Strong Electronic Oxide-Support Interaction over In2O3/ZrO2for Highly Selective CO2Hydrogenation to Methanol

Article abstract of DOI:10.1021/jacs.0c07195

Metal oxides are widely employed in heterogeneous catalysis, but it remains challenging to determine their exact structure and understand the reaction mechanisms at the molecular level due to their structural complexity, in particular for binary oxides. This paper describes the observation of the strong electronic interaction between In2O3 and monoclinic ZrO2 (m-ZrO2) by quasi-in-situ XPS experiments combined with theoretical studies, which leads to support-dependent methanol selectivity. In2O3/m-ZrO2 exhibits methanol selectivity up to 84.6% with a CO2 conversion of 12.1%. Moreover, at a wide range of temperatures, the methanol yield of In2O3/m-ZrO2 is much higher than that of In2O3/t-ZrO2 (t-: tetragonal), which is due to the high dispersion of the In-O-In structure over m-ZrO2 as determined by in situ Raman spectra. The electron transfer from m-ZrO2 to In2O3 is confirmed by XPS and DFT calculations and improves the electron density of In2O3, which promotes H2 dissociation and hydrogenation of formate intermediates to methanol. The concept of the electronic interaction between an oxide and a support provides guidelines to develop hydrogenation catalysts.

Full text of DOI:10.1021/jacs.0c07195

pubs.acs.org/JACS
Article
Strong Electronic Oxide−Support Interaction over In2O3/ZrO2 for
Highly Selective CO2 Hydrogenation to Methanol
Chengsheng Yang,€ Chunlei Pei,€ Ran Luo, Sihang Liu, Yanan Wang, Zhongyan Wang, Zhi-Jian Zhao,
and Jinlong Gong*
Cite This: https://dx.doi.org/10.1021/jacs.0c07195
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Metal oxides are widely employed in heterogeneous
catalysis, but it remains challenging to determine their exact structure
and understand the reaction mechanisms at the molecular level due to
their structural complexity, in particular for binary oxides. This paper
describes the observation of the strong electronic interaction between
In2O3 and monoclinic ZrO2 (m-ZrO2) by quasi-in-situ XPS
experiments combined with theoretical studies, which leads to
support-dependent methanol selectivity. In2O3/m-ZrO2 exhibits
methanol selectivity up to 84.6% with a CO2 conversion of 12.1%.
Moreover, at a wide range of temperatures, the methanol yield of
In2O3/m-ZrO2 is much higher than that of In2O3/t-ZrO2 (t-:
tetragonal), which is due to the high dispersion of the In−O−In structure over m-ZrO2 as determined by in situ Raman spectra. The
electron transfer from m-ZrO2 to In2O3 is confirmed by XPS and DFT calculations and improves the electron density of In2O3,
which promotes H2 dissociation and hydrogenation of formate intermediates to methanol. The concept of the electronic interaction
between an oxide and a support provides guidelines to develop hydrogenation catalysts.
hot electrons are generated via the surface electron excitation
and turn into low-energy electrons within the length scale of
the electron mean free path (electron transfer), which can
change the charge state of the surface adsorbate, activate
reaction intermediates, and mediate the activity or selectivity of
heterogeneous catalysts involving both ionic and covalent
pathways.16 However, little is known about this electronic
interaction between an oxide catalyst and oxide support.
Metal cations of the oxide catalyst can bind strongly to an
oxide support at the interface, forming binary oxide
nanostructures.17 When investigating the oxide−oxide inter-
faces at the atomic scale, the metal−O−metal bonding is a
crucial structural unit.18 In heterogeneous catalysis, different
metal−O−metal bonding can provide different chemical
environments, such as different valence, defect, and coordina-
tion statuses for these active centers, leading to unique catalytic
properties.19,20 Therefore, finding two adaptive oxide com-
pounds with catalytically active metal−O−metal bonding is a
potentially effective solution to enhance the catalytic perform-
ance of oxide−oxide systems.
INTRODUCTION
■
CO2 hydrogenation is an approach to produce clean fuels and
valuable chemicals from a gas mixture of CO2 and H2, which
can reduce the use of fossil fuels and control greenhouse gas
emissions. In particular, the methanol synthesis by CO2
hydrogenation is attractive because of its potential as a high-
energy-density fuel.1 Oxide-supported metal catalysts (Cu/
2
ZnO/Al2O3 and Pd/ZnO3) and oxide catalysts (In2O3,4
Ga2O3,5 ZnCr2O4,6 and ZnO−ZrO2 ) have been widely
7
applied in methanol synthesis and extensively studied for
decades. Generally, oxide catalysts are expected to be superior
to metal catalysts due to their higher selectivity to methanol,
durability, and stability, especially in a wide temperature range
(300−600 °C). However, the activation of H2 is difficult, and
CO2 conversion is low for oxide catalysts.8 The major
challenge of improving CO2 conversion for oxide catalysts
lies in understanding the correlation between the complex
oxide structure and methanol synthesis activity.
Improved catalytic performance for oxide-based catalysts can
be achieved by the efforts of (i) promoters and (ii) proper
supports.9−11 Recently, more attention has been focused on
tuning the local environment of an active oxide.12 For
supported metal nanoparticle (NP) catalysts, the metal−
support interaction plays an important role and has been
investigated in numerous processes, for example, the electronic
modulation effect,13 excellent metal dispersion,14 and the
encapsulation process.15 For exothermic chemical reactions,
Received: July 4, 2020
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
© XXXX American Chemical Society
A

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
decompressed to atmospheric pressure, and the reaction products
were analyzed with an online gas chromatograph (GC, Agilent
7890A) equipped with two detectors. CH3OCH3, CH4, and CH3OH
were determined by a flame ionization detector with an HP-FFAP
column using H2 as a carrier gas. A thermal conductivity detector with
columns of MS-5A and Hayesep Q using He as a carrier gas for other
gaseous products (including H2, CO2, N2, and CO) was used. The
carbon balance was found to be >95% in all the tests. CO2 conversion,
C-containing product selectivity, and the yield of products were
calculated as follows
A supported indium oxide (In2O3) catalyst is a promising
model system for methanol synthesis to investigate the surface
structure. Perez-Ramirez et al. claimed that supported In2O3
with abundant oxygen vacancies (VO) shows higher indium-
specific activity.4,21 VO on In2O3 can be modulated and
increased by CO treatment and interaction with a lattice-
mismatching zirconia carrier, which regulates the CO2
adsorption and is proposed as an active site for this reaction.
Muller and co-workers proposed In2O3−x is the active phase,
while the formation of In0 leads to catalyst deactivation.11
Nonetheless, the interface of binary oxides and the electronic
effect of a support during CO2 hydrogenation are still unclear
at the molecular level. Up to now, the application of this
catalyst is still in the bottleneck, due to the lack of knowledge
on both hydrogenation reaction mechanisms and the
fundamental understanding of the relationship between oxide
structure and catalytic activity.
FCO2,in − F
X
CO2(%) =
CO2,out *100
FCO2,in
(1)
%N
∑ ((%N))
SN(%) =
*100
(2)
(3)
yieldN(%) = XCO2*SN(%)*100
In this work, In2O3 supported on zirconium oxide (ZrO2)
with different crystalline phases (both m-: monoclinic and t-:
tetragonal phases) was prepared by Li doping into ZrO2. The
results show that the methanol selectivity and yield are strongly
related to the ZrO2 crystalline phase. The binary oxide, In2O3
supported on an m-ZrO2 catalyst, exhibited maximized
methanol selectivity at high CO2 single-pass conversion,
which achieved a dramatic enhancement in methanol yield in
comparison with In2O3/t-ZrO2. Herein, the experimental
observation shows the electron transfer from m-ZrO2 to
In2O3 due to the oxide−support interaction. The In2O3
structure and key intermediates during CO2 hydrogenation
were identified. In addition, the kinetic experiments confirmed
that the electron transfer between In2O3 and m-ZrO2 could
promote the hydrogenation process and account for the
excellent performance in methanol synthesis.
