Page 1
2116 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
Cite this: Catal. Sci. Technol., 2012, 2, 2116–2127
Synthesis of methanol and dimethyl ether from syngas over
Pd/ZnO/Al2O3 catalysts
Vanessa M. Lebarbier,*aRobert A. Dagle,*
aLibor Kovarik,
a
Jair A. Lizarazo Adarme,bDavid L. King
aand Daniel R. Palo
b
Received 11th May 2012, Accepted 16th June 2012
DOI: 10.1039/c2cy20315d
A Pd/ZnO/Al2O3 catalyst was developed for the synthesis of methanol and dimethyl ether (DME)
from syngas with temperatures of operation ranging from 250 1C to 380 1C. High temperatures
(e.g. 380 1C) are of interest when combining methanol and DME synthesis with a methanol to
gasoline (MTG) process in a single reactor bed. A commercial Cu/ZnO/Al2O3 catalyst, utilized
industrially for the synthesis of methanol at 220–280 1C, suffers from a rapid deactivation when
the reaction is conducted at high temperature (4 320 1C). On the contrary, a Pd/ZnO/Al2O3
catalyst was found to be highly stable for methanol and DME synthesis at 375 1C. The Pd/ZnO/
Al2O3 catalyst was thus further investigated for methanol and DME synthesis at P = 34–69 bar,
T = 250–380 1C, GHSV = 5000–18 000 h�1, and molar feeds H2/CO = 1, 2, and 3. Selectivity
to DME increased with decreasing operating temperature, and increasing operating pressure.
Higher space velocity and H2/CO syngas feed ratios also enhanced DME selectivity. Undesirable
CH4 formation was observed, however, it could be lessen through choice of process conditions
and by catalyst design. By studying the effect of the Pd loading and the Pd : Zn molar ratio the
formulation of the Pd/ZnO/Al2O3 catalyst was optimized. A catalyst with 5% Pd and a Pd : Zn
molar ratio of 0.25 : 1 has been identified as the preferred catalyst. Results indicate that PdZn
particles are more active than Pdo particles for the synthesis of methanol and less active for CH4
formation. A correlation between DME selectivity and concentration of acid sites has been established.
Hence, two types of sites are required for the direct conversion of syngas to DME: (1) PdZn particles
are active for the synthesis of methanol from syngas, and (2) acid sites which are active for the
conversion of methanol to DME. Additionally, CO2 formation was problematic as PdZn was found to
be active for the water-gas-shift (WGS) reaction, under all the conditions evaluated.
Introduction
Strained fossil fuel reserves and the increasing demand of
emergent countries make the use of renewable energies attrac-
tive. Sun, wind and biomass are promising sources of renew-
able energies. Several technologies exist for converting
biomass to energy such as heat, electricity or fuel. One
approach involves biomass gasification to produce syngas.
Syngas can be burned in gas turbines like natural gas, or it
can be converted to high quality chemicals and fuels, such as
methanol and dimethyl ether (DME). Methanol and DME are
key intermediates in the methanol-to-gasoline (MTG) and
methanol-to-olefins (MTO) processes.1,2
DME is often produced in a two-step process where methanol
is first generated from syngas followed by methanol dehydration
over a solid acid catalyst to produce DME. Alternatively, in a
single-step conversion, DME is produced directly from syngas
over a bifunctional or hybrid catalyst system employing both a
methanol synthesis function and a methanol dehydration
function.3,4 Producing DME directly from syngas has
many economic and technical advantages, provided suitable
catalyst(s) exist.5 In addition, thermodynamically, DME
production from syngas is favored over methanol. Direct
DME synthesis involves several competing reaction pathways.
Methanol synthesis (eqn (1) and (2)) and the water-gas-shift
(WGS, eqn (3)) are equilibrium reactions occurring over metal
or mixed metal catalysts:
CO + 2H2 2 CH3OH DH0 = �92.0 kJ mol�1
(1)
CO2 + 3H2 2 CH3OH + H2O DH0 = �49.5 kJ mol�1
(2)
CO + H2O 2 CO2 + H2 (WGS) DH0 = �41.1 kJ mol�1
(3)
a Institute for Integrated Catalysis, Pacific Northwest NationalLaboratory, Richland, WA 99352, USA.E-mail: [email protected] , [email protected]
bMicroproducts Breakthrough Institute, Pacific Northwest NationalLaboratory, Corvallis, Or 97330, USA
CatalysisScience & Technology
Dynamic Article Links
www.rsc.org/catalysis PAPER
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online / Journal Homepage / Table of Contents for this issue
Page 2
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2117
Dehydration of methanol to DME (eqn (4)) is also equilibrium
controlled and occurs over solid acid catalysts. By producing
DME in the same step, methanol is continuously consumed in
the reactor, pushing the conversion of syngas beyond that
which would be achieved with methanol synthesis alone.
2CH3OH 2 CH3OCH3 + H2O (dehydration)
DH0 = �23.6 kJ mol�1 (4)
Methane forming reactions could occur from the hydrogeno-
lysis of DME (eqn (5)), hydrogenolysis of methanol (eqn (6))
or from methanation reactions (eqn (7) and (8)) over metal
active sites. When operating at temperatures lower than
approximately 280 1C, as is typically the case for direct
DME synthesis, these reactions are not usually observed to
any great extent.5
CH3OCH3 + 2H2 - 2CH4 + H2O (hydrogenolysis)
DH0 = �207.5 kJ mol�1 (5)
CH3OH + H2 - CH4 + H2O (hydrogenolysis)
DH0 = �115.4 kJ mol�1 (6)
CO + 3H2 2 CH4 + H2O (CO methanation)
DH0 = �206 kJ mol�1 (7)
CO2 + 4H2 2 CH4 + 2H2O (CO2 methanation)
DH0 = �165 kJ mol�1 (8)
Combined methanol/DME synthesis typically necessitates the
use of a mixed catalyst system. Traditionally utilized are
CuZnAl-type methanol synthesis catalysts and acidic zeolite
or alumina to facilitate methanol dehydration.6,7 The system
operates at high pressure (5–10 MPa) and relatively low
temperature (200–280 1C). The well known Topsoe Integrated
Gasoline Synthesis (TIGAS) process has demonstrated
successful integration of methanol and DME synthesis.8 The
TIGAS process consists of the methanol/DME production in
one step and then in a second step methanol/DME is con-
verted to gasoline using the zeolite-catalyzed MTG process.8
Two separate synthesis reactors are used because the optimal
temperatures are different for the two operations. Methanol
synthesis favors temperatures around 250 1C, using con-
ventional Cu methanol catalysts, whereas MTG proceeds at
350–400 1C.2
Being a capital intensive process, combining methanol/
DME synthesis and gasoline production into a one synthesis
reactor operation would be economically attractive. Several
groups have attempted this integration using conventional
methanol synthesis and zeolite materials.9,10 However, several
technical hurdles exist. For one, there is a mismatch in
pressure and temperature when combining conventional
methanol synthesis and MTG. The high pressure (5–10 MPa)
operating conditions required to synthesize methanol/DME
will affect the final product distribution of the gasoline. The
conversion of DME to gasoline requires relatively high
temperature (350–400 1C),2 operating conditions at which
methanol/DME synthesis by itself is thermodynamically
unfavorable. Furthermore, the conventional Cu/ZnO/Al2O3
catalyst used for methanol synthesis suffers from instability at
temperatures greater than approximately 270 1C.11 Hence, a
catalyst both active and stable for the synthesis of methanol at
high temperature is lacking for the direct conversion of syngas
to gasoline.
