A Facet-Dependent Schottky-Junction Electron Shuttle in a BiVO4{010}–Au–Cu2O Z-Scheme Photocatalyst for Efficient Charge Separation

Article abstract of DOI:10.1002/adfm.201801214

Interfacial charge separation and transfer are the main challenges of efficient semiconductor-based Z-scheme photocatalytic systems. Here, it is discovered that a Schottky junction at the interface between the BiVO₄ {010} facet and Au is an eff

Full text of DOI:10.1002/adfm.201801214

FULL PAPER
Charge Separation
www.afm-journal.de
A Facet-Dependent Schottky-Junction Electron Shuttle
in a BiVO4{010}–Au–Cu2O Z-Scheme Photocatalyst
for Efficient Charge Separation
Chenguang Zhou, Shaomang Wang, Zongyan Zhao, Zhan Shi, Shicheng Yan,*
and Zhigang Zou
mimic natural photosynthesis,
a
two-
Interfacial charge separation and transfer are the main challenges of
efficient semiconductor-based Z-scheme photocatalytic systems. Here, it is
discovered that a Schottky junction at the interface between the BiVO4 {010}
facet and Au is an efficient electron-transfer route useful for constructing a
high-performance BiVO4{010}–Au–Cu2O Z-scheme photocatalyst. Spectro-
scopic and computational studies reveal that hot electrons in BiVO4 {010}
more easily cross the Schottky barrier to expedite the migration from BiVO4
{010} to Au and are subsequently captured by the excited holes in Cu2O.
This crystal-facet-dependent electron shuttle allows the long-lived holes and
electrons to stay in the valence band of BiVO4 and conduction band of Cu2O,
respectively, contributing to improved light-driven CO2 reduction. This unique
semiconductor crystal-facet sandwich structure will provide an innovative
strategy for rational design of advanced Z-scheme photocatalysts.
photon pathway, called
a
Z-scheme
photocatalytic system, was created with a
double-component photosystem and was
expected to easily achieve a high quantum
efficiency. The Z-scheme system success-
fully eliminates the inherent bandgap
engineering contradiction of single-com-
ponent photosystems, i.e., the trade-off
between an extended light-harvesting
range and degraded redox capacity.[3,4] Typ-
ically, a Z-scheme photocatalytic system
consists of an electron-transfer mediator
that connects two semiconductors, which
are called photosystem I (PS I) and II
(PS II). If PS I and II have a more positive
valence band (VB) level and more negative
conduction band (CB) level, respectively,
the light-excited electrons in the VBs of PS I and II moving into
their CBs will make the VB of PS I a strong electron acceptor
and the CB of PS II a strong electron donor when an electron-
transfer mediator is used to deliver electrons in the CB of PS I
into the VB of PS II. Therefore, both an excellent redox ability
and spatial separation of photogenerated charges can be created
in a Z-scheme photocatalytic system by combining two narrow-
bandgap semiconductors, and these properties are beneficial
for sufficient light harvesting.