Where N represents the carbon-containing species in the products,
including CO, CH3OCH3, CH4, and CH3OH. The results were
obtained when the reaction had reached a steady state.
Catalyst Characterization. Powder X-ray diffraction (XRD) of
different samples was recorded before and after reaction. XRD
patterns were performed with 2θ values between 20 and 90° using a
Bruker-D8 diffractometer employing graphite filtered Cu Kα radiation
(λ = 1.54056 Å). The experiment was operated at 40 kV and 40 mA
and employed a Vantec detector. Catalysts were characterized after
being filled in the glassy sample cell. The diameter of ZrO2
nanoparticles was calculated by the Sherrer equation. In the equation,
k stands for the Sherrer constant, λ is the wavelength of an X-ray, B is
the full width at half-maximum (fwhm) of the diffraction peak, and θ
is the angle of diffraction.
kλ
Bcos θ
D =
(4)
The fraction of m-ZrO2 in the ZrO2 support (Im‑Zr/IZr) was
calculated based on the intensity of the characteristic diffraction peaks
of t-ZrO2(101) (It‑Zr, ca. 30.0°) and m-ZrO2(111) (Im‑Zr, ca. 31.5°).
The equation was listed as follows
EXPERIMENTAL SECTION
■
Catalyst Preparation. The monoclinic and tetragonal ZrO2 were
synthesized by a coprecipitation method. First, the hydroxides were
precipitated by a dissolving solution of ZrOCl2·xH2O (Acros
Organics, 99.5%) and LiNO3 (Acros Organics, 99.5%) (molar ratio
of Li/Zr = 5:95) in a mixture of deionized water and ethanol (Sigma-
Aldrich, 99.8%), followed by the addition of NH4OH (Sigma-Aldrich,
25 wt % in H2O) until the pH reached 9. The resulting solids were
dried at 60 °C for 12 h prior to calcination in static air at 350 °C (5
°C·min−1) for 3 h. Other Li-doping ZrO2 catalysts were prepared by
the same procedure with a molar ratio of Li/Zr = 0:100, 20:80, and
40:60, respectively. In2O3 was supported on ZrO2 and Li−ZrO2,
nominally 8 wt % In by isovolumic impregnation. For the isovolumic
impregnation, In(NO3)3·xH2O was first dissolved in deionized water.
The solution was subsequently added drop by drop to the support,
and the resulting slurry was stirred ultrasonically treated for 2 h.
Finally, the impregnated extrudates were dried and calcined at 350 °C
(5 °C·min−1) for 3 h to get the In2O3/ZrO2 and In2O3/Li−ZrO2
samples.
Im‐Zr
Im‐Zr + It‐Zr
I
m‐Zr/IZr =
(5)
The electronic structure of In2O3/ZrO2 was studied by quasi-in-situ
X-ray photoelectron spectroscopy (XPS). The binding energies were
calibrated using the C 1s peak at 284.6 eV as a reference. Specifically,
the catalysts were pretreated in a reaction cell directly connected to
the spectrometer chamber, which allows sample transfer without
exposure to air. The CO2 hydrogenation reaction was performed
under the following conditions: CO2/H2 mixture (ratio = 1:3, flow
rate is 40 mL/min) for 0.5 h at 280 °C.
Diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) was used to record the vibrational spectra of species
adsorbed on the surface of catalysts under CO2 hydrogenation. In situ
DRIFTS experiments were performed on a Thermo Scientific Nicolet
IS50 spectrometer, equipped with a Harrick Scientific DRIFT cell and
a mercury-cadmium-telluride (MCT) detector cooled by liquid N2.
About 100 mg of sample was packed in the in situ chamber. All the
samples were pretreated at 350 °C in 10% H2/Ar flow (30 mL/min)
for 1 h and cooled to the desired temperature to obtain the
background spectrum, and the spectra of samples for each
measurement were then collected by subtracting the background
spectrum. The absorbance was normalized to the mass of the
catalysts. Generally, the in situ reactions were carried out under 0.8
MPa, 280 °C, 40 mL/min of CO2 + H2(D2), and CO2/H2(D2) = 1:3.
The spectra were recorded by collecting 64 scans at a resolution of 8
cm−1. The outlet of the DRIFT cell was connected to an online mass
spectrometer (MS), so the gas component in the effluent can be
monitored and recorded by the MS. Typically, the mass/charge ratio
Catalytic Evaluation. The catalytic activity was measured in a
fixed-bed reactor of a length of 400 mm and a diameter of 8 mm,
made of stainless steel. The bed consisted of 0.2 g of the catalyst (20−
40 mesh size distribution) and 2 g of quartz particles placed between
two layers of silica wool. First, the pressure of the reactor was
increased to 3.0 MPa with N2 flow, and the temperature was increased
to 350 °C. Then, the catalyst was reduced and activated in a stream of
diluted hydrogen (10 vol % H2/N2) at 350 °C for 1 h. Prior to each
experiment, the reactor was cooled down to a specific reaction
temperature (200−350 °C). The catalytic activity in the methanol
synthesis was tested under the following conditions: gas hourly space
velocity (GHSV) of the reactants flow = 12 000 h−1, H2/CO2 = 3, N2
flow rate = 10 mL/min. At the outlet of the reactor, the gases were
B
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(m/z) values are 33 for CH3OD, 34 for CH2DOD, 35 for CD2HOD,
and 36 for CD3OD in hydrogen isotope analysis.
DFT Method. The plane-wave-based Vienna Ab initio Simulation
package (VASP) was applied to perform our density functional theory
calculations.22 The electron exchange and correlation effects were
described by the generalized gradient approximation (GGA) in the
form of the PBE + U functional.23 A plane-wave basis set with a cutoff
energy of 400 eV was applied in all slab calculations, with a force
convergence criterium of 0.02 eV/Å. After convergence tests, k-point
grids of (2 × 2 × 1) for m-ZrO2(111) and (3 × 3 × 1) for t-
ZrO2(101), respectively, were used to sample the calculation space.
For the In−Zr binary model, we used the same k-point sets for m- and
t-ZrO2(111) as above. A 15 Å vacuum space is sufficiently applied in
all models to avoid periodic interaction along the Z-directions. In
addition, the transition states were located by the climbing-image
nudged elastic band method.24
Modeling of In2O3. The interaction between atomic cores and
electrons was described by the projector augmented wave (PAW)
method.25 The perfect In2O3(111) surface was modeled with a (1 ×
1) supercell, built from the optimized In2O3 bulk with lattice
parameters a = b = c = 10.34 Å, which is in good agreement with the
previous experimental results.26 And the surface of In2O3(111)
comprise 24 In atoms and 16 O atoms. To eliminate the periodic
interaction in the Z-direction, a sufficient 15 Å vacuum space was
applied. In all calculations, the bottom two layers were frozen at their
equilibrium bulk positions, whereas the top two layers together with
the adsorbates could relax.
