The Interplay between Structure and Product Selectivity of CO2 Hydrogenation

Article abstract of DOI:10.1002/anie.201904649

Identification of the active structure under reaction conditions is of great importance for the rational design of heterogeneous catalysts. However, this is often hampered by their structural complexity. The interplay between the surface structure of Co₃O₄ and the CO₂ hydrogenation is described. Co₃O₄ with morphology-dependent crystallographic surfaces presents different reducibility and formation energy of oxygen vacancies, thus resulting in distinct steady-state composition and product selectivity. Co₃O₄-0 h rhombic dodecahedra were completely reduced to Co⁰ and CoO, which presents circa 85 % CH₄ selectivity. In contrast, Co₃O₄-2 h nanorods were partially reduced to CoO, which exhibits a circa 95 % CO selectivity. The crucial role of the Co₃O₄ structure in determining the catalytic performance for higher alcohol synthesis over CuCo-based catalysts is demonstrated. As expected, Cu/Co₃O₄-2 h shows nine-fold higher ethanol yield than Cu/Co₃O₄-0 h owing to the inhibition for methanation.

Full text of DOI:10.1002/anie.201904649

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: The Interplay between Structure and Product Selectivity of CO2
Hydrogenation
Authors: Jinlong Gong
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201904649
Angew. Chem. 10.1002/ange.201904649
Link to VoR: http://dx.doi.org/10.1002/anie.201904649
http://dx.doi.org/10.1002/ange.201904649

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
The Interplay between Structure and Product Selectivity of CO
Hydrogenation
2
Chengsheng Yang, Sihang Liu, Yanan Wang, Jimin Song, Guishuo Wang, Shuai Wang, Zhi-Jian Zhao,
Rentao Mu and Jinlong Gong* [a]
Abstract: Identification of the active structure under the reaction
condition is of great importance for rational design of heterogeneous
catalysts, which, however, is often hampered due to their structural
complexity. This paper describes the interplay between the surface
morphologies, which can eliminate the influence of the catalyst
support. Through the combination of in situ experimental
characterizations and theoretical calculations, we show that the
Co
difference in product selectivity of CO
correlations were observed between the selectivity of CH
(methanation, CO +4H → CH + 2H O) and the reducibility of
Co . Key intermediates influencing the reaction route were
3
O
4
rhombic dodecahedra and nanorods present significant
structure of Co
3
O
4
and the CO
2
hydrogenation. Experimental results
with
2
hydrogenation. Strong
combined with theoretical calculations show that Co
O
3 4
4
morphology-dependent crystallographic surfaces presents different
reducibility and formation energy of oxygen vacancies, thus resulting
2
2
4
2
3
O
4
in distinct product selectivity. Co
3
O
4
-0h rhombic dodecahedra were
monitored under the reaction condition employing diffuse
reflectance infrared Fourier transform spectroscopy (DRIFTS).
0
completely reduced to Co and CoO, which presents ~ 85% CH
4
selectivity. In contrast, Co
CoO, which exhibits a ~ 95% CO selectivity. Furthermore, we
demonstrate the crucial role of the Co structure in determining
the catalytic performance for higher alcohol synthesis over CuCo-
based catalysts. As expected, Cu/Co -2h shows nine-fold higher
ethanol yield than Cu/Co -0h due to the inhibition for methanation.
3
O
4
-2h nanorods were partially reduced to
A
morphology-depended strategy was developed to
synthesize uniform Co-based catalysts. Capping ligands-free
Co rhombic dodecahedra (denoted Co -0h) and nanorod
(denoted Co -2h) mainly enclosed with {111}+{001} and
O
3 4
3
O
4
3 4
O
3
O
4
3
O
4
{110}+{001} planes, respectively, were synthesized by a
modified hydrothermal method (Figure 1).[ The {111} plane
12]
3
O
4
constitutes ~71.4% of the surface area for Co
110} plane occupies 52.6% for Co -2h. The morphology of
Co shows regular transformation from rhombic
dodecahedra to short rods, and the preferential orientation of the
Co changes from {111} plane to {110} plane with increasing
3 4
O -0h, whereas
{
3 4
O
Carbon dioxide (CO
change and ocean
anthropogenic CO will be continuously emitted due to the
increasing burning of fossil fuels. Among the different
approaches explored for controlling CO emission, chemical
conversion has attracted extensive attention.[1] Previous studies
have shown that geometric and electronic effects caused by
specific morphology of catalysts (such as Cu/ZnO nanoplates,
2
) is a greenhouse gas resulting in climate
3
O
4
a
acidification. However, massive
2
3
O
4
the time of the hydrothermal reaction (Figure S1). Vibrational
2
bands upon CO adsorption at ∼2000-2100 cm-1 are observed on
[13]
3 4 x
Co O and assigned to CO adsorption at CoO sites (x>1).
