Effects of Cu-ZnO content on reaction rate for direct synthesis of DME from syngas with bifunctional Cu-ZnO/γ-Al2O3 catalyst

Article abstract of DOI:10.1007/s10562-013-1022-6

The effects of Cu-ZnO content on the performance of bifunctional Cu-ZnO/γ-Al₂O₃ catalysts for dimethyl ether (DME) synthesis from syngas were investigated by varying the weight ratios of Cu-ZnO/γ-Al₂O₃ prepared by the coprecipitation of Cu-ZnO in a slurry of γ-Al₂O₃. A higher rate of DME production with CO conversion of 47.6 % and DME selectivity of 61.1 % was observed with the bifunctional catalyst at an optimal weight ratio of CuO/γ-Al₂O₃ of two, providing a higher surface area of metallic copper and an abundance of weak acid sites. The number of acidic sites on solid-acid γ-Al₂O₃ is a more crucial factor to enhance DME yield, due to the faster dehydration rate of methanol to DME than that of CO hydrogenation to methanol. Although the first step of methanol synthesis on active copper sites is a rate-limiting step with a low equilibrium value, the second step of the dehydration of methanol to DME on acid sites adjusts the overall rate by enhancing the forward reaction rate of CO hydrogenation to methanol with a simultaneous formation of surplus hydrogen by a water-gas shift reaction. Therefore, the proper design of a high surface area of metallic copper with larger acid sites on the bifunctional CuO-ZnO/γ-Al₂O₃ catalysts at an optimal ratio, produced by adjusting the weight ratio of CuO/γ-Al₂O ₃, is an important factor for improved catalytic performance.

Full text of DOI:10.1007/s10562-013-1022-6

Catal Lett (2013) 143:666–672
DOI 10.1007/s10562-013-1022-6
Effects of Cu–ZnO Content on Reaction Rate for Direct Synthesis
of DME from Syngas with Bifunctional Cu–ZnO/c-Al O Catalyst
2
3
Ji Woo Jeong
Soong Ho Um
•
Chang-Il Ahn
Jong Wook Bae
•
Dong Hyun Lee
•
•
Received: 8 January 2013 / Accepted: 7 May 2013 / Published online: 15 May 2013
Ó Springer Science+Business Media New York 2013
Abstract The effects of Cu–ZnO content on the perfor-
mance of bifunctional Cu–ZnO/c-Al O3 catalysts for
dimethyl ether (DME) synthesis from syngas were inves-
1 Introduction
2
The direct synthesis of dimethyl ether (DME) and higher
alcohols from syngas has been largely investigated as a
promising alternative method to produce renewable clean
energy [1]. Renewable energy resources like DME, derived
from biomass and natural gas, have a high potential to
tigated by varying the weight ratios of Cu–ZnO/c-Al O
2
3
prepared by the coprecipitation of Cu–ZnO in a slurry of
c-Al O . A higher rate of DME production with CO con-
2
3
version of 47.6 % and DME selectivity of 61.1 % was
observed with the bifunctional catalyst at an optimal
weight ratio of CuO/c-Al O of two, providing a higher
suppress the emission of CO and air pollutants [2]. The
2
commercialized process for DME production generally
involves two steps, methanol synthesis by hydrogenation
2
3
surface area of metallic copper and an abundance of weak
acid sites. The number of acidic sites on solid-acid c-Al O3
of CO , followed by consecutive DME synthesis by the
x
2
is a more crucial factor to enhance DME yield, due to the
faster dehydration rate of methanol to DME than that of
CO hydrogenation to methanol. Although the first step of
methanol synthesis on active copper sites is a rate-limiting
step with a low equilibrium value, the second step of the
dehydration of methanol to DME on acid sites adjusts the
overall rate by enhancing the forward reaction rate of CO
hydrogenation to methanol with a simultaneous formation
of surplus hydrogen by a water–gas shift reaction. There-
fore, the proper design of a high surface area of metallic
copper with larger acid sites on the bifunctional CuO–ZnO/
c-Al O catalysts at an optimal ratio, produced by adjust-
dehydration of methanol. However, great interest has been
focused on developing a single-step synthesis of DME
using a properly designed bifunctional catalyst that con-
tains dual functionalities for hydrogenation and dehydra-
tion [3, 4]. The direct synthesis of DME has the advantage
of a high equilibrium conversion of CO by increasing the
forward reaction of CO hydrogenation with surplus
hydrogen formation through a water–gas shift (WGS)
reaction. In addition, a single-step reaction is also suitable
to use for syngas having a lower H /CO ratio, such as that
2
frequently derived from biomass or coal gasification [4–7].