Modeling of ZrO2. The bulk structures of monoclinic and
tetragonal zirconia (m-ZrO2 and t-ZrO2) have been well optimized in
our calculation, with lattice parameters a = 5.19 Å, b = 5.25 Å, c = 5.29
Å for m-ZrO2 and a = b = 3.64 Å and c = 5.27 Å for t-ZrO2, consistent
with the experiments.27 According to the experiment part, we chose
m-ZrO2(111) and t-ZrO2(101) for further calculation, both with a (2
× 2) unit cell and a 15 Å vacuum space. After convergence tests, k-
point grids of (2 × 2 × 1) for m-ZrO2(111) and (3 × 3 × 1) for t-
ZrO2(101), respectively, were used to sample the calculation space.
Similar to In2O3(111), we fixed the bottom two layers and relaxed the
top two layers of the ZrO2 slabs with adsorbates in all the calculation.
Modeling of In−Zr Binary Structures. To elucidate the
electronic interactions between a large In2O3 particle and its support
ZrO2, we created ZrO2-supported In2O3 nanostripe models (32 In
atoms) to estimate the charge transfer between ZrO2 and In2O3. The
same method has been applied in ZnO−ZrO2.10
Figure 1. HRTEM images of In2O3/ZrO2 (A) and In2O3/5Li−ZrO2
(B). Aberration corrected scanning TEM−high-angle annular dark-
field images and element distribution of In2O3/ZrO2 (C) and In2O3/
5Li−ZrO2 (D).
Adjustment of ZrO2 Crystalline Structure via Li
Doping. To figure out the effect of doping on the structure
of catalysts, a series of spectroscopic techniques were carried
out. XRD patterns in Figure S2A demonstrate that the pure
ZrO2 is in tetragonal phase (30.0°, t-ZrO2(101)). When a
small amount of lithium is added (5Li−ZrO2), the crystal form
of ZrO2 is changed completely to monoclinic phase (28.1°, m-
ZrO2(−111) and 31.5°, m-ZrO2(111)). However, higher Li
content (20 to 40%) for Li−ZrO2 tends to transform the ZrO2
crystalline structure from monoclinic back to the tetragonal
phase.28
In situ Raman spectroscopy with different laser sources was
used to detect the structure of the In2O3/Li−ZrO2 catalyst in
different depths due to light absorption and light scattering.29
Generally, the phase in the bulk of catalysts (the depth is >50
nm) is sensitively detected by Raman spectroscopy with laser
sources at 532 nm.30 In Figure 2A, the appearance of Raman
peaks at 187, 334, and 380 cm−1 indicates that the bulk of
In2O3/5Li−ZrO2 is monoclinic ZrO2.31 The additional peak at
269 cm−1 became obvious for In2O3/ZrO2 and In2O3/40Li−
ZrO2, which is characteristic of the tetragonal ZrO2 phase.32
Both of these results are very consistent with the above-
discussed XRD analysis. Actually, the average Zr−O distance
in m-ZrO2 is 2.159 Å, whereas Zr−O distance in t-ZrO2 is
2.226 Å.33 During the doping process, the atomic arrangement
changes significantly, and this results in different Zr−O
distances and the t−m phase transition.
On the other hand, the phase near the skin layer (about 10
nm) is detected by ultraviolet Raman spectroscopy with a 325
nm excitation laser.30 Strong Eg vibrational modes of tetragonal
ZrO2 phases at 269 cm−1 were observed over In2O3/ZrO2. As
for In2O3/5Li−ZrO2, only monoclinic ZrO2 (Ag, 187 cm−1)
was detected. These results above demonstrate the existence of
the monoclinic phase through the particle, i.e., from the
RESULTS AND DISCUSSION
■
Synthesis and Morphology. ZrO2 was chosen as
supports for In2O3 nanoparticles, due to its inactivity for
CO2 hydrogenation and adjustable crystalline structure by
changing the Li-doping amount (xLi-ZrO2, x represents molar
percentage of Li, metal base). In this work, Li-doped ZrO2 was
prepared by ammonia coprecipitation. N2 sorption isotherms
exhibit similar surface areas for both ZrO2 and Li-doped ZrO2
(Table S1). In2O3 was supported on ZrO2 by an impregnation
method. Inductively coupled plasma (ICP) analyses show that
the indium content was ∼8 wt % for the samples (Table S2).
Moreover, the Zr/ In on the surface of the catalysts is about
2:1 as obtained from XPS (Table S2), suggesting the
distribution of In species is mainly concentrated on the
surfaces of catalysts.
High-resolution transmission electron microscopy
(HRTEM) images show In2O3 mostly localized on the surface
of ZrO2 (Figure 1). Most of the In2O3 nanoparticles in the
various samples have comparable average diameters, with
distributions in the range of 8−15 nm (Figures 1A,B and
S1A,B). Element distribution analysis shows that the In species
is dispersed evenly on ZrO2 and 5Li−ZrO2 (Figure 1C,D).
C
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 3. (A) CO2 conversion and (B) product distribution for CO2
hydrogenation over different catalysts. Methanol selectivity: solid bar,
CO selectivity: shaded bar. (C) Methanol yield for CO2 hydro-
genation over different catalysts at various temperatures. (D) The
proportion of In−O−In structure and methanol selectivity for CO2
hydrogenation over In2O3/ZrO2 and In2O3/Li−ZrO2 catalysts as a
function of Im‑Zr/IZr at 280 °C. Reaction conditions: P = 3.0 MPa,
CO2/H2 = 1:3, and GHSV = 12 000 h−1.
Figure 2. (A) In situ Raman spectra of catalysts with a 532 nm laser.
(B) In situ Raman spectra of catalysts with a 325 nm laser. Raman
shift from 250 to 400 cm−1 in Figure 2B for In2O3/ZrO2 (C), In2O3/
5Li−ZrO2 (D), and In2O3/40Li−ZrO2 (E).
particle core to the skin layer of In2O3/5Li−ZrO2. In contrast,
the bulk of In2O3/ZrO2 and In2O3/40Li−ZrO2 is primarily in
the tetragonal phase, whereas the surface is in the coexistence
of tetragonal and monoclinic phases. It is worth noting that the
tetragonal ZrO2 phase dominated in the skin layer of In2O3/
ZrO2. As shown in Figure 2B, the peaks at 305 and 502 cm−1
are assigned to δ(InO6) and v(InO6) of InO6 structure units,34
whereas the peak at 345 cm−1 is assigned to the stretching
vibrations of the In−O−In structures highly dispersed on the
surface of ZrO2.35 A higher proportion of the In−O−In
structure (vs(In−O−In))/δ(InO6) was found on the In2O3/
5Li−ZrO2, suggesting a stronger interaction between In2O3
and m-ZrO2 (Figure 2C−E and Table S2).
The phase transition of ZrO2 caused by Li doping is also
supported by the ultraviolet−visible (UV−vis) absorption
spectra and the Tauc plots. The supports show different
absorption edges at around 250 nm (Figure S1C), which is
associated with the indirect band gap energy (Eg) of Li-doped
ZrO2. According to the Tauc plots, the estimated Eg of 5Li−
ZrO2 is up to 4.96 eV, which is higher than that of ZrO2 (4.78
eV) and 40Li−ZrO2 (4.86 eV) (Figure S1D and Table S3),
suggesting that Li doping adjusted the crystalline form and
band structure of ZrO2.