In a Z-scheme system, electron migration between the two
photocatalysts is the rate-determining process.[5,6] Ionic redox
couples, such as Fe3+/Fe2+ and IO3−/I−, have been shown to
be effective in shuttling electrons in Z-scheme systems.[3,4]
Unfortunately, the effectiveness of ionic redox couples
depends on the reaction environment and is probably weak-
ened by back reactions. Alternatively, a series of solid-state
electron mediators with noble-metal nanoparticles (Au, Ag, Pt,
etc.) or metal-free conductive materials (carbon layer, reduced
graphene oxide, etc.) may be more favorable to ensure a con-
tinuous flow of electrons between the two photosystems.[6–9]
Great efforts have been devoted to improve intrinsic conduc-
tive properties and structural modifications. An Au layer or
1. Introduction
Artificial photosynthetic techniques to convert carbon dioxide
into carbon-based fuels have attracted attention as economical
and renewable strategies to meet ever-growing global energy
demands and solve associated environmental issues.[1,2] To
C. G. Zhou, Dr. S. M. Wang, Prof. S. C. Yan, Dr. Z. Shi
Eco-materials and Renewable Energy Research Center (ERERC)
Collaborative Innovation Center of Advanced Microstructures
College of Engineering and Applied Sciences
Nanjing University
No. 22, Hankou Road, Nanjing, Jiangsu 210093, P. R. China
E-mail: yscfei@nju.edu.cn
Prof. Z. Y. Zhao
Faculty of Materials Science and Engineering
Kunming University of Science and Technology
No. 61 Wenchang Road, Kunming, Yunnan 650093, P. R. China
Prof. Z. G. Zou
Macau Institute of Systems Engineering
Macau University of Science and Technology
Macau 999078, P. R. China
Prof. Z. G. Zou
Jiangsu Key Laboratory for Nano Technology
National Laboratory of Solid State Microstructures
Department of Physics
carbon film has been used as an electrical connection between
[8,9]
Mo-doped BiVO4 and (La, Rh)-codoped SrTiO3
to achieve
Nanjing University
No. 22, Hankou Road, Nanjing, Jiangsu 210093, P. R. China
highly efficient overall water splitting. However, studies on
the available semiconductor–conductor contact interface,
which plays a crucial role in manipulating the interfacial
charge transfer determined by the interface energetics, are
lacking.
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adfm.201801214.
DOI: 10.1002/adfm.201801214
©
Adv. Funct. Mater. 2018, 1801214
1801214 (1 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Theoretically, when the work function of a metal is greater
than that of an n-type semiconductor, a Schottky junction can
be established after their close contact to effectively hinder
the back transfer of photoexcited hot electrons trapped by the
metal.[4,10–12] Different atomic coordination environments on
different facets induce significant distinctions in work func-
tions.[12,13] Thus, the formation of a Schottky junction depends
on whether the facets on a semiconductor have an applicable
work function to establish a Schottky junction with a metal.
Recently, facet engineering was used to produce a reduction
facet for electron accumulation and an oxidation facet for hole
accumulation based on different surface work functions, which
led to different surface band bending.[12,14–19] Typically, obvious
work function differences are observed on adjacent facets, such
as {100} and {111} of Cu2O,[12] {110} and {010} of BiVO4,[14,15]
{001} and {110} of SrTiO3,[16] {100} and {111} of CeO2,[17] and
{001} and {101} of TiO2,[18,19] effectively inducing spatial charge
separation. Evidently, forming a Schottky junction on the reduc-
tion facet of a semiconductor by anchoring a metal is beneficial
for enhancing hot-electron injection into the metal because of
the high density of hot electrons on a reduction facet and the
enhanced surface electric states on a semiconductor due to metal
modification, which accelerates electron transfer to metal.[20] In
addition, the Schottky barrier offers a unidirectional electron-
transfer route from the semiconductor to the metal, enhancing
efficient charge separation. Thus, manipulating the metal con-
tact with a specific facet of a semiconductor will result in a pro-
nounced effect on the variable interface charge kinetics.[14,16]
Therefore, selectively assembling a metallic electron mediator
with a specific facet of a semiconductor in a Z-scheme photocat-
alytic system is an effective strategy to swiftly transfer electrons
and effectively suppress reverse electron transfer, enhancing the
spatial charge separation in the two photosystems.
Here, for the first time, we constructed an efficient electron-
migration pathway in a ternary BiVO4–Au–Cu2O heterostruc-
ture by anchoring an Au–Cu2O composite configuration onto
the {010} facet of a BiVO4 truncated octahedron through a
step-by-step synthetic route. The synthesized BiVO4{010}–Au–
Cu2O allows inert BiVO4 and Cu2O to be activated and exhibits
increased CO2 reduction activity, i.e., three and five times greater
than that of BiVO4{110}–Au–Cu2O and BiVO4{010}–Cu2O,
respectively. BiVO4{010}–Au–Cu2O simultaneously shows excel-
lent stability under prolonged irradiation due to inhibited photo-
corrosion in Cu2O. Single-particle and time-resolved photolumi-
nescence (PL) spectroscopy analyses and theoretical calculations
revealed that a desirable Schottky junction forms at the interface
between the BiVO4 {010} facet and Au. The photoexcited elec-
trons in the BiVO4 {010} facet can tunnel through the Schottky
barrier and directionally recombine with photogenerated holes
in Cu2O, leaving long-lived holes in the VB of BiVO4 and long-
lived electrons in the CB of Cu2O to increase photocatalysis.
