An intermetallic Pd2Ga nanoparticle catalyst for the single-step conversion of CO-rich synthesis gas to dimethyl ether

Article abstract of DOI:10.1016/j.apcata.2018.04.018

Well-defined Pd/Ga-nanoparticles were prepared and used as a precursor for the methanol active component in a bifunctional syngas-to-dimethyl ether catalyst. In situ X-ray absorption spectroscopy experiments were employed both to unravel the initial formation of the active catalyst phase in reductive H₂ atmosphere and to further monitor changes of the nanoparticles under conditions of dimethyl ether synthesis at a pressure up to 20 bar (250 °C). The catalytic studies were conducted using simulated biomass-derived, CO-rich syngas in a continuous-flow reactor, with the bifunctional catalyst offering the two types of active sites, i.e. for methanol synthesis (Pd/Ga nanoparticles) and its subsequent dehydration (γ-Al₂O₃), in close proximity. As compared to the conventional Cu/Zn-based reference catalyst prepared via a similar procedure, the Pd/Ga-based catalyst showed a promising activity together with a notable stability with time on stream and a high temperature tolerance (up to 300 °C). A kinetic model which considers the individual reactions involved in direct DME synthesis based on power law equations was used to fit the experimental data, and the apparent activation energies were compared to the Cu/Zn-based catalyst.

Full text of DOI:10.1016/j.apcata.2018.04.018

Accepted Manuscript
Title: An intermetallic Pd2Ga nanoparticle catalyst for the
single step conversion of CO-rich synthesis gas to dimethyl
ether
Authors: Manuel Gentzen, Dmitry E. Doronkin, Thomas L.
Sheppard, Jan-Dierk Grunwaldt, Jo¨rg Sauer, Silke Behrens
PII:
S0926-860X(18)30190-X
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.04.018
APCATA 16624
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
9-2-2018
9-4-2018
16-4-2018
Please cite this article as: Gentzen M, Doronkin DE, Sheppard TL, Grunwaldt J-Dierk,
Sauer J, Behrens S, An intermetallic Pd2Ga nanoparticle catalyst for the single step
conversion of CO-rich synthesis gas to dimethyl ether, Applied Catalysis A, General
(2010), https://doi.org/10.1016/j.apcata.2018.04.018
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

An intermetallic Pd2Ga nanoparticle catalyst for the single step conversion of CO-rich
synthesis gas to dimethyl ether
Manuel Gentzen,a Dmitry E. Doronkin,a,b Thomas L. Sheppard,b Jan-Dierk Grunwaldt,a,b Jörg
Sauer,a Silke Behrens*a
a
Institute of Catalysis Research and Technology, Karlsruhe Institute of Technology (KIT),
Herrmann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany;
b
Institute of Technical Chemistry and Polymer Chemistry, Karlsruhe Institute of
Technology (KIT), Engesserstr. 20, 76131 Karlsruhe, Germany
*Corresponding author: Silke Behrens, E-mail: silke.behrens@kit.edu
Graphical abstract
Highlights
• A synthetic procedure for uniform and well-defined Pd/Ga-based nanoparticles
was established.
• A bifunctional catalyst affording the close proximity of two catalytic functions
(i.e., for methanol synthesis and its consecutive dehydration to dimethyl ether)
was derived from the uniform Pd/Ga-based nanoparticle precursors.
• The active Pd/Ga phase was identified by in situ X-ray absorption spectroscopy
studies under reducing H2 atmosphere and reaction conditions (syngas, 250°C,
20 bar).
1

• The Pd/Ga nanoparticle-derived bifunctional catalyst shows significant promise
for the syngas-to-dimethyl ether process in terms of activity and stability.
Abstract
Well-defined Pd/Ga-nanoparticles were prepared and used as a precursor for the methanol
active component in a bifunctional syngas-to-dimethyl ether catalyst. In situ X-ray absorption
spectroscopy experiments were employed both to unravel the initial formation of the active
catalyst phase in reductive H2 atmosphere and to further monitor changes of the nanoparticles
under conditions of dimethyl ether synthesis at a pressure up to 20 bar (250 °C). The catalytic
studies were conducted using simulated biomass-derived, CO-rich syngas in a single
continuous-flow reactor and the bifunctional catalyst offering the two types of active sites, i.e.
for methanol synthesis (Pd/Ga nanoparticles) and its subsequent dehydration (γ-Al2O3), in
close proximity. As compared to the conventional Cu/Zn-based reference catalyst prepared
via a similar procedure, the Pd/Ga-based catalyst showed a promising activity together with a
notable stability with time on stream and a high temperature tolerance (up to 300°C). A
kinetic model considering the individual reactions involved in direct DME synthesis based on
power law equations was used to fit the experimental data, and the apparent activation
energies were compared to the Cu/Zn-based catalyst.
Keywords: Single-step dimethyl ether synthesis, simulated biomass-derived syngas, colloidal
nanoparticles, Pd2Ga, kinetic model
1. Introduction
Dimethyl ether (DME) exhibits unique properties and is used for a broad range of
applications, e.g. as propellant, as fuel for gas turbines, for domestic applications like cooking
and heating or as diesel substitute [1, 2]. DME synthesis from syngas is typically a two-step
process, where methanol is obtained over Cu/ZnO-based catalysts in the first stage, which is
successively dehydrated to DME over inorganic solid acids (e.g., -Al2O3 or zeolites) in the
second stage. Alternatively, DME can also be produced from synthesis gas in a single step in
the syngas-to-dimethyl ether (STD) process. The overall conversion of syngas to DME is
described by Equation (1).
3 CO + 3 H2 ⇌ CH3OCH3 + CO2
(1)
A renewable approach for supplying syngas follows the gasification of biomass-derived
feedstocks [3]. In contrast to fossil resources, the application of biomass as feedstock yields
CO-rich syngas and H2:CO ratios between 0.7 and 1.2 [4].
Usually, for the direct synthesis of DME from syngas, the methanol and the dehydration
catalyst are mechanically mixed [5, 6, 7]. However, these systems often suffer from the large
distance between the two catalytically active sites, low activities and low lifetimes [8]. The
lifetime of industrially applied catalysts for the synthesis of methanol is an important factor,
and Cu/ZnO-based methanol catalysts are known to deactivate with time on stream (TOS).
Several mechanisms have been suggested for catalyst deactivation, including sintering of Cu
particles [9], carbon deposition [10] and the formation of a dense crystalline ZnO overlayer on
the Cu surfaces [11]. Copper has a low Hüttig temperature which is reflected by a relatively
low melting point (i.e., bulk Cu 1083 °C), and thus, Cu catalysts are operated typically at
relatively low temperatures (240 to 260 °C). Higher temperatures may already cause severe
2