The vibrational bands of CO are shifted to higher wavenumbers
upon increasing in the hydrothermal time (Figure S2b). The
2 2 3 2
Cu/rod-like La O CO and Cu-Ni/CeO -nanotubes) can tune the
{
111} and {110} planes of Co
3
O
4
mainly expose Co termination
reaction routes and in turn the selectivity of CO
2
[12]
with +2 and +3 valence state, respectively.
located at 2032 cm-1 on Co
could fit well to CO adsorbed on Co termination of {111} and
Co3+ termination of {110}.
Thus, the peaks
hydrogenation.[ Other factors including metal particle size,
metal-support interaction and the chemical state of the active
component can also influence the product selectivity.
2]
-1
3
O
4
-0h and 2045 cm on Co
3
O
4
-2h
2
+
The investigation of base-reactions such as methanation and
reverse water gas shift (RWGS) remains essentially the
H
2
2
temperature-programmed reduction (H -TPR) was
conducted to investigate the reducibility of catalysts (Figure 2a).
The peak at low temperature is assigned to the reduction of
foundation for understanding the CO
Therefore, considerable efforts have been devoted to developing
2
conversion route.
Co
3
O
4
to CoO, while the peak at the temperature higher than
catalysts (e.g., Rh/Al
2
O [
3
3], Rh/TiO
2
[4], PtCo/TiO
2
or ZrO
2
,[5]
or
3
Co.
00 ̊C can be attributed to the reduction of CoO to metallic
6]
[7]
Ni/Al
2
O [
3
Ir/CeOH ) with high selectivity towards either CH
4
[14]
Co
3
O
4
-0h is reduced to metallic Co at 460
for Co
shows higher
̊
C while the
CO. Co-based catalyst has been widely used for multiple CO
2
reduction temperature increased to 540
demonstrating the {111} plane of Co O
̊
3 4
O
-2h,
hydrogenation reaction such as methanation,[8] methanol
synthesis[9] and C-C coupling-based reactions (e.g., Fischer-
Tropsch synthesis[10] and higher alcohol synthesis[11]). However,
little is known mechanistically about the structure-performance
relationship in Co-based catalytic systems.
3
4
reducibility than the {110} plane. The bulk structure of used
catalysts was investigated by XRD characterization (Figure 2b-c).
All the samples display the CoO diffraction peak at 42.5°, while
the diffraction characteristic peak of metallic Co decreases
gradually with hydrothermal time increasing from 0h to 3h. The
diffraction peak of Co(111) at 44° disappeared when
hydrothermal treatment for longer than 1h. These results also
3 4
We synthesized uniform Co O bulk catalysts with different
[a]
C. Yang, S. Liu, Y, Wang, Prof. Dr. Z. Zhao, J. Song, G. Wang, S.
Wang, Prof. Dr. R. Mu, Prof. Dr. J. Gong
illustrate the Co
longer hydrothermal time.
To examine the surface structure of the catalysts during the
reaction from 250 C to 350 C, XPS test was carried out (Figure
d-f). The binding energy (BE) of XPS Co 2p3/2 on fresh sample
3 4 3 4
O -0h can be easier reduced than Co O with
Key Laboratory for Green Chemical Technology of Ministry of
Education, School of Chemical Engineering and Technology, Tianjin
University; Collaborative Innovation Center of Chemical Science and
Engineering (Tianjin)
Tianjin 300072 (China)
E-mail: jlgong@tju.edu.cn
̊
̊
2
[
15]
is 779.5 eV, corresponding to Co
CO +H at 250 C, the BE of Co 2p3/2 peak shifts from 779.5 to
80.0 eV and a satellite peak at 786.0 eV appears, which
3
O
4
.
After the reaction of
Supporting information for this article is given via a link at the end of
the document.
