The construction of a wide-spectrum-responsive and high-activity photocatalyst, Bi25CoO40,viathe creation of large external dipoles

Article abstract of DOI:10.1039/d0ta11549e

Currently, a major issue in the field of photocatalysis is the insufficient utilization of solar energy by photocatalysts. Thus, to overcome this problem, herein, a new sillenite photocatalyst, Bi₂₅CoO₄₀, was designed to exhibit a nearly full visible-light spectrum response (up to 750 nm) and high activity. In particular, its photocatalytic activity for the degradation of typical wastewater pollutants (e.g., MB and 4-CP) reached up to 3.402 h^?1for MB under visible-light irradiation and 1.576 h^?1for 4-CP under full spectrum irradiation. The degradation rates areca.2-4 times higher than those of well-studied photocatalysts such as P25-TiO₂, Bi₂WO₆, and g-C₃N₄, and are also the best for organic pollutant removal among reported photocatalysts absorbing above 700 nm. Results showed that the remarkable degradation on Bi₂₅CoO₄₀is mainly due to the efficient charge separation resulting from the large external diploe moment between [Co^IIIO₄] and [Bi^VO₄] tetrahedrons. Results further illustrated that the VBM of new photocatalysts is composed of Co 3d-O 2p orbitals and exhibits a potential (ca.2 eV) positive enough to produce oxidative species for the degradation of organic pollutants. The construction of a large external dipole moment without sacrificing the valence band edge potential may provide some design rules for the development of more efficient photocatalysts with a wide-spectrum response.

Full text of DOI:10.1039/d0ta11549e

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
The construction of a wide-spectrum-responsive
and high-activity photocatalyst, Bi25CoO40, via the
creation of large external dipoles†
Cite this: J. Mater. Chem. A, 2021, 9,
3616
ab
ab
Zhenlin Wang,ab Yang Lou,ab Ying Zhang,ab Yuming Dong
*
Chengsi Pan,
and Yongfa Zhuabc
Currently, a major issue in the field of photocatalysis is the insufficient utilization of solar energy by
photocatalysts. Thus, to overcome this problem, herein, a new sillenite photocatalyst, Bi25CoO40, was
designed to exhibit a nearly full visible-light spectrum response (up to 750 nm) and high activity. In
particular, its photocatalytic activity for the degradation of typical wastewater pollutants (e.g., MB and 4-
CP) reached up to 3.402 hꢀ1 for MB under visible-light irradiation and 1.576 hꢀ1 for 4-CP under full
spectrum irradiation. The degradation rates are ca. 2–4 times higher than those of well-studied
photocatalysts such as P25-TiO2, Bi2WO6, and g-C3N4, and are also the best for organic pollutant
removal among reported photocatalysts absorbing above 700 nm. Results showed that the remarkable
degradation on Bi25CoO40 is mainly due to the efficient charge separation resulting from the large
external diploe moment between [CoIIIO4] and [BiVO4] tetrahedrons. Results further illustrated that the
VBM of new photocatalysts is composed of Co 3d–O 2p orbitals and exhibits a potential (ca. 2 eV)
positive enough to produce oxidative species for the degradation of organic pollutants. The construction
of a large external dipole moment without sacrificing the valence band edge potential may provide some
design rules for the development of more efficient photocatalysts with a wide-spectrum response.
Received 27th November 2020
Accepted 27th December 2020
DOI: 10.1039/d0ta11549e
rsc.li/materials-a
a narrow bandgap is dramatically reduced, which is probably
1. Introduction
due to the high eꢀ–h+ recombination rate derived from their low
excitation energy or insufficient band edge, making the
production of reactive species difficult.7,8 Therefore the devel-
opment of highly active photocatalysts with an adsorption band
edge above 600 nm is still a major challenge in the eld of
photocatalysis.
Bismuth-based photocatalysts, such as Bi2WO6,9 BiVO4,10
and BiOX (X ¼ Cl, Br, and I),11 are typical photocatalysts with
promising activity for environmental remediation. Their high
activity originates from the dipole moment-induced promotion
of the separation of photogenerated carriers.12 The dipole
moment existing inside distorted [BiO4] or [BiO5] polyhedrons is
created due to the existence of the Bi3+ 6s2 lone pair. The
internal dipole is supposed to solve the issue of the recombi-
nation of photogenerated carriers, leading to high activity for
Photocatalytic oxidation is an energy-saving and environmen-
tally friendly technology that can totally convert toxic organic
pollutants into CO2 and H2O.1 Since this technology is driven by
sunlight, photocatalysts that absorb most of the visible light will
be, in principle, highly efficient because visible light occupies
a large portion of solar energy (ca. 47%).2 Accordingly, many
narrow band gap photocatalysts have been developed with high
activity and an absorption band edge located generally at
around 400–600 nm. Photocatalysts such as sulde,3 nitride,4
iron-based photocatalysts5 and bismuth-based photocatalysts6
have been reported to exhibit superior performances to the
commercial photocatalyst-P25 (TiO2). Thus, currently, photo-
catalysts that respond to a higher wavelength (above 600 nm)
are being researched to achieve the maximum utilization of
solar energy. However, the activity of photocatalysts with
wide-spectrum-responsive photocatalysts (labs
> 600 nm).
However, developed Bi-based photocatalysts with a response at
a wavelength larger than 600 nm do not actually exhibit supe-
rior activity as expected because they suffer from other prob-
lems including multiple defects in their structure, acting as
recombination centres (e.g., Pr-doped Bi2O3 (ref. 13)), low band
edge potential, which is not suitable for producing sufficient
oxidation species (e.g., Bi25MnO40, V.B. ¼ 1.4 eV (ref. 14) or V-
doped Bi25FeO40, VB ¼ 0.8–1.1 eV (ref. 15)), and unstable
chemical components that are easily photo-corroded (e.g.,
aKey Laboratory of Synthetic and Biological Colloids, Ministry of Education, School of
Chemical and Material Engineering, Jiangnan University, Wuxi, Jiangsu, 214122,
China. E-mail: cspan@jiangnan.edu.cn
bInternational Joint Research Center for Photoresponsive Molecules and Materials,
Jiangnan University, Wuxi, Jiangsu 214122, China
cDepartment of Chemistry, Tsinghua University, Beijing 100084, China
† Electronic supplementary information (ESI) available: Detailed experimental
data and additional gures and tables. See DOI: 10.1039/d0ta11549e
3616 | J. Mater. Chem. A, 2021, 9, 3616–3627
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
BiOI16). Thus, considering the research gap between actual prepared via a hydrothermal method, which exhibited both
problems and potential activity advantages, the development of a wide spectrum response (up to 750 nm) and high activity. The
new bismuth-based photocatalysts with a wide spectrum formation of Bi25CoO40, which is associated with the structure
response and high-activity is still of signicant interest.