The reaction rate of DME synthesis by methanol dehy-
dration is much faster than that of methanol synthesis by
CO hydrogenation, with a higher equilibrium conversion
2
3
ing the weight ratio of CuO/c-Al O , is an important factor
2
3
for improved catalytic performance.
[
5]. The reactions involved and the heats of formation at
Keywords Dimethyl ether ꢀ Syngas ꢀ
Copper surface area ꢀ Acidic site ꢀ
298 K for the direct synthesis of DME from syngas can be
categorized as following reaction Eqs. (1) through (4), and
the overall reaction can be expressed as Eq. (5):
CuO–ZnO/c-Al O ꢀ Bifunctional catalyst
2
3
CO hydrogenation:COþ2H $ CH OH,
2
3
ð1Þ
ð2Þ
o
DH ¼ ꢁ90:8 kJ/mol
J. W. Jeong ꢀ C.-I. Ahn ꢀ D. H. Lee ꢀ S. H. Um ꢀ J. W. Bae (&)
School of Chemical Engineering, Sungkyunkwan University
(SKKU), Suwon, Kyonggi-do 440-746, Republic of Korea
e-mail: finejw@skku.edu
CO2 hydrogenation: CO þ 3H2 $ CH3OH þ H2O,
2
o
DH ¼ ꢁ49:6 kJ/mol
1
23

Effects of Cu–ZnO Content on Reaction Rate
667
Dehydration to DME: 2CH OH $ CH OCH þ H O,
number of acidic sites rather than a higher metallic copper
surface area, since the rate of methanol dehydration to
DME is faster than that of CO hydrogenation to methanol
3
3
3
2
o
DH ¼ ꢁ23:4 kJ/mol
ð3Þ
ð4Þ
ð5Þ
WGS reaction: H2O þ CO $ H2 þ CO2;
[
5, 13, 16].
o
DH ¼ ꢁ41:2 kJ=mol
In the present study, we investigated the effects of
Cu–ZnO content on the catalytic performance of Cu–ZnO/
Overall reaction: 3CO þ 3H2 $ CH3OCH3 þ CO2;
c-Al O bifunctional catalysts for the single-step synthesis
2
3
o
DH ¼ ꢁ246:0 kJ=mol
of DME from syngas, varying the copper surface areas and
the number of acidic sites using coprecipitation of Cu–ZnO
in a slurry of the solid acid component of c-Al O . The
The catalytic performances of the bifunctional catalyst
are generally explained by characterizing the dispersion of
copper crystallites, acidic properties of solid-acid catalyst,
and adsorption behavior of reactants [6–10]. Although the
reaction rates for the direct conversion of syngas to DME
could be controlled kinetically at low temperatures, the rates
tend to approach equilibrium values at high temperatures,
and CO conversions generally decrease with the increase in
temperature [4]. In addition, the hydrogenation of CO to
methanol is more thermodynamically favorable than that of
CO , and an increased CO concentration in syngas
2
3
catalytic activities were significantly affected by simulta-
neously varying the copper surface area and acid sites, as
mentioned in our previous work on the Cu–ZnO–Al O /Zr-
2
3
modified ferrierite catalytic system [14], and the role of
acidic sites was more dominant in the bifunctional catalysts
for a higher rate of DME synthesis. The optimum ratio of
active metals and solid acid components also depends on
the acid strength of the solid-acid catalysts, and a much
lower fraction of zeolite to active Cu–ZnO metals has been
reported as an optimal composition for a higher DME yield
on bifunctional catalysts compared with the solid-acid
catalyst of Al O [17]. The correlations of copper surface
2
2
generally decreases CO conversion at a fixed H /CO ratio
2
according to Le Chatelier’s principle, as shown in reaction
equation (5) [4]. Therefore, to enhance the rate of DME
production from syngas, it is crucial to design a proper
bifunctional catalyst that possesses a higher copper surface
area and a large number of acidic sites by simultaneously
modifying the strong acidic sites on the solid-acid
component [11, 12]. Many researchers have reported that
CO conversion is linearly correlated with metallic copper
surface areas on the admixed bifunctional catalyst, since the
CO hydrogenation reaction can be a rate-limiting step in the
consecutive DME synthesis reaction from syngas [13].