Catalytic Performance for CO2 Hydrogenation. The
catalytic performance for CO2 hydrogenation was tested at
200−350 °C and 3.0 MPa (Figure 3). As shown in Figure 3B,
the selectivity toward CH3OH decreased along with the
increasing of the reaction temperature. Therefore, a volcano
type of yield has been observed (Figure 3C). More
importantly, a significant improvement on the CH3OH yield
of In2O3/5Li−ZrO2 than that of In2O3/ZrO2 and In2O3/
40Li−ZrO2 was observed at a wide range of temperatures from
220 to 350 °C (Figure 3C). At 280 °C, the reaction
temperature with the highest CH3OH yield, In2O3/5Li−
ZrO2 produced CH3OH with 84.6% selectivity at 12.1% of
CO2 conversion (Figure S3A), which is much higher than that
obtained over In2O3/ZrO2 (61.6% of CH3OH selectivity at
6.0% conversion). As Li was further added into ZrO2 support
(20−40%), both CH3OH selectivity and CO2 conversion
decreased.
The XRD characterization after reaction shows that t-ZrO2
still dominates in used In2O3/ZrO2 , and m-ZrO2 dominates in
used In2O3/5Li−ZrO2 (Figure S2B). The fraction of m-ZrO2
in the ZrO2 support was obtained by normalizing the
diffraction peak intensity of m-ZrO2 and t-ZrO2 (Im‑Zr/IZr, eq
5 in Experimental Section). It is interesting to note that the
CH3OH selectivity exhibits a positive correlation with Im‑Zr/IZr
(Figure 3D). Thus, we propose the m-ZrO2 support plays an
important role in promoting methanol production. In addition,
In2O3/ZrO2 and In2O3/5Li−ZrO2 show different apparent
activation energies (Ea) of methanol, suggesting that there are
different active sites over In2O3/m-ZrO2 and In2O3/t-ZrO2
(Table S4 and Figure S3B,C). Based on the ultraviolet Raman
spectrum in Figure 2B, the proportion of the In−O−In
structure bonded to m-ZrO2 is correlated with the methanol
selectivity. That is, the In−O−In structure shows higher
reactivity. Therefore, the In−O−In structure could potentially
be the major active site for methanol synthesis (Figure 3D).
To eliminate the influence of Li serving as an electronic
additive, an In2O3 + 5Li/ZrO2 catalyst was prepared via co-
impregnation of In and a Li precursor, which exhibited lower
CO2 conversion (7.2%) and methanol selectivity (66.3%) than
In2O3/5Li−ZrO2 (Figure S3D,E). Therefore, the promotion in
catalytic performance for In2O3/5Li−ZrO2 should be
attributed to the existence of the monoclinic phase caused
by 5Li doping into ZrO2 as opposed to the Li additive.
Electronic Interaction between In2O3 and m-ZrO2. To
find out the origin of the high methanol yield for In2O3/5Li−
ZrO2, quasi-in-situ XPS was carried out to characterize the
surface structure of the catalyst during CO2 hydrogenation. As
shown in Figure S4A, In2O3/ZrO2 and In2O3/Li−ZrO2
catalysts showed peaks at ∼444.5−445.0 eV, corresponding
to In2O3.36 After the reaction, In2O3/5Li−ZrO2 showed a
lower binding energy of In 3d5/2 (444.4 eV) as opposed to
D
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. (A) Methanol yield and binding energy (BE) shift of In 3d5/2 (yellow) and Zr 3d5/2 (purple) among catalysts in the initial stage of the
reaction as a function of Im‑Zr/IZr. The BE of In 3d5/2 and Zr 3d5/2 over In2O3/ZrO2 was chosen as the reference. (B) ESR spectra of used catalysts.
(C) DFT simulation of charge transfer in the In2O3 stripe/m-ZrO2 and In2O3 stripe/t-ZrO2 catalyst models. In: light brown, Zr: mint green, O
atoms in ZrO2: red, and O atoms in In4O6: blue. The black numbers in the image are the total number of valence electrons on In atoms. (D) The
binding energy of In 3d5/2 plotted as a function of time on stream for In2O3/ZrO2 and In2O3/Li−ZrO2 catalysts. (E) The VO concentration plotted
as a function of time on stream for In2O3/ZrO2 and In2O3/Li−ZrO2 catalysts. (F) The methanol selectivity in the initial stage of the reaction and
binding energy of In 3d5/2 plotted as a function of VO concentration for In2O3/ZrO2 and In2O3/Li−ZrO2 catalysts.
other samples (Figure S4A). On the contrary, the binding
energy of Zr 3d5/2 in In2O3/5Li−ZrO2 (182.0 eV) is higher
than that in In2O3/ZrO2 (181.7 eV) (Figure S4B).37 This
phenomenon indicates In2O3 supported on 5Li−ZrO2 has a
higher electron density than that in In2O3/ZrO2 (Figure S4A−
C and Table S5). Also, there was no binding energy shift of Zr
3d5/2 observed for pure ZrO2 and Li−ZrO2 support after the
same treatment (Figure S4C), revealing that the addition of Li
has no influence on the electron density of Zr.
than the real reaction pressure. The low pressure may also
influence intermediate coverage and cause a different catalyst
surface from the actual situation. In contrast, the reaction for
quasi-in-situ XPS can be pressurized to higher pressure (0.1
MPa). Nevertheless, the combination of quasi-in-situ and AP-
XPS characterization suggests strong electron transfer electron
transfer from m-ZrO2 to In2O3 and the influence on the
reaction performance. As shown in Figure S6B−D, the
catalysts after a long-term reaction displayed similar trends of
binding energy shifts.
Figure 4A shows the correlation between the binding energy
of used catalysts (reference: used In2O3/ZrO2) and methanol
yield as a function of Im‑Zr/IZr. The binding energy of In
became lower, and the binding energy of Zr shifted to a higher
value with Im‑Zr/IZr increased, suggesting that stronger electron
transfer from m-ZrO2 to In2O3 leads to more electrons
accumulated on the In atom of In2O3/m-ZrO2 than that of
In2O3/t-ZrO2.38 XRD results show that after In2O3 was
supported onto different ZrO2, both the diffraction peaks of
t-ZrO2 and m-ZrO2 are shifted to a higher position (Figure
S5), meaning that a fraction of In is incorporated into the ZrO2
matrix by the form of In−O−Zr bonds.39 The In−O−In
structure may be bonded to m-ZrO2 via In−O−Zr bonds,
which changed the coordination of indium and enhanced the
electron transfer. As a result, the binding energy shift became
more obvious. Moreover, the methanol yield increased with
the strengthening of this electronic interaction (Figure 4A),
indicating electron-rich In2O3 favors the methanol formation.
Ambient pressure XPS (AP-XPS) experiments were also
carried out to eliminate possible error caused by measurement
under vacuum for quasi-in-situ XPS. Similarly, In2O3 on m-
ZrO2 shows a lower BE of In, i.e., higher electron density than
In2O3 on t-ZrO2 by AP-XPS (Figure S6A). It is worth noting
that the pressure of AP-XPS is 0.4 mbar, which is much smaller
Additionally, we applied Bader charge analysis based on
density functional theory (DFT) calculations to support the
electron transfer between the two oxides (Figure S7). The
model of In doped into ZrO2 show that In in m-ZrO2 can
enrich more electrons (0.04 e−) than In in t-ZrO2 (Figure S8).
Figure 4C shows that the In2O3 stripe model tends to accept
more electrons, ∼0.86 e−, on m-ZrO2 than that on t-ZrO2,
indicating In2O3 supported on m-ZrO2 should have greater
electron enrichment, which is in agreement with the XPS
results.