change resulting from the complete co-substitution method,
Sillenites are photocatalysts with a wide spectrum response, was examined via physical and spectroscopic methods. The
but their activity is strongly limited by their high rate of carrier photocatalytic activity of the prepared Bi25CoO40 photocatalyst
recombination. g-Bi2O3, as one of the simplest sillenite photo- was evaluated via the degradation of typical pollutants
catalyst, is reported to absorb most of the sunlight (labs < 550 including methylene blue (MB) and 4-chlorophenol (4-CP). In
nm) compared to other Bi2O3 phases, such as a-Bi2O3 (labs < 420 particular, the photocatalytic activity of Bi25CoO40 was not only
nm) and b-Bi2O3 (labs < 450 nm), and exhibits a positive valence stable, but also superior to that of well-known photocatalysts,
band edge (V.B. ¼ 3.0 eV).17,18 Thus, this photocatalyst is such as g-C3N4, P25-TiO2 and Bi2WO6. The relationship between
regarded as a parent material for studying other sillenites, the the excellent photocatalytic activity of Bi25CoO40 and complete
structure
of
which
is
.
reported
to
be co-substitution was studied in detail. Furthermore, the effect of
Bi12[(BixBiO3,O)0.8(,BiO4)0.2]O16
In this structure, the tetra- the construction of a large external dipole moment without
19
hedral sites are occupied by Bi3+ cations and Bi vacancies sacricing the valence band edge potential via the co-
(represented as ,), while the octahedral positions are occupied substitution method on the promotion of charge separation
by Bi3+ cations with a 6s2 lone pair of electrons. Both O and Bi and photocatalytic activity was elucidated.
vacancies exist in the tetrahedral sites to complete the coordi-
nation. These vacancies together with the weak driving force for
the separation of charge carriers resulting from the narrow
2. Experimental section
bandgap may lead to a high recombination rate of charge
carriers, and thus limit the activity of g-Bi2O3.
All chemicals were of analytic grade and purchased from the
National Chemical Company (Shanghai, China) and used as
received without further purication. Ultrapure water with
a resistivity of 18.20 MU was used in all experiments.
Regarding the structure difference between g-Bi2O3 and the
standard chemical formula of sillenite (Bi12(SiIVO4)O16),
complete cation co-substitution of the tetrahedral Bi3+ enables
the rearrangement of the tetrahedral atoms, which will ll the
vacancies, thus decreasing the number of recombination
centres. Moreover, a charge difference may form between the
co-substituted cation-centred tetrahedrons if co-substitution is
performed using paired cations with valence states higher or
lower than +4. The formed positive and negative tetrahedrons
will cause an external dipole moment in the structure besides
the internal dipole moment, resulting from the distortion of the
[BiO4] or [BiO5] polyhedrons, which may inhibit the rapid
recombination of charge carries in narrow-bandgap photo-
catalysts and signicantly enhance their photocatalytic activity.
On the other hand, appropriate co-substituted cations will also
narrow the bandgap. This effect has been reported in other
bismuth salts, such as La/Sr co-doped BiFeO3 (ref. 20) and Fe/
Mo co-doped BiVO4.21 Although partial substitution has been
intensively studied, complete cation co-substitution of Bi-based
photocatalysts has been rarely studied, and there is a lack of
2.1 Synthesis of sillenite Bi25CoO40
Bi25CoO40 was prepared via a hydrothermal method. In a typical
procedure, 2.5 mmol BiCl3 and 0.1 mmol Co(NO3)2$6H2O were
added to a 30 mL ethanol/water mixed solution (1 : 1 volume
ratio). Then 10 mL of aqueous 8 mM KOH was added with
stirring and a canary coloured suspension was formed. This
suspension was stirred for another 30 min at room temperature
and then transferred to a 50 mL Teon-sealed stainless steel
autoclave. The autoclave was maintained at 180 ꢁC for 24 h and
then cooled to room temperature. Aer the reaction, the
product was washed several times with water and ethanol until
ꢁ
pH ¼ 7. The product was nally dried at 60 C overnight and
ball-milled at 1400 rpm for 1 h before use. For comparison,
sillenite g-Bi2O3 was prepared in the same manner.
2.2 Characterization of Bi25CoO40
comprehensive understanding on the relationship between the The crystal structure was characterized via powder X-ray
enhancement of photocatalytic activity and the structural diffraction (XRD, Bruker D8-ADVANCE X-ray diffractometer
change resulting from complete cation co-substitution.
German) with a Cu Ka monochromator at 40 kV and 40 mA. The
Herein, Co3+ and Bi5+ were designed to completely co- morphology was characterized via scanning electron micros-
substitute for Bi3+ at the tetrahedral sites of g-Bi2O3 to form copy (SEM, Hitachi S-4800, Japan) and transmission electron
a newly designed sillenite photocatalyst, Bi25CoO40 (Bi24(BiV, microscopy (TEM, JEM 2100plus, JEOL, Japan). The Brunauer–
CoIII)O40). In this design, the 3d orbitals of Co3+ and 6s orbitals Emmett–Teller surface area was determined via nitrogen
of Bi5+ are proposed to contribute to the band edge and narrow absorption–desorption isotherm measurements at 77 K on
the bandgap, while Bi5+ and Co3+ co-substitution may ll the a nitrogen absorption apparatus (Micromeritics, ASAP2020 MP).
vacancies in g-Bi2O3, which will create an external dipole The composition of the product was studied via X-ray photo-
moment between the [CoIIIO4] and [BiVO4] tetrahedrons and electron spectroscopy (XPS, Axis Supra, Kratos, UK) with Al Ka
help enhance the separation of photogenerated charge carriers. monochromatized radiation (hn ¼ 1486.6 eV, 225 W) and ICP-
The crystal structure of Bi25CoO40 has been previously reported AES (PE optima 8000). The XPS spectrum was calibrated using
by Devalette et al.,22 but it has not been studied as a photo- the C 1s peak (284.8 eV) to eliminate the charge effect. The
catalyst to date. The newly designed sillenite photocatalyst was adsorption was recorded using a UV-vis spectrophotometer
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3617

View Article Online
Journal of Materials Chemistry A
Paper
(Persee TU-1901, China). Fluorescence spectra were recorded on cm) and dried in a vacuum oven overnight to form the
an FL spectrophotometer (FS5, PL, Edinburgh, UK).
Bi25CoO40/FTO electrode. The prepared Bi25CoO40/FTO elec-
trode was nally calcined at 300 ꢁC for 1 h before use. The
thickness of the prepared lm was estimated to be ca. 50–100
mm.
2.3 Evaluation of photocatalytic activity
The photocatalytic activity of the prepared samples was evalu-
ated via the photodegradation of methylene blue (MB, 100 mL, 2
ꢂ 10ꢀ5 M) and 4-chlorophenol (50 mL, 5 ppm) under a 300 W Xe
lamp (Perfect Light, Beijing, China) equipped with a 420 nm
cut-off lter. The average light intensity was measured to be 90
mW cmꢀ2. The concentration of the photocatalyst was 0.5 g Lꢀ1
in the suspension. Before the reaction, the suspension was
ultrasonicated for 10 min for dispersion and then stirred for
30 min in the dark to reach absorption–desorption equilibrium.
The saturated adsorption of the organic pollutants in the dark
was about 10% in all cases. During the degradation, 4 mL
suspension was sampled at certain intervals and centrifuged to
remove the contained photocatalyst particles. The concentra-
tion of MB solution was analyzed via UV-vis spectroscopy based
on its maximum absorption peak (663 nm). The concentration
of 4-CP was monitored at 225 nm by high performance liquid
chromatography (HPLC, Ultimate 3000 RS) equipped with a C18
reverse phase column.
