Spinel-Structured ZnCr2O4 with Excess Zn Is the Active ZnO/Cr2O3 Catalyst for High-Temperature Methanol Synthesis

Article abstract of DOI:10.1021/acscatal.7b01822

A series of ZnO/Cr₂O₃ catalysts with different Zn:Cr ratios was prepared by coprecipitation at a constant pH of 7 and applied in methanol synthesis at 260-300 °C and 60 bar. The X-ray diffraction (XRD) results showed that the calcined catalysts with ratios from 65:35 to 55:45 consist of ZnCr₂O₄ spinel with a low degree of crystallinity. For catalysts with Zn:Cr ratios smaller than 1, the formation of chromates was observed in agreement with temperature-programmed reduction results. Raman and XRD results did not provide evidence for the presence of segregated ZnO, indicating the existence of Zn-rich nonstoichiometric Zn-Cr spinel in the calcined catalyst. The catalyst with Zn:Cr = 65:35 exhibits the best performance in methanol synthesis. The Zn:Cr ratio of this catalyst corresponds to that of the Zn₄Cr₂(OH)₁₂CO₃ precursor with hydrotalcite-like structure obtained by coprecipitation, which is converted during calcination into a nonstoichiometric Zn-Cr spinel with an optimum amount of oxygen vacancies resulting in high activity in methanol synthesis. Density functional theory calculations are used to examine the formation of oxygen vacancies and to measure the reducibility of the methanol synthesis catalysts. Doping Cr into bulk and the (10-10) surface of ZnO does not enhance the reducibility of ZnO, confirming that Cr:ZnO cannot be the active phase. The (100) surface of the ZnCr₂O₄ spinel has a favorable oxygen vacancy formation energy of 1.58 eV. Doping this surface with excess Zn charge-balanced by oxygen vacancies to give a 60% Zn content yields a catalyst composed of an amorphous ZnO layer supported on the spinel with high reducibility, confirming this as the active phase for the methanol synthesis catalyst.

Full text of DOI:10.1021/acscatal.7b01822

Subscriber access provided by LONDON METROPOLITAN UNIV
Article
Spinel-Structured ZnCr2O4 with Excess Zn is the Active ZnO/
Cr2O3 Catalyst for High-Temperature Methanol Synthesis
Huiqing Song, Daniel Laudenschleger, John J Carey, Holger Ruland, Michael Nolan, and Martin Muhler
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b01822 • Publication Date (Web): 27 Sep 2017
Downloaded from http://pubs.acs.org on September 29, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 15
ACS Catalysis
1
2
3
4
5
6
7
8
9
Spinel-Structured ZnCr2O4 with Excess Zn is the Active ZnO/Cr2O3
Catalyst for High-Temperature Methanol Synthesis
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Huiqing Song†, Daniel Laudenschleger†, John J. Carey‡, Holger Ruland†, Michael Nolan‡*, and Martin
Muhler†*
†Laboratory of Industrial Chemistry, RuhrꢀUniversity Bochum, 44780 Bochum, North RhineꢀWestphalia, Germany
‡Tyndall National Institute, University College Cork, Cork T12R5CP, Munster, Ireland
ABSTRACT: A series of ZnO/Cr2O3 catalysts with different Zn:Cr ratios was prepared by coꢀprecipitation at a constant pH of 7
and applied in methanol synthesis at 260ꢀ300 °C and 60 bar. The Xꢀray diffraction (XRD) results showed that the calcined catalysts
with ratios from 65:35 to 55:45 consist of ZnCr2O4 spinel with low degree of crystallinity. For catalysts with Zn:Cr ratios smaller
than 1, the formation of chromates was observed in agreement with temperatureꢀprogrammed reduction results. Raman and XRD
results did not provide evidence for the presence of segregated ZnO indicating the existence of Znꢀrich nonꢀstoichiometric ZnꢀCr
spinel in the calcined catalyst. The catalyst with Zn:Cr=65:35 exhibits the best performance in methanol synthesis. The Zn:Cr ratio
of this catalyst corresponds to that of the Zn4Cr2(OH)12CO3 precursor with hydrotalciteꢀlike structure obtained by coꢀprecipitation,
which is converted during calcination into a nonꢀstoichiometric ZnꢀCr spinel with an optimum amount of oxygen vacancies resultꢀ
ing in high activity in methanol synthesis. Density functional theory calculations are used to examine the formation of oxygen vaꢀ
cancies and to measure the reducibility of the methanol synthesis catalysts. Doping Cr into bulk and the (10ꢀ10) surface of ZnO
does not enhance the reducibility of ZnO, confirming that Cr:ZnO cannot be the active phase. The (100) surface of the ZnCr2O4
spinel has a favorable oxygen vacancy formation energy of 1.58 eV. Doping this surface with excess Zn chargeꢀbalanced by oxygen
vacancies to give a 60% Zn content yields a catalyst composed of an amorphous ZnO layer supported on the spinel with high reducꢀ
ibility confirming this as the active phase for the methanol synthesis catalyst.
KEYWORDS: ZnO/Cr2O3, methanol synthesis, nonꢀstoichiometric spinel, oxygen vacancy, DFT+U
1. Introduction
calcination, the synergy observed between ZnO and Cr2O3 in
comparison to the pure component oxides was mainly ascribed
to an increase in the specific surface area.15 In addition, ZnO
had been generally identified as the only active component,
and the promoting role of Cr2O3 was attributed to spinel forꢀ
mation, which prevented the sintering of small ZnO crystalꢀ
lites by providing a considerably larger surface area and better
distribution of ZnO particles.8 It is generally accepted that
methanol synthesis over ZnO is a structureꢀsensitive reaction,
in which oxygen vacancies on ZnO (000ꢀ1) are assumed to be
the active sites. One model, which was reported by Boccuzzi
et al.16 in 1978, described an active site composed of three Zn
ions around an oxygen vacancy. Two years later, an active site
with similar geometry and composition was proposed by
Kung,17 who assumed that oxygen vacancies could assist in
the adsorption and activation of CO. He proposed a catalytic
cycle of methanol formation based on the oxygen vacancy as
active site. The reaction starts with the dissociative adsorption
of H2 on ZnO sites around the oxygen vacancy followed by the
adsorption of the CO molecule in the oxygen vacancy. The
reaction proceeds through hydrogenation of adsorbed CO by
the transfer of adsorbed hydrogen to form surface formyl
species (CHO)ads. Further hydrogenation of the formyl species
yields adsorbed formaldehyde (H2CO)ads, which is hydrogenꢀ
ated forming methoxy species (H3CO)ads. The hydrogenation
Methanol is one of the most important industrially produced
basic chemicals due to its application in a wide range of
fields.1ꢀ2 The investigation of methanol synthesis from a mixꢀ
ture of hydrogen and carbon oxides using metals and metal
oxides as catalyst dates back to the 1910s.3 In the 1920s BASF
issued two patents focusing on ZnO/Cr2O3 and Cu/ZnO cataꢀ
lysts, which were primarily used in the early period of the
industrial production of methanol.4ꢀ5 Later, the Cu/ZnO/Al2O3
catalyst with high activity and selectivity was patented by ICI
in the 1960s, which requires highly purified synthesis gas.6ꢀ7
Due to the lower operating temperature and pressure, the
Cu/ZnOꢀbased catalyst has been used extensively for methanol
production in the last decades. However, ZnꢀCr mixed oxides,
which are applied in the synthesis of methanol at high temperꢀ
atures and pressures,8 remain of interest owing to their reꢀ
sistance against sulfur impurities in syngas as well as the posꢀ
sibility of modifying their selectivities towards the synthesis of
higher alcohols.9ꢀ12 Furthermore, due to the high thermal staꢀ
bility, ZnO/Cr2O3 catalysts demonstrated great potential in
combination with microporous acidic catalysts in the oneꢀpot
synthesis of liquid hydrocarbons and short chain olefins from
syngas.13ꢀ14
In the early investigation on methanol synthesis over
ZnO/Cr2O3 catalysts prepared by coꢀprecipitation followed by
1
ACS Paragon Plus Environment

ACS Catalysis
of methoxy species leads to the product methanol. It was proꢀ
Page 2 of 15
ed an maximum specific activity for Zn:Cr = 2.56 (72:28),
which is in good agreement with Molstad and Dodge.25 Altꢀ
hough the influence of the Zn:Cr ratio on the performance of
ZnO/Cr2O3 catalyst has been investigated over decades, there
is still no agreement on the optimal Zn:Cr ratio.
1
2
3
4
5
6
7
8
posed that a change of the mechanism occurs when CO2 is
present, as the very stable formate (HCO2)ads species may be
formed in the defect site. This mechanism was confirmed by
Kurtz et al.,18 who found that in the presence of CO2 the rate
of methanol formation is significantly slower.
In this work, we studied the role of the Zn:Cr ratio in the cataꢀ
lytic activity of ZnO/Cr2O3 catalysts by a combined experiꢀ
mental and density functional theory (DFT) approach. A series
of ZnO/Cr2O3 catalysts with different Zn:Cr ratios varying
from 100:0 to 0:100 were prepared by coꢀprecipitation at 65°C
and pH 7 with Na2CO3 solution as the precipitating agent. The
resulting precipitates were calcined in air at 320 °C for 3 h and
characterized by N2 physisorption, XRD, temperatureꢀ
programmed reduction (TPR), XPS and Raman spectroscopy.
The catalytic testing was performed at 60 bar using a syngas
mixture with H2:CO = 1.5, and the reaction temperature was
varied in steps of 20 °C from 260 to 300 °C. The catalyst with
Zn:Cr=65:35 exhibits the best performance in methanol synꢀ
thesis.