However, in our recent research [14], the intrinsic catalytic
activity on Cu–ZnO-based bifunctional catalysts showed a
2
3
area and number of acid sites to catalyst performance using
a simple Cu–ZnO/c-Al O bifunctional catalytic system for
2
3
20 h were characterized by X-ray diffraction (XRD) anal-
ysis, N O titration, and the temperature-programmed
2
desorption of NH (NH -TPD) on bifunctional Cu–ZnO/
3
3
c-Al O catalysts.
2
3
2 Catalyst Preparation, Characterization and
Activity Tests
good correlation between the copper surface area (S ) and
Cu
The solid-acid catalyst of c-Al O for the methanol dehy-
2 3
the density of acidic sites (DAS), and the activity was
maintained at a constant value above optimum values of
dration reaction to DME has a surface area of around
2
230 m /g with an average pore diameter of 19 nm and was
(
SCu 9 DAS). This also suggests that the direct synthesis of
supplied by Saint-Gobain. The bifunctional catalysts were
prepared by the co-precipitation method in a slurry of
c-Al O at 70 °C using metal precursors of copper acetate
DME from syngas is a structurally insensitive reaction.
However, the effects of copper content on the catalytic
activity of Cu–ZnO/c-Al O bifunctional catalysts have not
2
3
2
3
and zinc acetate with Na CO precipitant at a fixed molar
2
3
been well explained in terms of copper surface area or
number of acidic sites. Specifically, the number of acidic
sites largely influences the overall reaction rate of CO
hydrogenation by promoting a faster dehydration rate of
methanol to DME, and the abundant presence of methanol
on the bifunctional catalyst surfaces could also enhance
the overall reaction rate due to the lowest activation bar-
rier for the pathway involving the adsorption of two
methanol molecules, as reported by Blaszkowski and van
Santen [15]. Therefore, the higher overall reaction rate of
DME synthesis from syngas on bifunctional catalysts
can be mainly attributed to the presence of a larger
ratio of CuO/ZnO = 3/1 and a final solution pH around 7.
The copper surface area and number of acid sites were
controlled by varying the CuO/c-Al O weight ratio from
2
3
1.13 to 6.75. The precipitate was aged for 3 h at 70 °C
followed by calcination at 300 °C for 5 h. The bifunctional
Cu–ZnO/c-Al O catalysts are denoted as CZA(X), where
2
3
C, Z, and A represent CuO, ZnO, and solid-acid c-Al O ,
2
3
respectively, and X denotes the weight ratio of CuO/c-
Al O such as CZA(1), CZA(2), CZA(4), and CZA(7) for
2
3
1.13, 1.75, 4.25, and 6.75, respectively. The catalytic
properties of the bifunctional catalysts are easily adjusted
by changing a weight ratio of CuO-ZnO/c-Al O , and the
2
3
123

6
68
J. W. Jeong et al.
coprecipitation of active metals in a slurry of the solid-acid
area of metallic copper on the bifunctional CZA catalysts
after reaction for 20 h was calculated by assuming
component of c-Al O is the more effective method com-
2
3
1
9
1.46 9 10 Cu atoms/m with a molar ratio of 0.5 for
2
pared with the physical mixing of two active catalysts, due
to the easy control of strong acidic sites through a depo-
sition–precipitation method [7, 10].
N O/Cu (where Cu is a Cu atom on the surface) [18].
2
s
s
The crystallite size of the copper species before and
The catalytic performance was tested in a fixed bed
tubular reactor with an outer diameter of 12.7 mm using a
catalyst loading of 0.2 g. Prior to reaction, the bifunctional
after reaction was also characterized by powder XRD
analysis using a Rigaku diffractometer with CuKa radia-
tion in order to identify the phases of metallic Cu, CuO,
ZnO, and c-Al O . The average crystallite size of the
catalysts were reduced in a flow of 5 vol% H balanced
2
2 3
with N at 300 °C for 5 h. The syngas was composed of an
2
copper species was calculated from the values of the full
width at half maximum (FWHM) of the XRD diffraction
peaks at 2h = 35.6° for CuO for fresh catalysts and
2h = 43.3° for the metallic copper (Cu ) on the used
catalysts.