̊
2
2
7
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
indicates the reduction of Co
temperature increased to 350
3
O
4
to CoO.[15] When the reaction
3 4
O
2.79 eV), illustrating Co O -0h tends to generate V (Figure S4-
6). Moreover, molecular dynamics simulation was conducted
over CoO slab to investigate the changes of surface structure in
̊
-0h, corresponding to Co (Figure 2d).[15] However,
0
over Co
3
O
4
0
there were almost no Co species over Co
3
O
4
-0.5h and Co
3
O
4
-
O
the presence of V upon the removal of the surface oxygen
2h in the temperature range of 250-350
̊
(Figure S7). V is apparently stable within a time period of 100ps
O
lattice O to Co (n
O
: nCo) of used catalysts were also obtained by
with even 50% of the surface oxygen atoms removed. However,
surface low-coordinated Co atoms gradually come closer to
each other and tend to agglomerate into larger clusters after
removing most of oxygen (>75%) in the top layer of CoO, in
quantitative XPS analysis. The n
from 1.37 to 1.01 after reaction at 350
Co -0h is reduced from 1.38 to 0.69 (Table S1).
O
3 4
: nCo of Co O -2h decreased
̊
3
O
4
agreement with the appearance of metallic Co over used Co
3 4
O -
0
h at 350 C in Figure 2c-d.
̊
Figure 1. Morphological characterizations of fresh catalysts. a), b) TEM, c), d)
HRTEM images and e), f) schematic representation of Co catalysts after
3 4
O
hydrothermal treatment for 0 h (a, c, e) and 2 h (b, d, f). Note, (hkl) indicates a
specified crystal plane, while {hkl} indicates a set of group crystal planes with
the same atomic configuration. Blue ball in (e, f) represents the Co atom, and
red ball in (e, f) represents O atom.
The O 1s peaks at ～529.5 and 531.5 eV should be assigned
to lattice oxygen and adsorbed oxygen species or OH groups,
respectively.[15a, 16] The amount of adsorbed oxygen at ~531.5
eV increases as the reaction temperature increases, and a
higher fraction of adsorbed oxygen was observed on the surface
of Co O -0h (70.1%) than Co O -2h (27.1%) (Figure S3a-c).
3 4 3 4
Moreover, the amount of adsorbed oxygen species displays a
downward trend with hydrothermal time increasing (Figure S3d).
Figure 2. Structural characterizations of various catalysts. a) H
for Co catalysts. b), c) XRD patterns of Co catalysts b) before reaction
and c) after reaction. d), e), f) The Co 2p XPS spectra of Co -0h (d), Co
0.5h (e), and Co -2h (f) catalysts during CO hydrogenation reaction.
Constrained fitting of spectra reveals the presence of three phases, Co
2
-TPR profiles
O
3 4
3 4
O
O
3 4
3 4
O -
O
3 4
2
O ,
3 4
CoO and Co, with their spectral contributions shown in blue, orange and green,
respectively.
O
Oxygen vacancies (V ) present excellent capacity to adsorb and
With hydrothermal time increasing, CO
from ~30% to 10% and product distribution changed greatly.
Specifically, CH selectivity decreases from ~85% to 5% and CO
selectivity increases from 15% to 95% upon increasing the
hydrothermal time to 3h (Figure 3a). The catalytic performance
2
conversion decreases
activate
oxygen-containing
intermediates
during
CO
-0h
, which favor the adsorption of oxygen-
containing species such as CO , because CO desorption
happened at higher temperature over Co -0h in CO -TPD
Figure S3e).[18]
Density functional theory (DFT) calculations were carried out
2
hydrogenation.[ It is assumed that the surface of Co
17]
O
3 4
4
tends to generate V
O
2
2
3
O
4
2
keeps stable for more than 4h (Figure S8). Besides, the CH
selectivity (87%) for Co -0h is still higher than Co -2h (2%)
4
(
3
O
4
3 4
O
at a comparable conversion (Figure S9a). We also investigate
the effects of surface crystallography on the catalytic properties
at different temperatures (Figure S9b). The CO selectivity of
to understand the formation of V
O
on Co
3
O
4
and the reaction
{111} terminated
{110} (2.20-
mechanism. The V formation energy on Co
O
3
O
4
3 4
with O is 0.96 eV, which is lower than that on Co O
Co
3
O
4
-2h exceeds 90% at
a
wide range of reaction
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
temperatures, whereas methanation process dominated on
activity to break the C-O bond and thus the high selectivity
towards methane.