3. Results and discussion
3.1 The successful formation of Bi25CoO40 via paired Co3+
and Bi5+ cation co-substitution at the tetrahedral sites of
sillenite
Fig. 1 shows the XRD patterns of Bi25CoO40 prepared via
hydrothermal treatment for different times. In the rst stage (0
h), several broad peaks were observed. All these peaks can be
assigned to bismuth basic nitrate ([Bi6O4.5(OH)3.5]2(NO3)11), as
previously reported,23 indicating that no target sillenite was
formed at this stage. With an increase in the reaction time from
0 h to 12 h, a peak belonging to the sillenite phase was observed.
Notably, the strongest sillenite peak located at ca. 27.5ꢁ was split
into two, indicating that the formed phase was a mixture. The
composition of the mixture is proposed to be sillenite Bi25CoO40
and g-Bi2O3 because the location of the two split peaks is
consistent with the standard patterns of Bi25CoO40 (PDF No. 39-
0871) and g-Bi2O3 (PDF No. 45-1344), respectively. With
a further increase in the reaction time to 24 h, the observed peak
2.4 Photoelectrochemical test
The photocurrent, open circuit potential (OCP) decay and splitting disappeared, indicating the successful formation of
electrochemical impedance spectroscopy (EIS) measurements pure sillenite Bi25CoO40
were performed on an electrochemical work station (CHI 750E, We also carried out control experiments without the addition
.
Chenhua, Shanghai, China). A three-electrode system was of EtOH and with 100% EtOH as the solvent. The XRD patterns
employed for the test with the photocatalyst as the working of the prepared samples are shown in the Supporting Infor-
electrode, a platinum net as the counter electrode, and the mation, Fig. S1.† It was found that without the addition of
standard Ag/AgCl electrode (E ¼ 0.210 V vs. NHE) as the refer- EtOH, XRD peaks belonging to monoclinic a-Bi2O3 instead of
ence electrode. During the test, 0.1 M Na2SO4 was used as the cubic sillenite Bi25CoO40 were observed. In the case of 100%
electrolyte, while a 300 W Xe lamp was used as the light source. EtOH, instead of Bi25CoO40, a metallic Bi phase was formed.
The working electrode of Bi25CoO40 was cast on FTO glass via Thus, the control experiments suggest that both H2O and EtOH
the drop-coating method. In particular, 2–3 mg Bi25CoO40 was are necessary for the formation of pure sillenite Bi25CoO40
dispersed in 100 mL H2O to form a homogeneous slurry. The under the used hydrothermal conditions. The existence of H2O
slurry was then carefully dropped onto an FTO glass (1 cm ꢂ 1.5 will stabilize the Bi3+, which is probably due to the ease of
Fig. 1 (a) XRD patterns of Bi25CoO40 prepared via hydrothermal treatment for different times. (b) XRD refinement for Bi25CoO40 prepared at
180 ꢁC for 24 h with an EtOH/H2O ratio of 1 : 1.
3618 | J. Mater. Chem. A, 2021, 9, 3616–3627
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
oxidation of EtOH by Bi3+ cations under conditions. The exis- result (0.475) and the theoretical ratio (0.5). The rest of the
tence of EtOH will facilitate the formation of Bi25CoO40. C2H5Oꢀ tetrahedral sites were found to be occupied by Bi cations, the
anions will form under basic conditions and coordinate with valence state of which must be 5+ to maintain charge balance.
Bi3+ 24 which may accelerate the exchange of the tetrahedral Bi3+ Therefore, the XRD renement conrmed that both the Bi5+ and
,
with Bi5+ and Co3+, benetting the formation of sillenite Co3+ cations completely substitute the tetrahedral Bi3+ and
Bi25CoO40 vacancies in the parent structure, sillenite g-Bi2O3, as we
.
Fig. 1b shows the XRD renement of Bi25CoO40 prepared at designed.
180 ꢁC for 24 h with an EtOH/H2O (V/V) ratio of 1 : 1. The
The existence of Bi5+ and Co3+ in the sillenite structure was
renement was performed using the Rietveld method and the further conrmed by XPS measurement. Fig. 2 shows the full
parameters are listed in the Supporting Information, Table S1.† XPS spectrum of Bi25CoO40 and narrow scans for Bi 4f, Co 3p,
In Fig. 1b, the circles, red line, orange line and black line and O 1s, respectively. A peak analysis of the narrow scan
represent the raw data, tting line, background and residuals, spectra was performed to better understand the existence of
respectively. The data shows that the XRD pattern is well tted each species. According to the full XPS spectrum in Fig. 2a, only
with few residuals. Rwp and c2, indicators of the tting good- Bi, Co, O and adventitious C signals can be observed. No K or Cl
ness, were calculated to be 9.73% and 1.80, respectively, which signals from the KOH or BiCl3 precursors are observed, indi-
are comparable to other XRD renement results for sillenite cating that the reaction was complete. According to the Bi 4f
structures, such as Bi12FeO20 (Rwp ¼ 6.48%, c2 ¼ 1.47)25 and XPS scan (Fig. 2b), two pairs of peaks can be observed at around
(Bi,Sm)24V2O40 (Rwp ¼ 6.44%, c2 ¼ 1.5).26 The lattice constant 158.6–163.9 eV and 165.4–160.1 eV, which can be assigned to
was calculated to be 10.137 A, which is slightly larger than the Bi3+ and Bi5+, respectively.27 In the Co 2p XPS scan (Fig. 2c), one
˚
˚
theoretical value (10.118 A). The renement results further pair of main peaks and one pair of satellites can be observed,
showed that Co occupies the tetrahedral sites of sillenite with which are located at around 794.9–779.4 eV and 804.9–788.8 eV,
a deduced occupation ratio of 0.492, which is close to our ICP respectively. The peak splitting between the main peaks and the
Fig. 2 XPS spectra of the prepared Bi25CoO40: (a) wide scan, (b) Bi 4f, (c) Co 2p, and (d) O 1s. The dots represent the raw data; the dash line
represents the fitting data for each species; and the red line is the whole fitting data.
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3619

View Article Online
Journal of Materials Chemistry A
Paper
satellites is ca. 9.7 eV. The positions of the main peaks and the where the lattice distance is measured to be 0.282 nm, corre-
peak splitting between the main peaks and the satellites indi- sponding to the (222) planes of the Bi25CoO40 crystals. The BET
cate the existence of Co3+ in the structure.28 Fig. 2d shows the O surface area of Bi25CoO40 was also measured, as shown in the
1s XPS scan, which can be tted into four peaks at 528.6 eV, Supporting Information, Fig. S3.† The data showed that the
529.4 eV, 530.5 eV and 532.4 eV. These peaks can be assigned as prepared sample exhibits a BET surface area of ca. 2.3 m2 gꢀ1
,
Bi–O–Bi, Bi–O–Co, surface O, and surface OH groups, respec- which is comparable to that of other high-activity Bi-salt pho-
tively, according to the literature.29 Thus, the Bi, Co and O XPS tocatalysts with a similar particle size such as BiPO4 (3.0 m2 gꢀ1
spectra demonstrate the successful co-substitution of Bi5+–Co3+ and 150 nm).32
at the tetrahedral sites in sillenite, which is consistent with the
XRD results.
Fig. shows the morphological characterization of
3.2 Wide-spectrum response and high valence band edge
potential of Bi25CoO40 resulting from the co-substitution
3
Bi25CoO40. Fig. 3a and b show the SEM images of the prepared
Bi25CoO40. Bi25CoO40 exhibits a particle-like structure with
a primary particle size of around 20–80 nm. Reported sillenite
particles such as g-Bi2O3 (ref. 30) and Bi25FeO40 (ref. 31) always
show a particle size larger than 1 mm. In our case, the addition
of EtOH clearly inhibited the growth of the primary particles,
resulting in a small size. Fig. 3c shows the TEM image of the
prepared Bi25CoO40, further conrming its morphology. The
TEM image shows that the primary particles easily agglomer-
ated together, forming large secondary particles with a size of
ca. 200 nm–1 mm. Notably, the size of the secondary particles
shown in the image was reduced by the ball milling method.
Without ball milling, the initial secondary particles were larger
(2–20 mm), as shown in the Supporting Information, Fig. S2.†
The lattice image of the prepared Bi25CoO40 is shown in Fig. 3d,
The band structure of Bi25CoO40 is shown in Fig. 4. Fig. 4a and
its inset show the UV-DRS spectra of sillenite Bi25CoO40 and g-
Bi2O3, respectively. The curves show that Bi25CoO40 absorbs up
to 750 nm, covering most of the UV-visible-light region of the
solar spectrum (300–800 nm), while g-Bi2O3 absorbs up to
540 nm. For Bi25CoO40, three peaks located at around 350 nm,
450 nm and 700 nm can be observed. The details of these
transitions will be discussed in the following paragraph. The
bandgap of sillenite Bi25CoO40 and g-Bi2O3 was determined
using the Tauc plots. Accordingly, the bandgap of sillenite
Bi25CoO40 was measured to be 1.70 eV, which is much narrower
than that of g-Bi2O3 (2.75 eV). Since these two compounds
exhibit the same sillenite structure, the narrowing bandgap is
probably due to the contribution of the valence orbitals of the
substituting Bi5+ or Co3+ cations.
Fig. 3 SEM (a and b) and TEM (c and d) images of the prepared Bi25CoO40
3620 | J. Mater. Chem. A, 2021, 9, 3616–3627
.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 4 (a) UV-vis spectra of sillenite Bi25CoO40 and g-Bi2O3. Inset is the Tauc plot of Bi25CoO40 and g-Bi2O3. (b) Experimental and calculated
band edge positions (versus NHE, pH ¼ 0) of sillenite Bi25CoO40 and g-Bi2O3. (c and d) DFT calculation for Bi25CoO40: crystal model (c) and PDOS
(d).
It should be noted that the bandgap of Bi25CoO40 was Bi2O3,38,39 here the value was determined to be 0.1 V. The
determined according to the rst tangent since the transition measured at band potentials for sillenite Bi25CoO40 and g-
does not originate from the impurities but due to the transi- Bi2O3 were 0.096 V and 0.122 V vs. Ag/AgCl, respectively.
tions between the hybrid orbitals formed by the 3d orbitals of Accordingly, the C.B. and valence band (V.B.) of sillenite
the Co3+ cations with O 2p or Bi 6p, as previously reported,33 Bi25CoO40 and g-Bi2O3 were found to be 0.21 V and 1.91 V, and
which will be detailed in the following discussion on the DFT 0.23 V and 2.98 V, respectively. The theoretical calculation was
results. To further conrm that the rst tangent does not orig- performed according to a method based on the absolute
inate from impurities such as vacancies as previously re- electronegativity suggested by Xu et al.37 and detailed in the
ported,34,35 EPR measurements were carried out, as shown in the Experimental section of the Supporting Information.† The
Supporting Information, Fig. S4a.† In the EPR spectra, calculation results show that the band edge positions for sil-
Bi25CoO40 shows a one sixth weaker intensity than that of g- lenite Bi25CoO40 and g-Bi2O3 are 0.35 V (C.B.) and 2.05 V (V.B.),
Bi2O3, indicating that less oxygen vacancies exist in the and 0.36 V (C.B.) and 3.11 V (V.B.), respectively. The calcula-
Bi25CoO40 lattice. Since the small amount of oxygen vacancies in tion results are similar to our experimental results, indicating
g-Bi2O3 only show a slight response (around 800 nm) in the UV- the good accuracy of the determination of the band edge
DRS spectrum, the much less oxygen vacancies in Bi25CoO40 will positions. The XPS valence band spectrum of Bi25CoO40 was
lead to a negligible absorbance.
further measured to conrm the measured band edges (shown
Fig. 4b shows the band edge positions for sillenite in the Supporting Information,† Fig. 4d). Again, similar band
Bi25CoO40 and g-Bi2O3, which were determined by the edge positions were obtained. Notably, the valence band edge
measurement of the at band potential36 and theoretical positions for sillenite Bi25CoO40 were close to 2 V by all the
calculation.37 The Supporting Information, Fig. S4b and c,† methods, indicating co-substitution will maintain the rather
shows the Mott–Schottky plots for sillenite Bi25CoO40 and g- strong oxidative ability of the new photocatalyst.
Bi2O3, respectively, revealing their at band potentials. The
The band structure of Bi25CoO40 was further studied via the
conduction band (C.B.) is suggested to be 0.1–0.3 V above the DFT method (detailed in the Supporting Information†). Fig. 4c
at band potentials.36 Referring to previous studies on and d show the crystal structure used for the calculation and the
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3621

View Article Online
Journal of Materials Chemistry A
Paper
partial density of state (PDOS) of Bi25CoO40. A homogeneous Bi
and Co distribution was used to simplify the calculation and
also based on our XRD renement results. Using this model, the
valence band maximum (VBM) was calculated to be mainly
composed of Co 3d and O 2p, while the conduction band
minimum (CBM) was mainly composed of O 2p–Bi 6p. Notably,
the 3d orbitals of the Co atoms in the [CoIIIO4] tetrahedron
contribute mainly to the valence band, and only a little to the
conduction band, which is oen seen in transition metal-based
sillenite (e.g. Bi12TiO20) and other Bi–Co–O systems (e.g.
BiCoO3).40,41 The 6s and 6p orbitals of the Bi atoms in the [BiVO4]
tetrahedron show a slight contribution to the VBM and CBM,
indicating the delocalization of their valence shell electrons.
The DFT results also explain the three peaks observed in the UV-
DRS spectrum of Bi25CoO40. The peak at 350 nm is similar to
that observed in g-Bi2O3, which is due to the band transition of
the O 2p to Bi 6p orbitals, as previously reported.42 The peaks at
450 nm and 700 nm are due to the band transitions from Co 3d–
O 2p to the higher energy Bi 6p orbitals (located above 2.0 eV)
and to lower-energy hybrid orbitals (located between 1.6–2.0
eV), that is, Bi 6p, O 2p, and a slight contribution from the Co 3d
orbitals, respectively. The DFT study indicated that upon light
irradiation, eꢀ is excited to the O 2p–Bi 6p orbitals, while h+
remains in the Co 3d–O 2p orbitals.
3.3 Remarkable degradation and complete mineralization of
MB and 4-CP on wide-spectrum-responsive Bi25CoO40
photocatalyst
Fig. 5 shows the photodegradation of MB and 4-CP on the
prepared photocatalysts. The typical model pollutants, MB and
4-CP, were used to evaluate the photocatalytic activity, as
previously reported.43,44 Fig. 5a and Supporting Information,
Fig. S5† show the degradation of MB on the Bi25CoO40, g-Bi2O3,
g-C3N4 and Bi2WO6 photocatalysts. g-C3N4 and Bi2WO6, which
were prepared according to our previously reported methods,
were employed for comparison because they are generally
regarded as photocatalysts with high visible-light activity.45,46
On sillenite Bi25CoO40, 95% of MB was degraded within 50 min,
while within the same time, only ca. 30% of MB was degraded
on g-Bi2O3 (Supporting Information, Fig. S5†). The kinetic rate
constant in the degradation process was calculated to be 3.402
hꢀ1 for sillenite Bi25CoO40, which is nearly 14 times that for g-
Bi2O3 (0.251 hꢀ1) (Fig. 5a). Remarkable activity enhancement
was observed on Bi25CoO40 relative to g-Bi2O3, indicating that
the co-substitution of Co3+ and Bi5+ for tetrahedral Bi3+ in g-
Bi2O3 may be benecial to enhance the photocatalytic activity.
On the other hand, MB was degraded quicker on Bi25CoO40
than that on g-C3N4 and Bi2WO6. Within 50 min irradiation, ca.
90% and 52% of MB was degraded on g-C3N4 and Bi2WO6,
Fig. 5 (a) Photodegradation rate of MB on Bi25CoO40, g-Bi2O3, g-C3N4 and Bi2WO6 (l > 420 nm). (b) Wavelength-dependent activity of
Bi25CoO40 for the degradation of MB. (c) Stability of Bi25CoO40 for the photodegradation of MB. (d) Photodegradation curves of 4-CP on
Bi25CoO40 and P25-TiO2 under full-spectrum irradiation and the concurrent TOC removal on Bi25CoO40 photocatalyst (Xe lamp, l > 300 nm).
3622 | J. Mater. Chem. A, 2021, 9, 3616–3627
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
respectively, while more than 95% of MB was degraded on sunlight, which was compared to the widely used photocatalyst,
Bi25CoO40 (Supporting Information, Fig. S5†). The kinetic rate P25-TiO2. 4-CP was employed as a general model for colourless
constant for the degradation of MB on Bi25CoO40 was 3.4 times and minimally biodegradable pollutants.1 The photo-
and 2.2 times higher than that on g-C3N4 (0.996 hꢀ1) and degradation of 4-CP and the concurrent total organic carbon
Bi2WO6 (1.576 hꢀ1), respectively (Fig. 5a). The photocatalytic (TOC) removal are shown in Fig. 5d and Supporting Information
activity of Bi25CoO40 is the best among the reported sillenite Fig. S8.† The HPLC results (Supporting Information, Fig. S8†)
photocatalysts with a visible-light response (generally, 4 h is and degradation results (Fig. 5d) show that the concentration of
needed for more than 90% degradation of pollutants, as 4-CP decreased rapidly aer 30 min irradiation; meanwhile, the
detailed in the Supporting Information, Table S2†). Also, our concentration of intermediates such as benzoquinone (BQ) and
photocatalyst exhibits better activity for pollutant oxidation hydroquinone (HQ) increased. With an increase in time, the
than other narrow bandgap photocatalysts absorbing above concentration of 4-CP and benzoquinone (HQ) decreased to
700 nm, such as Cu-salts and suldes (Supporting Information, nearly zero, while the concentration of HQ continued to
Table S3†). For the oxidation of organic pollutants, the activity increase due to the further oxidation of BQ. The accumulated
of Bi25CoO40 is only lower than the newly reported Bi8(CrO4)O11 HQ could be further degraded and converted into CO2 and H2O
photocatalyst, which contains Cr6+ in its lattice and has in 4 h according to the TOC results shown in Fig. 5d. For
a response slightly below 700 nm.
comparison, the photodegradation of 4-CP was also carried out
Fig. 5b and Supporting Information S6† show the on P25. In this case, the activity of Bi25CoO40 (2.28 hꢀ1) was
wavelength-dependent activity of Bi25CoO40 for the degradation nearly as twice that of P25 (1.05 hꢀ1), indicating its superior
of MB. Its activity is presented as series of kinetic rate constants, activity. In summary, Bi25CoO40 shows superior degradation
which were obtained under various wavelength cutoff lters, activity for both MB and 4-CP, indicating that the good activity
ranging from 300 nm to 730 nm. In the gure, the kinetic rate does not originate from specic pollutants, but due to the
constants for the degradation of MB decrease with an increase nature of the photocatalyst.
in the cut-off wavelength. The decreasing tendency is generally
the same as the absorption curves of Bi25CoO40 (Fig. 4a), indi-
cating the photocatalytic activity originates from the band
excitation of Bi25CoO40. Notably, Bi25CoO40 showed apparent
activity under the wavelengths larger than 730 nm. In this case,
3.4 The formation of large external dipoles in Bi25CoO40 and
their effect on the enhancement of charge separation and
photocatalytic activity
only Bi25CoO40 absorbs light, not MB. This demonstrates again Photocatalytic activity is related to multiple factors, such as
that the activity is due to band excitation, not the excitation of particle size, nanostructure, crystal structure, band edge
the MB dye. This wavelength-dependent activity suggests that potential, light adsorption, and separation efficiency of photo-
Bi25CoO40 is a promising photocatalyst with a wide-spectrum generated electrons and holes. In our case, Bi25CoO40 can be
response above 700 nm.
regarded as the co-substitution of Bi5+ and Co3+ cations for
Next, the stability of Bi25CoO40 was tested, as shown in tetrahedral Bi3+ in g-Bi2O3. The two compounds prepared via
Fig. 5c, which gure shows that its activity for the degradation a similar method exhibited the same crystal and morphological
of MB did not change aer 5 cycles, indicating its good stability. structure (Supporting Information, Fig. S9†). Besides, g-Bi2O3
According to the DFT results, Bi5+ and Co3+ contribute little to exhibited a more positive valence band edge potential than that
the CBM, and thus, electron-induced photo-corrosion may for Bi25CoO40 (Fig. 4). Therefore, g-Bi2O3 is expected to show
barely occur on Bi25CoO40 compared that in other photo- higher activity than Bi25CoO40. Nevertheless, Bi25CoO40 exhibi-
catalysts containing Bi5+ or Co3+, such as LaCoO3 and NaBiO3, ted much better activity than g-Bi2O3. One reason for this may
which may be responsible for the observed stability.
be that Bi25CoO40 absorbs more light than g-Bi2O3. However,
The oxidative species produced during the photodegradation this may not be the decisive factor for the observed activity
of MB were also studied (Supporting Information S7†). Research enhancement. Many reported sillenites (Supporting Informa-
has indicated that hydroxyl radicals (OHc), holes (h+) and super tion, Table S2†) also exhibit a narrower bandgap relative to g-
oxide radicals (O2ꢀc) are the typical oxidative species produced Bi2O3, but they do not show obviously activity enhancement.
on photocatalysts, which initiate the photocatalytic degrada- Aer excluding all the above factors, the most possible reason
tion. These species can be trapped by formic acid (FA), tert-butyl for the excellent activity of Bi25CoO40 is proposed to be the
alcohol (t-BuOH) and benzoquinone (BQ), respectively. In our suppression of the charge carrier recombination rate, which
case, the addition of t-BuOH and BQ did not inuence the may result from co-substitution, as discussed below.
activity much, whereas the addition of FA strongly inhibited the
activity. This suggests that h+, which is produced on Bi25CoO40
PEC measurements, such as transient photocurrent, elec-
trochemical impedance measurement, surface photovoltage
,
may be the main active species involved in the degradation of measurement and OCP decay, are regarded as sensitive indi-
MB. Simultaneously, eꢀ produced on Bi25CoO40 will form super cators for studying the recombination rate.1 Fig. 6 shows the
oxide radicals with O2, which are supposed to participate in the PEC results for Bi25CoO40 and g-Bi2O3 to clarify the difference in
oxidation process of the fragments derived from the decom- recombination rate between them. Fig. 6a shows the transient
posed pollutants.
photocurrent of Bi25CoO40 and g-Bi2O3 under illumination and
Encouraged by the superior activity of Bi25CoO40, its full- in the dark. The transient photocurrent signals were recorded at
spectrum activity was evaluated to simulate its activity under each light on/off interval (40 s). Under irradiation, the transient
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3623

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 6 (a) Transient photocurrent for Bi25CoO40 and g-Bi2O3, (b) electrochemical impedance measurement for Bi25CoO40 and g-Bi2O3, (c)
Surface P–V spectra for Bi25CoO40 and g-Bi2O3, and (d) open circuit potential (OCP) decay for Bi25CoO40 and g-Bi2O3.
photocurrent of Bi25CoO40 reached up to 1.1 mA cmꢀ2, while the to g-Bi2O3, suggesting that the former sample has better charge
current of g-Bi2O3 was reduced to 0.3 mA cmꢀ2. The higher transfer ability compared to g-Bi2O3 under illumination.
current indicates a higher separation efficiency of photo-
Fig. 6c shows the surface P–V spectra (SPV) for Bi25CoO40 and
generated carriers in Bi25CoO40 than that in g-Bi2O3.46 These g-Bi2O3. On the one hand, both Bi25CoO40 and g-Bi2O3 exhibit
results are also consistent with the LSV results shown in the a surface photovoltaic response together with their UV-DRS
Supporting Information, Fig. S10.†
spectrum, indicating that the observed photocatalytic activity
Fig. 6b shows the electrochemical impedance spectra (EIS) is due to band excitation. On the other hand, the SPV response
under illumination and in the dark for Bi25CoO40 and g-Bi2O3. A of photocatalysts was reported to result from the separation of
Randles circuit was used to simulate the impedance results, as the photogenerated electron–hole pairs.48 In our case, the SPV
shown in the inset of Fig. 6b, which has been veried to be of Bi25CoO40 is obviously higher than that of g-Bi2O3, suggesting
effective in many TiO2-based systems.47 The tting parameters that Bi25CoO40 has a better separation efficiency of photo-
are listed in the Supporting Information, Table S4.† In Fig. 6b, generated carriers relative to g-Bi2O3.
both Bi25CoO40 and Bi2O3 exhibit similar semicircles under
Fig. 6d shows the open circuit potential (OCP) decay curves for
illumination and in the dark, but with different radii. In Bi25CoO40 and g-Bi2O3. The OCP decay has been reported to be
particular, the samples show a smaller radius under illumina- capable of determining differences in the photogenerated charge
tion relative to that in the dark, indicating a decrease in the separation process.49 In particular, the decay in the OCP is due to
charge transfer resistance. This is because illuminating samples the recombination of photogenerated electron–hole pairs, which
will increase the generation of electrons and holes, and further leads to changes in the at band potential. The faster the OCP
increase the conductivity. The detailed tting results show that decays, the faster the photogenerated electron–hole pairs decay.
Rct, an indicator for the charge transfer in the powder electrode, During the measurement, Bi25CoO40 and Bi2O3 were rst irradi-
decreased from 29.3 kU in the dark to 6.12 kU under illumi- ated under illumination, and then kept in the dark to relax the
nation for Bi25CoO40, while for g-Bi2O3, it decreased from 66 kU OCP. As seen in Fig. 6d, the potential change upon UV illumi-
to 39 kU. Clearly, Bi25CoO40 shows a larger Rct change compared nation was ꢃ0.1–0.2 V for both photocatalysts, which is
3624 | J. Mater. Chem. A, 2021, 9, 3616–3627
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
commonly observed in photocatalysts.49 The increase in the OCP electrons coordinated. The metal cation with a valence state of
under illumination for Bi25CoO40 compared to g-Bi2O3 indicates +4 occupies a tetrahedral site with 4 O atoms coordinated. Each
that Bi25CoO40 excites more electrons than g-Bi2O3. The kinetic tetrahedron is linked by two octahedrons. In the case of g-Bi2O3,
constant for the decay process can be tted by using exponential the tetrahedral site is occupied by Bi3+ and Bi vacancy with
regression.49 Accordingly, the kinetic rate constants for the decay a total of 4 O2ꢀ and O vacancies coordinated. For Bi25CoO40, Bi3+
process of Bi25CoO40 and Bi2O3 were determined to be 7.80 ꢂ and the Bi vacancy are co-substituted by Co3+ and Bi5+ cations,
10ꢀ4 sꢀ1 and 1.47 ꢂ 10ꢀ3 sꢀ1, respectively. Bi25CoO40 exhibits while the O vacancy is occupied by O2ꢀ for charge balance.
a smaller kinetic rate constant compared to g-Bi2O3, suggesting Accordingly, the number of vacancies that act as recombination
that Bi25CoO40 has a better carrier separation efficiency.
centres greatly decrease, which may be partially benecial for
Fig. 7a and b show the PL and PL decay spectra for Bi25CoO40 charge separation.
and g-Bi2O3, respectively. The PL intensity reects the recom-
More importantly, the structure of Bi25CoO40 reveals that the
bination rate of photogenerated carriers in Bi25CoO40 and g- Co3+ and Bi5+ cations occupy the same tetrahedral sites but
Bi2O3 since they have the same structure and particle size. possess a lower and higher valence state relative to that in the
According to Fig. 7a, it can be found that g-Bi2O3 exhibits a PL standard sillenite (+4), respectively. This will form a negative
peak intensity (ca. 468 nm) four times higher than that for and positive net charge for the Co3+ and Bi5+-centred tetrahe-
Bi25CoO40, indicating that Bi25CoO40 has a slow recombination dron, respectively. The charge difference will cause the forma-
rate. The PL decay spectra for Bi25CoO40 and g-Bi2O3, as shown tion of an external dipole moment between the paired
in Fig. 7b, supply more direct evidence for the difference in tetrahedrons. The large external lattice dipole may cause a built-
charge separation efficiency because the measurement reects in electric led and facilitate charge separation, as discussed
the lifetime of the photogenerated carriers. The decay curves below.