In addition to ZnO, nonꢀstoichiometric ZnꢀCr spinel was reꢀ
ported as the active phase in methanol synthesis. Del Piero et
al.19 first reported that the nonꢀstoichiometric ZnꢀCr spinel was
formed by the dissolution of excess ZnO in ZnCr2O4, which
results from the solid reaction between ZnO and Cr2O3 during
calcination. They also pointed out that the dissolved Zn2+ ions
may be located in octahedral sites randomly substituting Cr3+
ions, leaving some of the tetrahedral sites vacant. The deactiꢀ
vation of the catalyst was observed to be accompanied by a
decrease in the nonꢀstoichiometry of the spinelꢀlike phase and
a corresponding increase in the amount of crystalline ZnO
detected by Xꢀray diffraction (XRD). Therefore, they assumed
the nonꢀstoichiometric spinelꢀlike phase to be the active phase
in the ZnO/Cr2O3 catalyst. More studies regarding the influꢀ
ence of the nonꢀstoichiometry of the ZnꢀCr spinel on its bulk
and surface properties were carried out by Bertoldi et al.,20
who proposed a general chemical formula of the nonꢀ
stoichiometric ZnꢀCr spinel as ZnxCr2/3(1ꢀx)O, where x ranges
from 0.25 to 0.40. They also found that the activity of the nonꢀ
stoichiometric ZnꢀCr spinel towards CO adsorption is much
higher than observed for pure ZnO, Cr2O3 and ZnCr2O4.
Giamello et al.21 and Riva et al.22 also confirmed the formation
of the nonꢀstoichiometric ZnꢀCr spinel based on methanol
temperatureꢀprogrammed desorption (TPD) and Xꢀray photoeꢀ
lectron spectroscopy (XPS) studies. Grimes et al.23 used atomꢀ
istic simulation to explain the mechanisms of the dissolution
of ZnO and suggested that one oxygen vacancy was created
for every three ZnO species dissolved.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Given that previous work shows the importance of oxygen
vacancies in methanol synthesis, we combine these experiꢀ
ments with DFT calculations of the reducibility of different
Cr/Znꢀcontaining systems, as measured by the oxygen vacancy
formation energy. For a catalyst composed of Cr doping into
ZnO (as excess Zn), the oxygen vacancy formation energies
are larger than undoped ZnO. The oxygen vacancy formation
energies for the pure ZnCr2O4 spinel are significantly lower
than the undoped ZnO. Doping the spinel structure with excess
Zn on Cr3+ octahedral sites gives compositions ranging from
30% to 60% Zn content. At the higher concentration of incorꢀ
porated Zn, the Zn dopants replace all surface Cr atoms (it is
less favorable to replace bulk Cr atoms) and the oxygen vaꢀ
cancy formation energy decreases indicating high reducibility
of this system. The structure of this catalyst can be thought of
as a thin, amorphous ZnOꢀstoichiometry layer supported on
the ZnCr2O4 spinel. In addition, the adsorption of CO and H2
on ZnCr2O4 with excess Zn is weaker than on stoichiometric
ZnCr2O4, allowing for increased reactivity. These DFT inꢀ
sights help to explain the experimental observations that both
ZnO and ZnCr2O4 are present in the catalyst. The reducibility
of and interaction with CO and H2 on ZnCr2O4 are the key to
increased methanol activity and a superior methanol synthesis
catalyst.
The Zn:Cr ratio was found to significantly influence the cataꢀ
lytic activity of the ZnO/Cr2O3 catalyst.24 The optimum Zn:Cr
ratio has been extensively investigated in the literature.8, 25ꢀ26
Molstad and Dodge25 reported that catalysts with Zn:Cr ratios
between 70:30 and 60:40 exhibit the highest yield of methanol
at relatively low temperatures of 300ꢀ325°C. This observation
is in good agreement with the results reported by Bone,24 who
found that the maximum activity is achieved with a Znꢀexcess
catalyst containing about 75 atom% of Zn. According to
Molstad and Dodge,25 the highest activity was obtained with
the catalyst containing 20ꢀ30 atom% of Cr at a reaction temꢀ
perature in the range from 350 °C to 425 °C. Additionally,
they observed that the catalysts with Zn:Cr > 75:25 lose their
activity rapidly, while the catalysts containing more Cr exhibit
a slight activity increase with reaction time. Furthermore, after
testing the catalysts had undergone volume shrinkage, which
increased with increasing Cr content. When using the already
shrunk catalyst, the maximum yield of methanol was obtained
with a Zn:Cr ratio of around 1:1. These results were also conꢀ
firmed by Errani et al.,26 however, with a different explanaꢀ
tion. Based on the investigation of nonꢀstoichiometric ZnꢀCr
spinel,19ꢀ22 the authors suggested that the activity increases up
to Zn:Cr = 50:50 due to the increasing nonꢀstoichiometry of
the spinelꢀtype phase, while the formation of a segregated ZnO
phase causes a decrease in activity. Nevertheless, no linear
correlation was found between the catalytic activity and the
excess zinc inside the nonꢀstoichiometric spinelꢀtype structure
up to Zn:Cr = 50 : 50. More recently, Bradford et al.27 reported
that ZnO/Cr2O3 catalysts prepared by coꢀprecipitation exhibitꢀ
2. Experimental
a. Catalyst preparation
Catalysts with different Zn:Cr ratios were prepared by a coꢀ
precipitation procedure adapted from Kurtz et al.18 A metal
salt solution (1 M) consisting of zinc and chromium nitrates
with the desired Zn:Cr ratio was added dropwise into a vessel,
which contained 50 ml HPLC water heated up to 65°C.The
precipitation was performed under continuously stirring at a
constant pH of 7 by simultaneous addition of Na2CO3 solution
(1.2 M) using an autotitrator while the precipitation temperaꢀ
ture was constantly kept at 65°C. The suspension after precipiꢀ
tation was aged at 65 °C for 2 h. Subsequently, the precipitate
was filtered and washed with HPLC water and then dried for
18 h at 120 °C. The dried precipitate was granulated and calꢀ
cined in synthetic air for 3 h with a heating rate of 2 °C min−1
to 320 °C and keeping this temperature for 3 h.
b. Characterization
2
ACS Paragon Plus Environment

Page 3 of 15
ACS Catalysis
The metal composition as well as the content of residual Na reaction temperature in the range from 260 °C to 300 °C. The
1
2
3
4
5
6
7
8
were determined by optical emission spectroscopy with inducꢀ
tively coupled plasma (ICPꢀOES) using an UNICAM PU
7000. Powder XRD was performed using a Panalytical MPD
diffractometer with Cu Kα radiation (λ = 0.154 nm) at 45 kV
and 40 mA without monochromator. A Ni filter was used to
suppress the Kß emission line.The XRD patterns were recordꢀ
ed over a 2θ range of 5–85° with a 0.0131 step width and 250
s collection time. Powder diffraction files (PDF) from the
International Center of Diffraction Data (ICDD) and
X’PertHighScore Plus software were applied for qualitative
phase analysis. Specific surface areas and pore structures were
measured by N2 physisorption at 77 K using a Quantachrome
Autosorbꢀ1ꢀMP instrument. The catalysts were degassed under
vacuum at 200 °C for 2 h prior to the measurement. Pore volꢀ
ume and average pore diameter were obtained using the Barꢀ
rettꢀJoynerꢀHalenda (BJH) method. TPR was carried out in a
Uꢀtube quartz reactor with a H2/Ar mixture containing 4.6%
H2 as reducing agent. About 100 mg catalysts (250–355 ꢁm
sieve fraction) were flushed with Ar (99.9995%) to remove
weakly bound water and then reduced using a total flow rate
of 84.1 Nml/min at a rate of 5 °C/min with a programmable
temperature controller. The samples were heated up to 800 °C
and temperature was held at 800 °C for 1 h. H2 consumption
was recorded by a thermal conductivity detector (TCD). The
effluent gas passed through a cold trap filled with isopropanol
and dry ice before entering the TCD in order to remove water
formed during reduction. XPS measurements were carried out
in an ultraꢀhigh vacuum setꢀup equipped with a highꢀ
resolution GammadataꢀScienta SES 2002 analyzer and a monꢀ
ochromatic Al Kα as the Xꢀray source (1486.6 eV, anode
operating at 14 kV and 30 mA). The base pressure in the
measurement chamber was maintained at about 7 x 10ꢀ10 mbar.
The measurements were performed in the fixed transmission
mode with a pass energy of 200 eV. Charging effects during
the measurements were compensated by applying a flood gun.
The measured spectra were calibrated based on the C 1s peak
at 284.5 eV originating from adsorbed hydrocarbons. The
CasaXPS software was used to analyze the XPS data.
desired pressure for the reaction was controlled by a pressure
regulator with proportionalꢀintegralꢀdifferential (PID) controlꢀ
lers. The outlet gas of the chosen reactor was introduced into
the GC by switching the multiꢀposition valve (VM), while the
gas mixtures from the other reactors were passed to the exꢀ
haust. In order to avoid product condensation in the gas mixꢀ
ture, all gas lines from the reactor outlet to the GC were inꢀ
stalled in a separate oven, in which the temperature was set to
155 °C. The reactors were made of stainless steel and coated
with glass on the inner wall. The catalyst was fixed between
two plugs of quartz wool in the middle of the reactor, and two
glass rods were placed in the reactors to reduce the dead volꢀ
ume.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
d. Catalytic tests
500 mg of catalyst with particle sizes of 250–355 ꢁm was
loaded in each glassꢀlined stainless steel reactor. No preꢀ
reduction was carried out prior to the catalytic tests. Methanol
synthesis was performed at 60 bar using a mixture of 36% CO,
54% H2 and 10% N2 at a volumetric feed rate of 40 ml min−1
−1.
(STP) resulting in a GHSV of 4800 ml h−1 gcat The reaction
temperature was varied in steps of 20 °C from 260 to 300 °C.
At each temperature the reaction was carried out for 20 h.