Co
of methane formation over Co
were estimated to be 77.7, 86.5 and 102.5 kJ·mol , respectively
Figure 3b), illustrating that easier methanation and stronger C-
O scission ability over the Co -0h catalyst. The results of
3
O
4
-0h and Co
3
O
4
-0.5h catalysts. The activation energies (E
a
)
3
O
4
-0h, Co -0.5h and Co -2h
3
O
4
3 4
O
-1
(
3
O
4
temperature programmed surface reaction (TPSR, Figure S10a-
b) and in situ DRIFTS (Figure S10c-d) also demonstrate the
3 4
inactivity of methanation reaction on Co O -2h catalyst, because
no signal of CH (m/z=16) and no infrared characteristic peak of
4
-
1
CH
the Co
reactivity.
4
(3016 cm ) can be observed over Co
3 4
O -2h. Conclusively,
3
O
4
catalysts exhibit morphology-dependent catalytic
Figure 3. Catalytic performance of Co
time. a) Catalytic activity and product selectivity of Co
CO hydrogenation. Black solid cycles represent CO
conditions: P = 1 atm, T = 350
C, 150 mg of catalyst, τ = 1.67 gcat·min·L−1
CO /H = 1/3. b) Activation energies of CH formation over the Co -0h,
Co -0.5h and Co -2h catalysts
3
O
4
catalysts with different hydrothermal
catalysts at 4h in
conversion. Reaction
3 4
O
2
2
̊
,
2
2
4
3 4
O
3
O
4
3 4
O
Figure 4. a) In situ DRIFTS spectra recorded during CO
CH OH reduction in the presence of over Co
catalysts. Condition: T=350 C, P=1 atm. b) The absorbance of CH
cm ) in in situ DRIFTS spectra recorded during CO hydrogenation at different
temperature. Condition: P=1 atm, CO /H /Ar = 1/3/1. c), d) CH OH-TPSR
2
, CO, HCOOH and
-0h and Co -2h
O* (1470
The surface of used Co
with graphitized carbon layers (Figure S11). In contrast, Co
h is intact and no coking formation can be observed on the
surface, consistent with results of temperature-programmed
oxidation (TPO) (Figure S12). The peaks <200 C and at 200-
00 C are attributed to the oxidization of surface adsorbed
species and amorphous coke, respectively. The peak at 550 C is
due to the oxidization of graphitized carbon observed on used
Co
-0h (Figure S11b).[19] Generally, the coking is generated
by cleavage of the C-O bond in H CO , dehydrogenation and
graphitization reactions.[6] Therefore, more graphitized coke
3
O
4
-0h becomes rough and is covered
3
H
2
3
O
4
3 4
O
̊
3
3
4
O -
-
1
2
2
2
2
3
profiles for c) used Co O -0 h and d) used Co O -2 h catalysts. e) Reaction
3
4
3
4
̊
energy diagram for the reaction routes of *CH
CoO(100) and CoO(100) with surface V
CH O, transition states (TS), and *CH on Co(111), CoO(100) and CoO(100)
with surface V , are shown as insets, Co: blue, O: red, C: gray, H: white.
3
O dissociation on Co(111),
O
. The DFT optimized geometries of
4
̊
*
3
3
̊
O
3
O
4
Although the C-O bond cleavage is hydrogen-assisted
experimentally, no direct evidence for the scission of CH O*
x
species has been reported.[23] Therefore, in situ DRIFTS
experiments under different reactant atmosphere were carried
out to monitor surface intermediates. Interestingly, similar
x
y
species over Co
scission ability.
3 4
O -0h are indicative of the stronger C-O
The rate of HD formation from the H
indicates that Co -0h has higher activity than Co
Figure S13 and Table S2). Thus, Co -0h catalysts present
better reactivity for H dissociation, which in turn assist C-O
bond cleavage and hydrogenation of *CH to produce methane.
methanation was widely
2
-D
2
exchange reaction
adsorption species were observed under CO
2
or CO
3
O
4
3
O
4
-2h
-1
hydrogenation condition (Figure 4a). The bands at 1375 cm
(
3
O
4
-1
and 1580 cm are assigned to the symmetric and anti-
2
symmetric stretching vibrational modes of OCO, respectively,
x
suggesting the presence of HCOO* intermediates on surface
The reaction mechanism of the CO
2
[24]
(
1
Table S3-S4). Furthermore, there was an additional peak at
examined on metal catalysts, such as Ni, Rh, Ru, Ir.[3, 7, 20-21] It is
-1
470 cm over Co
3
4
O -2h (Table S3). As the absorbance
widely accepted that the reaction proceeds via the formate
intensity of this band decreased after the feed gas was switched
from CO /H /Ar to CO /D /Ar (Figure S14), it could be assigned
pathway.[ Specifically, CH
1b]
formation occurs via the formate
COH* or H CO*.
