burn the catalyst. The catalysts prepared using the sol–gel method
are very air-sensitive after reduction.
3
2
3 2
glycol (EG). The amount of Cu(NO3)2 and/or Zn(NO3)2 depended
on the loading metal percent of the catalysts needed. AIP (purity
of Al[OCH(CH ) ] : 95%), EG (purity of HOCH CH OH: 99.5%,), cop-
3
2
3
2
2
per(II) nitrate trihydrate (purity of Cu(NO ) ·3H O: 99%), zinc(II)
3
2
2
nitrate hexahydrate (purity of Zn(NO ) ·6H O: 99%) used here
3
2
2
were manufactured by Wako Pure Chemical Industries, Ltd. (Wako).
AIP (∼10 g) was crushed using a mortar, and then dissolved and
◦
suspended in hot water (∼70 C, ∼200 ml), and a mixed aqueous
solution of Cu(NO ) and Zn(NO ) was added to the mixed hot
3
2
3 2
water. A small amount of EG (e.g., ∼5% of the total solution amount)
was also added as a sort of surfactant into the mixed hot water. This
◦
mixture was stirred and heated at ∼70 C for ∼30 min. Then, diluted
2.3. Apparatus and steam reforming of DME
nitric acid aqueous solution, made from HNO (69.5%, Wako), was
3
added every 15 min several times, and the pH of the mixture was
lowered with several additions until the pH decreased to 1–2.
Usually, all the steps of this process took ∼5 h. Regarding HNO3
addition, a clear-sol of boehmite was formed. Water in this sol was
evaporated under reduced pressure by using a rotary evaporator,
DME steam reforming was performed in a flow reactor (7.6 mm
i.d. Pyrex glass tube) using 0.10 g of the catalyst (d.m. <150 ␮m)
◦
at a temperature range from 200 to 500 C at atmospheric pres-
sure. Reactant (DME and water) flow with the Ar carrier gas was
adjusted using two mass flow controllers (Brooks 580E). DME was
diluted with Ar gas in a DME-Ar (5–95 vol.%) gas cylinder, and the
◦
and a gel was obtained. The obtained gel was dried at 170 C over
night. The dried gel was ground by using an agate mortar until the
−
1
diluted DME gas flow was 1.9 ml min . Water was supplied from
◦
diameter of each grain of powder was less than 150 ␮m. The pow-
a bubbler with water at 0 C (vapor pressure: 4.581 mmHg) by 50-
◦
−1
der was calcined at 500 C for 5 h, and ∼3-g alumina catalysts were
ml min
Ar gas flow. The reaction gas, a mixture of DME (flow
−
1
−1
−1 −1
usually obtained depending on the metal loading amount. Before
steam reforming of DME, the catalysts were reduced by flowing
of 2 mmol g−cat
h
) and water (flow of 6 mmol g−cat
h ), was
supplied to the catalyst layer. The DME concentration and the water
concentration of the inlet gas were 0.19% and 0.53%, respectively.
−
1
◦
H2 (99.99%, 10 ml min ) at 450 C for 10 h, and were evacuated at
3
◦
3
−1
,
00 C for 1 h, respectively. This treatment may sinter the metals of
The space velocity of all gases including Ar gas was 35 × 10 h

2
2
K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
-4.5
-5.0
-5.5
-6.0
-6.5
100
90
80
70
60
50
40
30
20
10
0
Cu amount of catalyst surface and Cu surface area of catalysts
were measured by N O titration method using BELCAT-B manufac-
2
tured by BEL Japan, Inc.
For the analysis of the reactants and products, two gas chro-
matographs (GCs) manufactured by Shimadzu Corporation were
used. One was a Shimadzu GC-6AM equipped with a thermal con-
ductivity detector (TCD), a methanizer (for CO analysis), and a flame
ionization detector (FID). The GC had an MS-5A stainless column
(
80–100 mesh, 5 m long, i.d. 3 mm) and its carrier gas was nitro-
gen. H , Ar (as the internal standard for GC analysis), CH , and CO
2
4
were quantitatively analyzed. The other was a Shimadzu GC-4C
with TCD and FID, and equipped with a Porapak Q stainless col-
umn (80–100 mesh, 1 m long, i.d. 3 mm) and a Porapak R stainless
column (80–100 mesh, 0.5 m long, i.d. 3 mm) in series. Its carrier gas
was helium. CH , CO , H O, DME, and methanol were quantitatively
4
2
2
analyzed.