The types of electron transfer in In2O3/ZrO2 were
characterized by electron spin resonance (ESR). The ESR in
Figure S4D suggests that the trapped electrons are absent on
pure ZrO2 supports, as no obvious peaks and troughs were
observed. After the loading of In2O3, the electron spin count at
g = 1.98−2.00 showed an obvious change for all samples
(Figure 4B). Thus, the trapped electrons were generated over
In2O3 or at the In−O−Zr interface, as opposed to over ZrO2.
Figure 4B reveals two intense signals for In2O3/5Li−ZrO2 at g
= 2.00 and g = 1.98, whereas only the former signal is detected
for In2O3/ZrO2 and In2O3/40Li−ZrO2. The signal at g = 2.00
has commonly been assigned to the delocalized electrons that
result from the abstraction of O atoms, probably oxygen
E
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
vacancies,40 and the signal at g = 1.98 represents unpaired
electrons,41,42 which may migrate from the conduction band of
m-ZrO2 to In2O3. It is therefore evident that the electronic
property of the In sites is modified by the neighboring
monoclinic Zr sites, when electrons are preferentially deposited
over In sites from m-ZrO2.
In the reducing reaction atmosphere, oxygen vacancies could
inevitably be formed on the In2O3 surface (Figure S9A-C). It is
widely discovered that the VO delocalization effect can reduce
the binding energy of the corresponding oxide.43 However, in
this study, the binding energy of In 3d5/2 on all samples
continued to decrease (Figure 4D) while the VO concentration
decreased (Figure 4E and Table S5) during the reaction. This
indicates that the electron transfer effect on In electron density
at the interface between m-ZrO2 and In2O3 is stronger than the
VO delocalization effect over the In2O3 surface. It is also found
that the VO concentration on the catalyst is not positively
correlated with the methanol selectivity (Figure 4F). In other
words, the electron transfer at the interface between m-ZrO2
and In2O3 plays a more significant role in the decrease of the
binding energy of In sites and promoting methanol synthesis
than VO.
Above all, the experimental evidence by XPS and ESR,
combined with theoretical calculations, proved the existence of
electron transfer from m-ZrO2 to In2O3 for In2O3/5Li−ZrO2,
which plays crucial roles in promoting catalytic CO2 to
methanol.
Figure 5. (A) XPS C 1s peaks of In2O3/ZrO2 and In2O3/5Li−ZrO2
before and after reaction. (B) In situ DRIFT spectra of different
catalysts at 280 °C. (C) Absorbance of generated intermediate species
over different catalysts after inletting CO2 + H2 at 280 °C. (D)
Absorbance of generated intermediate species over In2O3/5Li−ZrO2
after switching from CO2 + H2 to CO2 + D2 at 280 °C. (E) Simplified
models showing the main reaction mechanism on In2O3/t-ZrO2 and
In2O3/m-ZrO2.
Mechanism Investigation. To further understand the
effect of charge transfer between ZrO2 and In2O3 on catalytic
performance, the reaction mechanism was investigated.
Previous works have proposed that there are mainly two
reaction pathways to produce CH3OH via CO2 hydro-
genation.44 One is formed by the *CO intermediate, which
is produced from the reverse water−gas shift (RWGS, CO2 +
H2 → CO + H2O) reaction, and *CO is further hydrogenated
to CH3OH; the other pathway is designated as the formate
(HCOO*) pathway, where the HCOO* intermediate is
formed by *CO2 hydrogenation and further hydrogenated to
the *H2COOH intermediate, which eventually produces
CH2O* and CH3OH after breaking one C−O bond in
*H2COOH. The possible reaction pathway may be associated
with different catalyst systems, which show different adsorption
properties for reaction intermediates. The identification of the
active intermediates is essential to understand the preferred
pathways and the role of electron transfer in CO2 hydro-
genation.
cm−1) confirmed the presence of HCOO* species.45,47 As
shown in Figures 5C and S10A−C, HCOO* species appeared
at the beginning of the reaction owing to the adsorption and
hydrogenation of CO2. With the pulse of H2 flow, new bands
at 2830 and 1025 cm−1 were observed, which are assigned to
the stretching vibration of saturated C−H and C−O in
CH3O* species.10 Noticeably, the stronger IR peak intensity
(1025 cm−1) over In2O3/5Li−ZrO2 indicated more CH3O*
species were formed over In2O3/5Li−ZrO2 than other
samples, which should account for the high methanol yield.
Hence, HCOO* and H3CO* are key intermediates during
CO2 hydrogenation.
The used In2O3/ZrO2 and In2O3/5Li−ZrO2 were analyzed
by XPS C 1s spectra to identify reaction intermediates (Figure
5A). The peak at approximately 284.6 eV indicated the
presence of carbon on both surfaces. Upon reacting at 280 °C,
peaks at approximately 288.8 and 286.2 eV were observed for
In2O3/5Li−ZrO2 but not for In2O3/ZrO2. These peaks are
attributed to the surface formate (HCOO*)/carboxyl
(*COOH) species and methoxy (CH3O*) species respec-
tively, suggesting that abundant COO* or CH3O* species
were formed on In2O3/5Li−ZrO2 during CO2 hydrogena-
tion.45,46
To confirm the reaction route, DRIFTS was carried out
under CO2 flow with pulsed hydrogen, where the hydrogen
was insufficient to slow down the transformation and monitor
the change of intermediates clearly (Figure 5B). The
unsaturated C−H species (2876 cm−1) along with the
symmetric and asymmetric OCO stretches (1368 and 1574
For all the catalysts in Figure 5C, HCOO* species were
observed prior to the formation of CH3O* species. The IR
peak intensity of surface HCOO* became stable after a
reaction for 30 min, whereas it took more than 60 min for
H3CO* to reach steady state. When CO2 + H2 was substituted
for CO2 + D2 over In2O3/5Li−ZrO2 (Figure S10D), HCOO*
and CH3O* (2876 and 2830 cm−1, respectively) disappeared
gradually (Figure 5D). Meanwhile, DCOO* species (2176
cm−1) appeared before the CD3O* peak (2067 cm−1) arose.7
Both the sequences of generated intermediate species in in situ
DRIFTS and isotope labeling experiments suggest that CH3O*
species come from HCOO* hydrogenation, and the formate
route is the preferred pathway on the In2O3-based catalyst.
The apparent reaction orders of the reactants were
determined by altering the partial pressures of CO2 and
H2.48 As shown in Figure S11A, the formation rate of methanol
increased significantly with the increase of the partial pressure
F
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
of H2, while the methanol formation was suppressed in excess
of CO2. When the partial pressure of H2 is above the ratio of
H2/CO2 = 1:1, the orders of H2 to methanol were calculated at
1.2, 1.6, and 1.4 for In2O3/ZrO2, In2O3/5Li−ZrO2, and In2O3/
40Li−ZrO2, respectively (Figures S11B and S12A). However,
the orders of CO2 to methanol were about −0.5 for three
samples (Figures S11C and S12B). Consequently, the
activation and dissociation of H2 may be the limiting process
for methanol synthesis on In2O3-based catalysts rather than the
activation of CO2. Furthermore, the reaction order of H2 to
methanol is higher than that to CO (ca. 0.3, Figure S12C),
which indicates that the promotion of H2 activation has a more
positive effect on the formation of methanol than CO.
where this process over In2O3/5Li−ZrO2 (10 min) was much
faster than that over In2O3/ZrO2 (80 min) and In2O3/40Li−
ZrO2 (60 min). Since CH2DOD was produced via formate
reacted with dissociated D, this result affirmed that there is a
strong promotion effect on the H2 activation, facilitating
subsequent hydrogenation of intermediates over the electron-
rich In2O3 of In2O3/m-ZrO2.