Among all pathways, the C-O bond scission should be a critical
4
2
2
2
2
hydrogenation and C-O bond scission in H
2
3
3
to the vibrational features of the δ(CH *) modes in the adsorbed
[5]
methoxy (CH
the *CH
3
O*). A similar intensity of the bending vibration of
step and likely determines the overall CH
4
selectivity in CO
2
-1
3
(1439 cm ) in CH
calculations on CoO(100). When feeding CH
3
O* was also obtained by DFT
OH/Ar into the cell,
hydrogenation.[
7b, 20, 22]
3 4
In the case of Co O -0h, the specific
3
morphology improves the reducibility and lowers the
V
O
CH
3
OH was dissociated into *CH
3
which would be subsequently
formation energy. We suppose that the presence of CoO phase
with abundant V on the surface may be responsible for the high
-1
hydrogenated to produce CH
4
(3016 cm ) on Co
3
4
O -0h (Figure
O
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
4
a). These phenomena suggest abundant CH
occupy the surface of Co -2h.
The peak intensity of CH O* displays a declining trend when
temperature increased from 150 to 350 C over Co -0h, with
CH generating gradually (Figure 4b, Figure S15a). These
results show that the reaction towards CH goes through the C-
O dissociation of CH O* and CH * is hydrogenated to produce
CH for Co -0h when the reaction temperature was above
50 C. However, the amount of CH O* species keeps stable on
Co -2h with temperature increasing and become an inactive
intermediate due to its high stability (Figure 4b, Figure S15b).
The superior stability of CH O* may be attributed to lack of V
over used Co -2h. CH OH-TPSR was also conducted to
characterize the stability of CH O*(Figure 4c-d). An obvious CH
signal was observed on Co -0h as temperature increases,
while no CH signal can be observed on Co -2h. These
results confirm that the scission of C-O in CH O* is favored on
Co -0h.
CH
x
O* species
surface O to produce abundant V
bond scission, thereby promoting CH
only provides detailed understanding of structure-performance
relationship in CO hydrogenation over Co-based catalysts, but
also renders guidance for the rational design of catalyst for CO
hydrogenation. Here, we proposed the interplay between
structure and product selectivity, regarding that different types of
the original crystallographic planes of cobalt oxides could induce
O
, which favors the CH
3
-O*
3
O
4
4
formation. This work not
3
̊
3
O
4
2
4
2
4
3
3
4
3 4
O
2
̊
3
the different dynamic structural change under CO
2
3
O
4
hydrogenation atmosphere, which influence the conversion of
the reaction intermediates, and thus the product selectivity. We
also expect that the concept of interplay between surface
structure and catalytic performance can be widely applied in
other catalytic systems.
[
5]
3
O
3
O
4
3
3
4
3
O
4
4
3 4
O
3
Acknowledgements
3
O
4
4
is formed via the direct dissociation of CO or the splitting
We acknowledge the National Key R&D Program of China
of C-O bond in the CH
3
O* intermediate.[ As shown in Figure
25]
(
2016YFB0600901), the National Natural Science Foundation of
4
e and Figure S16, the activation barrier of the C-O bond
cleavage in the form of CH O* and CO over CoO(100) are 2.71
and 4.45 eV, respectively, while those over Co(111) are 1.45
and 2.57 eV, respectively. Thus, the CH O* intermediate may
China (21525626, 21761132023, 91645106), the Program of
Introducing Talents of Discipline to Universities (B06006) for
financial support.
3
3
act as another carbon source, making methane formation more
feasible than the direct dissociation of CO. Additionally, the
dissociation of CH O* to CH * has an activation barrier of 1.01
3 3
Keywords: CO
2
hydrogenation • crystallographic surfaces •
methoxy • reducibility • Co
3
O
4
eV at the oxygen vacancy site of CoO(100), much lower than
pristine CoO(100) (2.71 eV) (Figure 4e). More importantly, the
[
[
1]
2]
a) W. Wang, S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 2011, 40,
3703-3727; b) S. Kattel, P. Liu, J. G. Chen, J. Am. Chem. Soc. 2017,
activation barrier for dissociation of CH
3
O
O* at V of CoO(100)
139, 9739-9754.