sintering of the Cu nanoparticles. ZnO is known to display basic O2- sites and thus,
neutralization of some acid sites may occur when the Cu(ZnO) system is brought in intimate
contact with the dehydration catalyst. Moreover, migration and ion exchange of, e.g. Cu2+
species with the hydroxyl groups was shown to lead to a decrease in the concentration of
zeolite Brönsted acid sites and has also been considered as a potential mechanism for the
deactivation of STD catalysts during reaction [12]. Bifunctional catalysts with a close
proximity of the two types of catalytically active sites showed a superior performance as
compared to the physically admixed systems [13, 14]. Besides the conventional Cu/ZnO-
based materials, Pd-based catalysts are known to catalyze methanol formation from synthesis
gas [15, 16]. In particular, intermetallic Pd2Ga was reported to reach high activities in
methanol synthesis from H2/CO2 feedstocks with a molar H2:CO2 ratio of 3:1 [17, 18, 19].
Pd2Ga particles supported on SiO2, for example, were used for the low-pressure, gas-phase
hydrogenation of CO2 to methanol [19]. The formation of Ga-Pd intermetallics and their
effects on methanol synthesis by CO and CO2 hydrogenation were also studied by Behrens et
al. [20]. Pd2Ga-based catalysts have typically been prepared via impregnation [19, 21] or co-
precipitation [22], and, up to now, only few authors have described the colloidal synthesis of
Pd2Ga-based nanoparticles [23, 24]. These colloids were used, e.g., for the liquid-phase
hydrogenation CO2 to methanol [23].
In general, the application of colloidal nanoparticles for the synthesis of methanol or DME
has often been restricted to reactions in the liquid phase with few exceptions where, e.g., the
colloidal Cu/Zn nanoparticles were immobilized on a support [25, 26]. In the precursor
concept, uniform and well defined nanoparticles as obtained from colloidal preparation
techniques are applied as precursors for the preparation of heterogeneous catalysts.
Bifunctional STD catalysts have been prepared, e.g. via mixed metal impregnation, co-
precipitation impregnation, or co-precipitation sedimentation of the dehydration catalyst [14].
In these procedures, the dehydration catalyst is typically exposed to various metal ion
precursors and precipitating agents at different pH values which influence not only the
dispersion of the methanol active component but also the acid sites and thus, the dehydration
characteristics of the acid solid. Moreover, individually optimized synthetic recipes are
typically required which influence catalyst homogeneity and dispersion, and the resulting
differences may additionally complicate the comparison of the catalyst performance. By
exploiting colloidal approaches via the precursor concept, however, the size and shape of the
methanol active component may be separately tailored, which avoids mutual detrimental
interactions during catalyst preparation and reduces the influence of preparation history.
Therefore, the use of nanoparticle building units may further contribute to establish structure-
performance relationships and thus, to a more fundamental understanding in heterogeneous
STD catalysts.
The kinetics of methanol synthesis, methanol dehydration and the STD reaction have been
described in several studies, where complex models with rate expressions, e.g., based on
Langmuir-Hinshelwood or elementary steps were applied [27, 28, 29, 30]. Typically, higher
H2:CO ratios of 2 to 4 were employed in the case of the single step DME synthesis [31, 32,
33]. In general, a higher H2:CO ratio leads to higher CO-conversions since the methanol
synthesis is assumed to be the rate limiting step and has a higher reaction order than the
competing water gas shift reaction [7, 34]. Furthermore, the methanol active components used
in most studies were Cu/ZnO-based systems, which were either commercially available or
prepared via precipitation techniques [35, 36, 37].
In this study, bifunctional STD catalysts were prepared using colloidal Pd/Ga-based
nanoparticles as building units. A synthetic route was developed for well-defined and uniform
Pd/Ga-based nanoparticles which were subsequently immobilized on the dehydration catalyst.
3

This procedure enabled the intimate contact of the two types of active sites in the bifunctional
STD catalyst (i.e., the Pd/Ga-based nanoparticulate methanol active component and the -
Al2O3 dehydration catalyst). The nanoparticles and the resulting bifunctional catalysts were
characterized by transmission electron microscopy (TEM) with energy dispersive X-ray
spectroscopy (EDX), N2 physisorption, powder X-ray diffraction (XRD), and in situ X-ray
absorption spectroscopy (XAS). The formation of the active catalyst phase was monitored
under reductive H2 atmosphere by in situ XAS studies. For the first time, in situ XAS
experiments were also carried out at a pressure up to 20 bar in a quartz capillary reactor cell
closing the pressure gap in characterization under conditions of DME synthesis (15 vol.-% H2,
15 vol.-% CO). The catalytic performance in the STD process was investigated in a fixed bed
reactor using a simulated, biomass-derived syngas (H2:CO ratio of 1:1) between 250 and
300 °C. For the first time, we report on the modeling of the direct synthesis of DME for a
Pd/Ga-based catalyst. The experimental data were fitted using a reaction network based on
power-law models. The resulting kinetic parameters were compared to a catalyst of the
conventional Cu/ZnO-based reference system.
2. Materials and Methods
All procedures of Ga(butyl)3 and nanoparticle synthesis were carried out under argon
atmosphere using standard Schlenk techniques.
2.1 Chemicals
Palladium(II) acetylacetonate (Pd(acac)2, Alfa Aesar, ≥ 99.9 % trace metal basis), gallium
trichloride (GaCl3, ABCR, 99.999 %, ultra dry), toluene (Sigma Aldrich, 99.8 %), n-
butyllithium (Li(C4H9), ACROS Organics, 2.7 M in toluene, 19.8-20.2 (as BuLi)), γ-Al2O3
(Alfa Aesar) were used as received without any further purification.
2.2 Synthesis of Ga(butyl)3
Ga(butyl)3 was obtained using a modified procedure originally described by Kovar and
coworkers [38]. Briefly, GaCl3 (21.1 g, 120 mmol) was dissolved in 100 mL toluene. A
solution of n-butyl lithium in toluene (134 mL, 362 mmol n-BuLi) was added over a period of
90 minutes at 90 °C. The resulting mixture was stirred for another 20 h at 90 °C. The
precipitated LiCl was removed by filtration over a glass-frit (Por. 4). The product was
obtained as a yellow, clear solution in toluene (solution 1) and used without further
purification.
2.3 Synthesis of Pd/Ga-based nanoparticles
Pd(acac)2 (3.02 g, 10 mmol) was suspended in 200 mL toluene. 57 mL of solution 1 was
added over a period of 90 minutes at 40 or 60 °C. After a few minutes, the reaction mixture
turned black and was stirred for another 24 h at the reaction temperature. Eventually, the
solvent was removed in vacuum, and the Pd/Ga-based nanoparticles were obtained as black,
viscous product.
2.4 Preparation of the bifunctional STD catalysts
To prepare the bifunctional STD catalysts, the Pd/Ga-based nanoparticles were redispersed in
50 mL toluene, and the γ-Al2O3 dehydration catalyst was successively added to the
nanoparticle suspension. After stirring for 15 min, the solvent was removed in vacuum, and
the resulting powder was calcined in air at 350 °C for 4 h (heating rate 110 °C/h). The final
catalyst (with 7.6 wt.-% Pd and 16.5 wt.-% Ga according to EDX analysis) was obtained by
4

