CuO-decorated dual-phase TiO2 microspheres with enhanced activity for photocatalytic CO2 reduction in liquid–solid regime

Article abstract of DOI:10.1016/j.cplett.2019.04.010

Copper oxide (CuO)-decorated anatase and TiO₂(B) dual-phase titania microsphere (CuO/TiO₂(AB)) photocatalysts were synthesized using a facile two-step approach. The photocatalytic reduction of CO₂ to form methyl formate (MF) in methanol was evaluated. The results indicated that the CuO-decorated dual-phase TiO₂ microspheres displayed MF yield of 1800 μmol·g^?1 under 4 h illumination, higher than other reference photocatalysts, due to its mixed-phase heterojunction structure and larger surface area, thus leading to an enhanced UV-light response, effective separation and low recombination rate of the photoinduced charge carriers. CuO/TiO₂(AB) also showed good stability with the MF yield reproducibility greater than 90% in cyclic runs.

Full text of DOI:10.1016/j.cplett.2019.04.010

Accepted Manuscript
Research paper
CuO-decorated dual-phase TiO2 microspheres with enhanced activity for pho-
tocatalytic CO2 reduction in liquid–solid regime
Yunxia Zhao, Jiaxin Chen, Wei Cai, Yunfei Bu, Qiong Huang, Tao Tao,
Jiangang Lu
PII:
S0009-2614(19)30281-7
DOI:
https://doi.org/10.1016/j.cplett.2019.04.010
CPLETT 36364
Reference:
Chemical Physics Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
27 December 2018
2 April 2019
4 April 2019
Please cite this article as: Y. Zhao, J. Chen, W. Cai, Y. Bu, Q. Huang, T. Tao, J. Lu, CuO-decorated dual-phase
TiO2 microspheres with enhanced activity for photocatalytic CO2 reduction in liquid–solid regime, Chemical Physics
Letters (2019), doi: https://doi.org/10.1016/j.cplett.2019.04.010
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CuO-decorated dual-phase TiO2 microspheres with
enhanced activity for photocatalytic CO2 reduction in
liquid–solid regime
Yunxia Zhao, Jiaxin Chen, Wei Cai*, Yunfei Bu, Qiong Huang, Tao Tao, Jiangang Lu
Jiangsu Collaborative Innovation Center of Atmospheric Environment & Equipment Technology,
Jiangsu Key Laboratory of Atmospheric Environment Monitoring and Pollution Control, School
of Environmental Science and Engineering, Nanjing University of Information Science &
Technology, Nanjing 210044, PR China.
Abstract
Copper oxide (CuO)-decorated anatase and TiO2(B) dual-phase titania microsphere
(CuO/TiO2(AB)) photocatalysts were synthesized using a facile two-step approach. The
photocatalytic reduction of CO2 to form methyl formate (MF) in methanol was evaluated. The
results indicated that the CuO-decorated dual-phase TiO2 microspheres displayed MF yield of
1800 μmol•g-1 under 4 h illumination, higher than other reference photocatalysts, due to its
mixed-phase heterojunction structure and larger surface area, thus leading to an enhanced
UV-light response, effective separation and low recombination rate of the photoinduced charge
carriers. CuO/TiO2(AB) also showed good stability with the MF yield reproducibility greater than
90% in cyclic runs.
Keywords: Photocatalytic CO2 reduction, Anatase TiO2, TiO2 (B), CuO loading, Heterojunctions
1. Introduction
*Corresponding author
E-mail: caiwei_19880105@163.com; Tel: +86 25 58731090

Carbon dioxide (CO2) is a carbon source that possesses a very high energy potential and
could serve as a sustainable energy source. The photocatalytic conversion of waste CO2 into
energy-rich hydrocarbon fuels or more valuable chemicals is a promising method to address two
major problems: the imminent energy crisis and anthropogenic climate change. Titanium dioxide
(TiO2), the earliest studied photocatalyst [1], has advantages such as being nontoxic, lossless and
environmentally friendly. Among the many well-studied photocatalysts, TiO2 has attracted much
attention and has been the focus of research due to its high activity, high thermal stability,
[2]
durability, low cost, nontoxicity and other characteristics.
However, as a single-component
semiconductor photocatalyst, TiO2 has an inherent drawback, i.e., its rapid photogenerated
electron/hole (e−/h+) pair recombination (~10 ns); hence, only a fraction of the generated e−/h+
pairs are available for photoreaction [3, 4]. To further improve the photocatalytic performance of
TiO2, numerous methods, including metal/nonmetal doping, metal loading, heterojunction
construction and coupling with carbon materials, have been proposed [5]. Jiang et al. [6] assembled
a TiO2/active carbon-Ag ternary composite photocatalyst for the photocatalytic conversion of CO2
using H2O as the reducing agent. The composite photocatalyst exhibited a six-fold increase in the
CO yield compared to that obtained with pristine TiO2. Li et al. [7] reported a promising N-rich
carbon quantum dot/TiO2 photocatalyst for CO2 photoreduction by gaseous water under simulated
solar irradiation. As a result, 1.838 μmol CO and 1.195 μmol CH4 were obtained from CO2
[8]
reduction on 50 mg of photocatalyst after 6 h of irradiation. Zhao et al.
synthesized a
graphene-wrapped Pt/TiO2 photocatalyst with high photocatalytic activity and selectivity for CO2
conversion; the formation rate of CH4 was 41.3 μmol•g-1•h-1, and the selectivity of the CO2
conversion to CH4 was 99.1%.

