Ultrasmall C-TiO2?x nanoparticle/g-C3N4 composite for CO2 photoreduction with high efficiency and selectivity

Article abstract of DOI:10.1039/c8ta08091g

The photoreduction of CO₂ to CO offers a promising sustainable and clean approach for a global new energy program. Coupling this reductive process with a matched water photo-oxidation pathway is an attractive avenue to accelerate the half-reaction of CO₂ reduction. Herein, we propose a three-component photocatalyst design strategy for reducing CO₂ to CO coupled with water oxidation via a two-electron/two-step pathway. Employing polyoxotitanium ([Ti₁₇O₂₄(OPrⁱ)₂₀]) as a titanium source, ultrasmall TiO_2?x nanoparticles coated with ultrathin carbon layers (C-TiO_2?x) were fabricated and loaded on to a g-C₃N₄ matrix through chemical bonding (C-TiO_2?x@g-C₃N₄) for the first time. The optimized C-TiO_2?x@g-C₃N₄ photocatalyst showed a very high activity of 12.30 mmol g^?1 (204.96 mmol g_TiO2^?1) CO generation within 60 h visible-light irradiation, which represents the highest CO production rate to date among the reported TiO₂-based materials under similar conditions. The excellent adsorption capability of C-TiO_2?x@g-C₃N₄ for photons, H⁺ protons, and CO₂ molecules together with efficient charge separation and the two-electron/two-step oxidative pathway lead to the high reactivity.

Full text of DOI:10.1039/c8ta08091g

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Ultrasmall C-TiO2ꢀx nanoparticle/g-C3N4
composite for CO2 photoreduction with high
efficiency and selectivity†
Cite this: J. Mater. Chem. A, 2018, 6,
21596
a
a
a
a
Jie Zhou, Han Wu, Chun-Yi Sun, Cheng-Ying Hu,a Xin-Long Wang,
*
*
b
a
*
Zhen-Hui Kang
and Zhong-Min Su
The photoreduction of CO2 to CO offers a promising sustainable and clean approach for a global new
energy program. Coupling this reductive process with a matched water photo-oxidation pathway is an
attractive avenue to accelerate the half-reaction of CO2 reduction. Herein, we propose a three-
component photocatalyst design strategy for reducing CO2 to CO coupled with water oxidation via
a two-electron/two-step pathway. Employing polyoxotitanium ([Ti17O24(OPri)20]) as a titanium source,
ultrasmall TiO2ꢀx nanoparticles coated with ultrathin carbon layers (C-TiO2ꢀx) were fabricated and
loaded on to a g-C3N4 matrix through chemical bonding (C-TiO2ꢀx@g-C3N4) for the first time. The
optimized C-TiO2ꢀx@g-C3N4 photocatalyst showed a very high activity of 12.30 mmol gꢀ1 (204.96 mmol
Received 20th August 2018
Accepted 11th October 2018
gTiO ꢀ1) CO generation within 60 h visible-light irradiation, which represents the highest CO production
2
rate to date among the reported TiO2-based materials under similar conditions. The excellent adsorption
capability of C-TiO2ꢀx@g-C3N4 for photons, H+ protons, and CO2 molecules together with efficient
charge separation and the two-electron/two-step oxidative pathway lead to the high reactivity.
DOI: 10.1039/c8ta08091g
rsc.li/materials-a
oxidation pathway is a promising avenue to boost the reductive
half-reaction activity.9 For water oxidation, the challenge mainly
Introduction
lie in O2 release, which usually involves a four-proton coupled
four-electron transfer process.10 Although the four-electron
process is thermodynamically more favorable than the two-
electron way, two electrons participating in the process is
kinetically favored.11 A stepwise two-electron/two-step pathway
is proposed as an optimum approach between the thermody-
namic and kinetic constraints for water oxidation.9 A signicant
issue toward achieving the stepwise pathway is to incorporate
a co-catalyst to decompose H2O2 to O2 and H2O.
As a classical photocatalyst, titanium dioxide (TiO2) has been
widely used in photocatalysis.12–14 Beneting from its excellent
stability, cheap, and favorable band edges, TiO2 is also
a popular semiconductor photocatalyst in CO2 reduction.15 It is
worth noting that TiO2 is one of the few photocatalysts able to
reduce CO2 into CO under H2O conditions, albeit with a low
efficiency (below several micromole).15–18 Signicant efforts have
been devoted to enhancing the performance of TiO2, mainly
focused on increasing the harvesting of visible light and charge
separation.15 Unfortunately, the rate of reduction of CO2 to CO
is still limited to tens of micromoles per hour.19 For CO2
photoreduction, the current TiO2-based catalysts still face
several challenges: (i) a low proportion of active surface area
(random aggregates of nanoparticles or crystals with a large
size); (ii) poor stability of the doping and sensitization agents;
(iii) negligible hydrogen proton adsorption, which is necessary
to trigger CO2 reduction reactions via a proton-coupled
Increasing fossil fuel consumption in human activities has led
to a global energy crisis and high emissions of carbon dioxide,
which contribute more than 60% to global warming.1–3
Sustainable CO2 reduction into useful chemicals with solar
energy as a power source is regarded as an economic, safe, and
clean avenue to combat these issues simultaneously.4–6 As one
of the main products from CO2 reduction, CO has aroused great
attention because it is regarded as one of the most valuable raw
materials of syngas.7,8 As we know, syngas is an important
feedstock for the generation of synthetic fuels and industrial
chemicals, potentially replacing traditional production tech-
nology that usually relies on the reformation of fossil fuels.8
Therefore, exploring a cheap, efficient, and durable photo-
catalytic system for converting CO2 into CO is crucial for
a global new energy program in a clean and sustainable way.
Coupling the CO2 reductive process with a matched water
aNational & Local United Engineering Laboratory for Power Batteries, Key Laboratory
of Polyoxometalate Science of Ministry of Education, Department of Chemistry,
Northeast Normal University, Changchun, 130024 Jilin, People's Republic of China.
E-mail: suncy009@nenu.edu.cn; wangxl824@nenu.edu.cn
bJiangsu Key Laboratory for Carbon-Based Functional Materials & Devices, Institute of
Functional Nano & So Materials (FUNSOM), Soochow University, 199 Ren'ai Road,
Suzhou 215123, Jiangsu, China. E-mail: zhkang@suda.edu.cn
† Electronic supplementary information (ESI) available: Materials and methods,
experimental details including synthesis, experimental procedure and
supporting data (Fig. S1–S14 and Table S1). See DOI: 10.1039/c8ta08091g
21596 | J. Mater. Chem. A, 2018, 6, 21596–21604
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
pathway; (iv) an insufficient oxidation pathway. Herein, we In addition, its excellent proton-accumulation capability (300
propose a design strategy involving a highly durable TiO2-based mmol gꢀ1 minꢀ1) provided a reliable guarantee for CO2 reduc-
three-component photocatalyst for the efficient and selective tion through a low-energy proton-coupled pathway. Under the
generation of CO with H2O and visible-light irradiation, assistance of the carbon layer, water oxidation was performed
comprising coupling an active CO2 reductive photocatalyst with via an efficient two-electron/two-step pathway. The thin carbon
an efficient water oxidation catalyst via a two-electron/two-step layer served as an electron reservoir and played a crucial role in
oxidative pathway. For water oxidation, we adopted a low-cost facilitating the generated electron–hole (e–h+) pairs separation.