pelletizing, crushing, and sieving to the fraction of 80 – 160 µm. The Cu/Zn-based reference
catalyst was obtained using a similar procedure [25, 26].
2.5 Characterization
The nanoparticles were analyzed with a transmission electron microscope (TEM, FEI Tecnai
F20 ST TEM, operating voltage 200 kV) and elemental analysis (EDX). The TEM was
equipped with a field emission gun and an EDAX EDS X-ray spectrometer (Si(Li) detecting
unit, super ultra-thin window, active area 30 mm2, resolution 135 eV (at 5.9 keV)). For TEM
analysis, a small droplet of the dispersed nanoparticles was deposited on amorphous carbon-
coated 400 mesh Cu grids and dried in air. The mean particle diameter was calculated based
on TEM images by measuring the size of at least 100 particles. The composition of the
bifunctional catalyst was analyzed with a DSM 982 Gemini SEM, equipped with a Schottky-
type thermal field emission cathode (Zeiss corp., Germany) and an EDX.
Powder x-ray diffraction (XRD) patterns were recorded on a PANalytical X’Pert Pro X-ray
diffractometer employing a Bragg-Brentano geometry with Cu Kα radiation and a Ni filter.
The range between 5° and 120° was measured within 1 h. The diffraction peaks were
compared to reference compounds reported in the Inorganic Crystal Structure Database
(ICSD) and the Joint Committee of Powder Diffraction Standards (JCPDS). The crystallite
size Lhkl was calculated for the reflection at 48.3° (2θ) of the Pd2Ga phase using the Scherrer
equation
0.9 λ
(2)
Lhkl=
β cos⁡(θ)
where β is the full width at half maximum and θ the Bragg angle. LaB6 (NIST) was used as a
standard in order to account for instrumental broadening. After reaction, the catalyst was
deposited without air contact in the anoxic glove box onto a XRD sample holder which was
covered by Kapton foil.
The specific total surface areas of the catalysts were determined by N2-physisorption
measurements using a Quantachrome Nova 2000e instrument. The calcined catalyst was
degassed in vacuum at 230 °C for at least 12 h. Temperature programmed ammonia
desorption was performed (Micromeritics Autochem 2950) on the initially reduced catalysts
(2 °C/min, 250 °C, 2 vol.-% H2 in Ar). The catalysts were loaded with ammonia (1.221 ±
0.024 vol.-% NH3 in He, 30 mL/min) at 100 °C. Afterwards, physisorbed ammonia was
removed at 120 °C in He flow (30 mL/min). Ammonia desorption was monitored with a mass
spectrometer (MS, MKS Cyrrus2) between 120 and 750 °C (4 °C/min) in a He flow (30
mL/min).
In situ X-ray absorption spectra (XAS) in terms of the X-ray absorption near-edge structure
(XANES) and extended X-ray absorption fine structure (EXAFS) were measured at the CAT-
ACT beamline of KIT’s synchrotron radiation source ANKA in fluorescence mode using a
PIPS diode [39]. The incident energy was selected by a double crystal monochromator with a
pair of Si (111) or Si (311) crystals to scan energies around Ga K or Pd K edges,
correspondingly. Higher harmonics were rejected by a pair of Si or Pt/Ir coated mirrors for Ga
K or Pd K edges, respectively. The beam size at the sample was set to 1 (horizontal) × 0.5
(vertical) mm2. The catalysts (approx. 6 mg, pressed and sieved to 80-160 μm grains) were
placed in in situ microreactors (quartz capillary with a plug flow reactor geometry, 1.5 mm
diameter, 20 μm wall thickness, catalyst bed length approx. 5 mm) between two quartz wool
plugs and mounted horizontally over a hot air blower (FMB Oxford GSB-1300) [40].
Temperature was calibrated using readings of an external thermocouple at the top and the
bottom of the capillary microreactor in a separate test without X-ray beam. He or 5 vol.%
5

H2/He were dosed using individual mass flow controllers at a flow rate of 50 mL/min, and the
outlet gas was analyzed using a Pfeiffer Vacuum OmniStar GSD320 mass spectrometer.
Initially, XANES and EXAFS spectra were recorded at 25 °C in He flow, then the gas
mixture was switched to 5 vol.-% H2/He (50 mLNTP/min, ambient pressure), and XANES
spectra were recorded continuously during heating to 250 °C at a ramp rate 1 °C/min. After
reaching 250 °C, the catalyst was kept for 60 min in 5 vol.-% H2/He, after which another set
of XANES and EXAFS spectra were recorded (5 vol.-% H2/He, 250 °C). The experiments
were repeated twice while recording either spectra at Pd K edge (the catalyst was not diluted
in this case) or at Ga K edge (the catalyst was diluted 2 times with γ-Al2O3 prior to pressing
and sieving to diminish incident beam self-absorption). After the TPR experiments, the
reaction conditions of the direct DME synthesis were applied to the in situ microreactor
(15 vol.-% H2, 15 vol.-% CO in He, 250 °C, 18 mLNTP/min, 20 bar) and maintained for one
hour. The product formation was monitored by a mass spectrometer. After 1 h, an additional
XAS spectrum at the Pd K edge was recorded. The spectra were normalized, and background
subtracted to extract EXAFS using the ATHENA program from the IFFEFIT software
package [41]. The spectra of the catalyst measured at Ga K edge were corrected for incident
beam self-absorption using the Fluo algorithm built in the ATHENA program [41]. The k1-,
k2-, and k3-weighted EXAFS functions were Fourier transformed in the k range of 3.0 –
11 Å−1 and multiplied by a Hanning window with sill size of 1 Å-1. The structural models
were based on bulk Pd metal (ICSD collection code 52251) and Pd2Ga alloy (ICSD collection
code 409939). The structure refinement was performed using ARTEMIS software (IFFEFIT)
[41]. For this purpose, the corresponding theoretical backscattering amplitudes and phases
were calculated by FEFF 6.0 [42]. The theoretical data were then adjusted to the experimental
spectra by a least square method in R space between 1.5 and 3.0 Å-1. First, the amplitude
2
reduction factors (S0 = 0.58 for Pd) were calculated using the Pd foil reference spectra, and
then, the coordination numbers (CN), interatomic distances, energy shift (δE0) and mean
square deviation of interatomic distances (σ2) were refined. The absolute misfit between
theory and experiment was expressed by ρ.
2.6 Catalytic testing
The catalytic properties were evaluated in a plug flow reactor consisting of a stainless steel
tube with a length of 250 mm and an inside diameter of 8 mm. An alloy (stainless steel
V4A 1.4571) was chosen that doesn’t promote methanation reactions. Gases (CO 3.7, H2 6.0,
Ar 6.0, N2 6.0) were dosed using mass flow controllers (Bronkhorst). Effluent gases were
analyzed with a gas chromatograph (GC, Hewlett Packard 6890). Two different columns
(RESTEK RT®-U-Bond, RESTEK RT-Msieve) as well as two detectors (thermal
conductivity detector, flame ionization detector) were employed to analyze the gas mixtures.
The catalysts were activated in the reactor employing diluted H2 and low heating rates to limit
adiabatic temperature rise, which may cause particle sintering. For each test, 2 g of the
calcined catalyst was initially activated in situ via the following reduction program: (1)
heating (17 °C/h) in 2 vol.-% H2 in Ar (100 mLNTP/min) to 200 °C which was maintained for
1 h, (2) heating (17 °C/h) to 240 °C which was maintained for 1 h, and (3) heating (10 °C/h)
in pure H2 (50 mLNTP/min) to 250 °C which was maintained for 1 h. Subsequently the reaction
pressure was set to 50 bar, and the educt gas (total flow 50 mLNTP/min, composition of
Ar:N2:H2:CO = 5:2:1.5:1.5) was introduced. The temperature was increased every 6 h by
10 °C (heating rate 10 °C/h) until the final temperature of 300 °C was reached. The axial
temperature profile of the catalytic bed was monitored with a thermocouple which was
applied within a guiding sleeve in the middle of the reactor (maximum temperature deviation
± 0.2 K). The reactor, the thermowell and the inert material (SiC) did not reveal any catalytic
activity in synthesis gas conversion. The catalytic performance was evaluated using the
following equations:
6