Furthermore, HCOOH was chosen as reactant, which can
provide COO* intermediates, to verify the COO* hydro-
genation step. Figures 6E and S15A,B show the varying
tendency of the preadsorbed COO* with time after H2
inletting into reaction cell. The IR intensity of adsorbed
COO* (1589 and 1450 cm−1) over In2O3/5Li−ZrO2 dropped
rapidly under H2, accompanied by the formation of CH3O*
species at 1025 cm−1. However, the signal of adsorbed COO*
over In2O3/ZrO2 remains barely changed in intensity.
Additionally, there was no CH3O* formed, indicating the
COO* species were difficult to be converted to CH3O* over
In2O3/ZrO2. For the formate pathway, the binding strength of
HCOO* should be moderate, and CH3O* should be
destabilized on the surface to allow the production of
CH3OH. Based on the DRIFTS results, the hydrogenation of
HCOO* to CH3O* is easy for In2O3/5Li−ZrO2, but HCOO*
is overstabilized on In2O3/ZrO2, poisoning the surface and
thus blocking the formation of CH3O*. Overall, the electronic
interaction over In2O3/m-ZrO2 (In2O3/5Li−ZrO2) leads to a
higher methanol yield, which is due to the stronger H2-
dissociation activity to convert the formate intermediate to
CH3O* and thus methanol (Figure 5E).
H2-temperature-programmed reduction (H2-TPR) and D2-
temperature-programmed surface reaction (D2-TPSR) showed
that H2 dissociation and HD formation happened at lower
temperatures over In2O3/5Li−ZrO2 than In2O3/ZrO2, In2O3 +
5Li/ZrO2, and In2O3/40Li−ZrO2 (Figures S13A,B and 6A).
CONCLUSION
■
In summary, we describe a strategy to support In2O3 over ZrO2
with different crystallization phases. The best catalytic
performance was achieved over In2O3/m-ZrO2, which exhibits
high activity and selectivity in the selective CO2 hydrogenation
to methanol. Compared with In2O3/t-ZrO2, the improved
catalytic performance is potentially due to the strong
interaction between In2O3 and monoclinic ZrO2, where
In2O3 is dispersed mainly in the form of In-O-Inand electrons
were transferred from monoclinic ZrO2 to In2O3, generating
the electron-rich In2O3. The highly dispersed In−O−In
structure is proposed as a major active site to convert CO2
to methanol. Also, the electron transfer from m-ZrO2 to In−
O−In plays a more significant role in methanol synthesis than
VO. It is confirmed experimentally that the production of
methanol from CO2 hydrogenation over an In2O3-based
catalyst is via formate route. Accordingly, electron-rich In2O3
was proven to promote the H2 dissociation and help HCOO*
convert into CH3O* by hydrogenation. This work extends the
application of oxide catalysts in CO2 hydrogenation to
methanol and also reveals the crucial effect of the electronic
interaction between the oxide and support in the hydro-
genation reaction, which opens an avenue for developing
multicomponent oxide catalysts for CO2 conversion.
Figure 6. (A) HD signals of different catalysts in D2-TPSR. DRIFT-
MS of CO2 + H2 subsequently switched to CO2 + D2 at 100 min over
In2O3/ZrO2 (B), In2O3/5Li−ZrO2 (C), and In2O3/40Li−ZrO2 (D).
(E) Absorbance of generated intermediate species over In2O3/ZrO2
and In2O3/5Li−ZrO2 after HCOOH preadsorption and subsequent
H2 sweeping at 280 °C.
Moreover, In2O3/5Li−ZrO2 presented a higher conversion of
H2 to HD in the H2−D2 exchange reaction (Figure S13C) than
other samples. This conclusion is also supported by the
calculation in Figure S14, since H2 dissociation on In2O3/m-
ZrO2 (0.37 eV) shows a lower energy barrier than that on
In2O3/t-ZrO2 (0.59 eV). Both tests indicated that electron-rich
In2O3 of In2O3/m-ZrO2 exhibits higher activity in the H−H
bond dissociation and increases the amount of dissociated H
species. Since the orders of H2 are positive, the increased H
species can accelerate the rate of methanol formation.
To investigate the effect of this electron transfer on the
hydrogenation process of intermediates, the deuterium-
containing products of isotope labeling experiments were
analyzed by mass spectrometry connected with DRIFTS. As
shown in Figure 6B−D, the time for CH2DOD stabilization
after switching CO2 + H2 to CO2 + D2 reflects the H2-
dissociation ability of catalysts under reaction conditions,
ASSOCIATED CONTENT
* Supporting Information
■
sı
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.0c07195.
Further experimental details, catalyst performance data,
quasi-in-situ XPS, ESR, kinetics of hydrogenation, and in
situ DRIFTS characterization data (PDF)
G
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
for financial support. We highly appreciate Dr. Yi Cui,
Zhongmiao Gong, and Hao Li (Suzhou Institute of Nano-
Tech and Nano-Bionics, Chinese Academy of Sciences) for the
support of AP-XPS.
AUTHOR INFORMATION
Corresponding Author
■
Jinlong Gong − Key Laboratory for Green Chemical
Technology of Ministry of Education, School of Chemical
Engineering and Technology; Collaborative Innovation
Center of Chemical Science and Engineering, Tianjin
University, Tianjin 300072, China; Joint School of National
University of Singapore and Tianjin University, International
Campus of Tianjin University, Fuzhou 350207, China;
orcid.org/0000-0001-7263-318X; Email: jlgong@
tju.edu.cn
REFERENCES
■
(1) Wang, W.; Wang, S.; Ma, X.; Gong, J. Recent Advances in
Catalytic Hydrogenation of Carbon Dioxide. Chem. Soc. Rev. 2011, 40
(7), 3703−3727.
(2) Behrens, M.; Studt, F.; Kasatkin, I.; Kuhl, S.; Havecker, M.;
Abild-Pedersen, F.; Zander, S.; Girgsdies, F.; Kurr, P.; Kniep, B. L.;
Tovar, M.; Fischer, R. W.; Norskov, J. K.; Schlogl, R. The Active Site
of Methanol Synthesis over Cu/ZnO/Al2O3 Industrial Catalysts.
Science 2012, 336 (6083), 893−897.
(3) Bahruji, H.; Bowker, M.; Hutchings, G.; Dimitratos, N.; Wells,
P.; Gibson, E.; Jones, W.; Brookes, C.; Morgan, D.; Lalev, G. Pd/ZnO
Catalysts for Direct CO2 Hydrogenation to Methanol. J. Catal. 2016,
343, 133−146.
(4) Martin, O.; Martin, A. J.; Mondelli, C.; Mitchell, S.; Segawa, T.
F.; Hauert, R.; Drouilly, C.; Curulla-Ferre, D.; Perez-Ramirez, J.
Indium Oxide as a Superior Catalyst for Methanol Synthesis by CO2
Hydrogenation. Angew. Chem., Int. Ed. 2016, 55 (21), 6261−6265.
(5) Choi, H.; Oh, S.; Trung Tran, S. B.; Park, J. Y. Size-Controlled
Model Ni Catalysts on Ga2O3 for CO2 Hydrogenation to Methanol. J.
Catal. 2019, 376, 68−76.