3
. Results and discussion
3
.1. Steam reforming of DME over the two-layer catalyst, the
mixed catalyst, and the single-use catalyst prepared using the
sol–gel method
Fig. 1. DME steam reforming over Cu–Zn(4.5–0.5 wt.%)/Al2O3 catalyst prepared
using the sol–gel method.
Catalyst weight: 0.1 g; reaction gas flow: DME-water = 2–6 mmol g−cat
In accordance with the SRDME reaction mechanism (Eqs.
−
1
−1
h .
(
(
1)–(3)), ␥-A O (a DME hydrolysis catalyst), Cu–Zn(9–1 wt.%)/SiO
2
3
2
an SRM catalyst), and Cu–Zn(4.5–0.5 wt.%)/Al O3 catalysts pre-
2
pared by the sol–gel method were used for the three different
catalyst types. One was the two-layer type catalyst consist-
ing of the ␥-Al O3 catalyst (0.05 g) layer (the 1st layer) and
2
the Cu–Zn(9–1 wt.%)/SiO2 catalyst (0.05 g) layer (the 2nd layer).
Another was the physically mixed catalysts consisting of the
␥
-Al O catalyst (0.05 g) and the Cu–Zn(9–1 wt.%)/SiO2 cata-
2 3
lyst (0.05 g). The other was the single-use type catalyst, the
Cu–Zn(4.5–0.5 wt%)/Al O3 catalyst (0.1 g) alone, without mixing
2
of acid catalysts. These catalysts were used for SRDME. The phys-
ical data of the catalysts and the part of the results are in Table 1.
All catalysts produced H , CO , small amount of CO, and much
2
2
smaller amount of CH4 and methanol. Example of the production
details is shown Fig. 1. The order of the H2 production rate and H2
yield over the catalysts were as follows: single-use type catalyst
of an SRDME catalyst prepared using the sol–gel method
(
Cu–Zn(4.5–0.5 wt.%)/Al O ) >physically mixed catalyst consisting
2 3
of a DME hydrolysis catalyst and an SRM catalyst (␥-Al O3 and
2
Cu–Zn/SiO2 physically mixed) »two-layer type catalyst consisting
of a DME hydrolysis catalyst and an SRM catalyst (␥-Al O3 and
2
Cu–Zn/SiO2 used separately) as shown in Fig. 2 and Table 1. The
order by the DME conversion is also the same order. From the pro-
duction rate of H2 and the Cu amount of catalyst surface, turn over
frequency (TOF) is estimated, and the data of the three catalysts are
shown in Table 1. In the cases of TOF, H2 production rate per Cu sur-
Fig. 2. H2 production rate and H2 yield of DME steam reforming over a single-
use type catalyst of an SRDME catalyst prepared using the sol–gel method
face area, and H production rate per catalyst surface area, the each
2
(
Cu–Zn(4.5–0.5 wt.%)/Al2O3 (0.1 g)) (ᮀ),
a physically mixed catalyst consist-
order is also as follows: single-use type catalyst of an SRDME cat-
ing of DME hydrolysis catalyst and an SRM catalyst (␥-Al2O3 (0.05 g) and
a
alyst (Cu–Zn(4.5–0.5 wt.%)/Al O3 (Sol)) »physically mixed catalyst
2
Cu–Zn(9–1 wt.%)/SiO2 (0.05 g) physically mixed) ( ), and a two-layer type catalyst
consisting of a DME hydrolysis catalyst and an SRM catalyst (␥-Al2O3 (0.05 g) and
Cu–Zn(9–1 wt.%)/SiO2 (0.05 g) used separately) (ꢀ).
consisting of a DME hydrolysis catalyst (␥-Al O ) and an SRM
2
3
catalyst (Cu–Zn/SiO ) »two layer catalyst consisting of a DME
2
−
1
−1
h .
Reaction gas flow: DME-water = 2–6 mmol g−cat
hydrolysis catalyst (␥-Al O ) and an SRM catalyst (Cu–Zn/SiO ).