Online gas analysis was performed with an Agilent gas chroꢀ
matograph equipped with two thermal conductivity detectors
using He as carrier gas.
e. Computational Methodology
All calculations are carried out using density functional theory
(DFT) with the generalized gradient approximation (GGA)28
and the PerdewꢀBurkeꢀErnzerhof (PBE)29 exchange correlaꢀ
tion functional as implemented in the Vienna ab initio Simulaꢀ
tion Package (VASP).30ꢀ31 The valence electrons are expanded
using a plane wave basis set, and the interactions between the
core (Cr:[Ar], Zn:[Ar], O:[He]) and the valence electrons
(Cr:[3p6, 3d5, 4s1], Zn:[3d10, 4s2], O:[2s2, 2p4]) are treated
using the projected augmented wave (PAW) method.32 To
describe the strong electron correlation of Cr cations in Crꢀ
ZnO and in the spinel structure, a Hubbard +U correction
(giving DFT+U)33ꢀ34 is applied to the Cr 3d states, for which
the value of U is 5 eV.35ꢀ36 Due to doping of lower valent Zn2+
cations onto Cr3+ sites, oxygen hole states will form and in
describing these states we apply U=5.5 eV36 to the O 2p states.
Raman spectra were measured using a WITec confocal Raman
microscope Alpha300 R/S/A. The samples excitation was
carried out by using a 532 nm frequencyꢀdoubled Nd:YAG
laser. A 50x/0.4 NA objective was used to focus the laser onto
the sample. The backꢀscattered Raman light was collected by
the same objective and transferred via a 50 ꢁm multimode
fiber to the spectrometer unit consisting of a 600 lines/mm
diffraction grating and a 1600x200 pixels electron multiplying
chargeꢀcoupled device camera operating at 60 °C. The spectral
resolution was approximately 4 cmꢀ1. 10 Raman spectra of
each sample were recorded with an integration time of 30 s for
each spectrum at a laser power of 20 mW.
The low energy structures for the ZnO bulk and ZnCr2O4
spinel bulk are obtained by relaxing a series of structures from
ꢀ4% to +4% of the experiment lattice constant where the atomꢀ
ic positions, cell vectors and angles were allowed to relax at a
constant volume. The energies obtained from each structure
were fitted to the Murnaghan Equation of State37 to obtain the
minimum energy structure. This was carried out for different
plane wave cutꢀoff energies of 400 eV, 450 eV and 500 eV
where the Brillouin Zone integrations were sampled using the
MonkhorstꢀPack method38 at different kꢀpoint sampling grids
of 2x2x2, 4x4x4, and 6x6x6. The structures were deemed
converged when the forces on the atoms were less than 0.02
eV/Å. For both bulk structures, the parameters for the bulk
structure are a plane wave cutꢀoff energy of 450 eV and a kꢀ
point sampling grid of 4x4x4 which gives a deviation of 1.6%
from the experimental lattice constants for both bulk strucꢀ
tures.39ꢀ40
c. Continuous methanol synthesis setꢀup
The setꢀup for methanol synthesis from syngas (Figure S1)
consists of the gas supply unit, the reactor unit and the gas
chromatograph as analytical device. The gas supply includes
syngas, Ar, H2, and H2 diluted in He. In the present work,
syngas with a ratio of H2:CO =1.5 containing 10% N2 as interꢀ
nal standard was used for the catalytic tests.
In the reactor unit the main gas line was split into 4 branch
lines connected to 4 separate reactors. Each branch line conꢀ
sisted of a MFC, a valve, a backꢀpressure valve and two filters.
All 4 reactors were mounted in an oven, which established the
The ZnO bulk structure was expanded to a 3x3x3 supercell
(Zn54O54) and relaxed using a reduced Gamma centered kꢀ
point sampling grid of 1x1x1 with the same energy cutꢀoff,
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 15
while the ZnCr2O4 spinel bulk was expanded to a 1x1x2 energy of the free molecule and E(ZnCr2O4) is the total energy
1
2
3
4
supercell structure (Zn16Cr32O64) with a reduced kꢀpoint samꢀ
pling grid of 4x4x2. The bulk structures were expanded to
accommodate point defects under periodic boundary condiꢀ
tions for investigations of Cr doping and oxygen vacancy
formation in each of the systems of interest.
of the spinel (100) surface.
The changes in oxidation states of the cations in the bulk
structures was investigated using Bader’s atoms in molecules
method as implemented in VASP by Henkelman et al.45ꢀ47 and
spin magnetization values.
5
6
7
8
9
The nonꢀpolar ZnO (10ꢀ10) and ZnCr2O4 (100) surfaces are
also examined as model catalysts to explore different possible
structures and compositions, with Cr:Zn ratios ranging from 0
– 100%. The surfaces are generated from the lowest energy
bulk structures and modelled using the slab method. For ZnO
(10ꢀ10) and ZnCr2O4 (100), slab thicknesses of 13 Å and 10 Å,
with 9 and 11 atomic layers, are used. The vacuum region
above the surface is 13 Å. The bottom 3 layers of each slab are
fixed while the rest of the slab is allowed to fully relax. The
surfaces are converged using the same parameters as the bulk
calculations
3. Results
a. Characterization of the catalyst precursors
The actual metal compositions of the calcined precursors
determined by ICPꢀOES as well as the textural properties
determined by N2 physisorption are summarized in Table 1.
Small variations between the nominal and actual Zn:Cr ratio
can be observed for all the catalysts. More specifically, the
actual amount of Cr in the catalysts is always less than exꢀ
pected. Based on the calculation of the maximal concentration
of ionic Cr after precipitation, the possibility of incomplete
precipitation of Cr3+ can be excluded. Hence, the loss of Cr
probably occurred during the filtration. The Crꢀcontaining
precipitates formed with a very small particle size during the
coꢀprecipitation were most likely Cr(OH)3 nanoparticles,48 and
a part of these precipitates penetrated the glass fiber filter
during the filtration. For the catalysts containing high amounts
of Cr, a considerably high amount of Na residue was found to
be present. This difficulty of removing Na may be also caused
by the smallꢀsized Cr(OH)3 nanoparticles formed by coꢀ
precipitation. The reaction between acidic Cr(OH)3 and the
basic Na residue presumably prevents the complete removal of
Na.49 For the catalysts with lower Zn:Cr ratio, more smallꢀ
sized Cr(OH)3 was formed, resulting in more difficult filtration
and washing.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The different compositions of the Cr:ZnO systems in the exꢀ
perimental catalysts are modelled by Cr3+ doping in ZnO and
Zn2+ doping on a Cr site in ZnCr2O4 spinel using the bulk and
surface structures. A key property identified experimentally
for methanol synthesis catalyst is the reducibility, which is
modelled by oxygen vacancy formation. We investigate the
effect of doping on the reducibility. The doping of these strucꢀ
tures is carried out in a similar manner to previous studies for
charge compensating mechanisms for doping in Cr2O3,36
44
CeO2,41ꢀ43 or TiO2 by first examining any charge compensatꢀ
ing oxygen vacancies in the doped system and then calculating
the formation energy of the next oxygen vacancy, which reꢀ
duces the oxide. We also investigate the interaction with CO
and H2 by adsorbing the molecules on different sites of the
stoichiometric ZnCr2O4 and Znꢀrich ZnCr2O4 (100) surfaces,
computing the adsorption energy of each molecule as follows:
The specific surface area strongly depends on the Zn:Cr ratio
of the catalysts. With decreasing nominal Zn:Cr ratio from
100:0 to 60:40 the specific surface area increases, and for
Zn:Cr= 60:40 the specific surface area reaches a maximum of
117 m2 gꢀ1. A further decrease of the nominal Zn:Cr ratio from
60:40 to 0:100 results in a decrease in specific surface area.
E
ads = E(CO/H2ꢀZnCr2O4) – [E(CO/H2) + E(ZnCr2O4)]
Where E(CO/H2ꢀZnCr2O4) is the total energy of CO or H2
adsorbed at the spinel (100) surface, E(CO/H2) is the total
Table 1. Nominal and actual metal composition as well as the textural properties of the catalysts after calcination.