2
H /CO molar ratio of 2.0 with an internal standard gas of
5
.6 mol% N2 based on total syngas. The reaction was
0
carried out for around 20 h on stream with the following
reaction conditions: T = 270 °C, P = 3.5 MPa, and space
velocity = 2,000 ml/gcat /h. The CO conversion and
product distribution were obtained from the steady-state
average values over 5 h after 15 h of reaction. The prod-
ucts were analyzed using an online gas chromatograph
3 Results and Discussion
(
Younglin GC, YL6100) using a Porapack-Q column
3.1 Catalytic Performance and Characteristics
of Bifunctional CZA Catalysts
connected to a thermal conductivity detector (TCD) to
analyze N , H , CO, and CO and a GS-Q column con-
2
2
2
nected to a flame-ionized detector to analyze methanol,
DME, and hydrocarbon byproducts.
The results of the catalytic performances at steady state are
summarized in Table 1. CO conversion and DME selec-
tivity were found to be highest for CZA(2), with values of
47.6 and 61.1 mol%, respectively. The CO conversion was
somewhat lower than the calculated equilibrium CO con-
version of around 75 % at 270 °C and 3.5 MPa, which was
calculated using the HSC Chemistry 7.0 simulator. The
lowest CO conversion and DME selectivity were observed
for CZA(7), which contained a higher CuO/c-Al O ratio,
The bifunctional CZA catalysts were characterized by
temperature-programmed reduction (TPR) experiments.
Each catalyst was pretreated with a He flow up to 200 °C
for 1 h to remove the adsorbed water and other contami-
nants, followed by cooling to 50 °C. A 5 vol% H /He
2
mixture was introduced in the BELCAT instrument at a
flow rate of 30 ml/min with a heating rate of 10 °C/min up
to 500 °C. The effluent gas was passed through a molecular
sieve to remove water formed during the TPR experiment,
and it was analyzed by TCD. The temperature-programmed
2
3
with values of 19.0 and 24.7 mol%, respectively. The
overall reaction rates (reacted CO mol/g /s) were found to
cat
be in the range of 0.57–1.42. The reaction rate was
desorption of ammonia (NH -TPD) was also carried out on
3
increased with an increase of CuO–ZnO/c-Al O up to
2 3
fresh bifunctional CZA catalysts to determine the surface
acidity. Around 0.1 g of catalyst was initially flushed with
a He flow at 250 °C for 2 h, then cooled to 100 °C and
saturated with NH . After the NH exposure, the catalyst
two, and it was decreased to 0.57 for CZA(7). Therefore,
the overall reaction rate and yield of DME were maxi-
mized for CZA(2), and the methanol selectivity was
inversely proportional to CO conversion on the bifunc-
3
3
was purged under a He flow until it reached an equilibrium
point and was then subjected to a TPD experiment in the
temperature range of 100–600 °C with a heating rate of
0 °C/min using the same BELCAT instrument.
The Brunauer–Emmett–Teller (BET) surface area of the
tional CZA catalysts. CO selectivity was also proportional
2
to the extent of CO conversion due to enhanced WGS
reaction activity. The simultaneously formed surplus
hydrogen on the CZA catalysts can enhance the reaction
1
rate and CO selectivity together, and the byproduct for-
2
bifunctional CZA catalysts was measured by the nitrogen
adsorption method at -196 °C using a constant-volume
adsorption apparatus (Micromeritics, ASAP-2400). The
surface area of the metallic copper was measured by the
N O surface titration method. Prior to N O titration, the
mation was related to the CO conversion by the possible
reforming reaction of the products [6, 10]. Although the
performance of the CZA catalysts seems to have no sig-
nificant relationship with copper content, the CO conver-
sion and DME yield on the CZA catalysts may be strongly
related to two factors, the surface area of metallic copper
for CO hydrogenation and the number of acidic sites for
the dehydration of methanol in a consecutive reaction of
direct DME synthesis from syngas through the intermedi-
ate formation of methanol.