were tted according to exponential regression (detailed in the
The electron density difference was calculated via the DFT
Supporting Information†) and plotted as solid lines. As shown method to show effect of the above-mentioned external dipole
in Fig. 7b, the PL peak intensity of Bi25CoO40 decays slower than moment on the charge separation. (Detailed in the Experi-
that of g-Bi2O3. Notably, a hump was observed at 2 ns, which is mental section of the Supporting Information†). Fig. 8b shows
due to the photomultiplier effect of the instrument, as previ- the electron density difference on the (110) planes of sillenite
ously reported.50 In particular, Bi25CoO40 and g-Bi2O3 showed Bi25CoO40. The (110) plane is shown because the Co3+- and
a decay time of 2.76 ns and 0.53 ns, respectively. The longer Bi5+-centred tetrahedrons are both located in this plane and
decay time of Bi25CoO40 indicates the longer lifetime of its the difference between them is the largest. According to
photogenerated carriers and slow recombination rate.
Fig. 8b, it is clear that the valence shell electrons distribute
Thus, the PEC measurement and PL results conrm that the mostly around the Co3+-centred tetrahedrons and O2ꢀ, while
excellent activity in sillenite Bi25CoO40 is due to its efficient the positive charge must distribute around the Bi5+-centred
charge separation. Since the only difference between sillenite tetrahedrons to maintain the charge balance. Thus, an
Bi25CoO40 (a high-activity photocatalyst) and g-Bi2O3 (a low- external dipole moment will form between [CoIIIO4] and
activity photocatalyst) is the existence of an external dipole [BiVO4], as we supposed. The value of the lattice dipole was
moment between the [CoIIIO4] and [BiVO4] tetrahedrons in the calculated to be 30.1 D, which is considerably large compared
former structure, rapid charge separation may result from this, to that of reported bismuth-salt photocatalysts, such as BiPO4
as discussed below.
(10.5 D) and Bi8CrO4O11 (22.23 D).51 In the previous reports,
Fig. 8a details the structural change between sillenite distortion of the [VO4] or [TiO6] polyhedrons formed an
Bi25CoO40 and g-Bi2O3. In the common sillenite structure, Bi3+ internal dipole moment and facilitate charge separation.52,53
occupies an octahedral site with 5 O atoms and a lone pair of In layered photocatalysts such as Bi2WO6, the dipole moment
Fig. 7 (a) PL spectra of Bi25CoO40 and g-Bi2O3 and (b) PL decay spectra of Bi25CoO40 and g-Bi2O3. Excitation wavelength: 350 nm.
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3625

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 8 (a) Structure difference between sillenite Bi25CoO40 and g-Bi2O3. (b) Electron density difference on the (110) plane of the Bi25CoO40
lattice (2 ꢂ 1 ꢂ 1), and schematic relationship between the excitation of photogenerated carriers and the dipole moment in sillenite Bi25CoO40
.
between the [Bi2O2] and [WO4] layers were also reported to activity is the best among the reported photocatalysts respon-
facilitate charge separation.54 In our case, when the compound sive above 700 nm and also superior to other well-known pho-
was excited, electrons are transferred from the Co 3d–O 2p tocatalysts, such as P25-TiO2, g-C3N4 and Bi2WO6. The excellent
orbitals to the Bi 2p–O 2p orbitals, namely, from the [CoIIIO4] performance was demonstrated to result from the large external
tetrahedrons to the [BiIIIO5] octahedrons (Fig. 8b), according dipole created between the [CoIIIO4] and [BiVO4] tetrahedrons
to our PDOS results, as shown in Fig. 4. The formed external associated with a considerably positive valence band edge (ca. 2
dipole moment may help the charge transfer by keeping the eV). The development of Bi25CoO40 as a wide-spectrum-
holes in the [CoO4] tetrahedrons and repelling the electrons to responsive and high-activity photocatalyst and the under-
the [BiO5] octahedrons, leading to efficient charge separation. standing of its high-activity mechanism may give some guide-
Also, although the Co3+ and Bi5+ co-substitution in Bi25CoO40 lines for the design of more efficient photocatalysts with
did not dramatically decrease the valence band potential, a narrow bandgap in the future.
Bi25CoO40 still maintained a rather positive valence band edge
potential, which was sufficient to produce oxidative species for
degrading organic pollutants. Thus, the highly efficient charge
separation induced by the external dipole associated with the
The authors claim that there are no conicts to declare.
positive valence band edge potential may be responsible for
Conflicts of interest
the activity observed.
Acknowledgements
4. Conclusion
This work was nancially supported by the Natural Science
In summary, a newly designed sillenite, Bi25CoO40, was reported Foundation of Jiangsu Province (BK20201345), National Natural
for the rst time as a wide-spectrum-responsive and high- Science Foundation of China (21908079, 21676123, 21902009,
activity photocatalyst for the degradation of organic pollut- 21707052), Startup Funding at Jiangnan University
ants. The photocatalyst exhibited a wavelength response up to (1045210322190170, 1045281602190010, 1042050205204100,
750 nm, and a quick degradation rate for MB and 4-CP. The 1045219039200010).
3626 | J. Mater. Chem. A, 2021, 9, 3616–3627
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
26 G. M. Kuz'micheva, T. I. Mel'nikova, I. A. Kaurova and
N. B. Bolotina, Mater. Lett., 2017, 199, 75–78.
27 C. Lv, G. Chen, X. Zhou, C. Zhang, Z. Wang, B. Zhao and
D. Li, ACS Appl. Mater. Interfaces, 2017, 9, 23748–23755.
28 C. Pan, R. G. Nuzzo and A. A. Gewirth, Chem. Mater., 2017,
29, 9351–9359.
29 T. D. Nguyen, Q. T. P. Bui, T. B. Le, T. M. Altahtamouni,
K. B. Vu, D.-V. N. Vo, N. T. H. Le, T. D. Luu, S. S. Hong and
K. T. Lim, RSC Adv., 2019, 9, 23526–23534.
References
1 J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang,
Y. Horiuchi, M. Anpo and D. W. Bahnemann, Chem. Rev.,
2014, 114, 9919–9986.
2 Z. Zou, J. Ye, K. Sayama and H. Arakawa, Nature, 2001, 414,
625–627.
3 L. Cheng, Q. Xiang, Y. Liao and H. Zhang, Energy Environ.
Sci., 2018, 11, 1362–1391.
30 H. Jing, X. Chen and X. Jiang, Micro & Nano Lett., 2012, 7,
357–359.
4 C. Pan, T. Takata, M. Nakabayashi, T. Matsumoto,
N. Shibata, Y. Ikuhara and K. Domen, Angew. Chem., Int.
Ed. Engl., 2015, 54, 2955–2959.