2
3
2
The Cu–Zn(4.5–0.5 wt.%)/Al O catalyst was used singly and H was
2
3
2
3
−1
.
and that of the reactant gas (DME and water) was 0.27 × 10 h
produced well, and from the catalyst contents; each particle of the
Cu–Zn/Al2O3 catalyst has both of alumina sites for DME hydrolysis
and copper sites for SRM. These results and the order suggest that
two sites for DME hydrolysis and SRM should be the closer for the
higher TOF, the faster H2 production rate, and the better H2 yield by
SRDME. The sol–gel method is very optimum catalyst preparation
method for these effects.
The contact time was 0.10 s.
The reactor was part of a closed circulation system. After the
above-mentioned reduction and evacuation in the rector, and
before the reaction, the Brunaure–Emmett–Teller (BET) specific
surface area of the catalyst in the reactor (in situ) without exposure
◦
to air was measured using N2 gas at −196 C.

K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
23
Fig. 3. SEM image of Cu–Zn/Al2O3 catalyst prepared using the sol–gel method.
(a) Image by low magnification.
(b) Image by high magnification.
3
.2. Surface analysis of Cu–Zn/Al O3 catalyst prepared using the
in Fig. 3. The catalyst was ground by using an agate mortar until
the diameter of each grain of powder was less than 150 ␮m. From
the SEM photos the catalyst is ground with cleavage into several
sizes such as 0.5-120 ␮m. On the large size catalysts, there are very
small size catalysts. Each catalyst shows knobs and hills. Some pho-
tos of the SEM-ESD analysis are shown in Fig. 4. As shown in the
photo of the Cu analysis (b), red, yellow, and green mean detec-
tion of Cu, so such knobs and hills are Cu. As shown in the photo
of the Al analysis (c), red, yellow, and green mean detection of Al,
so Al (Al O ) exists on the whole of the catalyst surface. The each
2
sol–gel method
Cu–Zn/Al O catalysts prepared by the sol–gel method singly
2
3
use produce hydrogen more effectively than physically mixed cat-
alysts with DME hydrolysis catalysts and SRM catalysts. This is
probably the reason that Al O3 for DME hydrolysis to methanol
and Cu for SRM to H2 and CO2 coexist closely and are well dis-
persed on the catalyst surface. In the case of mixed catalysts, each
particle of the catalysts has each reaction sites of DME hydrolysis
and SMR, respectively. This means that there is a greater distance
2
2
3
color, red, yellow, green, and blue part in the Cu analysis (b) pattern
between in the Al analysis (c) are almost opposite, so it is specu-
between the DME hydrolysis sites (Al O3 sites) and the SRM sites
2
(
Cu sites) than that between the two sites on the Cu–Zn/Al O3
lated that Cu is dotted within Al O . The all clusters of Cu are putting
2
2 3
catalyst prepared using the sol–gel method. In order to confirm
this speculation, SEM-EDS analysis was performed on the surface
of the Cu–Zn/Al O catalysts. Some of the SEM photos are shown
in/on Al O ; it means that Cu atoms which are the active sites for
2 3
methanol steam reforming and Al O molecules that are the active
2
3
sites for DME hydrolysis are contacting and dispersed well. Zn was
2
3
Fig. 4. SEM-EDS image of Cu–Zn/Al2O3 catalyst prepared using the sol–gel method.
(
(
(
a) SEM light image. The same part of the EDS analyses.
b) EDS Cu analysis.
c) EDS Al analysis. (For interpretation of the references to colour in the text, the reader is referred to the web version of this article.)