(
1.01 eV) is even lower than that of Co(111) (1.45 eV),
a) S. Du, W. Tang, X. Lu, S. Wang, Y. Guo, P.-X. Gao, Adv. Mater.
Inter. 2017, 1700730; b) M. Duyar, C. Tsai, J. L. Snider, J. A. Singh, A.
Gallo, J. S. Yoo, A. J. Medford, F. Abild-Pedersen, F. Studt, J.
Kibsgaard, S. Bent, J. K. Norskov, T. Jaramillo, Angew. Chem. Int. Ed.
2018, 57, 15045-15050; c) K. Chen, X. Duan, H. Fang, X. Liang, Y.
Yuan, Catal. Sci. Tech. 2018, 8, 1062-1069; d) Q. Tan, Z. Shi, D. Wu,
Ind. Eng. Chem. Res. 2018, 57, 10148-10158; e) P. Gao, F. Li, H. Zhan,
N. Zhao, F. Xiao, W. Wei, L. Zhong, H. Wang, Y. Sun, J. Catal. 2013,
suggesting the dissociation of CH
more feasible than Co(111). This indicates that the V
could promote the CH
methanation.
3
O* at V
O
of CoO(100) is even
over CoO
3
O* intermediate dissociation and
O
We further applied the as-prepared Co
3
O
4
catalysts to
alcohols synthesis under higher pressures (1-30 bar). The
-1
-1
highest space time yield of methanol (6.21 mmol g·cat ·h )
Figure S17, Table S5) was achieved over Co -2h at 250
and 6 bar. The accumulation of CH O* on surface of Co -2h
may account for the higher methanol yield. Co -supported
0wt%Cu catalyst was synthesized (Figure S18) for higher
298, 51-60.
(
3
O
4
̊C
[
[
3]
4]
D. Heyl, U. Rodemerck, U. Bentrup, ACS Catal. 2016, 6, 6275-6284.
J. C. Matsubu, V. N. Yang, P. Christopher, J. Am. Chem. Soc. 2015,
137, 3076-3084.
x
3 4
O
3
O
4
1
[5]
S. Kattel, W. Yu, X. Yang, B. Yan, Y. Huang, W. Wan, P. Liu, J. G.
Chen, Angew. Chem. Int. Ed. Engl. 2016, 55, 7968-7973.
L. R. Winter, E. Gomez, B. Yan, S. Yao, J. G. Chen, Appl. Catal., B
3 3 4
alcohol synthesis. More CH O* was observed over Cu/Co O -2h
[
[
6]
7]
3 4
than Cu/Co O -0h by DRIFTS (Figure S19). As expected,
2018, 224, 442-450.
Cu/Co
3
O
4
-1
-2h exhibited nine-fold higher ethanol yield (1.87 mmol
a) S. Li, Y. Xu, Y. Chen, W. Li, L. Lin, M. Li, Y. Deng, X. Wang, B. Ge,
C. Yang, S. Yao, J. Xie, Y. Li, X. Liu, D. Ma, Angew. Chem. Int. Ed.
Engl. 2017, 56, 10761-10765; b) X. D. Chen, X. Su, H. Y. Su, X. Y. Liu,
S. Miao, Y. H. Zhao, K. J. Sun, Y. Q. Huang, T. Zhang, ACS Catal.
2017, 7, 4613-4620.
-1
-1 -1
3 4
g·cat ·h ) than Cu/Co O -0h (0.19 mmol g·cat ·h ) due to the
inhibition for methanation (Table S6). These results demonstrate
the crucial role of the Co structure in CuCo-based catalysts
3
O
4
in determining the catalytic performance for higher alcohol
synthesis. However, the detailed mechanism needs to be further
studied.
[8]
W. Li, X. Nie, X. Jiang, A. Zhang, F. Ding, M. Liu, Z. Liu, X. Guo, C.
Song, Appl. Catal., B 2018, 220, 397-408.
[
9]
a) M. U. Khan, L. Wang, Z. Liu, Z. Gao, S. Wang, H. Li, W. Zhang, M.