2
2
catalyst was reduced at 300 °C for 5 h with 5 vol% H /N
2
2
mixed gases at a flow rate of 30 ml/min, and consumption
of N O with a concomitant release of N onto the metallic
2
2
copper sites (N2O þ 2Cu ¼ Cu2O þ N2) was analyzed by a
BELCAT instrument equipped with a TCD. The surface
1
23

Effects of Cu–ZnO Content on Reaction Rate
669
Table 1 Catalytic performances over the bifunctional Cu–ZnO/c-Al
O
2 3
catalysts
a
b
Reaction rate /10
-2
Notation
CO conv.
mol%)
Selectivities (mol%)
Yield of DME
(
c
Methanol
DME
CO
2
BP
CZA(1)
CZA(2)
CZA(4)
CZA(7)
a
42.7
47.6
28.4
19.0
1.27
1.42
0.85
0.57
28.0
17.3
27.3
67.6
56.2
61.1
53.3
24.7
12.2
17.4
14.8
5.3
3.6
4.1
4.6
2.4
24.0
29.1
15.1
4.7
The bifunctional catalysts are denoted as CZA(X), where C, Z, and A represents CuO, ZnO, and solid-acid c-Al
ratio of CuO/c-Al at a fixed CuO/ZnO molar ratio of three
The reaction rates of DME synthesis from syngas are defined as reacted CO mol/gcat/s, which are the average values at steady-state
2
O
3
, and X denotes a weight
2 3
O
b
c
4 2
Byproducts (BP) mainly include CH and a trivial amount of C hydrocarbons
Table 2 The characteristics of the bifunctional Cu–ZnO/c-Al
2
O
3
catalysts
Notation Weight ratio of (CuO/ Surface XRD (crystallite
c-Al
Acid sites (D(AS))
a
Cu surface area (S(Cu)
b
2
2
O
3
)
area
(
size, nm)
(mmol NH /g)
3
(m /gCu
)
2
m /g)
CuO
(
Metallic Peak T
fresh) Cu
(
1
Peak T
(300–500 °C)
2
Total
2
(T ? T )
(\300 °C)
1
used)
CZA(1) 1.13
CZA(2) 1.75
CZA(4) 4.52
CZA(7) 6.75
a
95.3
91.8
88.1
85.7
11.8
10.6
11.7
14.5
32.3
24.2
27.8
36.3
0.514
0.468
0.355
0.321
0.428
1.038
1.011
0.437
0.942
1.506
1.466
0.758
3.47
6.25
5.48
5.90
The acidic site density (D(AS) was calculated from the desorption peak areas for NH
) and 300–500 °C (T ), respectively
The copper surface area (defined as SCu = exposed metallic copper area (m /g) per gram of catalyst) of S(Cu)) after reaction was measured by
3
-TPD, and they are assigned at the regions of 100–300 °C
(
b
T
1
2
2
2
N O titration method
As shown in Table 2, the surface areas of the bifunc-
2
tional CZA catalysts were in the range of 85–96 m /g, and
they decreased slightly with the increase in the CuO/
c-Al O weight ratio due to the possible blockage of
2
3
mesopores of the solid-acid c-Al O by the deposition of
2
3
CuO–ZnO crystallites on surfaces. As shown in Fig. 1,
TPR profiles of the CZA catalysts exhibited only one broad
reduction peak at a maximum reduction peak of around
1
2
65 °C, without displaying any shoulder peaks below
50 °C. However, the peak did get broader in the shape of
a skewed Gaussian distribution with the increase in the
CuO/c-Al O weight ratio, and this could suggest the for-
2
3
mation of a larger size of heterogeneously-distributed
CuO–ZnO crystallites on the surface of c-Al O [6],
2
3
especially in CZA(7), which had a high temperature
reduction peak due to its large copper content. The lowest
reduction temperature was observed for CZA(2), and this
also suggests a homogeneous dispersion of CuO–ZnO
crystallites due to its Gaussian distribution. TPR profiles of
the CZA catalysts also suggest that the CuO species were
well distributed on the c-Al O surface by the formation of
2 3
Fig. 1 TPR profiles on the bifunctional Cu–ZnO/c-Al O catalysts
responsible for the suppressed aggregation of copper
crystallites during the 20 h reaction duration. The crystal-
lite size of CuO on CZA catalysts was characterized by
XRD analysis on fresh CZA catalysts, and the results are
summarized in Table 2 and Fig. 2a. The crystalline phases
2
3
a CuO–ZnO matrix [10, 11], and this homogeneous dis-
tribution with a small crystallite size on CZA(2) was
123