̇
̇
NCO,0 - NCO
(3)
XCO=
∙100
̇
NCO,0
̇
̇
with NCO,0 molar flow of carbon monoxide at reactor inlet, NCO molar flow of carbon
monoxide at reactor outlet, and XCO CO conversion in %
̇
̇
ξi(Ni −⁡Ni,0)
(3)
Si=
∙100
̇
̇
NCO,0 - NCO
̇
̇
with Si selectivity of component i, Ni mole flow of component i at reactor outlet, Ni,0 mole
flow of component i at reactor inlet and ξi number of carbon atoms in a molecule of
component i.
2.7 Kinetic modelling
The following set of reaction equations was considered for fitting the experimental data using
Matlab® (version 2016b, The Mathworks GmbH).
kJ
R1
CO + 2 H2 ⇌ CH3OH
2 CH3OH ⇌ CH3OCH3 + H2O
CO + H2O ⇌ CO2 + H2
(5)
(6)
(7)
(8)
(9)
∆RH0 = - 90.5
∆RH0 = - 23.9
∆RH0 = - 41.2
∆RH0 = - 206.1
∆RH0 = - 164.9
mol
kJ
R2
R3
mol
kJ
mol
kJ
R4
R5
CO + 3 H2 ⇌ CH4 + H2O
CO2 + 4 H2 ⇌ CH4 + 2 H2O
mol
kJ
mol
Equation (10) shows a differential balance for a component i in a small volume element dV of
the fixed bed reactor.
∂Ci
∂t
dmcat
dV
(
)
= - div Ci∙u⃗ + div(Di,effgrad Ci) + (∑(νi,j∙rm,j))
(10)
0
j
with Ci concentration of component i, u0 superficial velocity, Di,eff effective diffusion
coefficient of component i, νi,j stoichiometric coefficient of component i in reaction j, rm,j
specific reaction rate of reaction j, mcat catalyst mass and V reactor volume.
In an ideal plug flow reactor diffusion effects and radial flows can be neglected. The modified
residence time (τmod) is defined by Equation (11).
7

mcat
τmod
=
(11)
Areactor ∙ u0
where Areactor is the cross sectional area of the reactor.
These assumptions and inserting (11) into (10) lead to equation (12).
dCi
= (∑(νi,j∙rm,j))
(12)
dτmod⁡
j
For the reaction rates the following power law models based on the fugacities (fi) were
applied.
fMeOH
γMeOH
2
(13)
(14)
(15)
(16)
rR1= k1 ∙
rR2= k2 ∙
rR3= k3 ∙
rR4= k4 ∙
f
CO ∙ fH
-
2
K1 ∙
2
γCO ∙ γH
2
(
)
fDME ∙ fH O
2
2
f
⁡-
γH O ∙ γDME
MeOH
2
K2 ∙
2
γMeOH
(
)
fCO ∙ fH
2
γCO ∙ 2γH
f
CO ∙ fH O⁡-
2
2
2
K3 ∙
γCO∙ γH O
2
(
)
fH O ∙ fCH
3
2
4
f
CO ∙ fH ⁡-
γH O⁡∙ γCH
2
2
4
K4 ∙
3
γCO∙ γH
2
(
)
fH O2∙fCH
4
2
4
rR5= k5 ∙ fCO ∙ fH
-
(17)
γH O2∙ γCH
2
2
2
4
K5 ∙
4
γCO ∙ γH
(
)
2
2
8

The equilibrium constants were multiplied with the fugacity coefficients to account for the
influence of the non-ideal behavior at high reaction pressures.
- ∆RG
(18)
K = exp (
)
R ∙ T
Equilibrium constants were calculated for all the reactions considered within this model
according to equation (18). The equilibrium constants K1-K5 were then included in the
expressions for the reaction rates (13)-(17) of the corresponding reaction equations and the
kinetic expressions were fitted to the experimental molar fractions of the gaseous components.
The activation energies were calculated from Arrhenius plots with a coefficient of
determination of at least 0.98.
3. Results and discussion
A synthetic procedure was developed for colloidal Pd/Ga-based nanoparticles by reacting
Pd(acac)2 with Ga(butyl)3. The synthetic procedure yielded particles in the lower nanometer
size range and could easily be scaled-up to the necessary catalyst quantities. No additional
ligands, polymers or surfactants were needed to control particle nucleation and growth and to
finally stabilize the nanoparticles. After addition of Ga(butyl)3 to the suspension of Pd(acac)2
in toluene, a black, colloidal solution containing small Pd/Ga-based nanoparticles was formed
from which the nanoparticles were eventually isolated as a black viscous product. The particle
size was controlled via the reaction temperature (i.e., 40 and 60 °C). A TEM image of the
nanoparticles which were obtained in the particle synthesis at 40 °C is shown in Figure 1a.
The mean diameter of the particles was 4.6 (± 1.1) nm. The TEM image displays the presence
of some particles with irregular shape and a few agglomerates. If the synthesis of the Pd/Ga-
based nanoparticles was carried out at 60 °C, well defined nanoparticles of spherical shape
were obtained (Figure 1b), and the mean particle diameter decreased to 3.4 (± 0.8) nm. No
agglomerates were observed in this case. Previously, we have described the synthesis of
Cu/Zn-based nanoparticles via a similar synthetic approach [25], where the presence of a Zn-
containing organic shell stabilized the metallic Cu nanoparticles. For the Pd/Ga-based
nanoparticles, EDX analysis also revealed a close contact between Pd and Ga (Figure 1c),
suggesting metallic Pd nanoparticles stabilized by an organic Ga-containing shell.
(a)
(b)
(c)
C
O
Ga
Cu
Cu
Pd
Ga
Cu
Si
Ga
0
2
4
6
8
10
12
Energie in keV
Figure 1. TEM images of the Pd/Ga nanoparticles synthesized at 40 °C (a) and 60 °C (b). STEM-EDX analysis of Ga/Pd
nanoparticles synthesized at 60 °C (c).
These uniform nanoparticles were then employed as building blocks for a catalyst following
the precursor concept [23, 39]. A bifunctional catalyst for the direct synthesis of DME from
9

synthesis gas was prepared via immobilization of the Pd/Ga nanoparticles on γ-Al2O3. The
supported nanoparticles were calcined to remove any organic residues from nanoparticle
synthesis. The elemental composition of the calcined, bifunctional STD catalyst was
determined by EDX analysis to 7.6 wt.-% Pd and 16.5 wt.-% Ga. The molar ratio of Pd and
Ga was 3.3, which was consistent with the Pd/Ga ratio of 3 of the precursors applied in
nanoparticle synthesis. Moreover, the EDX spectra revealed no chlorine signals for the
bifunctional catalyst, further indicating the complete reaction of the GaCl3 with butyl lithium
to Ga(butyl)3 in the first step of nanoparticle preparation. TEM images of the calcined and
reduced bifunctional catalyst (Fig. S1) again displayed the presence of individual
nanoparticles distributed over the support. The specific surface area was measured by N2
physisorption (BET method) to 131 m2/g. If only the amount of γ-Al2O3 in the bifunctional
catalyst (56.3 wt.-%) was considered, the experimental surface area was in good agreement
with the pristine γ-Al2O3 support (221 m²/g). Ammonia desorption experiments revealed
160.9 µmolNH /gcat with a main desorption temperature below 500 °C which was assigned to
3
the weakly to medium acidic sites of γ-Al2O3.
XRD analysis of the calcined, bifunctional STD catalyst showed reflections of a tetragonal
PdO phase (JCPDS 00-043-1024) and a cubic Pd phase (JCPDS 01-087-0638) (Figure 2)
indicating the partial oxidation of the immobilized Pd nanoparticles during the calcination
procedure. The remaining reflections could be assigned either to the γ-Al2O3 support or a
Ga2O3 phase (JCPDS 00-020-0426). The Ga2O3 phase featured broad reflections of low
intensity at 36.2°, 44.1°, 58.4°, and 64.2° (2θ), indicating low crystallinity and/or small
crystallite size. The reflections of the γ-Al2O3 support at 37.6°, 45.9°, 60.9°, and 67.0° (2θ)
were broad and of low intensity. After in situ activation under reducing H2 atmosphere and
catalytic testing, the catalyst was isolated in an anoxic glove box, and the diffraction pattern
of the spent catalyst was again recorded. The reflections at 39.7°, 44.7°, and 48.3° (2θ) could
be assigned to the orthorhombic Pd2Ga phase (ICSD 98-040-9939) and confirmed the
formation of an intermetallic Pd2Ga phase. The formation of a Pd2Ga phase after reduction of
Pd/Ga2O3 catalysts was previously described by Oyola-Rivera et al. [44]. Reflections of a
metallic Pd phase were not detected for the spent catalyst. The size of the Pd2Ga particles of
the spent bifunctional catalyst was estimated using the Scherrer equation to 19.8 nm
(Equation 2).
10