Nominal
Actual
Na
(wt %)
ABET
rp
(nm)
Vp
Catalyst
100Zn
Zn:Cr ratio Zn:Cr ratio
(m2 gꢀ1)
(cm3 gꢀ1)
(atomic)
100:0
(atomic)
100:0
83:17
71:29
67:33
62:38
57:43
51:49
46:54
35:65
28:72
0:100
0.1
34
82
95
106
117
112
85
34
18
2
6.8
2.4
2.9
2.7
1.8
2.4
1.9
1.9
1.9
2.4
9.2
0.14
0.11
0.18
0.18
0.17
0.18
0.12
0.07
0.05
0.00
0.03
80Zn20Cr 80:20
70Zn30Cr 70:30
65Zn35Cr 65:35
60Zn40Cr 60:40
55Zn45Cr 55:45
50Zn50Cr 50:50
40Zn60Cr 40:60
33Zn66Cr 33:66
20Zn80Cr 20:80
0.2
0.03
0.04
0.01
0.3
0.7
2.8
4.1
7.2
100Cr
0:100
8.1
3
4
ACS Paragon Plus Environment

Page 5 of 15
ACS Catalysis
When the composition of the catalyst is changed from excess 70:30, the reflections of ZnO generally become broader and
1
2
3
4
5
6
7
8
of Zn to excess of Cr, the specific surface area undergoes a
drastic decline. In this work, the catalysts containing high
levels of Cr generally exhibit lower surface areas and smaller
pore volumes compared with the catalysts containing high
amounts of Zn due to the high contents of Na residue. This
trend in the specific surface area of the catalysts scales with
the catalytic activity shown in section b. Catalytic tests, indiꢀ
cating that surface area is a major factor that influences the
catalytic activity of the ZnO/Cr2O3 catalysts. It should also be
noted that both pure ZnO and Cr2O3 exhibit a significantly
smaller surface area than the ZnO/Cr2O3 catalysts with high
Zn:Cr ratio between 80:20 and 50:50. Therefore, in the
ZnO/Cr2O3 catalyst system, Cr2O3 probably does not act as a
traditional support material that directly increases the surface
of the catalyst. Instead, ZnCr2O4 is the actual support that
increases the surface area of the catalyst and improves the
distribution of ZnO.8
weaker, and a further decrease in the Zn:Cr ratio leads to the
disappearance of the ZnO reflections. The ZnCr2O4 spinel
phase is found to exist in the catalysts with the Zn:Cr ratios
from 70:30 to 20:80. The catalysts containing high levels of Cr
show sharper reflections of ZnCr2O4 compared with the cataꢀ
lysts containing more Zn. Furthermore, the sharp reflections of
Cr2O3 can only be found in the XRD pattern of 100Cr. Addiꢀ
tionally, a series of small and sharp reflections located beꢀ
tween 2θ = 14° and 32° can be observed in the catalysts conꢀ
taining a high amount of Cr, which can be mostly assigned to
zinc and sodium chromate. The presence of sodium chromates
in the calcined catalyst corresponds to the results obtained by
ICPꢀOES, which show a certain amount of Na residue to be
present in the catalyst containing excess of Cr. Obviously, the
residual Na compounds such as Na2CO3 react with Cr2O3 and
O2 forming sodium chromates. In addition to the formation of
sodium chromates, Cr2O3 can also be oxidized yielding
ZnCrO4.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The XRD patterns of the hydroxycarbonate precursors before
calcination are shown in Figure 1. For 100Zn, most of the
reflections can be assigned to zinc hydroxycarbonate. A few
unidentified reflections possibly originate from Zn(OH)2
or/and ZnCO3. For the precursors containing both Cr and Zn,
the existence of the ZnꢀCr hydrotalciteꢀlike phase
Zn4Cr2(OH)12CO3 can be confirmed by identification of the
corresponding characteristic reflections located at 2θ = 34.5°
and 2θ = 61.2°.50ꢀ51 With decreasing nominal Zn:Cr ratio from
80:20 to 20:80, these two reflections become broader and their
intensity decreases, which indicates that the precursors beꢀ
come more amorphous as more Cr is present. The XRD patꢀ
tern of 100Cr shows no identifiable reflections at all. With
decreasing Zn:Cr ratio, more Cr(OH)3 can be formed during
the coꢀprecipitation, which results in a more amorphous state
of the precursors. It has to be pointed out that Cr2(CO3)3 is not
the product of the precipitation of Cr3+ using Na2CO3 soluꢀ
tion.52
Figure 2. XRD patterns of the calcined ZnO/Cr2O3 catalysts.
The catalysts with a Zn:Cr ratio in the range from 55:45 to
33:66 only exhibit the broad reflections of the ZnCr2O4 phase
in addition to the chromates phases. According to Table 1, a
considerable amount of ZnO should be present in these cataꢀ
lysts due to the 1:2 ratio of Zn:Cr in ZnCr2O4, even if all Cr2O3
reacted to form ZnCr2O4 during the calcination, but no characꢀ
teristic ZnO reflections were found. The absence of ZnO in the
ZnO/Cr2O3 catalyst with low Zn:Cr ratio is in agreement with
the results reported by Del Piero et al.,19 who suggested that
ZnO can be dissolved in the ZnCr2O4 spinel to form a nonꢀ
stoichiometric ZnꢀCr spinel. Hence, a small shift of the
ZnCr2O4 reflections is supposed to be detected. The quantity
of nonꢀstoichiometric ZnꢀCr spinel can be estimated by deterꢀ
mination of the lattice parameter of the spinel phase using
quantitative XRD analysis.19,26 However, due to the broad
reflections, this method cannot be applied in the present work.
Figure 1. XRD patterns of the hydroxycarbonate precursors.
Figure 3 shows the effect of the Zn: Cr ratio on the formation
of chromates and spinels during calcination more clearly.
After calcination at 250 °C and 320 °C, the 65Zn catalyst
exhibits a series of broad reflections indicating the existence
of amorphous material. A further increase in temperature led
to the appearance of the sharp reflections of ZnO and
ZnCr2O4. Based on these observation and the XRD patterns of
the used catalyst, it can be assumed that the catalyst contains
5
Figure 2 shows the XRD patterns of the calcined ZnO/Cr2O3
catalysts. It can be seen that the hydroxycarbonates with hyꢀ
drotalciteꢀlike structure were decomposed by calcination at
320 °C for 3 h. The XRD pattern of 100Zn exhibits clear and
sharp reflections of ZnO indicating the formation of relatively
large ZnO particles. With decreasing Zn:Cr ratio from 100:0 to
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 15
amorphous ZnO and ZnCr2O4. The 33Zn catalyst shows sharp phonon scattering processes of ZnO. This observation indiꢀ
1
2
3
4
5
6
7
8
reflections of zinc chromate at the calcination temperature of
250 °C. At higher calcination temperatures spinel reflections
can be found in the XRD patterns. The differences in the XRD
patterns of 65Zn35Cr and 33Zn66Cr at the same calcination
temperature provide a strong indication that the Zn:Cr ratio
plays a crucial role in the phase composition of the calcined
catalyst and the morphology of the each component. For cataꢀ
lysts with high Zn:Cr ratio, amorphous ZnO and spinel are
formed during calcination. Considering the stoichiometry of
the ZnꢀCr hydrotalciteꢀlike compound Zn4Cr2(OH)12(CO3)
found in the precursor, it is assumed that for catalysts with
Zn:Cr ratios between 100:0 and 66:33, the main phases existꢀ
ing in the precursor are ZnꢀCr hydroxycarbonates and
cates that for catalysts with Zn.Cr ratios higher than 33:66,
part of the additional ZnO is dissolved in ZnCr2O4 forming the
nonꢀstoichiometric ZnꢀCr spinel. The bands at around 871
cm−1, 890 cm−1 and 950 cm−1 are ascribed to the Cr–O stretchꢀ
ing vibration of chromates. Catalysts with low Zn:Cr ratio
seem to exhibit more intense bands at these positions. This
observation is consistent with the XRD results indicating that
the formation of chromates during calcination is more severe
leading to larger chromate particle sizes for the Crꢀrich cataꢀ
lysts.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
53
Zn(OH)2 or ZnCO3. During calcination hydrotalciteꢀlike Znꢀ
Cr hydroxycarbonate decomposes to form ZnꢀCr spinel and
ZnO, which is also the product of the decomposition of
Zn(OH)2 and ZnCO3. For catalysts with Zn:Cr ratios smaller
than 66:33, a considerable amount of Cr(OH)3 is also present
in the precursor. Hence, ZnCrO4 was formed by oxidation of
Cr2O3 in the present of ZnO and O2. At high temperatures
ZnCrO4 can decompose back to ZnO and spinel with a relaꢀ
tively large particle size. Thus, the higher Cr content the cataꢀ
lyst has, the more ZnCr2O4 is formed through this pathway
resulting in a lower surface area and more ZnCrO4 in the cataꢀ
lyst.
Figure 4. Raman spectra of the ZnO/Cr2O3 catalysts with
different Zn:Cr ratios.
The XP spectra of the calcined ZnO/Cr2O3 catalysts with difꢀ
ferent Zn:Cr ratios are shown in Figure 5. For all the catalysts,
the peaks were observed at almost the same position. The Zn
2p3/2 peak at 1021.2 eV and the Zn 2p1/2 peak at 1044.2 eV
originate from the Zn2+ species in the nearꢀsurface region of
the catalysts. It is not possible to determine whether the Zn2+
ions are present in ZnO or ZnCr2O4 due to their similar peak
position.54 The Cr 2p3/2 peak at 576.3 eV and the Cr 2p1/2 peak
at 585.9 eV indicate the presence of Cr3+ species as in
ZnCr2O4.55 The Cr 2p3/2 peak at 579.2 eV confirms the presꢀ
ence of a small amount of surface chromate species resulting
from the oxidation during calcination. The atomic Zn:Cr ratio
both in the bulk and on the surface of the catalysts are summaꢀ
rized in Table 2. None of the catalysts exhibits significant zinc
enrichment in the nearꢀsurface region. Instead, a nonꢀ
significant chromium enrichment of the surface is observed.
Therefore, a coreꢀshell structure with a thick amorphous or
highly dispersed ZnO shell on ZnCr2O4 can be excluded. The
ZnꢀCr spinel is assumed to be the dominant state of Zn2+ in the
nearꢀsurface region. This spinel structure is nonꢀstoichiometric
due to the dissolution of ZnO as indicated by the Zn:Cr ratios
higher than 1:2 derived from the XPS results.
Figure 3. XRD patterns of the ZnO/Cr2O3 catalysts with Zn:Cr
ratio of 65:35 (left) and 33:66 (right) calcined at 250 °C, 320 °C
and 500 °C for 3 h.
Figure 4 shows the Raman spectra of the catalysts with differꢀ
ent Zn:Cr ratios. The observed Raman modes and data reportꢀ
ed in literature are summarized in Table S1. Based on the
selection rules, nonꢀdefective ZnCr2O4 spinel should exhibit
three Raman active modes, namely A1g, Eg, and F2g, which are
observed in Figure 4. The sharp peak at 175 cmꢀ1 originates
from the Zn–O bending vibration of the ZnCr2O4 spinel. The
band with medium intensity located at 499 cm−1 and the less
intense bands at 599 cmꢀ1 and 586 cmꢀ1 can be assigned to the
Cr–O stretching vibration (F2g). The intense band at 666 cm−1
is attributed to the symmetric Cr–O stretching vibration of A1g
symmetry originating from the CrO6 groups in the spinel strucꢀ
ture. Only weak and broad peaks are found at 300 cm−1 beꢀ
longing to the vibration mode associated with the multipleꢀ
6
ACS Paragon Plus Environment

Page 7 of 15
ACS Catalysis
Table 2. Bulk and surface composition of the calcined
catalysts with different Zn:Cr ratios before and after reac-
tion.