and described in detail in the Experimental Section. Briefly,
BiVO4 truncated octahedron with coexposed {010} and {110}
facets was prepared via a solid–liquid state reaction.[21] Au nan-
oparticles were selectively deposited on the surfaces of {010}
BiVO4 through a photoreduction method using HAuCl4·4H2O
as a precursor because photogenerated electrons mostly accu-
mulate on the {010} surface of BiVO4. To obtain BiVO4{110}–
Au, Au4+ was reduced into Au particles on the electron-deficient
BiVO4 {110} facets because the BiVO4 {010} facets were coated
by a polyvinyl pyrrolidone (PVP) insulating layer to hinder the
reduction of Au4+. A hemispherical Cu2O shell layer was sub-
sequently coated on the Au nanoparticles by a facile aqueous
solution approach.[22,23] The X-ray diffraction (XRD) patterns
of these as-prepared BiVO4-based samples exhibited all the
identified peaks of BiVO4 crystals, which were indexed to
single-phase, monoclinic BiVO4 (JCPDS NO.14-0688). A vis-
ible diffraction peak with 2θ = 38.2° was observed for both
BiVO4–Au and BiVO4–Au–Cu2O materials, indicating that Au
particles were successfully deposited on the BiVO4 surfaces.
The weak diffraction peak of Cu2O was apparent (JCPDS
NO.65-3288), indicating Cu2O was effectively coated onto the
BiVO4 and BiVO4–Au crystals (Figure 1b).
The microstructures and morphologies of the as-prepared
samples were examined by field-emission scanning electron
microscopy (FE-SEM) and transmission electron microscopy
(TEM) observations. As shown in Figure 1c–e and Figure S1
(Supporting Information), truncated octahedron-shaped BiVO4
crystals with a length of ≈1 µm and a smooth surface are
symmetrically enclosed by the {010} and {110} facets. After
photoreduction Au deposition, Au particles (30–60 nm) were
evenly distributed on the {010} facets of BiVO4, and the {110}
facets remained clear, implying that the electrons mostly accu-
mulate on the {010} facet under irradiation. In contrast, Au
particles with similar sizes were mainly deposited on the {110}
facet when PVP surfactant was preadsorbed on the {010} facet,
resulting in Au4+ ions being captured only by electrons from
the {110} facets. Figure 1f,g shows that epitaxial Cu2O pre-
fers to anchor on the Au surface to establish hemispherical,
core–shell Au–Cu2O on the BiVO4 {010} and {110} facets,
respectively. Further observations by TEM indicated that the
Cu2O shell with a thickness of ≈200 nm is tightly coated on
the Au particle surface. In addition, the high-resolution TEM
(HRTEM) images show that the lattice fringe spacing of the
shell is 0.290 nm, which can be assigned to the lattice pitch
of the Cu2O (110) atomic plane (Figure 1h).[24] Figure 1i shows
the elemental mapping for the Au–Cu2O core–shell structure,
further revealing the spatial distribution of Bi, Au, Cu, and O
in BiVO4–Au–Cu2O. As a reference, the BiVO4{010}–Cu2O het-
erostructure (Figure S2, Supporting Information) and Cu2O
particles were also synthesized through a photoreduction route
using CuCl2 as the precursor. The compositions of various
BiVO4-based samples were further confirmed by X-ray photo-
electron spectroscopy (XPS) analyses (Figure S3, Supporting
Information). UV–vis absorption measurements (Figure S4,
Supporting Information) showed that the similar visible-light
absorption of BiVO4{010}-Au and BiVO4{110}–Au was broader
relative to that of the pristine BiVO4 crystal, which was ascribed
to localized surface plasmon resonance absorption from the
Au particles.[5,25] Moreover, the absorption spectrum of BiVO4
2. Results and Discussion
2.1. Synthesis and Characterization of the Z-Scheme
Photocatalysts
The facet-dependent Z-scheme BiVO4–Au–Cu2O heterostruc-
tures were synthesized as illustrated schematically in Figure 1a
©
Adv. Funct. Mater. 2018, 1801214
1801214 (2 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 1. a) Schematic illustration of the synthetic process for the BiVO4-based heterostructures. b) XRD patterns of the as-prepared samples. SEM
images of the as-synthesized c) BiVO4 truncated octahedron, d) BiVO4{010}–Au, e) BiVO4{110}–Au, f) BiVO4{010}–Au–Cu2O, and g) BiVO4{110}–
Au–Cu2O. h) TEM and i) elemental mapping images of the Au–Cu2O core–shell heterostructure, and the corresponding HRTEM image is shown in
the inset of (h).