6
70
J. W. Jeong et al.
above 800 °C is assigned to the desorption of water mol-
ecules by possible structural collapse of the main frame-
works of c-Al O and CuO–ZnO. Consequently, only the
2
3
T and T peaks were assigned to the active sites for
1
2
methanol dehydration to DME. The concentration of weak
acid sites for the T1 peak decreased from 0.514 to
0
.321 mmol NH /g with an increase in the CuO/c-Al O
3
2 3
weight ratio by the deposition on c-Al O surface, and this
2
3
was also confirmed by the observed decrease in BET sur-
face area. The T peak assigned to the strong acidic sites
2
was maximized for CZA(2), with 1.038 mmol NH /g, and
3
the summation of the acidic sites (T ? T ) [denoted as
1
2
D(AS)] for the active sites of methanol dehydration to
DME was also maximized on CZA(2), with 1.506 mmol
NH /g, showing the shape of a volcano curve. This is
3
mainly attributed to the degree of dispersion of the copper
crystallites and the pore blockage of the Al O surface, and
2
3
the smaller crystallite size of copper oxide and a lower
degree of pore blockage were characteristic of CZA(2).
3.2 Effects of the Copper Surface Area and Acidic
Sites
Although the characteristics of bifunctional CZA catalysts
before reaction are important to describe the catalytic
activity at the beginning of reaction, the physicochemical
properties of the used bifunctional CZA catalysts are more
convincing to explain the different catalytic performances.
Therefore, the used bifunctional CZA catalysts were fur-
ther characterized by XRD and N O titration methods to
2
measure their copper surface areas and crystallite sizes. On
industrial Cu–ZnO–Al O catalyst for CO hydrogenation
Fig. 2 XRD patterns of the fresh and used bifunctional Cu–ZnO/
c-Al catalysts a fresh CZA catalysts, b used CZA catalysts
2
3
2 3
O
to methanol, metallic coppers are known to be active sites,
and ZnO and Al O act as structural promoters to enhance
2
3
of CuO, ZnO, and c-Al O were observed on all fresh
the dispersion of copper crystallites and to suppress the
aggregation of copper crystallites, which is affected by the
composition of Cu–ZnO–Al O and the synthesis condi-
2
3
catalysts, and the calculated crystallite size was in the
range of 10.6–14.5 nm. The smallest crystallite size of
CuO was observed on CZA(2), with a size around 10.6 nm,
which corresponded to the lower reduction temperature
with a Gaussian distribution in the TPR experiment.
The acidic properties of bifunctional CZA catalysts were
2
3
tions [19, 20]. As shown in Fig. 2b and Table 2, the copper
crystallites were significantly aggregated after 20 h of
reaction, as indicated by the increased peak intensity of the
XRD patterns, and the calculated crystallite size of metallic
copper was in the range of 24.2–36.3 nm. The smallest
crystallite size of 24.2 nm was observed on CZA(2), with a
low sintering character, and the larger crystallite formation
was observed on CZA(7), which had a higher copper
content. This is also supported by the observed high surface
characterized by NH -TPD, and the summarized results are
3
also included in Table 2. The desorption temperatures of
NH were correlated with the extent of surface acidity for
3
the activity for methanol dehydration to DME. The
desorption patterns of NH3 (not included) showed two
distinguishable peaks, the T peak below 300 °C and the T
area of metallic copper [denoted as S(Cu)] with a value of
2
6.25 m /g on CZA(2) after reaction, as measured by the
1
2
peak in the range of 300–500 °C. According to our previ-
Cu
ous investigations [7–10], the T peak corresponds to the
N O titration method and summarized in Table 2. The
2
1
acidity attributed to the weakly acid sites from the c-Al O
surface area of metallic copper was in the range of
2
3.47–6.25 m /g , and the area was in the shape of a vol-
2
3
support matrix, and the T2 peak corresponds to the
desorption of strongly adsorbed NH3 on CuO–ZnO or
c-Al O acidic sites. In addition, the observed sharp peak
Cu
cano curve, showing a slight decrease above the CuO/
c-Al O weight ratio of 1.75. Although the surface area of
2 3
2
3
1
23

Effects of Cu–ZnO Content on Reaction Rate
671
copper as well as the number of acidic sites were found to
be in the shape of a volcano curve, the decline in acidic site