graphitic carbon nitride (g-C3N4) coupled amorphous carbon
layer as the catalyst. By cooperating with carbon dots, this
material has exhibited excellent performance in water splitting
via a two-electron pathway.11 Furthermore, the stable meso-
Experimental section
Synthesis procedure of the photocatalysts
All the reagents and solvents used in the synthesis and photo-
catalysis were purchased from commercial sources and used
without pretreatment.
porous material of g-C3N4 is a good substrate for the dispersion
of TiO2 nanoparticles and as an adsorbent for CO2 molecules20
and H+ protons. An ultrasmall black TiO2 (TiO2ꢀx) nanoparticle-
coated carbon layer was designed for the photocatalyst for
converting CO2 to CO, based on the following consideration: (i)
TiO2ꢀx possesses excellent visible-light adsorption ability and
could be a rich electron source for the reduction reaction; (ii) its
ultrasmall size guarantees a large reactive surface area; (iii) the
carbon coating could serve as an electron reservoir to promote
electron transfer and boost CO2 chemisorption.21 Further, the
spatial separation of reductive and oxidative sites may exclude
the possibility of re-oxidation of reduction products.
Synthesis of C-TiO2ꢀx
To prepare C-TiO2ꢀx, [Ti17O24(OPri)20] was synthesized rst.
Titanium isopropoxide (8.4 mL, 0.0285 mol) was added to acetic
acid (2.1 mL, 0.037 mol). Then, the mixture was transferred and
ꢁ
sealed in a Teon reactor with 23 mL capacity at 150 C over 5
days. Aer cooling to room temperature, a white powder of
[Ti17O24(OPri)20] was obtained with a yield of 38.6%.24 C-TiO2ꢀx
was prepared by adding [Ti17O24(OPri)20] (200 mg) into diac-
etone (10 mL) and then heating to 80 ꢁC until the solution
became clear. Then the temperature was kept at 60 ꢁC to
evaporate most diacetone to form a slurry. The slurry was
heated to 350 ꢁC with a heating rate of 10 ꢁC minꢀ1 in a muffle
furnace with a cap for 2 h. Aer cooling to room temperature,
black powder was obtained.
During this study, polyoxotitanium [Ti17O24(OPri)20]22 was
selected as a precursor of TiO2 for its carbon-rich and small-
sized-cluster structure. Through heating a mixture of [Ti17-
O24(OPri)20] and g-C3N4, the targeted three-component
composite C-TiO2ꢀx@g-C3N4 was obtained, in which TiO2ꢀx
nanoparticles (5–8 nm) were coated by ultrathin carbon layers.
The resultant C-TiO2ꢀx@g-C3N4 material exhibited signicant
adsorption capabilities for photons, CO2, and H+ protons. CO2
photoreduction experiments under visible-light irradiation in
water vapor showed around 12.30 mmol gꢀ1 (204.96 mmol
Fabrication of the C-TiO2ꢀx@g-C3N4 composites
The used g-C3N4 was obtained using
a thermal poly-
condensation method, which was prepared by heating urea in
an aluminaꢀc1rucible with a cover to 550 ꢁC with a heating rate of
10 ꢁC min in a muffle furnace for 2 h. Aer allowing to
naturally cool to room temperature, g-C3N4 with a faint yellow
color was obtained and then ground into powder. The C-
TiO2ꢀx@g-C3N4 composite photocatalyst was prepared as
follows: rst, an appropriate amount of [Ti17O24(OPri)20] was
added into diacetone solution and heated to 80 ꢁC until the
solution became clear. Then, the right amount of g-C3N4 in
acetonitrile solution was added. Aer ultrasonication for
60 min, the above mixture was transferred into an alumina
gTiO ꢀ1) CO generation within 60 h reaction (Fig. 1). This
2
generation rate is one order higher than most TiO2-based
catalyst and presents the highest amount of CO production to
date among the reported TiO2-based materials under similar
conditions. Mechanism studies revealed that the reductive and
oxidative sites were spatially separated on C-TiO2ꢀx and g-C3N4.
ꢁ
crucible with a cover under stirring. A water bath of 60 C was
used to evaporate the solvent to form a slurry. Finally, the
mixture was heated to 350 ꢁC with a heating rate of 10 ꢁC minꢀ1
in a muffle furnace for 2 h. A brown or grey solid was obtained
and then ground into powder.
Photocurrent measurements
Photocurrent measurements were performed on an electro-
chemical workstation (CHI660e, CH Instruments, Inc., USA)
with a three-electrode conguration. Our sample, Ag/AgCl
electrode, and Pt-wire electrode were employed as the
working, reference, and counter electrodes, respectively. Irra-
diation was carried out using a 300 W xenon lamp. The Na2SO4
Fig. 1 Schematic illustration on the photocatalytic process in the
system of C-TiO2ꢀx@g-C3N4.
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A, 2018, 6, 21596–21604 | 21597

View Article Online
Journal of Materials Chemistry A
Paper
solution (0.1 M) was used as the electrolyte. The working elec- was 1.9 min. To detect the formation of hydrogen from the reac-
trodes (1 cm ꢂ 1 cm) were prepared by spreading aqueous tion mixture, 1000 mL of the headspace of the test tube was taken
˚
slurries of the various samples on an ITO glass substrate.
out by syringe and injected into a TCD detector, using a 5 A
molecular sieve column and argon as the carrier gas and reference
gas. The volume of the hydrogen produced was calculated by
comparing the integrated area of the signals of hydrogen with
a calibration curve. The injector and detector temperatures were
set to 60 ꢁC. The retention time of hydrogen was 0.7 min.
Electrochemical impedance measurements
Electrochemical impedance spectroscopy was performed on g-
C3N4, C-TiO2ꢀx, and C-TiO2ꢀx@g-C3N4-3. To obtain anodes
materials, the above samples were mixed with conductive
carbon black (CB) and poly(vinylidene uoride) (PVDF),
respectively, with a mass ratio of 7 : 2 : 1. Then, the obtained
anodes materials were dispersed uniformly into N-methyl-2-
pyrrolidinone (NMP). Aer that, a paste with appropriate
viscosity was formed. Then, the paste was coated on copper
akes mildly and vacuum-dried at 50 ꢁC for 24 h. A button cell
was assembled with the sample anode, Li cathode, and
approximately 50 mL typical electrolyte, here 1 M LiPF6 in diethyl
carbonate (DEC) and ethylene carbonate (EC) with a ratio of
50 : 50 (w/w). The electrochemical workstation produced by
Princeton Applied Research (Germany) was employed to record
the electrochemical impedance spectroscopy of the batteries at
constant ambient temperature. The frequency limits were set at
105 to 10ꢀ2 Hz with an amplitude of 5 mV.