2
4
K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
Fig. 5. Comparison of reaction scheme of DME steam reforming over the mixed catalysts and the singly use Cu/Al2O3 catalyst prepared using the sol–gel method.
not detected by the EDS, although a much metal loading catalyst,
Cu–Zn(25–15 wt.%)/Al O prepared using the sol–gel method, was
2
3
used for the easy detection of Cu and Zn. Zn may be dispersed well
in and/or on the catalyst, while Cu makes sort of clusters and/or
small aggregates. If Al O is compared to seawater, Cu is observ-
2
3
able as Cu is exposed from sea level (Al O3 surface) like an iceberg
2
and is dotted on the surface of the sea. Al O and Cu coexist and
2
3
are distributed with sufficient balance as predicted. It is considered
that DME hydrolysis into methanol occurs on Al O3 and the SRM
2
to H2 and CO2 immediately takes place on Cu very near Al O . In
2
3
the case of mixed catalysts, there are some distance between the
DME hydrolysis catalysts and the SRM catalysts, so it is difficult to
conduct these two reactions immediately. Fig. 5 shows the above-
mentioned concepts. It is concluded that the Cu–Zn/Al O3 catalyst
2
prepared by the sol–gel method in single use has these two reac-
tions occurring sequentially and efficiently, and a quick hydrogen
production rate is obtained by the sol–gel Cu–Zn/Al O catalyst.
2
3
The sol–gel method is very optimum catalyst preparation method
for this reaction system of DMESR, and introduces the faster H2
production rate and higher H2 yield of SRDME.
Fig. 6. H2 yield and CO concentration in the products of DME steam reforming at
◦
2
50 C over some Cu/Al2O3 catalysts prepared using the sol–gel method.
−
1
−1
h .
Catalyst weight: 0.1 g; reaction gas flow: DME-water = 2–6 mmol g−cat
3.3. Effect of Cu loading amount for SRDME
3.4. Effect of Zn addition amount for SRDME
Sol-gel preparation methods are usually used for a low load-
ing metal amount such as less than 10 wt.%. However, the activity
of SRM is related to the surface area of Cu, and previous results
show that the rate of H2 production increases with an increase in
the Cu loading percent on the copper catalysts prepared using the
sol–gel method [24–26]. Based on the SRM results, the dependence
of the Cu loading percent for SRDME was examined. The results are
shown in Fig. 6. It is clear that 30 wt.% is the most suitable load-
ing percent of Cu for H2 production by SRDME. It is speculated
The effect of Zn addition into Cu/Al O (70 wt.%) catalysts was
2
3
examined. Zn addition into Cu/Al O3 and Cu/SiO2 catalysts pre-
2
pared using the sol–gel method catalysts promotes hydrogen
production by SRM [4,24–28]. This time Zn was added into Cu/Al O3
2
catalysts prepared using the sol–gel method for hydrogen produc-
tion by SRDME. From the result of Fig. 6, the percent of alumina
in the catalyst was settled at 70 wt.%, and the total amount of Cu
and Zn metals was settled at 30 wt.%. Fig. 7 shows the relation
between loading Zn percent and the H2 yield, or the CO concen-
that 70 wt.%Al O -composition has the most comfortable surface
2
3
percent of Al O for DME hydrolysis, and the 30 wt.%Cu-loading
2
3
tration in the total amount of products. The H yield could be made
2
amount is the most suitable surface percent of Cu for SRM, because
the rate of H2 production is the fastest among the different loaded
Cu alumina catalysts.
higher by adding Zn, the most effective amount was 1 wt.%. Zn is
a promoter of SRM (Eq. (3)). ZnO may improve the dispersion and
the reducibility of Cu species [27–29]. SRDME (Eq. (1)) consists
2+
CO that is harmful to fuel cells was also produced with the high-
of DME hydrolysis (Eq. (2)) and SRM. DME hydrolysis is the rate-
determining step. Produced methanol by the DME hydrolysis will
be used for methanol to hydrogen by SRM, DME hydrolysis equi-
librium (CH OCH + H O→2CH OH (Eq. (2)) will be broken to right
est percent over the Cu(30 wt.%)/Al O catalyst. It is considered that
2
3
2
CO is formed by the reverse water-gas-shift reaction of H and CO2
produced from SRDME.
3
3
2
3

K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
25
Fig. 7. H2 yield and CO concentration in the all products of DME steam reforming at
◦
2
75 C over some Cu–Zn(Total 30 wt.%)/Al2O3 catalysts prepared using the sol–gel
method.
Catalyst weight: 0.1 g; reaction gas flow: DME-water = 2–6 mmol g−cat h−1.
Fig. 8. DME steam reforming over Cu–Zn(29–1 wt.%)/Al2O3 catalyst prepared using
the sol–gel method.