¨
31 R. Koferstein, T. Buttlar and S. G. Ebbinghaus, J. Solid State
Chem., 2014, 217, 50–56.
32 C. Pan, D. Li, X. Ma, Y. Chen and Y. Zhu, Catal. Sci. Technol.,
2011, 1, 1399–1405.
33 S. Sun, W. Wang, L. Zhang and M. Shang, J. Phys. Chem. C,
2009, 113, 12826–12831.
34 A. Kumar, A. Kumar and V. Krishnan, ACS Catal., 2020, 10,
10253–10315.
5 M. Mishra and D.-M. Chun, Appl. Catal., A, 2015, 498, 126–141.
6 X. Meng and Z. Zhang, J. Mol. Catal. A: Chem., 2016, 423, 533–549.
7 S. K. Loeb, P. J. J. Alvarez, J. A. Brame, E. L. Cates, W. Choi,
J. Crittenden, D. D. Dionysiou, Q. Li, G. Li-Puma, X. Quan,
D. L. Sedlak, T. David Waite, P. Westerhoff and J.-H. Kim,
Environ. Sci. Technol., 2019, 53, 2937–2947.
8 F. Zhang, A. Yamakata, K. Maeda, Y. Moriya, T. Takata,
J. Kubota, K. Teshima, S. Oishi and K. Domen, J. Am.
Chem. Soc., 2012, 134, 8348–8351.
35 Y. Huang, N. Zhang, Z. Wu and X. Xie, J. Mater. Chem. A,
2020, 8, 4978–4995.
36 Z.-H. Cui and H. Jiang, J. Phys. Chem. C, 2017, 121, 3241–3251.
37 X. Yong and A. A. S. Martin, Am. Mineral., 2000, 85, 543–556.
38 S. R. Alharbi and A. F. Qasrawi, Mater. Res., 2019, 22,
e201807222.
39 M. Hojamberdiev, G. Zhu, H. Lu, M. Kumar, M. Wang and
J. Gao, J. Mater. Sci.: Mater. Electron., 2018, 30, 2610–2621.
40 D. Hou, X. Hu, Y. Wen, B. Shan, P. Hu, X. Xiong, Y. Qiao and
Y. Huang, Phys. Chem. Chem. Phys., 2013, 15, 20698–20705.
41 U. Koroglu, S. Cabuk and E. Deligoz, Solid State Sci., 2014,
34, 1–7.
42 H. Tian, F. Teng, J. Xu, S. Lou, N. Li, Y. Zhao and M. Chen,
Sci. Rep., 2015, 5, 7770.
43 R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga,
Science, 2001, 293, 269–271.
44 Y. Cui, J. Huang, X. Fu and X. Wang, Catal. Sci. Technol.,
2012, 2, 1396–1402.
45 H. Fu, C. Pan, W. Yao and Y. Zhu, J. Phys. Chem. B, 2005, 109,
22432–22439.
9 L. Zhang and Y. Zhu, Catal. Sci. Technol., 2012, 2, 694–706.
10 A. Malathi, J. Madhavan, M. Ashokkumar and
P. Arunachalam, Appl. Catal., A, 2018, 555, 47–74.
11 S. Garg, M. Yadav, A. Chandra and K. Hernadi, J. Nanosci.
Nanotechnol., 2019, 19, 280–294.
12 H. Deng, W. Hao, H. Xu and C. Wang, J. Phys. Chem. C, 2011,
116, 1251–1255.
13 G. Wang, X. Chen, L. Li and G. Shi, Micro & Nano Lett., 2018,
13, 657–660.
14 X. Wu, M. Li, J. Li, G. Zhang and S. Yin, Appl. Catal., B, 2017,
219, 132–141.
15 Y. Ni, D. Shi, B. Luo, Z. Yang, Z. Xiong, C. Lu, L. Fang, Y. Xia,
J. Gong, L. Cao and Z. Gao, Inorg. Chem., 2019, 58, 6966–6973.
16 D. S. Bhachu, S. J. A. Moniz, S. Sathasivam, D. O. Scanlon,
A. Walsh, S. M. Bawaked, M. Mokhtar, A. Y. Obaid,
I. P. Parkin, J. Tang and C. J. Carmalt, Chem. Sci., 2016, 7,
4832–4841.
17 M. Weber, R. D. Rodriguez, D. R. T. Zahn and M. Mehring,
Inorg. Chem., 2018, 57, 8540–8549.
18 W. Hao, Y. Gao, X. Jing, W. Zou, Y. Chen and T. Wang, J.
Mater. Sci. Technol., 2014, 30, 192–196.
19 S. F. Radaev, V. I. Simonov and Y. F. Kargin, Acta Crystallogr.,
Sect. B: Struct. Sci., 1992, 48, 604–609.
20 S. Irfan, L. Li, A. S. Saleemi and C.-W. Nan, J. Mater. Chem. A,
2017, 5, 11143–11151.
21 R. Liu, J. Ren, D. Zhao, J. Ning, Z. Zhang, Y. Wang, Y. Zhong,
C. Zheng and Y. Hu, Inorg. Chem. Front., 2017, 4, 2045–2054.
22 M. Devalette, J. Darriet, M. Couzi, C. Mazeau and
P. Hagenmuller, J. Solid State Chem., 1982, 43, 45–50.
23 E. A. Abdullah, A. H. Abdullah, Z. Zainal, M. Z. Hussein and
T. K. Ban, E-J. Chem., 2012, 9, 1885–1896.
46 C. Pan, J. Xu, Y. Wang, D. Li and Y. Zhu, Adv. Funct. Mater.,
2012, 22, 1518–1524.
ˆ
˜
47 J. Angelo, P. Magalhaes, L. Andrade and A. Mendes, Appl.
Surf. Sci., 2016, 387, 183–189.
48 T.-f. Xie, D.-j. Wang, L.-j. Zhu, T.-j. Li and Y.-j. Xu, Mater.
Chem. Phys., 2001, 70, 103–106.
49 N. Baram and Y. Ein-Eli, J. Phys. Chem. C, 2010, 114, 9781–
9790.
50 C. Lewis, W. R. Ware, L. J. Doemeny and T. L. Nemzek, Rev.
Sci. Instrum., 1973, 44, 107–114.
51 Y. Zhu, X. Ma, Y. Xu and X. Chen, Natl. Sci. Rev., 2020, 7, 652–
659.
52 S. Tokunaga, H. Kato and A. Kudo, Chem. Mater., 2001, 13,
4624–4628.
53 J. Sato, H. Kobayashi, K. Ikarashi, N. Saito, H. Nishiyama and
Y. Inoue, J. Phys. Chem. B, 2004, 108, 4369–4375.
54 K. Morita, J.-S. Park, S. Kim, K. Yasuoka and A. Walsh, J.
Phys. Chem. C, 2019, 123, 29155–29161.
24 K. Kato, K. Suzuki, K. Tanaka, D. Fu, K. Nishizawa and
T. Miki, Integr. Ferroelectr., 2010, 62, 133–140.
25 T. I. Mel’nikova, G. M. Kuz’micheva, N. B. Bolotina and
N. V. Sadovskaya, Crystallogr. Rep., 2014, 59, 155–159.
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 3616–3627 | 3627