All of the above co-catalysts can extend the light absorption range and suppress the
recombination and transfer of charge carriers, resulting in superior CO2 photoconversion
performance. However, the limited reserves, inaccessibility and high cost of these co-catalysts,
often due to the use of noble metals, limit their practical applications. As an alternative, the
incorporation of CuO into TiO2 is a less expensive option [9-12]. CuO has a band gap of 1.6 eV that,
in principle, should favour the absorption of visible light and improve the photocatalytic activity.
Gao et al. [13] designed an innovative noble metal-free inorganic nanocomposite (hollow TiO2/CuO)
that is capable of displaying both photochromism and catalysis effects under solar irradiation.
[14]
Taufik et al.
successfully synthesized ternary CuO/TiO2/ZnO nanocomposites to investigate
their photocatalytic degradation of methylene blue. Under optimal conditions, 100% and 98% of
the methylene blue was removed within 2 h under UV and visible light, respectively. The results
[15]
showed that holes were the main active species in the degradation process. Razali et al.
reported that CuO-loaded TiO2 nanotube photocatalysts demonstrated a higher CO2 conversion
efficiency than TiO2 nanotubes because the former could effectively separate photogenerated
electron-hole pairs due to the presence of CuO.
Given the above, CuO can serve as an effective co-catalyst for CO2 photoreduction. The main
catalyst, TiO2, has at least 11 different phases. [16] The most well-known phases are anatase, rutile,
and brookite. Anatase is the most investigated allotrope due to its relatively high photocatalytic
activity. TiO2(B) is another polymorph of TiO2 and possesses a more negative conduction band
[7]
(CB) edge than anatase[17]. To date, polyphase TiO2 has rarely been studied for photocatalytic
CO2 reduction. By the definition of a heterojunction, the interface between two different phases
with unequal band structures can enhance the optical adsorption of photocatalysts. Si et al. [18]

prepared a novel composite with anatase TiO2 quantum dots grown on TiO2 (B) nanosheets for
photocatalytic H2 production, which showed a high photocatalytic activity, approximately 45
times higher than that of rutile-anatase mixed Degussa P25. The anatase and B mixed phase
promoted efficient charge separation at the interface, while a relatively high conduction band edge
of TiO2 (B) with respect to the redox potential level of H+/H2 leads to a further enhancement of the
phtocatalytic activity.
In this paper, CuO-decorated anatase and TiO2 (B) dual-phase titania microspheres were
prepared for the photocatalytic conversion of CO2 by methanol. To the best of our knowledge,
there has been no report on this catalyst for CO2 photoreduction. Three other materials, namely,
pure anatase TiO2, CuO-loaded anatase TiO2 and anatase-B polyphase TiO2, were compared with
the objective catalyst to highlight its unique synergistic advantage of mixed phases and CuO
loading. The optical catalysis reaction system was a catalyst-methanol slurry regime at room
temperature and atmospheric pressure. Methanol was used as a reducing agent because it has a
higher CO2 absorption capacity than H2O. Meanwhile it was a sacrificial agent to avoid
introducing other toxic sacrificial agent such as acetonitrile. Improving the CO2 absorption
capacity, extending the light absorption range and intensity and suppressing carrier recombination
are expected to enhance the efficiency for the photocatalytic conversion of CO2.
2. Experimental Section
2.1. Photocatalyst preparation
First, single-component anatase TiO2, referred to as TiO2(A), was synthesized by a sol-gel
process [19-21]. A total of 10 g tetrabutyltitanate was dissolved in 10 mL ethanol by vigorous
stirring for 30 min, and 1.3 g CTAB was dissolved in 10 mL ethanol using the ultrasonic