Catalyst weight: 0.1 g; reaction gas flow: DME-water = 2–6 mmol g-cat
−
1
−1
h .
side by SRM. Therefore, SRM-promoter, Zn, will enhance the SRDME
and H2 production with near to 100% of DME conversion and H2
yield.
The addition of Zn works for enhancement of H production, but
2
it didn’t significantly work to restrain CO that is harmful to fuel cells.
The other product was methane, but this amount was very low.
Details of the products are in Fig. 8. The Cu–Zn(29–1 wt.%)/Al O
2
3
−
1
−1
◦
catalyst produced H at a rate of 11.5 mmol g−cat
h
at 275 C, a
2
very lower temperature than the temperature of methane and LPG
−
1
−1
introduces a value of H2
reforming. A value of 11.5 mmol g−cat
h
yield (95%), because the theoretical maximum amount of H2 pro-
−
1
duction from the reactant flow (DME-water = 2–6 mmol g−cat h
)
−
1
−1
is 12 mmol g−cat
h . The DME conversion was 95%, and the con-
centration of CO was restrained at 0.8 mol.%. Compared between
Fig. 8 and Fig. 1, it is clear that SRDME catalysts are improved very
well by these optimizations of Cu and Zn composition.
We have already reported about the DME steam reforming and
the Cu–Zn/Al O3 catalysts prepared using the sol–gel method in
2
the previous paper [4]. This time, Cu–Zn/Al O3 catalysts prepared
2

2
6
K. Takeishi, Y. Akaike / Applied Catalysis A: General 510 (2016) 20–26
using the sol–gel method and the ratio of Cu/Zn have optimized
for the higher production of H2 by SEDME. The H2 yield increased
[4] K. Takeishi, H. Suzuki, Appl. Catal. A: Gen. 260 (2004) 111–117.
[
[
[
[
[
5] K. Takeishi, H. Suzuki, Y. Yamamoto, N. Konagai, F. Kimata, Japan. Kokai
Tokkyo Koho (2002), JP 2002263504; Japan Patent No. 4106645 (2008).
6] V.V. Galvita, G.L. Semin, V.D. Belyaev, T.M. Yurieva, V.A. Sobyanin, Appl. Catal.
A: Gen. 216 (2001) 85–90.
7] T. Matsumoto, T. Nishiguchi, H. Kania, K. Utani, Y. Matsumura, S. Imamura,
Appl. Catal. A: Gen. 276 (2004) 267–273.
◦
◦
◦
from 94% at 350 C and 53% at 250 C to 95% at 275 C and 68% at
◦
2
50 C. We have developed the catalyst for hydrogen production
by SRDME at the lower reforming temperature using the sol–gel
method.
8] Y. Tanaka, R. Kikuchi, T. Takeguchi, K. Eguchi, Appl. Catal. B: Environ. 57
(
2005).
9] T. Mathew, Y. Yamada, A. Ueda, H. Shioyama, T. Kobayashi, Appl. Catal. A: Gen.
86 (2005) 11–22.
4
. Conclusions
2
[
10] T. Mathew, Y. Yamada, A. Ueda, H. Shioyama, T. Kobayashi, Appl. Catal. A: Gen.
286 (2005) 11–22.
It is found that the sol–gel method is very optimum catalyst
preparation method for DME steam reforming, because the sol–gel
preparation method makes two sites for DME hydrolysis and SRM
very close on the catalyst surface, and also because the method
disperse the pairs of the two sites well. The optimization of the
content ratio of Cu and Zn improves the catalyst activity compared
with the previous results.
[11] T.A. Semelsberger, K.C. Ott, R.L. Borup, H.L. Greene, Appl. Catal. B: Environ. 65
2006) 291–300.
(
[
12] T. Kawabata, H. Matsuoka, T. Shishido, D. Li, Y. Tian, T. Sano, K. Takehira, Appl.
Catal. A: Gen. 308 (2006) 82–90.
[13] T.A. Semelsberger, K.C. Ott, R.L. Borup, H.L. Greene, Appl. Catal. A: Gen. 309
2006) 210–223.
(
[
14] K. Faungnawakij, Y. Tanaka, N. Shimoda, T. Fukunaga, R. Kikuchi, K. Eguchi,
Appl. Catal. B: Environ. 74 (2007) 144–151.
