Isolation of Cu Atoms in Pd Lattice: Forming Highly Selective Sites for Photocatalytic Conversion of CO2 to CH4

Article abstract of DOI:10.1021/jacs.7b00452

Photocatalytic conversion of CO₂ to CH₄, a carbon-neutral fuel, represents an appealing approach to remedy the current energy and environmental crisis; however, it suffers from the large production of CO and H₂ by side reactions. The design of catalytic sites for CO₂ adsorption and activation holds the key to address this grand challenge. In this Article, we develop highly selective sites for photocatalytic conversion of CO₂ to CH₄ by isolating Cu atoms in Pd lattice. According to our synchrotron-radiation characterizations and theoretical simulations, the isolation of Cu atoms in Pd lattice can play dual roles in the enhancement of CO₂-to-CH₄ conversion: (1) providing the paired Cu-Pd sites for the enhanced CO₂ adsorption and the suppressed H₂ evolution; and (2) elevating the d-band center of Cu sites for the improved CO₂ activation. As a result, the Pd₇Cu₁-TiO₂ photocatalyst achieves the high selectivity of 96% for CH₄ production with a rate of 19.6 μmol g_cat^-1 h^-1. This work provides fresh insights into the catalytic site design for selective photocatalytic CO₂ conversion, and highlights the importance of catalyst lattice engineering at atomic precision to catalytic performance.

Full text of DOI:10.1021/jacs.7b00452

Subscriber access provided by UNIV OF ARIZONA
Article
Isolation of Cu Atoms in Pd Lattice: Forming Highly
Selective Sites for Photocatalytic Conversion of CO2 to CH4
Ran Long, Yu Li, Yan Liu, Shuangming Chen, Xusheng Zheng, Chao Gao, Chaohua
He, Nanshan Chen, Zeming Qi, Li Song, Jun Jiang, Junfa Zhu, and Yujie Xiong
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b00452 • Publication Date (Web): 09 Mar 2017
Downloaded from http://pubs.acs.org on March 9, 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.
Journal of the American Chemical Society 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 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Isolation of Cu Atoms in Pd Lattice: Forming Highly Selective
Sites for Photocatalytic Conversion of CO2 to CH4
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
Ran Long, Yu Li, Yan Liu, Shuangming Chen, Xusheng Zheng, Chao Gao, Chaohua He,
Nanshan Chen, Zeming Qi, Li Song, Jun Jiang, Junfa Zhu, and Yujie Xiong*
Hefei National Laboratory for Physical Sciences at the Microscale, iChEM (Collaborative Innovation Center of
Chemistry for Energy Materials), School of Chemistry and Materials Science, Hefei Science Center (CAS), and
National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei, Anhui 230026, P.
R. China.
KEYWORDS. Photocatalysis; carbon dioxide; methane; active sites; copper.
Supporting Information Placeholder
ABSTRACT: Photocatalytic conversion of CO2 to CH4 – a carbon-neutral fuel represents an appealing approach to remedy
the current energy and environmental crisis; however, it suffers from the large production of CO and H2 by side reactions.
The design of catalytic sites for CO2 adsorption and activation holds the key to address this grand challenge. In this article,
we develop highly selective sites for photocatalytic conversion of CO2 to CH4 by isolating Cu atoms in Pd lattice. According
to our synchrotron-radiation characterizations and theoretical simulations, the isolation of Cu atoms in Pd lattice can play
dual roles in the enhancement of CO2-to-CH4 conversion: (1) providing the paired Cu-Pd sites for the enhanced CO2
adsorption and the suppressed H2 evolution; and (2) elevating the d-band center of Cu sites for the improved CO2 activation.
As a result, the Pd7Cu1-TiO2 photocatalyst achieves the high selectivity of 96 % for CH4 production with a rate of 19.6 μmol
-1
gcat h-1. This work provides fresh insights into the catalytic site design for selective photocatalytic CO2 conversion, and
highlights the importance of catalyst lattice engineering at atomic precision to catalytic performance.
suffer from oxidation. Although Cu nanoparticles can be
stabilized by capping agents such as oleylamine,16 the Cu
atoms would be rapidly oxidized once the capping agents
are removed prior to catalysis. Given this situation, we
decide to adopt an alloying strategy for overcoming this
limitation. Alloying Cu with Pd can significantly enhance
the resistance of nanoparticles to oxidation.17, 18 Another
important reason for choosing Pd in the alloys is the
limited activity of Pd in H2 evolution.19-21 The strong
binding of H atoms to Pd would suppress the side reaction
of H2 production during the photocatalytic conversion of
CO2 with H2O.
INTRODUCTION
A significant increase in atmospheric carbon dioxide (CO2)
levels during the past decades has been widely recognized
as a global environment problem.1 For this reason, a large
number of investigations have been performed for the
reduction of CO2.1-6 In particular, the photocatalytic
conversion of carbon dioxide with water into methane
(CO2 + 8H+ + 8e− → CH4 + 2H2O) is an appealing approach,
which allows the transformation of CO2 into a carbon-
neutral fuel by harvesting solar energy.7-13 Despite the
promising future, this approach has faced a grand
challenge in terms of reaction activity and selectivity. The
activity and selectivity have yet to be significantly
improved for two reasons: (1) CO2 molecules possess very
low reactivity in chemical transformations, limiting the
catalytic activity; and (2) multiple side reactions such as
the reduction of CO2 to CO and the H2 evolution from H2O
reduction simultaneously take place, lowering the reaction
selectivity.