{010}–Cu2O improved after the deposition of Cu2O on the
BiVO4 surface, which was due to the characteristic absorption
of Cu2O. Obviously, the surface plasmon resonance absorption
of Au particles and the hybrid absorption of Cu2O and BiVO4
were incorporated into absorption spectrum of the BiVO4–Au–
Cu2O composite. A visible absorption peak at 750 nm in the
UV–vis absorption spectrum of Cu2O results from the narro-
wband gap absorption of CuO.[26] An XPS analysis (Figure S5,
Supporting Information) showed that a Cu+ oxidation state
with a binding energy of 952.3 eV for Cu 2p1/2 existed on
the surface of Cu2O, which indicated that a small amount of
amorphous CuO formed on the Cu2O due to inevitable surface
oxidation.
the weak hole oxidation ability of the Cu2O VB, which result in
the inert photocatalytic activities of BiVO4 and Cu2O even when
their surfaces are modified with Au particles.[6,8,9,27] The photo-
catalytic activities of these as-synthesized BiVO4-based samples
were then investigated for CO2 photoreduction under visible-
light illumination (λ ≥ 420 nm). As seen in Table 1, BiVO4,
Cu2O, BiVO4{010}–Au, and BiVO4{110}–Au exhibited no or
slight photocatalytic activities for CO2 reduction. CH4 and CO
evolved over the Cu2O catalyst, which was in good agreement
with a previous report on the generation of various products
over Cu-based catalysts.[27] The heterostructure BiVO4{010}–
Cu2O composite created by coating Cu2O on the BiVO4{010}
facet obviously had a significant enhancement in CH4 and CO
evolution.
Under irradiation, if the BiVO4{010}–Cu2O composites
conform to traditional type-II heterojunction band alignment
(Figure 2a), the photoinduced holes in the VB of BiVO4 will
transfer to the VB of Cu2O, and electrons will flow through
their CBs in the reverse direction, leading to poor photocatalytic
performance of the heterostructures because the accumulated
electron reduction level in the CB of BiVO4 is insufficient for
CO2 photoreduction. However, the significant increase in CH4
2.2. Photocatalytic CO2 Reduction
According to the valence band XPS and UV–vis diffuse reflec-
tance spectra (Figure S6, Supporting Information), the CB and
VB positions in BiVO4 and Cu2O were estimated to be –0.1 and
2.43 eV and –1.65 and 0.35 eV, respectively. This result sug-
gested the weak electron reduction ability of the BiVO4 CB and
©
Adv. Funct. Mater. 2018, 1801214
1801214 (3 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Table 1. Photocatalytic CO2 reduction tests of the BiVO4-based Z-scheme photocatalysis systems under visible-light irradiation.
Entrya)
PS l
Electron mediator
PS ll
AQEb)
Activity [µmol h1 g−1
]
CH4
–
CO
–
1
2
3
4
5
6
7
8
BiVO4
–
–
Cu2O
–
–
–
–
–
0.03
0.04
0.04
1.22
3.14
0.59
3.15
0.02
–
BiVO4{010}
BiVO4{110}
BiVO4{110}
BiVO4{010}
BiVO4{010}
BiVO4{010}
Au
Au
Au
Au
–
–
–
–
–
Cu2O
Cu2O
Cu2O
Cu2O–Au
1.12
2.02
0.45
2.08
0.38
0.43
0.32
0.44
Au
a)Conditions: photocatalyst (0.1 g); light source, 300 W Xe lamp with a cut-off filter (λ ≥ 420 nm); top-irradiation cell with a Pyrex window; b)Apparent quantum efficiency
(AQE) for the BiVO4-based heterostructures irradiated under 500 nm monochromatic light.