Cu–Zn/Al O catalysts prepared using the sol–gel method singly
[15] S.D. Badmaev, G.G. Volkova, V.D. Belyaev, V.A. Sobyanin, React. Kinet. Catal.
Lett. 90 (2007) 205–211.
2
3
use produce hydrogen more effectively at a lower reaction tem-
perature than physically mixed catalysts with DME hydrolysis
catalysts and SRM catalysts. SEM-EDS analysis suggests that this
is the reason that Al O for DME hydrolysis to methanol and Cu for
[
[
16] K. Faungnawakij, T. Fukunaga, R. Kikuchi, K. Eguchi, J. Catal. 256 (2008) 37–44.
17] T. Fukunaga, N. Ryumon, S. Shimazu, Appl. Catal. A: Gen. 348 (2008) 193–200.
[18] N. Shimoda, K. Faungnawakij, R. Kikuchi, T. Fukunaga, K. Eguchi, Appl. Catal.
A: Gen. 365 (2009) 71–78.
2
3
[
19] N. Shimoda, K. Faungnawakij, R. Kikuchi, K. Eguchi, Appl. Catal. A: Gen. 378
(2010) 234–242.
SRM to H2 and CO2 coexist closely and are well dispersed on the
catalyst surface. In the case of mixed catalysts, there is a longer dis-
tance between the DME hydrolysis catalysts and SRM catalysts than
that of the two sites on the Cu–Zn/Al O catalyst prepared using the
[20] X. Wang, X. Pan, R. Lin, S. Kou, W. Zou, J.-X. Ma, Int. J. Hydrogen Energy 35
2010) 4060–4068.
21] X. Long, Q. Zhang, Z.-T. Liu, P. Qi, J. Lu, Z.-W. Liu, Appl. Catal. B: Environ.
34–135 (2013) 381–388.
(
[
2
3
1
sol–gel method. Therefore, two reactions of DME hydrolysis over
alumina sites and SRM over Cu sites occur more sequentially in the
case of sol–gel alumina catalysts than in the case of mixed catalysts
of DME hydrolysis catalysts and SRM catalysts.
[22] G. Volkova, S. Badmaev, V. Belyaev, L. Plyasova, A. Budneva, E. Paukshtis, V.
Zaikovsky, V. Sobyanin, Stud. Surf. Sci. Catal. 167 (2007) 445–450.
[
23] 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.
[24] K. Takeishi, H. Suzuki, A. Ueno, 79th Spring Meet. Chem. Soc. Jpn. (2001) p382,
Abstracts.
The addition of Zn into Cu/Al O (70 wt.%) enhances the SRDME
2
3
[
25] K. Takeishi, H. Suzuki, A. Ueno, 80th Fall Meet. Chem. Soc. Jpn. (2001) p216,
Abstracts.
activities. The Cu–Zn(29–1 wt.%)/Al O catalyst show excellent
2
3
◦
activity of SRDME at the lower reforming temperature of 275 C;
the DME conversion was 95%; the H2 yield was 95%; and the CO
concentration was 0.8 mol.%.
[
26] K. Takeishi, H. Suzuki, A. Ueno, 81st Spring Meet. Chem. Soc. Jpn. (2002) p577,
Abstracts.
[
[
[
27] Y. Matsumura, H. Ishibe, Appl. Catal. B: Environ. 86 (2009) 114–120.
28] Y. Matsumura, H. Ishibe, J. Catal. 268 (2009) 282–289.
29] M.M. Günter, T. Ressler, R.E. Jentoft, B. Bems, J. Catal. 203 (2001) 133–149.
References
[1] T.H. Fleisch, A. Basu, M.J. Gradassi, J.G. Masin, Stud. Surf. Sci. Catal. 107 (2000)
1
17–125.
[
[
2] T.H. Fleisch, A. Basu, R.A. Sills, J. Nat. Gas Sci. Eng. 9 (2012) 94–107.
3] O. Okada, T. Takemoto, K. Tsukahara, H. Hashimoto, N. Matsuda, S. Suzuki, J.
Jpn. Inst. Energy 84 (2005) 331–334.