PdO
Pd
(a)
Pd2Ga
Ga2O3
(b)
clacined
spent
-Al2O3
40
60
80
Angle (2)
Figure 2 (a) XRD patterns of the γ-Al2O3 support, the calcined and spent bifunctional STD catalyst. (b) Illustration of the
crystal structure of the intermetallic Pd2Ga phase (Pd atoms blue, Ga atoms red).
The size of the initial Pd/Ga-based nanoparticles was 3.4 (± 0.8) nm according to TEM
analysis which suggests some particle growth during the consecutive steps of calcination,
activation and catalyst testing.
The formation of the active phase of the bifunctional catalysts in reducing H2 atmosphere was
investigated by in situ XAS experiments. Initially, a XAS spectrum was recorded at room
temperature (5 % H2/He), where PdO was completely reduced to Pd. Reduction of PdO in
SiO2 and Al2O3-supported Pd catalysts occurs between -63 and 27 °C [45]. The room
temperature XANES spectrum showed similar features as a β-PdHx phase, indicating that the
reduction of Pd proceeded via β-PdHx intermediates [46]. During the heating process, the
position of the Pd edge shifted towards lower energies for temperatures up to 150 °C
(Figure 3), indicating the change in electron density by filling of the valence d band of Pd by
covalent interaction between Pd and Ga. At the same time, the β-PdHx phase was decomposed
[46]. At temperatures above 150 °C, hardly any further changes were observed at the Pd K
edge. Measurements at the Ga K edge, however, revealed the reduction of Ga during the
whole activation process. A gradual Ga-enrichment of the nanoparticles was detectable down
to a Pd0/Ga0 ratio close to 1:1 (Pd71Ga61). Ga-enrichment seemed to be a rather gradual
process involving H2 consumption rather than a well-resolved reduction step. Fourier-
transformed, extended X-ray absorption fine structure (EXAFS) spectra are shown in Figure 4
and the refined structural parameters are summarized in Table S1. The closest neighbors in
the Pd2Ga model structure are located at 2.55 Å (Pd-Ga) and 2.82 Å (Pd-Pd), whereas the
activated bifunctional STD catalyst showed nearest neighbors at 2.53 ± 0.01 Å (Pd-Ga) and
2.79 ± 0.03 Å (Pd-Pd). Similar distances of 2.51 ± 0.01 Å (Pd-Ga) and 2.81 ± 0.02 Å (Pd-Pd)
have been reported by Fiordaliso et al. for XAS measurements of Pd2Ga compounds, which
were prepared by impregnation [19]. At 250 °C, the conditions of the STD reaction were
applied, and the educt gases were introduced at a pressure of 20 bar (15 vol.-% CO, 15 vol.-%
H2 in He). Morphological changes of the catalyst were monitored by XAS measurement (Pd
K edge) after 1 h under reaction conditions. Moderate changes included a more intense 3rd
(Ga) shell and a decrease of the Pd-Pd distance (2.77 Å, towards Pd-Pd in metallic Pd). Both
effects may be attributed to sintering and / or dealloying of the PdGa nanoparticles. XANES
11

region around Pd K edge did not show any changes, which excludes PdHx formation or
decomposition at this stage. The reoxidation of Ga species and the formation of Pd2Ga under
reaction conditions has previously been suggested to occur for Pd/Ga-based catalysts in
methanol synthesis from CO2 by Fiordaliso et al. [19].
(b)
Ga K-edge
(a)
Pd K-edge
44 °C
21 °C
73 °C
85 °C
150 °C
209 °C
250 °C
147 °C
210 °C
250 °C
24300
24350
24400
24450
10350
10400
10450
10500
Energy (eV)
Energy (eV)
Figure 3 XANES spectra of the in situ TPR experiments at Pd K (a) and Ga K (b) edges between RT and 250 °C in 5 vol.-%
H2/He (heating rate 1 °C/min).
(a)
(b)
Ga K-edge
Pd K-edge
PG (8)-A reacted
PG (8)-A reduced
PG (8)-A reduced
Pd-Pd (Metal)
PG (8)-A calcined
Ga-Ga (metal)
PG (8)-A calcined
Ga
Ga-O
Pd-O
Pd-Pd (Oxide)
Pd
Ga-Ga (oxide)
Ga2O3
PdO
0
1
2
3
4
5
0
1
2
3
4
5
R (Å)
R (Å)
Figure 4 k²-weighted Fourier-transformed EXAFS spectra of the in situ XAS experiments at the Pd K (a) and Ga K (b) edges
for the calcined catalyst, the reduced catalyst at 250 °C and after 1 hour under reaction conditions (15 vol.-% H2, 15 vol.-%
CO in He, 250 °C, 20 bar) with the corresponding reference spectra.
The catalytic properties of the Pd2Ga-based, bifunctional catalyst in the direct synthesis of
DME from syngas were evaluated in a fixed-bed reactor. A total pressure of 50 bar,
temperatures between 250 °C and 300 °C and 70 vol.-% dilution (Ar/N2) of the educt gases
(i.e., to avoid the formation of hotspots in the reactor) were employed, leading to partial
pressures of H2 and CO of 7.5 bar, respectively. Pd-based catalysts have been reported as
highly active catalysts in the synthesis of methanol from CO2, however, these catalysts remain
largely unexplored for the direct synthesis of DME from syngas. It has to be noted that H2
was typically employed in excess (H2:CO2 ratio of 3:1) in these previous experiments. To
compare the 8 wt.-% Pd/Ga-based catalyst with a conventional Cu/Zn-based catalyst with a
higher, 12 wt.-% Cu loading, reaction rates were normalized over mole of metal content in the
STD catalyst. This Cu/ZnO/-Al2O3 catalyst (12 wt.-% Cu) was prepared from colloidal,
Cu/Zn-based nanoparticles using a similar synthetic procedure and was used as a reference
-1
[25, 26]. The reaction rate based on moles of metal was 0.46 mmolCO∙s-1∙molPd for the
12