dissolution method. Then, the above two solutions were mixed together and stirred for 30 min.
After ageing for 12 h at room temperature, a sol formed. The sol was dried at 100 ℃ for 12 h and
then calcined at 450 ℃for 6 h with a heating rate of 2 ℃/min.
[22]
The anatase and B dual-phase TiO2 were synthesized by the solvothermal method
.
Tetrabutyltitanate (0.1 mL) was added to 10 mL acetic acid and vigorously stirred for 10 min at
room temperature. Then, the solution was transferred into a 100-mL Teflon-lined stainless-steel
autoclave and maintained at 180 ℃ for 12 h. The reaction vessel was cooled naturally to room
temperature, and then the precipitate was collected by suction filtration. After washing repeatedly
with ethanol and drying at 80 ℃ overnight, the powder was calcined at 370 ℃ for 2 h and named
TiO2(AB).
CuO-incorporated TiO2 composites were prepared using an impregnation method. They were
named CuO-TiO2(A) and CuO/TiO2(AB). Taking CuO/TiO2(AB) as an example, 0.1 g of
as-prepared TiO2(AB) was added into a certain concentration solution of Cu(NO3)2. A
homogeneous phase system was made by ultrasonic dispersion technology. Then, the above
suspension was dried in a constant-temperature bath followed by further drying at 120 ℃ for 12 h
in an electric oven. The final product was obtained by calcination in a muffle furnace at 400 ℃
with a heating rate of 1 ℃•min-1 for 2 h.
2.2. Photocatalyst characterization
X-ray powder diffraction (XRD) analysis was carried out using a Philips X'Pert Pro XRD
instrument with Cu Kα X-ray radiation. The Raman spectra were recorded using a RENISHAW
spectrometer with an excitation laser of 532 nm. Scanning electron microscopy (SEM) was
performed using a Hitachi S4800 microscope. Before the test, the samples were treated by gold

coating. High-resolution transmission electron microscopy (HR-TEM) studies were performed on
a JEM-2100 (JEOL, Japan) instrument equipped with a slow-scan CCD camera and an
accelerating voltage of 200 kV. N2 adsorption-desorption isotherms were measured at -196 ℃
using a Micromeritics ASAP 2020 physisorption instrument. The surface areas and porosity were
obtained by Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halender (BJH) method. X-ray
photoelectron spectroscopy (XPS) analyses were performed on an ESCALAB 250 spectrometer
(Thermo, U.S.A.). All binding energies (BE) were referenced to the adventitious C1s at 284.4 eV
[23, 24]. The X-ray source utilized was Al Kα X-ray (hν=1486.6 eV) radiation. Ultraviolet-visible
diff use reflectance spectroscopy (DRS) was performed on a Shimadzu UV-2600 spectrometer.
Photoluminescence (PL) spectra were recorded on a Hitachi F-4500 instrument with an excitation
wavelength of 325 nm at room temperature. Electrochemical impedance spectroscopy (EIS)
measurements and the photocurrents were obtained using a CHI760 electrochemical working
station with a standard three-electrode system and 0.5 mol·L-1 Na2SO4 as the electrolyte solution.
FTO glass was dip-coated with a mixed solution of the photocatalyst, naphthol, deionized water
and alcohol and used as the work electrode.
2.3. Photocatalytic activity tests
The photocatalytic efficiency of the obtained samples was evaluated by the production of
HCOOCH3 using a 50 mL quartz tube as a photocatalytic reactor. A 250-W Hg lamp (BL-GHX-V,
Shanghai BiLon, China) was positioned parallel to and 15 cm away from one side of the
photocatalytic reactor as a light source. The irradiation intensity was measured about 3900
μW/cm2 at 365 nm which is the main radiation wavelength of the lamp. Typically, 30 mg of the
photocatalyst was added into a pure CH3OH solution (30 mL), and pure CO2 gas was then injected

into the closed reaction system for 30 min. Before the light irradiation, the solution was saturated
with CO2 gas to ensure that an adsorption-desorption equilibrium was achieved. The magnetic
stirrer at the bottom provided ideal mixing. The product from the photocatalytic reaction,
HCOOCH3, was analysed using gas chromatography (GC 9790Ⅱ, China) with a flame ionization
detector.
3. Results and Discussion
3.1. Characterization of photocatalysts
3.1.1. Structural properties
The phase structures of these photocatalysts were investigated by XRD, as shown in Fig. 1
(ⅰ). The XRD patterns of TiO2(A) and CuO-TiO2(A) match well with that of anatase-phase TiO2
(PDF# 21-1272) with diffraction peaks at 25.3°, 36.9°, 37.8°, 38.6°, 48.0°, 53.9°, 55.1°, 62.1°,
62.7°, 68.7°, 70.3°, 74.0°, 75.0° and 76.0° that can be indexed to the (101), (103), (004), (112),
(200), (105), (211), (213), (204), (116), (220), (107), (215) and (301) crystal facets, respectively.
The XRD patterns of TiO2(AB) and CuO/TiO2(AB) are similar to the standard spectrum of the
monoclinic polymorph TiO2(B) (PDF# 46-1237). However, a very obvious peak at 37.5° and a
less obvious peak at 70.0° indicate that the TiO2(AB) and CuO/TiO2(AB) are not single-phase
materials but contain some anatase phase. Based on the XRD patterns, TiO2(A) and CuO-TiO2(A)
have a single phase, while TiO2(AB) and CuO/TiO2(AB) have two phases. CuO is not detected in
the composites, which might be due to its low content and high dispersity. However, the
intensities of the peaks increase with the incorporation of CuO, indicating the influence of CuO on
the crystallization of TiO2.
Figure 1 (ⅱ) shows the Raman spectra of the prepared samples. As reported in the

literature[8,18,25], the characteristic Raman peaks of the anatase phase are located at 395, 516, and
638 cm-1 and those of TiO2(B) are located at 210, 256, 382, 422, 477, 553, 632, and 828 cm-1 [18]
.
For TiO2(A) and CuO-TiO2(A), only the Raman peaks corresponding to anatase TiO2 can be
detected, indicating the formation of a pure anatase phase. For TiO2(AB) and CuO/TiO2(AB),
peak differentiation was used to confirm their phase formation due to undistinguishable peaks.
The results show the presence of characteristic peaks for both anatase and TiO2(B), implying that
mixed phases are present. Based on the intensity of these peaks, the contributions of both phases
are considerable. The XRD results are confirmed here, including the changes in the peak
intensities after the CuO addition. Similarly, the Raman peaks of CuO are not detected in the
composites of CuO and TiO2.
(i)
(ii)
Fig. 1. (ⅰ) XRD patterns and (ⅱ) Raman spectra of single-component TiO2 and its composites
with CuO. (A and B refer to the anatase and TiO2(B) phases, respectively.)
The morphologies of the as-synthesized pure TiO2 and CuO/TiO2 composites were
investigated by SEM, as shown in Fig. 2. The samples were not treated by grinding, ultrasound or
other size reduction procedures before the test. The images show that TiO2(A) and CuO-TiO2(A)

have anomalistic bulk structures with non-uniform sizes and agglomerated nanoparticles.
TiO2(AB) and CuO/TiO2(AB) have highly dispersed microsphere structures with diameters of ~2
μm. The microspheres are composed of numerous well-defined, smooth nanoplates with sharp
edges. The acetic acid present in the reaction system could promote the formation of this
hierarchical structure[26]. The size of TiO2(AB) is more evenly distributed than that of
CuO/TiO2(AB). In short, the dual-phase TiO2 samples possess more order and better dispersal
than the single-phase TiO2 samples. Moreover, the smaller particle size is more beneficial to
electron transfer[27]
.
Fig. 2. SEM images of all the prepared samples. a) TiO2(A), b) CuO-TiO2(A), c, e) TiO2(AB), and
d, f) CuO/TiO2(AB).
Figure 3a displays a TEM image of the CuO/TiO2(AB) microspheres consisting of

aggregated nanosheets. The HRTEM image (Fig. 3b) shows lattice fringes with d spacings of
0.356, 0.352 and 0.233 nm. According to the XRD standard PDF cards, the former two spacings
can be ascribed to the (110) plane of TiO2(B) and (101) plane of anatase TiO2, respectively. The
0.233 nm spacing may correspond to the (111) plane of CuO or the (112) plane of anatase TiO2.
This image cannot confirm the presence of crystalline CuO in the dual-phase TiO2 composites.
Therefore, further characterization is required to determine the overall chemical composition.
Fig. 3. (a) TEM and (b) HRTEM images of CuO/TiO2(AB).
Fig. 4 presented the nitrogen sorption isotherms and BJH pore size distribution curves of the
as-prepared samples. Their typical type IV isotherms with a distinct hysteresis loop are the
characteristic feature of mesopores[11,14]. The mixed-phase TiO2 samples show a broad pore size
distribution than the anatase samples, as indicated in Fig. 4(ii). The calculated BET surface area
and pore volume of all these samples are summarized in Table S1. Obviously, the polyphase TiO2
samples possess higher surface area and pore volume than the anatase samples due to smaller
particle size and richer porosity.

(i)
(ii)
Fig. 4. (ⅰ) N2 adsorption-desorption isotherms and (ⅱ) BJH pore size distribution curves of
single-component TiO2 and its composites with CuO.
3.1.2. Chemical composition
Figure S1 shows the EDS and elemental mapping results for CuO/TiO2(AB), confirming the
presence of Ti, O, and Cu finely dispersed on the surface of the dual-phase TiO2. Table S2
summarizes the weight percentages and atomic percentages of the above elements in
CuO/TiO2(AB). The Cu content in the composites is approximately 6 wt.%, and the atomic
percent of O matches well with the atomic ratios of TiO2 and CuO.
The surface chemical compositions and valence states of various elements in the synthesized
materials were examined through XPS, as shown in Fig. 5. The survey scan spectra (Fig. 5ⅰ)
indicate the presence of C, O and Ti in the two TiO2 samples and Cu in the TiO2 and CuO
composites. Among these elements, C is contamination, as explained in the experimental section.
Fig. 5ii displays the O 1s spectra; the spectra can be separated into two peaks at approximately
532.0 and 530.0 eV that can be assigned to O-H and Ti-O bonds, respectively, which are attributed
to Ti-OH linkages and Ti-O-Ti (lattice O). As reported in the literature[17, 18, 21], the bridging
hydroxyl groups (Ti-O-H) are known to come from the dissociative adsorption of water on the
TiO2 surface. In Fig. 5ii, the intensity of this shoulder peak for CuO/TiO2(AB) is higher than that

of the other three samples. The hydroxyl oxygen could capture electrons to produce active species
(•OH and •O2), and these species would accelerate the photoreaction[15]. The Ti 2p XPS spectra
(Fig. 5iii) of all synthesized samples are almost the same; the Ti 2p1/2 and Ti 2p3/2 peaks are
centred at binding energy values of 464.5 and 458.7 eV, respectively, which correspond to Ti4+
species[17, 21]. The binding energies of Cu 2p3/2 and Cu 2p1/2 are 932.6-933.1 eV and 952.7-952.8
eV, respectively, for the two CuO-loaded composites (Fig. 5iv), suggesting that the existence of
element Cu [28]. Although the binding energy of CuO and Cu2O is very close and difficult to
distinguish, the absence of shake-up satellite peaks between the two Cu 2p peaks implies the
presence of Cu2O [12], which disagrees with the results from EDS. It has been explained by that
diff erent chemical states of copper were caused by the sample being modified in the high vacuum
environment of XPS and simultaneous exposure to X-rays [12]. Thus, the Cu2+ translated into Cu+
under the crucial working environment in XPS analysis. The binding energy of Cu 2p in
CuO/TiO2(AB) has a positive shift, which can be seen as evidence of the different chemical
environments of Cu in the ternary system.

(ⅰ)
(ⅱ)
(ⅲ)
(ⅳ)
Fig. 5. (ⅰ) XPS survey spectra of the synthesized samples and high-resolution spectra of (ⅱ) O 1s
(ⅲ) Ti 2p and (ⅳ) Cu 2p.
3.1.3. Optical properties
The optical properties of the as-prepared photocatalysts were measured by UV-vis DRS, as
shown in Fig. 6. As one can see in Fig. 6ⅰ, after modification with CuO, the single-phase and
dual-phase TiO2 samples show enhanced light absorption in both the ultraviolet and visible-light
ranges. A higher absorption intensity indicates a higher rate of electron-hole pair formation on the
photocatalyst surface, resulting in the photocatalyst exhibiting a higher photocatalytic activity.

Compared to the single-phase samples, the UV–vis DRS spectra of the dual-phase samples shift to
shorter wavelengths, which indicates that the incorporation of B phase TiO2 into anatase TiO2
increases the light absorption threshold. The adsorption wavelength is related to the band gap. As
reported in the literature [18], the conduction band (CB) level of B phase TiO2 is higher than that of
anatase TiO2. Based on the Kubelka-Munk function, the UV-vis spectra were converted into Tauc
plots, as shown in Fig. 6ⅱ. The corresponding plots of (αhν)2 versus the photon energy (hν) can be
used to determine the band gaps by extrapolating the maximum slope to the x axis. The optical
band gaps were 3.22 eV, 3.20 eV, 3.37 eV and 3.35 eV for TiO2(A), CuO-TiO2(A), TiO2(AB) and
CuO/TiO2(AB), respectively. Since the band gap of TiO2(B) is larger than that of TiO2(A),
TiO2(AB) as the composite of anatase TiO2 and TiO2(B) shows a higher band gap value. Similarly,
the band gap values of the samples with CuO loading are lower due to the small band gap of CuO.
The increase in the band gap energy probably causes a decrease in the valence band (VB) edge or
an increase in the CB edge, which would lead to a decrease in the photoinduced carrier
[18]
recombination rate[19] and provide a larger driving force for the reduction of CO2
.
(i)
(ii)
Fig. 6. (ⅰ) UV-vis diffuse reflectance spectra, (ⅱ) plots of (αhν)2 versus the photon energy (hν) of
the as-synthesized photocatalysts.

The PL spectra of all samples were obtained to determine their rate of photogenerated
electron and hole recombination. Higher PL peaks indicate a higher probability of recombination,
which will result in a lower photocatalytic activity [21]. As shown in Fig. 7, the emission peak of
pure anatase TiO2 is at approximately 378 nm and exhibits the highest PL intensity. The
CuO-TiO2(A), TiO2(AB) and CuO/TiO2(AB) samples have emission peaks with similar intensities
that are much weaker than that of TiO2(A), which indicates that the recombination of
photogenerated electrons and holes is largely inhibited. Fig. 7 clearly shows that either CuO
loading or mixed-phase TiO2 can decrease the rate of charge carrier recombination. All three of
these samples can be considered composites. Their inner electric field changes, which helps to
minimize/inhibit charge carrier recombination. Among all the samples, the photogenerated
electrons and holes in the CuO/TiO2(AB) composite have the longest lifetime. This confirms that
a heterojunction was formed at the interface between TiO2 and CuO. The photo-induced electrons
would be shifted from the higher CB of TiO2 to CuO, which serves as a reservoir[16]
.
Fig. 7. PL spectra measured at room temperature for TiO2(A), CuO-TiO2(A), TiO2(AB) and
CuO/TiO2(AB) photocatalysts.

The photogenerated charge transfer efficiency was evaluated by a photoelectrochemical
method. As displayed in Fig. 8ⅰ, the photocurrents of the as-prepared samples were measured
during on and off cycles of irradiation. The obtained photocurrent densities decrease in the
following order: CuO/TiO2(AB) > TiO2(AB) > CuO-TiO2(A) > TiO2(A). The rapid photoinduced
e- and h+ separation efficiency corresponds to a high photocurrent density. Therefore,
CuO/TiO2(AB), a ternary hybrid structure, has a higher separation efficiency for photoinduced
charge carriers under light illumination. The results of the photocurrent response versus time
curves show that CuO loading or the incorporation of B phase TiO2 into anatase TiO2 can promote
the efficient separation of photogenerated electron-hole pairs to further enhance the photocatalytic
activity of TiO2(A). As shown in Fig. 8ⅰ, the photocurrent densities of TiO2(AB) and
CuO-TiO2(A) are close and higher than that of TiO2(A). CuO/TiO2(AB) has the highest value. To
further elucidate the separation and transfer efficiency of the photogenerated electrons and holes,
the prepared catalysts were also investigated by EIS. In theory, the arc radius in the EIS Nyquist
plot represents the electron transfer resistance. A smaller arc radius indicates a better migration
and transfer efficiency of the photoexcited e--h+ pairs. The EIS (Fig. 8ⅱ) results are consistent
with the transient photocurrent responses (Fig. 8ⅰ). As observed in Fig. 8ⅱ, the arc radius of
CuO/TiO2(AB) is the smallest, indicating that it has the lowest charge transfer resistance on the
electrode surface. A higher conductivity is conducive to the transfer of photoinduced carriers,
resulting in better photocatalytic activity.

(ⅰ)
(ⅱ)
Fig. 8. (ⅰ) Photocurrent response versus time curves; (ⅱ) EIS results of the as-prepared catalysts.
3.2. Photocatalytic CO2 reduction performance
The photocatalytic activities of the TiO2 and TiO2/CuO composites were evaluated by the
photoreduction of CO2 into methyl formate (MF) in methanol. MF is an important chemical
intermediate. It is a considerable and promising way to synthesize MF using clean energy under
moderate conditions. Fig. 9 shows the MF yields under ultraviolet radiation for 4 h. To investigate
the effect of CO2, a contrasting experiment replacing CO2 with N2 in the case of CuO/TiO2(AB)
showed the MF yield slowing down from 1800 to 350 μmol•g-1. This finding suggests that CO2
made a significant contribution to the formation of MF. In Fig. 9, the MF yield clearly decreases
in the following order: CuO/TiO2(AB) > TiO2(AB) > CuO-TiO2(A) > TiO2(A). The
CuO/TiO2(AB) composite possesses a higher photocatalytic activity than the other three samples
for the same catalyst dosage, and this can be attributed to the effects of the mixed phases and CuO
loading. The presence of two crystal phases might be equivalent to a complex of two
semiconductors, promoting the effective separation of the photogenerated electrons and holes, i.e.,
the so-called “mixed-phase effect”. Although the UV-vis diffuse reflectance spectrum (Fig. 6ⅰ) of
TiO2(AB) shows a blueshift compared to that of TiO2(A), the peaks are still in a superior light
intensity range for the high-pressure mercury lamp (Table S3) used here. More importantly, this

type of mixed-phase crystal is more effective in suppressing the recombination of photogenerated
electrons and holes as well as transporting photogenerated carriers between two semiconductors.
The above beneficial effect was largely brought by higher BET surface area of mixed-phase TiO2
as shown in Table S1. The higher surface area is favourable for electron transport and also
provides more surface reactive sites for photocatalytic CO2 reduction process[11,14,27]. Leaving
aside surface area, formation rate of MF in unit of μmol/m2/h is shown in Fig.S2. The results of
TiO2(A) samples are superior to those of TiO2(AB) samples because of lower band gap of the
former with faster light response. Meanwhile, the MF yield results of the samples with CuO are
better than those of the samples without CuO in Fig.S2. The presence of CuO with a low band gap
leads to an increase in the full-spectrum light absorption and decreases the charge carrier
recombination, allowing higher electron mobility in the composites. The synergistic effect of the
above two factors (mixed phases and CuO loading) was obvious in CuO/TiO2(AB), which exhibits
the highest photocatalytic MF production yield of ~1800 μmol•g-1. This value is higher than that
obtained with metal sulphides, as reported by Xin et al.[29-31]. Although the MF formation rate here
[33]
is somewhat lower than that obtained with Bi2S3-ZnIn2S4[32], CdS/g-C3N4
and CuO-TiO2
composites[34] with optimal formulas, the design and optimization of the component proportions of
CuO/TiO2(AB) will be a research priority in the future.

Fig. 9. The MF yields over various catalysts.
A possible mechanism for the CO2 photoreduction to MF in CH3OH over CuO/TiO2(AB)
catalysts under ultraviolet-light irradiation was proposed based on the above experimental results,
as displayed in Fig. 10. The CB edges of anatase TiO2, TiO2(B) and CuO are -0.32, -0.90 and 0.46
eV, respectively, with respect to the normal hydrogen electrode (NHE), while the VB edges of
anatase TiO2, TiO2(B) and CuO are +2.90, +2.60 and +2.16 eV, respectively [15, 18, 21]. Under
UV-light irradiation, all of the anatase TiO2, TiO2(B) and CuO can be excited to generate
photoelectrons and holes when absorbing photon energy. The electrons in the CB of TiO2(B) are
driven by the potential energy to transfer to the CB of anatase TiO2 and then to CuO. CuO is used
to trap electrons as a p-type semiconductor [12, 21, 35]. Conversely, the photogenerated hole can
move from the VB of anatase TiO2 to that of TiO2(B) and CuO [21]. The CuO/TiO2(AB) composite
[21]
acts as a ‘‘staggered’’ type II heterojunction
system, which leads to the effective separation of
the excited electrons and holes at the TiO2(B)–TiO2(A)–CuO interfaces. As reported, the VB
potential of all components is more positive than E0(CH3OH/•CH2OH) (0.927 V) [20, 32-34], and
their holes can oxidize methanol into formaldehyde with H+ formation. The electrons accumulate
in CuO and reduce part of CuO into Cu2O, forming Cu2O@CuO/TiO2(AB). The CB of Cu2O is

more negative than E0(CO2/HCOOH) (-0.61 V) [20, 32-34] and can reduce CO2 into formic acid in the
presence of H+. After donating electrons, Cu2O oxidizes back into CuO. This is consistent with the
XPS result and the experiment phenomena. The catalyst was initially an off-white color and then
darkened after illumination for a while in the suspension. Eventually, the catalyst changed back to
off-white after recovery. Similar explanations have been reported in the literature [12,21]. MF can be
produced through the esterification of formic acid and methanol and the dimerization of
formaldehyde via the Tishchenko reaction [32-34]. The synergistic effect of both mixed phases and
CuO loading on CuO/TiO2(AB) contributes to the improvement in the photocatalytic performance.
Fig. 10. Schematic of the mechanism for the CO2 photoreduction to MF in CH3OH over
CuO/TiO2(AB) catalysts.
The reusability of the photocatalysts was evaluated after recovering the catalyst from solution
by centrifugation. As displayed in Fig. 11(i), the yields of MF are still above 90% under UV-light
irradiation with the four recycled samples. Moreover, the two dual-phase TiO2 microsphere
photocatalysts have a better performance. When compared with those reported in the literatures
[29,30,33], the stability of our photocatalyst samples is still in an acceptable range. Time course of
MF production over CuO/TiO2(AB) was further investigated as shown in Fig. 11(ii), and the
photocatalyst was recovered every 4 h. Continuous increases of MF amount during every round
can be observed upon up to 4 h illumination. With the increase in cycle, the recyclability of

photocatalytic CO2 reduction in methanol to MF remains above 90% of last round. CuO/TiO2(AB)
exhibits a slightly decrease in photocatalytic CO2 reduction performance. In addition, XRD and
SEM analyses were carried out on the CuO/TiO2(AB) catalysts after the first cycle of
photoreduction, as displayed in Fig. S3. The XRD results show that all the diffraction peaks of the
CuO/TiO2(AB) composites are basically the same before and after the photocatalytic reaction.
Only the intensity of the peak at 25.3° in the XRD pattern of CuO/TiO2(AB) after use decreases,
as shown in Fig. S3. In the recovery operations, the structure or components of the catalyst
samples might be affected. The highly dispersed microsphere morphology did not change, as
shown in the SEM image, after the photocatalytic CO2 reduction reaction. All of the above results
demonstrate relatively good stability of CuO/TiO2(AB) during the photocatalytic reaction
processes. In the future, optimization of the recovery operations will be another research point for
further improving the stability of the photocatalysts in cyclic runs.
(ⅰ)
(ⅱ)
Fig. 11. (ⅰ) Reusability of the as-prepared catalysts for photocatalytic CO2 reduction in methanol;
(ii) Time dependence of MF yield over CuO/TiO2(AB) in three cyclic runs.
4. Conclusions
CuO-decorated anatase and B dual-phase TiO2 microsphere (CuO/TiO2(AB)) photocatalysts
were synthesized using a facile two-step method and used for the photoreduction of CO2 in
methanol. The structural and chemical composition analyses confirmed the existence of

mixed-phase TiO2 with highly dispersed CuO on the surface. Compared with anatase TiO2,
CuO-loaded anatase and dual-phase TiO2 without CuO, CuO/TiO2(AB) had the highest
photocatalytic MF production yield and can be considered a mixed-phase heterojunction
composite. The synergistic effect of both mixed phases and CuO loading contributed to the high
photocatalytic activity, which was a result of the enhanced light response to UV light, and
effective separation and reduced recombination of photoinduced charge carriers. CuO/TiO2(AB)
also showed good reusability and stability during the photocatalytic reaction process.
Conflicts of interest
There are no conflicts of interest to declare.
Acknowledgements
This work was supported by the National Natural Science Foundation of China (51802160 and
51808296), the Natural Science Foundation of Jiangsu Province of China (BK20150892 and
BK20170938), the Natural Science Foundation of Jiangsu Higher Education Institution of China
(15KJB610012), and the 63rd Chinese Postdoctoral Science Foundation (2018M630487). Dr. Y.
Zhao and Dr. Wei Cai are grateful for the Startup Foundation for Introducing Talent of NUIST
(S8113082001 and 2243141601034). In addition, Wei Cai would also like to acknowledge support
and helpful discussion from the WECHAT group named “Energetic and Progressive Mightyboy”!
Supporting Information
Supplementary material associated with this article can be found in the online version.
References
[1] A. Fujishima, K. Honda. Nature, 5358, (1972)37.