Electron-transfer number (n) calculation
The electron-transfer number (n) was calculated as follows:
n ¼ 4Idisk/(Idisk + Iring/N)
where N is the RRDE collection efficiency, measured to be 0.4
using the system by Liu et al.9 In detail, the RRDE electrode was
dipped into 0.01 mol Lꢀ1 K3[Fe(CN)6] in 0.1 mol Lꢀ1 KCl solu-
tion and rotated under different rotation rates (u ¼ 400, 900,
1600 2500 rpm). The disk potential Edisk was scanned from 0.2 V
to ꢀ1.0 V vs. Ag/AgCl at a scan rate of 10 mV sꢀ1, with the ring
potential Ering xed to 0.454 V vs. Ag/AgCl (the reduced product
[Fe(CN)6]4ꢀ can be oxidized at this potential), and the current–
potential voltammograms were then recorded during the elec-
trode rotation. The RRDE measurement is shown in Fig. S13
and S14.† The ratio of Iring/Idisk was almost constant under
various u. The calculation method of the electron-transfer
number (n) was also considered by Liu et al.9 The result of the
calculation was that n equals 2.
Cyclic voltammetry
Cyclic voltammetry tests of g-C3N4, C-TiO2ꢀx@g-C3N4-3, and C-
TiO2ꢀx were performed using an electrochemical workstation
(CHI 760e, China) with a three-electrode system. The working
electrode was a GC electrode (d ¼ 3 mm), with platinum wire as
the auxiliary electrode and an Ag/AgCl electrode as the reference
electrode. First, 5 mg of the samples and 5 mg of carbon black
Results and discussion
(Vulcan XC-72R) were dispersed in 0.6 mL of solution contain- Synthesis and structural analysis of the C-TiO2ꢀx@g-C3N4
ing 300 mL of 0.5 wt% Naon solution, 150 mL of water and 150 composite
mL of ethanol, followed by sonication for 60 min until
For the synthesis of the targeted three-component photocatalyst,
a suspension was formed. Aer sonication, 5 mL of the as-
g-C3N4 (ref. 23) and [Ti17O24(OPri)20]22,24 were rst prepared
prepared suspension was drop-cast onto the GC electrode.
following the reported literature. Through one-pot pyrolyzing the
Then, the working electrode was dried spontaneously under
mixtures of [Ti17O24(OPri)20] and g-C3N4 with different weight
ratios under 350 ꢁC and an air atmosphere, the three-component
composites of C-TiO2ꢀx@g-C3N4 with different contents of C-
ambient temperature. The nal mass loading of the catalysts
was about 0.59 mg cmꢀ2. The electrochemical measurements
were tested in methyl cyanide solution.
TiO2ꢀx were fabricated (Fig. 2). Bare C-TiO2ꢀx was obtained by
directly heating [Ti17O24(OPri)20] under similar conditions. Their
Photocatalytic activity measurements
photographs are shown in Fig. 3g. As determined by inductively
The as-prepared C-TiO2ꢀx@g-C3N4 catalyst (1 mg) was dispersed coupled plasma mass spectrometry (ICP-MS), the corresponding
in a mixture of 2 mL deionized water and ethanol (v/v ¼ 1 : 1) TiO2ꢀx content in the resultant C-TiO2ꢀx@g-C3N4 composites was
under sonication. Aer heating to 70 ꢁC, the suspension was drop- respectively 1 wt%, 4.5 wt%, 6 wt%, and 10 wt% based on the Ti
coated onto a glass of 1 cm ꢂ 2.5 cm. Then, the glass was moved content and the samples were labeled as C-TiO2ꢀx@g-C3N4-1, C-
to a quartz photoreactor and 0.1 mL deionized water was added to TiO2ꢀx@g-C3N4-2, C-TiO2ꢀx@g-C3N4-3, and C-TiO2ꢀx@g-C3N4-4.
keep the glass surrounded by water vapor. Aer bubbling through The powder X-ray diffraction (PXRD) patterns of the as-prepared
with water-vapor-saturated CO2 for 30 min, the quartz photo- g-C3N4, C-TiO2ꢀx, and C-TiO2ꢀx@g-C3N4 composites (Fig. 3f)
reactor was put into a water-cooled reactor and irradiated under showed that the pure C-TiO2ꢀx possessed similar patterns with
visible light using a PLS-SXE300 Xe lamp with a 420 nm cutoff anatase TiO2. C-TiO2ꢀx@g-C3N4 composites exhibited the char-
lter. To detect the gas product of CO, 500 mL gas in the middle of acteristic PXRD peaks of both g-C3N4 and C-TiO2ꢀx, with the
the test tube was taken out by syringe and injected into FID intensity of the characteristic peaks of C-TiO2ꢀx increasing with
detector, using argon as carrier gas. The volume of the CO the enhancement of C-TiO2ꢀx content (Fig. S1†).
produced was calculated by comparing the integrated area of the
SEM and TEM images of C-TiO2ꢀx@g-C3N4 composite were
signals of CO with a calibration curve. The injector and detector taken to directly analyze its structure (Fig. 3a–d). As shown in
ꢁ
temperatures were set to 60 C. The retention time of hydrogen the SEM image (Fig. 3a), the layered structure of g-C3N4 was
21598 | J. Mater. Chem. A, 2018, 6, 21596–21604
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
could be an efficient and effective method to obtain ultrasmall
TiO2ꢀx nanoparticle-based composite materials. High-
resolution TEM (HRTEM) was employed to identify the phase
of the TiO2ꢀx nanoparticles in the C-TiO2ꢀx@g-C3N4 composite.
As shown in Fig. 3d, the distance between the adjacent lattice
planes was 0.35 nm along [101] direction, which is a typical
lattice plane distance for anatase TiO2.20 The structure of the
TiO2ꢀx-coated amorphous carbon layer can be clearly observed
from Fig. 3e. Furthermore, the carbon layer was the link con-
necting g-C3N4 and TiO2ꢀx nanoparticles. For the electron pool
character of the carbon layer, such a structure will facilitate
photoinduced electron transfer. In order to illuminate the
source of amorphous carbon, an HRTEM image of bare C-
TiO2ꢀx was collected. A clear carbon coating around TiO2ꢀx
nanoparticles was found, which strongly excluded the possi-
bility of the carbon source being from g-C3N4.
X-ray photoelectron spectroscopy (XPS) was further
employed to determine the atomic valence states and bonding
environment in C-TiO2ꢀx and C-TiO2ꢀx@g-C3N4. Fig. 3 shows
the XPS of C-TiO2ꢀx, in which the high-resolution spectra of Ti
2p located at 465.0, 464.2, 459.2, and 458.7 eV could be assigned
to Ti4+ 2p1/2, Ti3+ 2p1/2, Ti4+ 2p3/2, and Ti3+ 2p3/2, respectively.25
The C 1s had three peaks at 284.7, 286.1, and 288.5 eV, corre-
sponding to C–C, C–O, and C–O–Ti,26 respectively, which
provided strong evidence for TiO2ꢀx connecting to the carbon
layer with chemical bonds. The O 1s had two peaks located at
531.2 eV and 529.5 eV, attributing to the O–H and Ti–O bonds,27
respectively. The formation of Ti3+ and C might originate from
the reduction effect of H2 and CO28 gas, which may come from
the hydrolysis and decomposition of diacetone under high
temperature. Control experiments were carried out to illumi-
nate these issues. Without diacetone, only white TiO2 was ob-
tained, implying the indispensable role of diacetone during the
formation of C-TiO2ꢀx. When using titanium isopropoxide to
replace [Ti17O24(OPri)20], yellow TiO2 emerged, revealing
Fig. 2 Schematic illustration of the fabrication of C-TiO2ꢀx@g-C3N4
composites. The cluster in green represents [Ti17O24(OPri)20], the layer
structure is representative of g-C3N4, the ball in yellow is on behalf of
TiO2ꢀx and the fullerene-like structure with grey shell is presented for
the carbon layer.