Pd/Ga-based catalyst at 250 °C and thus, higher than the conventional Cu/Zn-based reference
catalyst (0.37 mmolCO∙s-1∙molCu-1). These reaction rates may be increased by increasing the
H2:CO ratio, as well as the partial CO and H2 pressure (i.e., by reducing the amount of N2 in
the feed). For both catalysts, the methanol selectivity was low at all reaction temperatures
applied in this study, indicating the complete dehydration over γ-Al2O3 under these
conditions. Consequently, methanol formation seemed to be the rate limiting step in both
cases. Equilibrium CO conversion can be reached by increasing the reaction temperature and
therefore, increasing the reaction rate of R1 (Equation 4). Since the same -Al2O3 dehydration
catalyst was employed for the preparation of both catalysts, the difference in catalytic
performance may be attributed to the metal active component (i.e., Cu/Zn and Pd/Ga). The
equilibrium of the water gas shift reaction (Equation 6) is on the right side under these
conditions (CO2 and H2). Both catalysts were active in the WGS reaction and showed high
CO2 selectivities (i.e., SCO2 (250 °C) 31.3 % Cu/Zn-based catalyst, SCO2 (250 °C) 33.3 %
Pd/Ga-based catalyst). This is in good agreement with the CO2 selectivity which is expected
according to the total reaction equation for the direct synthesis of DME (Equation 1).
However, as compared to the conventional Cu/Zn-catalyst, the Pd/Ga-based catalyst showed a
significantly different behaviour regarding methane formation.
CO conversions and carbon-based product selectivities of the Pd2Ga-based STD catalyst are
summarized in Table 1. By successively increasing the reaction temperature from 250 to
300 °C, the CO conversion of the Pd/Ga catalyst increased from 11 to 34 % and the DME
selectivity decreased from 53 to 10 %, whereas the methane selectivity rose from 12 to 43 %,
accordingly (Table 1). This was not observed for the Cu/Zn-based catalyst where the methane
selectivity was significantly lower (< 4 % below 280 °C) and increased from 1.2 % at 250°C
to 8.8 % at 300°C. In both cases, the methane selectivity strongly increased with increasing
temperature. The Pd/Ga-based catalyst, however, revealed significantly higher methane
selectivities over the whole temperature range. In general, reactions forming methane may
occur via DME or methanol hydrogenolysis or from CO and/or CO2 methanation reactions.
When operating at low temperatures ( 280 °C) usually employed for DME synthesis, these
reactions are typically not observed to any great extent for the conventional Cu/Zn-based
catalysts. However, methanation reactions may be catalyzed over metallic catalysts,
particularly at higher temperatures. The methane selectivity was also considerable (up to
25%), e.g., in the CO hydrogenation to methanol at 260°C using Pd nanoparticles supported
on ZrO2, TiO2, and Al2O3 [47]. Pd0 is known to promote the formation of methane over
methanol. Although the authors did not show any analytical data for the activated catalysts,
the formation of alloyed, Pd-based catalyst species can also not be excluded under these
activation conditions (350 °C, H2 atmosphere).
In view of methane selectivity, the future optimization of the elemental composition of the
Pd-based catalyst (in terms of the Pd/Ga ratio or the replacement of Ga by, e.g., In or Zn) may
be a measure to further decrease methane selectivity in the STD reaction while maintaining
high reaction rates and longevity. The formation of Ga-Pd intermetallics and their effects on
methanol synthesis have been previously described by Zhang et al. [20]. In this case,
hydrogenation of CO did not occur, but a considerable methanol yield from a CO2/H2 feed
was observed for Pd2Ga/α-Ga2O3. It has to be noted that even though CO/H2 was applied as
educt gas in the STD reaction in the present study, CO2 was formed in situ via the water gas
shift reaction. In methanol synthesis from H2/CO2 mixtures, the increase in reaction
temperature from 190 to 240 °C led to a decrease in methanol selectivity to approximately 40
% [23] and 20 % [19], accordingly. Moreover, the Pd/Ga ratio was reported to influence the
catalytic performance of colloidal, quasi homogenous nanoparticles in methanol synthesis,
and an increase in the rate of methanol formation was observed with increasing Ga content of
13

the nanoparticles [23]. A maximum in methanol formation rate was described for a Pd/Ga
ratio of 1:2, which was also the highest Pd/Ga ratio tested in this study.
Table 1 CO conversion and carbon-based product selectivities in the direct synthesis of DME for the Pd/Ga catalyst.
(reaction conditions: pressure 50 bar, educt gas composition H2:CO:Ar:N2 = 3:3:10:4, total flow 50 mlNTP/min; reaction
temperatures between 250 °C and 300 °C; XEq: equilibrium CO conversion under these reaction conditions according to
equation (1))
Reaction
temperature
(°C)
XCO
(%)
XEq
(%)
SDME
(%)
SCO2
(%)
SMeOH
(%)
SCH4
(%)
250
260
270
280
290
300
10.9
14.6
19.6
25.2
30.2
34.0
66.4
60.5
54.1
47.3
40.4
33.6
52.4
46.2
38.2
29.2
19.6
10.3
33.6
35.1
37.6
40.3
42.8
45.0
1.9
1.8
1.7
1.5
1.3
1.0
12.1
16.9
22.2
28.6
35.8
43.1
Most notably, the Pd/Ga-based catalyst showed an excellent stability under reaction
conditions with no sign of deactivation and thus, outperformed the Cu/Zn/-Al2O3 catalysts in
view of longevity in the STD reaction. In this context, the Pd/Ga-based STD catalyst was
tested for three modified residence times (mod 0.8, 1.2, and 2.4 g s/mLNTP) increasing the
reaction temperature successively from 250 °C to 300 °C (i.e., every 6 h the reaction
temperature was increased by 10 °C (heating rate 10 °C/h) from 250 °C to 300 °C)
(Figure S2). After testing the catalyst for each modified residence time in the temperature
range from 250 °C to 300 °C, the temperature of the catalyst bed was again decreased to
250 °C, and both the CO conversion and the DME selectivity of the original, activated STD
catalyst at 250 °C were recovered. Any variation in CO conversion was within the error of the
GC measurement.
A kinetic model based on power law equations and five independent reaction equations was
used to fit the experimental data. Using this approach the individual reaction rates in methanol
synthesis, methanol dehydration, WGS reaction and methanation from CO and CO2 were
considered, instead of applying the total reaction equation of the direct synthesis of DME
from CO according to Equation (1).
0.18
0.17
0.16
0.15
0.14
0.13
0.18
0.16
0.14
0.12
300 °C
250 °C
CO
H2
CO
H2
0.10
0.04
0.03
0.02
0.01
0.00
-0.01
CO2
CH4
0.02
0.00
DME
CO2
CH4
DME
MeOH
MeOH
0.0
0.5
1.0
1.5
2.0
-1
2.5
0.0
0.5
1.0
1.5
2.0
-1
2.5
Modified residence time (gsml
)
Modified residence time (gsml )
Figure 5 Comparison of experimental (symbols) and simulated molar fractions (lines) for Pd/Ga (8)-γ-Al2O3 at 250 °C (left)
and 300 °C (right).
14

The H2:CO ratio was kept constant at 1:1. Figure 5 shows the simulated and experimental
molar fractions dependent on the modified residence time at 250 and 300 °C (260 to 290 °C,
Figure S3). The trends of CO/H2 conversion and of DME and byproduct formation can be
simulated using this kinetic model. An increase in reaction temperature leads to a higher
reaction rate (i.e., in particular at the entrance of the catalyst bed) which is reflected by a
higher slope at 300 °C for low modified residence times. The CO conversions during the
measurements were relatively high (34 % at 300 °C) and close to equilibrium conversions of
the direct synthesis of DME under these conditions. Nevertheless, this model can predict CO
concentration and product distribution as a consequence of the equilibrium constants
calculated (Equation 18) and implemented in the expressions for the reaction rates. Although
the direct synthesis of DME is a strongly exothermic reaction, hotspots were successfully
avoided during the catalytic experiments by dilution with inert gases.
The quality of the fitting was verified using parity plots. For CO, this plot is shown in Figure
6. All experimental values are in good agreement with the model predictions and within a
range of 10 % compared to the calculated molar fractions.
0.18
PG (8)-A
0.16
0.14
0.12
0.10
0.08
0.08
0.10
0.12
0.14
0.16
0.18
yCO, simulated
Figure 6 Parity plot of CO molar fractions for Pd/Ga (8)-γ-Al2O3; dashed lines indicate a deviation of 10 % from the
predicted values using the kinetic model.
For DME and CH4, these plots are shown in Figure S4. The experimental molar fractions of
CH4 are also in agreement with the model. Notable deviations can only be observed for DME
at low molar fractions. These deviations can be ascribed to the dilution with inert gases
(70 vol.-% Ar+N2) and low DME selectivities, which lead to low molar fractions of DME.
Table 2 summarizes the apparent activation energies of the considered reactions for the
bifunctional Pd/Ga-based STD catalyst (this work) and the Cu/Zn-based reference STD
catalyst [25, 26]. In both cases, the as-prepared nanoparticles were directly immobilized on γ-
Al2O3 which served as dehydration catalyst. The activation energies were derived from the
slopes of the corresponding Arrhenius plots (Figure S5). Different catalytically active sites of
the Cu/Zn and the Pd/Ga catalyst (originating from Cu0 and Pd2Ga surface) may account for
the need of the additionally considered reaction equation R5 (methanation of CO2) in case of
the Pd-based catalyst. The activation energies were comparable to the ones previously
reported by others [27, 31, 33, 34]. Activation energies of 66.2kJ/mol [37] and 90.0 kJ/mol
[33] were demonstrated for methanol synthesis by Hadipour et al. and Ereña et al.,
respectively, and 69.2 kJ/mol and 63.5 kJ/mol for its dehydration to DME, accordingly.
15

Table 2 Apparent activation energies of the Pd/Ga-based STD catalyst (this work) and the Cu/Zn-
based reference STD catalyst ([25])
Activation energies (kJ/mol)
Catalyst
R1
R2
R3
R4
R5
52.8 ± 3.6
33.7 ± 1.5
34.8 ± 0.5 34.8 ± 0.5 143.7 ± 2.8
-
Cu/Zn (12)-γ-Al2O3
Pd/Ga (8)-γ-Al2O3
28.2 ± 0.4 29.4 ± 0.2 106.0 ± 2.0 107.1 ± 5.1
However, the activation energy was shown to depend on the kinetic model and activation
energies were calculated for the methanation reaction, ranging from 15.9 to 54.3 kJ/mol [31].
The activation energy of the methanation reaction R4 was 106 kJ mol-1 for Pd/Ga (8)-γ-Al2O3
and thus, lower than the one of the conventional Cu/Zn-based reference catalyst (i.e., Cu/Zn
(12)-γ-Al2O3 144 kJ mol-1). This is in agreement with the favored production of methane. In
general, the lower apparent activation energy corresponds to a lower temperature dependency
under the chosen reaction conditions. Moreover, the lower activation energy for R1 of the
Pd/Ga catalyst may also be attributed to different active sites present on the surface of this
catalyst (e.g., Pd and Ga sites) and a lower number of "methanol active sites" as compared to
the Cu/Zn catalyst. Overall, our kinetic model can be applied to describe the DME synthesis
for these reaction conditions with appropriate accuracy even for different methanol active
components (i.e., Cu/Zn or Pd/Ga) and for CO conversion close to thermodynamic
equilibrium. Further modifications of the model and the reaction conditions taking into
account, e.g., the different metal loadings might extend the range of validity, which will be a
subject of future studies.
Conclusion
A Pd/Ga-based catalyst was prepared via an organometallic colloidal approach. Initially,
small and uniform Pd/Ga-based nanoparticles were obtained as a suspension in toluene in a
synthetic procedure by reacting Pd(acac)2 with Ga(butyl)3. The synthesis yielded spherical
particles in the lower nanometer size range. The 3.4 nm size Pd/Ga nanoparticles were
successively employed as building blocks for the preparation of a bifunctional STD catalyst
where the nanoparticles were directly supported on the -Al2O3 dehydration catalyst. The
Pd/Ga particles were nicely distributed over the dehydration catalyst which enabled a close
proximity of the two catalytic functions in the bifunctional STD catalyst. The formation of the
active Pd/Ga phase in reductive H2 atmosphere was confirmed by both ex situ XRD and in
situ XAS studies. In situ XAS experiments were carried out for the first time under reaction
conditions of the STD reaction at elevated pressure (CO/H2, 20 bar, 250°C, 1 h) and showed
slight morphological changes of the Pd/Ga particles during the first hour of reaction, most
probably indicating some dealloying under reaction conditions. Overall, the activity and
stability of the new Pd/Ga nanoparticle-derived STD catalysts are promising for the STD
process. The activity of the Pd/Ga-based catalyst was comparable to the Cu/Zn-based STD
catalyst which was obtained via a similar synthetic procedure. When normalized by active
metal content, the reaction rate of the Pd2Ga-based catalyst was even higher than the one of
the conventional Cu/Zn-based reference. In contrast to the Cu/Zn-based system, the Pd/Ga-
based catalyst, however, revealed high methane selectivity over the whole temperature range
investigated in this study which will require further optimization in future studies. Notably,
the Pd/Ga-based catalyst revealed an excellent stability in the STD process with no signs for
catalyst deactivation during catalytic testing. After testing the catalyst at temperatures up to
300 °C, the original CO conversion and DME selectivity of the STD catalyst were recovered
after decreasing the temperature of the catalyst bed again to 250 °C. The experimental results
were fitted for simulated biomass-deriveded CO-rich syngas (H2:CO ratio 1:1) and
16

temperatures between 250 and 300 °C. Therefore, a kinetic model based on power law
equations for the individual reactions involved in the direct DME synthesis was developed.
Methane formation observed in case of Pd2Ga may be suppressed by further optimizing the
catalyst composition of the Pd-based STD catalyst. It is expected that the synthetic
methodology should be applicable to a wide range of Pd-based catalysts and provide not only
high performance bifunctional Pd-based catalysts but further contribute to a more
fundamental understanding in view of synthesis-structure performance relationships in this
reaction.
Acknowledgements
M.G. kindly acknowledges financial support of the Helmholtz Research School Energy-
Related Catalysis. Further acknowledgments are given to KIT’s synchrotron radiation source
ANKA for providing beamtime at CAT-ACT beamline and Dr. Anna Zimina and Dr. Tim
Prüßmann for their support during the measurements. Doreen Neumann-Walter and Dr.
Thomas Otto are acknowledged for their support with the BET measurements.
Literature
[1]
[2]
[3]
O. Yotaro, Y. Masahiro, S. Tsutomu, I. Osamu, O. Takashi, I. Norio, JFE Technical
Report 8 (2006) 34.
T. A. Semelsberger, R. L. Borup, H. L. Greene, Journal of Power Sources 156 (2006)
497.
N. Dahmen, E. Henrich, E. Dinjus, F. Weirich, Sustainability and Society 2:3 (2012)
1.
[4]
[5]
P. Lv, Z. Yuan, L. Ma, C. Wu, Y. Chen, J. Zhu, Renewbale Energy 32 (2007) 2173.
F. S. Ramos, A. M. Duarte de Farias, L. E. P. Borges, J. L. Monteiro, M. A. Fraga, E.
F. Sousa-Aguiar, L. G. Appel, Catal. Tod. 101 (2005) 39.
[6]
[7]
[8]
[9]
D. Mao, J. Xia, Q. Chen, G. Lu, Catal. Commun. 10 (2009) 620.
M. Stiefel, R. Ahmad, U. Arnold, M. Döring, Fuel Proc. Technol. 92 (2011) 1466.
J. Sun, G. Yang, Y. Yoneyama, N. Tsubaki, ACS Catal. 4 (2014) 3346.
H. H. Kung, Catal. Tod. 11 (1992) 443.
[10] A. T. Aguayo, J. Erena, I. Sierra, M. Olazar, J. Bilbao, Catal. Tod. 106 (2005) 265.
[11] T. Lunkebein, F. Girgsdies, T. Kandemir, N. Thomas, M. Behrens, R. Schlögl, E. Frei,
Angew. Chem. Int. Ed. 55 (2016) 12708.
[12] V.V. Ordomsky, M. Cai, V. Sushkevich, S. Moldovan, O. Ersen, C. Lancelot, V.
Valtchev, A.Y. Khodakov, Applied Catalysis A: General 486 (2014) 266.
[13] Q. Ge, Y. Huang, F. Qiu, S. Li, Appl. Catal. A 167 (1998) 23.
[14] R. Ahmad, D. Schrempp, S. Behrens, J. Sauer, M. Döring, U. Arnold, Fuel Process.
Technol. 121 (2014), 38.
17

[15] F. Fajula, R. G. Anthony, J. H. Lunsford, J. Catal. 73 (1982) 237.
[16] D. Lee, G. S. Jung, H. C. Lee, J. S. Lee, Catal. Tod. 111 (2006) 373.
[17] A. Ota, E. L. Kunkes, I. Kasatkin, E. Groppo, D. Ferri, B. Poceiro, R. M. Navarro
Yerga, M. Behrens, J. Catal. 293 (2012) 27.
[18] T. Fujitani, M. Saito, Y. Kanai, T. Watanabe, J. Nakamura, T. Uchijima, Appl. Catal.
A 125 (1995) 199.
[19] E. M. Fiordaliso, I. Sharafutdinov, H. W. P. Carvalho, J.-D. Grunwaldt, T. W. Hansen,
I. Chorkendorff, J. B. Wagner, C. D. Damsgaard, ACS Catal. 5 (2015) 5827.
[20] L. Li, B. Zhang, E. Kunkes, K. Föttinger, M. Armbrüster, D. S. Su, W. Wie, R.
Schlögl, M. Behrens, ChemCatChem 4 (2012) 1764.
[21] S. E. Collins, J. J. Delgado, C. Mira. J. Calvino, S. Bernal, D. L. Chiavassa, M. A.
Baltanás, A. L. Bonivardi, J. Catal. 292 (2012) 90.
[22] A. Ota, M. Armbrüster, M. Behrens, D. Rosenthal, M. Friedrich, I. Kasatkin, F.
Girgsdies, W. Zhang, R. Wagner, R. Schlögl, J. Phys. Chem. C 115 (2011) 1368.
[23] A. García-Trenco, E. R. White, A. Regoutz, D. J. Payne, M. S. P. Shaffer, C. K.
Williams, ACS Catal. 7 (2017) 1186.
[24] M. Armbrüster, G. Wowsnick, M. Friedrich, M. Heggen, R. Cardoso-Gil, J. Am.
Chem. Soc. 133 (2011) 9112.
[25] M. Gentzen, W. Habicht, D. E. Doronkin, J.-D. Grunwaldt, J. Sauer, S. Behrens.,
Catal. Sci. Technol. 6 (2016) 1054.
[26] M. Gentzen, D. E. Doronkin, T. L. Sheppard, J. D. Grunwaldt, J. Sauer, S. Behrens,
Appl. Catal. A: General 557 (2018), 99.
[27] G. H. Graaf, E. J. Stamhuis, A. A. C. M. Beenackers, Chem. Eng. Sci. 43 (1988) 3185.
[28] G. Bercic, J. Levec, Ind. Eng. Chem. Res. 32 (1993) 2478.
[29] T. N. Otto, C. Mas, H. Ederer, M. Stiefel, M. Döring, E. Dinjus, Chem. Ing. Tech. 83
(2011) 874.
[30] S. K. Wilkinson, L.G.A. van de Water, B. Miller, M.J.H. Simmons, E.H. Stitt, M.J.
Watson, J. Catal. 337 (2016) 208.
[31] A. T. Aguayo, J. Erena, D. Mier, J. M. Arandes, M. Olazar, J. Bilbao, Ind. Eng. Chem.
Res. 46 (2007) 5520.
[32] I. Sierra, J. Ereña, A. T. Aguayo, M. Olazar, J. Bilbao, Ind. Eng. Chem. Res. 49 (2010)
481.
[33] J. Ereña, , I. Sierra, A. T. Aguayo, A. Ateka, M. Olazar, J. Bilbao, Chem. Eng. J. 174
(2011) 660.
[34] G.R. Moradi, J. Ahmadpour, F. Yaripour, Chem. Eng. J. 144 (2008) 88.
18

[35] W. Ding, M. Klumpp, S. Lee, S. Reuß, S. A. Al-Thabaiti, P. Pfeifer, W. Schwieger, R.
Dittmeyer, Chem. Ing. Tech. 87 (2015) 702.
[36] Y. I. Pyatnitskii, P. E. Strizhak, N. K. Lunev, Theor. Exp. Chem. 45 (2009) 325.
[37] A. Hadipour, M. Sohrabi, Chem. Eng. J. 137 (2008) 294.
[38] R. A. Kovar, G. Loaris, H. Derr, J. O. Callaway, Inorg. Chem. 13 (1974) 1476.
[39] A. Zimina, K. Dardenne, M.A. Denecke, D.E. Doronkin, E. Huttel, H. Lichtenberg, S.
Mangold, T. Pruessmann, J. Rothe, T. Spangenberg, R. Steininger, T.Vitova, H.
Geckeis, J.-D. Grunwaldt., Rev Sci Instrum. 88 (2017) 113113.
[40] J.-D. Grunwaldt, M. Caravati, S. Hannemann, A. Baiker, Phys. Chem. Chem. Phys. 6
(2004) 3037.
[41] B. Ravel, M. Newville, ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-
ray absorption spectroscopy using IFEFFIT, J. Synchrotron Radiat. 12 (2005) 537–
541.
[42] J.J. Rehr, R.C. Albers, Rev. Mod. Phys. 72 (2000) 621.
[43] H. Bönnemann, W. Brijoux, R. Brinkmann, E. Dinjus, T. Joussen, B. Korall, Angew.
Chem. Int. Ed. Engl. 30 (1991) 1312.
[44] O. Oyola-Rivera, M. A. Baltanás, N. Cardona-Martínez, J. CO2 Utilization 9 (2015) 8.
[45] K.-I. Muto, N. Katada, M. Niwa, Catal. Tod. 35 (1997) 145.
[46] A. L. Bugaev, A. A. Guda, K.l A. Lomachenko, V. V. Srabionyan, L. A. Bugaev, A.
V. Soldatov, C. Lamberti, V. P. Dmitriev, J. A. van Bokhoven, J. Phys. Chem. C 118
(2014) 10416.
[47] W. Kim, H. Hayashi, M. Kishida, H. Nagata, K. Wakabayashi, Applied Catalysis A:
General 169 (1998) 157
19