[2] B. Dong, T. Liu, C. Li, et al. Chinese Chem. Lett., 5(2018) 671.
[3] S.Y. Lee, S.J. Park. J. Int. Eng. Chem. 6(2013) 1761.
[4] Z. Xiong, H. Wang, N. Xu, et al. Int. J. Hydrogen. Energ. 32 (2015)10049.
[5] K. Qi, B. Cheng, J. Yu, et al. Chinese J. Catal. 12(2017) 1936.
[6] Z. Jiang, X. Zhang, Z. Yuan, et al. Chem. Eng. J. 348(2018) 592.
[7] M. Li, M. Wang, L. Zhu, et al. Appl. Catal. B-Environ. 231(2018) 269.
[8] Y. Zhao, Y. Wei, X. Wu, et al. Appl. Catal. B-Environ. 226( 2018) 360.
[9] B. Li, Y. Hao, B. Zhang, et al. Appl. Catal. A-Gen. 531(2017) 1.
[10] P. Chaudhari, S. Mishra. Measurement. 90(2016)468.
[11] I. H. Chowdhury, S. Ghosh, S. Basak, et al. J. Phys. Chem. Solids.104(2017) 103.
[12] Y. T. Liao, Y. Y. Huang, H. M. Chen, et al. ACS Appl. Mater. Inter. 49(2017) 42425.
[13] M. Gao, S. W. L. Ng, L. Chen, et al. J. Mater. Chem. A. 22(2017) 10909.
[14] A. Taufik, A. Albert, R. Saleh. J. Photoch. Photobio. A-Chem. 344(2017) 149.
[15] M. H. Razali, M. Yusoff. Mater. Lett. 221( 2018) 168.
[16] M. K. Hossain, A. R. Koirala, U. S. Akhtar, et al. Chem. Mater. 27(2015) 6550.
[17] T. Di, J. Zhang, B. Cheng, et al. Sci. China Chem. 3(2018) 344.
[18] J. Si, Y. Wang, X. Xia, et al. J. Power Sources, 360(2017) 353.
[19] P. Periyat, K. V. Baiju, P. Mukundan, et al. Appl. Catal. A: Gen., 349(2008) 13-19.
[20] S. Qin. Photoreduction of CO2 in methanol to prepare methyl formate using CuO-TiO2.
Tianjin University, 2010.
[21] M. Golestanbagh, M. Parvini, A. Pendashteh. Catal. Lett. 148 (2018) 2162.
[22] C. Wang, X. Zhang, Y. Liu. Nanoscale. 10(2014) 5329.

[23] W. Cai, B. Y u, Y. Zhao, et al. Chem. Eng. J. 317(2017) 376.
[24] W. Cai, Y. Zhao, M. Chen, et al. Chem. Eng. J. 333(2018) 414.
[25] M. Edelmannová, K. Y. Lin, J. C. S. Wu, et al. Appl. Surf. Sci. 454(2018) 313.
[26] Z. Q. Li, L. Mo, W. Chen, et al. ACS Appl. Mater. Interfaces, 9(2017) 32026.
[27] J. Zheng, L. Mo, W. Chen, et al. Nano Research, 10(2017) 3671.
[28] J. Dong, J. Ye, D. Ariyanti, et al. Chemosphere. 204(2018) 193.
[29] J. Chen, F. Xin, S. Qin, et al. Chem. Eng. J., 230(2013) 506.
[30] J. Chen, S. Qin, G. Song, et al. Dalton T. 42(2013) 15133.
[31] J. Chen, F. Xin, X. Yin, et al. RSC Adv. 5(2015) 3833.
[32] J. Chen, F. Xin, H. Niu, et al. Mater. Lett.198(2017) 1.
[33] X. Yang, W. Xin, X. Yin, et al. Chem. Phys. Lett. 651(2016) 127.
[34] S. Qin, F. Xin, Y. Liu, et al. J. Colloid Interf. Sci. 1(2011) 257.
[35] K. K. Mandari, J. Y. Do, S. V. P. Vattikuti, et al. J. Alloy. Compd. 750 (2018) 292.

Graphical Abstract

Highlights
1. CuO-decorated anatase and B dual-phase TiO2 microsphere (CuO/TiO2(AB)) photocatalysts
were successfully synthesized.
2. CuO/TiO2(AB) photocatalysts exhibited high catalytic activity for CO2 reduction to HCOOCH3
in CH3OH solution.
3. CuO/TiO2(AB) can efficiently enhance the photogenerated charge separation.
4. CuO/TiO2(AB) acted as a mixed-phase heterojunction composite.