Wang, Z. Wang, C. Ma, J. Zeng, Angew. Chem. Int. Ed. Engl. 2016, 55,
In summary, we demonstrated that selectivity of CH
and selectivity of CO increased dramatically with the
crystallographic planes changed from {111} to {110}. Co {111}
showed higher reducibility and lower V formation energy than
Co {110}. CH O* was found to be the key intermediate and
the accumulation of CH O* on the surface inhibit the C-O bond
scission and be responsible for the high reactivity for RWGS.
We propose that the {111} plane of Co is easier to lose
4
reduced
9548-9552; b) C. S. Li, G. Melaet, W. T. Ralston, K. An, C. Brooks, Y.
3
O
4
Ye, Y. S. Liu, J. Zhu, J. Guo, S. Alayoglu, G. A. Somorjai, Nat.
Commun. 2015, 6, 6538.
O
3
O
4
3
[
10] a) W. Li, A. Zhang, X. Jiang, M. J. Janik, J. Qiu, Z. Liu, X. Guo, C. Song,
J. CO2 Util. 2017, 23, 219-225; b) C. Xie, C. Chen, Y. Yu, J. Su, Y. Li, G.
A. Somorjai, P. Yang, Nano Lett. 2017, 17, 3798-3802.
3
3
O
4
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
[
11] a) Z. He, Q. Qian, J. Ma, Q. Meng, H. Zhou, J. Song, Z. Liu, B. Han,
Angew. Chem. Int. Ed. Engl. 2016, 55, 737-741; b) L. Wang, L. Wang, J.
Zhang, X. Liu, H. Wang, W. Zhang, Q. Yang, J. Ma, X. Dong, S. J. Yoo,
J. G. Kim, X. Meng, F. S. Xiao, Angew. Chem. Int. Ed. Engl. 2018, 57,
[19] a) M. Larsson, M. Hultén, E. A. Blekkan, B. Andersson, J. Catal. 1996,
164, 44-53; b) F. Jiang, L. Zeng, S. Li, G. Liu, S. Wang, J. Gong, ACS
Catal. 2014, 5, 438-447; c) N. Martín, M. Viniegra, E. Lima, G.
Espinosa, In. Eng. Chem. Res. 2004, 43, 1206-1210.
6
104-6108; c) B. Ouyang, S. Xiong, Y. Zhang, B. Liu, J. Li, Appl. Catal.,
[20] a) X. Wang, Y. Hong, H. Shi, J. Szanyi, J. Catal. 2016, 343, 185-195; b)
A. Sápi, G. Halasi, J. Kiss, D. G. Dobó, K. L. Juhász, V. J. Kolcsár, Z.
Ferencz, G. Vári, V. Matolin, A. Erdőhelyi, Á. Kukovecz, Z. Kónya, J.
Phy. Chem. C 2018, 122, 5553-5565.
A 2017, 543, 189-195.
[
[
12] X. W. Xie, Y. Li, Z. Q. Liu, M. Haruta, W. J. Shen, Nature 2009, 458,
7
46-749.
13] a) H. K. Lin, C. B. Wang, H. C. Chiu, S. H. Chien, Catal. Lett. 2003, 86,
3-68; b) Y. Lou, J. Ma, X. Cao, L. Wang, Q. Dai, Z. Zhao, Y. Cai, W.
[21] a) H. C. Wu, Y. C. Chang, J. H. Wu, J. H. Lin, I. K. Lin, C. S. Chen,
Catal. Sci. Tech. 2015, 5, 4154-4163; b) P. Panagiotopoulou, Appl.
Catal., A 2017, 542, 63-70; c) A. Kim, C. Sanchez, G. Patriarche, O.
Ersen, S. Moldovan, A. Wisnet, C. Sassoye, D. P. Debecker, Catal. Sci.
Tech. 2016, 6, 8117-8128; d) J. C. Matsubu, S. Zhang, L. DeRita, N. S.
Marinkovic, J. G. Chen, G. W. Graham, X. Pan, P. Christopher, Nat.
Chem. 2017, 9, 120-127; e) X. Wang, H. Shi, J. Szanyi, Nat. Commun.
2017, 8, 513.
6
Zhan, Y. Guo, P. Hu, G. Lu, Y. Guo, ACS Catal. 2014, 4, 4143-4152; c)
M. Haruta, J. Catal. 1993, 144, 175-192.
[
[
14] Z. Wang, J. J. Spivey, Appl. Catal., A 2015, 507, 75-81.
