Elucidating the Copper-H?gg Iron Carbide Synergistic Interactions for Selective CO Hydrogenation to Higher Alcohols

Article abstract of DOI:10.1021/acscatal.7b01469

CO hydrogenation to higher alcohols (C₂₊OH) provides a promising route to convert coal, natural gas, shale gas, and biomass feedstocks into value-added chemicals and transportation fuels. However, the development of nonprecious metal catalysts with satisfactory activity and well-defined selectivity toward C₂₊OH remains challenging and impedes the commercialization of this process. Here, we show that the synergistic geometric and electronic interactions dictate the activity of Cu⁰-χ-Fe₅C₂ binary catalysts for selective CO hydrogenation to C₂₊OH, outperforming silica-supported precious Rh-based catalysts, by using a combination of experimental evidence from bulk, surface-sensitive, and imaging techniques collected on real and high-performance Cu-Fe binary catalytic systems coupled with density functional theory calculations. The closer is the d-band center to the Fermi level of Cu⁰-χ-Fe₅C₂(510) surface than those of χ-Fe₅C₂(510) and Rh(111) surface, and the electron-rich interface of Cu⁰-χ-Fe₅C₂(510) due to the delocalized electron transfer from Cu⁰ atoms, facilitates CO activation and CO insertion into alkyl species to C₂-oxygenates at the interface of Cu⁰-χ-Fe₅C₂(510) and thus enhances C₂H₅OH selectivity. Starting from the CHCO intermediate, the proposed reaction pathway for CO hydrogenation to C₂H₅OH on Cu⁰-χ-Fe₅C₂(510) is CHCO + (H) → CH₂CO + (H) → CH₃CO + (H) → CH₃CHO + (H) → CH₃CH₂O + (H) → C₂H₅OH. This study may guide the rational design of high-performance binary catalysts made from earth-abundant metals with synergistic interactions for tuning selectivity. (Chemical Equation Presented).

Full text of DOI:10.1021/acscatal.7b01469

Subscriber access provided by UNIV OF SOUTHERN QUEENSLAND
Article
Elucidating the Copper–Hägg Iron Carbide Synergistic
Interactions for Selective CO Hydrogenation to Higher Alcohols
Yongwu Lu, Riguang Zhang, Baobao Cao, Binghui Ge, Franklin (Feng) Tao, Junjun Shan,
Luan Nguyen, Zhenghong Bao, Tianpin Wu, Jonathan W. Pote, Baojun Wang, and Fei Yu
ACS Catal., Just Accepted Manuscript • Publication Date (Web): 21 Jun 2017
Downloaded from http://pubs.acs.org on June 21, 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 13
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Elucidating the Copper–Hägg Iron Carbide Synergistic Interactions for
Selective CO Hydrogenation to Higher Alcohols
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Yongwu Lu,†,⊗ Riguang Zhang, Baobao Cao, Binghui Ge, Franklin (Feng) Tao, Junjun Shan,
‡,⊗
§
‖
⏊
⏊
⏊
†
#
†
‡,*
†,*
Luan Nguyen, Zhenghong Bao, Tianpin Wu, Jonathan W. Pote, Baojun Wang, and Fei Yu
†Department of Agricultural and Biological Engineering, Mississippi State University, Mississippi State, MS 39762,
United States
‡Key Laboratory of Coal Science and Technology of Ministry of Education and Shanxi Province, Taiyuan University of
Technology, Taiyuan, Shanxi 030024, P.R. China
§School of Materials Science and Engineering, Southwest Jiaotong University, Chengdu, Sichuan 610031, P. R. China
‖Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Science, Beijing
1
00190, P.R. China
⏊Department of Chemical and Petroleum Engineering and Department of Chemistry, University of Kansas, Lawrence,
KS 66045, United States
#X-ray Science Division, Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439, United
States
ABSTRACT: CO hydrogenation to higher alcohols (C2+OH) provides a promising route to convert coal, natural gas, shale gas,
and biomass feedstocks into value-added chemicals and transportation fuels. However, the development of nonprecious
metal catalysts with satisfactory activity and well-defined selectivity towards C2+OH remains challenging and impedes the
commercialization of this process. Here we show that the synergistic geometric and electronic interactions dictate the activ-
0
ity of Cu –χ-Fe
5
2
C binary catalysts for selective CO hydrogenation to C2+OH, outperforming silica-supported precious Rh-
based catalysts, by using a combination of experimental evidence from bulk, surface-sensitive, and imaging techniques col-
lected on real and high-performance Cu-Fe binary catalytic systems coupled with density functional theory calculations. The
0
closer the d-band center to the Fermi level of Cu –χ-Fe
5
C
2
(510) surface than those of χ-Fe
5
C
2
(510) and Rh(111) surface, and
0
0
the electron-rich interface of Cu –χ-Fe
5
C
2
(510) due to the delocalized electron transfer from Cu atoms, which facilitate CO
activation and CO insertion into alkyl species to C
0
2
-oxygenates at the interface of Cu –χ-Fe
5
C
2
(510) and thus enhance
OH on
OH. This study may guide the
C
2
H
5
OH selectivity. Starting from CHCO intermediate, the proposed reaction pathway for CO hydrogenation to C
2
H
5
0
Cu –χ-Fe
5
C
2
(510) is CHCO+(H)→CH
2
CO+(H)→CH
3
CO+(H)→CH
3
CHO+(H)→CH
3
CH
2
O+(H)→C
2
H
5
rational design of high-performance binary catalysts made from earth-abundant metals with synergistic interactions for
tuning selectivity.
KEYWORDS: CO hydrogenation, higher alcohols, copper, Hägg iron carbide, synergistic effect, reaction mechanism
To decrease the dependence on crude oil and adhere
to strict global environmental stipulations, the selective
INTRODUCTION
Binary metal materials have gained a crucial interest
1
–
conversion of syngas (CO + H
2
) into clean fuels and value-
as functional materials for several emerging technologies.
8,9
7
added chemicals is regarded a primary scientific target.
Syngas can be derived from various carbonaceous re-
sources, such as coal, natural gas, shale gas, municipal solid
waste, or biomass feedstocks by using gasification or re-
As catalysts, they often exhibit electronic and chemical
properties that are distinct from those of their parent met-
1
,2
als, thereby enabling the development of novel catalysts
with enhanced selectivity, activity, and stability through
8,9
1
2
2,3
3
forming technologies. The synthesis of ethanol and high-
bifunctional, ligand, geometric, electronic, or lattice
4
er alcohols (C2+OH) from syngas leaps out as a catalytic
route of prominent interest, because these compounds can
be used as hydrogen carriers, fuels, fuel additives in gaso-
line, precursors for major platform chemicals such as ole-
fins, and reagents for manufacturing pharmaceuticals,
cosmetics, lubricants, plasticizers, and detergents.5
Additionally, the reaction is a prototypical example that
involves the synergistic roles between proximate active
strain effect. The synergistic interaction in binary metal
catalyst represents an efficient approach towards desira-
ble properties by either the creation of hybrid sites or the
concerted action of concomitant functionalities.1,5–7 Binary
metal catalysts also show promise of replacing precious
–7,10,11
metal catalysts with equally active and selective catalysts
composed of earth-abundant metals.3,5,6
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 13
sites with diverse functionality,5–7,10,11 as both CO dissocia-
tion (direct or H-assisted dissociation) and CO/CHO inser-
tion into CH (x = 1–3) species are required to occur simul-
taneously to form CH CO/CH CHO (x = 1–3), thereafter
undergoing a stepwise hydrogenation to selectively pro-
tively. The indexed selected area electron diffraction pat-
tern (Figure S4b) further manifest the existence of CuO
and Fe
Catalytic Performance. 3DOM Cu-Fe catalysts were re-
duced in situ using H /CO = 1 at 300 ℃ for 48 h. The cata-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
x
3
4
O .
x
x
2
5
–7,10,11
duce C2+OH.
In contrast to methanol synthesis based
lytic tests were performed under a condition of T = 260 ,
9
–1
on Cu-based catalyst or Fischer–Tropsch synthesis (FTS)
of hydrocarbons by using Fe- or Co-based catalysts,
P = 700 psig, H
2
/CO = 1, GHSV = 2000 h , and time-on-
8
,12–14
stream (TOS) of 120 h after achieving steady state. The
selectivity of alcohols, hydrocarbons, and CO is presented
currently the development of heterogeneous catalysts with
high selectivity and space-time yield to achieve commer-
2
in Figure 1a. The representative gas chromatography–
mass spectrometry (GC–MS) analysis of aqueous and liquid
organic products from CO hydrogenation over 3DOM
5
–
cial-scale C2+OH production remains a crucial challenge.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
,10,11
Rh is the only single metal reported that can selective-
Cu
2
Fe
1
catalyst are shown in Figure S5. The 1-alcohols and
ly produce C2+OH from syngas, however, its prohibitive
cost and scarcity make a large-scale heterogeneous pro-
hydrocarbons distribution agree with Anderson-Schulz-
7
,8
Flory (ASF) distribution (Figure S6), and the calculated
ASF chain growth probability (α) of 1-alcohols and hydro-
carbons are summarized in Tables S1–S2.
1
1,15–19
cess impossible.
This has urged the search of inex-
pensive binary catalysts containing earth-abundant metals
such as Cu-Fe catalyst,1
1,20,21
because of its high activity and
8
0
a
Alcohols
Hydrocarbons
CO2
r
C
2+OH
40
30
20
10
0
b
d
6
5
4
3
2
1
0
selectivity that is reported comparable to Rh-based cata-
lysts. However, the origin of the synergistic interaction, the
nature of the active sites, and the reaction mechanism for
CO hydrogenation to C2+OH over Cu-Fe binary catalysts
that play a pivotal role in rational development of robust
catalysts for industrial application are not fully understood
at the fundamental level. This is partly due to the complex-
ity of the CO hydrogenation reaction on Cu-Fe binary cata-
lyst, whose reactivity is affected by many factors such as
geometric, electronic, or bifunctional effect.
Here we show that three-dimensionally ordered
macroporous (3DOM) Cu-Fe catalysts exhibit higher per-
formance than that of silica-supported Rh-based catalysts
for selective CO hydrogenation to C2+OH. Cs-corrected
high-angle annular dark-field scanning transmission elec-
tron microscopy (HAADF-STEM) imaging and electron
energy loss spectroscopy (EELS) elemental mapping is
employed to study the Cu-Fe binary active sites. In situ
surface chemistry studies are tracked with ambient pres-
sure X-ray photoelectron spectroscopy (AP-XPS) to identi-
fy the active sites of Cu-Fe catalyst. In situ X-ray absorption
spectroscopy is conducted to reveal the bulk structure of
Cu-Fe active sites. Additionally, we perform density func-
tional theory (DFT) calculations to explore the origin of the
synergistic interactions, unravel the nature of the active
sites, and elucidate the underlying reaction mechanism of
Cu-Fe catalyst for selective CO hydrogenation to C2+OH at a
molecular scale.
70
Alcohols
C2+OH
60
50
40
30
20
10
0
e
1
2F
e
1
e
1
u1Fe
3
2
Fe
u3
F
u
u1
F
M
-P
1
C
C
C
3
C
O
M
M
C
e
M
3DOM 3DOM
M
D
3
D
O
u2
F
O
O
3
D
D
60
c
3
DOM Cu
2
Fe1
70 3DOM Cu
2Fe1
Hydrocarbons
Alcohols
50
60
3DOM Cu
3
Fe1
50
40
3DOM Cu
1
Fe1
40
30
20
3DOM Cu
1Fe2
30
C2+OH alcohols
3
DOM Fe
1
0
3DOM-PM Cu2Fe1
CO
2
20
0
0
20 40 60 80 100 120
Time on stream (h)
0
20 40 60 80 100 120
Time on stream (h)
Figure 1. 3DOM Cu-Fe catalysts (a) The selectivity of alcohols,
hydrocarbons, and CO
2
. (b) The formation rate of C2+OH
–1
(
mmol gcat h ) (scatter), the selectivity to alcohols and C2+OH
alcohols (column). (c) Time-on-stream (TOS) evolution of CO
conversion. 3DOM Cu Fe catalyst (d) TOS evolution of prod-
uct selectivity. (Reaction conditions: T = 260 ℃, P = 700 psig,
/CO = 1, and GHSV = 2000 h–1).
2
1
H
2
3
DOM Cu catalyst shows poor activity (Figure 1b).
3
4
DOM Fe catalyst displays a total alcohol selectivity of
.3%, a C2+OH selectivity of 1.3%, and a C2+OH formation
–1 –1
rate of 0.11 mmol gcat
h (Figure 1b). Importantly, 3DOM
Cu-Fe catalysts exhibit a higher activity and selectivity to-
wards C2+OH (Figure 1b). At a Cu/(Cu+Fe) atomic ratio of
RESULTS AND DISCUSSION
Synthesis and Characterization. 3DOM Cu-Fe binary cat-
0
.67 (3DOM Cu
2
1
Fe ), the selectivity to alcohols (31.3%)
and C2+OH (26.1%) is maximum; meanwhile the formation
alysts were synthesized by a “glyoxylate route” poly (me-
–1 –1
rate of C2+OH reaches 5.65 mmol gcat h , ca. one order of
2
1
thyl methacrylate) colloidal crystal template method as
described in the section of methods. 3DOM Cu Fe catalyst
magnitude higher than that of 3DOM Fe catalyst (Figure
2
1
1
a). The CO conversion is 58.4% for 3DOM Cu
without deactivation during TOS of 120 h (Figure 1c). The
carbon selectivity to CO (11.5%), hydrocarbons (57.2%),
2
Fe catalyst
1
with hierarchically macroporous structure contains peri-
odic voids with an average diameter of (200±10) nm and a
wall thickness of (50±5) nm (Figure S1). X-ray diffraction
2
alcohols (31.3%), and C2+OH (26.1%) for 3DOM Cu
catalyst for TOS of 120 h is presented in Figure 1d.
2
Fe
1
(
(
XRD) patterns show that fresh catalysts consist of CuO
JCPDS 48-1548) and Fe (JCPDS 65-3107) (Figure S2).
3
O
4
In comparison, we prepared a (5wt.%)Rh/SiO
and tested it in the same reaction condition with that of
3DOM Cu Fe catalyst. The XRD patterns of the fresh and
reduced (5wt.%)Rh/SiO are shown in Figure S7, where Rh
species is identified as RhO (JCPDS 21-1315) and Rh
JCPDS 05-0685), respectively. The catalytic performance
2
catalyst
XPS at the Cu 2p and Fe 2p core-levels further confirm the
existence of CuO and Fe (Figure S3). HRTEM image of
fresh 3DOM Cu Fe catalyst is shown in Figure S4a. The
lattice fringes of 4.87 and 2.32 Å are characteristic inter-
layer spacing of Fe (111) and CuO (111) planes, respec-
3
O
4
2
1
2
1
2
2
3
O
4
(
ACS Paragon Plus Environment

Page 3 of 13
ACS Catalysis
of (5wt.%)Rh/SiO
2
catalyst shows a CO conversion of 6.3%,
gistic interfacial interactions, we prepared a physical mix-
ture (PM) of 3DOM Fe and 3DOM Cu catalyst, denoted as
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
C
0
2+OH selectivity of 11.2 %, and C2+OH formation rate of
–1 –1
.63 mmol gcat
h
(Table S3). Thus, the catalytic perfor-
Fe shows a CO conversion of 58.4%,
2+OH selectivity of 26.1%, and C2+OH formation rate of
3DOM PM-Cu
tion with those of 3DOM Cu-Fe catalysts. 3DOM PM-Cu
catalyst shows a CO conversion of 24.5% (Figure 1c),
2
Fe
1
, and tested it in the same reaction condi-
mance of 3DOM Cu
C
2
1
2
Fe
1
5.65 mmol gcat h–1, which is much higher than that of
(5wt.%)Rh/SiO , and a select of other silica-supported Rh-
based catalysts (Table S3) reported in the literatures.
–1
C
2+OH selectivity of 1.1%, and C2+OH formation rate of 0.11
–1
–1
2
mmol gcat
h
(Figure 1b), which is much less active and
Fe
catalyst. Furthermore, the calculated intrinsic activity of
15–18
selective for C2+OH production than that of 3DOM Cu
2
1
The catalytic performances of 3DOM Cu-Fe catalysts
indicate that the activity and selectivity may attribute to
the synergistic interfacial interactions of Cu-Fe binary ac-
tive sites. To further validate the hypothesis of the syner-
2
–2 –1
3DOM Cu
2
Fe
1
catalyst is 6.6ꢀ10– mmol m
h (Table S4),
which is one order of magnitude higher than that of 3DOM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
–2 –1
PM-Cu
2
Fe
catalyst (2.0ꢀ10– mmol m h ).
1
(111)
a
ꢁ
b
•
experiment
fit
3DOM Fe
ꢁ
Cu
-Fe
JCPDS 004-0836
JCPDS 051-0997
χ
χ
χ
-Fe
5
C
2
2
(II)
(I)
(III)
-Fe5
C
C
ꢀ
χ
5C
2
-Fe
5
2
3
+
Fe spm
3DOM Cu1Fe2
(
200)
ꢁ
(
220)
(311)
ꢁ
ꢁ
(222)
ꢁ
3
DOM Cu
3
Fe
Fe
Fe
1
3DOM Cu1Fe1
3DOM Cu
2
1
3
DOM Cu
1
1
2
3DOM Cu2Fe1
3DOM Cu
1
Fe
3DOM Cu3Fe1
ꢀ
ꢀ
3DOM Fe
ꢀ
ꢀꢀ
ꢀ
ꢀꢀ
30
40
50
60
70
80
90
100
-10
-5
0
5
10
o
2
θ
Velocity (mm/s, relative to α-Fe)
Figure 2. 3DOM Cu-Fe catalysts after reduction (a) XRD pattern and (b) Mössbauer spectra. 3DOM Cu
2
Fe
1
catalyst after reduction
(
c) HAADF-STEM image and (d–g) EELS mapping of the selected region (green box) showing elemental distribution of C (d), Fe (e),
0
Cu (f), and their overlap (g). Cs-corrected HAADF-STEM images (h) Cu –χ-Fe
5
C
2
(510) that involves Cu step and twin boundary,
0
inset the fast Fourier transform of the selected region (red dashed box) showing Cu is aligned along the [110] zone axis. (i) The
magnification of white dashed box in (h).
Structural Characterization of the Catalytic Active Sites.
XRD and Mössbauer spectroscopy were used to identify
the active sites of 3DOM Cu-Fe catalysts after reduction at
indicates the selected region for EELS elemental mapping.
As shown in EELS elemental maps (Figure 2d–g), the cata-
lyst reduction treatment ended finally with Cu species in-
timate contact with iron carbide species. Cs-corrected
HAADF-STEM image (Figure 2h) displays the lattice fringe
3
4
00℃ using H
2
/CO = 1 for 48 h. The long reduction time of
8 h was applied to ensure the complete reduction and
carburization of iron oxide in the 3DOM catalysts into iron
carbide. XRD pattern of the reduced catalysts (Figure 2a)
of 2.05 Å from χ-Fe
5
C (510) (JCPDS 051-0997) is closely
2
0
adjacent to Cu to generate intimate and extensive
interface. The inset in Figure 2h is the fast Fourier trans-
form of the selected region (red dashed box), indicating
0
indicate that CuO is reduced to Cu (JCPDS 004-0836), and
Fe is reduced and carburized into Hägg carbide (χ-Fe
JCPDS 051-0997). Figure 2b shows Mössbauer spectra of
reduced catalysts that can be fitted with three sextets and
one doublet, representing χ-Fe with different hyperfine
The fiting Mössbauer parameters are
summarized in Table S5. Figure 2c shows HAADF-STEM
image of reduced 3DOM Cu Fe catalyst, and the green box
3
O
4
5
2
C )
0
0
(
that Cu is aligned along the [110] zone axis. Cu involves
typical Cu step and twin boundary (Figure 2h). Figure 2i is
the magnification of the selected region (white dashed box)
C
5 2
8
,22
parameters.
in Figure 2h. Because neither Cu nor χ-Fe
5
2
C alone could
achieve the high selectivity of C2+OH in 3DOM Cu-Fe cata-
0
2
1
lysts (Figure 1), we advocate the formation of the Cu –χ-
ACS Paragon Plus Environment

ACS Catalysis
binary active sites (Figure 2i) and their synergistic
Page 4 of 13
Fe
5
C
2
charge transfer between the transition metal 3d and sur-
rounding ligand oxygen 2p orbitals, and they do not pre-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
interactions that result in the observed catalytic perfor-
mance.
0
sent in Cu
2
O and Cu because of their fully filled 3d orbit-
2
3
In situ Studies of Surface Chemistry Using AP-XPS. To
achieve a deep insight into the Cu-Fe binary active sites for
CO hydrogenation, we performed in situ AP-XPS. During
the data acquisition of in situ studies, the gaseous reactants
remained around catalyst at a certain temperature. The
als. The Cu 2p3/2 peak at binding energy of 933.6 eV (Fig-
2
4
ure 3a) is thus indicative of CuO. The broadening of main
Cu LMM peak at 918.0 eV to lower kinetic energies was
found during in situ reduction up to 100℃ (Figure 3b). The
drop of shakeup satellite intensity and the shift of Cu 2p3/2
peak to lower binding energy indicate the formation of
fresh 3DOM Cu
2
1
Fe catalyst was pretreated in situ with 1.0
+
24
mbar O at 300℃ for 3 h to remove any residual carbona-
ceous surface species (Figure 3). The Cu 2p3/2 and Cu 2p1/2
peaks are accompanied with distinct shakeup satellites at
2
monovalent copper (Cu ). Upon heating to 150℃, the
appearance of Cu LMM peaks at 918.6 and 921.3 eV can be
attributed to metallic copper; meanwhile, no shakeup sat-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
+
24–26
binding energies of 942 and 962 eV after O
2
pretreatment
ellite of Cu species was detected in Cu 2p spectra.
(
Figure 3a). These characteristic satellites stem from
0
a
0
+
Cu
c Fe 2p
Csurf
Cu 2p
(Cu ,Cu )
32.4
2p3/2
b
Cu (LMM)
707.3
d C 1s
9
18.6
720.3
2p
2p
284.6
9
+
3/2
9
52.2
p1/2
Cu
916.4
°C
0
Cu
921.3
1/2
χ-Fe5C2
CO+H
2
400
°
C
C
2
CO+H
2
400
300
CO+H
CO+H
2
400
300
°
C
C
2
83.5
CO+H
2
2
°C
iron carbide
CO+H
2
300
°
CO+H
2
400
°
C
2
°
CO+H
250
°C
CO+H
2
300°
C
CO+H
2
2
250
200
°
C
CO+H
CO+H
2
250
°
C
CO+H
2
200°C
2
200
°
C
CO+H
CO+H
2
250
°
C
CO+H
°C
7
22.6
711.7
710.7.
CO+H
2
150
100
°
C
2+
CO+H
CO+H
2
2
150
°C
Fe
+Fe
Fe
Satellite peak
3
O
4
2
200°
C
Cu
9
CO+H
CO+H
2
2
150
°
C
724.0
18.0
100
°C
Fe
2
O3
3
O4
CO+H
2
2
150
100
°C
CO+H
2
°
C
100
°
C
1.0mbar O
2
300
°
C
2O3
CO+H
°
C
1
.0mbar O
2
300°C
1
.0 mbar O2 300°C
2
+
Cu2+
Cu
Satellite peaks
1
.0 mbar O2 300°C
933.6
9
60 950 940 930
912 914 916 918 920 922 924
Kinetic energy (eV)
730
720
710
290 288 286 284 282 280
Binding energy (eV)
Binding energy (eV)
Binding energy (eV)
Figure 3. In situ AP-XPS studies of 3DOM Cu
2
Fe
1
catalyst surface with O
2
pretreatment at different temperatures in H and CO
2
(
H
2
/CO=1) at mbar pressure range. (a) Cu 2p, (b) Cu LMM, (c) Fe 2p and (d) C 1s.
After in situ reduction with H /CO = 1 at 100 and Edge Structure (XANES) and Extended X-ray Absorption
50℃, there are no measurable carbon species presented Fine Structure (EXAFS) studies for 3DOM Cu Fe catalyst at
on the catalyst surface (Figure 3d). Upon heating to 200℃,
the iron oxide after O pretreatment was partially
carburized into iron carbide, as indicated by the appear-
2
1
2
1
Cu and Fe K-edge with H
2
/CO = 1 (Figure 4). The fitted
3
3
2
k ꢁ(k) (k-space plot), Fourier transforms of [k ꢁ(k)] (R-
–3
space plot), and Re[ꢁ(q)] (Å ) (q-space plot) spectra for
the Cu K-edge and Fe K-edge signals of 3DOM Cu Fe cata-
lyst at RT and 300℃ are illustrated in Figures S10–S11,
respectively, and the fitted parameters are summarized in
Table S6.
2
7
ance of a peak at 283.5 eV in C 1s spectra. The shift of the
binding energy of Fe 2p3/2 peak from 710.7 to 707.3 eV
2
1
(
Figure 3c) further corroborates that the formed iron car-
1
4,28
bide is χ-Fe
5
C
2
,
which is consitent with Mössbauer
spectra analysis (Figure 2b). Upon heating to 250℃, the
The XANES spectra of 3DOM Cu
2
Fe
1
catalyst
carbide peak is associated with a shoulder at 284.6 eV
(CO+H @RT) at Cu K-edge (8979 eV) (Figure 4a) illustrate
2
(
Figure 3d), which is assigned to the occurrence of generic
that there are two characteristic peaks for CuO. The first
one at 8978 eV originated from the weak quadrupole-
allowed Cu 1s-to-3d transition.
shoulder peak at 8986 eV, could be ascribed from the di-
pole-allowed 1s-to-4p “shakedown” transition because of
the interaction between the metal and ligand.
27
non-oxygenated surface carbon species (Csurf). The evolu-
tion of Cu/(Cu+Fe) atomic ratio under various reaction
conditions (Figure 3) is close to nominal composition of
29–31
The second one, a
0
.67 (Figure S8), indicating no preferential surface en-
richment of Cu or Fe species upon in situ reduction. We
also performed in situ AP-XPS studies for 3DOM Cu Fe
catalyst without O pretreatment (Figure S9): (I) in UHV,
(II) in H /CO = 1 at RT, (III) in H /CO = 1 at 100–500℃, and
(IV) cooling to 280, 260, and 220℃ in H /CO = 1. The cop-
29,31
The fit-
@RT) is pre-
3
2
1
ting FT[k ꢁ(k)] of Cu K-edge EXAFS (CO+H
2
2
sented in Figure S10c. The peak at 1.94 Å for CuO is as-
signed to the first Cu–O coordination shell. Two peaks de-
termined at 2.90 Å and 3.45 Å are associated with the Cu–
2
2
2
3
2,33
per phase remains metallic state (Figure S9a–b) during
Cu coordination shells.
Apart from these peaks, a broad
catalysis at 280, 260, and 220℃. The iron phase keeps χ-
peak observed at 5.74 Å is attributed to the third Cu–Cu
coordination shell. The fitted scattering paths at 1.942 Å
Fe
5
C state (Figure S9c), albeit the existence of residual
2
2
+
2+
carbonaceous surface species (at 284.5 eV) before catalysis
(Cu–O bond) and 2.849 Å (Cu –Cu bond) (CO+H
are characteristic of CuO (Table S6).
2
@RT)
After in situ reduc-
tion (≥250℃), the XANES spectra are similar with Cu foil
spectrum (Figure 4a). The EXAFS spectra (CO+H ≥250℃)
3
2,33
renders the iron carbide peak at 283.5 eV of C 1s spectra
0
(
Figure S9d) indiscernible. Cu and χ-Fe
5
C
2
are thus
elucidated as the binary active sites for CO hydrogenation.
2
3
4,35
In situ X-ray Absorption Spectroscopy Studies. The re-
were fitted with theoretical model of Cu.
FT[k ꢁ(k)] of Cu K-edge EXAFS spectra (CO+H
The fitting
@300℃) is
0
3
vealing of Cu and χ-Fe
5
C
2
binary active sites from in situ
2
AP-XPS are corroborated by in situ X-ray Absorption Near
presented in Figure S10d. The fitted scattering path at
ACS Paragon Plus Environment

Page 5 of 13
ACS Catalysis
2
.543 Å (Table S6) can be assigned to the Cu–Cu bond of
2.6 Å and a broad shoulder at 3.1 Å.36,37 The second peak
consists of several Fe–Fe (Fetet–Fetet, Fetet–Feoct, and Feoct
Feoct) and Fe–O (Fetet–O and Feoct–O) scattering contribu-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
34,35
metallic copper.
–
The XANES spectrum (CO+H
eV) are within the range expected for Fe
Fe has the cubic inverse spinel structure with Fe cati-
ons in octahedral sites and Fe cations in both octahedral
and tetrahedral sites, surrounded by oxygen ions. Both
octahedral and tetrahedral iron atoms contribute to R-
space plot. A theoretical model of Fe
paths was used to fit the EXAFS spectra (CO+H
There are two main features for Fe in R-space plot (Fig-
ure 4d), a peak centered at 1.5 Å consisting of nearest-
neighbor oxygen coordination shell contributions (Fetet–O
and Feoct–O), and a second peak with maximum intensity at
2
@RT) at Fe K-edge (7112
3
6,37
3
O
4
(Figure 4c).
tions.
The fitted scattering path at 1.932 Å (Feoct–O
2
+
3
O
4
bond) and 3.319 Å (Fetet–Fetet bond) (Figure S11c, Table S6)
3
+
are characteristic of Fe
3
O
4
for 3DOM Cu
2
Fe
1
catalyst
3
6
36,37
(CO+H
2
@RT).
After in situ reduction (≥300℃), the
spectrum (Figure
≥300℃) were fitted with
XANES spectra are similar with χ-Fe
4c). The EXAFS spectra (CO+H
C
5 2
3
O
4
with five scattering
2
3
8
2
≤250℃).
theoretical model of χ-Fe
lyst (CO+H @300℃) as an example (Figure S11d), the fit-
ted scattering paths at 1.645 Å (Fe–C bond) and 2.578 Å
5
C
2
. Taking 3DOM Cu
2
Fe
1
cata-
3
O
4
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
8–40
(Fe–Fe bond) (Table S6) attribute to χ-Fe
5
C
2
.
c
Fe K-edge
a
b Cu K-edge
Fe K-edge
d
Cu K-edge
3DOM Cu2Fe
1
3DOM Cu2Fe1
Cu foil
Fe foil
Cu foil
400
50
00
250
°
C
χ-Fe5C2
Cu
2
O
3
3
°
C
C
C
400
°C
Cu2O
Fe foil
-Fe
°
CuO
Fe
3
O
4
350
°
C
CuO
°
300
250
200
°
°
°
C
C
C
χ
5
C
2
2
00
150
100
0
RT
°
C
CO + H
2
CO + H
CO + H
CO + H
CO + H
2
CO + H
CO + H
CO + H
2
400°C
350°C
300°C
Fe
400
350
300
250
200
150
3
O
4
2
°
°
C
C
CO + H
CO + H
CO + H
CO + H
2
2
2
2
2
150
°
C
CO + H
CO + H
CO + H
2
°
°
°
C
C
C
2
5
°
C
2
100°
C
2
2
2
50
RT
°
C
CO + H
CO + H
CO + H
CO + H
CO + H
CO + H
2
250°C
CO + H
CO + H
CO + H
CO + H
CO + H
2
CO + H
CO + H
CO + H
CO + H
CO + H
CO + H
2
°
C
CO + H
CO + H
2
2
200°C
150°C
100°C
50°C
RT
2
°
C
2
2
2
2
2
°
C
2
2
2
2
100
50
RT
°
C
CO + H
CO + H
2
2
2
3DOM Cu
2
Fe
1
3DOM Cu
2
Fe1
2
°
C
2
2
7050
7100
7150
7200
7250
0
1
2
3
4
5
6
7
8
9 10
8950
9000
9050
0
1
2
3
4
5
6
7
8
9 10
Photon energy (eV)
R (Å)
Photon energy (eV)
R (Å)
Figure 4. In situ X-ray absorption spectroscopy studies for 3DOM Cu
K-edge XANES, (b) Fourier transform magnitude of k -weighted of Cu K-edge EXAFS spectra, (c) Fe K-edge XANES, (d) Fourier
transform magnitude of k -weighted of Fe K-edge EXAFS spectra.
2
Fe
1
catalyst at different temperatures with H /CO = 1. (a) Cu
2
3
3
0
The Cu –χ-Fe
alcohols were mainly produced on Cu –χ-Fe
catalysts based on the GC–MS results (Figure S5). We envi-
sioned the formation of higher alcohols via the hydro-
formylation of the olefin intermediates from FTS reac-
tion, because otherwise, it would produce branch-alcohols.
Additionally, the alcohol and hydrocarbon distributions on
Cu –χ-Fe
follow ASF distributions. The FTS reaction on χ-Fe
catalysts follow ASF distribution as well (Figure S6). The
similar ASF distribution of 1-alcohols and hydrocarbons
found on Cu –χ-Fe
chain propagation may occur primarily on χ-Fe
ing the active site for which CO dissociates and hydrogen-
ates to alkyl species, thereby initiating the chain growth
through CH
lyst was well reported for its selective synthesis of linear
hydrocarbons in FTS reaction,
the present finding of linear 1-alcohols formed on Cu –χ-
Fe
5
C
2
Synergistic Interactions. Linear 1-
ure 5) based on the HAADF-STEM result (Figure 2i) in
comparison to χ-Fe (510) surface model (Figure S12).
Reaction Pathway on χ-Fe (510). The χ-Fe (510)
0
5
C
2
binary
C
5 2
5
C
2
C
5 2
surface morphology and its adsorption sites is illustrated
in Figure S12. The most stable adsorption configuration of
reactants, products, and possible intermediates involved in
6
,7
CO hydrogenation to C
2
-hydrocarbon and C
2
-oxygenates on
0
5
C
2
catalysts (Figure S6) are found similar and
χ-Fe (510) are shown in Figure S13, and the correspond-
5 2
C
7,8
5
C
2
ing adsorption energy and key geometrical parameters are
listed in Table S7. The routes of CO activation, and the for-
mation of CH
oxygenates on χ-Fe
S21. All possible elementary reactions involved in CO hy-
drogenation to C -hydrocarbons and C -oxygenates to-
gether with activation barriers and reaction energies on χ-
Fe (510) are listed in Table S8. The reaction pathway for
CO hydrogenation on χ-Fe (510) is shown in Figure S22.
On χ-Fe (510), CO direct dissociation and CO hydro-
x
(x = 1–3), CH
3
OH, C
2
-hydrocarbons, and C
2
-
0
5
C
2
binary catalysts implies that the
, provid-
C
5 2
(510) are presented in Figures S14–
C
5 2
2
2
+ CH
x
(x = 1–3) coupling.11 The χ-Fe
5
C
2
cata-
C
5 2
x
C
5 2
8
,12,14,22
which corroborates
C
5 2
0
genation to CHO are more favorable than CO hydrogena-
tion to COH (Figure S14). Meanwhile, CO direct dissocia-
tion and CO hydrogenation to CHO are kinetically competi-
tive due to their close activation barriers. C and CHO are
thus the major intermediates of CO activation. Starting
from the initiate states of C+H, CHO, and CHO+H, C hydro-
genation to CH is more favorable than the other pathways
(Figure S15). This suggests that C hydrogenation mainly
contributes to CH formation, which is kinetically competi-
0
0
5
C
2
catalysts. Thus, the essence of Cu in Cu –χ-Fe
5
C cat-
2
alysts could be rationalized by providing the active sites
for CO associative adsorption, in combination with the lin-
ear hydrocarbon intermediates formed on the χ-Fe
at the interface of Cu –χ-Fe
ates, which are hydrogenated to selectively produce linear
5
C part
2
0
5
C to generate linear oxygen-
2
1
1
0
1
-alcohols. To shed further light on the Cu –χ-Fe
5
2
C syn-
ergistic interactions for selective CO hydrogenation to
higher alcohols, we resorted to density functional theory
tive with CH
of CH+H, CHO+H, and CH
2
O formation. Starting from the initiate states
O, both CH hydrogenation and
0
(DFT) calculation on Cu –χ-Fe
5
C
2
(510) surface model (Fig-
2
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 13
CH
to CH
is more favorable than CH
ing from the initial states of CH
hydrogenation mainly results in CH
more favorable than CH OH formation (Figure S17). CH is
thus the most favorable CH (x = 1–3) species on χ-
Fe (510) (Figure S18). Additionally, CH formation is
more favorable than CH OH formation, suggesting that χ-
Fe (510) exhibits higher selectivity to CH species instead
of CH OH (Figure S18 and Figure S22).
On χ-Fe (510), CH prefers coupling to C
than dissociation to C, hydrogenation to CH , or being in-
serted by CO/CHO to C -oxygenates (Figure S19). CH pre-
fers dissociation to CH rather than hydrogenation to CH
coupling to C , or being inserted by CO/CHO to C
oxygenates (Figure S20). CH prefers dissociation to CH
instead of hydrogenation to CH , coupling to C
inserted by CO/CHO to C -oxygenates (Figure S21). Once
CH and CH are formed on χ-Fe (510), both would dis-
sociate to CH, further confirming that CH is the most favor-
able CH (x = 1–3) species. Our calculation results are in
excellent agreement with the experimental results
2
O dissociation with H-assisted significantly contribute
formation (Figure S16). Furthermore, CH formation
O or CH OH (Figure S16). Start-
+H, CH O+H, and CH O, CH
formation, which is
ically difficult owing to their higher activation barriers
than CO insertion into CH (x = 1–3) to CH CO (x = 1–3)
(Figures S19–S21). Therefore, on Cu –χ-Fe (510), we
(x = 1–3) species
prefers
are
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
x
x
0
3
2
C
5 2
2
2
3
2
will only consider CO insertion into CH
x
0
3
to C
dissociation to CH
found to be the most favorable CH
2
-oxygenates. On Cu –χ-Fe
5
C
2
(510) surface, CH
3
3
2
(Figure S26). Thus, CH and CH
2
0
x
x
species on Cu –χ-
(510). The adsorbed CO at the interface of Cu –χ-
0
5
C
2
Fe
Fe
5
C
C
2
3
5
2
(510) inserts into CH
(510) surface to
CO) with the activation barriers of 64.8 and
x
(x = 1, 2) that adsorbed at the
5
C
2
Fe sites of χ-Fe
(CHCO/CH
5
C
2
C
2
-oxygenates
3
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–1
5
C
2
2
H
2
rather
32.5 kJ·mol , respectively (Figure 6a). Whereas on χ-
Fe (510), the activation barriers are 144.5 and 113.3
kJ·mol , respectively, which are much higher by 79.7 and
2
C
5 2
–1
2
2
,
80.8 kJ·mol– than those on Cu –χ-Fe
1
0
5
2
C (510) (Figure 6a).
3
2
H
4
2
-
Additionally, the previous DFT calculation reported by
4
5
3
2
Zhao et al. on Rh(111) for CO hydrogenation to C
2
-
4
2
H
6
, or being
oxygenates indicated that CO insertion into CH (x = 1, 2)
x
–
1
2
had activation barriers of 129.3 and 120.6 kJ·mol (Table
S9), respectively, which are much higher by 64.5 and 88.1
2
3
C
5 2
–1
0
kJ·mol than those on Cu –χ-Fe
5
C
2
(510). Thus, the smaller
(x =
x
activation barriers required for CO insertion into CH
1, 2) to C
x
14,22
0
and
exhibits good catalytic
2
-oxygenates (CHCO/CH CO) suggest that Cu –χ-
2
41–43
DFT calculations
performance for CO dissociation and carbon-chain prop-
agation for Fe-based FTS catalysts. Thus, CH is the most
that χ-Fe
5
C
2
12
Fe (510) is indeed favorable for higher alcohols for-
5 2
C
mation compared with χ-Fe
5
2
C (510) and Rh(111).
favorable species on χ-Fe
5
C
2
(510), and χ-Fe
5
C
2
(510) main-
ly contributes to the formation of C
2
H
2
via CH coupling
(Figure S22, Supplementary Movie 1).
0
The Cu –χ-Fe
5
C
2
(510) Surface Model. To further eluci-
0
date the synergistic interactions of Cu –χ-Fe
5
C binary cat-
2
alyst for selective CO hydrogenation to higher alcohols, we
0
resorted to DFT calculation on a Cu –χ-Fe
5
2
C (510) surface
model (Figure 5) based on HAADF-STEM result (Figure 2i).
We construct a Cu strip adsorbed on a p(2×1) χ-Fe
5
C (510)
2
0
surface to optimize and represent Cu –χ-Fe
5
2
C binary cata-
0
lyst. The Cu –χ-Fe
5
2
C (510) surface morphology and its
adsorption sites are also shown in Figure 5.
0
The Geometric Effect of Cu –χ-Fe
5
2
C (510). The geomet-
ric effect plays a crucial role in catalytic performance, as it
associates with the atomic arrangement at the active
2
,3,44
site.
The way the active site is configured can exert a
3,44
significant effect on the binding strength of adsorbates.
The calculated geometric results show that only CO prefers
adsorption at the interface of Cu –χ-Fe
0
5
C
2
(510), while CH
x
(
x = 1−3) and H species prefer adsorption at the top Fe
0
sites on Cu –χ-Fe
5
C
2
(510) surface (Figure S23). Moreover,
as shown in Table S9, the calculated adsorption energies of
0
Figure 5. The Cu –χ-Fe
5
C
2
(510) surface model. A Cu strip
(510) surface and its adsorption sites
with top and side view. T1–T4 denote the top Fe site of χ-
Fe (510) surface, and T 5–T 10 denote the top Fe or Cu sites
of Fe–Cu interface; B1–B6 and B7–B12 denote the Fe–Fe
bridge and Cu–Cu bridge sites, respectively; B 1–B 3 denote
the Fe–Cu bridge sites of Fe–Cu interface; F1–F4 denote the 3-
fold Fe sites of χ-Fe (510) surface, F5 and F6, F7 denote the
4-fold and 3-fold Cu sites over Cu strip, respectively; F 1 and
2 denote the 3-fold sites of Fe–Cu interface. The interface
CH
x
(x = 1−3) and H species decrease and the calculated CO
adsorbed on χ-Fe
C
5 2
0
adsorption energy increases on Cu –χ-Fe
5
C (510) in com-
2
parison to those on χ-Fe
reactions related to CH (x = 1–3) (Figures S24–S26), the
reactions of their dissociation, hydrogenation, and cou-
5
C
2
(510). Additionally, among all
5
C
2
I
I
x
I
I
pling to C
on Cu –χ-Fe
2
-hydrocarbons prefer occurring at the Fe sites
0
5
C
2
2
(510). However, only CO insertion into CH
-oxygenates occur at the interface of Cu –χ-
x
C
5 2
0
(
x = 1–3) to C
Fe (510) (Figures S24–S26). Thus, χ-Fe
mainly contributes to C -hydrocarbons formation, and the
geometric effect at the interface of Cu –χ-Fe
an important role in C -oxygenates formation.
On χ-Fe (510), it is noteworthy that CHO insertion
into CH (x = 1–3) to CH CHO (x = 1–3) are all more kinet-
I
5
C
2
5
C
2
(510) alone
F
I
0
between Cu and χ-Fe
5
2
C (510) is circled by the red dotted
2
0
lines. Purple, gray and orange balls denote Fe, C, and Cu at-
oms, respectively.
5
2
C (510) plays
2
C
5 2
The potential energy profile of CH
x
+ CH
x
(x = 1, 2)
x
x
coupling and CO insertion into CH (x = 1, 2) reactions on
x
ACS Paragon Plus Environment

Page 7 of 13
ACS Catalysis
(510) is pre-
0
Cu –χ-Fe
sented in Figure 6a. The activation barriers for C
hydrocarbons (C /C and -oxygenates
(CHCO/CH CO) formation over Cu –χ-Fe (510) are re-
markably lower than those on χ-Fe
CHCO/CH
5
C
2
(510) in comparison to χ-Fe
5
C
2
C
2
-hydrocarbons and C
Fe (510). The activation barrier difference between
(C ) and CHCO(CH CO) formation decreases to
2
-oxygenates in comparison to χ-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
-
C
5 2
2
H
2
2
H
4
)
C
2
C
2
H
2
2
H
4
2
0
–1
0
2
5
C
2
44.7(18.5) kJ·mol on Cu –χ-Fe
5
C (510) from 108.6(65.1)
(510), indicating that the selectivity of
2
kJ·mol– on χ-Fe
1
C
5
C
2
(510), especially for
5
2
0
0
2
CO formation, suggesting that Cu –χ-Fe
5
C
2
(510)
C
2
-oxygenates on Cu –χ-Fe
5
C
2
(510) are higher than that on
for CH CO formation.
exhibits a higher catalytic activity towards the formation of
χ-Fe
5
C
2
(510),
especially
2
a 1
50
TS 144.5
TS 113.3
100
CHCO
20.5
108.6
1
65.1
50
CH
61.7
2CO
TS 48.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH+CO
TS 35.9
CH2+CO
0
CH+CH 0.0
C2H
2
CH2+CH2 0.0
−
1.0
-50
C2H4
−
31.9
-100
χ
-Fe5
C
2
(510)
χ-Fe5C2(510)
1
50
00
1
TS 64.8
44.7
5
0
TS 32.5
TS 14.0
CHCO
9.4
CH
2CO
CH+CO
CH
2
+CO
18.5
7.8
0
TS 20.1
CH+CH 0.0
CH2
+CH
2 0.0
-
50
C
2
H2
C2H4
−97.9
0
−96.1
0
-100
Cu −χ-Fe
5
C
2(510)
Cu −χ-Fe
5
C
2
(510)
Reaction coordinate
b
0
−
2.02
Cu −χ-Fe
5
C
2
(510)
χ
-Fe
5
C
2
(
510
)
−2.11
Rh(111)
14 -12
−
2.40
-
-10
-8
-6
-4
-2
0
2
E−
Ef (eV)
Figure 6. (a) The potential energy profiles of CH
x
+ CH
x
(x = 1, 2) coupling and CO insertion into CH
x
(x = 1, 2) on χ-Fe
5
C
2
(510) and
0
0
Cu –χ-Fe
5
C
2
(510), respectively. (b) Projected densities of states for the surface d-orbital center of Cu-Fe, Fe, and Rh over Cu –χ-
Fe
5
C
2
(510), χ-Fe
5
C (510), and Rh(111), respectively, where the vertical black dashed lines denote the d-band center and the verti-
2
cal blue dashed line is the Fermi level. It is noted that for calculating the d-band center, surface Fe and Cu atoms are selected over
0
Cu –χ-Fe
5
C
2
(510), surface Fe atoms and Rh atoms are selected over χ-Fe
5
C
2
(510) and Rh(111) surface, respectively. (c) Top view
(510) surface. The blue and yellow shaded regions
0
0
and (d) side view of the charge density difference of Cu atoms for Cu –χ-Fe
denote electronic charge depletion and charge accumulation, respectively. Purple, gray, and orange balls denote Fe, C, and Cu at-
oms, respectively.
C
5 2
0
The Electronic Effect of Cu –χ-Fe
ric structure of the catalytic active site is intrinsically
linked to its electronic structure. To gain a deep insight
into the electronic effect of Cu –χ-Fe
ed the calculation of projected densities of states (PDOS),
d-band center (ꢂ : the average energy of the d-band),
charge density difference, and Bader charge. The ꢂ char-
acterizes the ability to eject an electron to the adsorbate
5
C
2
(510). The geomet-
surface (Figure 6b). For CO insertion into CH
oxygenates, a C−C bond is formed when the CH
migrate to CO, the doubly occupied 5σ CO orbital interacts
with the doubly occupied σ-CH orbitals to generate doubly
occupied bonding and anti-bonding orbitals, thereby giv-
x
to C
2
-
x
species
3
0
5
C
2
(510), we conduct-
x
4
7
0
ꢃ
ing the repulsion. The upward shift of ꢂ on Cu –χ-
ꢃ
ꢃ
Fe
states, which could accept more electrons from CO and CH
fragments orbitals than that on χ-Fe (510) surface,
thereby reducing the repulsion and facilitating CO inser-
5
C
2
(510) surface (Figure 6b) empties more anti-bonding
x
46
from the d-band of the metal. The electronic effect on the
C
5 2
binding strength of adsorbate attributes to the change in
the electronic structure of a catalyst. For transition metals,
the way their d-band interacts with the adsorbate deter-
mines the binding strength.3 The trend is that the more
low-lying (relative to the Fermi level) the d-band, the
weaker the binding owing to the occupancy of anti-
3,47
tion into CH
x
.
In addition, the downward shift of ꢂ (0.38
ꢃ
eV) on Rh(111) surface (Figure 6b) can accept less elec-
,44
trons from CO and CH
x
fragments orbitals than that on
0
Cu –χ-Fe
5
C (510) surface, thereby increasing the repulsion
2
3
,47
and retarding CO insertion into CH
insertion into CH (x = 1, 2) to C -oxygenates on Rh(111)
x
,
in line with that CO
3
,44
bonding states.
The PDOS for the ꢂ of metallic Cu-Fe, Fe, and Rh on
x
2
4
5
ꢃ
surface has higher activation barriers (Table S9) than
those on Cu –χ-Fe
0
0
Cu –χ-Fe
5
C
2
(510), χ-Fe
5
C
2
(510), and Rh(111) surfaces are
5
2
C (510) surface.
0
shown in Figure 6b. The ꢂ of Cu –χ-Fe
5
C
2
(510) surface is
(510)
The calculated density of states shows that χ-Fe
metallic in nature (Figure S27), where Fe is a cation with a
5
C
2
is
ꢃ
0
.09 eV closer to the Fermi level than that of χ-Fe
5 2
C
ACS Paragon Plus Environment

ACS Catalysis
Bader charge of 0.40 e, and C is anion with Bader charge of
Page 8 of 13
0
Reaction Pathway on Cu –χ-Fe
of C -oxygenates is preferred by CO insertion into CH
1, 2) on Cu –χ-Fe
ly, C -oxygenates undergo a stepwise hydrogenation to
5
C
2
(510). The formation
(x =
(510) (Figures S24–S26), subsequent-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
–
0.99 e. The plotted charge density difference of Cu atoms
2
x
0
0
for Cu –χ-Fe
5
C
2
(510) surface with top view and side view
5 2
C
in Figure 6c–d indicates that the electron-rich interface of
2
0
Cu –χ-Fe
5
C
2
0
(510) is due to the delocalized electron trans-
C
2
H
5
OH (Figure 7a). The proposed reaction pathway for CO
hydrogenation to C
0
fer from Cu atoms, which facilitates CO activation and CO
2
H
5
OH over Cu –χ-Fe
5
C (510) is shown
2
insertion into CH
x
(x = 1–3) to C
2
-oxygenates at the inter-
in Figure 7a and Supplementary Movie 2. The potential
energy profile with calculated structural parameters of
initial states (ISs) and final states (FSs), and transition
states (TSs) in the reaction pathway are displayed in Fig-
ure 7a–b. The catalytic elementary steps, activation barri-
ers and reaction energies are listed in Table S10.
0
48,49
face of Cu –χ-Fe
5
C
2
(510).
Indeed, the delocalized elec-
0
0
tron transfer from Cu atoms to the interface of Cu –χ-
Fe
5
C (510) is also confirmed by calculating Bader charge
2
0
0
analysis of Cu atoms for Cu –χ-Fe
charge of 0.08 e. Therefore, the synergistic geometric and
5
2
C (510) with a positive
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
9
0
electronic effects of Cu –χ-Fe
the increased activity and selectivity towards
oxygenates in comparison to χ-Fe and Rh catalyst.
The Quantitative Energy Descriptor for Catalytic Activ-
ity and Selectivity. To quantify the activity and selectivity
5
C
2
binary catalyst account for
Starting from CHCO intermediate (Figure 7a), one re-
action pathway is CHCO hydrogenation [CHCO+H(1)] to
2
C -
5
C
2
CH
ergy of 37.5 and 24.1 kJ·mol , respectively; in TS1, CHCO
is adsorbed at the 3-fold F 2 site of Fe–Cu interface, and H
2
CO via TS1 with an activation barrier and reaction en-
–
1
I
for CO hydrogenation to
oxygenates, we calculated the effective barriers
C
2
-hydrocarbons and
C
2
-
atom is adsorbed at the top T3 Fe site with the C–H dis-
tance of 1.693 Å. The other is CHCO hydrogenation
[CHCO+H(2)] to CHCHO via TS2 with an activation barrier
(
ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢈꢉ and ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢋ) to evaluate the reaction rate
ꢌꢇꢈꢉꢊꢇꢈꢉ and ꢌꢇꢈꢉꢊꢇꢋ) of CH + CH (x = 1, 2) coupling and
(x = 1, 2). Additionally, we calculated
the effective barrier difference (ΔEeff = ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢈꢉ
ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢋ) between CH + CH (x = 1, 2) coupling and CO
insertion into CH (x = 1, 2) to describe the selectivity be-
tween C -hydrocarbons and C -oxygenates. In principle,
the higher ΔEeff represents the higher selectivity of C
oxygenates and the lower selectivity of C -hydrocarbons.
Furthermore, the coverage of C species CH
ꢍꢇꢈꢉ) and CO (ꢍꢇꢋ) is considered.
–1
(
x
x
and reaction energy of 74.6 and 5.9 kJ·mol , respectively.
Thus, CHCO prefers hydrogenation to CH
Starting from CH CO, one reaction pathway is CH
hydrogenation [CH CO+H(1)] to CH CO via TS3 with an
activation barrier and reaction energy of 34.5 and –3.1
kJ·mol , respectively; in TS3, CH
fold F 1 site of Fe–Cu interface, H atom is adsorbed at the
top T4 Fe site, and the C–H distance decreases to 1.476 Å
from 2.585 Å in CH CO+H(1). The second one is CH CO
hydrogenation [CH CO+H(2)] to CH CHO via TS4 with an
activation barrier and reaction energy of 92.9 and –66.1
2
CO.
CO insertion into CH
x
2
2
CO
−
2
3
x
x
x
–
1
2
CO is adsorbed at the 3-
2
2
I
2
-
2
2
2
1
x
(x = 1, 2)
2
2
(
–1
Table 1. The effective barrier (ꢄꢅꢆꢆ) of CH
coupling and CO insertion into CH
barrier difference (∆ꢄꢅꢆꢆ) between CH
insertion into CH (x = 1, 2) on χ-Fe
x
+ CH
x
(x = 1, 2)
kJ·mol , respectively. The third one is CH
tion CH CO+H(3) to CH COH via TS5 with an activation
barrier and reaction energy of 128.1 and 76.1 kJ·mol ,
2
CO hydrogena-
x
(x = 1, 2), and the effective
2
2
x
+ CH
x
coupling and CO
(510) and Cu –χ-
–1
0
x
5
C
2
respectively. Thus, CH
2
CO prefers hydrogenation to CH
CO, one reaction pathway is CH
CHO via TS6 with an
activation barrier and reaction energy of 47.1 and–14.5
3
CO.
–
1
Fe
5
C
2
(510), respectively. (Unit: kJ·mol )
Starting from CH
3
3
CO
CH species
CH
2
species
hydrogenation [CH
3
CO+H(1)] to CH
3
0
0
5
χ-Fe C
2
(510)
Cu –χ-Fe
5
C
2
(510)
5
χ-Fe C
2
(510)
Cu –χ-Fe
5
C
2
(510)
–1
ꢄ
ꢅꢆꢆ,ꢇꢈꢉꢊꢇꢈꢉ
kJ·mol , respectively; in TS6, CH
fold F 1 site of Fe–Cu interface, H atom is adsorbed at the
Fe–Fe bridge site, and the C–H distance decreases to 1.528
Å from 2.188 Å in CH CO+H(1). The other reaction path-
way is CH CO hydrogenation [CH CO+H(2)] to CH COH via
TS7 with an activation barrier and reaction energy of
3
CO is adsorbed at the 3-
48.9
151.0
−102.1
41.2
75.3
241.3
209.8
31.5
227.9
139.4
88.5
I
ꢄ
ꢅꢆꢆ,ꢇꢈꢉꢊꢇꢋ
∆ꢄꢅꢆꢆ
−34.1
3
3
3
3
The effective barriers (ꢄꢅꢆꢆ
coupling to C are 48.9 and 241.3 kJ·mol on χ-
(510) (Table 1), which decrease to 41.2 and 227.9
)
of CH coupling to C
2
H
2
and
–1
CH
Fe
kJ·mol on Cu –χ-Fe
2
2
H
4
–1
1
40.7 and 87.8 kJ·mol , respectively.
As CH CO prefers hydrogenation to CH
subsequently, CH CHO hydrogenation [CH
CH CH O via TS8 has an activation barrier of 40.8 kJ·mol
and a reaction energy of –28.4 kJ·mol , which is more fa-
vorable than CH CHO hydrogenation [CH CHO+H(2)] to
CH CHOH via TS9 with an activation barrier and reaction
energy of 155.8 and 36.2 kJ·mol , respectively; in TS8,
CH CHO is adsorbed at the B9 Cu–Cu bridge site of Fe–Cu
interface, H atom is adsorbed at the top T 5 site of Fe–Cu
interface, and the C–H distance reduces to 1.631 Å from
.859 Å in CH CHO+H(1). Eventually, CH CH O hydrogena-
tion to C OH via TS10 is exothermic by 11.1 kJ·mol
with an activation barrier of 114.2 kJ·mol ; in TS10,
CH CH O is adsorbed at the top T 7 Cu site of Fe–Cu inter-
C
5 2
3
3
CHO via TS6,
CHO+H(1)] to
–1
0
5
C (510), respectively. This indicates
2
3
3
that the formation rate of (C
on Cu –χ-Fe
estingly, the effective barriers of CO insertion into CH
2
H
2
/C
2
H
4
) is slightly improved
(510). Inter-
(x =
–
1
3
2
0
5
C
2
(510) compared with χ-Fe
C
5 2
–1
x
–
3
3
1
1
, 2) remarkably decrease to 75.3 and 139.4 kJ·mol on
3
0
–1
Cu –χ-Fe
5
C
2
(510) from 151.0 and 209.8 kJ·mol on χ-
–1
Fe (510) (Table 1), respectively. This suggests that the
5 2
C
3
formation rate of CH
hanced on Cu –χ-Fe
x
CO (x = 1, 2) can be significantly en-
(510) compared with χ-Fe (510).
I
0
5
C
2
C
5 2
More importantly, the effective barrier differences (∆ꢄꢅꢆꢆ)
between CH + CH (x = 1, 2) coupling and CO insertion into
CH (x = 1, 2) on Cu –χ-Fe C (510) are much higher by 68.0
2
3
3
2
x
x
–
1
2
H
5
0
x
5
2
–
1
–1
and 57.0 kJ·mol than those on χ-Fe
respectively, confirming that Cu –χ-Fe
ly improves the C -oxygenates selectivity.
5
C
5 2
C
2
(510) (Table 1),
(510) significant-
3
2
I
0
face, H atom is adsorbed at the B12 Cu–Cu bridge site of
Fe–Cu interface, and the O–H distance decreases to 1.680 Å
2
from
3.531
Å
in
CH
3
CH O+H.
2
ACS Paragon Plus Environment

Page 9 of 13
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
0
Figure 7. The reaction pathway for CO hydrogenation to C
2
H
5
OH on Cu –χ-Fe
5
C (510) surface proposed from DFT calculation. (a)
2
The reaction pathway and potential energy profile with calculated structural parameters of initial states and final states. (b) Tran-
sition states in the reaction pathway. Bond lengths are in Å. Orange ball denotes Cu atom; Purple and gray balls denote Fe and C
atoms of χ-Fe
5
C
2
(510), respectively; Dark green, red, and white balls denote C, O, and H atoms of adsorbates, respectively.
Catalyst Synthesis. 3DOM Cu-Fe catalysts were prepared by
poly(methyl methacrylate) (PMMA) colloidal crystal template
(CCT) method, using ethylene glycol (EG)–methanol solution of
metal nitrates Cu(NO ·6H O and Fe(NO ·9H O as precursors.
CONCLUSION
21
In summary, we advocate that the synergistic geomet-
ric and electronic interactions determine the activity of
)
3 2
2
)
3 3
2
0
Mono-disperse PMMA microspheres were synthesized using an
emulsion technique. The PMMA was centrifuged to form CCT. The
stoichiometric amounts of mixed metal nitrates were dissolved
with 15 mL of EG by stirring in a beaker (100 mL) at RT for 2 h,
and the mixed solution was poured into a volumetric flask (50
mL). Methanol (6 mL) and EG were added to achieve the solution
with 12 vol.% of methanol. The mixed Cu-Fe precursor was added
to CCT, permeated the voids between the close-packed spheres,
and condensed into a hard and inorganic framework upon frying.
Excessive liquid was removed from the impregnated CCT via a
Buchner funnel connected to vacuum. The infiltered template was
dried in a desiccator at RT overnight. Finally, the dried sample
was mixed with γ-alumina sphere (0.125 inch) and heated in a
quartz tube at 1 ℃ min from RT to 450 ℃ in air flow for 5 h. The
γ-alumina sphere helped the removal of heat produced by oxida-
tive decomposition of PMMA during the calcination, and it was
separated from the catalyst after calcination and discarded.
Measurements of Catalytic Performance. The catalytic tests
were carried out in a half-inch fixed-bed reactor. 1 g catalyst di-
luted with quartz sand was loaded in the catalyst bed. First, the
catalysts were in situ reduced at atmospheric pressure by passing
Cu –χ-Fe
5
2
C for selective CO hydrogenation to higher alco-
hols, outperforming than silica-supported precious Rh-
based catalysts, by using a combination of experimental
evidence from bulk, surface-sensitive, and imaging tech-
niques collected on real and high-performance Cu-Fe bina-
ry catalytic systems coupled with DFT calculations. The
0
closer the d-band center of Cu –χ-Fe
5
C
2
(510) surface to the
Fermi level than those of χ-Fe
C
5 2
(510) and Rh(111)
surface (0.09 and 0.38 eV, respectively), and the elecron-
0
rich interface of Cu –χ-Fe
5
0
C (510) owing to the delocalized
2
electron transfer from Cu atoms, which thereby promote
CO activation and CO insertion into alkyl species to C
2
-
–
1
0
oxygenates at the interface of Cu –χ-Fe
5
2
C (510) and thus
improve C
2
5
H OH selectivity. Starting from CHCO interme-
diate, the proposed reaction pathway for CO hydrogena-
0
tion
CHCO+(H)→CH
O+(H)→C OH. This study may pave a way for the ra-
to
C
2
H
5
OH
on
3
Cu –χ-Fe
5
C
2
(510)
is
2
CO+(H)→CH
CO+(H)→CH
3
CHO+(H)→CH
3
C
H
2
2
H
5
tional design of high-performance binary catalysts made
from earth-abundant metals with synergistic interaction
for tuning selectivity.
with H
ture was increased to 300 ℃ and maintained for 48 h. Second, the
temperature was lowered to 25 ℃ with syngas (H /CO =1, 6% N
as internal standard) and the reactor pressure was slowly in-
2
/CO = 1.0. During in situ reduction process, the tempera-
2
2
EXPERIMENTAL PROCEDURES
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 13
creased to 700 psig. Then, the temperature in the catalyst bed was
between the external wall of the catalytic reactor and the internal
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
increased from 25 ℃ to the target temperature (e.g. 260 ℃). Once
the target temperature was achieved, the reaction was proceeded
for a period of 10–15 h to ensure steady state of the catalyst activ-
ity. The time-on-stream after achieving steady state is 120 h.
The outlet gaseous products were analyzed on-line using
Agilent 7890 gas chromatograph (GC) provided with two thermal
conductivity detectors (TCD) and a flame ionization detector
wall of the UHV chamber. The gaseous side was the internal wall
of the reactor. Au thin film (0.4 mm thick, 99.99%, VWR) was
used as a substrate to load a catalyst. Au foil was deliberately
roughened using a SiC knife to increase adhesion. A certain
amount of 3DOM Cu
2
Fe catalyst was suspended in 100% ethanol
1
and deposited on pre-cleaned Au foil. Ethanol left in the sample
on the Au foil was vaporized by placing Au foil in a vacuum oven
(FID). C
Plot capillary column (50 mꢎꢀꢎ530 μm ID) with a FID and N
rier. CO, CO , and N were analyzed using molecular sieve-packed
columns with TCD and He carrier. H was analyzed using molecu-
lar sieve-packed columns with TCD and N carrier. Liquid prod-
ucts were collected using a cold trap kept at –5 ℃. Alcohols and
hydrocarbons were analyzed off-line using a GC coupled to a mass
spectrometer (GC–MS) equipped with an Agilent 6890 Series gas
chromatograph system, an Agilent 7683B Series Injector, a 5973
Mass Selective Detector, and a FID detector. An Agilent DB-
WAXetr (50 mꢀ0.32 mm I.D., 1.0μm) capillary column was used
for analyzing aqueous product that included oxygenated com-
pounds and water, and water was quantified using an external
standard method. An Agilent HP-5 capillary column was used for
analyzing hydrocarbons.
1
–C
4
ranged alkanes and alkenes were analyzed using a HP
at 60℃. The in situ reaction medium was syngas (H /CO = 1). The
2
2
car-
reaction pressure was 1 mbar. The temperature varied from RT to
500℃, and each target temperature was held for 1 h before spec-
trum collection. High resolution spectra of Cu 2p, Cu (LMM), Fe
2p, and C 1s were gathered using an average of 5–35 scans with a
pass energy of 23.5 eV and a step size of 0.05 eV. All spectra were
calibrated to Au 4f7/2 binding energy (84.0 eV).
In situ X-ray Absorption Spectroscopy. In situ XAS measure-
ments were performed at beamline 9-BM-C of the Advanced Pho-
ton Sources at Argonne National Laboratory. All Cu K-edge and Fe
K-edge measurements were carried out in transmission mode. A
copper or iron foil spectrum was acquired through a third ion
chamber simultaneously with each measurement for energy cali-
bration. Harmonic rejection was accomplished using a Rhodium-
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
coated harmonic rejection mirror. The CuO, Cu
2
O, Fe
3 4
O , and ꢁ-
CO conversion (%) was defined as: COꢎꢎconversionꢎꢎꢏ%ꢎꢐ =
Fe standards, and catalyst samples were diluted with boron
5 2
C
ꢑ
ꢒꢓ
ꢔꢕꢖ,ꢗꢘꢙꢑꢚꢛꢜ
ꢔ
ꢕꢖ,ꢝꢞꢟ
–1
nitride (BN) then pressed into a pellet, which could simultaneous-
ly hold six samples. The sample thickness was chosen to give a
total absorbance (μx) at Cu K-edge or Fe K-edge between 1 and 2
absorption lengths, and edge steps (∆μx) around 0.3–0.5. For in
situ measurements, the sample holder was placed into a con-
trolled atmosphere quartz tube, equipped with thermocouple to
monitor the temperature of samples, Swagelok Ultra-Torr fitting
and Kapton windows for sealing, and shut-off valves for gas
treatments and isolation of the samples after in situ treatment.
The quartz tube was placed within a tube furnace controlled by a
programmable temperature controller. The sample was heated
ꢀ
100, where ꢠꢡꢢ and ꢠꢣꢤꢥ (mol h ) were the
ꢑ
ꢒꢓ ꢕꢖ,ꢗꢘ
ꢔ
total molar flow rates of the reactor inlet syngas and outlet gas,
respectively, ꢦꢏꢧꢐ,ꢧꢨ and ꢦꢏꢧꢐ,ꢩꢪꢫ are the molar fraction of compo-
nent i in the reactor inlet syngas and outlet gas, respectively.
Product selectivity was defined as: Selectivityꢎꢏmol%ꢎꢐ =
ꢨ
ꢬ
ꢑ
ꢝꢞꢟ
ꢔ
ꢬ
ꢀ
100, where ꢭ represents the number of
ꢮ
ꢑ
ꢒꢓ
ꢔ
ꢕꢖ,ꢗꢘꢙꢑꢚꢛꢜ
ꢔ
ꢕꢖ,ꢝꢞꢟ
ꢎ
carbon atoms contained in product j, and ꢦ represents the molar
ꢮ
fraction of product j. The formation rate of C2+OH was defined as:
ꢇ
ꢯꢰꢋꢈꢎꢏꢹꢹꢣꢺꢐ
ꢌꢇꢯꢰ ꢱꢲꢲꢳꢴꢎgꢙꢵ hꢙꢵꢸ = ꢻ
.
ꢋꢈ
ꢶꢷꢫ.
ꢅꢡꢼꢽꢥꢎꢣꢆꢎꢇꢾꢥꢾꢺꢿꣀꢥꢎꢏꢼꢐꢀꣁꢡꢹꢅꢎꢏꢽꢐ
–
1
–1
from RT to 400 ℃ at a rate of 3℃ min under a 10 mL min flow
Mössbauer Spectroscopy. Mössbauer experiments were con-
57
of 3%H /3%CO balanced with helium, and held at each desired
2
ducted using a Co/Rh source in a constant acceleration trans-
temperature for 1 h except 300℃ for 24 h. Data were acquired at
RT in helium flow. Trace oxidants in helium were removed by
passing through a Matheson PUR0 Gas Triple Purifier Cartridge.
The IFEFFIT package was used for data normalization and pro-
mission spectrometer. The spectra were recorded at 27 ℃. The
spectrometer was calibrated using a standard α-Fe foil and the
reported isomer shifts were relative to the center of the α-Fe
spectrum. The WinNormos-for-Igor 3.0 program was used to
determine Mössbauer parameters. A nonlinear least-squares fit-
ting procedure with a set of independent Lorentzian lines that
models the spectra as a combination of singlets, quadruple dou-
blets and/or magnetic sextets was used for data analysis. The
spectra components were identified based on their isomer shift,
quadruple splitting, and magnetic hyperfine field. Magnetic hy-
perfine fields were calibrated with 330 kOe field of α-Fe at 27 ℃.
Cs-corrected HAADF-STEM Imaging and EELS Elemental
Mapping. HAADF-STEM images were taken by a JEOL ARM 200
equipped with a probe corrector and a cold field emission gun
operated at 200 eV with a spatial resolution of 0.08 nm, and EELS
elemental maps were taken by using Gatan Quantum 965 with an
energy resolution of 0.6 eV without a monochromator.
50,51
cessing.
The normalized, energy-calibrated XANES spectra
were obtained by standard methods. Standard procedures based
on IFEFFIT were used to process the EXAFS data.
COMPUTATIONAL METHODOLOGY
All calculations were performed by using the Vienna Ab Initio
52,53
Simulation Package (VASP),
tions were expressed by projector-augmented wave (PAW) meth-
in which the electron-ion interac-
54,55
od.
The generalized gradient approximation (GGA) proposed
by Perdew−Burke−Ernzerhof (PBE)
56,57
was used to describe the
exchange-correlation energies and potential. Owing to the mag-
58
netic properties of Fe, all calculations were spin-polarized with
a plane wave cutoff energy of 400 eV. The Brillouin zone was
sampled by a 2×2×2 k-points grid generated via the Monkhorst-
In situ Ambient Pressure X-ray Photoelectron Spectroscopy.
Gas flowed through the reactor and exited through the exit port
and an aperture that interfaces the gaseous environment of the
pre-lens. Flow rate in the reactor was measured using a mass flow
meter installed between each gas source and the entrance of the
flow reaction cell. The flow rate of pure gas was in the range of 3–
59,60
Pack procedure.
The geometry optimization was converged
when the energy differences between two electronic optimization
–5
steps were smaller than 10 eV, and the forces for ions were less
than 0.03 eV/Å. To study the minimum energy reaction pathways,
61,62
the Climbing-Image Nudged Elastic Band method (CI-NEB)
5
mL pure gas per minute. The total pressure of the mixture gas of
was employed to find saddle points between the known reactants
and products, and the transition states were optimized using the
reactor was measured using a capacitance gauge installed at the
entrance. The pressure at the exit was measured using another
capacitance gauge. An average of the pressures at entrance and
exit was defined as the pressure above the catalyst in the catalytic
reactor that was integrated with a monochromatic Al Kα (hυ =
63,64
65
dimer method.
Bader charge analysis and charge density
66
difference analysis was conducted for discussing charge transfer.
The optimized transition state structures were converged when
the forces for all atoms were less than 0.05 eV/Å. In all calculated
energy data, the zero-point energy (ZPE) has been considered.
1
486.7 eV) X-ray source and a different pumping stage.
The catalyst was heated by heating the vacuum side of a sam-
χ-Fe
5
C has a monoclinic structure with C2/c crystallographic
2
67
ple stage using e-beaming heater installed in the vacuum section
symmetry, including 20 Fe and 8 C atoms per unit cell. The cal-
ACS Paragon Plus Environment

Page 11 of 13
ACS Catalysis
culated lattice parameters (a = 11.588 Å, b = 4.579 Å, c = 5.059 Å,
and βꢎ=97.7°) and magnetic moment (1.73 μB) agree well with the
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI:.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
68,69
previous experimental data.
The χ-Fe
5
C (510) is modeled
2
using a p(2ꢀ1) supercell slab with three-layered iron and six-
0
Catalyst characterizations, catalytic performance data,
DFT calculation results, additional experimental methods,
including SI Figures S1–S27 and Tables S1–S10 (PDF)
Supplementary Movies, showing the optimum reaction
pathway from proposed DFT for CO hydrogenation on χ-
layered carbon. For χ-Fe
5
C
2
(510) and Cu –χ-Fe
5
C (510) surface
2
models, the bottom one-layered iron and two-layered carbon are
fixed in their bulk position during all the calculations, whereas the
top two-layered iron, four-layered carbon, and adsorbates are
relaxed. The vacuum spacing between slabs is 10 Å to ensure no
significant interaction between the slabs.
0
Fe C (510) and Cu –χ-Fe C (510) surface (AVI)
5
2
5 2
46
The d-band center (ꢂ ) is calculated via Eq. (1), where ꣂꢃ
ꢃ
represents the density of states projected onto the metal atom’s d-
AUTHOR INFORMATION
Corresponding Author
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
band and E
f
is the Fermi energy.
꣇
꣈
ꣃ
꣄ꣅ
꣆
ꢏ꣄ꢐꢃ꣄
꣉
꣊
ꢂꢃ = ꣃ
(1)
* wangbaojun@tyut.edu.cn
꣇
꣈
ꣅ
꣆
꣊
ꢏ꣄ꢐꢃ꣄
꣉
Based on the DFT calculation results, adsorption energies,
reaction energies, and activation barriers are used to describe the
thermodynamic and kinetic properties of the reaction. The ad-
sorption energy is defined as Eq. (2), where E(adsorbates/slab) is the
total energy of the slab with adsorbates, E(slab) is the total energy
of the slab, and E(adsorbates) is the total energy of free adsorbates.
Therefore, the more negative the Eads, the stronger the adsorption.
* fyu@abe.msstate.edu
ORCID
Yongwu Lu: 0000-0003-0858-8315
Franklin (Tao) Feng: 0000-0002-4916-6509
Fei Yu: 0000-0001-5595-6147
The reaction energy (∆꣋) and activation barrier (E ) are calculat-
a
Author Contributions
⊗Y. L. and R. Z. contributed equally to this work.
ed by Eqs. (3) and (4), respectively, where E(IS), E(FS), and E(TS)
are the energies of the corresponding initial state (IS), final state
(FS), and transition state (TS), respectively.
Notes
E
ads = E(adsorbates/slab) – [E(slab) + E(adsorbates)
]
(2)
(3)
(4)
The authors declare no competing financial interest.
∆
E
꣋ = ꢄꢏFSꢐ − ꢄꢏISꢐ
= E(TS) – E(IS)
a
ACKNOWLEDGMENTS
The coverage of C
1
species CH (i =1, 2) (ꢍꢇꢈꢉ) can be ex-
x
pressed with respect to the C coverage70 (ꢍ ) by Eq. (5), where ꢄ꣌
The experimental work was supported by US Department of
Agriculture under the Award Number of 2012-10008-20302.
The theoretical work was supported by the National Natural
Science Foundation of China (Nos. 21476155, 21276003), the
Natural Science Foundation of Shanxi Province (No.
2014011012-2), the Program for the Top Young Academic
Leaders of Higher Learning Institutions of Shanxi, and the Top
Young Innovative Talents of Shanxi. Work performed at Ar-
gonne National Laboratory and use of the Advanced Photon
Source were supported by the US Department of Energy, Of-
fice of Science, Office of Basic Energy Sciences, through con-
tract No. DE-AC02-06CH11357.
ꢇ
is the relative stability of CH
is the energy difference between adsorbed CH
ꢍ ⁄ꢍ , ꢍ and ꢍ are the coverage of H and free surface site, re-
x
with respect to C on surfaces, which
x
and C + xH; ꣍ is
ꢈ
∗
ꢈ
∗
7
1–75
spectively. Based on the previous DFT calculation studies,
reaction rate of CH + CH (x = 1, 2) coupling and CO insertion into
CH (x = 1, 2) can be derived and expressed as Eqs. (6) and (7),
respectively,
ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢋ = ꢄꢷ
x = 1, 2) coupling and CO insertion into CH
the
x
x
x
ꢉ ꢉ
ꢇꢈ ꢊꢇꢈ
where ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢈꢉ = ꢄꢷ
+ 2ꢄ꣌
,
and
+ CH
(x = 1, 2) reactions,
are the activation barriers of
(x = 1, 2) coupling and CO insertion into CH (x = 1, 2)
reactions, respectively. Combining Eqs. (6) and (7), the ratio of
reaction rate for the CH + CH (x =1, 2) coupling and CO insertion
into CH (x = 1, 2) reactions can be expressed as Eq. (8), where
ꢄꢅꢆꢆ = ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢈꢉ − ꢄꢅꢆꢆ,ꢇꢈꢉꢊꢇꢋ is the effective barrier difference
between CH + CH (x = 1, 2) coupling and CO insertion into CH (x
1, 2) reactions. It is noteworthy that both ꣍ and ꣎ have little
ꢉ
ꢇꢈ ꢊꢇꢋ
+ ꢄ are the effective barriers of CH
x
x
꣌
(
x
ꢇꢈ
ꢉ
ꢊꢇꢈ
ꢉ
ꢉ
ꢇꢈ ꢊꢇꢋ
respectively. ꢄꢷ
CH + CH
and ꢄꢷ
x
x
x
REFERENCES
(1) Shi, J. Chem. Rev., 2013, 113, 2139–2181.
(2) Yu, W.; Porosoff, M. D.; Chen, J. G. Chem. Rev., 2012, 112, 5780–
x
x
x
∆
5
781.
x
x
x
꣎
꣏
(3) Nørskov, J. K.; Bligaard, T.; Rossmeisl, J.; Christensen, C. H. Nat.
Chem., 2009, 1, 37–46.
=
꣏꣐
effect on the selectivity74,75 in comparison to ∆ꢄꢅꢆꢆ with exponen-
tial impact. Thus, ∆ꢄꢅꢆꢆ can be used as a quantitative energy de-
(4) Strasser, P.; Koh, S.; Anniyev, T.; Greeley, J.; More, K.; Yu, C.; Liu,
Z.; Kaya, S.; Nordlund, D.; Ogasawara, H.; Toney, M. F.; Nilsson, A.
Nat. Chem., 2010, 2, 454–460.
scriptor to evaluate the FTS selectivity between CH
) coupling and CO insertion into CH (x = 1, 2) reactions. In prin-
ciple, the higher ∆ꢄꢅꢆꢆ represents the higher selectivity of C
oxygenates and the lower selectivity of C -hydrocarbons.
x
+ CH
x
(x = 1,
2
x
(
5) Prieto, G.; Beijer, S.; Smith, M. L.; He, M.; Au, Y.; Wang, Z.; Bruce,
2
-
D. A.; de Jong, K. P.; Spivey, J. J.; de Jongh, P. E. Angew. Chem. Int.
Ed., 2014, 53, 6397–6401.
2
꣇
ꢉ
꣇ꢉ
ꢍꢇꢈꢉ = ꣑ꢏꢙ꣒꣓ ꢍ ꢏꢍ ⁄ꢍ ꢐ = ꣑ ꣒꣓ ꢍ ꣍
ꢐ
꣌
ꢏꢙ
ꢐ
꣌
(
6) Pei, Y. -P.; Liu, J. -X.; Zhao, Y. -H.; Ding, Y. -J.; Liu, T.; Dong, W. -
(5)
ꢇ
ꢈ
∗
ꢇ
꣏꣗ꢉꢰ꣏꣗ꢉ
꣏꣗ꢉꢰ꣏꣗ꢉ
D.; Zhu, H. -J.; Su, H. -Y.; Yan, L.; Li, J. -L.; Li, W. -X. ACS Catal., 2015,
5, 3620–3624.
꣇
꣇
ꢰꢯ꣇ꢉ
꣖
꣖
꣕
ꢙ
꣘
꣕ꢙ
꣘
꣒꣓
꣙
꣒꣓
꣙
꣙꣌
ꢌꢇꢈꢉꢊꢇꢈꢉ = ꣔꣑
ꢍ
= ꣔꣑
ꢍ ꣍
ꢇ
ꢇꢈ
ꢉ
(
7) Xiang, Y.; Kruse, N. Nat. Commun., 2016, 7, 13058.
꣇
꣚
꣛꣛,꣏꣗ꢉꢰ꣏꣗ꢉ
ꢐ
꣙
꣙꣌
=
꣔꣑ꢏꢙ
꣒꣓
ꢍ ꣍
(6)
(8) Torres Galvis, H. M.; Bitter, J. H.; Khare, C. B.; Ruitenbeek, M.;
Iulian Dugulan, A.; de Jong, K. P. Science 2012, 335, 835–838.
(9) Behrens, M.; Studt, F.; Kasatkin, I.; Kühl, S.; Hävecker, M.;
Abild-Pedersen, F.; Zander, S.; Girgsdies, F.; Kurr, P.; Kniep, B. -L.;
Tovar, M.; Fischer, R. W.; Nørskov, J. K.; Schlögl, R. Science 2012,
336, 893–897.
ꢇ
꣏
꣗ꢉꢰ꣏꣐
꣏꣗ꢉꢰ꣏꣐
꣄
꣄
ꢊ꣄
ꢉ
꣖
꣖
꣜ꢙ
꣟
꣜ꢙ
꣟
꣝
꣞
꣝꣞
ꢍ ꢍ ꣍꣌
ꢇ ꢇꢋ
ꢌꢇꢈꢉꢊꢇꢋ = ꣔꣑
ꢍꢇꢈꢉꢍꢇꢋ = ꣔꣑
꣇
꣚
꣛꣛,꣏꣗ꢉꢰ꣏꣐
= ꣔꣑ꢏꢙ
꣍꣌
ꢐ
ꢍ ꢍ ꣍꣌
(7)
(8)
꣒꣓
ꢇ
ꢇꢋ
∆
꣇
꣚꣛꣛
꣠
꣏꣗ ꢰ꣏꣗
꣡
ꢉ
ꢉ
꣏
꣑ꢏꢙ ꣒꣓
ꢐ
=
꣠
ꢉ
꣏꣗ ꢰ꣏꣐
꣡
꣏꣐
(10) Xiang, Y.; Chitry, V.; Liddicoat, P.; Felfer, P.; Cainey, J.; Ringer,
S.; Kruse, N. J. Am. Chem. Soc., 2013, 135, 7114–7117.
ASSOCIATED CONTENT
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 13
(
11) Luk, H. T.; Mondelli, C.; Ferré, D. C.; Stewart, J. A.; Pérez-
Ramírez, J. Chem. Soc. Rev., 2017, 46, 1358–1426.
12) Wang, D.; Chen, B.; Duan, X.; Chen, D.; Zhou, X. J. Energy Chem.,
016, 25, 911–916.
13) Zhong, L.; Yu, F.; An, Y.; Zhao, Y.; Sun, Y.; Li, Z.; Lin, T.; Lin, Y.;
Qi, X.; Dai, Y.; Gu, L.; Hu, J.; Jin, S.; Shen, Q.; Wang, H. Nature 2016,
38, 84–87.
(40) Zhai, P.; Xu, C.; Gao, R.; Liu, X.; Li, M.; Li, W.; Fu, X.; Jia, C.; Xie,
J.; Zhao, M.; Wang, X.; Li, Y. -W.; Zhang, Q.; Wen, X. -D.; Ma, D. An-
gew. Chem. Int. Ed., 2016, 55, 9902–9907.
(41) Zhao, S.; Liu, X. -W.; Huo, C. -F.; Li, Y. -W.; Wang, J.; Jiao, H. J.
Catal., 2012, 294, 47–53.
(42) Pham, T. H.; Duan, X.; Qian, G.; Zhou, X.; Chen, D. J. Phys. Chem.
C 2014, 118, 10170−10176.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
2
(
5
(14) Yang, C.; Zhao, H.; Hou, Y.; Ma, D. J. Am. Chem. Soc., 2012, 134,
15814–15821.
(43) Olus Ozbek, M.; Niemantsverdriet, J. W. J. Catal., 2014, 317,
158–166.
(44) Kim, D.; Resasco, J.; Asiri, A. M.; Yang, P. Nat. Commun., 2014,
5, 4948.
(45) Zhao, Y. -H.; Sun, K.; Ma, X.; Liu, J.; Sun, D.; Su, H. -Y.; Li, W. -X.
Angew. Chem. Int. Ed., 2011, 50, 5335–5338.
(46) Li, J.; Croiset, E.; Ricardez-Sandoval, L. J. Mol. Catal. A Chem.,
2012, 365, 103–114.
(47) Choi, Y. M.; Liu, P. J. Am. Chem. Soc., 2009, 131, 13054–13061.
(48) van Steen, E.; Claeys, M. Catal. Struct. React., 2015, 1, 11–18.
(49) Tian, X.; Wang, T.; Yang, Y.; Li, Y. -W.; Wang, J.; Jiao, H. J. Phys.
Chem. C 2015, 119, 7371–7385.
(50) Newville, M. J. Synchrotron. Radiat., 2001, 8, 322–324.
(51) Ravel, B.; Newville, M. J. Synchrotron. Radiat., 2005, 12, 537–
541.
(52) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169–
11186.
(53) Kresse, G.; Furthmüller, J. Comput. Mater. Sci., 1996, 6, 15–50.
(54) Blöchl, P. E. Phys. Rev. B 1994, 50, 17953–17979.
(55) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758–1775.
(56) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett., 1996,
77, 3865−3868.
(57) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett., 1997,
78, 1396.
(58) Louie, S. G.; Froyen, S.; Cohen, M. L. Phys. Rev. B 1982, 26,
1738–1742.
(59) Ge, Q.; Neurock, M.; Wright, H. A.; Srinivasan, N. J. Phys. Chem.
B 2002, 106, 2826–2829.
(60) Joos, L.; Filot, I. A. W.; Cottenier, S.; Hensen, E. J. M.; Waro-
quier, M.; Speybroeck, V. V.; van Santen, R. A. J. Phys. Chem. C
2014, 118, 5317–5327.
(61) Sheppard, D.; Xiao, P.; Chemelewski, W.; Johnson, D. D.; Hen-
kelman, G. J. Chem. Phys., 2012, 136, 074103.
(62) Sheppard, D.; Terrell, R.; Henkelman, G. J. Chem. Phys., 2008,
128, 134106.
(63) Henkelman, G.; Jónsson, H. J. Chem. Phys., 1999, 111, 7010–
7022.
(64) Olsen, R. A.; Kroes, G. J.; Henkelman, G.; Arnaldsson, A.;
Jónsson, H. J. Chem. Phys., 2004, 121, 9776–9792.
(65) Henkelman, G.; Arnaldsson, A.; Jónsson, H. Comput. Mater.
Sci., 2006, 36, 354–360.
(66) Teng, B. -T.; Zhao, Y.; Wu, F. -M.; Wen, X. -D.; Chen, Q. -P.;
Huang, W. -X. Surf. Sci., 2012, 606, 1227–1232.
(67) Storsæter, S.; Chen, D.; Holmen, A. Surf. Sci., 2006, 600,
2051−2063.
(68) Steynberg, P. J.; van de Berg, J. A.; Janse van Rensburg, W. J.
Phys.: Condens. Matter 2008, 20, 064238.
(
15) Mao, W.; Su, J.; Zhang, Z.; Xu, X. -C.; Dai, W.; Fu, D.; Xu, J.; Zhou,
X.; Han, Y. -F. Chem. Eng. Sci. 2015, 135, 312–322.
16) Mei, D.; Rousseau, R.; Kathmann, S. M.; Glezakou, V. -A.;
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Engelhard, M. H.; Jiang, W.; Wang, C.; Gerber, M. A.; White, J. F.;
Stevens, D. J. J. Catal., 2010, 271, 325–342.
(17) Liu, J.; Tao, R.; Guo, Z.; Regalbuto, J. R.; Marshall, C. L.; Klie, R.
F.; Miller, J. T.; Meyer, R. J. ChemCatChem 2013, 5, 3665–3672.
(
2
(
18) Yu, J.; Mao, D.; Han, L.; Guo, Q.; Lu, G. Catal. Commun. 2012,
7, 1–4.
19) Yang, N.; Medford, A. J.; Liu, X.; Studt, F.; Bligaard, T.; Bent, S.
F.; Nørskov, J. K. J. Am. Chem. Soc., 2016, 138, 3705–3714.
20) Lu, Y.; Yu, F.; Hu, J.; Liu, J. Appl. Catal. A 2012, 429–430, 48–58.
(
(21) Lu, Y.; Cao, B.; Yu, F.; Liu, J.; Bao, Z.; Gao, J. ChemCatChem
2014, 6, 473–478.
(
22) Herranz, T.; Rojas, S.; Pérez-Alonso, F. J.; Ojeda, M.; Terreros,
P.; Fierro, J. L. G. J. Catal., 2006, 243, 199–211.
(
(
23) Kim, K. S. J. Electron. Spectrosc., 1974, 3, 217–226.
24) Svintsitskiy, D. A.; Kardash, T. Y.; Stonkus, O. A.; Slavinskaya,
E. M.; Stadnichenko, A. I.; Koscheev, S. V.; Chupakhin, A. P.; Boro-
nin, A. I. J. Phys. Chem. C 2013, 117, 14588–14599.
(25) Natesakhawat, S.; Lekse, J. W.; Baltrus, J. P.; Ohodnicki, Jr., P.
R.; Howard, B. H.; Deng, X.; Matranga, C. ACS Catal., 2012, 2, 1667–
1
(
676.
26) Ye, Y.; Wang, L.; Zhang, S.; Zhu, Y.; Shan, J.; Tao, F. Chem.
Commun., 2013, 49, 4385–4387.
27) de Smit, E.; van Schooneveld, M. M.; Cinquini, F.; Bluhm, H.;
(
Sautet, P.; de Groot, F. M. F.; Weckhuysen, B. M. Angew. Chem. Int.
Ed., 2011, 50, 1584–1588.
(
9
(
28) Kuivila, C. A.; Butt, J. B.; Stair, P. C. Appl. Surf. Sci., 1988, 32,
9–121.
29) Tranquada, J. M.; Heald, S. M.; Moodenbaugh, A. R. Phys. Rev.
B 1987, 36, 5263–5274.
30) Kau, L. S.; Hodgson, K. O.; Solomon, E. I. J. Am. Chem. Soc.,
1989, 111, 7103–7109.
31) Su, J.; Zhang, Z.; Fu, D.; Liu, D.; Xu, X. -C.; Shi, B.; Wang, X.; Si,
R.; Jiang, Z.; Xu, J.; Han, Y. -F. J. Catal., 2016, 336, 94–106.
32) Martens, G.; Rabe, P.; Schwentner, N.; Werner, A. Phys. Rev. B
978, 17, 1481–1488.
33) Sankar, G.; Vasudevan, S.; Rao, C. N. R. J. Chem. Phys., 1986,
(
(
(
1
(
85, 2291–2299.
(34) Wang, X.; Hanson, J. C.; Frenkel, A. I.; Kim, J. Y.; Rodriguez, J. A.
J. Phys. Chem. B 2004, 108, 13667–13673.
(
35) Frenkel, A. I.; Wang, Q.; Marinkovic, N.; Chen, J. G.; Barrio, L.;
Si, R.; López Cámara, A.; Estrella, A. M.; Rodriguez, J. A.; Hanson, J.
C. J. Phys. Chem. C 2011, 115, 17884–17890.
(
36) Bock, D. C.; Pelliccione, C. J.; Zhang, W.; Wang, J.; Knehr, K. W.;
(69) Retief, J. J. Powder Diffr. 1999, 14, 130−132.
(70) Hofer, L. J. E. Cohn, E. M. J. Am. Chem. Soc. 1959, 81,
1576−1582.
(71) Cheng, J.; Hu, P.; Ellis, P.; French, S.; Kelly, G.; Lok, C. M. J. Phys.
Chem. C 2009, 113, 8858–8863.
(72) Cheng, J.; Gong, X. -Q.; Hu, P.; Lok, C. M.; Ellis, P.; French, S. J.
Catal., 2008, 254, 285–295.
(73) Cheng, J.; Hu, P.; Ellis, P.; French, S.; Kelly, G.; Lok, C. M. J. Phys.
Wang, J.; Wang, F.; West, A. C.; Marschilok, A. C.; Takeuchi, K. J.;
Takeuchi, E. S. ACS Appl. Mater. Interfaces 2016, 8, 11418–11430.
(
37) Thüne, P.; Moodley, P.; Scheijen, F.; Fredriksson, H.; Lancee,
R.; Kropf, J.; Miller, J.; Niemantsverdriet, J. W. H. J. Phys. Chem. C,
012, 116, 7367–7373.
38) Ribeiro, M. C.; Jacobs, G.; Pendyala, R.; Davis, B. H.; Cronauer,
2
(
D. C.; Kropf, A. J.; Marshall, C. L. J. Phys. Chem. C, 2011, 115, 4783–
4792.
(39) Park, J. C.; Yeo, S. C.; Chun, D. H.; Lim, J. T.; Yang, J. -I.; Lee, H. -
T.; Hong, S.; Lee, H. M.; Kim, C. S.; Jung, H. J. Mater. Chem. A 2014,
Chem. C 2010, 114, 1085–1093.
(74) Pham, T. H.; Qi, Y.; Yang, J.; Duan, X.; Qian, G.; Zhou, X.; Chen,
D.; Yuan, W. ACS Catal., 2015, 5, 2203–2208.
(75) Cheng, J.; Hu, P.; Ellis, P.; French, S.; Kelly, G.; Lok, C. M. J. Phys.
Chem. C, 2008, 112, 9464–9473.
2
, 14371–14379.
ACS Paragon Plus Environment

Page 13 of 13
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
13
ACS Paragon Plus Environment