and CO generation implied that the charge-transfer process
between Cu2O and BiVO4 is not a type-II heterojunction band
alignment. A reasonable Z-scheme charge-transfer route is that
the photogenerated electrons in the CB of BiVO4 transfer to the
contact interface and recombine with holes in the VB of Cu2O
(Figure 2b), leading to long-lived holes and electrons in the VB of
BiVO4 and CB of Cu2O, respectively. This charge-transfer process
is beneficial for both enhancing the charge-separation efficiency
and maintaining the strong redox ability of the photocatalysts.
The CO2 reduction experiment performed in the dark or in the
absence of a photocatalyst resulted in no detected products,
which indicated that the CO2 reduction process with the photo-
catalyst is driven by light. Both CH4 and CO were detected over
the Cu2O or Cu2O-containing composite catalysts, which may
imply that CO2 reduction occurs on the surface of Cu2O.
The CH4 and CO generation rates over BiVO4{010}–Au–
Cu2O were ≈5.3 and 4.5 times higher, respectively, than those
over BiVO4{010}–Cu2O, demonstrating that Au particles offer
effective channels as electron shuttles to promote high-mobility
directional electron migration between BiVO4 and Cu2O
(entries 6 and 7 in Table 1). To further prove the Z-scheme
transfer mechanisms, the oxidation and reduction activities
of holes and electrons in the BiVO4–Au–Cu2O heterostruc-
tures were investigated by PL detection with the assistance
of scavengers. Using terephthalic acid as a probe molecule, a
single-particle PL technique can effectively detect •OH in solu-
tion since •OH easily reacts with terephthalic acid to produce
2-hydroxyterephthalic acid, a fluorescent product with a PL
peak at 425 nm.[28,29] In aqueous solutions, •OH, the primary
oxidant, is generated via direct hole oxidation[30] or photo-
generated electron-induced multistep reduction of O2 (O2
+
e− → •O2, •O2 + e− + 2H+ → H2O2, H2O2 + e− → •OH + OH−).[31]
As seen in Figure S7a,b (Supporting Information), when
Ethylenediaminetetraacetic acid disodium salt (EDTA-Na2)
was used as an efficient hole scavenger, no •OH formed on
BiVO4, implying that •OH cannot be generated by a multistep
reduction of O2 because the CB potential of BiVO4 is probably
−
more positive than that of the O2/•O2 couple (−0.046 V vs
NHE).[32] However, in the presence of EDTA-Na2, PL signals
were detected at 425 nm for Cu2O, BiVO4{010}–Cu2O and
BiVO4{010}–Au–Cu2O, and the PL intensity followed the order
of BiVO4{010}–Au–Cu2O > BiVO4{010}–Cu2O > Cu2O. This
result shows that the electron potential of the Cu2O CB is suf-
−
ficiently negative to reduce O2 into •O2 , generating •OH by a
multistep reduction of O2. Similarly, when p-benzoquinone
was used as an electron scavenger, Cu2O did not exhibit a PL
Figure 2. A scheme to describe the possible charge-transfer mechanism between BiVO4 and Cu2O. a) Type-II heterojunction and b) Z-scheme system.
©
Adv. Funct. Mater. 2018, 1801214
1801214 (4 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
signal since the VB potential of Cu2O is more negative than
the Eo of the •OH /H2O couple (2.38 V vs NHE).[32] The PL
peak intensity followed the order of BiVO4{010}–Au–Cu2O >
BiVO4{010}–Cu2O > BiVO4 (Figure S7c,d, Supporting Informa-
tion), confirming efficient charge separation by Au and that the
holes on the BiVO4 VB are positive enough to oxidize H2O into
•OH. These results indicated that Au acts as a carrier mediator
to promote electron migration between BiVO4 and Cu2O via
the Z-scheme mechanism. Indeed, BiVO4{010}–Au–Cu2O pos-
sessed the highest photoreduced activity (entries 5 and 6 in
Table 1), which is 2.6 and 1.8 times higher for CH4 and CO,
respectively, than those of BiVO4{110}–Au-Cu2O. Thus, delib-
erately coating Au–Cu2O onto the different facets of BiVO4 may
potentially be responsible for the unusual enhancement in the
photocatalytic performance of BiVO4 {010}–Au–Cu2O.