maintained in C-TiO2ꢀx@g-C3N4. Energy dispersive X-ray (EDX)
elemental mappings and the spectrum (Fig. 3b) of the selected
area showed that the Ti and O elements were highly dispersed
on the g-C3N4 surface. TEM images of C-TiO2ꢀx@g-C3N4 in
Fig. 3c reveal the size of TiO2ꢀx nanoparticles were mainly
distributed between 6 and 9 nm. Compared with the 590 nm
particle size of TiO2 in TiO2@g-C3N4 (Fig. S2†), which was
synthesized by a traditional hydrothermal method,13 the size of
TiO2ꢀx in C-TiO2ꢀx@g-C3N4 was reduced by two orders of
magnitude. Therefore, employing polyoxotitanate as a Ti source
Fig. 3 Morphology, microstructure, and optical property. (a–d) SEM, EDX elemental mappings, TEM with the size distribution of C-TiO2ꢀx and
HRTEM images of C-TiO2ꢀx@g-C3N4-3. (e) HRTEM images of C-TiO2ꢀx. (f) PXRD of g-C3N4, C-TiO2ꢀx@g-C3N4-3, C-TiO2ꢀx, and anatase TiO2.
(g) UV-visible adsorption spectra and photographs of C-TiO2ꢀx@g-C3N4 composites, C-TiO2ꢀx, and g-C3N4.
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A, 2018, 6, 21596–21604 | 21599

View Article Online
Journal of Materials Chemistry A
Paper
[Ti17O24(OPri)20] clusters may be the main source of amorphous 0.4 eV shied compared with the corresponding peak in pure g-
carbon. C3N4 (Fig. S5†), which might due to the formation of chemical
The XPS of C-TiO2ꢀx@g-C3N4 is presented in Fig. S3,† and bonds.31 The N 1s could be tted into four peaks, located at
the high-resolution XPS results of C 1s, N 1s, O 1s, and Ti 2p are 403.8, 400.8, 399.6, and 398.2 eV. The peak at 398.2 eV is
exhibited in Fig. 4a–d. Compared to the peaks of Ti 2p in pure C- assigned to sp2-hybridized aromatic N bound to C atoms (C]
TiO2ꢀx (Fig. S4†), the corresponding peaks in C-TiO2ꢀx@g-C3N4 N–C), while the signal at the binding energy of 399.6 eV indi-
are a little shied, centering around 464.2, 463.3, 458.3, and cates tertiary nitrogen N–(C)3. The peaks at 400.8 and 403.8 eV
457.8 eV. The O 1s of C-TiO2ꢀx@g-C3N4 has two peaks at 532.3 are assigned to C–N–H groups and charging effects.32
and 529.5 eV, ascribed to the O–H group and Ti–O bond.
Compared with C-TiO2ꢀx, the peak of O–H group is shied Adsorption study
a little, which should be attributed to the chemical environment
The N2 adsorption behavior was measured to reveal the change in
surface of the C-TiO2ꢀx@g-C3N4 composites. The adsorption–
desorption isotherms at 77 K (Fig. S6†) of C-TiO2ꢀx@g-C3N4-2, 3, 4
change.29 The C 1s XPS spectrum exhibits anisomerous peaks
with an obvious shoulder. Aer tting, ve peaks could be
observed with the binding energies of 289.9, 288.6, 287.6, 286.0,
and 284.6 eV. The peaks located at 287.6 and 286.0 eV are
showed a gradual decrease in BET surface areas (SBET) with values
of 94.1, 87.2, and 77.8 m2 gꢀ1, respectively, while the SBET of bare
attributed to hybridized carbon atoms containing the C–(N)3
g-C3N4 was relatively high (99.4, m2 gꢀ1). Compared with the SBET
and N–C–N group of g-C3N4, while the one at 284.6 eV is
of bare g-C3N4, the decrease in composite materials accounted for
the loading of C-TiO2ꢀx, whereby with the increase in loading
content, the decrement enhances. The CO2 adsorption ability of
C-TiO2ꢀx@g-C3N4 was also measured. As shown in Fig. 4e, the
assigned to the C–C coordination, originating from adventitious
carbon of the XPS instrument itself.30 The peak emerging at
288.6 eV is matched with the Ti–O–C. The le peak at 289.9 eV is
maximum CO2 uptake at 273 K for C-TiO2ꢀx@g-C3N4-2, 3, 4 was
17.7, 15.9, and 9.2 cm3 gꢀ1, respectively, which are different to
that of pure g-C3N4 (11.5 cm3 gꢀ1). In contrast to N2 adsorption,
the improvement in CO2 adsorption of C-TiO2ꢀx@g-C3N4-2 and 3
probably comes from the increased interaction of C-TiO2ꢀx with
CO2 molecules, while the decrease in C-TiO2ꢀx@g-C3N4-4 may be
due to the sharply decreased surface area.
As H+ protons have a signicant inuence on the CO2 reduc-
tion activity, the H+ proton adsorption capability of C-TiO2ꢀx@g-
C3N4 was studied with HCl as the H+ source. The H+ adsorption
curve of C-TiO2ꢀx@g-C3N4 is presented in Fig. 4f. It was found
that at the initial 5 min, 1.5 mmol gꢀ1 H+ was adsorbed by C-
TiO2ꢀx@g-C3N4, corresponding to 300 mmol gꢀ1 minꢀ1. Aer 3 h,
it reached saturated adsorption and the amount was 4.46 mmol
gꢀ1, which is 2.8-fold higher than bare g-C3N4 (1.55 mmol gꢀ1).
Such a high protons accumulation capability of C-TiO2ꢀx@g-C3N4
may signicantly boost the reductive performance of C-
TiO2ꢀx@g-C3N4 via the low-energy proton-coupled pathway.
Optical property
UV-Visible adsorption spectra can provide insightful informa-
tion into the interactions of photocatalytic materials with
photons of different energy. Fig. 3g shows the optical absorp-
tion spectra of g-C3N4, C-TiO2ꢀx and their composite materials.
Pure C-TiO2ꢀx has strong absorption within the whole visible
light, while bare g-C3N4 was only able to absorb light with lm
#
450 nm. All C-TiO2ꢀx@g-C3N4 composites showed an obvious
light absorption from 300 to 800 nm, and with increase in
doping concentration of TiO2ꢀx, the absorption intensity region
from 460 to 800 nm was improved. Therefore, we could expect
that the C-TiO2ꢀx@g-C3N4 may exhibit an outstanding perfor-
mance in visible-light photocatalysis.
Fig. 4 XPS, adsorption, and photoelectrical properties. (a–d) XPS of C-
TiO2ꢀx@g-C3N4. (e) CO2 adsorption of C-TiO2ꢀx@g-C3N4 composites
and g-C3N4. (f) H+ adsorption of C-TiO2ꢀx@g-C3N4-3, and g-C3N4. (g
and h), transient photocurrent responses and electrochemical
Electrochemical property
As for heterojunction semiconductor, the energy level of each
impedance spectroscopy of C-TiO2ꢀx@g-C3N4, g-C3N4, C-TiO2ꢀx
,
and TiO2.
component has
a signicant inuence on the catalytic
21600 | J. Mater. Chem. A, 2018, 6, 21596–21604
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
reactivity. The energy level of the synthesized g-C3N4 was well (hydrogen production is shown in Fig. S9†). Compared with
studied in our former work33 and the band gap was 2.73 eV, with pure g-C3N4 and C-TiO2ꢀx, a higher photocatalytic activity was
the conduction band energy (ECB) and the valence band energy achieved by C-TiO2ꢀx@g-C3N4 composites and the content of C-
(EVB) at ꢀ0.75 eV and +1.98 eV, respectively, vs. a normal TiO2ꢀx showed a considerable inuence on the photocatalytic
hydrogen electrode (NHE). For our materials, EVB was deter- activity: (I) the catalyst with 1% content of Ti (C-TiO2ꢀx@g-C3N4-
mined from cyclic voltammograms (CV, Fig. S7†) and ECB was 1) showed an obvious improvement in CO generation (up to
calculated based on the optical band gap (Eg).9 Using analyses 0.412 mmol gꢀ1), while H2 evolution was almost unchanged
and calculations based on Fig. S8,† the EVB energy of C-TiO2ꢀx (0.0181 mmol gꢀ1), (II) enhancing the amount of C-TiO2ꢀx to
was 1.63 eV. Eg was estimated from the Tauc plot (Fig. S8†), 6 wt% of Ti (C-TiO2ꢀx@g-C3N4-3) resulted in an evident
which was determined to be 2.18 eV. According to the formula enhancement in both yield and selectivity. CO generation was
ECB ¼ EVB ꢀ Eg, the ECB of C-TiO2ꢀx was estimated to be ꢀ0.55 V. around 5 and 15 times higher than that of bare C-TiO2ꢀx and g-
The energy levels of both g-C3N4 and C-TiO2ꢀx were appropriate C3N4, respectively, keeping in 2.460 mmol gꢀ1, while the yield of
for reducing CO2 to CO and water oxidation.
H2 was scarcely increased to about 0.0741 mmol gꢀ1, corre-
Given that charge separation has a signicant effect on the sponding to 97% selectivity of CO. When normalizing the
redox reaction occurring at a photocatalyst surface, the tran- quality of TiO2, the generation of CO in C-TiO2ꢀx@g-C3N4-3
sient photocurrent responses and electrochemical impedance remained at 40.995 mmol gTiO ꢀ1, (III) further enhancement of
2
spectroscopy (EIS) of C-TiO2ꢀx@g-C3N4-3, g-C3N4, and C-TiO2ꢀx the C-TiO2ꢀx content (C-TiO2ꢀx@g-C3N4-4) led to a decrease in
were collected.34,35 In Fig. 4g, the photocurrent value of C- CO generation, which might result from the decreased CO2
TiO2ꢀx@g-C3N4-3 can be seen to be rapidly elevated with the adsorption capacity. These results reveal that the combination
light turned on and then it gradually climbs up to the highest of C-TiO2ꢀx with g-C3N4 led to a superior enhancement in both
current level (up to 0.83 mA cmꢀ2) during irradiation. When the production and selectivity toward CO and showed that the
irradiation was turned off, a sharp decrease emerged. This optimal catalyst was C-TiO2ꢀx@g-C3N4-3.
switch loop could be implemented repeatedly multiple times. A
To illuminate the source of generated CO, isotopic experi-
similar phenomenon occurred with pure C-TiO2ꢀx and bare g- ments were carried out using 13CO2 as the gas atmosphere33
C3N4 materials during on–off irradiation; however, the photo- under similar photocatalytic conditions. As exhibited in Fig. 5g,
current densities of them were 4-fold and 2.5-fold lower than C- peaks at 1.88 min and 13.08 min with m/z 29 and 45 were
TiO2ꢀx@g-C3N4-3. Additionally, the photocurrent value of pure assigned to 13CO and 13CO2, respectively. No signal at m/z 28
TiO2 was 2-fold lower than that of C-TiO2ꢀx, illuminating that
the C doping plays a crucial role in facilitating the charge
separation generated by TiO2ꢀx
.
Under visible light irradiation, the EIS of C-TiO2ꢀx@g-C3N4-3
heterjunction, bare g-C3N4, and C-TiO2ꢀx and TiO2 were
measured. As shown in Fig. 4h, the charge-transfer resistance (Rct)
of C-TiO2ꢀx@g-C3N4-3 was 47 U cmꢀ2, which was considerably
lower than those of bare C-TiO2ꢀx (108 U cmꢀ2) and g-C3N4 (82 U
cmꢀ2). These results indicate that the C-TiO2ꢀx@g-C3N4-3
composite performed a higher separation efficiency of the pho-
togenerated charge carriers and a relatively lower recombination
rate of electron–hole pairs under visible light. Simultaneously, Rct
of pure TiO2 was 172 U cmꢀ2, which was 1.6-fold higher than C-
TiO2ꢀx. This further revealed that the carbon layer plays a positive
role in promoting photoinduced charge separation.
Photocatalytic CO2 reduction study
The CO2 reduction experiment was performed following the
photocatalytic condition using (Au, Cu)/TiO2 as the catalyst re-
ported by Neatu and co-worker, in which only light, CO2, and
H2O were used in the system.36 Our experiment was carried out
employing C-TiO2ꢀx@g-C3N4 as a photocatalyst under a visible-
light-driven system (l $ 420 nm) with CO2 atmosphere and
Fig. 5 Photocatalytic performance for CO generation. (a) Different
water vapor. As a control, the photocatalytic performances of catalysts under 12 h visible-light irradiation. (b) Different catalysts
under 12 h visible light or whole light irradiation. (c) The generation in
(b) after normalizing the quality of TiO2ꢀx. (d) Recycling experiments
bare g-C3N4 and C-TiO2ꢀx were investigated. Fig. 5a shows that
CO gas was generated as the main reductive reaction products
using C-TiO2ꢀx@g-C3N4-3 as catalyst. (e) Time course of the CO and
aer 12 h irradiation. Clearly, pure C-TiO2ꢀx showed around
H2 with the C-TiO2ꢀx@g-C3N4-3 as catalyst under 12 h irradiation. (f)
0.575 mmol gꢀ1 CO and 0.0252 mmol gꢀ1 H2 while bare g-C3N4
GC-Mass result of isotopic experiment under 13CO2 atmosphere. (g)
presented 0.164 mmol gꢀ1 CO and 0.0144 mmol gꢀ1 H2 The corresponding GC spectrum of (f).
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A, 2018, 6, 21596–21604 | 21601

View Article Online
Journal of Materials Chemistry A
Paper
was found (Fig. 5f). This information provided unambiguous
To further expose the photocatalytic ability of C-TiO2ꢀx@g-
evidence that the produced CO originated from the photore- C3N4, the CO2 reduction reaction was performed under the
duction of CO2 rather than from the decomposition of any other irradiation of whole light (Fig. 5b). Under irradiation of 12 h,
organic species in the system.
around 4.500 mmol gꢀ1 CO was produced by C-TiO2ꢀx@g-C3N4-
In order to further reveal the importance of the composite 3. The yields of C-TiO2ꢀx and g-C3N4 were 2.900 and 0.600 mmol
structure of C-TiO2ꢀx@g-C3N4 in CO2 photoreduction, control gꢀ1, respectively. These results suggest that in C-TiO2ꢀx@g-
experiments employing different TiO2-based composites as pho- C3N4, C-TiO2ꢀx may be responsible for the active CO2 reductive
tocatalysts were carried out under similar conditions (Table 1). site. For TiO2ꢀx, commercial P25, and TiO2@g-C3N4, the yields
When using commercial P25 TiO2 (entry 2) as the photocatalyst, were 0.800, 0.050, and 0.180 mmol gꢀ1, respectively. When
around 0.0301 mmol gꢀ1 CO was generated, which was about 20- normalizing the quality of TiO2, the generation of CO in C-
ꢀ1
fold lower than that of C-TiO2ꢀx. When reducing TiO2 to TiO2ꢀx
,
TiO2ꢀx@g-C3N4-3 stayed at 78.98 mmol gTiO
(Fig. 5c), which
2
the generation of CO was enhanced to 0.255 mmol gꢀ1 (entry 3); was around 70 times higher than that of bare TiO2ꢀx
however, this was still 2.3-fold lower than that of C-TiO2ꢀx. These The reactivity of the C-TiO2ꢀx@g-C3N4-3 catalyst in different
.
results implied the carbon layer played a positive part in TiO2 in time periods was revealed through detecting the amount of CO
CO2 reduction. With TiO2@g-C3N4 (entry 4) as the catalyst, made every other hour (Fig. 5e). No apparent CO was detected in the rst
from P25 and g-C3N4, the yield was higher than bare P25 but was hour, illustrating this photocatalytic system needed an activation
68 times lower than that from C-TiO2ꢀx@g-C3N4-3. Replacing P25 process for light excitation and adsorption equilibrium of the
with titanium isopropoxide as the starting material, the resultant starting materials on the catalyst. Aer that, the yield of CO per
TiO2@g-C3N4 (entry 5) exhibited a sharply increased reactivity, hour almost increased at a constant speed in a linear fashion.
with up to 0.274 mmol gꢀ1 CO generation. However, this reactivity
The stability of a photocatalyst is vital to its practical and
was one order of magnitude lower than that of C-TiO2ꢀx@g-C3N4- industrial application. The stability of C-TiO2ꢀx@g-C3N4-3 was
3. These results illuminated that the carbon layer and excellent studied through ve-run cycling photocatalytic experiments
light adsorption of TiO2ꢀx play a pivotal role in the photocatalysis. (Fig. 5d). Aer 60 h visible-light irradiation, almost no obvious
Using black TiO2 nanobelts-doped g-C3N4 (ref. 25) as the catalyst decrease was observed in CO generation, and the total production
(entry 6), 0.535 mmol gꢀ1 CO was produced, which was 4.6-fold amount reached 12.30 mmol gꢀ1, corresponding to 204.96 mmol
lower than that from C-TiO2ꢀx@g-C3N4-3, revealing a small size of gTiO ꢀ1. Aer catalysis, PXRD spectroscopies (Fig. S10†) and UV-
2
nanoparticles was indispensable. Besides these TiO2-based cata- visible adsorption spectra (Fig. S11†) of the sample were
lysts, the CO generation rate of C-TiO2ꢀx@g-C3N4 was one order collected, which revealed there was little change in the structure of
higher than most the reported TiO2-based catalysts and it out- the catalyst and thus is demonstrated a reliable photostability.
performed any other reported TiO2-based material under similar Furthermore, we also investigated the inuence of the amount of
conditions (Table S1†).25,37
C-TiO2ꢀx@g-C3N4-3 catalyst on the photocatalytic reactivity. When
increasing the catalyst amount to 5, 10, or 30 mg, nearly the same
catalytic activity to 1 mg was observed and their CO yields were
close at 2.39, 2.35, 2.51 mmol gꢀ1. These results demonstrate that
with the increase in C-TiO2ꢀx@g-C3N4-3 amount, the high reac-
tivity rate can still be maintained (Fig. S12†).
Table 1 Variation of the reference experiments conditionsa
Entry
CO (mmol gꢀ1
)
H2 (mmol gꢀ1
)
Sel. (%)b
Control experiments were performed to elucidate the
importance of the components in the photoreduction system.
Through mechanically mixing C-TiO2ꢀx (6 wt%) with g-C3N4
(entry 7), the yield was similar to bare g-C3N4. The signicant
decreases implied that a loose contact is insufficient for the
reduction of CO2. Without either C-TiO2ꢀx@g-C3N4-3 (entry 8)
or light (entry 9), no gas was detected, which indicated that the
other components cannot arouse CO2 reduction, and thus again
proved that the photogenerated charge pairs originated from C-
TiO2ꢀx@g-C3N4-3. Once CO2 gas was changed to Ar (entry 10),
the generation of CO also stopped, therefore further excluding
the source being from the degradation effects of organic
species. In the absence of vapor (entry 7), the yield of CO
dramatically decreased to 0.0997 mmol gꢀ1, implying vapor
might be essential as an electron/proton donor.
1c
2.4560
0.0301
0.2550
0.0365
0.2740
0.5350
0.1670
0.0997
<0.0001
<0.0001
<0.0001
0.0741
0.0056
0.0813
0.0141
0.0556
0.0685
0.0394
<0.0001
<0.0001
<0.0001
<0.0001
97.0
84.3
75.8
72.1
82.9
88.6
80.9
—
2d
3e
4f
5g
6h
7i
8j
9k
10l
11m
—
—
—
a
Reaction conditions: H2O (0.20 mL), CO2 (1 bar), photocatalyst (1 mg),
25 ꢁC, 12 h, and l > 420 nm. b Selectivity ¼ nCO/n(CO+H ) ꢂ 100. c Using C-
2
d
TiO2ꢀx@g-C3N4-3 as catalyst. Employing commercial P25 as catalyst.
e
Applying TiO2ꢀx (made from reduced P25 by NaBH4) as catalyst.
Employing TiO2@g-C3N4 as photocatalyst, prepared by mixing g-C3N4
f
and P25 in Teon reactor, heating in an oven at 180 ꢁC for 12 h.
g
Using TiO2@g-C3N4 as photocatalyst, which was prepared with
a similar procedure to that in (f), except P25 was replaced by with
h
titanium isopropoxide.
Black TiO2 nanobelts@g-C3N4 acted as
Possible photocatalytic mechanism
photocatalyst, which was prepared following the procedure reported
i
by Zhou et al. Using directly mixed C-TiO2ꢀx and g-C3N4 to replace C-
The performance of a photocatalytic system in CO2 reduction is
dictated by a balance between the thermodynamics and
kinetics. A famous mechanism path is to rst convert CO2 into
j
k
TiO2ꢀx@g-C3N4. Without H2O vapour. Without C-TiO2ꢀx@g-C3N4.
l
m
In dark. Using Ar to replace CO2. Detailed synthesis procedure of
these catalysts please see the ESI.
21602 | J. Mater. Chem. A, 2018, 6, 21596–21604
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
Journal of Materials Chemistry A
CO2cꢀ, which needs a high negative equilibrium potential of
Based on the above discussion, a possible catalytic mecha-
1.9 V versus NHE.38,39 However, the photoinduced electrons in nism for CO2 reduction over C-TiO2ꢀx@g-C3N4 is proposed
the conduction band edge of all semiconductors cannot provide (Fig. S15†). First, CO2 is adsorbed on the surface of TiO2ꢀx
enough drivꢀing force to carry out the one-electron reduction of spheres and unoccupied g-C3N4. Aer irradiation, charge
CO2 to CO2c .40,41 Another mechanistic route is through proton- separation occurs in g-C3N4 and TiO2ꢀx, producing the elec-
coupled electron transfer processes, which could bypass the tron–hole pairs. The generated electrons transfer to the amor-
formation of CO2cꢀ. Through transferring the protons with phous carbon and trigger the reductive reaction under the
electrons, the large activation barriers can be readily avoided.42 assistance of a H+ proton from water to produce CO, while the
However, the proton-participating route has a signicant remaining holes on g-C3N4 with the assistance of the carbon
kinetic dependence on the electron density on the surface of layers perform the oxidative reaction and generate O2 via a two-
semiconductors catalysts. As discussed above, the initial H+ electron/two-step pathway.
protons capture-speed of C-TiO2ꢀx@g-C3N4-3 is 300 mmol
gꢀ1 minꢀ1, which is two orders of magnitude higher than the
Conclusions
generation speed of CO (3.41 mmol gꢀ1 minꢀ1). The affluent
protons accumulated on the surfaces of photocatalyst provide In conclusion, we propose a design strategy of a low-cost, highly
a guarantee for the successful reduction of CO2 through the durable, and efficient TiO2-based three-component photo-
proton-coupled pathway.
catalyst C-TiO2ꢀx@g-C3N4 for the generation of CO coupled with
As for the other half reaction, the reactivity of water oxidation water oxidation. In this strategy, polyoxotitanium, [Ti17O24
triggered by photogenerated holes has a considerable inuence (OPri)20], is ingeniously employed as the precursor of TiO2. In
on the efficiency of CO2 reduction in H2O system. For water the composite photocatalyst, an ultrasmall TiO2ꢀx nanoparticle
oxidation, the challenge lies mainly in the release of diatomic (5–8 nm) coated ultra-thin carbon layer was fabricated as the
O2. The stepwise two-electron/two-step pathway is proposed as catalyst for converting CO2 toward CO, and a g-C3N4 coupled
an optimum approach between thermodynamic and kinetic amorphous carbon layer was adopted as a catalyst for water
constraints for water oxidation.9,16 In our photocatalytic system, oxidation via a two-electron/two-step pathway. Such a delicate
g-C3N4 incorporated with amorphous carbon layers was structure widens the light adsorption into the whole visible
designed to full the two-electron/two-step oxidation pathway. light, meanwhile it was shown to possess a good adsorption
The oxidation product of O2 around 1.22 mmol gꢀ1 and negli- capability for CO2 (17.7 cm3 gꢀ1). Under visible-light irradiation,
gible H2O2 was detected in the catalytic system using C- C-TiO2ꢀx@g-C3N4 showed around 12.30 mmol gꢀ1
TiO2ꢀx@g-C3N4-3 as the photocatalyst. Rotating ring-disk elec- (204.96 mmol gTiO ꢀ1) CO generation within 60 h reaction,
2
trode (RRDE) experiments (Fig. S13 and S14†) directly corresponding to 0.205 mmol gꢀ1 hꢀ1. When switching the
conrmed that the oxidation reaction undergoes a two-electron condition of visible light into whole light, the photocatalytic
process.
activity displayed a 1-fold enhancement. This CO generation
In photocatalysis, three universally accepted steps are rate is one order higher than the reported TiO2-based catalyst
involved: light harvesting, charge separation, and trans- under similar conditions. The experiments for H+ protons
portation to surface sites, thus triggering redox reactions.43,44 As adsorption showed the capture speed of C-TiO2ꢀx@g-C3N4
for CO2 reduction, two more steps should be taken into account, was 300 mmol gꢀ1 minꢀ1, which was two orders of magnitude
they are CO2 and proton adsorption. CO2 adsorption through higher than the generation speed of CO (3.41 mmol
interactions with surface atoms facilitates the formation of gꢀ1 minꢀ1). Such rich protons accumulation provides a guar-
a partially charged species, which could lower the energy barrier antee for CO2 reduction through the low-energy proton-
and lead to increased reactivity. Proton participation could also coupled pathway. Coupling the reductive process with
conduct the reduction in a low-energy way.45
a matched water oxidation route is an attractive avenue to
In CO2 reduction with C-TiO2ꢀx@g-C3N4-3 as the photo- boost the reductive half-reaction. Rotating ring-disk electrode
catalyst, C-TiO2ꢀx contributes most of the adsorbed photons, (RRDE) experiments indicated that the oxidative reaction in
while g-C3N4 could absorb the light before 450 nm. The ultrathin the C-TiO2ꢀx@g-C3N4 system was carried out via the efficient
carbon layer serving as an electron reservoir takes an important two-electron/two-step pathway with O2 as the nal oxidative
role in accumulating the photoinduced electrons, therefore effi- product. Photocurrent studies revealed the thin carbon layer
ciently suppressing the recombination of the generated electron– serving as electron reservoir played an important role in
hole (e–h+) pairs. The ultrasmall size of C-TiO2ꢀx supplies suffi- facilitating the generated electron–hole (e–h+) pairs separa-
cient reductive sites. Owing to the slower recombination speed of tion. Because of the slower recombination speed of electron–
electron–hole pairs of g-C3N4 than that of C-TiO2ꢀx, g-C3N4 could hole pairs of g-C3N4 than that of C-TiO2ꢀx, the photoexcited
supply photoexcited electrons to C-TiO2ꢀx. The mesoporous electrons on the conduction band of g-C3N4 could transfer to
structure of g-C3N4 provides an appropriate accumulation space that of C-TiO2ꢀx, therefore enhancing the change separation
for CO2 and H+ protons. Under the assistance of the carbon layer, and further increasing the photocatalytic activity. This work
g-C3N4 becomes an active photocatalyst for water oxidation might open up an alternative method to produce ultrasmall
through an efficient two-electron/two-step pathway. The spatially black TiO2 nanoparticles based materials and provides novel
separated reductive and oxidative sites could exclude the possi- insights for the design of high-performance and low-cost
bility of re-oxidation of the reduction products.
photocatalysts for converting CO2 to CO.
This journal is © The Royal Society of Chemistry 2018
J. Mater. Chem. A, 2018, 6, 21596–21604 | 21603

View Article Online
Journal of Materials Chemistry A
Paper
21 H. Li, Z. Kang, Y. Liu and S.-T. Lee, J. Mater. Chem., 2012, 22,
24230.
Conflicts of interest
22 W. H. Fang, L. Zhang and J. Zhang, J. Am. Chem. Soc., 2016,
138, 7480–7483.
There are no conicts to declare.
23 S. Cao, J. Low, J. Yu and M. Jaroniec, Adv. Mater., 2015, 27,
2150–2176.
Acknowledgements
This work was nancially supported by the NSFC of China (No.
21671034, 21471027), the Fundamental Research Funds for the
Central Universities (No. 2412016KJ041), Changbai Mountain
Scholars of Jilin Province, Foundation of Jilin Educational
Committee (No. 2016498).
24 N. Steunou, G. Kickelbick, K. Boubekeur and C. Sanchez, J.
Chem. Soc., Dalton Trans., 1999, 3653–3655.
25 L. Shen, Z. Xing, J. Zou, Z. Li, X. Wu, Y. Zhang, Q. Zhu,
S. Yang and W. Zhou, Sci. Rep., 2017, 7, 41978.
26 W. Wang, D. Xu, B. Cheng, J. Yu and C. Jiang, J. Mater. Chem.
A, 2017, 5, 5020–5029.
27 J. Yu, S. Wang, J. Low and W. Xiao, Phys. Chem. Chem. Phys.,
2013, 15, 16883–16890.
28 Z. Li and W. T. Tysoe, Surf. Sci., 2012, 606, 1934–1941.
29 S. Pany and K. M. Parida, Phys. Chem. Chem. Phys., 2015, 17,
8070–8077.
Notes and references
1 J. L. White, M. F. Baruch, J. E. Pander, Y. Hu, I. C. Fortmeyer,
J. E. Park, T. Zhang, K. Liao, J. Gu, Y. Yan, et al., Chem. Rev.,
2015, 115, 12888–12935.
2 A. Goeppert, M. Czaun, J. P. Jones, G. K. Surya Prakash and
G. A. Olah, Chem. Soc. Rev., 2014, 43, 7995–8048.
3 M. Mikkelsen, M. Jørgensen and F. C. Krebs, Energy Environ.
Sci., 2010, 3, 43–81.
30 X. Song, Y. Hu, M. Zheng and C. Wei, Appl. Catal., B, 2016,
182, 587–597.
31 Y. Chen, W. Huang, D. He, Y. Situ and H. Huang, ACS Appl.
Mater. Interfaces, 2014, 6, 14405–14414.
4 N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A.,
2006, 103, 15729–15735.
32 W. J. Ong, L. L. Tan, Y. H. Ng, S. T. Yong and S. P. Chai,
Chem. Rev., 2016, 116, 7159–7329.
5 S. N. Habisreutinger, L. Schmidt-Mende and J. K. Stolarczyk,
Angew. Chem., Int. Ed., 2013, 52, 7372–7408.
6 Z. Y. Sun, T. Ma, H. C. Tao, Q. Fan and B. X. Han, Chem, 2017,
3, 560–587.
7 A. d. Klerk, Energy Environ. Sci., 2011, 4, 1177.
8 V. R. Calderone, N. R. Shiju, D. Curulla-Ferre, S. Chambrey,
A. Khodakov, A. Rose, J. Thiessen, A. Jess and
G. Rothenberg, Angew. Chem., Int. Ed., 2013, 52, 4397–4401.
9 J. Liu, Y. Liu, N. Y. Liu, Y. Z. Han, X. Zhang, H. Huang,
Y. Lifshitz, S. T. Lee, J. Zhong and Z. H. Kang, Science,
2015, 347, 970–974.
33 J. Zhou, W. C. Chen, C. Y. Sun, L. Han, C. Qin, M. M. Chen,
X. L. Wang, E. B. Wang and Z. M. Su, ACS Appl. Mater.
Interfaces, 2017, 9, 11689–11695.
34 M. Sun, Y. Fang, Y. Kong, S. Sun, Z. Yu and A. Umar, Dalton
Trans., 2016, 45, 12702–12709.
35 Z. Lu, L. Zeng, W. L. Song, Z. Y. Qin, D. W. Zeng and C. S. Xie,
Appl. Catal., B, 2017, 202, 489–499.
36 S. Neatu, J. A. Macia-Agullo, P. Concepcion and H. Garcia, J.
Am. Chem. Soc., 2014, 136, 15969–15976.
37 S. Xie, Q. Zhang, G. Liu and Y. Wang, Chem. Commun., 2016,
52, 35–59.
10 X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110,
6503–6570.
38 B. Lin, S. Golshanshirazi and G. Guiochon, J. Phys. Chem.,
1989, 93, 3363–3368.
11 K. Domen, S. Naito, M. Soma, T. Onishi and K. J. Tamaru,
catalyst, J. Chem. Soc., Chem. Commun., 1980, 12, 543–544.
12 J. Tian, Z. Zhao, A. Kumar, R. I. Boughton and H. Liu, Chem.
Soc. Rev., 2014, 43, 6920–6937.
13 X. Chen, L. Liu and F. Huang, Chem. Soc. Rev., 2015, 44,
1861–1885.
14 X. Chen and S. S. Mao, Chem. Rev., 2017, 107, 2891–2959.
15 Y. Sohn, W. Huang and F. Taghipour, Appl. Surf. Sci., 2017,
396, 1696–1711.
39 B. A. Rosen, A. Salehi-Khojin, M. R. Thorson, W. Zhu,
D. T. Whipple, P. J. Kenis and R. I. Masel, Science, 2011,
334, 643–644.
40 H. A. Schwarz and R. W. Dodson, J. Phys. Chem., 1989, 93,
409–414.
41 E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja,
Chem. Soc. Rev., 2009, 38, 89–99.
42 T. Inoue, A. Fujishima, S. Konishi and K. Honda, Nature,
1979, 277, 637–638.
16 X. Chang, T. Wang and J. Gong, Energy Environ. Sci., 2016, 9,
2177–2196.
43 T. Hisatomi, J. Kubota and K. Domen, Chem. Soc. Rev., 2014,
43, 7520–7535.
17 S. Xie, Y. Wang, Q. Zhang, W. Deng and Y. Wang, ACS Catal.,
2014, 4, 3644–3653.
18 A. Pougin, M. Dilla and J. Strunk, Phys. Chem. Chem. Phys.,
2016, 18, 10809–10817.
19 X. Zhang, Y. Wang, B. Liu, Y. Sang and H. Liu, Appl. Catal., B,
2017, 202, 620–641.
44 D. Kang, T. W. Kim, S. R. Kubota, A. C. Cardiel, H. G. Cha
and K. S. Choi, Chem. Rev., 2015, 115, 12839–12887.
45 L. Zhang, D. W. Jing, X. L. She, H. W. Liu, D. J. Yang, Y. Lu,
J. Li, Z. F. Zheng and L. J. Guo, J. Mater. Chem. A, 2014, 2,
2071–2078.
20 J. Zhao, Y. Li, Y. Zhu, Y. Wang and C. Wang, Appl. Catal., A,
2016, 510, 34–41.
21604 | J. Mater. Chem. A, 2018, 6, 21596–21604
This journal is © The Royal Society of Chemistry 2018