(6) Song, H. Q.; Laudenschleger, D.; Carey, J. J.; Ruland, H.; Nolan,
M.; Muhler, M. Spinel-Structured ZnCr2O4 with Excess Zn Is the
Active ZnO/Cr2O3 Catalyst for High-Temperature Methanol Syn-
thesis. ACS Catal. 2017, 7 (11), 7610−7622.
(7) Wang, J.; Li, G.; Li, Z.; Tang, C.; Feng, Z.; An, H.; Liu, H.; Liu,
T.; Li, C. A Highly Selective and Stable ZnO-ZrO2 Solid Solution
Catalyst for CO2 Hydrogenation to Methanol. Sci. Adv. 2017, 3 (10),
e1701290.
(8) Alvarez, A.; Bansode, A.; Urakawa, A.; Bavykina, A. V.;
Wezendonk, T. A.; Makkee, M.; Gascon, J.; Kapteijn, F. Challenges
in the Greener Production of Formates/Formic Acid, Methanol, and
DME by Heterogeneously Catalyzed CO2 Hydrogenation Processes.
Chem. Rev. 2017, 117 (14), 9804−9838.
(9) Chen, T. Y.; Cao, C. X.; Chen, T. B.; Ding, X. X.; Huang, H.;
Shen, L.; Cao, X. Y.; Zhu, M. H.; Xu, J.; Gao, J.; Han, Y. F. Unraveling
Highly Tunable Selectivity in CO2 Hydrogenation over Bimetallic In-
Zr Oxide Catalysts. ACS Catal. 2019, 9 (9), 8785−8797.
(10) Wang, Y.; Kattel, S.; Gao, W.; Li, K.; Liu, P.; Chen, J. G.; Wang,
H. Exploring the Ternary Interactions in Cu-ZnO-ZrO2 Catalysts for
Efficient CO2 Hydrogenation to Methanol. Nat. Commun. 2019, 10
(1), 1166.
(11) Tsoukalou, A.; Abdala, P. M.; Stoian, D.; Huang, X.; Willinger,
M. G.; Fedorov, A.; Muller, C. R. Structural Evolution and Dynamics
of an In2O3 Catalyst for CO2 Hydrogenation to Methanol: An
Operando XAS-XRD and In Situ TEM Study. J. Am. Chem. Soc. 2019,
141 (34), 13497−13505.
Authors
Chengsheng Yang − Key Laboratory for Green Chemical
Technology of Ministry of Education, School of Chemical
Engineering and Technology; Collaborative Innovation
Center of Chemical Science and Engineering, Tianjin
University, Tianjin 300072, China
Chunlei Pei − Key Laboratory for Green Chemical Technology
of Ministry of Education, School of Chemical Engineering and
Technology; Collaborative Innovation Center of Chemical
Science and Engineering, Tianjin University, Tianjin 300072,
China
Ran Luo − Key Laboratory for Green Chemical Technology of
Ministry of Education, School of Chemical Engineering and
Technology; Collaborative Innovation Center of Chemical
Science and Engineering, Tianjin University, Tianjin 300072,
China
Sihang Liu − Key Laboratory for Green Chemical Technology
of Ministry of Education, School of Chemical Engineering and
Technology; Collaborative Innovation Center of Chemical
Science and Engineering, Tianjin University, Tianjin 300072,
China
Yanan Wang − Key Laboratory for Green Chemical
Technology of Ministry of Education, School of Chemical
Engineering and Technology; Collaborative Innovation
Center of Chemical Science and Engineering, Tianjin
University, Tianjin 300072, China
Zhongyan Wang − Key Laboratory for Green Chemical
Technology of Ministry of Education, School of Chemical
Engineering and Technology; Collaborative Innovation
Center of Chemical Science and Engineering, Tianjin
University, Tianjin 300072, China
Zhi-Jian Zhao − Key Laboratory for Green Chemical
Technology of Ministry of Education, School of Chemical
Engineering and Technology; Collaborative Innovation
Center of Chemical Science and Engineering, Tianjin
University, Tianjin 300072, China; orcid.org/0000-
0002-8856-5078
(12) Jia, J.; Qian, C.; Dong, Y.; Li, Y. F.; Wang, H.; Ghoussoub, M.;
Butler, K. T.; Walsh, A.; Ozin, G. A. Heterogeneous Catalytic
Hydrogenation of CO2 by Metal Oxides: Defect Engineering-
Perfecting Imperfection. Chem. Soc. Rev. 2017, 46 (15), 4631−4644.
(13) Liao, F.; Zeng, Z.; Eley, C.; Lu, Q.; Hong, X.; Tsang, S. C.
Electronic Modulation of A Copper/zinc oxide Catalyst by A
Heterojunction for Selective Hydrogenation of Carbon Dioxide to
Methanol. Angew. Chem., Int. Ed. 2012, 51 (24), 5832−5836.
(14) Yang, B.; Liu, C.; Halder, A.; Tyo, E. C.; Martinson, A. B. F.;
Seifert, S.; Zapol, P.; Curtiss, L. A.; Vajda, S. Copper Cluster Size
Effect in Methanol Synthesis from CO2. J. Phys. Chem. C 2017, 121
(19), 10406−10412.
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.0c07195
Author Contributions
€C.Y. and C.P. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the National Key R&D Program of China
(2016YFB0600901), the National Natural Science Foundation
of China (21525626, 21761132023), and the Program of
Introducing Talents of Discipline to Universities (BP0618007)
̈
(15) Lunkenbein, T.; Schumann, J.; Behrens, M.; Schlogl, R.;
Willinger, M. G. Formation of a ZnO Overlayer in Industrial Cu/
ZnO/Al2O3 Catalysts Induced by Strong Metal-Support Interactions.
Angew. Chem. 2015, 127 (15), 4627−4631.
H
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(16) Park, J. Y.; Baker, L. R.; Somorjai, G. A. Role of Hot Electrons
and Metal-oxide Interfaces in Surface Chemistry and Catalytic
Reactions. Chem. Rev. 2015, 115 (8), 2781−2817.
(37) Sotiropoulou, D.; Ladas, S. An XPS and XAES Study of the Ni/
ZrO2 Interface. Surf. Sci. 1998, 408 (1−3), 182−189.
(38) Mullins, W. M.; Averbach, B. L. Bias-Reference X-Ray
Photoelectron Spectroscopy of Sapphire and Yttrium Aluminum
Garnet Crystals. Surf. Sci. 1988, 206 (1−2), 29−40.
(17) Vannice, M.; Garten, R. Metal-support Effects on the Activity
and Selectivity of Ni Catalysts in CO/H2 Synthesis Reactions. J. Catal.
1979, 56 (2), 236−248.
(39) Behrens, M.; Zander, S.; Kurr, P.; Jacobsen, N.; Senker, J.;
Koch, G.; Ressler, T.; Fischer, R. W.; Schlogl, R. Performance
Improvement of Nanocatalysts by Promoter-Induced Defects in the
Support Material: Methanol Synthesis over Cu/ZnO:Al. J. Am. Chem.
Soc. 2013, 135 (16), 6061−6068.
(18) Lin, L.; Yao, S.; Rui, N.; Han, L.; Zhang, F.; Gerlak, C. A.; Liu,
Z.; Cen, J.; Song, L.; Senanayake, S. D.; Xin, H. L.; Chen, J. G.;
Rodriguez, J. A. Conversion of CO2 on A Highly Active and Stable
Cu/FeOx/CeO2 Catalyst: Tuning Catalytic Performance by Oxide-
Oxide Interactions. Catal. Sci. Technol. 2019, 9 (14), 3735−3742.