In recent years a Pd/ZnO-type catalyst has been developed
for the methanol steam reforming reaction (MSR). Pd/ZnO
has been shown to be quite active and especially selective to
direct production of H2 and CO2.12,13 Beyond methanol
steam reforming, the PdZn-type catalyst was also found to
be very active for both water gas shift14 and also for methanol
synthesis,15 the latter having operating as high as 350 1C.
Pd/ZnO catalysts have been shown to be significantly more
stable for methanol steam reforming at higher temperatures
(e.g. 4280 1C) as compared to the conventional CuZnAl-type
methanol catalyst.16 Hence, the more stable PdZn-type
catalyst is a potential candidate for use in the direct synthesis
of gasoline from syngas.
The present study investigated a Pd/Zn/Al2O3 bifunctional
catalyst for the direct conversion of syngas to methanol and
DME over a wide temperature range, 250–380 1C. The higher
temperatures incorporate a regime that is not optimal for the
methanol synthesis reaction since this reaction is thermo-
dynamically limited. However, in a single step gasoline
synthesis process equilibrium constraint to hydrocarbon
product is alleviated, as described in more detail elsewhere
(e.g. integrating Pd/ZnO/Al2O3 with zeolite).17,18 The, objec-
tives here are to (1) demonstrate that the Pd/ZnO/Al2O3
catalyst is more stable than the Cu/ZnO/Al2O3 catalyst for
methanol synthesis at high temperature (375 1C), and (2) to
study the effect of the Pd loading and Pd : Zn molar ratio on
catalyst performance in order to optimize catalyst formulation
and understand what parameters affect the selectivity.
Operating temperature, pressure, gas hour space velocity
(GHSV), and syngas ratio (H2/CO) were process variables
explored. Catalytic compositions were altered by varying Pd
and Zn loadings on the alumina substrate, affecting catalytic
activity and selectivity. The relationship of the bifunctional
PdZn metal and acid sites and their impact on catalytic
performance were investigated. A combination of synthesis
experimentation and material characterization has led to
insights regarding the reaction pathways.
Experimental section
1. Catalyst preparation
Two series of Pd/ZnO/Al2O3 catalysts were prepared by
incipient wetness impregnation of an Al2O3 support (Engelhard,
AL-3945E) with a Pd nitrate solution (21.21 wt% Pd in nitric
acid) to which a Zn nitrate precursor (Sigma Aldrich) was
added, as reported in a previous study.19 The catalysts were
dried at 110 1C for 8 h and calcined at 350 1C for 3 h. For the
first series, the Pd loading was equal to 8.8 wt% and the
Pd : Zn molar ratio varied from 0.25 : 1 to 0.75 : 1. For the
second series, the Pd loading varied from 2.5 to 20 wt% and
the Pd : Zn molar ratio was kept constant and equal to 0.25 : 1.
A Pd/Al2O3 catalyst with 8.8 wt% Pd was prepared according
to the same method with no addition of Zn nitrate precursor
to the Pd solution. The Pd/ZnO/Al2O3 catalysts were labeled
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 3
2118 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
xPd/ZnO/Al2O3 � y were x stands for the Pd loading and y
stands for the Pd/Zn molar ratio. For example, 8.8Pd/ZnO/
Al2O3�0.25 indicates a Pd loading of 8.8 wt% and a
Pd : Zn content of 0.25 : 1 (molar). For comparison purpose
a commercial Cu/ZnO/Al2O3 catalyst (Synetix, F51-8 PPT)
was tested under the same conditions as the supported Pd
catalysts. Note that the term ‘‘spent’’ refers to the catalyst
after reaction.
2. BET surface area
Nitrogen adsorption was measured at 77 K with an automatic
adsorptiometer (Micromeritics ASAP 2000). The samples were
pretreated at 150 1C for 12 h under vacuum. The surface areas
were determined from adsorption values for five relative
pressures (P/P0) ranging from 0.05 to 0.2 using the BET
method. The pore volumes were determined from the total
amount of N2 adsorbed between P/P0 = 0.05 and P/P0 =
0.98. Prior to BET measurements the catalysts have been
reduced under 10% H2/N2 at 400 1C for 2 h.
3. X-ray Diffraction (XRD)
XRD analysis of the spent catalysts (i.e. after methanol
synthesis reaction conditions) was conducted using a Philips
X’pert MPD (Model PW3040/00) diffractometer with a copper
anode (Ka1 = 0.15405 nm) and a scanning rate of 0.011
per second between 2y = 101–701. The diffraction patterns
were analyzed using Jade 5 (Materials Data Inc., Livermore,
CA) and the Powder Diffraction File database (International
Center for Diffraction Data, Newtown Square, PA). Particle
sizes of the samples were determined from the XRD patterns
using the Debye-Sherrer relation (d = 0.89l/ Bcos y, where lis the wavelength of Cu-Ka radiation, B is the calibrated half-
width of the peak in radians, and y is the diffraction angle of a
crystal face). The metal dispersion was estimated from the
particle size by assuming hemispherical geometry using the
equationD=1/d (D=dispersion and d=metal particle size).20
4. Scanning transmission electron microscopy (S/TEM)
Scanning Transmission Electron Microscopy (S/TEM) was
performed with FEI Titan 80–300 operated at 300 kV. The
FEI Titan is equipped with CEOS GmbH double-hexapole
aberration corrector for the probe-forming lens, which allows
imaging with B0.1 nm resolution in scanning transmission
electron microscopy (STEM) mode. The STEM images were
acquired on High Angle Annular Dark Field (HAADF) with
inner collection angle of 52 mrad. In general, the TEM sample
preparation involved mounting of powder samples on copper
grids covered with lacey carbon support films and immediate
loading them into the TEM airlock to minimize an exposure to
atmospheric O2. Note that the samples were analyzed by
S/TEM after ex situ reduction under 10% H2/N2 at 400 1C
for 2 h.
5. Infrared spectroscopy
IR spectra were recorded with a Bruker spectrometer,
equipped with a MCT detector (resolution: 4 cm�1, 256 scans).
The samples pressed into a pellet were first pretreated under
H2 for 2 h at 400 1C. During this pretreatment, the sample was
alternatively exposed to H2 for 30 min and evacuated under
vacuum for 15 min to simulate flow conditions. After that, the
temperature was cooled down to room temperature and small
doses of CO were progressively added until saturation of the
catalyst surface occurred.