density was more significant with the increase in CuO/
c-Al O weight ratio. This could be attributed mainly to the
density. The increased CO conversion on CZA(2) is mainly
attributed to a larger number of acidic sites rather than a
higher surface area of copper, since the decrease in surface
area with the increase in the CuO/c-Al O weight ratio was
2
3
2 3
abundant presence of CuO–ZnO crystallites on the outer
not significant, but the increase in the number of acid sites
was significant. Therefore, the overall rate of CO conver-
sion in this consecutive DME synthesis reaction could be
predominantly controlled by the rate of methanol dehy-
dration to DME, as compared to the CO hydrogenation rate
to methanol. As suggested by Blaszkowski and van Santen
RA [15], the lowest activation barrier for an adsorption
pathway of two methanol molecules is responsible for the
facile formation of DME by the dehydration of methanol
compared to methoxy formation from one adsorbed
methanol molecule. Therefore, the observed increase in
DME selectivity with a higher CO conversion for CZA(2)
could be attributed to the fast dehydration rate of methanol
to DME due to the abundant formation of methanol on the
active copper metal surfaces. Although the copper surface
area on the bifunctional CZA catalysts could be signifi-
cantly enhanced by the small crystallite size of the copper
species with a high surface area on CZA(2), the total
number of acidic sites significantly contributed to the
overall reaction rate for direct DME synthesis from syngas.
This is also supported by the observation of a lower rate
surface of c-Al O , where they blocked mesopores.
2
3
Therefore, a larger metallic surface area of copper and
larger number of acidic sites on CZA catalysts could be
simultaneously responsible for a higher catalytic activity
[
13, 14].
On CZA(2), which showed a higher CO conversion and
DME selectivity, the number of acidic sites assigned to
peaks of I and II (below 500 °C) and the surface area of
metallic copper after reaction were the largest among the
2
catalysts, with values of 1.506 mmol NH /g and 6.25 m /g,
3
respectively. The enhanced CO conversion on the bifunc-
tional CZA catalysts was due to the fast consumption rate
of methanol during its dehydration to DME, with the sur-
plus hydrogen formation through the WGS reaction over-
coming an equilibrium conversion of CO. The observed
lower CO conversion with CZA(7) could be attributed to a
lower number of acidic sites for a lower rate of methanol
dehydration. In addition, the hydrocarbon formation was
slightly greater for CZA(2) and CZA(4) due to the abun-
dant presence of acidic sites, which are possible active sites
for methanol or DME reforming reactions [6, 10, 14, 21,
with CZA(7), which possessed a higher metallic surface
2
area of copper (5.90 m /g ) and a smaller number of
2
2]. Therefore, the performance of the CZA catalysts
Cu
depends simultaneously on the surface area of metallic
copper and the number of acidic sites, but their contribu-
tions could be somewhat different.
acidic sites (0.758 mmol NH /g). These characteristics of
3
CZA(7) were possibly induced by acidic site blockage by
the deposition of small Cu–ZnO crystallites on c-Al O
2
3
As shown in Fig. 3, the reaction rate of the CO con-
version to DME was strongly related to the number of
acidic sites as well as the metallic copper surface area. The
maximum reaction rate was observed with CZA(2), which
possessed a higher copper surface area and acidic site
surface, which resulted in a lower overall CO conversion.
The simultaneous consideration of the surface area of
metallic copper and the number of acidic sites showed a
good correlation for the simply prepared Cu–ZnO/c-Al O
2
3
catalysts, which is in agreement with the results of our
previous work [14], and the correlations of the values of
S(Cu) 9 D(AS) with the reaction rate are well explained
for the bifunctional CZA catalysts, rather than the corre-
lations with the values of S(Cu) or D(AS) alone, as shown
in Fig. 3. In addition, the number of acidic sites exposed on
solid-acid c-Al O is a more crucial factor to obtain a
2
3
higher DME yield on bifunctional CZA catalysts than the
surface area of metallic copper due to the fast dehydration
rate of methanol to DME.
4
Conclusions
CO conversion and DME selectivity with bifunctional
Cu–ZnO/c-Al O catalysts for the direct synthesis of DME
2
3
Fig. 3 Correlation of reaction rates versus the number of acid sites
from syngas were strongly related to the surface area of
metallic copper and the number of acidic sites; however,
the overall conversion of CO in this consecutive reaction is
(
D(AS)), surface area of metallic copper after reaction (S(Cu)), and
their product values (D(AS) 9 (S(Cu)) on bifunctional Cu–ZnO/c-
Al catalysts
2 3
O
123

6
72
J. W. Jeong et al.
predominantly controlled by the rate of methanol dehy-
dration to DME as compared to hydrogenation rate of CO
to methanol. From the present observation, the number of
acidic sites is a more crucial factor to enhance DME yield
with bifunctional Cu–ZnO/c-Al O catalysts. The signifi-
3. Gao Z, Huang W, Yin L, Hao L, Xie K (2009) Catal Lett 127:354
4
. Chen WH, Lin BJ, Lee HM, Huang MH (2012) Appl Energy
8:92
9
5
. Aguayo AT, Erena J, Mier D, Arandes JM, Olazar M, Bilbao J
(2007) Ind Eng Chem Res 46:5522
6. Pokrovski KA, Bell AT (2006) J Catal 241:276
2
3
7
. Bae JW, Potdar HS, Kang SH, Jun KW (2008) Energy Fuels
2:223
cant increase in the reaction rate for the direct CO con-
version to DME is mainly attributed to the presence of a
large number of acidic sites created by optimizing the CuO/
c-Al O weight ratio at around two and by enhancing the
2
8
. Kang SH, Bae JW, Kim HS, Dhar GM, Jun KW (2010) Energy
Fuels 24:804
9. Prasad PSS, Bae JW, Kang SH, Lee YJ, Jun KW (2008) Fuel
Process Technol 89:1281
0. Bae JW, Kang SH, Lee YJ, Jun KW (2009) J Ind Eng Chem
2
3
consumption rate of the methanol dehydration to DME.
The design of a high surface area of metallic copper with
abundant acidic sites during the coprecipitation step are
important factors for obtaining a higher catalytic perfor-
mance, and the contribution of acidic sites is more domi-
nant for the overall reaction rate with bifunctional
Cu–ZnO/c-Al O catalysts.
1
1
5:566
11. Kang SH, Bae JW, Jun KW, Potdar HS (2008) Catal Commun
9:2035
2. Okamoto Y, Fukino K, Imanaka T, Teranishi S (1983) J Phys
Chem 87:3740
3. Venugopal A, Palgunadi J, Jung KD, Joo OS, Shin CH (2009)
J Mol Catal A 302:20
4. Jung JW, Lee YJ, Um SH, Yoo PJ, Lee DH, Jun KW, Bae JW
(2012) Appl Catal B 126:1
1
1
1
2
3
Acknowledgments The authors would like to acknowledge the
financial support of a National Research Foundation (NRF) of Korea
Grant funded by the Korean government (MEST; 2011-0009003 and
15. Blaszkowski SR, van Santen RA (1997) J Phys Chem B 101:2292
16. Mao D, Yang W, Xia J, Zhang B, Song Q, Chen Q (2005) J Catal
230:140
2
012R1A2A2A02013876). This work was supported by the Korea
Institute of Energy Technology Evaluation and Planning (KETEP)
under ‘‘Energy Efficiency & Resources Programs’’ with Project
number of 2011T100200023. This work was also supported by the
grant from the Industrial Source Technology Development Programs
17. Abu-Dahrieh J, Rooney D, Goguet A, Saih Y (2012) Chem Eng J
203:201
18. Jensen JR, Johannessen T, Livbjerg H (2004) Appl Catal A
266:117
(2012-10042712) by the Korea government Ministry of Trade,
Industry and Energy.
19. Baltes C, Vukojevic S, Schuth F (2008) J Catal 258:334
20. Behrens M (2009) J Catal 267:24
2
1. Baek SC, Kang SH, Bae JW, Lee YJ, Lee DH, Lee KY (2011)
Energy Fuels 25:2438
2. Bae JW, Kang SH, Lee YJ, Jun KW (2009) Appl Catal B 90:426
2
References
1
2
. Gupta M, Smith ML, Spivey JJ (2011) ACS Catal 1(6):641
. Ng KL, Chadwick D, Toseland BA (1999) Chem Eng Sci
5
4:3587
1
23