15] a) L. F. Liotta, G. Di Carlo, G. Pantaleo, A. M. Venezia, G. Deganello,
Appl. Catal., B 2006, 66, 217-227; b) M. C. Biesinger, B. P. Payne, A. P.
Grosvenor, L. W. M. Lau, A. R. Gerson, R. S. Smart, Appl. Surf. Sci.
2
011, 257, 2717-2730; c) Z. Chen, C. X. Kronawitter, B. E. Koel, Phys.
[22] R. Mutschler, E. Moioli, W. Luo, N. Gallandat, A. Züttel, J. Catal. 2018,
366, 139-149.
Chem. Chem. Phys. 2015, 17, 29387-29393; d) T. J. Chuang, C. R.
Brundle, D. W. Rice, Surf. Sci. 1976, 59, 413-429.
[23] a) E. Shustorovich, J. Catal. 1988, 113, 341-352; b) F. Solymosi, J.
Catal. 1990, 122, 166-177; c) L. C. Grabow, M. Mavrikakis, ACS Catal.
2011, 1, 365-384; d) Y. Guo, S. Mei, K. Yuan, D.-J. Wang, H.-C. Liu,
C.-H. Yan, Y.-W. Zhang, ACS Catal. 2018, 8, 6203-6215..
[
16] a) D. A. Pawlak, M. Ito, M. Oku, K. Shimamura, T. Fukuda, J. Phys.
Chem. B 2002, 106, 504-507; b) M. Konsolakis, Z. Ioakimidis, T. Kraia,
G. E. Marnellos, Catalysts 2016, 6, 39; c) G. A. Carson, M. H. Nassir, M.
A. Langell, J. Vac. Sci. Technol. A 1996, 14, 1637-1642; d) J. L.
Ayastuy, A. Gurbani, M. P. González-Marcos, M. A. Gutiérrez-Ortiz, Int.
J. Hydrogen Energy 2012, 37, 1993-2006; e) Y. Zheng, K. Li, H. Wang,
Y. Wang, D. Tian, Y. Wei, X. Zhu, C. Zeng, Y. Luo, J. Catal. 2016, 344,
[24] T. G. Kelly, A. L. Stottlemyer, H. Ren, J. G. Chen, J. Phy. Chem. C
2011, 115, 6644-6650.
[25] a) A. J. Medford, A. C. Lausche, F. Abild-Pedersen, B. Temel, N. C.
Schjødt, J. K. Nørskov, F. Studt, Top. Catal. 2013, 57, 135-142; b) M. P.
Andersson, F. Abild-Pedersen, I. N. Remediakis, T. Bligaard, G. Jones,
J. Engbæk, O. Lytken, S. Horch, J. H. Nielsen, J. Sehested, J. Catal.
2008, 255, 6-19; c) A. C. Lausche, A. J. Medford, T. S. Khan, Y. Xu, T.
Bligaard, F. Abild-Pedersen, J. K. Nørskov, F. Studt, J. Catal. 2013,
307, 275-282.
365-377.
[
[
17] L. Li, J. Yan, T. Wang, Z. J. Zhao, J. Zhang, J. Gong, N. Guan, Nat.
Commun. 2015, 6, 5881.
18] a) O. Martin, A. J. Martin, C. Mondelli, S. Mitchell, T. F. Segawa, R.
Hauert, C. Drouilly, D. Curulla-Ferre, J. Perez-Ramirez, Angew. Chem.
Int. Ed. Engl. 2016, 55, 6261-6265; b) N. Rui, Z. Wang, K. Sun, J. Ye,
Q. Ge, C.-j. Liu, Appl. Catal., B 2017, 218, 488-497.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201904649
COMMUNICATION
COMMUNICATION
This communication describes the
correlation of Co O catalysts with
3 4
different crystallographic surfaces
with reducibility and formation
energy of oxygen vacancies,
Chengsheng Yang, Sihang Liu,
Yanan Wang, Jimin Song, Guishuo
Wang, Shuai Wang, Zhi-Jian Zhao,
Rentao Mu and Jinlong Gong*
resulting
in
distinct
product
Page No. – Page No.
selectivity of CO
The key step (CH
breaking) which affect C
distribution was identified by the in
situ DRIFTS and DFT, suggesting
that oxygen vacancies can promote
2
hydrogenation.
The Interplay between Structure
3
-O* bond
product
and Product Selectivity of CO
Hydrogenation
2
1
3
the CH -O* bond breaking.
This article is protected by copyright. All rights reserved.