6. NH3 temperature programmed desorption (TPD)
NH3-TPD experiments were performed on an automated
catalyst characterization unit (Micromeritics Autochem
2910) equipped with a TCD detector. The catalyst (0.1 g)
was loaded in a U-type quartz tube. Then a 10% H2/Ar
mixture was passed through the sample starting from 20 1C
and heating up to 400 1C with a ramp of 5 1C min�1 and held
at this temperature for 2 h under 10% H2/Ar mixture and one
more hour under He. The temperature was cooled down to 100 1C
under He flow and the adsorption of NH3 (16% NH3/He) was
carried out at 100 1C for 2 h. After that, He flowed for 2 h at
the same temperature, to remove the physisorbed NH3 from
the surface of the catalyst. The catalyst was then heated to
650 1C (ramp 5 1C min�1) and held at this temperature for 1 h.
7. Catalytic activity
Catalytic activity tests were conducted in a 7.8 mm inner
diameter fixed-bed reactor. The catalyst (0.6 g), diluted with
SiC (3 g), was loaded between two layers of quartz wool inside
the reactor. A dual K-type thermocouple was placed in the
reactor for the measurement of inlet and catalyst bed
temperatures. The catalyst was reduced at 400 1C for 2 h,
using 10% H2/N2 gas mixture, prior to the test. A premixed
gas containing H2, CO, CO2 and N2 was fed into the system
using a Brooks Mass Flow Controller (5850E series). Four
different premixed gas compositions with syngas ratios H2/CO=
1, 2 or 3 (mol) were used and are listed in Table 1. The
catalysts were tested at temperatures between 250–380 1C over
a range of gas hour space velocity (5000–20 000 h�1) and
pressures (34.5–69 bar). The gas products were separated using
MS-5A and PPU columns and analyzed on-line by means of
an Agilent Micro GC equipped with a TCD detector. The
catalysts activities were compared using the following definitions:
CO conversion ð%Þ ¼ moles of CO reacted
moles of CO fed� 100
Selectivity ð%Þ ¼ moles Pi � uiP
i
moles of Pi � ui� 100
where Pi is a certain product and u is the number of carbon
atoms/molecule in Pi. e.g. if P = CO2, uCO2= 1, while for P=
CH3OCH3, uCH3OCH3= 2. Chemcad (version 5.6) was used to
estimate the equilibrium CO conversion.
Table 1 Premix gas compositions
H2/CO H2 (%) CO (%) CO2(%) N2 (%)
Premix 1 2 59.73 32.18 4.15 3.94Premix 2 1 41.50 41.44 12.97 4.09Premix 3 2 58.03 28.98 8.89 4.10Premix 4 3 66.11 21.94 7.75 4.20
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 4
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2119
Results and discussion
1. Catalyst compositions and textural properties
Compositional information and the surface area and pore
volume of the catalysts are shown in Table 2. For a given
Pd loading, the surface area and pore volume increase with an
increase of Pd : Zn ratio (i.e. decrease in the ZnO content).
This could be due to a blocking of the pores of the Al2O3
support by ZnO and by the PdZn particles. The PdZn particle
size decreases with the Pd : Zn ratio (see Table 3). Similarly,
the decrease of the surface area and pore volume with the
increase of the Pd loading from 2.5 to 20% is likely due to the
increase of the PdZn particle size and ZnO content.
2. X-ray diffraction
For the 8.8Pd/ZnO/Al2O3�0.38 catalyst exposed to methanol
synthesis reaction conditions, the results have shown that the
PdZn particles size increases during the first 12 h on stream
(from 4.0 nm to 7.5 nm) but does not significantly increase for
TOS 4 12 h. We have thus examined the spent catalysts by
XRD to determine the PdZn particle size. Fig. 1(a) shows the
XRD patterns for the spent Pd/ZnO/Al2O3 catalysts with
different Pd loading and same Pd : Zn molar ratio, measured
at 2y= 401 to 481. For the samples with a Pd loading4 2.5%,
peaks characteristic of bimetallic PdZn at 41.21 and 44.11 are
observed. When the Pd loading increases these peaks become
more intense and their bandwidths decrease indicating an
increase of the PdZn particle size (see Table 3). For the
2.5Pd/ZnO/Al2O3�0.25 catalyst, one broad peak is detected
between 2y = 40.51 and 42.51. It is likely that this peak is
characteristic of Al2O3. However, the presence of small PdZn
particles that could contribute to this broad peak is not ruled out.
None of the XRD patterns shows peaks characteristic of Pdo
(expected at 40.21), suggesting that the samples present only
bimetallic PdZn particles. Fig. 1(b) displays the XRD patterns
for the Pd/ZnO/Al2O3 catalysts with 8.8% Pd and different
Pd :Zn molar ratios. The XRD pattern obtained for the 8.8Pd/
Al2O3 catalyst is presented in Fig. 1(b) as well. The XRD
patterns for the Pd/ZnO/Al2O3 catalysts show only peaks char-
acteristic of PdZn, again suggesting the absence of metallic Pdo
particles. Note that the bandwidth of the PdZn peak at 41.21
increases with the Pd :Zn ratio, indicating a decrease in the
bimetallic PdZn particle size. The PdZn particle size and
dispersion calculated from these XRD measurements are pre-
sented in Table 3. The dispersion increases with the increase in
Pd :Zn molar ratio and decreases with an increase in the Pd
loading.
3. Scanning transmission electron microscopy (STEM)
The Pd/ZnO/Al2O3 catalysts were analyzed using STEM.
Fig. 2a shows bimetallic PdZn particles supported by ZnO/
g-Al2O3 for the 2.5Pd/ZnO/Al2O3�0.25 catalyst. Detailed
high-resolution imaging in Fig. 2b confirms that the particles
are PdZn bimetallic and have an ordered tetragonal structure
with L10 type ordering. Nevertheless, some of the PdZn
particles, such as the one shown in Fig. 2b, exhibit a contrast
variation at the nanoscale indicating some compositional or
structural inhomogeneities. The observations of bimetallic
PdZn particles with tetragonal L10 type ordering are common
for all of the analyzed Pd/ZnO/Al2O3 catalysts.