More importantly, the alloying strategy would also
provide an opportunity for improving the activity and
selectivity of CO2 conversion from the angle of lattice
engineering. The surface and electronic structures related
to catalytic sites are two versatile knobs for tuning the
catalytic activity and selectivity, both of which can be
tailored by lattice engineering. The metal alloys can supply
various local atomic arrangements on surface, which offers
the capability of maneuvering adsorption configuration.22-
Like many other catalytic systems, the efficiency of CO2-
to-CH4 conversion relies on the catalytic sites on
photocatalyst surface.7 Cu has been previously identified
as a metal which can provide active sites for CO2
conversion.1,4,14,15 In addition to the activity and selectivity
issues, the metallic Cu materials at the nanoscale heavily
24
Meanwhile, the compositions in alloys hold the key to
tuning the d-band centers of active sites.5,25-27 Taken
together, controllable lattice engineering in Cu-based
alloys should be a promising approach to the giant
enhancement on CO2 conversion activity and selectivity.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
Herein, we report that the PdxCu1 alloys with Cu atoms R3000 electron energy analyzer, a twin-anode X-ray
1
2
3
4
5
6
7
8
isolated in Pd lattice can provide highly active sites for the
enhanced CO2 adsorption and activation. The active sites,
the paired Cu-Pd atoms with tunable d-band centers,
enable the dramatically enhanced activity and selectivity
in the photocatalytic reaction of CO2 with H2O. As a proof
of concept, the Pd7Cu1-TiO2 photocatalyst achieves the
high selectivity of 96 % for CH4 production rate with a rate
of 19.6 μmol gcat-1 h-1. This work offers a lattice engineering
approach to boosting the selectivity in photocatalytic CO2
conversion.
source(Mg Ka and Al Ka), a UV light source, a rear-view
optics for low-energy electron diffraction (LEED), and a
high-precision manipulator with four-degree-of-freedom
(X, Y, Z, R for changing polar angle ()). The preparation
chamber comprised an Ar+ sputter ion gun, a quartz crystal
microbalance (QCM), a quadrupole mass spectrometer
(Pfeiffer QMS220), a high precision manipulator with four-
degree-of-freedom (X, Y, Z and R), and several evaporators.
The high-pressure reactor performed the in-situ high-
pressure experiments under different atmosphere
conditions with a maximum range of 2 MPa and a highest
temperature of 650 °C.
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
EXPERIMENTAL SECTION
In-situ DRIFTS characterization. In-situ diffuse
reflectance infrared Fourier-transform spectroscopy
(DRIFTS) measurements were performed using a Bruker
IFS 66v Fourier-transform spectrometer equipped with a
Harrick diffuse reflectance accessory at the Infrared
Spectroscopy and Microspectroscopy Endstation (BL01B)
in National Synchrotron Radiation Laboratory (NSRL),
Hefei. Each spectrum was recorded by averaging 256 scans
at a 2 cm−1 spectral resolution. The samples were held in a
custom-fabricated IR reaction chamber which was
specifically designed to examine highly scattering powder
samples in the diffuse reflection mode. The chamber was
sealed with two ZnSe windows. During the in-situ
characterization, 0.2 MPa of CO2 was introduced into the
chamber, and the peaks of free molecular CO2 were set as
a reference between the samples.
Photocatalytic CO2 conversion measurement. The
measurement was performed using our previously
developed stainless-steel photocatalytic reactor.30 In the
reactor, the quartz tube had a length of 320 mm and an
inner diameter of 20 mm. The total volume of the reactor
was about 100 mL. A 300-W Xe lamp (SolarEdge700, China)
with a 400-nm short-wave-pass cutoff filter (i.e., λ < 400
nm) was used as the light source and positioned above the
photocatalytic reactor. The power density was measured
to be 2 mW·cm-2. The catalysts (5 mg of TiO2 loaded with
0.01 mmol of metal atoms) were mounted above a small
quartz beaker containing 1 mL of water and placed in a
quartz tube. The photocatalytic reaction was typically
performed for 2 h with 0.2 MPa of CO2 to calculate the
average rates for gas production. The amounts of CH4, CO
and H2 evolved were determined using gas chromatograph
(GC, 7890A, Ar carrier, Agilent). H2 was detected using a
thermal conductivity detector (TCD). The CH4 was
measured by a flame ionization detector (FID). The CO
was converted to CH4 by a methanation reactor and then
analyzed by the FID. The isotope-labelled experiments
were performed using 13CO2 instead of 12CO2, and the
products were analyzed using gas chromatography-mass
spectrometry (GC-MS, 7890A and 5975C, Agilent).
Synthesis of PdxCu1-TiO2 hybrid structures. In a
typical synthesis, 8 mL of an aqueous solution containing
100 mg of TiO2 nanosheets was sonicated for 20 min in a
50-mL glass vial. Then 105 mg of polyvinylpyrrolidone
(PVP, M.W. = 55,000), 120 mg of ascorbic acid (AA), and
300 mg of KBr were added into the glass vial under
magnetic stirring in an oil bath preset to 80 °C for 10 min.
Subsequently, certain amounts of K2PdCl4 (0.100 mmol,
0.150 mmol, 0.167 mmol, 0.175 mmol, 0.180 mmol, and 0.183
mmol in 3 mL water) and CuCl2·2H2O (0.100 mmol, 0.050
mmol, 0.033 mmol, 0.025 mmol, 0.020 mmol, and 0.0167
mmol in 1 mL water) solutions were injected into the
reaction solution. The reaction mixture was heated at
80 °C in air for 3 h. The samples were washed with water
several times to remove most of the AA and PVP by
centrifugation. The obtained samples were denoted as
Pd1Cu1-, Pd3Cu1-, Pd5Cu1-, Pd7Cu1-, and Pd11Cu1-TiO2,
respectively, according to the amounts of added K2PdCl4
and CuCl2 solutions.
XAFS characterization. Cu K-edge and Pd K-edge X-
ray absorption fine structure (XAFS) measurements were
performed at the beamline 14W1 in Shanghai Synchrotron
Radiation Facility (SSRF), China.
The X-ray was
Si(311)
monochromatized by double-crystal
a
monochromator. The storage ring of SSRF was operated at
3.5 GeV with the current of 300 mA. The acquired
extended XAFS (EXAFS) data were processed according to
the standard procedures using the WinXAS3.1 program.28
Theoretical amplitudes and phase-shift functions were
calculated with the FEFF8.2 code using the crystal
structural parameters of the Cu foil, CuO and Pd foil.29
XPS characterization.
X-ray photoelectron
spectroscopy (XPS) experiments were performed at the
Photoemission Endstation at the BL10B beamline in the
National Synchrotron Radiation Laboratory (NSRL) in
Hefei, China. This beamline was connected to a bending
magnet and equipped with three gratings that cover
photon energies from 100 to 1000 eV with a typical photon
flux of 1×1010 photons s-1 and a resolution (E/ΔE) better than
1,000 at 244 eV. The endstation comprised four chambers
including analysis chamber, preparation chamber, high-
pressure reactor and load-lock chamber. The base
pressures for the four chambers were 2×10−10, 2×10−10, 5×10−9,
and 5×10−9 mbar, respectively. The analysis chamber was
connected to the beamline and equipped with a VG Scienta
First-principles simulations. Spin-polarized density
functional theory (DFT) calculations were performed with
the Vienna ab initio simulation package (VASP).31 Perdew-
Burke-Ernzerhof (PBE) exchange-correlation functional
and projector augmented-wave (PAW) potential were
employed. An energy cutoff of 400 eV was used for the
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
plane-wave expansion of the electronic wave function. The energy-dispersive X-ray spectroscopy (EDS) line scans on
1
2
3
4
5
6
7
8
force and energy convergence criterion were set to be 0.01
eV Å -1 and 10-5 eV, respectively. A 2 × 2 slab with five atomic
layers was employed to model the Pd-Cu (100) surfaces, of
which two bottom layers were fixed to their bulk positions
during the geometry optimizations. A 7 × 7 × 1 k-points
grid was adopted for the first Brillouin zone using gamma
center sampling scheme. The periodic boundary condition
(PBC) was set with a 25 Å vacuum layer in the c direction
to isolate surfaces in adjacent slabs. The adsorption energy
the alloy nanocrystals. As revealed by Figures 1e and 1f, Pd
and Cu elements are nearly distributed at the same
locations, suggesting the formation of alloys.
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
Eads is defined as: Eads = Etotal – Esurface – ECO2, where Etotal
,
Esurface, and ECO2 stand for the energy of adsorption
configurations, the energy of metallic surfaces, and the
energy of molecular CO2, respectively. In the simulation,
we have considered several configurations for the adsorbed
CO2 molecule. According to the adsorption energy, the
most favorable configurations are presented.
RESULTS AND DISCUSSION
The PdxCu1 alloy nanocrystals with a face-centered cubic
(fcc) phase are in-situ grown on anatase TiO2 nanosheets
through an aqueous solution synthesis similarly to the
protocol for Pd-TiO2 hybrid structures,32. 33 in which the
atomic ratios of Pd to Cu can be controlled by adjusting the
concentrations of Pd and Cu precursors. As shown in
Figures 1 and S1, the namely PdxCu1-TiO2 hybrid structures
have PdxCu1 alloy nanocrystals supported on TiO2
nanosheets. The PdxCu1 alloy nanocrystals display a
truncated cubic profile with the average edge length of 6
nm, and the TiO2 nanosheets have edge lengths of 50-70
nm and thickness of 6-10 nm. While their crystal phases
are confirmed by X-ray diffraction (XRD, Figure S2), we
employ inductively-coupled plasma mass spectrometry
(ICP-MS) to determine the molar ratios of Pd to Cu in the
samples. As listed in Table S1, the content of Cu in PdxCu1
alloys can be controlled in a certain range up to 50 %,
beyond which the formed product becomes unstable. As a
result, we obtain a series of samples (namely, Pd1Cu1-,
Pd3Cu1-, Pd5Cu1-, Pd7Cu1-, Pd9Cu1-, and Pd11Cu1-TiO2) with
various atomic ratios of Pd to Cu and comparable total
molar contents of Pd and Cu (11-13 %). XPS measurements
(Figure S3) indicate that the Pd and Cu elements in the
samples are of zero valence; however, too high Cu content
makes the PdxCu1 nanocrystals more susceptible to
oxidation.
Figure 1. TEM images of (a) Pd1Cu1-TiO2 and (b) Pd7Cu1-
TiO2 hybrid structures. HAADF-STEM images of (c) Pd1Cu1
and (d) Pd7Cu1. The insets of c and d show the atomic-
resolution images taken from the regions marked by the
boxes. EDS line mapping profiles of (e) Pd1Cu1 and (f)
Pd7Cu1 along the directions marked by the yellow lines in c
and d, respectively.
To precisely resolve the local structural information for
Pd and Cu atoms, the samples are further characterized by
synchrotron radiation-based X-ray absorption fine
structure (XAFS) spectroscopy. Figures 2a and 2b show the
Cu and Pd K-edge X-ray absorption near-edge structure
(XANES) spectra of Pd1Cu1, Pd7Cu1 and Pd11Cu1 in reference
to standard Cu and Pd foil. Processed through a Fourier-
transform (FT), the Cu and Pd K-edge extended XAFS
(EXAFS) spectra are obtained. The main peak in the Cu K-
edge spectra of PdxCu1 samples (Figure 2c) fits the
characteristics of Pd-Pd bond instead of Cu-Cu bond,
indicate that Cu atoms are mostly coordinated with Pd
atoms in the Pd lattice. Moreover, the emerging shoulder
peaks in the Cu K-edge spectra of PdxCu1 samples indicate
that the surface Cu atoms are slightly oxidized particularly
for the Pd1Cu1 with relatively high Cu content. In
comparison, the Pd K-edge spectra of PdxCu1 samples
(Figure 2d) do not show distinct difference from that of Pd
foil, except for the slightly reduced amplitude. This local
Figures 1a and 1b display the transmission electron
microscopy (TEM) images of two typical samples with
different Cu contents – Pd1Cu1-TiO2 and Pd7Cu1-TiO2,
showing that the PdxCu1 alloy nanocrystals with
comparable shapes and sizes are loaded on TiO2
nanosheets regardless the content of Cu atoms. Atomic-
resolution aberration-corrected high-angle annular dark-
field scanning TEM (HAADF-STEM, Figures 1c and 1d)
images show that the Pd1Cu1 and Pd7Cu1 have the lattice
fringes with a spacing of 1.9 and 2.0 Å , respectively, which
can be assigned to the (200) planes of fcc Pd-Cu alloys. The
incorporation of more Cu atoms into Pd lattice slightly
reduces the lattice constant. To resolve the spatial
distribution of Pd and Cu elements, we have collected
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
structural information indicates that the Pd atoms mainly cocatalysts for improving charge separation and supplying
1
2
3
4
5
6
7
8
retain the original Pd lattice.
reaction sites. Figure S5 shows the photocurrents of all
samples which can offer an informative evaluation for the
efficiency of electron-hole separation. It demonstrates
that the addition of PdxCu1 and Pd nanocrystals
significantly enhances the photocurrent of TiO2,
suggesting the role of co-catalysts in separating the
electron-hole pairs photogenerated in TiO2 nanosheets.
Nevertheless, various PdxCu1 and Pd co-catalysts show very
limited difference in the photocurrent enhancement
factors, which are in the range of 1.8 to 2.3 times. It
indicates that all the co-catalysts can establish valid
Schottky junctions with TiO2 to improve the electron-hole
separation. Thus the volcanic relationship between
cocatalyst compositions and photocatalytic activities
should be directly related to the active sites.
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
Table 1. Fitting results of Cu and Pd K-edge EXAFS data.
The lengths of Cu-O, Cu-Cu, Cu-Pd, Pd-Cu, and Pd-Pd
bonds and coordination numbers of Cu and Pd atoms are
extracted from the curve-fitting for Cu and Pd K-edge
EXAFS data (Figure 2), respectively. CN, the coordination
numbers of Cu and Pd atoms; R, the lengths of Cu-O, Cu-
Cu, Cu-Pd, Pd-Cu, and Pd-Pd bonds;σ2, the Debye-Waller
factor. Error bounds (accuracies) are estimated as CN,
5 %; R, 1 %; 2, 1 %.
Figure 2. Normalized (a) Cu K-edge and (b) Pd K-edge
XANES spectra of Pd1Cu1, Pd7Cu1 and Pd11Cu1 in reference
to Cu foil and Pd foil. k3-weighted Fourier-transform (c)
Cu K-edge and (d) Pd K-edge EXAFS spectra.
To look into bonding information, we extract metal-
metal bond lengths and metal coordination numbers from
EXAFS curve fitting. It is known that the atoms inside an
fcc lattice are fully coordinated to have the coordination
number of 12, while the surface atoms are coordinately
unsaturated with a coordination number of 7, 8, and 9 for
(110), (100), and (111) facet, respectively. With the crystal
sizes are shrunk to the nanoscale, the proportion of surface
atoms should be boosted. In our case, the average particle
size of PdxCu1 alloy nanocrystals is about 6 nm, which
makes the total coordination number substantially less
than 12 particularly for Cu atoms. As listed in Table 1, the
oxidation of Cu obviously occurs when the Cu
concentration is at a relatively high level in Pd1Cu1. As the
content of Cu is reduced below 12.5 % (Pd7Cu1 and Pd11Cu1),
the oxidation state of Cu are significantly suppressed.
Moreover, the Cu-Cu bonds disappear in these two
samples, suggesting that the Cu atoms are nearly isolated
in Pd lattice when the Cu/Pd ratio is less than 1/7.
Sampl
EXAFS
Bond
CN R (Å ) σ2 (10-3 Å 2)
e
2.0
0.4
3.3
2.3
7.1
1.2
4.7
1.1
1.91
2.59
2.66
2.66
2.72
1.92
2.68
2.68
2.73
1.89
2.69
2.69
2.73
2.54
2.74
4.3
9.9
11.0
11.0
7.3
3.9
9.1
Cu-O
Cu-Cu
Cu-Pd
Pd-Cu
Pd-Pd
Cu-O
Cu K
Pd1Cu1
Pd K
Cu K
Pd K
Cu K
Pd K
Cu-Pd
Pd-Cu
Pd-Pd
Cu-O
Pd7Cu1
Pd11Cu1
9.1
7.7
0.4
6.3
0.7
8.0
12
6.8
4.0
8.4
8.4
6.4
7.4
5.4
Cu-Pd
Pd-Cu
Pd-Pd
Cu-Cu
Pd-Pd
Upon acquiring the fine structures, we are now in a
position to investigate the efficacy of atomic isolation in
the enhancement of photocatalytic CO2 reduction. Figure
3a shows the average production rates of CH4 and CO in
photocatalytic CO2 reduction with H2O by various TiO2-
based samples. By loading PdxCu1 and Pd nanocrystals on
TiO2, the photocatalytic activities have been obviously
Cu foil
Pd foil
Cu K
Pd K
12
enhanced.
At the same loading amounts, the
Among the photocatalysts, Pd7Cu1-TiO2 achieves the
highest performance with a selectivity of 95.9 % and a rate
photocatalytic production rates show a strong correlation
with the chemical compositions of the loaded metal
nanocrystals.
nanocrystals has not altered light absorption (Figure S4),
this performance enhancement should be related to the
functions of cocatalysts.
-1
of 19.6 μmol gcat h-1 for CH4 production (Table S2). To
Given that the addition of metal
verify the origins of CH4 and CO, we have traced the carbon
sources in the reactions using a 13C isotopic label. Figure
3b shows the mass spectra of CO and CH4 products using
13CO2 as substrate under identical photocatalytic reaction
In principle, the metal
nanocrystals integrated with TiO2 can play a role as
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
conditions. The peaks at m/Z = 29 and m/Z = 17 can be the reduced number of Cu atoms on surface. To support
assigned to 13CO and 13CH4, indicating that the carbon
source of CO and CH4 indeed originate from the used CO2.
Notably the H2 evolution from H2O reduction has been
effectively eliminated (Table S2), demonstrating the niche
of Pd matrix in cocatalysts. In sharp contrast, bare TiO2
exhibits a relatively low selectivity of 57.6 % and a
this argument, we normalize the production rates by the
amount of Cu atoms. As shown in Figure 3c, the reaction
activity per gram of Cu is kept consistent once the Cu
atoms are isolated in lattice (i.e., from Pd7Cu1 to Pd11Cu1).
The isolated Cu atoms can achieve a CH4 production rate
of about 1538 μmol gCu-1 h-1, which turns out to be stable in
a continuous 20-hour test (Figure 3d). As indicated by
Figure S6, the Pd7Cu1-TiO2 photocatalyst shows excellent
durability in seven 4-hour cycles (i.e., 28 hours in total),
which can be ascribed to the well-maintained Cu isolation
and resistance to oxidation during the photocatalysis
(Figure S7 and Table S3).
1
2
3
4
5
6
7
8
-1
production rate of 0.66 μmol gcat h-1 for CH4 production.
When the Pd-rich nanocrystals are used as cocatalysts, the
enhancement of photocatalytic activities is mainly
reflected by the facilitated CO production. For instance,
the integration of TiO2 with Pd1Cu1 nanocrystals
substantially promotes the CO production rate from 0.94
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
-1
to 2.10 μmol gcat h-1, while the production rate for CH4
Since high Cu content makes the PdxCu1 nanocrystals
more susceptible to oxidation (Figure S3), a question
naturally arises whether the oxidation of Cu sites causes
the relatively low activity in photocatalytic CO2 conversion
by Pd1Cu1-TiO2. To examine this possibility, we treat the
Pd1Cu1-TiO2 sample with H2 reduction at 200 C, which can
fully reduce the oxidized Cu atoms to Cu(0) but does not
form palladium hydride (Figures S8 and S9). It turns out
that the H2-treated Pd1Cu1-TiO2 still exhibits the same low
activity in photocatalytic CO2 conversion. It indicates that
the low activity of Pd1Cu1 in CO2 reduction is not caused by
the surface oxidation.
-1
stays at 0.70 μmol gcat h-1. As a result, the selectivity for
CH4 production drops to 35.4 % (Table S2).
Figure 3. (a) Average production rates of CH4 and CO in
photocatalytic CO2 reduction with H2O by bare TiO2,
Pd1Cu1-TiO2, Pd3Cu1-TiO2, Pd5Cu1-TiO2, Pd7Cu1-TiO2,
Pd9Cu1-TiO2, Pd11Cu1-TiO2 and Pd-TiO2 hybrid structures
under UV-light (λ < 400 nm) irradiation with a power
density of 2 mW·cm-2. The catalysts are 5 mg of TiO2
loaded with 0.01 mmol of metal atoms. (b) Mass spectra of
13CH4 (m/z = 17) and 13CO (m/z = 29) produced over Pd7Cu1-
TiO2 in photocatalytic reduction of 13CO2. (c) Average
production rates of CH4 and CO calculated by the amounts
of Cu atoms in the PdxCu1-TiO2 catalysts. (d) Time-
dependent production of CH4 and CO in photocatalytic
CO2 reduction with H2O by Pd7Cu1-TiO2 up to 20 h.
Figure 4. In-situ DRIFTS spectra for the adsorption and
activation of CO2 in the presence of H2O under Xe-lamp
irradiation: (a) Pd1Cu1-TiO2 at 3550-3750 cm-1, (b) Pd7Cu1-
TiO2 at 3550-3750 cm-1, (c) Pd1Cu1-TiO2 at 1000-2250 cm-1,
and (d) Pd7Cu1-TiO2 at 1000-2250 cm-1.
Given this situation, we look into the effects of PdxCu1
atomic arrangements on CO2 reduction reactivity by
employing in-situ DRIFTS spectroscopy. The in-situ
DRIFTS spectroscopy is a powerful technique to gain the
local information for reaction species and their
intermediates. In our measurement, a reaction chamber
for photocatalytic reactions is used. The baselines are
deducted from all spectra to exclude environment factors,
and the peaks of free molecular CO2 are set as a reference.
Based on the identical environment, the intensity
difference between the samples can be qualitatively
analyzed. The chamber is firstly filled by 0.2-MPa CO2 for
60 min without incident light. The DRIFTS spectra of both
Pd1Cu1-TiO2 (Figure 4a) and Pd7Cu1-TiO2 (Figure 4b) show
The CH4 production can be dramatically promoted by
isolating Cu in Pd lattice (Table S2). As the Cu content is
diluted to less than 12.5 % (i.e., Pd7Cu1), the selectivity for
CH4 production is elevated to the level of about 96 %.
According to the XAFS characterization (Table 1), Cu
atoms are nearly isolated by the surrounding Pd atoms
when reducing the Cu concentrations down below 12.5 %.
Thus the isolation of Cu sites in lattice appears to be a
powerful approach to improve the selectivity for CH4
production. However, the production rates are gradually
reduced with the further decrease in Cu contents beyond
12.5 %. This decay in production rates should be related to
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
broad peaks at 3735 and 3704 cm-1 owing to the stretching
To gain insight into the CO2 adsorption and activation
vibrations of surface OH-Ti4+ and OH-Ti3+, respectively,34
in addition to the vibrations of gas-phase CO2 (Figures 4c
and 4d). The bands located at 3626 and 3610 cm-1 can be
assigned to the OH stretching vibrations of free HCO3- and
H2O, respectively.35,36
with Cu atomic isolation, we have employed first-
principles simulation31 to examine the CO2 adsorption on
various PdxCu1 surfaces. In the simulation, we first
examine the adsorption states of CO2 at Pd surfaces with 1
or 3 atoms replaced by Cu, which represent the surfaces
with fully isolated Cu sites and neighbored Cu atoms,
respectively. With the Cu atom fully isolated by Pd lattice,
the CO2 optimal adsorption configuration takes place at
the neighbored Cu-Pd pair with the adsorption energy of -
0.463 eV (Figure 5a). In sharp contrast, the adsorption of
CO2 to the neighbored Cu-Cu pair is substantially weaker
(-0.308 eV, Figure 5b). It demonstrates that a Cu-Pd
atomic pair, which is formed by isolating Cu atoms in Pd
lattice, can serve as a highly active site for CO2 adsorption.
1
2
3
4
5
6
7
8
The chamber is then irradiated by 300-W simulated
solar Xe lamp for another 60 min. In the presence of CO2
and incident light, the vibrations of OH show no distinct
changes for both Pd1Cu1-TiO2 and Pd7Cu1-TiO2 hybrid
structures (see Figures 4a and 4b). However, the spectra
exhibit some significant differences in the range of 1000-
1800 cm-1 (see Figures 4c and 4d). The exposure of Pd7Cu1-
TiO2 to CO2 and incident light leads to the appearance of
several bands, whereas the changes for Pd1Cu1-TiO2 are
significantly weaker. On the surface of Pd7Cu1-TiO2, the
peaks at 1179 and 1650 cm-1 can be assigned to surface
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
In addition to the surface structures, electronic
structures are believed to play an additional role in tuning
catalytic activities. We calculate the d-band centers of Cu
atoms in two ordered lattices – Pd7Cu1 (Figure 5c) and
Pd1Cu1 (Figure 5d), which reveals that the Cu d-band center
of Pd7Cu1 is substantially higher than that of Pd1Cu1 (-1.161
versus -1.452 eV). The Cu d-band centers are tuned
through the electronic structure coupling with the
surrounding Pd atoms. When a Cu atom is surrounded by
more Pd atoms, the electronic structure will be more
affected by the Pd atoms to better elevate the d-band
center of each Cu site. The elevated Cu d-band center of
Pd7Cu1 would increase the catalytic activity of Cu sites. As
compared with CO production (CO2 + 2H+ + 2e− → CO +
H2O), the conversion of CO2 to CH4 (CO2 + 8H+ + 8e− →
CH4 + 2H2O) requires an 8-electron process to which the
enhanced catalytic activity of reaction sites can make an
important contribution.
-
bicarbonate HCO3 species.34,35 Note that the broadened
peaks may originate from a variety of intermediates formed
during the photocatalytic CO2 conversion. Nevertheless,
the peaks at 1294 and 1434 cm-1 are in good agreement with
-
the reported values for the generation of CO2 and
carbonate-like (CO3=) species.36,37 The absence of signals in
the range of 1800-2000 cm-1 indicates that no CO has been
detected in gas phase or at catalytic surface.33,34 Based on
the DRIFTS observations, we can conclude that the
monatomic Cu sites in Pd matrix lattice significantly
-
-
enhances the formation of HCO3 , CO2 and CO3= species,
which are the most important intermediate states in the
reactions of CO2 conversion as reported in literature.35-37 In
comparison, the signal changes during all adsorption and
reaction processes in the spectra for Pd1Cu1-TiO2 are quite
weak, which well aligns with its low performance in
photocatalytic CO2 conversion.
To further demonstrate the importance of Cu isolation
to CO2 activation, we have annealed our Pd1Cu1-TiO2
sample at 300 C in H2 atmosphere (see Figure S10 and
Table S4). Although this annealing process does not
produce the intermetallic B2 phase where Cu sites are
completely isolated as reported in literature,5 a small
number of Cu atoms have migrated to become relatively
separated. As a result, the annealing treatment has slightly
promoted the formation of CH4 in photocatalysis. This
suggests that the isolation of Cu sites should be beneficial
to the activation of CO2 molecules. Our investigation
indicates that the isolation of Cu sites can promote the
photocatalytic CO2 reduction regardless of photocatalyst
semiconductors. As a demonstration, we have employed
BiVO4 as a visible-light photocatalyst for integration with
our PdxCu1 cocatalysts. As shown in Figure S11, Pd7Cu1-
BiVO4 can effectively convert CO2 into CH4 with a small
amount of CO via photocatalysis under visible irradiation.
Figure 5. The most favorable configurations and
adsorption energies of CO2 at (a) an isolated Cu atom (i.e.,
a Cu-Pd pair), and (b) two neighbored Cu atoms (i.e., a Cu-
Cu pair). The structural models and Cu d-band centers for
ordered (c) Pd7Cu1 and (d) Pd1Cu1 lattices. All the results
are obtained by first-principles simulations.
CONCLUSION
In conclusion, we have developed highly selective sites
for photocatalytic conversion of CO2 to CH4 by isolating Cu
atoms in Pd lattice.
Our synchrotron-radiation
characterizations and theoretical simulations reveal that
the isolation of Cu atoms in Pd lattice can play multiple
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
6. Liu, R.; Stephani, C.; Han, J. J.; Tan, K. L.; Wang, D. W. Angew. Chem.
roles in the enhancement of CO2-to-CH4 conversion. The
Int. Ed. 2013, 52, 4225.
7. Roy, S. C.; Varghese, O. K.; Paulose, M.; Grimes, C. A. ACS Nano
2010, 4, 1259.
8. Habisreutinger, S. N.; Schmidt-Mende, L.; Stolarczyk, J. K. Angew.
Chem. Int. Ed. 2013, 52, 7372.
9. Tu, W.; Zhou, Y.; Zou, Z. Adv. Mater. 2014, 26, 4607.
10. White, J. L.; Baruch, M. F.; Pander III, J. E.; Hu, Y.; Fortmeyer, I. C.;
Park, J. E.; Zhang, T.; Liao, K.; Gu, J.; Yan, Y.; Shaw, T. W.; Abelev,
E.; Bocarsly, A. B. Chem. Rev. 2015, 115, 12888.
11. Zhang, H.; Wei, J.; Dong, J.; Liu, G.; Shi, L.; An, P.; Zhao, G.; Kong,
J.; Wang, X.; Meng, X.; Zhang, J.; Ye, J. Angew. Chem. Int. Ed. 2016,
55, 14310.
1
2
3
4
5
6
7
8
incorporation of trace Cu atoms in Pd matrix can form the
paired Cu-Pd sites for enhanced CO2 adsorption and
elevate the d-band center of Cu sites for improved CO2
activation. The Pd matrix also makes a contribution to the
suppression of a frequently observed side reaction – H2
evolution. Remarkably, the Pd7Cu1 alloy cocatalysts
integrated with TiO2 photocatalysts have achieved the high
selectivity of 96 % for CH4 production with a rate of 19.6
-1
μmol gcat h-1. This work represents a step towards the
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
high-selectivity conversion of CO2 to carbon-neutral fuels.
The concept demonstrated here calls for future efforts on
lattice engineering at atomic precision for catalysts design
towards solar-driven CO2 conversion.
12. Meng, X. G.; Wang, T.; Liu, L. Q.; Ouyang, S. X.; Li, P.; Hu, H. L.;
Kako, T.; Iwai, H.; Tanaka, A.; Ye, J. H. Angew. Chem. Int. Ed. 2014,
53, 11478.
13. Choi, K. M.; Kim, D.; Rungtaweevoranit, B.; Trickett, C. A.;
Barmanbek, J. T.; Alshammari, A. S.; Yang, P.; Yaghi, O. M. J. Am.
Chem. Soc. 2016, 139, 356.
14. Cao, Z.; Kim, D.; Hong, D.; Yu, Y.; Xu, J.; Lin, S.; Wen, X.; Nichols,
E. M.; Jeong, K.; Reimer, J. A.; Yang, P.; Chang, C. J. J. Am. Chem.
Soc. 2016, 138, 8120.
15. Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo, T. F. Energ. Environ.
Sci. 2012, 5, 7050.
16. Jin, M.; He, G.; Zhang, H.; Zeng, J.; Xie, Z.; Xia, Y. Angew. Chem. Int.
Ed. 2011, 50, 10560.
17. Friedrich, M.; Penner, S.; Heggen, M.; Armbruster, M. Angew. Chem.
Int. Ed. 2013, 52, 4389.
18. Long, R.; Huang, H.; Li, Y.; Song, L.; Xiong, Y. Adv. Mater. 2015, 27,
7025.
19. Fan, Z.; Luo, Z.; Huang, X.; Li, B.; Chen, Y.; Wang, J.; Hu, Y.; Zhang,
H. J. Am. Chem. Soc. 2016, 138, 1414.
20. Durst, J.; Simon, C.; Hasche, F.; Gasteiger, H. A. J. Electrochem. Soc.
2015, 162, F190.
21. Kyriakou, G.; Boucher, M. B.; Jewell, A. D.; Lewis, E. A.; Lawton, T.
J.; Baber, A. E.; Tierney, H. L.; Flytzani-Stephanopoulos, M.; Sykes,
E. C. Science 2012, 335, 1209.
22. Xu, Y.; Ruban, A. V.; Mavrikakis, M. J. Am. Chem. Soc. 2004, 126,
4717.
23. Bai, S.; Xie, M.; Kong, Q.; Jiang, W.; Qiao, R.; Li, Z.; Jiang, J.; Xiong,
Y. Part. Part. Syst. Charact. 2016, 33, 506.
24. Nørskov, J. K.; Bligaard, T.; Rossmeisl, J.; Christensen, C. H. Nat.
Chem. 2009, 1, 37.
25. Du, N.; Wang, C.; Wang, X.; Lin, Y.; Jiang, J.; Xiong, Y. Adv. Mater.
2016, 28, 2077.
26. Stamenkovic, V. R.; Mun, B. S.; Arenz, M.; Mayrhoffer, K. J. J.; Lucas,
C. A.; Wang, G.; Ross, P. N.; Markovic, N. M. Nat. Mater. 2007, 6,
241.
27. Greeley, J.; Stephens, I. E. L.; Bondarenko, A. S.; Johansson, T. P.;
Hanse, H. A.; Jaramillo, T. F.; Rossmeisl, J.; Chorkendorff, I.; Nørskov,
J. K. Nat. Chem. 2009, 1, 552.
28. Ressler, T. J. Synchrotron Radiat. 1998, 5, 118.
29. Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Phys. Rev.
B 1998, 58, 7565.
30. Li, R.; Hu, J.; Deng, M.; Wang, H.; Wang, X.; Hu, Y.; Jiang, H. L.;
Jiang, J.; Zhang, Q.; Xie, Y.; Xiong, Y. Adv. Mater. 2014, 26, 4783.
31. Kresse, G.; Furthmüller, J.; Phys. Rev. B 1996, 54, 11169.
32. Long, R.; Mao, K.; Gong, M.; Zhou, S.; Hu, J.; Zhi, M.; You, Y.; Bai,
S.; Jiang, J.; Zhang, Q.; Wu, X.; Xiong, Y. Angew. Chem. Int. Ed. 2014,
53, 3205.
33. Yu, J.; Low, J.; Xiao, W.; Zhou, P.; Jaroniec, M. J. Am. Chem. Soc.
2014, 136, 8839.
34. Wu, W. Q.; Bhattacharyya, K.; Gray, K.; Weitz, E. J. Phys. Chem. C
2013, 117, 20643.
35. Garand, E.; Wende, T.; Goebbert, D. J.; Bergmann, R.; Meijer, G.;
Neumark, D. M.; Asmis, K. R. J. Am. Chem. Soc. 2010, 132, 849.
36. Neatu, S.; Macia-Agullo, J. A.; Concepcion, P.; Garcia, H. J. Am. Chem.
Soc. 2014, 136, 15969.
37. Su, W.; Zhang, J.; Feng, Z.; Chen, T.; Ying, P.; Li, C. J. Phys. Chem.
C 2008, 112, 7710.
ASSOCIATED CONTENT
Supporting Information. Detailed experimental section,
characterization methods, and additional material
characterizations. This material is available free of charge via
the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*yjxiong@ustc.edu.cn
Author Contributions
These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This work was financially supported in part by 973 Program
(No. 2014CB848900), NSFC (No. 21471141, U1532135,
21601173), CAS Key Research Program of Frontier Sciences
(QYZDB-SSW-SLH018), Recruitment Program of Global
Experts, CAS Hundred Talent Program, Anhui Provincial
Natural Science Foundation (1608085QB24), and China
Postdoctoral Science Foundation (2015T80660). XAFS
measurements were performed at the beamline BL14W1 in
the Shanghai Synchrotron Radiation Facility (SSRF), China.
DRIFTS and XPS measurements were performed at the
Infrared Spectroscopy and Microspectroscopy Endstation
(BL01B) and the Photoemission Endstation (BL10B) in the
National Synchrotron Radiation Laboratory (NSRL) in
Hefei, China, respectively.
REFERENCES
1. Aresta, M.; Dibenedetto, A.; Angelini, A. Chem. Rev. 2014, 114, 1709.
2. Maki-Arvela, P.; Simakova, I. L.; Salmi, T.; Murzin, D. Y. Chem. Rev.
2014, 114, 1909.
3. Kothandaraman, J.; Goeppert, A.; Czaun, M.; Olah, G. A.; Prakash, G.
K. J. Am. Chem. Soc. 2016, 138, 778.
4. Posada-Perez, S.; Ramirez, P. J.; Evans, J.; Vines, F.; Liu, P.; Illas, F.;
Rodriguez, J. A. J. Am. Chem. Soc. 2016, 138, 8269.
5. Ma, S.; Sadakiyo, M.; Heima, M.; Luo, R.; Haasch, R. T.; Gold, J. I.;
Yamauchi, M.; Kenis, P. J. J. Am. Chem. Soc. 2017, 139, 47.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
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
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
SYNOPSIS TOC
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
ACS Paragon Plus Environment