(19) Liu, Z. W.; Zhang, H. J.; Chen, Z. Q. Monolayer Dispersion of
CoO on Al2O3 Probed by Positronium Atom. Appl. Surf. Sci. 2014,
293, 326−331.
(40) Liao, F.; Huang, Y.; Ge, J.; Zheng, W.; Tedsree, K.; Collier, P.;
Hong, X.; Tsang, S. C. Morphology-Dependent Interactions of ZnO
with Cu Nanoparticles at the Materials’ Interface in Selective
Hydrogenation of CO2 to CH3OH. Angew. Chem., Int. Ed. 2011, 50
(9), 2162−2165.
(20) Nagai, Y.; Hirabayashi, T.; Dohmae, K.; Takagi, N.; Minami,
T.; Shinjoh, H.; Matsumoto, S. Sintering Inhibition Mechanism of
Platinum Supported on Ceria-Based Oxide and Pt-Oxide-Support
interaction. J. Catal. 2006, 242 (1), 103−109.
́
(41) Block, D.; Herve, A.; Cox, R. T. Optically Detected Magnetic
Resonance and Optically Detected ENDOR of Shallow Indium
Donors in ZnO. Phys. Rev. B: Condens. Matter Mater. Phys. 1982, 25
(9), 6049−6052.
(21) Frei, M. S.; Mondelli, C.; Cesarini, A.; Krumeich, F.; Hauert,
(42) Ischenko, V.; Polarz, S.; Grote, D.; Stavarache, V.; Fink, K.;
Driess, M. Zinc Oxide Nanoparticles with Defects. Adv. Funct. Mater.
2005, 15 (12), 1945−1954.
́
R.; Stewart, J. A.; Curulla Ferre, D.; Perez-Ramírez, J. Role of Zirconia
Carrier on Indium Oxide-Catalyzed CO2 Hydrogenation to Methanol.
ACS Catal. 2020, 10 (2), 1133−1145.
(43) Kowalski, P. M.; Camellone, M. F.; Nair, N. N.; Meyer, B.;
Marx, D. Charge Localization Dynamics Induced by Oxygen
Vacancies on the TiO2(110) Surface. Phys. Rev. Lett. 2010, 105
(14), 146405.
(22) Kresse, G.; Hafner, J. Ab Initio Molecular Dynamics for Liquid
Metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47 (1), 558−
561.
(23) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient
Approximation Made Simple. Phys. Rev. Lett. 1996, 77 (18), 3865−
3868.
(44) Kattel, S.; Liu, P.; Chen, J. G. Tuning Selectivity of CO2
Hydrogenation Reactions at the Metal/Oxide Interface. J. Am. Chem.
Soc. 2017, 139 (29), 9739−9754.
́
(24) Henkelman, G.; Uberuaga, B. P.; Jonsson, H. A Climbing
(45) Kattel, S.; Yu, W.; Yang, X.; Yan, B.; Huang, Y.; Wan, W.; Liu,
P.; Chen, J. G. CO2 Hydrogenation over Oxide-Supported PtCo
Catalysts: The Role of the Oxide Support in Determining the Product
Selectivity. Angew. Chem., Int. Ed. 2016, 55 (28), 7968−73.
(46) Graciani, J.; Mudiyanselage, K.; Xu, F.; Baber, A. E.; Evans, J.;
Senanayake, S. D.; Stacchiola, D. J.; Liu, P.; Hrbek, J.; Sanz, J. F.;
Rodriguez, J. A. Highly Active Copper-Ceria and Copper-Ceria-
Titania Catalysts for Methanol Synthesis from CO2. Science 2014, 345
(6196), 546−550.
Image Nudged Elastic Band Method for Finding Saddle Points and
Minimum Energy Paths. J. Chem. Phys. 2000, 113 (22), 9901−9904.
(25) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the
Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter
Mater. Phys. 1999, 59 (3), 1758−1775.
(26) Marezio, M. Refinement of the Crystal Structure of In2O3 at
Two Wavelengths. Acta Crystallogr. 1966, 20 (6), 723−728.
(27) Teufer, G. The Crystal Structure of Tetragonal ZrO2. Acta
Crystallogr. 1962, 15 (11), 1187−1187.
(28) Keramidas, V. G.; White, W. B. Raman Scattering Study of the
Crystallization and Phase Transformations of ZrO2. J. Am. Ceram. Soc.
1974, 57 (1), 22−24.
(29) Shi, L.; Tin, K. C.; Wong, N. B. Thermal Stability of Zirconia
Membranes. J. Mater. Sci. 1999, 34 (14), 3367−3374.
(30) Kim, H.; Kosuda, K. M.; Van Duyne, R. P.; Stair, P. C.
Resonance Raman and Surface- and Tip-Enhanced Raman Spectros-
copy Methods to Study Solid Catalysts and Heterogeneous Catalytic
Reactions. Chem. Soc. Rev. 2010, 39 (12), 4820−4844.
(31) Li, M. J.; Feng, Z. C.; Ying, P. L.; Xin, Q.; Li, C. Phase
Transformation in the Surface Region of Zirconia and Doped Zirconia
Detected by UV Raman Spectroscopy. Phys. Chem. Chem. Phys. 2003,
5 (23), 5326−5332.
(47) Heyl, D.; Rodemerck, U.; Bentrup, U. Mechanistic Study of
Low-Temperature CO2 Hydrogenation over Modified Rh/Al2O3
Catalysts. ACS Catal. 2016, 6 (9), 6275−6284.
(48) Frei, M. S.; Capdevila-Cortada, M.; García-Muelas, R.;
́
́
́
Mondelli, C.; Lopez, N.; Stewart, J. A.; Curulla Ferre, D.; Perez-
Ramírez, J. Mechanism and Microkinetics of Methanol Synthesis via
CO2 Hydrogenation on Indium Oxide. J. Catal. 2018, 361, 313−321.
(32) Kim, B.-K.; Hahn, J.-W.; Han, K. R. Quantitative Phase Analysis
in Tetragonal-Rich Tetragonal/Monoclinic Two Phase Zirconia by
Raman Spectroscopy. J. Mater. Sci. Lett. 1997, 16 (8), 669−671.
(33) Levin, I.; Cockayne, E.; Lufaso, M. W.; Woicik, J. C.; Maslar, J.
E. Local Structures and Raman Spectra in the Ca(Zr,Ti)O3 Perovskite
Solid Solutions. Chem. Mater. 2006, 18 (3), 854−860.
(34) Kaur, M.; Jain, N.; Sharma, K.; Bhattacharya, S.; Roy, M.;
Tyagi, A. K.; Gupta, S. K.; Yakhmi, J. V. Room-Temperature H2S Gas
Sensing at Ppb Level by Single Crystal In2O3 Whiskers. Sens.
Actuators, B 2008, 133 (2), 456−461.
(35) Baszczuk, A.; Jasiorski, M.; Nyk, M.; Hanuza, J.; Maczka, M.;
Strek, W. Luminescence Properties of Europium Activated SrIn2O4. J.
Alloys Compd. 2005, 394 (1−2), 88−92.
(36) Poznyak, S. K.; Golubev, A. N.; Kulak, A. I. Correlation
between Surface Properties and Photocatalytic and Photoelectro-
chemical Activity of In2O3 Nanocrystalline Films and Powders. Surf.
Sci. 2000, 454, 396−401.
I
https://dx.doi.org/10.1021/jacs.0c07195
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX