Photocatalytic C-C coupling from carbon dioxide reduction on copper oxide with mixed-valence copper(I)/copper(II)

Article abstract of DOI:10.1021/jacs.1c00206

To realize the evolution of C₂₊ hydrocarbons like C₂H₄ from CO₂ reduction in photocatalytic systems remains a great challenge, owing to the gap between the relatively lower efficiency of multielectron transfer in photocatalysis and the sluggish kinetics of C-C coupling. Herein, with Cu-doped zeolitic imidazolate framework-8 (ZIF-8) as a precursor, a hybrid photocatalyst (CuO_X@p-ZnO) with CuO_X uniformly dispersed among polycrystalline ZnO was synthesized. Upon illumination, the catalyst exhibited the ability to reduce CO₂ to C₂H₄ with a 32.9% selectivity, and the evolution rate was 2.7 μmol·g^-1·h^-1 with water as a hole scavenger and as high as 22.3 μmol·g^-1·h^-1 in the presence of triethylamine as a sacrificial agent, all of which have rarely been achieved in photocatalytic systems. The X-ray absorption fine structure spectra coupled with in situ FT-IR studies reveal that, in the original catalyst, Cu mainly existed in the form of CuO, while a unique Cu⁺ surface layer upon the CuO matrix was formed during the photocatalytic reaction, and this surface Cu⁺ site is the active site to anchor the in situ generated CO and further perform C-C coupling to form C₂H₄. The C-C coupling intermediate *OC-COH was experimentally identified by in situ FT-IR studies for the first time during photocatalytic CO₂ reduction. Moreover, theoretical calculations further showed the critical role of such Cu⁺ sites in strengthening the binding of *CO and stabilizing the C-C coupling intermediate. This work uncovers a new paradigm to achieve the reduction of CO₂ to C₂₊ hydrocarbons in a photocatalytic system.

Full text of DOI:10.1021/jacs.1c00206

pubs.acs.org/JACS
Article
Photocatalytic C−C Coupling from Carbon Dioxide Reduction on
Copper Oxide with Mixed-Valence Copper(I)/Copper(II)
Wei Wang, Chaoyuan Deng, Shijie Xie, Yangfan Li, Wanyi Zhang, Hua Sheng,* Chuncheng Chen,
and Jincai Zhao
Cite This: J. Am. Chem. Soc. 2021, 143, 2984−2993
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: To realize the evolution of C2+ hydrocarbons like C2H4
from CO2 reduction in photocatalytic systems remains a great challenge,
owing to the gap between the relatively lower efficiency of multielectron
transfer in photocatalysis and the sluggish kinetics of C−C coupling.
Herein, with Cu-doped zeolitic imidazolate framework-8 (ZIF-8) as a
precursor, a hybrid photocatalyst (CuOX@p-ZnO) with CuOX uniformly
dispersed among polycrystalline ZnO was synthesized. Upon illumination,
the catalyst exhibited the ability to reduce CO2 to C2H4 with a 32.9%
selectivity, and the evolution rate was 2.7 μmol·g−1·h−1 with water as a
hole scavenger and as high as 22.3 μmol·g−1·h−1 in the presence of
triethylamine as a sacrificial agent, all of which have rarely been achieved
in photocatalytic systems. The X-ray absorption fine structure spectra
coupled with in situ FT-IR studies reveal that, in the original catalyst, Cu
mainly existed in the form of CuO, while a unique Cu+ surface layer upon
the CuO matrix was formed during the photocatalytic reaction, and this surface Cu+ site is the active site to anchor the in situ
generated CO and further perform C−C coupling to form C2H4. The C−C coupling intermediate *OC−COH was experimentally
identified by in situ FT-IR studies for the first time during photocatalytic CO2 reduction. Moreover, theoretical calculations further
showed the critical role of such Cu+ sites in strengthening the binding of *CO and stabilizing the C−C coupling intermediate. This
work uncovers a new paradigm to achieve the reduction of CO2 to C2+ hydrocarbons in a photocatalytic system.
similar function to form C2+ hydrocarbons as that in
electrochemical systems.14−16 This difference likely originates
from the distinct reaction driving forces between photo-
catalysis and electrocatalysis: in the former, electrons with
adequate capability to reduce CO2 are provided by semi-
conductor catalysts excited under illumination, with far smaller
densities of transferred electrons than from the latter under
bias voltage.17−19 In the generally proposed pathway for
generation of the C2+ products (Scheme 1), CO2 is first
deoxygenated to *CO (* indicates the adsorbed intermediates
or products), and then two *CO molecules on the neighboring
catalytic sites are coupling, forming the key intermediate
*OC−CO or its protonated form *OC−COH, with the
subsequent sequential multielectron and -proton transfer
completing the CO2 reduction to generate C2+ products such
as ethylene, ethane, or ethanol.20−24 In the photocatalytic
INTRODUCTION
■
Artificial photosynthesis has raised significant current interest
due to its potential application in converting the greenhouse
gas CO2 to fuels or industrial feedstocks with solar energy as
the driving force.1 Currently, the products of light-driven CO2
reduction are mainly limited to two-electron-reduced CO,2,3
while the generation of further reduced hydrocarbons such as
methane4,5 or methanol6 are also reported. Among all the
possible CO2 reduction products, the C−C coupling products,
such as ethylene, that are formed via the sluggish multielectron
reduction and C−C bond formation have been regarded as the
most promising ones, possessing both high energy densities
and market prices.7,8 However, the efficient generation of C−C
coupling products such as ethylene has rarely been achieved in
photocatalytic CO2 reduction systems.9,10
In the electrochemical reduction of CO2, Cu-based catalysts
have been recognized as the only metal catalysts to reduce CO2
into C2+ hydrocarbons.11−13 Moreover, recent studies have
indicated that through the modulation of the oxidation states
and morphologies, the generation of ethylene on Cu-based
catalysts with considerable selectivity (60%−80% Faradaic
efficiency) can be achieved.12,13 However, the utilization of Cu
as a cocatalyst in photocatalytic systems has rarely exhibited a
Received: January 7, 2021
Published: February 11, 2021
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993
© 2021 American Chemical Society
2984

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
32.9% selectivity. In comparison, pristine p-ZnO without Cu
showed no CO2 reduction activity beyond CO. The combined
mechanistic studies by X-ray absorption fine structure spectra
(XAFS), in situ FT-IR spectra and theoretical calculations
revealed that the Cu2+ sites in the surface layer of the CuOX@
p-ZnO catalyst were first reduced to Cu+ in the initial stage
and then being stabilized during the subsequent photocatalytic
CO2 reduction. Such specific surface Cu+ sites on the CuO
matrix were active for CO2 reduction to C2H4, which tightly
bound the in situ generated CO and stabilized the *OC−COH
intermediate, realizing the efficient C−C coupling. This result
sheds light on the significance of tuning the oxidation state of
Cu and provides strategies for the generation of C2+ products
with high selectivity in photocatalytic systems.
Scheme 1. Generally Proposed Pathway for CO2 Reduction
to CO, CH4, and C2H4
system, even in the presence of a Cu cocatalyst that should
thermodynamically favor C−C coupling, the slower electron
transfer rate together with the sluggish kinetics for C−C bond
formation may lead to the release of *CO from the surface
before it can accept the subsequent electrons to be further
reduced to C2+ products.9 Therefore, the success in producing
C2+ hydrocarbon in a photocatalytic system remains a
significant challenge to date, and the search for strategies to
bridge the gap between the lower efficiency of multielectron
transfer and sluggish kinetics for *CO coupling is still
underway.
EXPERIMENTAL SECTION
■
Chemicals. Zinc acetate dihydrate (99%), methanol (≥99.9%), 2-
methylimidazole (98%), copper nitrate (98%), and triethylamine
(TEA) (99%) were purchased from Sigma-Aldrich. Carbon dioxide
(≥99.999%) and argon gas (≥99.999%) were purchased from Beijing
Zhongke Tailong Electronic Technology Co., Ltd. 13CO2 (13C, 99%;
18O, <2%) was purchased from Cambridge Isotope Laboratories.
Recently, the oxidation state of Cu has been proposed to
significantly affect the product selectivity in electrochemical
CO2 reduction.24−27 The oxide-derived Cu catalyst, which is
prepared by the reduction of thermally oxidized Cu, displays
both improved Faradaic efficiency and reduced overpotential
for C2+ products generation, compared with the ordinary Cu0
catalyst. It has been proposed that the residual Cu+ in the
oxide-derived Cu catalyst plays a critical role in both
strengthening the binding of *CO to the catalytic site and
stabilizing the C−C coupling intermediate *OC−CO.27 On
the basis of the above propositions, the effect of the oxidation
state of Cu should be even more crucial in photocatalytic CO2
reduction than in the electrochemical procedures, as the less
efficient electron transfer in the photocatalytic system requires
a longer residence time of *CO on the catalytic sites to accept
the subsequent electrons. Moreover, oxidized Cu+ sites can be
more stable under the reducing condition of photocatalysis
than that of electrochemical procedures. In electrochemical
CO2 reduction, the electrons are transferred from the electrode
to the Cu catalyst and then to the surface adsorbed CO2, so the
applied negative bias potential leads to the fast reductive
depletion of most of the residual Cu+ to Cu0 in oxide-derived
Cu catalysts before the CO2 reduction is initiated. By contrast,
in photocatalysis, the reduction of Cu+ sites or surface
adsorbed CO2 is both driven by the electron transfer from
the external semiconductor photosensitizers, so there is a
competition between Cu+ reduction and CO2 reduction. Also
considering the relatively smaller amounts of electrons
transferred at a time in photocatalysis, it is very possible that
the electrons mainly contribute to the surface CO2 reduction
before reducing the surface Cu+ sites to Cu0. Therefore, tuning
the oxidation state of the Cu cocatalyst would potentially be a
feasible strategy to achieve the multielectron reduction of CO2
to C2+ hydrocarbons in a photocatalytic system, which,
however, has never been systematically explored before.
Herein, we developed a CuOX@p-ZnO hybrid catalyst with
Cu-doped zeolitic imidazolate framework-8 (ZIF-8) as
precursor, in which the initial state of copper was mainly the
oxidized form of CuO, and the polycrystalline ZnO (p-ZnO)
was utilized as semiconductor sensitizer to absorb light and
provide electrons while the copper oxide was the catalytic
center for CO2 reduction. The CO2 reduction products on this
hybrid photocatalyst are not only limited to CO and CH4, the
C−C coupling products of C2H4 were also generated with
Ultrapure water was filtered by equipment (Millipore, Milli-RO Plus)
in the laboratory. All chemicals were used as received without further
purification.
Synthesis of Polycrystalline ZnO (p-ZnO) Derived from ZIF-
8. ZIF-8 was prepared by microwave-assisted thermal synthesis.
Specifically, 9 mmol of zinc acetate dihydrate and 36 mmol of 2-
methylimidazole were dissolved in 80 and 100 mL of methanol,
respectively, to form solution A and B. Solution A was poured into
solution B under vigorous stirring. Subsequently, the resulting mixture
was transferred into the microwave chemical reactor (MCR-3) and
stirred at 50 °C for 2 h under microwave irradiation. The as-prepared
ZIF-8 was collected by centrifugation and washed several times with
methanol. After being dried at 60 °C for 12 h, ZIF-8 was transformed
to p-ZnO by calcination under air at 450 °C for 3 h with a heating
rate of 2 °C·min−1, followed by slow cooling to room temperature.
Synthesis of CuOX@p-ZnO Derived from Cu-Doped ZIF-8
(Cu-ZIF-8). For the synthesis of Cu-ZIF-8, the identical procedure
was employed as described for ZIF-8, except for the introduction of
Cu2+ to solution A (3 mmol of copper nitrate and 6 mmol of zinc
acetate dihydrate in 80 mL of methanol). The collected Cu-ZIF-8 was
then transformed into CuOX@p-ZnO via an identical thermal
procedure as for p-ZnO.
Characterization. Transmission electron microscopy (TEM)
images, high-resolution transmission electron microscopy
(HRTEM) images, and energy dispersive X-ray (EDX) analysis
elemental maps were taken on a JEOL-2100F microscope operating at
an accelerating voltage of 200 kV. X-ray diffraction (XRD) patterns
were obtained on a Bruker D8 Focus X-ray diffractometer with Cu Kα
radiation (λ = 1.5405 Å). X-ray photoelectron spectroscopy (XPS)
was conducted on a VG ESCALAB MKII X-ray photoelectron
spectrometer with a non-monochromatized Al Kα X-ray source (hν =
1486.7 eV). X-ray absorption fine structure spectra (XAFS) were
obtained at the Beijing Synchrotron Radiation Facility (1W1B). UV−
vis absorbance spectra were acquired on a Hitachi U-3900
spectrometer.
Activity Tests of Photocatalytic CO2 Reduction on p-ZnO
and CuOX@p-ZnO. The photocatalytic CO2 reduction performance
for each sample was evaluated with a 300 W Xe lamp (Perfect Light,
Microsolar 300, 320−780 nm, 100 mW·cm−2) as the light source. In a
typical activity test, 5 mg of the photocatalyst was dispersed in
deionized water and then dripped onto a microfiber paper 3 cm in
diameter (Whatman, QMA 1851-047). After naturally drying in the
air, the microfiber paper loaded with the photocatalysts was placed in
a 100 mL reactor equipped with a quartz window on the top. Prior to
irradiation, the photoreactor was subjected to vacuum pumping three
times and then refilled by flowing water-vapor-saturated CO2 gas until
2985
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
the pressure reached 0.5 MPa. During the reaction, the gas products
were analyzed by a gas chromatograph (GC7920-TF2Z) equipped
with thermal conductivity and flame ionization detectors, by which
H2, CO, CH4, C2H4, and C2H6 are well-resolved. 13CO2 isotope
labeling experiments were conducted under the same conditions,
except though triethylamine (50 vol % in water) as an extra sacrificial
agent to enhance the yield of the CO2 reduction products, and the gas
products were quantified by GC−MS (gas chromatography−mass
spectrometry, Agilent Technologies 7890A-5957C) with a triple-axis
detector.
In Situ FT-IR Measurement. In situ FT-IR spectra were collected
on a Bruker Vertex 70 V FT-IR spectrometer equipped with a narrow-
band HgCdTe detector and a transmission reaction chamber
(Harrick) connected to an evacuation line (∼10−7 mbar). Five
milligrams of the sample powder were pressed into a self-supported
pellet (7.0 mm in diameter) and placed in the transmission chamber.
In a typical in situ FT-IR measurement, 0.375% CO (diluted by N2)
was first introduced as a probe to identify the states of the surface Cu
sites. Then, the transmission chamber was evacuated to remove all of
the adsorbed and gaseous CO and subsequently purged with water-
vapor-saturated 25% CO2 (12CO2 or 13CO2, diluted by N2) gas flow.
After the equilibrium of CO2 and water adsorption on the catalyst
pellet was reached, the transmission chamber was sealed for the
subsequent photocatalytic reactions. A 100 mW continuous diode
laser (355 nm) was employed as the light source, and a chopped
illumination program with 180 cycles of 20 s illumination and 6 s dark
was performed by periodically intercepting the laser beam with a
mechanical shutter (Vincent Associates, model Uniblitz). The shutter
was controlled by a BNC pulse/delay generator model 565 and
synchronized with the data acquisition in an FT-IR spectrometer. The
operando IR spectrum with a spectral resolution of 4 cm−1 and
scanning velocity of 160 Hz was collected in each 6 s dark period of
the chopped illumination program. Finally, after the photocatalytic
reaction, the chamber was evacuated and then purged with 0.375%
CO again to explore the changes in the surface Cu sites.
energy of the adsorbate before adsorption, Esur is the energy of the
surface before adsorption, and Esys is the energy of the system after
adsorption. Gibbs free energies for each gaseous and adsorbed species
were calculated at 298 K, according to the expression GE = EDFT
+
EZEP − TS, where EDFT is the electronic energy calculated with VASP,
EZPE is the zero-point energy, and TS is the entropy contribution.39
RESULTS
■
p-ZnO was prepared by the aerobic thermal treatment of
pristine ZIF-8, while the hybrid CuOX@p-ZnO photocatalyst
was obtained via the identical pyrolysis procedure but from
Cu-doped ZIF-8 (Cu-ZIF-8). For Cu-ZIF-8, some of the Zn2+
in ZIF-8 were replaced by Cu2+, as indicated by the overall
right shift of the characteristic XRD peaks [Figure S1a,
Supporting Information (SI)], while the dodecahedral top-
ography remained the same as that of ZIF-8 (Figure S2, SI).
After the aerobic pyrolysis, the XRD patterns of the as-
prepared CuOX@p-ZnO and p-ZnO (Figure S1b, SI)
demonstrated the collapse of the topological structure of Cu-
ZIF-8 and ZIF-8 and the formation of the metal oxides. All
characteristic diffraction peaks of ZnO were observed in both
CuOX@p-ZnO and p-ZnO,40 and in the former one, in
addition to the ZnO peaks, the characteristic peaks at 2θ =
35.4° and 38.7° can be readily assigned to the (002) and (111)
planes of CuO, respectively,41 which indicated that the Cu in
CuOX@p-ZnO mainly existed in the form of CuO. The
dominant Cu2+ valence state was also revealed by the Cu 2p
XPS spectrum in Figure S3a (SI) by the observation of satellite
features at 943.0 eV.42
On the basis of TEM images in Figures 1a and S4a (SI),
after pyrolysis, both CuOX@p-ZnO and p-ZnO maintained
Calculation Details. The first-principles calculations were carried
out with the Vienna ab initio simulation package (VASP).28−30 The
interaction between ions and valence electrons was described by using
projector augmented wave (PAW) potentials, and the exchange−
correlation between electrons was treated using the generalized
gradient approximation (GGA) in the Perdew−Burke−Ernzerhof
(PBE) form.31,32 Moreover, the van der Waals interactions are also
taken into account using DFT-D2 corrections.33,34 The plane wave
cutoff energy was 400 eV and a 2 × 2 × 1 sheet k-point mesh was
used. The Γ-point was used for gas-phase molecules. Ionic relaxations
were carried out under conventional energy (10−5 eV) and force (0.03
eV Å−1) convergence criteria. Since conventional density functional
theory (DFT) functionals are unable to describe the strong
correlation effect among the partially filled Cu 3d states in Cu2O@
CuO, the Hubbard parameter, U, is introduced for Cu 3d electrons to
describe the on-site Coulomb interaction, giving the well-known GGA
+ U method. The values of U = 7 eV and J = 0 eV were used for the
Cu2O@CuO material and spin-polarized calculations were performed
since bulk Cu2O@CuO has an anti-ferromagnetic ground state.35
Cu2O(111) and CuO(111) are the most stable low-index surfaces of
the Cu2O and CuO crystals, which could best represent the surface
properties, consistent with previous theoretical results.20,35 For the
original Cu2O, the Cu2O(111) model was established with nine
atomic layers, including three copper atomic layers and six oxygen
atomic layers. Each copper atomic layer was sandwiched by two
oxygen atomic layers.36 For the Cu2O@CuO model, CuO(111) was
first established to be composed of nine atomic layers, including three
copper atomic layers and six oxygen atomic layers, and then half of the
oxygen atoms on the surface layers of CuO(111) were removed to
simulate the feature of the reduced surface Cu+ upon CuO matrix
(Cu2O@CuO).37 These models were further expanded by the
supercell method into a 2 × 2 cell. A 15 Å vacuum layer was placed
above the surface slab to avoid the interference from imaging surface
slabs.38 The adsorption energy (Ead) denotes the interaction between
the surface and is defined as Ead = Eads + Esur − Esys, where Eads is the
Figure 1. (a) TEM, (b) HRTEM, (c) annular dark-field TEM images,
and (d−f) the corresponding elemental mapping images of CuOX@p-
ZnO.
their dodecahedral topography from the Cu-ZIF-8 and ZIF-8
precursors. The HRTEM (Figure S4b, SI) of p-ZnO showed
that the dodecahedron was made up of ZnO nanoparticles with
different orientations. In the hybrid CuOX@p-ZnO (Figure
1b), CuOX nanoparticles were well-dispersed among the ZnO
nanoparticles. The EDX elemental mapping (Figure 1c−f)
further indicated such uniform distribution of Cu and Zn in
the hybrid catalyst, and the atomic concentrations of Cu and
Zn in the near-surface region were calculated to be 7.32 and
42.24 atom %, respectively, with a Cu/Zn atomic ratio of
2986
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
nearly 1:6 (Figure S5, SI). Since the feeding Cu/Zn ratio in the
reactants was 1:1, this would suggest that the current
incorporation of Cu in the CuOX@p-ZnO hybrid catalyst
had reached the maximum. The C elemental mapping (Figures
S4e and S6, SI) showed that only trace amounts of C remained
after the aerobic calcination, while N, another element from
the organic framework of the ZIF precursor, was not detected,
since no notable N 1s peak was observed in the XPS spectra
(Figure S7, SI), demonstrating that the vast majority of the
organic frameworks was transformed to gaseous products and
was removed from the photocatalyst during the aerobic
pyrolysis. On the basis of all these characterizations, it can
be determined that through the aerobic thermal treatment of
the Cu-ZIF-8 precursor, a hybrid catalyst with uniformly
dispersed Cu in the ZnO matrix was obtained.
and C2H4 were also detected; the evolution of the latter has
rarely been reported in photocatalytic systems. The generation
rates of CH4 and C2H4 were of the same magnitude as that of
CO at rates of 2.2 and 2.7 μmol·g−1·h−1, respectively. The
selectivity of hydrocarbon production reached 59.8%, and the
C2H4 production selectivity was 32.9%. H2 was not detected in
our gas−solid reaction system, which could originate from the
advantages in the reagent concentration of gaseous CO2
compared to water vapor in the gas−solid reaction that
remarkably pressed the H2 production.4 On both p-ZnO and
CuOx@p-ZnO, the generation of O2 was detected (Figure S9,
SI), consolidating the role of water vapor as hole scavenger
during the photocatalytic reactions. The hybrid catalyst
exhibited excellent stability over reaction time; the cycling
experiment demonstrated that even in the fourth cycle of
photoreactions (Figure 2c), the yield of each CO2 reduction
product remained >90% of that in the first cycle, and the
overall selectivity of hydrocarbons also remained constant.
The activity test was also performed in the presence of
triethylamine (TEA) as a sacrificial agent (Figure S10, SI).
Since TEA is a better hole scavenger than water, it should
facilitate the separation of the photogenerated carriers and
then enhance the density of the photogenerated electrons. In
the presence of TEA, the photocatalytic CO2 reduction
significantly accelerated on both p-ZnO and CuOX@p-ZnO.
On p-ZnO, the generation rate of CO increased to 9.3 μmol·
g−1·h−1. It is worth noting that with the enhanced electron
density, multielectron-reduced CH4 (2.1 μmol·g−1·h−1) was
formed, suggesting again that the electron density is the
bottleneck for the photocatalytic CO2 reduction to hydro-
carbon. In sharp contrast, despite the enhanced electron
densities in the presence of TEA, C2H4 with more sluggish
generation kinetics was still absent on p-ZnO, stressing the
irreplaceable role of the CuOX moiety in the formation of
C2H4. On CuOX@p-ZnO, regardless of the remarkably
enhanced yield for each CO2 reduction product (CO, 27.3
μmol·g−1·h−1; CH4, 17.9 μmol·g−1·h−1; and C2H4, 22.3 μmol·
g−1·h−1), the product distribution was still the same as that in
the absence of TEA, which suggested that the addition of TEA
enhanced only the reaction rate but did not changed the
reaction pathway.
Then the photocatalytic CO2 reduction performance on p-
ZnO and CuOX@p-ZnO was also tested. As shown in Figure
2a, in the presence of water vapor, the only CO2 reduction
A series of control experiments were perfomed to determine
the origin of the CO2 reduction products on CuOX@p-ZnO.
As shown in Figure S11 (SI), after a 12 h reaction, no CO or
hydrocarbon products were detected when the reaction system
was in absence of illumination, catalyst, or CO2, separately.
Trace products (<0.02 μmol·g−1·h−1) were detected when
replacing the water-saturated CO2 atmosphere with dry CO2,
since the limited amounts of the adsorbed water on the catalyst
may act as hole scavengers. The experiment using isotope-
labeled 13CO2 was also conducted with TEA as a hole
scavenger, and the products were measured by GC−MS. The
GC−MS peak sequences of CO, CH4, CO2, and C2H4 are
shown in Figure S12 (SI), and the ion fragment analysis results
of each peak are displayed in Figure 2d−f. The ion fragment
peaks for CO, CH4, or C2H4 under 13CO2 were clearly shifted
compared with those under 12CO2, demonstrating that the
generated CO and hydrocarbons on CuOX@p-ZnO were
definitely derived from the reduction of CO2 rather than other
sources.
Figure 2. Products of photocatalytic CO2 reduction for (a) CuOX@p-
ZnO and (b) p-ZnO; error bars represent the SD of three
independent measurements using 5 mg of fresh sample for each
measurement. (c) Cycling measurements for CO2 photoreduction of
CuOX@p-ZnO (using a 5 mg sample for this measurement; when a
new catalytic cycle begins, the reactor is pumped and refilled with
pure CO2 and H2O). GC−MS spectra of (d) CO, (e) CH4, and (f)
C2H4 from the photocatalytic reduction of 12CO2 or 13CO2 on
CuOX@p-ZnO.
product on p-ZnO upon illumination was CO, which
accumulated linearly with the reaction time at a rate of 1.8
μmol·g−1·h−1, and no CH4 or C2H4 was detected during the 8
h illumination period. On CuOX@p-ZnO (Figure 2b), under
identical conditions, the rate of generation of CO was almost
doubled, with an average generation rate of 3.3 μmol·g−1·h−1.
More intriguingly, multielectron-reduced hydrocarbons of CH4
Regardless of the identity of the hole scavenger in the
system, the comparison of the performance of CuOX@p-ZnO
and p-ZnO in photocatalytic CO2 reduction clearly demon-
2987
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
strates the key role of the CuOX moiety in the formation of
hydrocarbons, especially the C−C coupling products of C2H4.
Therefore, further mechanistic exploration was conducted to
understand the origin of the extraordinary activities of C2H4
generation on our CuOX@p-ZnO photocatalyst.
Cu+.44,46 Similar to the XPS results, the variations between the
8 and 4 h XANES spectra were unnoticeable, and when
comparing between the 8 and 0 h spectra, the main
characteristic CuO features were retained. The result was
further verified by the Fourier-transformed K2-weighted χ(k)
function of EXAFS. As exhibited in Figure 3b, before
photoreaction, the main backscattering peak of CuOX@p-
ZnO around 1.5 Å corresponded to the Cu−O shell in CuO,
and as the reaction progressed, the peak gradually shifted to
smaller radial distance, indicating the partial reduction of Cu2+,
since Cu2O has the main Cu−O peak at 1.4 Å. It should be
noted that, after 4 h photoreactions, the average Cu−O
distance in CuOX@p-ZnO was very close to that in pristine
Cu2O, whereas the shift in the average distance of the Cu−Cu
shell was quite limited and still closer to that in pristine CuO.
Such a discrepancy would suggest that the structure of the in
situ formed surface Cu2O layers is distinctive from the pure
lattice of Cu2O: the reduction of the original CuO surface layer
leads to the removal the surface O atom, resulting in the
variation in the average distance of the Cu−O shell, but the
Cu2+ reduction should be limited to the surface layer;
therefore, the surface Cu atoms can be still anchored to the
bulk CuO matrix with subsurface O atoms, and then the
change in the distance of the Cu−Cu shell should not be
remarkable.43,45
Since the catalytic behaviors of CO2 reduction on Cu could
be significantly affected by the oxidation states of Cu, the
stabilized valence state of Cu during photocatalytic CO2
reduction on CuOX@p-ZnO was first determined. XPS and
XAFS characterizations were employed to identify the valence
state of Cu in the hybrid photocatalyst that occurred during
different reaction times. According to Cu 2p XPS data (Figure
S13, SI), as the reaction progressed, the characteristic peak of
Cu2+ at 933.5 eV slightly shifted to a lower binding energy,
which would indicate the partial reduction of Cu2+ to Cu+.43
The shift was more prominent within the first 4 h, while no
further alternations in the shape of the ∼933.5 eV peak
occurred when the illumination prolonged, suggesting that no
more transformation from Cu2+ to Cu+ transpired when the
illumination time exceeded 4 h. On the other hand, the
retained collection satellite features at 943.0 eV characteristic
for CuO after 10 h reactions would suggest that the majority of
Cu2+ was unchanged during the illumination and only a small
fraction was reduced to Cu+.
The more definitive evidence for this judgment was from
semi-in-situ XAFS characterization. Cu K-edge X-ray absorp-
tion near-edge structure (XANES) spectra for the hybrid
CuOX@p-ZnO collected after different reaction times are
shown in Figure 3a, and as a comparison, the reference spectra
On the basis of the fitting of data in Figure 3a (Figures S14−
17, SI), Figure 3c revealed the time-dependent variations of the
relative contents of CuO and Cu2O in the near-surface region
of the samples: the relative content of Cu2O increased from the
original 4.1% to 22.5% after 2 h and 31.8% after 4 h, but the
difference between the content at 4 and 8 h (32.1%) was quite
small.
According to the XPS and XAFS data, it can be clearly
concluded that, as the reaction progressed, the surface Cu2+ in
CuOX@p-ZnO was reduced to Cu+, and the Cu2+ reduction
was more significant in the first 2 h and came to an end after 4
h. Then, the most possible structure of the CuOX moiety after
stabilization would be the reduced surface Cu+ layer on the
CuO matrix. Despite the conduction band level of ZnO being
sufficiently high to fully reduce Cu2+ to Cu+ and even Cu0,47
the majority of Cu in the hybrid catalyst still exists in the form
of CuO, and the most likely explanation is the competition
between Cu2+ reduction and CO2 reduction on the catalyst
surface that impeded further reduction of the Cu2+ in the
subsurface layer and bulk. The detailed reaction mechanism
will be subsequently explored by in situ FT-IR spectra.
Time-dependent in situ FT-IR spectra collected during
photocatalytic reactions are employed not only to identify the
reaction intermediate from the IR absorption features but also
to trace the generation kinetics of both intermediates and
products, which provides additional information to uncover
the detailed reaction mechanism.48−51 Before and after the in
situ photocatalytic reactions, external CO molecules were
introduced as probes. CO probes are frequently combined with
infrared spectra to identify the states of the surface sites, as the
IR vibration frequency of bound CO is very sensitive to the
valence state and coordination environment of the surface
metal atoms.23,52,53 As shown in Figure 4a, when CO probes
were introduced before the photocatalytic reaction, only one
band from the adsorbed CO was observed on CuOX@p-ZnO
at 2085 cm−1. Since no CO band was obtained on ZnO, this
band should be assigned to the CO adsorbed on the CuOX
moiety.58 The frequency of CO coordinated to Cu2+ was
Figure 3. (a) Copper K-edge X-ray absorption near-edge structure
(XANES) spectra on pristine Cu0 foil, Cu2O, CuO, and CuOX@p-
ZnO collected after 0, 2, 4, and 8 h photoreactions. (b) Fourier-
transformed K2-weighted χ(k) function of EXAFS and (c) relative
content of CuO and Cu2O on CuOX@p-ZnO after 0, 2, 4, and 8 h
photoreactions.
collected on standard CuO, Cu2O, and Cu foil are also
exhibited. The original catalyst (0 h) mainly showed the
characteristic peaks of CuO at 8985 and 8997 eV, assigned to
1s → 4pz and 1s → 4px,y of Cu2+, respectively.44,45 As the
reaction time increased (2 and 4 h), the characteristic shoulder
peak of Cu2O at 8982 eV, assigned to 1s → 4pz of Cu+, became
prominent; meanwhile, the characteristic peak of CuO at 8997
eV slightly shifted to 8995 eV, ascribed to 1s → 4px,y of
2988
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. Identification of surface Cu sites on CuOX@p-ZnO with
CO as a molecular probe: adsorption of CO before (a) and after (b) 1
h of chopped illumination (the black dashed line is the original
absorption peak, and the other lines are the fitted peaks). The
background for both spectra was the one collected on CuOX@p-ZnO
under vacuum before the photocatalytic reaction. For clarification, the
P and R branches at 2173 and 2120 cm−1 for gaseous CO were
subtracted using the standard spectra collected from the sample-free
chamber containing an identical concentration of gaseous CO.
Figure 5. (A) The chopped illumination program synchronized with
the in situ FT-IR measurement for both ZnO and CuOX@ZnO. (B)
(a) In-situ FT-IR spectra of the photocatalytic CO2 reduction on ZnO
recorded during illumination cycles of 1, 4, 10, 20, 40, 80, 120, and
180 (bottom to top); (b) the FT-IR spectrum collected on CuO@
ZnO in the 180th illumination cycle exhibited for the convenience of
comparison. The background was collected immediately right before
illumination for each sample.
reported to be in the region of 2200−2190 cm−1; however, due
to the weak interaction with Cu2+, such a band was not
observed here.54,55 The 2085 cm−1 band can be attributed to
CO adsorption on the minor Cu+ sites (Cu+2085) that originally
existed in the CuOX@p-ZnO catalyst.56 When the in situ
photocatalytic reactions were ended, the reaction chamber was
evacuated and filled with external CO probe again to identify
the changes on the surface sites after reactions. As shown in
Figure 4b, besides the 2085 cm−1 band, another band at 2102
cm−1 was observed, which can be assigned to adsorbed CO on
newly formed Cu+ that is under different coordination
circumstance with regard to Cu+2085. Corresponding to the
+
XAFS and XPS results, this in situ formed Cu2102 should be
originating from the photocatalytic reduction of the surface
layer Cu2+ in CuOX@p-ZnO.
The CO probe experiment displays accordant results with
XAFS and XPS for the in situ generation of surface Cu+ sites in
CuOX@p-ZnO, and then the detailed reaction procedure was
uncovered by analyzing the in situ IR spectra collected during
the chopped illuminations. A chopped illumination program
synchronized with in situ FT-IR measurements was employed,
as shown in Figure 5A. In each illumination−measurement
cycle, one IR spectrum was collected after 20 s of illumination.
As a control, photocatalytic CO2 reduction on p-ZnO was
first explored. As shown in Figure 5B-a, illumination by a 355
nm laser induced an IR featureless background absorbance that
monotonically increased from 4000 to 1250 cm−1, which was
intensified with prolonged illumination. A similar background
shift has been reported on TiO2 and was assigned to the
accumulation of photogenerated electrons in the conduction
band.57 Simultaneously, with the increase in the background
absorbance, depletions at 1416, 1518, and ∼3190 cm−1 are
observed. There should be another diminished band at
approximately 1643 cm−1 that is blurred by the dramatically
increasing background absorbance (which was more prom-
inent on CuOX@p-ZnO in Figures 5B-b and 6). These
depletion bands can be all assigned to the adsorbed water
molecules on p-ZnO: the broad band centered at 3190 cm−1
for the O−H stretching vibration (νOH), the 1643 cm−1 band
for the bending vibration of the molecular adsorbed water
(δH2O), and the 1518 and 1416 cm−1 bands for the bending
vibration of the dissociated adsorbed water (δZn−OH).58,59
Additionally, considering the relatively low CO2 reduction rate
Figure 6. In situ FT-IR spectra of the photocatalytic (a) 12CO2 or (b)
13CO2 reduction on CuOX@p-ZnO recorded in the illumination cycle
of 1, 2, 4, 10, 20, 40, 80, 120, and 180 (bottom to up).
on p-ZnO in the activity test, it is suggested that on p-ZnO,
water oxidation was much faster than the CO2 reduction half
of the reaction; therefore, since the holes and electrons were
generated stoichiometrically, the dramatic consumption of
holes by water oxidation leaves the residual electrons
accumulated in the conduction band, observed as the
background shift in the infrared spectra.
However, such an increase in the IR background absorbance
disappeared when applying the identical illumination program
on CuOX@p-ZnO, as indicated by the sharp contrast between
Figure 5B-b and the top curve in Figure 5B-a, which are the
2989
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
infrared spectra collected in the 180th illumination cycle on
CuOX@p-ZnO and p-ZnO, respectively. The lack of
accumulated electrons indicated the efficient electron transfer
from the p-ZnO to the CuOX moiety, and the latter part acted
as the reduction reaction center to rapidly consume the
electrons for either the reduction of the adsorbed CO2 or the
self-reduction to Cu+.57 On the basis of this consideration, on
CuOX@p-ZnO, all of the observed CO2-reduction-related IR
bands should be attributed to the intermediates or products on
the CuOX moiety rather than the p-ZnO moiety.
The variations in the infrared spectra with the illumination
time on CuOX@p-ZnO are shown in Figure 6a, while in Figure
6b, 13CO2 was used instead of 12CO2. It can be observed that,
during illumination, the depletion of the molecular adsorbed
water at 1643 cm−1 was greater than that on p-ZnO and was
intensified with prolonged illumination, confirming the
consumption of water molecules as hole scavengers. On the
other hand, in Figure 6a, the growth of the IR bands at 1590,
1439, and 1321 cm−1 was prominent even from the first cycle
of illumination, and all of the three bands had their 13C-
counterparts at 1552, 1410, and 1286 cm−1, respectively
(Figure 6b). According to the literature reports, we would
assign the 1552 and 1286 cm−1 bands to the asymmetric and
symmetric vibration of surface-adsorbed carbonate, while the
1439 cm−1 band could be attributed to the one-electron-
reduced *COOH.22,23,60
and co-workers, who reported the IR wavenumbers of 1584
cm−1 for CO stretching and 1191 cm−1 for C−OH
stretching during electrochemical IR measurements on a
copper electrode.52,60 The band positions were very close to
their theoretical calculation results of 1576 and 1235 cm−1.
The v-QM theoretical calculation of Cu(001) from Goddard
and co-workers also predicted similar wavenumbers at 1548
and 1189 cm−1,22 respectively. On the basis of these literature
reports and the spectral kinetics observed here, we assigned
our 1566 and 1180 cm−1 bands with identical growth kinetics
to *OC−COH intermediates adsorbed on Cu2102+, which
provides the first experimental identification of a C−C
coupling intermediate during the photocatalytic CO2 reduc-
tion. On the other hand, the band at 1407 cm−1 has been
reported as the IR frequency of *CHO, the most reported
intermediate for CH4, on a copper electrode.60 It is also
correlated to our observed spectral kinetics to assign 1407
cm−1 to that active intermediate for CH4.
DISCUSSION
■
According to the in situ FT-IR studies, a detailed reaction
process on CuOX@p-ZnO can be proposed. In the initial stage,
the surface Cu2+ sites on the hybrid catalyst are first reduced to
+
Cu2102 by electron transfer from excited p-ZnO. After the
surface Cu2+ is reduced to Cu+, which is more active for CO2
reduction than Cu2+, surface CO2 reduction on Cu+ dominates
and primarily consumes the electrons transferred from p-ZnO
and thus prevents the reduction of Cu2+ in the subsurface layer
and bulk; therefore, a specific structure of a Cu+ surface layer
upon the CuO matrix is formed, as proven by the XAFS and
CO probing experiments. For the CO2 reduction conducted
on Cu2102+, after the two-electron reduction to CO, some part
of the generated *CO desorbed to form gaseous CO, and
other *CO species were trapped on Cu2102+, leading to
subsequent electron transfer to further reduce the surface-
bound *CO into CH4 and C2H4 via the intermediates *CHO
and *OC−COH, respectively.
In the initial stage of the photocatalytic reactions, no CO
band was observed in the region of 2200−2000 cm−1; after
only an induction period of 10 cycles of illumination, the IR
band of Cu+−CO at 2102 cm−1 with its 13C-counterpart at
2053 cm−1 became notable and grew sharply under the
subsequent illuminations. The appearance of 2102 cm−1 bands
+
gave two indications: first, the Cu2102 was generated; second,
it thereafter acted as the reaction center to reduce CO2 to CO,
and then the generated CO was subsequently in situ trapped.
Intriguingly, no CO band at 2085 cm−1 was detected during
the whole photocatalytic reaction processes, suggesting that
+
the in situ generated Cu2102 was a more reactive surface site
On the basis of this proposition, theoretical calculations
were conducted to determine whether such a specific structure
of surface Cu+ sites favored the formation of C2H4. To
simulate the C−C coupling step on the CuOX catalyst with
mixed Cu(I)/Cu(II) valence states, half of the O atoms were
removed from the surface slab of CuO(111) to form the
surface layers of Cu+(Cu2O@CuO), which is consistent with
the EXAFS results, while the ordinary Cu2O(111) structure
was employed as a comparison (Figure S18, SI). As shown in
Figure 7, the adsorption energy of *CO on Cu2O@CuO (3.29
+
for CO2 reduction than the originally existing Cu2085
.
Almost simultaneously with the appearance of Cu+−CO at
2102 cm−1 was the change in the IR features in the region of
1600−1550 cm−1 (Figure 6a): a new peak at 1566 cm−1
started to grow and the absorption maximum in this region
gradually shifted from the carbonate band of 1590 cm−1 to this
new band with prolonged illumination. The IR bands at 1180
and 1407 cm−1 exhibited growth kinetics identical to those of
the 1566 cm−1 band: an induction period in the first 10 cycles
of illumination was also observed. In the 13CO2 isotope
experiment (Figure 6b), these three bands shifted to 1535,
1374, and 1154 cm−1, respectively, with similar spectral
kinetics.
Distinctive with the carbonate or *COOH bands growing
without an induction period, the evolution of the CO2-related
bands of 1566, 1407, and 1180 cm−1 exhibited growth kinetics
similar to that of the Cu2102+−CO band, suggesting that the
formation of these bands needs the generation of Cu2102+−CO
as a premise. Since, in the activity tests under similar reaction
conditions, the generation of C2H4 and CH4 was confirmed,
both of which were proposed to be further reduced from *CO
(Scheme 1), the 1566, 1407, and 1180 cm−1 bands could be
assigned to the intermediate between *CO and hydrocarbon
C2H4 or CH4. The only spectroscopic identification of *OC−
COH, one of the C−C coupling intermediates, was by Koper
Figure 7. (a) The initial ball-and-stick structural models of the
theoretical calculations of the adsorption energy of *CO on pristine
Cu2O and Cu2O@CuO. (b) First-principles calculations of the C−C
coupling step for Cu2O and Cu2O@CuO. Color code: Cu, pink; O,
red; C, gray.
2990
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
eV) is significantly higher than that on ordinary Cu2O (1.24
eV) (Figure 7a), and the change in the Gibbs free energy (ΔG)
for the formation of the *OC−CO intermediate on Cu2O@
CuO (0.58 eV) is far smaller than the ΔG of 2.29 eV on
pristine Cu2O(111) (Figure 7b). Therefore, it verifies that
compared with Cu+ in the pure Cu2O lattice, the in situ
generated surface Cu+ upon the CuO matrix in CuOX@p-ZnO
should be more conducive to both the trapping of CO and the
subsequent formation of key intermediates in the C−C
coupling. The strengthened binding of *CO and the
remarkable reduced formation energy of the C−C coupling
intermediate during the CO2 reduction are both critical to
achieve the formation of C2H4 in the photocatalytic system:
the former prolongs the residence time of *CO on the catalytic
center, and the latter suggests the acceleration in the critical
but sluggish step of C−C coupling, both of which guarantee
the completion of the multielectron reduction of *CO before it
releases from the catalytic sites, especially when considering
the relatively slower electron transfer rate in the photocatalytic
system. Thus, the theoretical calculations further confirmed
that the structure of in situ generated Cu+ in CuOX@p-ZnO
played a vital role in the photocatalytic CO2 reduction to
C2H4.
AUTHOR INFORMATION
Corresponding Author
■
Hua Sheng − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China; orcid.org/0000-0002-5605-2630;
Email: hsheng@iccas.ac.cn
Authors
Wei Wang − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China
Chaoyuan Deng − Key Laboratory of Photochemistry,
Institute of Chemistry, Chinese Academy of Sciences, Beijing
100190, PR China; University of Chinese Academy of
Sciences, Beijing 100190, PR China
Shijie Xie − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China
Yangfan Li − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China
Wanyi Zhang − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China
Chuncheng Chen − Key Laboratory of Photochemistry,
Institute of Chemistry, Chinese Academy of Sciences, Beijing
100190, PR China; University of Chinese Academy of
Sciences, Beijing 100190, PR China; orcid.org/0000-
0003-4034-8063
Jincai Zhao − Key Laboratory of Photochemistry, Institute of
Chemistry, Chinese Academy of Sciences, Beijing 100190, PR
China; University of Chinese Academy of Sciences, Beijing
100190, PR China; orcid.org/0000-0003-1449-4235
CONCLUSION
■
With Cu-doped ZIF-8 as a precursor, a photocatalyst with
uniformly dispersed copper oxide among polycrystalline ZnO
is synthesized, and upon illumination, this catalyst exhibits the
capability to reduce CO2 to the hydrocarbons CH4 and C2H4.
This ability has rarely been achieved in photocatalytic systems.
The self-preactivation in the initial stage of photocatalysis
builds up the unique surface reaction sites: the surface Cu2+ is
reduced to a Cu+ surface layer upon the CuO matrix, and such
a structure has been demonstrated to be critical for trapping
the in situ generated CO and the subsequent catalytic C−C
coupling for C2H4 production. Further improvement in the
efficiency of CO2 reduction to C2H4 can be expected by
optimizing the assembling of the semiconductor sensitizer and
CuOX cocatalyst. Our study sheds light on the roles of the
valence state of Cu in tuning the reaction pathway of CO2
reduction and opens up a new paradigm to achieve the
reduction of CO2 to C2+ hydrocarbons in a photocatalytic
system.
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c00206
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the National Key Research and
Development Program of China (No. 2020YFA0710303), the
National Natural Science Foundation of China (Nos.
22076193 and 21827809), and the “Strategic Priority Research
Program B” of the Chinese Academy of Sciences (Grant No.
XDB36000000). The authors thank the photoemission beam-
line 1W1B station of the Beijing Synchrotron Radiation
Facility (BSRF) for XAFS measurements.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c00206.
XRD patterns; TEM images; XPS spectra; EDX;
element distribution of C; VB-XPS spectra; O2 evolution
during the photocatalytic CO2 reduction; yield of each
CO2 reduction; reaction behaviors in the CuOX@p-
ZnO system; GC−MS spectra peak sequence of CO,
CH4, CO2, and C2H4 during the photocatalytic CO2
reduction; fitting of relative content of CuO and Cu2O
to XANES spectra; and the initial structural models of
Cu2O, Cu2O−CuO, Cu2O−*OCCO, and Cu2O@
CuO−*OCCO (Figures S1−S19 and Table S1)
(PDF)
REFERENCES
■
(1) Hepburn, C.; Adlen, E.; Beddington, J.; Carter, E. A.; Fuss, S.;
Mac Dowell, N.; Minx, J. C.; Smith, P.; Williams, C. K. The
Technological and Economic Prospects for CO2 Utilization and
Removal. Nature 2019, 575, 87−97.
(2) Jiang, Z.; Xu, X.; Ma, Y.; Cho, H. S.; Ding, D.; Wang, C.; Wu, J.;
Oleynikov, P.; Jia, M.; Cheng, J.; Zhou, Y.; Terasaki, O.; Peng, T.;
Zan, L.; Deng, H. Filling Metal−Organic Framework Mesopores with
TiO2 for CO2 Photoreduction. Nature 2020, 586, 549−554.
2991
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(3) Xiong, X.; Mao, C.; Yang, Z.; Zhang, Q.; Waterhouse, G. I. N.;
Gu, L.; Zhang, T. Photocatalytic CO2 Reduction to CO over Ni
Single Atoms Supported on Defect-Rich Zirconia. Adv. Energy Mater.
2020, 10, 2002928.
(4) Li, X.; Sun, Y.; Xu, J.; Shao, Y.; Wu, J.; Xu, X.; Pan, Y.; Ju, H.;
Zhu, J.; Xie, Y. Selective Visible-Light-Driven Photocatalytic CO2
Reduction to CH4 Mediated by Atomically Thin CuIn5S8 Layers.
Nat. Energy 2019, 4, 690−699.
(5) Rao, H.; Schmidt, L. C.; Bonin, J.; Robert, M. Visible-Light-
Driven Methane Formation from CO2 with a Molecular Iron Catalyst.
Nature 2017, 548, 74−77.
(6) Wu, Y. A.; McNulty, I.; Liu, C.; Lau, K. C.; Liu, Q.; Paulikas, A.
P.; Sun, C. J.; Cai, Z.; Guest, J. R.; Ren, Y.; Stamenkovic, V.; Curtiss,
L. A.; Liu, Y.; Rajh, T. Facet-Dependent Active Sites of a Single Cu2O
Particle Photocatalyst for CO2 Reduction to Methanol. Nat. Energy
2019, 4, 957−968.
(7) Nitopi, S.; Bertheussen, E.; Scott, S. B.; Liu, X.; Engstfeld, A. K.;
Horch, S.; Seger, B.; Stephens, I. E. L.; Chan, K.; Hahn, C.; Nørskov,
J. K.; Jaramillo, T. F.; Chorkendorff, I. Progress and Perspectives of
Electrochemical CO2 Reduction on Copper in Aqueous Electrolyte.
Chem. Rev. 2019, 119, 7610−7672.
(8) Chen, C.; Khosrowabadi Kotyk, J. F.; Sheehan, S. W. Progress
toward Commercial Application of Electrochemical Carbon Dioxide
Reduction. Chem 2018, 4, 2571.
(9) Ran, J.; Jaroniec, M.; Qiao, S. Z. Cocatalysts in Semiconductor-
Based Photocatalytic CO2 Reduction: Achievements, Challenges, and
Opportunities. Adv. Mater. 2018, 30, 1704649.
(10) Albero, J.; Peng, Y.; García, H. Photocatalytic CO2 Reduction
to C2+ Products. ACS Catal. 2020, 10, 5734−5749.
(11) Jiang, K.; Sandberg, R. B.; Akey, A. J.; Liu, X.; Bell, D. C.;
Nørskov, J. K.; Chan, K.; Wang, H. Metal Ion Cycling of Cu Foil for
Selective C−C Coupling in Electrochemical CO2 Reduction. Nat.
Catal. 2018, 1, 111−119.
(12) Zhou, Y.; Che, F.; Liu, M.; Zou, C.; Liang, Z.; De Luna, P.;
Yuan, H.; Li, J.; Wang, Z.; Xie, H.; Li, H.; Chen, P.; Bladt, E.;
Quintero-Bermudez, R.; Sham, T. K.; Bals, S.; Hofkens, J.; Sinton, D.;
Chen, G.; Sargent, E. H. Dopant-Induced Electron Localization
Drives CO2 Reduction to C2 Hydrocarbons. Nat. Chem. 2018, 10,
974−980.
(13) Dinh, C.; Burdyny, T.; Kibria, G.; Seifitokaldani, A.; Gabardo,
C. M.; García de Arquer, F. P.; Kiani, A.; Edwards, J. P.; De Luna, P.;
Bushuyev, O. S.; Zou, C.; Quintero-Bermudez, R.; Pang, Y.; Sinton,
D.; Sargent, E. H. CO2 Electroreduction to Ethylene via Hydroxide-
Mediated Copper Catalysis at an Abrupt Interface. Science 2018, 360,
783−787.
(14) Wang, G.; He, C. T.; Huang, R.; Mao, J.; Wang, D.; Li, Y.
Photoinduction of Cu Single Atoms Decorated on UiO-66-NH2 for
Enhanced Photocatalytic Reduction of CO2 to Liquid Fuels. J. Am.
Chem. Soc. 2020, 142, 19339−19345.
(15) Wang, Y.; Chen, Z.; Han, P.; Du, Y.; Gu, Z.; Xu, X.; Zheng, G.
Single-Atomic Cu with Multiple Oxygen Vacancies on Ceria for
Electrocatalytic CO2 Reduction to CH4. ACS Catal. 2018, 8, 7113−
7119.
(21) Jiang, M. P.; Huang, K. K.; Liu, J. H.; Wang, D.; Wang, Y.;
Wang, X.; Li, Z. Da; Wang, X. Y.; Geng, Z.-B.; Hou, X. Y.; Feng, S. H.
Magnetic-Field-Regulated TiO2 {100} Facets: A Strategy for C-C
Coupling in CO2 Photocatalytic Conversion. Chem. 2020, 6, 2335−
2346.
(22) Cheng, T.; Fortunelli, A.; Goddard, W. A. Reaction
Intermediates during Operando Electrocatalysis Identified from Full
Solvent Quantum Mechanics Molecular Dynamics. Proc. Natl. Acad.
Sci. U. S. A. 2019, 116, 7718−7722.
(23) Sheng, H.; Oh, M. H.; Osowiecki, W. T.; Kim, W.; Alivisatos,
A. P.; Frei, H. Carbon Dioxide Dimer Radical Anion as Surface
Intermediate of Photoinduced CO2 Reduction at Aqueous Cu and
CdSe Nanoparticle Catalysts by Rapid-Scan FT-IR Spectroscopy. J.
Am. Chem. Soc. 2018, 140, 4363−4371.
(24) Lee, S. Y.; Jung, H.; Kim, N. K.; Oh, H. S.; Min, B. K.; Hwang,
Y. J. Mixed Copper States in Anodized Cu Electrocatalyst for Stable
and Selective Ethylene Production from CO2 Reduction. J. Am. Chem.
Soc. 2018, 140, 8681−8689.
(25) Chou, T. C.; Chang, C. C.; Yu, H. L.; Yu, W. Y.; Dong, C. L.;
Velasco-Vélez, J. J.; Chuang, C. H.; Chen, L. C.; Lee, J. F.; Chen, J.
M.; Wu, H. L. Controlling the Oxidation State of the Cu Electrode
and Reaction Intermediates for Electrochemical CO2 Reduction to
Ethylene. J. Am. Chem. Soc. 2020, 142, 2857−2867.
́
(26) Aran-Ais, R. M.; Scholten, F.; Kunze, S.; Rizo, R.; Roldan
Cuenya, B. The Role of in Situ Generated Morphological Motifs and
Cu(I) Species in C2+ Product Selectivity during CO2 Pulsed
Electroreduction. Nat. Energy 2020, 5, 317−325.
(27) Lum, Y.; Ager, J. W. Evidence for Product-Specific Active Sites
on Oxide-Derived Cu Catalysts for Electrochemical CO2 Reduction.
Nat. Catal. 2019, 2, 86−93.
(28) Hohenberg, P.; Kohn, W. Inhomogeneous Electron Gas. Phys.
Rev. 1964, 136, B864−B871.
(29) Kresse, G.; Hafner, J. Ab initio molecular dynamics for liquid
metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1993, 47, 558−561.
(30) Damia, G.; Rossanda, M. Prime Esperienze Con Fluothane in
Chirurgia Generale. Minerva Anestesiol. 1959, 25, 19−25.
(31) Kresse, G.; Furthmuller, J. Efficiency of ab-initio total enery
calculations for metals and semiconductors using a plane-wave basis
set. Comput. Mater. Sci. 1996, 6, 15−50.
(32) Yiu, G. Rapid Communications: Antiperspirant Induced DNA
Damage in Canine Cells by Comet Assay. Toxicol. Mech. Methods
2004, 15, 25−28.
(33) Lee, K.; Lundqvist, B. I. Higher-Accuracy van der Waals
Density Functional. Phys. Rev. B: Condens. Matter Mater. Phys. 2010,
82, 081101.
̌
(34) Klimes, J.; Bowler, D. R.; Michaelides, A. Van Der Waals
Density Functional Applied to Solids. Phys. Rev. B: Condens. Matter
Mater. Phys. 2011, 83, 195131.
(35) Yu, X.; Zhang, X. High Coverage Water Adsorption on CuO
(011) Surface. Phys. Chem. Chem. Phys. 2017, 19, 18652−18659.
(36) Sun, B. Z.; Chen, W. K.; Xu, Y. J. Reaction Mechanism of CO
Oxidation on Cu2O(111): A Density Functional Study. J. Chem. Phys.
2010, 133, 154502.
(37) Maimaiti, Y.; Nolan, M.; Elliott, S. D. Reduction Mechanisms
of the CuO (111) Surface through Surface Oxygen Vancancy
Formation and Hydrogen Adsorption. Phys. Chem. Chem. Phys.
2014, 16, 3036−3046.
(38) Zhang, J.; Zhang, R.; Wang, B.; Ling, L. Insight into the
Adsorption and Dissociation of Water over Different CuO(111)
Surfaces: The Effect of Surface Structures. Appl. Surf. Sci. 2016, 364,
758−768.
(39) Li, Q.; Fu, J.; Zhu, W.; Chen, Z.; Shen, B.; Wu, L.; Xi, Z.;
Wang, T.; Lu, G.; Zhu, J. J.; Sun, S. Tuning Sn-Catalysis for
Electrochemical Reduction of CO2 to CO via the Core/Shell Cu/
SnO2 Structure. J. Am. Chem. Soc. 2017, 139, 4290−4293.
(40) Tian, Z. R.; Voigt, J. A.; Liu, J. U. N.; Mckenzie, B.; Mcdermott,
M. J.; Rodriguez, M. A.; Konishi, H.; Xu, H. Complex and Oriented
ZnO Nanostructures. Nat. Mater. 2003, 2, 821−826.
(16) Yuan, L.; Hung, S. F.; Tang, Z. R.; Chen, H. M.; Xiong, Y.; Xu,
Y. J. Dynamic Evolution of Atomically Dispersed Cu Species for CO2
Photoreduction to Solar Fuels. ACS Catal. 2019, 9, 4824−4833.
(17) Xu, S.; Carter, E. A. Theoretical Insights into Heterogeneous
(Photo)Electrochemical CO2 Reduction. Chem. Rev. 2019, 119,
6631−6669.
(18) Ozin, G. A. Throwing New Light on the Reduction of CO2.
Adv. Mater. 2015, 27, 1957−1963.
(19) Wu, J.; Huang, Y.; Ye, W.; Li, Y. CO2 Reduction: From the
Electrochemical to Photochemical Approach. Adv. Sci. 2017, 4,
1700194.
(20) Xiao, H.; Goddard, W. A.; Cheng, T.; Liu, Y. Cu Metal
Embedded in Oxidized Matrix Catalyst to Promote CO2 Activation
and CO Dimerization for Electrochemical Reduction of CO2. Proc.
Natl. Acad. Sci. U. S. A. 2017, 114, 6685−6688.
2992
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(41) Bersani, M.; Gupta, K.; Mishra, A. K.; Lanza, R.; Taylor, S. F.
R.; Islam, H.; Hollingsworth, N.; Hardacre, C.; de Leeuw, N. H.; Darr,
J. A. Combined EXAFS, XRD, DRIFTS, and DFT Study of Nano
Copper-Based Catalysts for CO2 Hydrogenation. ACS Catal. 2016, 6,
5823−5833.
(42) Velu, S.; Suzuki, K.; Gopinath, C. S.; Yoshida, H.; Hattori, T.
XPS, XANES and EXAFS Investigations of CuO/ZnO/Al2O3/ZrO2
Mixed Oxide Catalysts. Phys. Chem. Chem. Phys. 2002, 4, 1990−1999.
(43) Kim, J. Y.; Rodriguez, A.; Hanson, J. C.; Frenkel, A. I.; Lee, P.
L. Reduction of CuO and Cu2O with H2: H Embedding and Kinetic
Effects in the Formation of Suboxides. J. Am. Chem. Soc. 2003, 125,
10684−10692.
(59) Yu, X.; Roth, J. P.; Wang, J.; Sauter, E.; Nefedov, A.; Heißler, S.;
̈
Pacchioni, G.; Wang, Y.; Woll, C. Chemical Reactivity of Supported
ZnO Clusters: Undercoordinated Zinc and Oxygen Atoms as Active
Sites. ChemPhysChem 2020, 21, 2553−2564.
(60) Perez-Gallent, E.; Marcandalli, G.; Figueiredo, M. C.; Calle-
Vallejo, F.; Koper, M. T. M. Structure- and Potential-Dependent
Cation Effects on CO Reduction at Copper Single-Crystal Electrodes.
J. Am. Chem. Soc. 2017, 139, 16412−16419.
(44) Zhang, W.; Huang, C.; Xiao, Q.; Yu, L.; Shuai, L.; An, P.;
Zhang, J.; Qiu, M.; Ren, Z.; Yu, Y. Atypical Oxygen-Bearing Copper
Boosts Ethylene Selectivity toward Electrocatalytic CO2 Reduction. J.
Am. Chem. Soc. 2020, 142, 11417−11427.
(45) Kau, L.; Spira-Solomon, D. J.; Penner-Hahn, J. E.; Hodgson, K.
O.; Solomon, E. I. X-ray Absorption Edge Determination of the
Oxidation State and Coordination Number of Copper: Application to
the Type 3 Site in Rhus Vernicifera Lacease and Its Reaction with
Oxygen. J. Am. Chem. Soc. 1987, 109, 6433−6442.
(46) Macedo, L. J. A.; Hassan, A.; Sedenho, G. C.; Crespilho, F. N.
Assessing electron transfer reactions and catalysis in multicopper
oxidases with operando X-ray absorption spectroscopy. Nat. Commun.
2020, 11, 316.
(47) Kattel, S.; Ramírez, P.; Chen, J.; Rodriguez, J.; Liu, P. Active
Sites for CO2 Hydrogenation to Methanol on Cu/ZnO Catalysts.
Science 2017, 355, 1296−1299.
(48) Wu, J. C. S.; Huang, C. W. In Situ DRIFTS Study of
Photocatalytic CO2 Reduction under UV Irradiation. Front. Chem.
Eng. China 2010, 4, 120−126.
(49) Nomura, N.; Tagawa, T.; Goto, S. In Situ FTIR Study on
Hydrogenation of Carbon Dioxide over Titania-Supported Copper
Catalysts. Appl. Catal., A 1998, 166, 321−326.
(50) Corson, E. R.; Kas, R.; Kostecki, R.; Urban, J.; Smith, W. A.;
Mccloskey, B. D.; Kortlever, R. In Situ ATR - SEIRAS of Carbon
Dioxide Reduction at a Plasmonic Silver Cathode. J. Am. Chem. Soc.
2020, 142, 11750−11762.
(51) Lee, B. J.; Kuo, M. C.; Chien, S. H. In Situ FT-IR Studies of
NO Decomposition on Pt/TiO2 Catalyst under UV Irradiation. Res.
Chem. Intermed. 2003, 29, 817−826.
(52) Perez-Gallent, E.; Figueiredo, M. C.; Calle-Vallejo, F.; Koper,
M. T. Spectroscopic Observation of a Hydrogenated CO Dimer
Intermediate During CO Reduction on Cu(100) Electrodes. Angew.
Chem., Int. Ed. 2017, 56, 3621−3624.
(53) Abdel-Mageed, A. M.; Klyushin, A.; Rezvani, A.; Knop-Gericke,
A.; Schlogl, R.; Behm, R. J. Methanol Synthesis Negative Charging of
Au Nanoparticles during Methanol Synthesis from CO2/H2 on a Au/
ZnO Catalyst: Insights from Operando IR and Near-Ambient-
Pressure XPS and XAS Measurements. Angew. Chem., Int. Ed. 2019,
58, 10325−10329.
(54) Kampshoff, E.; Hahn, E.; Kern, K. Correlation Betvveen Surface
Stress and the Vibrational Shift of CO Chemisorbed on Cu Surfaces
Elisabeth. Phys. Rev. Lett. 1994, 73, 704−707.
(55) Hollins, P.; Pritchard, J. Infrared Studies of Chemisorbed
Layers on Single Crytals. Prog. Surf. Sci. 1985, 19, 275−349.
(56) Nielsen, N. D.; Smitshuysen, T. E. L.; Damsgaard, C. D.;
Jensen, A. D.; Christensen, J. M. Characterization of Oxide-Supported
Cu by Infrared Measurements on Adsorbed CO. Surf. Sci. 2021, 703,
121725.
(57) Sheng, H.; Zhang, H.; Song, W.; Ji, H.; Ma, W.; Chen, C.;
Zhao, J. Activation of Water in Titanium Dioxide Photocatalysis by
Formation of Surface Hydrogen Bonds: An In Situ IR Spectroscopy
Study. Angew. Chem., Int. Ed. 2015, 54, 5905−5909.
(58) Kim, Y.; Park, S.; Shin, S.-J.; Choi, W.; Min, B. K.; Kim, H.;
Kim, W.; Hwang, Y. J. Time-Resolved Observation of C−C Coupling
Intermediates on Cu Electrodes for Selective Electrochemical CO2
Reduction. Energy Environ. Sci. 2020, 13, 4301−4311.
2993
https://dx.doi.org/10.1021/jacs.1c00206
J. Am. Chem. Soc. 2021, 143, 2984−2993