Figure S2 shows the H2 TPR profiles of the calcined catalysts
1
2
3
4
5
6
7
8
with different Zn:Cr ratios. Pure ZnO (100Zn) exhibits no
reduction peak, while the ZnO/Cr2O3 catalysts with Zn:Cr
ratios from 80:20 to 55:45 show a single wellꢀdefined reducꢀ
tion peak at about 310 °C. The single reduction peak of
40Zn60Cr and 33Zn66Cr shifts to 353 °C and 384 °C, respecꢀ
tively. It can be observed that its intensity increases with deꢀ
creasing Zn:Cr ratio. For 20Zn, three broad reduction peaks
located at 457 °C, 512 °C and 670 °C are found. The H2 TPR
profile of 100Cr consists of overlapping peaks at 576 °C and
673 °C. In addition, two sharp shoulders are found between
550 °C and 580 °C. In literature,56ꢀ58 the single reduction peak
for catalysts with Zn:Cr ratios from 80:20 to 40:60 was asꢀ
cribed to the reduction of the ZnꢀCr spinel. Based on the XRD,
XPS and Raman results in the present work, this peak more
likely originates from the reduction of zinc chromates. For
40Zn60Cr and 33Zn66Cr, the shift of this reduction peak to
the highꢀtemperature region is caused by the larger chromate
particle size. The complicated peaks at high temperature for
20Zn80Cr and 100 Cr originate from the reduction of sodium
chromates, which can be further confirmed by the XRD results
of these samples after the TPR experiments (Figure S3). The
observed reflections of sodium chromite, which is the reducꢀ
tion product of sodium chromate, prove the existence of sodiꢀ
um chromate. Furthermore, the total H2 consumption (Table
S2) increases with decreasing Zn:Cr ratio indicating that the
catalysts with high levels of Cr contain a higher amount of
chromates species.
Zn:Cr ratio
before reacꢀ
tion (bulk)
Zn:Cr ratio
before reacꢀ
Zn:Cr ratio
Sample
after reaction
tion (surface) a(surface)
70Zn30Cr
65Zn35Cr
55Zn45Cr
71:29
67:33
57:43
63:37
59:41
53:47
58:42
56:44
50:50
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
At 60 bar and 260 °Cꢀ300 °C for 60 h (20 h for each temꢀ
perature step) with a space velocity of 4800 ml h−1g−1. A synꢀ
gas mixture with H2:CO:N2=54:36:10 was used as feed.
Figure 5: Zn 2p (left) and Cr 2p (right) XP spectra of the calcined
ZnO/Cr2O3catalysts with different Zn:Cr ratios.
Figure 6. Methanol productivity of catalysts with different Zn:Cr ratios at 260ꢀ300 °C, 60 bar H2:CO:N2=54:36:10 and a flow of 4800
ml h−1g−1.
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 15
b. Catalytic tests
c. Characterization of the catalysts after methanol syntheꢀ
1
sis
The effect of the Zn:Cr ratio on the performance of the cataꢀ
lysts was studied in methanol synthesis. The carbon balance
for all the catalytic tests was between 99% and 101%. At all
three reaction temperatures the highest methanol productivity
was achieved with 65Zn35Cr (Figure 6). The selectivities to
methanol and the undesired product hydrocarbons as well as
the degrees of CO conversion are summarized in Table 3. For
all the catalysts, CO conversion increases with increasing
reaction temperature, but the rate of hydrocarbon formation
increases more rapidly with rising reaction temperature than
that of methanol formation resulting in a lower methanol seꢀ
lectivity at 300 °C. The formation of hydrocarbons over
ZnO/Cr2O3 catalysts has also been observed by Jiao et al.14
When the reaction temperature was high enough (400 °C), Znꢀ
Cr oxide catalysts prepared by coꢀprecipitation exhibited alꢀ
most 100% selectivity towards hydrocarbons and CO2 comꢀ
bined. In contrast to the 100Cr catalyst without Zn, the 100Zn
catalyst without Cr achieves a significant methanol productiviꢀ
ty. The catalysts with Zn:Cr ratios in the range from 70:30 to
50:50 reach high degrees of conversion, high methanol selecꢀ
tivity and low selectivity to hydrocarbons in comparison to the
catalysts containing an excess of Cr. Generally, a high Cr
content leads to high acidity of the catalyst, which favors the
formation of hydrocarbons. Water as the coupled product of
hydrocarbon formation further enhances the formation of CO2.
As a result of the enhanced formation of both hydrocarbons
and CO2, the selectivity to methanol significantly decreases.
2
3
4
5
6
7
8
9
The XRD patterns of the ZnO/Cr2O3 catalysts after reaction
are shown in Figure S7. Compared with the fresh catalysts
(Figure 2), the used catalysts exhibit slightly sharper reflecꢀ
tions, which indicate a small increase of the particle size, but
no severe sintering. Analogous to the fresh catalysts, the used
catalysts also exhibit a change of the dominating phase from
ZnO to ZnCr2O4 with decreasing Zn:Cr ratio. Correspondingꢀ
ly, under reaction conditions neither ZnO nor ZnCr2O4 were
decomposed or transformed into other phases. Furthermore, no
characteristic reflections of Cr2O3 can be found in all the XRD
patterns for the catalyst containing both Zn and Cr, presumaꢀ
bly due to the small particle size of Cr2O3. For the catalysts
with an excess of Cr, the small sharp reflections of the chroꢀ
mates phase can no longer be seen indicating the reduction of
chromates during methanol synthesis. For 100Cr containing
the highest level of residual Na, the sharp reflections suggest
the presence of Naꢀcontaining compounds. Figure 7 demonꢀ
strates more clearly that after reaction weak reflections of ZnO
can be identified for the Zn:Cr =65:35 and 60:40 catalysts.
With further decreasing Zn:Cr ratio, the catalysts only exhibit
more intensive spinel reflections. The XP spectra of the
ZnO/Cr2O3 catalysts after reaction are shown in Figure S8.
The Zn 2p3/2 peak at 1021.1 eV and the broad Cr 2p3/2 peak at
576.2 eV reveal that after reaction Zn2+ and Cr3+ are still the
primary states of Zn and Cr species in the nearꢀsurface region
of the catalyst. The peak originating from the surface chromate
species can no longer be observed indicating that these species
were fully reduced during the catalytic tests, which is in a
good agreement with the XRD results of the catalyst after
reaction (Figures S7 and 7). In no case significant Zn enrichꢀ
ment in the nearꢀsurface region of the catalysts was observed
after reaction (Table 2). These observations support the hyꢀ
pothesis that the nonꢀstoichiometric ZnꢀCr spinel is the active
phase in ZnO/Cr2O3 catalyst for methanol synthesis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Furthermore, the high content of Na residue may also play an
important role in the low conversion and high selectivity toꢀ
ward hydrocarbons for the catalysts containing excess of Cr.
As shown in Table 1, with decreasing Zn:Cr ratio from 50:50
to 0:100 the content of the Na residue increases drastically
leading to the formation of more chromate species, which
further results in the rapid decrease in surface area. As a result,
the degree of conversion for these catalysts decreases with
increasing Na content. In addition, the selectivity towards high
products also significantly increases with increasing Na conꢀ
tent. This observation corresponds to the promoting effect of
alkali metals on the formation of longꢀchain products, which
has been reported in both FischerꢀTropsch synthesis59ꢀ60and
higher alcohol synthesis.54ꢀ55 Additionally, a series of Naꢀfree
catalysts was synthesized by using (NH4)2CO3 as the precipitaꢀ
tion reagent, while other preparation and test conditions were
not changed. As shown in Figure S4, the Na residue was conꢀ
firmed to cause the drastic decrease in surface area. Correꢀ
spondingly, for catalysts with Zn:Cr <50:50, the application of
a Naꢀfree precipitation reagent can significantly improve the
methanol productivity (Figure S5) due to the absence of sodiꢀ
um chromate species (Figure S6). More importantly, the cataꢀ
lyst with Zn:Cr=65:35 still exhibits the highest methanol
productivity and the dependence of catalytic activity of the
catalyst on the Zn:Cr ratio is not significantly changed.
Figure 7. XRD patterns of the ZnO/Cr2O3 catalysts with Zn:Cr
ratios from 65:35 to 50:50 before and after reaction at 60 bar and
260 °Cꢀ300 °C for 60 h (20 h for each temperature step) with a
space velocity of 4800 ml h−1g−1. A syngas mixture with
H2:CO:N2=54:36:10 was used as feed.
8
ACS Paragon Plus Environment

Page 9 of 15
ACS Catalysis
Table 3. The catalytic performance of the ZnO/Cr2O3 catalysts with different Zn:Cr ratio in methanol synthesis.
1
260 °C
280 °
300 °C
2
3
4
5
6
7
8
9
a
a
a
catalyst
SMeOH
(%)
SHC
(%)
SMeOH
(%)
SHC
(%)
SMeOH
(%)
SHC
(%)
XCO
(%)
XCO (%)
XCO (%)
100Zn
1.3
1.9
1.5
1.8
1.7
1.7
1.3
0.5
0.4
0.3
0.2
52.1
20.4
23.2
9.6
2.9
3.6
2.9
3.3
3.2
3.1
2.3
0.9
0.6
0.8
0.5
40.9
25.0
25.3
16.3
10.0
10.9
12.4
6.2
6.0
6.8
5.6
5.9
5.8
5.7
4.0
1.9
0.9
1.1
1.2
27.5
32.3
29.6
21.1
11.3
13.2
16.8
7.2
80Zn20Cr
75Zn25Cr
65Zn35Cr
60Zn40Cr
55Zn45Cr
50Zn50Cr
40Zn60Cr
33Zn66Cr
20Zn80Cr
100Cr
46.3
79.6
86.7
81.5
82.7
86.7
46.7
13.6
3.5
41.8
66.3
77.6
76.1
72.6
87.1
55.8
13.4
4.7
34.6
56.9
75.2
69.8
63.2
83.7
53.9
14.5
7.2
6.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6.8
7.9
6.6
27.1
45.1
63.8
66.4
20.7
45.9
60.0
65.1
21.1
45.1
56.1
67.1
4.6
6.8
5.2
P = 60 bar, H2:CO:N2 = 54:36:10, GHSV = 4800 ml h−1 g−1
a Hydrocarbons with carbon numbers of 1–6.
d. Density functional theory calculations
The formation of an oxygen vacancy reduces Zn cations, reꢀ
gardless of the Cr doping concentration.
To aid in determining the active phase of the ZnO/Cr2O3
methanol synthesis catalyst we have investigated different
compositions of ZnO/Cr2O3 using density functional theory
calculations. We firstly consider Cr doping in bulk ZnO and in
the nonpolar (10ꢀ10) ZnO surface by replacing Zn2+ lattice
cation sites with Cr3+ cations at different concentrations. The
first Cr3+ cations substitutionally dope a Zn2+ cation site for a
2% doping concentration in either bulk or surface. However
for larger concentrations of Cr3+, different Zn2+ substitutional
sites are explored to determine the relative distribution of Cr in
the ZnO structures. Once the lowest energy configurations for
different concentrations are established, then oxygen vacancy
formation is examined by removing different oxygen atoms
from the doped structures to determine the lowest oxygen
vacancy formation energy. This procedure is followed for all
doped systems.
Doping both the ZnO bulk and surface with Cr does not imꢀ
prove the oxygen vacancy formation as the calculated forꢀ
mation energies are always larger than the undoped structures.
The origin of this result is that the Cr3+ dopant on a Zn2+ site
introduces an extra electron into the system, and as the forꢀ
mation of an oxygen vacancy is also a reducing process (reꢀ
leasing two electrons into the system), it becomes increasingly
more difficult to form oxygen vacancies with increasing the
number of Cr dopants in the structures. Therefore, the Cr
doped ZnO structure as a possible composition with an excess
of Zn over Cr cannot be the active phase for methanol syntheꢀ
sis, as the formation of active oxygen vacancies has a high
energy cost at the optimum experimentally determined ratio of
65:35 Zn:Cr.
The lowest energy configurations for the Cr distributions for
1% ꢀ 10% doping in bulk ZnO and the (10ꢀ10) surface of ZnO,
and the positions of the oxygen vacancies are shown in the
Supporting Information, Figures S9 and S10. The lowest oxyꢀ
gen vacancy energies are shown as a function of the Cr doping
concentration in Figure 8 to examine the effect of bulk and
surface Cr doping of ZnO on the reducibility. For Cr doping in
bulk ZnO, the oxygen vacancy formation energy increases
with increasing Cr doping concentration up to 4% and remains
constant thereafter, showing no change up to 10%. Larger
doping concentrations beyond this value are expected to show
no enhancement in the oxygen vacancy formation energy. The
oxygen vacancy formation energy for Cr doping in the ZnO
surface follows a similar trend to the bulk, increasing up to 8%
Cr doping while at 10% the oxygen vacancy formation energy
is similar to the undoped surface.
Figure 8. The oxygen vacancy formation energies for Cr dopꢀ
ing in the ZnO bulk and the (10ꢀ10) surface.
Computed Bader charges and spin magnetizations for the Cr
cations are between 10.65 and 10.70 electrons and 3 spins
when Cr is doped into ZnO. These are typical of Cr3+ cations,
and this oxidation state is maintained up to 10 % Cr doping.
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 15
The formation of oxygen vacancies was investigated in the
Table 4. The calculated formation energies for compensat-
1
2
3
4
5
6
7
8
ZnCr2O4 spinel bulk and the (100) surface in a similar fashion.
The (100) surface is studied as recent work indicates that this
is the stable surface of the ZnꢀCr spinel rather than the (111)
surface described in older studies using interatomic
potentials.61 The calculated oxygen vacancy formation energy
for the bulk ZnCr2O4 spinel is +4.41 eV, which is larger than
ZnO and Cr doped ZnO. Doping the ZnCr2O4 spinel with
excess Zn replaces an octahedrally coordinated Cr3+ cation
with a Zn2+ cation which is a lower valence dopant. Two Zn2+
must then be incorporated into the spinel structure (forming
two holes on neighboring oxygen atoms) with charge compenꢀ
ing, E[comp] and reducing E[reducing] oxygen vacancies
for different concentrations of Zn in the ZnCr2O4 spinel
surface.
E[comp]
(eV)
E[reducing]
(eV)
0% Zn doping
ꢀꢀꢀ
+1.59
(30% Zn content)
6.66% Zn doping
(40% Zn content)
13.33% Zn doping
(46.66% Zn content)
20% Zn doping
9
+0.41
ꢀ0.05
ꢀ0.02
ꢀ0.25
+2.85
+2.02
+1.58
+1.55
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
sation by oxygen vacancy formation, described as a [2ꢀꢁꢂꢄ ꢃ
+
ꢅ∙∙] species in KrögerꢀVink notation; this is similar to Zn
ꢆ
doped bulk Cr2O3.36 The calculated oxygen vacancy formation
energy for Zn doping in the spinel is +3.49 eV showing an
improvement over the undoped spinel bulk, but the energy
cost is higher than for ZnO. Upon further Zn doping of the
bulk spinel the oxygen vacancy formation energy increases.
(53.33% Zn content)
26.67% Zn doping
(60% Zn content)
In studying the formation of oxygen vacancies in the (100)
surface of the ZnCr2O4 spinel we examined surface and subꢀ
surface oxygen vacancies and we find that surface oxygen
vacancy formation is preferred with a calculated formation
energy of +1.59 eV. This is significantly lower than oxygen
vacancy formation in the Cr doped ZnO systems and the spinel
bulk indicating that the spinel surface is more reducible.
The structures for the (100) surface of the spinel with varying
Zn content are shown in Figure 9. The Zn2+ dopants preferenꢀ
tially replace all the available surface Cr3+ lattice sites up to
46.66% Zn content (Figure 9c). Further Zn doping then reꢀ
places Cr cation sites in the subꢀsurface layer and for 60% Zn
content (26.66% doping) all Cr cations in the surface and subꢀ
surface layers have been replaced by Zn cations. The compenꢀ
sating oxygen vacancies are preferentially removed from the
surface and subꢀsurface regions, resulting in a 1:1 ratio of
Zn:O in this region in which all Zn have a 2+ oxidation state.
This structure has the appearance of a thin amorphous
ZnO layer on the ZnCr2O4 spinel structure, which accom-
modates the excess Zn on the original Cr3+ sites which has
an overall stoichiometry similar to experiment. This region
contains undercoordinated Zn and O sites on the surface
and promotes oxygen vacancy formation at the surface.
Thus we can unravel, for the first time, the precise atomic
level nature of the active, Znꢀrich ZnCr2O4 spinel catalyst.
Zn doping in the (100) ZnCr2O4 spinel surface is investigated
to model the incorporation of excess Zn in the nonꢀ
stoichiometric methanol catalyst described in the experimental
studies. Similar to the spinel bulk, the Zn2+ dopants replace
Cr3+ cations on their lattice sites and are compensated by forꢀ
mation of charge balancing oxygen vacancies; Table 4 shows
that the formation energies for the charge compensating oxyꢀ
gen vacancies are close to 0 eV which is typical for spontaneꢀ
ous formation of a charge compensating oxygen vacancy. In
the (100) spinel surface, Zn dopants can substitute at surface
or subꢀsurface Cr sites or in the bulkꢀlike region of the surface
slab and we examined these sites for Zn doping. We also inꢀ
vestigate the possible compensating oxygen vacancy sites.
Once the lowest energy charge balanced Zn doping configuraꢀ
tion is established, the reducing oxygen vacancy is investigatꢀ
ed to determine the reducibility of the surface compared to the
undoped surface. Zn doping contents up to 60% Zn are studꢀ
ied, which is close to the optimum experimental Zn:Cr ratio.
We note from Table 4 that both the stoichiometric and the Znꢀ
excess spinel structures have low formation energies for the
reducing oxygen vacancy. However, this process is not the
only important process in the methanol synthesis. The initial
adsorption and activation of CO and hydrogen on the catalyst
surface are also important. Figure 10 shows the adsorption
structures for a H atom and the CO molecule adsorbed on the
stoichiometric and Znꢀexcess spinel surface. On ZnCr2O4, the
computed H and CO adsorption energies are ꢀ1.18 eV and ꢀ
2.08 eV, relative to 1/2H2 and free CO. With excess Zn, the
corresponding adsorption energies are ꢀ0.99 eV and ꢀ1.14 eV
for H and CO. On both surfaces CO and H adsorb favorably
and the further reaction of the molecules to hydrogenate CO to
methanol depends on the strength of the initial adsorption. We
therefore propose that the weaker, but still favorable, H and
CO adsorption energies on the nonꢀstoichiometric ZnCr2O4
allow for more favorable hydrogenation of CO, thus enhancꢀ
ing the production of methanol.
The calculated formation energies for the reducing oxygen
vacancies are shown in Table 4. These oxygen vacancy forꢀ
mation energies show that initial Zn doping at 6.66% increases
the vacancy formation energy. However, further Zn doping of
the surface significantly enhances oxygen vacancy formation
in the surface and thus increases the surface reducibility. For
20 – 26.67% Zn doping, the oxygen vacancy formation enerꢀ
gies are similar to the undoped spinel surface.
10
ACS Paragon Plus Environment

Page 11 of 15
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. The calculated local structure for (a) 0%, (b) 6.67%, (c)
13.33%, (d) 20% and (e) 26.67% Zn doping in the (100) surface
of the ZnCr2O4 spinel.
Figure 10. The calculated atomic structure for (a) H atom adsorpꢀ
tion and (b) CO adsorption on the stoichiometric ZnCr2O4 spinel.
Panels (c) and (d) show the atomic structure for H atom and CO
adsorption on the Znꢀrich ZnCr2O4 spinel from our model of the
Znꢀrich ZnCr2O4 spinel, which is shown in Figure 9e. In this
Figure, H is the white sphere, C is the grey sphere. The arrow and
the orange ring indicate the adsorbates.
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 15
4. Discussion
As a result, the catalyst with a ratio of Zn:Cr=65:35 presumaꢀ
1
2
3
4
5
6
7
8
bly contains the highest amount of ZnCr2O4, implying that the
highest amount of nonꢀstoichiometric spinel can be formed.
Indeed, the catalyst with Zn:Cr=65:35 exhibits the highest
productivity in methanol synthesis. With further increasing
Zn:Cr ratio from 65:35, the catalyst contain less ZnCr2O4 and
the larger amount of Znꢀcontaining hydroxycarbonate phases
favors the formation of ZnO. The important role of
Zn4Cr2(OH)12CO3 is that both Zn2+ and Cr3+ can exist in the
same phase, which is beneficial for the formation of ZnꢀCr
nonꢀstoichiometric spinel during calcination.
In recent literature the nonꢀstoichiometric ZnꢀCr spinel has
been repeatedly reported as the active phase of ZnO/Cr2O3
catalysts applied in methanol synthesis,18ꢀ22, 26, 55ꢀ58 which origꢀ
inates from the dissolution of excess ZnO in ZnCr2O4. In the
present work, the diffraction patterns of the calcined Znꢀrich
catalysts with Zn:Cr ratios in the range from 65:35 to 50:50
shown in Figure 2 do not indicate the presence of ZnO, altꢀ
hough a considerable amount of ZnO should be present in
these catalysts according to their composition. In the Raman
spectra shown in Figure 4, the weak and broad band originatꢀ
ing from the multipleꢀphonon scattering processes of ZnO
further confirms that the amount of detected ZnO was always
smaller than that expected for a simple phase mixture of ZnO
and ZnCr2O4 with Zn:Cr ratios in the range from 65:35 to
50:50. The existence of a substantial amount of highly disꢀ
persed or amorphous ZnO, which is not detectable by XRD
and Raman spectroscopy, has been excluded by the XPS reꢀ
sults, in which no Zn enrichment was observed for the calꢀ
cined catalysts before or after reaction (Table 2). All these
observations indicate the presence of nonꢀstoichiometric Znꢀ
Cr spinel formed by the dissolution of ZnO in ZnCr2O4. For
catalysts even with a Zn:Cr ratio of 65:35, essentially all Zn2+
ions are present in the nonꢀstoichiometric ZnꢀCr spinel due to
the high solubility of ZnO in ZnCr2O4.23,26 According to the
DFT results this Znꢀrich nonꢀstoichiometric ZnꢀCr spinel
system is in fact composed of a thin ZnOꢀlike layer supported
on the spinel surface, that results from replacement of surface
and subꢀsurface Cr cations with Zn, rather than bulk doping of
the spinel.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT calculations confirm for the first time that the active
phase of the catalyst is not the ZnO structure doped with Cr, as
the formation of oxygen vacancies is not favorable and thus
this composition will not be active for methanol synthesis. The
calculations show that the (100) surface of the ZnCr2O4 spinel
has favorable oxygen vacancy formation energies, which has
been shown to be a key property for methanol synthesis. For
the nonꢀstoichiometric spinel structure with excess Zn (60%)
content, the nonꢀstoichiometric surface is terminated by an
amorphousꢀlike thin layer of ZnO. This is the active phase for
the methanol synthesis catalyst, as it has a low oxygen vacanꢀ
cy formation energy, favorable, but not too strong H and CO
adsorption energies and therefore has activity required for
methanol synthesis.
The drastic loss of activity for the catalysts containing an
excess of chromium can be explained analogously. Due to
stoichiometric constraints the higher Cr content leads to less
hydrotalciteꢀlike Zn4Cr2(OH)12CO3 during coꢀprecipitation and
more Cr being presumably present in the form of amorphous
Cr(OH)3. Furthermore, due to the acidity of Cr(OH)3, the
sodium residues cannot be completely removed by washing
after coꢀprecipitation (Table 1). Hence, after calcination the
amount of amorphous ZnO and spinel generated from decomꢀ
position of the Zn4Cr2(OH)12CO3 decreases, and more Cr2O3
formed by decomposition of Cr(OH)3 can react with ZnO or
sodium residues in the present of O2 to form chromates (Figꢀ
ure 2). This competitive reaction can negatively influence the
formation of the nonꢀstoichiometric spinel, and the formed
chromate can partially decompose or be reduced to form spiꢀ
nel again, but with high crystallinity and large particle sizes.
Further studies are in progress aiming at further increasing the
amount of nonꢀstoichiometric ZnꢀCr spinel by improving the
precipitation and postꢀtreatment conditions and at identifying
the amorphous thin ZnO layer on the ZnꢀCr spinel by low
energy ion scattering spectroscopy and highꢀresolution transꢀ
mission electron microscopy. Furthermore, since the formation
of the ZnꢀCr nonꢀstoichiometric spinel increases the amount of
Zn2+ cations located at octahedral sites, Xꢀray absorption specꢀ
troscopy may be a suitable characterization method to deterꢀ
mine its amount.
In spite of the absence of bulk ZnO, these catalysts achieve
high activity in methanol synthesis, suggesting the nonꢀ
stoichiometric ZnꢀCr spinel to be the active phase. After reacꢀ
tion weak reflections of ZnO are found for the catalysts with
Zn:Cr ratios of 65:35 and 60:40 (Figure 8), indicating a minor
increase of the amount of ZnO in the catalyst. This segregation
seems to be correlated with the slight deactivation of the cataꢀ
lysts with time on stream, during which a small amount of
nonꢀstoichiometric ZnꢀCr spinel changed towards its stoichiꢀ
ometric form accompanied by the formation of ZnO.19 Furꢀ
thermore, as shown in Figure 3, high calcination temperatures
up to 500 °C can accelerate the segregation of the nonꢀ
stoichiometric spinel structure resulting in the stoichiometric
ZnꢀCr spinel ZnCr2O4 and crystalline ZnO.19, 55
The Zn:Cr ratio can significantly influence the amount of the
catalytically active nonꢀstoichiometric ZnꢀCr spinel in the
catalyst due to its effect on the structure of the coꢀprecipitated
catalyst precursor. In Figure 1, the reflections of
Zn4Cr2(OH)12CO3 can be observed for Zn:Cr ratios in the
range from 80:20 to 20:80, indicating that the formation of this
hydrotalciteꢀlike ZnꢀCr hydroxycarbonate is favored during
coꢀprecipitation. Due to the stoichiometric constraint of
Zn:Cr=2:1 in Zn4Cr2(OH)12CO3, the highest amount of
Zn4Cr2(OH)12CO3 can be obtained with Zn:Cr=65:35, and a
further increase or decrease in the Zn:Cr ratio leads to the
formation of Znꢀrich hydroxycarbonates and amorphous
Cr(OH)3, respectively. As shown in Figure 2,
Zn4Cr2(OH)12CO3 decomposes during calcination forming
ZnO and ZnCr2O4. It is worth noting that the formation of
ZnCr2O4 via a solidꢀstate reaction of ZnO and Cr2O3 would
require much high temperatures. Hence, Zn4Cr2(OH)12CO3 is
assumed to be the major source for ZnCr2O4 under the applied
mild calcination conditions in synthetic air at 320 °C for 3 h.62
5. Conclusions
A series of ZnO/Cr2O3 catalysts with different Zn:Cr ratios
was prepared by coꢀprecipitation at pH 7 and calcination at
320 °C. The Zn:Cr ratio was found to have a significant influꢀ
ence on the structural and catalytic properties. The catalysts
with Zn:Cr ratios between 70:30 and 55:45 had relatively high
specific surface areas of about 100 m2 gꢀ1, whereas the small
specific surface area of Crꢀrich catalysts is mainly attributed to
the formation of chromates during calcination in the presence
of sodium residues after coꢀprecipitation. XRD, XPS and
12
ACS Paragon Plus Environment

Page 13 of 15
ACS Catalysis
Raman results confirm that the nonꢀstoichiometric ZnꢀCr
AUTHOR INFORMATION
1
2
3
4
5
6
7
8
spinel is present in the calcined ZnO/Cr2O3 catalysts. The
catalyst with a Zn:Cr ratio of 65:35 exhibits the highest methꢀ
anol productivity of about 105 g (kgcat·h)ꢀ1 at 300°C. This ratio
corresponds to the stoichiometric Zn:Cr ratio of the hyꢀ
drotalcite–like precursor Zn4Cr2(OH)12CO3 formed during coꢀ
precipitation. It is suggested that 65:35 is the optimal Zn:Cr
ratio yielding the highest amount of Zn4Cr2(OH)12CO3, which
further decomposes during calcination forming the highest
amount of the nonꢀstoichiometric ZnꢀCr spinel as active phase
of ZnO/Cr2O3 catalysts applied in methanol synthesis. Density
functional theory calculations have characterized the nonꢀ
stoichiometric ZnꢀCr catalyst with excess Zn as an amorphous
ZnO layer supported on the ZnCr2O4 spinel structure thus
confirming this system as the active methanol synthesis cataꢀ
lyst with the lowest formation energy for oxygen vacancies.
Corresponding Authors
*muhler@techem.rub.de
*michael.nolan@tyndall.ie
Author Contributions
The manuscript was written through contributions of all authors.
All authors have given approval to the final version of the manu-
script.
9
Funding Sources
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
European FP7 NMP 2013 project BIOGO (grant no: 604296)
ACKNOWLEDGMENT
This work was supported by the European FP7 NMP 2013 project
BIOGO (grant no: 604296) (Catalytic partial oxidation of biogas
and reforming of pyrolysis oil for synthetic gas production and
conversion into fuels, www.biogo.eu). J.J.C and M.N would like
to acknowledge use of the Science Foundation Ireland (SFI)
funded computational resources at Tyndall and at the Irish Centre
for High End Computing (ICHEC), and support from the COST
action CM1104.
ASSOCIATED CONTENT
Supporting Information. Flow scheme of the highꢀpressure setꢀ
up. H2 TPR profiles and H2 consumption. Observed Raman
modes, results obtained with (NH4)2CO3 as precipitating agent,
XRD patterns and XP spectra after reaction. The various tested
sites for Cr doping in ZnO by DFT+U calculations. This material
is available free of charge via the Internet at http://pubs.acs.org.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
Wender, I. Cat. Rev. - Sci. Eng. 1984, 26, 303ꢀ321.
Calkins, W. H. Cat. Rev. - Sci. Eng. 1984, 26, 347ꢀ358.
Mittasch, A.; Schneider, C. US1201850, 1916.
Mittasch, A.; M. Pier; Winkle, K. US1558559, 1925.
Mittasch, A.; Pier, M. US1569775, 1926.
Gallagher, J.; Kidd, J. GB1159035, 1969.
Cornthwaite, D. GB1296212, 1972.
Natta, G. In Catalysis; Emmet, P. H., Eds.; Reinhold Publ.
21.
A. J. Chem. Soc., Faraday Trans. I 1989, 85, 237ꢀ249.
22. Riva, A.; Trifiro, F.; Vaccari, A.; Mintchev, L.; Busca, G.
J. Chem. Soc., Faraday Trans. I 1988, 84, 1423ꢀ1435.
23. Grimes, R. W.; Binks, D. J.; Lidiard, A. B. Philos. Mag. A
1995, 72, 651ꢀ668.
Giamello, E.; Fubini, B.; Bertoldi, M.; Busca, G.; Vaccari,
24.
25.
Bone, W. A. Trans. Inst. Chem. Eng. 1930, 8, 98ꢀ106.
Molstad, M. C.; Dodge, B. F. Ind. Eng. Chem. 1935, 27,
Corp.: New York, 1955; vol. 3, p 349.
134ꢀ140.
26.
9.
Riva, A.; Trifiro, F.; Vaccari, A.; Busca, G.; Mintchev, L.;
Errani, E.; Trifiro, F.; Vaccari, A.; Richter, M.; Del Piero,
Sanfilippo, D.; Manzatti, W. J. Chem. Soc., Faraday Trans. I 1987,
83, 2213ꢀ2225.
G. Catal. Lett. 1989, 3, 65ꢀ72.
27.
Bradford, M. C. J.; Konduru, M. V.; Fuentes, D. X. Fuel
10.
Epling, W. S.; Hoflund, G. B.; Hart, W. M.; Minahan, D.
Process. Technol. 2003, 83, 11ꢀ25.
M. J. Catal. 1997, 169, 438ꢀ446.
28.
29.
Perdew, J. P.; Yue, W. Phys. Rev. B 1986, 33, 8800ꢀ8802.
Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett.
11.
Epling, W. S.; Hoflund, G. B.; Hart, W. M.; Minahan, D.
M. J. Catal. 1997, 172, 13ꢀ23.
1996, 77, 3865ꢀ3868.
12.
Minahan, D. M.; Epling, W. S.; Hoflund, G. B. Appl.
30.
15ꢀ50.
31.
Kresse, G.; Furthmüller, J. Comput. Mater. Sci. 1996, 6,
Catal. A 1998, 166 , 375ꢀ385.
13.
Ereña, J.; Arandes, J. M.; Bilbao, J.; Aguayo, A. T.; de
Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169ꢀ
Lasa, H. I. Ind. Eng. Chem. Res. 1998, 37, 1211ꢀ1219.
11185.
32.
14.
Jiao, F.; Li, J.; Pan, X.; Xiao, J.; Li, H.; Ma, H.; Wei, M.;
Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758ꢀ1775.
Pan, Y.; Zhou, Z.; Li, M.; Miao, S.; Li, J.; Zhu, Y.; Xiao, D.; He, T.;
Yang, J.; Qi, F.; Fu, Q.; Bao, X. Science 2016, 351, 1065ꢀ1068.
33.
Dudarev, S. L.; Liechtenstein, A. I.; Castell, M. R.; Briggs,
G. A. D.; Sutton, A. P. Phys. Rev. B 1997, 56, 4900ꢀ4908.
34. Dudarev, S.; Botton, G.; Savrasov, S.; Humphreys, C.;
Sutton, A. Phys. Rev. B 1998, 57, 1505ꢀ1509.
15.
867ꢀ871.
16.
Cryder, D. S.; Frolich, P. K. Ind. Eng. Chem. 1929, 21,
Boccuzzi, F.; Garrone, E.; Zecchina, A.; Bossi, A. J. Catal.
35.
125426.
36.
Rohrbach, A.; Hafner, J.; Kresse, G. Phys. Rev. B 2004, 70,
1978, 51, 160ꢀ168.
17.
18.
Kung, H. H. Cat. Rev. - Sci. Eng. 1980, 22, 235ꢀ259.
Kurtz, M.; Strunk, J.; Hinrichsen, O.; Muhler, M.; Fink, K.;
Carey, J. J.; Legesse, M.; Nolan, M. J. Phys. Chem. C
2016, 120, 19160ꢀ19174.
37. Murnaghan, F. Proc. Natl. Acad. Sci. U.S.A. 1944, 30, 244ꢀ
247.
Meyer, B.; Wöll, C. Angew. Chem. Int. Ed. 2005, 44, 2790ꢀ2794.
19. Piero, G. D.; Trifiro, F.; Vaccari, A. J. Chen. Soc. Chem.
Commun. 1984, 656ꢀ658.
38.
Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188ꢀ
20.
Bertoldi, M.; Fubini, B.; Giamello, E.; Busca, G.; Trifiro,
5192.
F.; Vaccari, A. J. Chem. Soc., Faraday Trans. I 1988, 84, 1405ꢀ1421.
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 15
39.
Moureen, C. K.; Phillip, T. B.; Stephanie, L. M.; Michael,
53.
Choudhary, V. R.; Chaudhari, P. A.; Narkhede, V. S.
W. G.; Joshua, A. K.; Ram, S.; Matthew, R. S.; YoungꢀIl, K. J.
Phys.:Cond. Matt. 2013, 25, 326001.
Catal. Commun. 2003, 4, 171ꢀ175.
1
2
3
4
5
6
7
8
54.
Epling, W. S.; Hoflund, G. B.; Minahan, D. M. J. Catal.
40.
41.
42.
Reeber, R. R. J. Appl. Phys. 1970, 41, 5063ꢀ5066.
Carey, J. J.; Nolan, M. Appl. Catal. B 2016, 197, 324ꢀ336.
Carey, J. J.; Nolan, M. Catal. Sci. Technol. 2016, 6, 3544ꢀ
1998, 175, 175ꢀ184.
55.
Tian, S.; Wang, S.; Wu, Y.; Gao, J.; Bai, Y.; Wang, P.;
Xie, H.; Han, Y.; Tan, Y. J. Mol. Catal. A 2015, 404, 139ꢀ147.
3558.
43.
56.
Vitidsant, T.; Tsubaki, N. Appl. Catal. A 2015, 505, 141ꢀ149.
57. Tian, S.; Wang, S.; Wu, Y.; Gao, J.; Wang, P.; Xie, H.;
Yang, G.; Han, Y.; Tan, Y. Catal. Sci. Technol. 2016, 6, 4105ꢀ4115.
58. Tian, S.; Wang, S.; Wu, Y.; Gao, J.; Xie, H.; Li, X.; Yang,
G.; Han, Y.; Tan, Y. RSC Adv. 2015, 5, 89273ꢀ89281.
59. NgantsoueꢀHoc, W.; Zhang, Y.; O’Brien, R. J.; Luo, M.;
Davis, B. H. Appl. Catal., A 2002, 236, 77ꢀ89.
60. Riedel, T.; Claeys, M.; Schulz, H.; Schaub, G. Appl. Catal.
A 1999, 186, 201ꢀ213.
Tan, L.; Yang, G.; Yoneyama, Y.; Kou, Y.; Tan, Y.;
Nolan, M. J. Phys. Chem. C 2011, 115, 6671ꢀ6681.
44.
Iwaszuk, A.; Nolan, M. J. Phys.:Cond. Matt. 2011, 23,
334207.
45.
Henkelman, G.; Arnaldsson, A.; Jónsson, H. Comput.
9
Mater. Sci. 2006, 36, 354ꢀ360.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
46.
Sanville, E.; Kenny, S. D.; Smith, R.; Henkelman, G. J.
Comput. Chem. 2007, 28, 899ꢀ908.
47.
Tang, W.; Sanville, E.; Henkelman, G. J. Phys.:Cond.
Matt. 2009, 21, 084204.
48.
Music, S.; Maljkovic, M.; Popovic, S.; Trojko, R. Croat.
61.
Rasmussen, M. K.; Foster, A. S.; Hinnemann, B.; Canova,
Chem. Acta. 1999, 72, 789ꢀ802.
F. F.; Helveg, S.; Meinander, K.; Martin, N. M.; Knudsen, J.; Vlad,
A.; Lundgren, E.; Stierle, A.; Besenbacher, F.; Lauritsen, J. V. Phys.
Rev. Lett. 2011, 036102.
49.
345ꢀ349.
50.
1178.
51.
Rai, D.; Sass, B. M.; Moore, D. A. Inorg. Chem. 1987, 26,
Parida, K.; Mohapatra, L. Dalton Trans. 2012, 41, 1173ꢀ
del Arco, M.; Rives, V.; Trujillano, R.; Malet, P. J. Mater.
62.
Lal, M.; Howe, A. J. Solid State Chem. 1981, 39, 368ꢀ376.
Chem. 1996, 6, 1419ꢀ1428.
52.
W. E. Motzer, In Chromium(VI) Handbook; Guertin, J.;
Jacobs, J. A.; Avakian, C. P., Eds.; CRC Press: Boca RatonꢀLondonꢀ
New YorkꢀWashington D.C., 2004; p 23.
14
ACS Paragon Plus Environment

Page 15 of 15
ACS Catalysis
TOC Graphic
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
ACS Paragon Plus Environment