4. Infrared spectroscopy
Fig. 3(a) shows the infrared spectra recorded between
1800–2125 cm�1, after saturation of the catalyst surface with
CO at room temperature, for the Pd/ZnO/Al2O3�0.25 cata-
lysts with different Pd loadings. For all the catalysts, the IR
spectra present one main band between 2069–2077 cm�1
ascribed to the vibration of CO linearly adsorbed on the PdZn
alloy particles.21,22 The shift observed between the spectra for
the different catalysts for the band at 2069–2077 cm�1 is not
understood yet. In addition, a closer look at the spectrum
obtained for the 2.5Pd/ZnO/Al2O3�0.25 catalyst shows one
broad band between 1800–2000 cm�1 due to multi-bonded
CO species21 and characteristic of Pd1 particles.22 Note that
the band at 1800–2000 cm�1 is hardly detectable for the
2.5Pd/ZnO/Al2O3�0.25 catalyst, suggesting that the amount
of Pd1 is low compared to the amount of bimetallic PdZn
particles. Note that the presence of a band due to linearly CO
species adsorbed on Pd1 for the catalysts with Pd 4 2.5% can
be ruled out since for Pd1/ZnO/Al2O3 catalyst band due to
linearly adsorbed CO between 2100–2000 cm�1 is accompa-
nied by a more intense band between 2000–1800 cm�1.22
Fig. 3(b) shows the infrared spectra recorded between
1800–2125 cm�1, for the 8.8Pd/ZnO/Al2O3 catalysts with
different Pd : Zn molar ratios. All the spectra present one band
at 2069–2077 cm�1 characteristic of PdZn alloy. The spectrum
recorded for the 8.8Pd/ZnO/Al2O3�0.75 catalyst also shows
Table 2 Catalysts composition and textural properties
CatalystPd loading(wt%)
Pd : Zn(molar)
Surface area(m2 g�1)
Pore volume(cm3 g�1)
8.8Pd/ZnO/Al2O3�0.25
8.8 0.25 : 1 171.8 0.37
8.8Pd/ZnO/Al2O3�0.38
8.8 0.38 : 1 183.4 0.43
8.8Pd/ZnO/Al2O3�0.75
8.8 0.75 : 1 197.2 0.46
8.8Pd/Al2O3 8.8 1 : 0 229.2 0.552.5Pd/ZnO/Al2O3�0.25
2.5 0.25 : 1 213.6 0.59
5Pd/ZnO/Al2O3�0.25
5.0 0.25 : 1 192.8 0.5
20Pd/ZnO/Al2O3�0.25
20 0.25 : 1 81.05 0.12
Table 3 Particles size of the Pd/ZnO/Al2O3 catalysts determined byXRD and dispersion measurements
Catalyst PdZn Particle size (nm) Dispersiona (%)
8.8Pd/ZnO/Al2O3�0.25 13.8 7.28.8Pd/ZnO/Al2O3�0.38 8.7 11.58.8Pd/ZnO/Al2O3�0.75 7.3 13.72.5Pd/ZnO/Al2O3�0.25 o4b NA5Pd/ZnO/Al2O3�0.25 8.6 11.620Pd/ZnO/Al2O3�0.25 14.8 6.8
a dispersion calculated from PdZn particle size using the equation
D = 1/d with D = dispersion and d = PdZn particle size. b Peaks
characteristic of bimetallic PdZn particles were not detected by XRD
indicating that the particle size is below the XRD detection limit
(i.e. o 4 nm).
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 5
2120 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
one more band between 1800–2000 cm�1 which is attributed to
Pd1. These results suggest that the amount of Pd1 increases
with the Pd : Zn molar ratio. Note that contrary to the IR
measurements, the XRD patterns did not indicate the presence
of Pd1 for the 8.8Pd/ZnO/Al2O3�0.75. It can be due to the
fact that the Pd1 particles are too small to be detected by XRD
or to the fact that IR spectroscopy is sensitive to the surface
composition of the catalyst, whereas XRD technique informs
on the structure/bulk composition of the catalyst.
5. Catalytic activity
5.1 Thermodynamics of methanol synthesis and dehydration
reactions. Fig. 4(a) presents the equilibrium CO conversion
for the synthesis of methanol at T = 225–400 1C and P =
34.5–69 bar and H2/CO/CO2 of 2/1/0.13 (premix 1 in Table 1).
For these calculations two cases are considered: (1) methanol
was the only product, and (2) both methanol and DME as
products. For the first case, equilibrium CO conversion
decreases with increasing temperature from 75% at 225 1C,
to 0% at 400 1C. For the second case, with both methanol and
DME as products, the same trend of decreasing conversion
with increasing temperature is seen, but overall conversions
are higher. This demonstrates the benefit in thermodynamic
driving force when both methanol synthesis and methanol
dehydration are employed in tandem. Fig. 4(a) also shows how
equilibrium CO conversion increases with pressure. For
example, at 375 1C, CO conversion increases from 10% at
P = 34.5 bar, to 38% at P = 69 bar.
Fig. 4(b) shows the equilibrium CO conversion and selectiv-
ities to methanol, DME, and CO2 at T= 225–400 1C and P=
69 bar, when considering both methanol and DME as products.
Equilibrium selectivity to DME decreases from 64% at 225 1C,
Fig. 1 XRD patterns for the spent Pd/ZnO/Al2O3 catalysts with different Pd loading and same Pd : Zn molar ratio (a) and for the catalysts with
8.8% Pd and different Pd : Zn molar ratio (b).
Fig. 2 (a) General view of the supported PdZn particles for the spent
2.5% Pd/ZnO/Al2O3 0.25 : 1 catalyst. (b) High resolution HAADF image
revealing the crystallographic nature of the PdZn intermetallic particles.
Fig. 3 Infrared spectra recorded after CO adsorption at room temperature, after saturation of the surface by CO for the Pd/ZnO/Al2O3 catalysts
with different Pd loading and Pd : Zn molar ratio.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 6
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2121
to 40% at 400 1C. On the other hand, equilibrium selectivity to
CO2 increases from 31% at 225 1C, to 54% at 400 1C.
Equilibrium methanol selectivity remains quite level across
the entire temperature range at approximately 5%.
5.2 Comparison of catalytic activity for commercial Cu/
ZnO/Al2O3 and 8.8Pd/ZnO/Al2O3 catalysts for methanol and
DME synthesis from syngas. Industrially, the Cu/ZnO/Al2O3
catalyst is used for the synthesis of methanol from syngas.11
We have compared a commercial Cu/ZnO/Al2O3 catalyst to
the 8.8Pd/ZnO/Al2O3�0.38 catalyst for the synthesis of
methanol, focusing on catalyst stability at relatively high
pressure (i.e. 69 bar) and temperature (i.e. 375 1C). These
reaction conditions are suitable for the direct conversion of
syngas17 to gasoline and the temperature is significantly higher
than that employed in conventional methanol synthesis. Fig. 5
presents CO conversion versus time for a period of 125 h on
stream for the two catalysts. Note that the stability test was
conducted at higher GHSV (i.e. 8340 h�1) for the 8.8Pd/ZnO/
Al2O3�0.38 catalyst to ensure the CO conversion to be below
the equilibrium CO conversion. It is clear that the Cu/ZnO/
Al2O3 suffers from rapid deactivation under these conditions.
This deactivation is not surprising and is due to sintering of
the Cu particles.23 This is in stark contrast with the trend
observed for the 8.8Pd/ZnO/Al2O3�0.38 catalyst. CO conver-
sion is quite stable with time-on-stream for the supported
PdZn catalyst. Catalytic activity as a function of temperature
is shown in Fig. 6(a) for both catalysts. Note that for each
catalyst, the temperature was increased progressively from
250 1C to 380 1C. The activity was measured after a 12 h
plateau at 250 1C and a 3–4 h plateau at 310 1C, 330 1C, 355 1C
and 380 1C. For the 8.8Pd/ZnO/Al2O3�0.38 catalyst, CO
conversion increases with temperature until 360 1C, beyond
which it decreases due to equilibrium constraints. At tempera-
ture Z 330 1C the CO conversion is greater for the 8.8Pd/
ZnO/Al2O3�0.38 catalyst than for Cu/ZnO/Al2O3. For
both catalysts, CO2, C2H6, CH4, methanol and DME were
produced during reaction. For Cu/ZnO/Al2O3, the CO con-
version is relatively flat, as compared to the 8.8Pd/ZnO/
Al2O3�0.38 catalyst. This is due to a progressive deactivation
occurring during data collection. The activity was measured
over a 24 h period during which the catalyst deactivates as
evidenced from Fig. 5.
Fig. 6(b) shows the evolution of the methanol, DME, and
CH4 selectivities as a function of temperature. The remaining
selectivity (not shown) is for CO2. Note that for simplification
and because both are the desired products, methanol
and DME are lumped together and shown in the graph as
‘‘methanol + DME’’. The compositional breakdown between
methanol and DME is described below and is also shown in
Table 4. As shown in Fig. 6(b) selectivity to both methanol
and DME decreases dramatically with increasing temperature.
For the Cu/ZnO/Al2O3 catalyst, methanol and DME selecti-
vity decreases from 75% at 240 1C, to 4.3% at 380 1C. For the
8.8Pd/ZnO/Al2O3�0.38 catalyst, the selectivity to methanol
Fig. 4 (a) Evolution of the equilibrium CO conversion as a function of the temperature for methanol synthesis considering the formation of
CH3OH and DME as products for P = 69 bar, P = 51.8 bar, P = 34.5 bar, and considering only the formation of CH3OH for P = 69 bar,
H2/CO/CO2 = 2/1/0.13 (premix 1, Table 1). (b) Evolution of the CO conversion, CO2 selectivity, DME selectivity and CH3OH selectivity at
equilibrium as a function of the temperature for P = 69 bar and H2/CO = 2, considering the formation of CH3OH and DME as products.
Fig. 5 Evolution of the CO conversion with time on stream for the
Cu/ZnO/Al2O3 and the 8.8P/ZnO/Al2O3�0.38 catalysts. Reaction
T = 375 1C, P = 69 bar, H2/CO = 2 (premix 3), GHSV = 3500 h�1
for Cu/ZnO/Al2O3 and GHSV = 8340 h�1 for 8.8Pd/ZnO/Al2O3�0.38.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 7
2122 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
and DME decreases from 36% at 240 1C, to 26% at 380 1C,
but is significantly higher than for the Cu/ZnO/Al2O3 catalyst
at 380 1C. At 380 1C, the selectivity to methanol is 1.8% and
4.5% for the Cu/ZnO/Al2O3 and 8.8Pd/ZnO/Al2O3�0.38catalysts, respectively, whereas the DME selectivity is 2.5
and 21.5%, respectively. This represents a seven-fold selectiv-
ity advantage for the 8.8Pd/ZnO/Al2O3�0.38 catalyst at
380 1C. For both catalysts, selectivity to undesirable CH4
increases with temperature and reaches 32% and 21.4% at
380 1C for the Cu/ZnO/Al2O3 and 8.8Pd/ZnO/Al2O3�0.38catalysts, respectively. CH4 production is an undesired bypro-
duct and could also lead to the formation of coke. At
temperatures above 350 1C, the Cu/ZnO/Al2O3 catalyst
produces a substantially greater amount of CH4 and lesser
amount of methanol and DME compared to the 8.8Pd/ZnO/
Al2O3�0.38 catalyst.
These results clearly show the 8.8Pd/ZnO/Al2O3�0.38catalyst to be preferred over the Cu/ZnO/Al2O3 catalyst for
methanol and DME synthesis from syngas at temperatures
above 350 1C. In contrast to the Cu/ZnO/Al2O3 catalyst, the
8.8Pd/ZnO/Al2O3�0.38 catalyst does not suffer from deacti-
vation at these relatively high temperatures. In addition,
higher selectivity to desirable DME and methanol, and lower
selectivity to undesired CH4 is observed for the 8.8Pd/ZnO/
Al2O3�0.38 catalyst. Optimizing the formulation of the
Pd/ZnO/Al2O3 catalyst in order to potentially suppress the
production of CH4 while facilitating methanol and DME
formation appears thus very relevant.
5.3 Comparison of catalytic activity for the commercial Cu/
ZnO/Al2O3 catalysts, 8.8Pd/ZnO/Al2O3 catalyst and Al2O3
support for methanol dehydration to DME. As discussed above
for both the Cu/ZnO/Al2O3 and 8.8Pd/ZnO/Al2O3�0.38catalysts, formation of DME was observed. It is well known
that dehydration of methanol proceeds over acidic sites offered
by solid acid catalysts such as alumina and zeolites.24 Thus,
the acid sites of the Al2O3 support likely promote DME
formation via methanol dehydration once methanol is formed
from the syngas. To investigate this further, we compared
catalytic activity for the methanol-to-DME dehydration
reaction (eqn (4)) at 250–425 1C and at 1 bar. Activities of
Cu/ZnO/Al2O3, 8.8Pd/ZnO/Al2O3�0.38, and Al2O3 alone
(the support used for the PdZn catalysts) were compared.
Fig. 7 presents methanol conversion versus temperature for the
three catalysts. Conversion increases dramatically with
increasing temperature for both Cu/ZnO/Al2O3 and 8.8Pd/
ZnO/Al2O3�0.38 and complete methanol conversion was
achieved at approximately 350 1C for both. For the Al2O3
support, conversion increases just slightly with temperature,
from 29% to 35%. The significant increase of the conversion
with the temperature for the Cu/ZnO/Al2O3 and 8.8Pd/ZnO/
Al2O3�0.38 catalysts, compared to the Al2O3 alone, is due to a
considerable increase of their activity for methanol decom-
position. Fig. 8(a) and (b) present selectivities for the different
products (DME, CH4, CO and CO2) as a function of the
temperature. For the Al2O3 support, as expected, DME is the
only product observed up to 400 1C. At 425 1C, CH4 is produced
Fig. 6 Evolution of (a) the CO conversion and (b) the methanol, DME and CH4 selectivity as a function of the temperature for the Cu/ZnO/
Al2O3 (Cu) and the 8.8Pd/ZnO/Al2O3�0.38 catalysts. opened symbols: 8.8Pd/ZnO/Al2O3�0.38 (PdZn), filled symbols: Cu/ZnO/Al2O3. P= 69 bar,
GHSV = 10 000 h�1 and H2/CO = 2 (premix 1).
Table 4 Effect of the temperature, pressure and GHSV, on the conversion and selectivity for the 8.8Pd/ZnO/Al2O3�0.38 catalyst
Temperature (1) Pressure (bar) GHSV (h�1) CO conversion (%)
Selectivity (%)
CO2 CH4 C2H6 Methanol DME
307 69 10 000 22.9 54.2 7.6 2.3 8.8 27.1332 69 10 000 41.9 53.5 12.9 3.9 8.4 21.1352 69 10 000 49.7 51.6 15.8 4.4 6.0 22.0380 69 10 000 44.2 52 17.5 4.5 4.5 21.4380 34.5 10 000 20.6 66.1 16.4 2.2 3.3 12380 51.8 10 000 39 57.4 19.6 4.2 4.1 14.7380 69 5000 63.7 57.7 28.4 8.0 2.1 3.8380 69 18 000 40.3 52.8 15.1 3.2 5.2 23.7
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 8
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2123
in addition to DME. Also at this temperature a small amount
of H2 and CO2 was detected. It is thus possible that CH4 was
produced, at least in part, from DME hydrogenolysis or CO2
methanation. Note that H2 and CO2 were likely produced
from methanol steam reforming rather than methanol decom-
position since no CO was detected. This is in contrast with the
results obtained for the Cu/ZnO/Al2O3 catalyst where CH4 is
produced along with CO and CO2, with little DME formation.
It appears that methanol is primarily decomposed to CO and
H2 (the reverse of eqn (1)). The small amount of CO2 formed is
likely due to the water-gas-shift reaction (eqn (3)). The CH4 is
produced from CO (and/or CO2) methanation, DME decom-
position or DME hydrogenolysis. Interestingly, for the
Pd/ZnO/Al2O3�0.38 catalyst, formation of CH4 is not
observed over the entire range investigated. However, DME
formation is observed, with an optimum temperature at
approximately 310 1C (20% selectivity) and decreases with
increasing temperature. Like the Cu/ZnO/Al2O3 catalyst, the
majority product formed over the entire temperature range is
CO, as a result of methanol decomposition.
These results indicate that the Pd/ZnO/Al2O3�0.38 catalyst
is more active for the dehydration of methanol to DME as
compared to commercial Cu/ZnO/Al2O3. This explains why
when syngas feed is used (see Fig. 6b), the selectivity to DME
is higher for the Pd/ZnO/Al2O3�0.38 catalyst than the
Cu/ZnO/Al2O3 catalyst. As seen in Fig. 8, the selectivity
toward DME is lower than the selectivity toward CO and
CO2 whatever the temperature (between 250–410 1C), for the
Pd/ZnO/Al2O3�0.38 catalyst. This is due to the fact that the
methanol dehydration reaction experiments were conducted
at atmospheric pressure. At high pressure (i.e. 69 bar), equili-
brium selectivity to methanol from syngas is favored via
methanol synthesis (eqn (2)). Thus, with increased methanol
synthesis, as opposed to methanol reforming, increased DME
production will result.
5.4 Effect of the temperature, pressure and gas hour space
velocity (GHSV) on the syngas conversion and selectivity for the
8.8Pd/ZnO/Al2O3�0.38 catalyst. The effects of the tempera-
ture, pressure, and GHSV on the reactivity for syngas
conversion to methanol and DME were examined for the
8.8Pd/ZnO/Al2O3�0.38 catalyst and the results are presented
in Table 4. The effect of temperature on conversion and
selectivity has already been discussed above. As expected,
CO conversion increases with pressure. Keeping the operating
temperature constant at 380 1C, CO conversion increases from
20.6% to 44.2% when increasing the pressure from 34.5 bar to
69.0 bar. Selectivity to methanol and DME does increase
somewhat with pressure. Methanol selectivity increases from
3.3% to 4.5% and DME increases from 12.0% to 21.4%. An
increase in methanol production is indeed predicted as the
forward methanol synthesis reaction rates (eqn (1) and (2)) are
favored with increasing pressure. CH4 and C2H6 production
did not increase substantially with pressure. When the pressure
increased from 34.5 bar to 69 bar the CH4 and C2H6
selectivities increased only slightly from 16.4% to 17.5% and
from 2.2% to 4.5%, respectively.
GHSV was varied to determine the time dependence of
conversion and selectivity, with the results shown in Table 4.
As expected, CO conversion decreases with increasing GHSV.
GHSV’s of 5000, 10 000, and 18 000 h�1 resulted in CO
conversions of 63.7%, 44.2%, and 40.3%, respectively.
Selectivities to CO2, CH4 and C2H6 increase with decreasing
GHSV, whereas selectivity toward DME and methanol
Fig. 7 Methanol-to-DME reaction. Evolution of the CH3OH con-
version with the reaction temperature for P=1 bar, GHSV= 5000 h�1
and. CH3OH = 36.1% in N2.
Fig. 8 Methanol-to-DME reaction. Evolution of the DME selectivity (a), CH4 selectivity (a), CO selectivity (b) and CO2 selectivity (b) as a
function of the temperature for the dehydration of methanol. P = 1 bar, GHSV = 5000 h�1, CH3OH = 36.1% in N2. Filled symbols represent
DME (a) and CO (b) selectivities. Opened symbols represent CH4 (a) and CO2 (b) selectivities.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 9
2124 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
decrease. For example, at GHSVs of 5000 and 18 000 h�1
selectivity to DME was 3.8% and 23.7%, respectively. Hence,
shorter residence time favor the formation of DME relative to
side products. Methane formation reactions such as DME
hydrogenolysis (eqn (5)) become increasingly dominant at
longer residence time. The fact that increased throughputs
enhance selectivity to methanol and DME is an important
finding from a practical standpoint.
One can see from Table 4 that a considerable amount of
CO2 is produced during reaction. Indeed, whatever the
reaction temperature (300–380 1C), pressure (34–69 bar) and
GHSV (5000–18 000 h�1), the selectivity to CO2 is always
approximately 50%. Under the present reaction conditions,
the water-gas-shift activity is significant, limiting methanol
and DME formation.
5.5 Effect of the Pd :Zn molar ratio on the reactivity of the
Al2O3 supported catalysts with 8.8% Pd. Changes in Pd/Zn
composition and the resulting effect on catalytic activity were
also examined. Fig. 9(a) presents the evolution of the conver-
sion as a function of the Pd : Zn molar ratio at 380 1C and 69
bar. CO conversion increases from 36% to 44% when the
Pd : Zn molar ratio increases from 0.25 : 1 to 0.38 : 1. CO
conversion decreases for higher Pd : Zn molar ratios and is
equal to 26% for the 8.8Pd/Al2O3 catalyst with a Pd : Zn =
1 : 0. As evident from Fig. 9(a) there exists an optimum in
Pd : Zn ratio for CO conversion. The higher CO conversion
observed for the 8.8Pd/ZnO/Al2O3�0.38 catalyst, compared
to the 8.8Pd/ZnO/Al2O3�0.25 catalyst, is probably due to a
higher PdZn dispersion (see Table 3). From Fig. 9(a), it can be
seen that for Pd : Zn 4 0.25 : 1, the CO conversion decreases
with an increase in Pd : Zn ratio. The IR measurements have
shown an increase of the amount of Pdo with the increase of
the Pd : Zn ratio. Consequently, these results strongly suggest
Pdo particles to be less active than PdZn particles for the
synthesis of methanol.
The effect of Pd : Zn molar ratio on product selectivity is
displayed in Fig. 9(b). The CO2 selectivity is similar for all
Pd : Zn molar ratios investigated and is equal to B54%. One
can also see that the CH4 and DME selectivities follow
opposite trends. Indeed, the DME selectivity decreases from
24.5% to 10.5% with increasing Pd : Zn molar ratio. Consis-
tent with this, a significant increase of the CH4 selectivity,
from 13 to 33%, is observed with increasing Pd : Zn molar
ratio. As the Pd : Zn ratio increases from Pd : Zn = 0.38 to
1.0 more Pdo sites are present on the surface of the catalyst, as
shown by the IR results described above. Hence, these results
show Pdo to facilitate CH4 formation.
Methanol dehydration to DME is catalyzed by acid cata-
lysts. Al2O3 alone is active for the formation of DME from
methanol.25 However, Pd/ZnO is inactive for the dehydration
of methanol to DME.26,27 It is thus reasonable to assume that
for the Pd/ZnO/Al2O3 catalysts, the alumina support is the
source of acidity. NH3-TPD experiments were conducted to
determine the concentration of the acid sites for the
8.8Pd/ZnO/Al2O3 catalysts and the 8.8Pd/Al2O3 catalyst. The
NH3-TPD profiles (not shown) have indicated the presence of
one single peak located at 180 1C for all the Pd/ZnO/Al2O3
catalysts. Fig. 11(a) shows the evolution of the amount of NH3
desorbed as a function of the Pd : Zn ratio (for the catalyst
with 8.8% Pd loading). The amount of NH3 desorbed
decreases with increasing Pd : Zn ratio and is the lowest for
the 8.8Pd/Al2O3 sample (with Pd : Zn = 1 : 0). This signifies
that the concentration of acid sites decreases with increasing
Pd : Zn ratio. Since the dispersion increases with the Pd : Zn
ratio (see Table 3), the coverage of the Al2O3 support increases
and the number of accessible acid sites decreases. Note that
there is a correlation between the DME selectivity and the
amount of acid sites. It indicates that for the 8.8Pd/ZnO/Al2O3
catalysts, the acid sites of the Al2O3 support are active for the
dehydration of methanol to DME.
5.6 Effect of the Pd/Zn loading on the reactivity of the
Pd/ZnO/Al2O3 catalysts with Pd :Zn = 0.25 : 1. Catalytic
activity of several Pd/ZnO/Al2O3 catalysts with the same
Pd : Zn molar ratio (Pd : Zn = 0.25 : 1) and different Pd
loadings, varying from 2.5 to 20 wt%, were examined. The
results obtained at 380 1C and P = 69 bar are presented in
Fig. 10. CO conversion ranges from 41% to 47% for all the
Pd loadings tested. As shown in Fig. 10(b) the CO2 selec-
tivity is somewhat stable for all the Pd loadings tested and
methanol selectivity is less than 7% for all the catalysts.
Fig. 9 Evolution of (a) the conversion and (b) the selectivity with the Pd : Zn molar ratio for the supported 8.8% Pd catalysts. Reaction T =
380 1C, P = 69 bar, H2/CO = 2 (premix 1) and GHSV = 10 000 h�1.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 10
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2125
Interestingly, DME selectivity goes through a maximum of
28% at 5% Pd loading. This trend in DME selectivity is
opposite to the trend in CH4 selectivity. The CH4 selectivity
is at its lowest (11.2%) for 5Pd/ZnO/Al2O3�0.25. Contrary to
5Pd/ZnO/Al2O3�0.25, the IR spectra recorded for 2.5Pd/
ZnO/Al2O3�0.25 suggest the presence of Pd1. The higher
CH4 selectivity observed for 2.5Pd/ZnO/Al2O3�0.25, com-
pared to 5Pd/ZnO/Al2O3�0.25 is thus attributed to the
presence of Pd1. Since the CH4 selectivity increases with the
Pd loading for Pd Z 5% and no Pd1 was detected by IR
Fig. 10 Evolution of (a) the conversion and (b) the products selectivity as a function of the Pd loading for the Pd/ZnO/Al2O3 catalysts with a
Pd : Zn molar ratio equal to 0.25 : 1. Reaction T = 380 1C, P = 69 bar, H2/CO = 2 (premix 1) and GHSV = 10000 h�1.
Fig. 11 Evolution of the DME selectivity and the amount of NH3 desorbed from the catalysts surface as a function of (a) the Pd : Zn molar ratio
and (b) the Pd loading. Reaction T = 380 1C, P = 69 bar, GHSV = 10 000 h�1 and H2/CO = 2 (premix 1).
Fig. 12 Effect of the feed gas composition on (a) the conversion and (b) on the selectivity for the 5Pd/ZnO/Al2O3�0.25 catalyst. For H2/CO= 1
(premix 2), for H2/CO = 2 (premix 3) and for H2/CO = 3 (premix 4). Reaction T = 380 1C, P = 69 bar and GHSV = 10 000 h�1.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 11
2126 Catal. Sci. Technol., 2012, 2, 2116–2127 This journal is c The Royal Society of Chemistry 2012
spectroscopy for the higher loadings one can speculate that the
CH4 formation is facilitated on bigger PdZn particles.
Fig. 11(b) shows the NH3 TPD results as a function of Pd
loading. The amount of NH3 desorbed increases with the Pd
loading from 2.5 to 8.8% Pd and decreases for the highest
loading. One can see that the DME selectivity and the amount
of NH3 desorbed (i.e. amount of acid sites) from the catalyst
surface, follow the same trend with the increase of the Pd
loading. This further confirms the acid sites of the catalysts are
active for the production of DME.
From the analysis of the influence of the Pd loading and
Pd : Zn molar ratio on the reactivity of the Pd/ZnO/Al2O3
catalysts, we can conclude that under these reaction conditions
the sample with 5% Pd and a Pd : Zn molar ratio equal to
0.25 : 1 is preferred for the syngas conversion to methanol
and DME.
5.7 Effect of the feed syngas ratio on the catalytic activity of
the 5Pd/ZnO/Al2O3�0.25 catalyst. The effect of feed syngas
ratio on the catalytic performance of this composition is
examined on 5Pd/ZnO/Al2O3�0.25, identified above as the
most selective catalyst under these conditions. H2/CO molar
ratios of 1.0, 2.0, and 3.0 were evaluated and the results are
shown in Fig. 12. CO conversion increases whereas the CO2
selectivity decreases with the increase of the H2/CO ratio
[Fig. 12(a)]. DME selectivity increases from approximately
11% to 32% when increasing the H2/CO ratio from 1.0 to 3.0.
The methanol selectivity is approximately 4% at H2/CO = 1
then approaches zero at higher H2/CO ratios. Interestingly,
CH4 production also decreases with feed syngas ratio, from
approximately 6% to 2% for H2/CO= 1.0 and H2/CO= 3.0,
respectively. Methanol synthesis and subsequent methanol
dehydration is thus favored over CO methanation or DME
hydrogenolysis under these higher hydrogen-containing feeds.
Conclusion
In this work we have demonstrated the use of a Pd/ZnO/Al2O3
catalyst for the high-temperature production of methanol and
DME from syngas and contrasted its activity with that of a
commercial Cu/ZnO/Al2O3 catalyst. The PdZn-formulation
outperforms the Cu-based formulation at these conditions
which are directly relevant to the potential single-step syngas-
to-gasoline concept, which combines methanol/DME synth-
esis with MTG in a single reactor. By studying the influence of
the Pd loading and Pd : Zn molar ratio, a catalyst with 5% Pd
and a Pd : Zn molar ratio of 0.25 : 1 has been identified as the
best performing catalyst. Since Pdo promotes the formation of
methane over methanol it is critical that catalyst synthesis
avoid generation of Pdo. A direct relationship between DME
selectivity and concentration of acid sites was shown. Hence,
two types of sites are required for the direct conversion of
syngas to DME: (1) PdZn particles which are active sites for
the synthesis of methanol from syngas, and (2) acid sites which
are active for the conversion of methanol to DME. The results
of this study have shown that a non negligible amount of
undesired CH4 is produced during reaction at high tempera-
ture of operation. Also, under the conditions tested, the PdZn
particles were quite active for the water-gas-shift reaction,
leading to the formation of a large amount of CO2 at the
expense of DME formation. To consider Pd/ZnO/Al2O3 as a
catalyst for direct conversion of syngas to gasoline it will be
necessary to further investigate the parameters that could
favor the methanol synthesis reaction and further lower
methane formation.
Acknowledgements
The authors gratefully acknowledge financial support for this
work provided by the Energy Conversion Initiative, funded
internally by the Pacific Northwest National Laboratory
(PNNL). This work was also supported by the National
Advanced Biofuels Consortium (NABC) which is funded by
the Department of Energy’s Office of Biomass Program with
recovery act funds. PNNL funding was provided under
contract DE-AC05-76RL01830. Finally, the authors would
like to acknowledge that a portion of this work was done in
the Environmental Molecular Sciences Laboratory (EMSL),
a DOE sponsored user facility located at PNNL in
Richland, WA.
References
1 F. Keil, Microporous Mesoporous Mater., 1999, 29, 49–66.2 T. Mokrani and M. Scurell, Catal. Rev. Sci. Eng., 2009, 51, 1–145.3 J. Hu, K. Brooks, J. Holladay, D. Howe and T. Simon, Catal.Today, 2007, 125, 103–110.
4 S. H. Ahn, S. H. Kim, K. B. Jung and H. S. Hahm, Korean J.Chem. Eng., 2008, 25, 466–470.
5 C. Arcoumanis, C. Bae, R. Crookes and E. Kinoshita, Fuel, 2008,87, 1014–1030.
6 S. Jiang, J. S. Hwang, T. Jin, C. Tianxi, W. Cho, Y. S. Baek andS. E. Parkd, Bull. Korean Chem. Soc., 2004, 25, 185–189.
7 M. Mollavali, F. Yaripou, H. Atashi and S. Sahebdelfar, Ind. Eng.Chem. Res., 2008, 47, 3265–3273.
8 J. Topp-Jorgensen, in Methane Conversion, ed. D. M. Bibby,C. D. Chang, R. F. Howe and S. Yurchak, Elsevier SciencePublishers, 1988, p. 293.
9 K. Fujimoto, H. Saima and H. Tominaga, J. Catal., 1985, 94,16–23.
10 J. Erena, J. M. Arandes, J. Bilbao, M. Olazar and H. I. de Lasa,J. Chem. Technol. Biotechnol., 1998, 72, 190–196.
11 C. Satterfield, in Heterogeneous Catalysis in Industrial Practice, ed.C. Satterfield, Krieger Publishing Company, Malabar, Florida,2nd edn, 1996, p. 446.
12 Y.-H. Chin, R. Dagle, J. Hu, A. C. Dohnalkova and Y. Wang,Catal. Today, 2002, 77, 79–88.
13 N. Iwasa, S. Masuda, N. Ogawa and N. Takezawa,Appl. Catal., A,1995, 125, 145.
14 R. Dagle, A. Platon, D. Palo, A. Datye, J. Vohs and Y. Wang,Appl. Catal., A, 2008, 342, 63–68.
15 J. Hu, R. Dagle, J. Holladay, C. CaO, Y. Wang, J. White,E. Douglas and D. Stevens, U.S. Patent, 7 858 667, 2010.
16 D. Palo, R. Dagle and J. Holladay, Chem. Rev., 2007, 107,3992–4021.
17 R. A. Dagle, J. A. Lizarazo, L. V. M., M. J. Gray, J. F. White,D. L. King and D. R. Palo, Submitted to Applied Catalysis A:General, 2012.
18 Y. Zhu, S. B. Jones, M. J. Biddy, R. A. Dagle and D. R. Palo,Bioresour. Technol., 2012, 117, 341–351.
19 G. Xia, J. D. Holladay, R. A. Dagle, E. O. Jones and Y. Wang,Chem. Eng. Technol., 2005, 28, 515.
20 M. Boudart and G. Djega-Mariadassou, The Kinetics of Hetero-geneous Catalytic Reactions, Princeton University Press, Princeton,NJ, 1984.
21 N. Sheppard and T. T. Nguyen, in Advances in Infrared and Ramanspectroscopy, ed. R. E. Hester and R. J. H. Clarke, Heyden andSon, London, 1978, vol. 5, p. 67.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online
Page 12
This journal is c The Royal Society of Chemistry 2012 Catal. Sci. Technol., 2012, 2, 2116–2127 2127
22 V. Lebarbier, R. A. Dagle, T. Conant, J. M. Vohs, A. Datye andY. Wang, Catal. Lett., 2008, 122, 223–227.
23 J. T. Sun, I. S. Metcalfe and M. Sahibzada, Ind. Eng. Chem. Res.,1999, 38(10), 3868–3872.
24 F. S. Ramos, A. M. D. de Farias, L. E. P. Borges, J. L. Monteiro,M. A. Fraga, E. F. Sousa-Aguiar and L. G. Appel, Catal. Today,2005, 101, 39–44.
25 F. Yaripour, F. Baghaei, I. Schmidt and J. Perregaard, Catal.Commun., 2005, 6, 147–152.
26 N. Iwasa, O. Yamamoto, T. Akazawa, S. Ohyama andN. Takezawa, J. Chem. Soc., Chem. Commun., 1991, 18,1322–1323.
27 Y. A. Ryndin, R. F. Hicks, A. T. Bell and Y. I. Yermakov,J. Catal., 1981, 70, 287–297.
Publ
ishe
d on
18
June
201
2. D
ownl
oade
d on
04/
11/2
014
13:4
8:21
. View Article Online