The amount of CO2 physically adsorbed on BiVO4{010}–
Au–Cu2O, BiVO4{110}–Au–Cu2O, BiVO4{010}–Cu2O, and
BiVO4 was evaluated by the Brunauer–Emmett–Teller (BET)
method to be 1.35, 1.18, 1.02, and 0.91 mg g−1, respectively.
The amounts of CO2 adsorbed on these samples were normal-
ized by the specific surface areas to be 0.41–0.43 mg m−2. Thus,
the CO2 adsorption, which was nearly the same for all sam-
ples, did not induce the difference in the CO2 photoreduction
activity (Figures S8 and S9 and Table S1, Supporting Informa-
tion). Transient photocurrent response measurements showed
that the current density on the BiVO4{010}–Au–Cu2O electrode
was obviously higher than that on the BiVO4{110}–Au-Cu2O or
BiVO4{010}–Cu2O electrode, implying a higher separation effi-
ciency of photoexcited electron–hole pairs in the BiVO4{010}–
Au–Cu2O sample (Figure S10, Supporting Information). The
CO2 photoreduction apparent quantum efficiency (AQE) was
measured with monochromatic light at 500 nm and calculated
according to the following equation (1):
catalysts exhibited the same radiative PL processes. The PL
intensity followed the order of BiVO4 > BiVO4{010}–Cu2O >
BiVO4{011}–Au > BiVO4{010}–Au > BiVO4{110}–Au–Cu2O >
BiVO4{010}–Au–Cu2O, confirming that Au and/or Cu2O
modification enhanced the charge-separation efficiency. Direct
observation of the change in the PL intensity on BiVO4 crystal
facets (see squares enclosed by broken lines in Figure 3b–h)
was carried out to further understand the charge behavior.
Figure 3b shows a typical PL image of a BiVO4 decahedron
under irradiation from a 488 nm laser. The slightly inhomoge-
neous distribution of the PL intensity on the crystal indicated
the spatial variation of the emission sites, which was ascribed
to the intrinsic anisotropic distribution of charges on the BiVO4
crystal facets. When Au particles were selectively deposited on
{010} or {110} facets, apparent fluorescence quenching phe-
nomena were observed (Figure 3c,d) under the same experi-
mental conditions, revealing that the PL intensity of the BiVO4
{110} facet on the BiVO4{010}–Au catalyst is weaker than that
of the BiVO4 {010} facet on the BiVO4{110}–Au catalyst. The
decreased PL intensities can be ascribed to the injection of
photogenerated electrons from BiVO4 to Au particles prior to
electron–hole pair capture by trapping sites, effectively sup-
pressing radiative charge recombination. By coating Cu2O
on the Au particles, the PL intensities of the BiVO4 {110}
facet on BiVO4{010}–Au–Cu2O and the BiVO4 {010} facet
on BiVO4{110}–Au–Cu2O further decreased but retained the
order of BiVO4{110} > BiVO4{010} (Figure 3f,g). No distinct
fluorescence signals were found in the PL image of Cu2O par-
ticles, indicating that the detected PL signals were all from
BiVO4 (Figure 3e). These results demonstrated that electrons
migrating to the Au surface were instantly consumed by the
holes from Cu2O, which reduced their chance of returning to
the BiVO4 trapping sites. On the other hand, the PL intensity
of the BiVO4{110} facet on BiVO4{010}–Cu2O showed only a
small decline, suggesting that the embedded Au particles pri-
marily contribute to charge transfer at the interface between
BiVO4 and Cu2O (Figure 3h). Given the significant distinction
in PL quenching between {110} and {010} facets, the electron-
transfer efficiency from BiVO4 to Au heavily relies on variation
of the contact interfaces.
To acquire more information about this optical response, we
measured the PL decay profiles of various BiVO4-based pho-
tocatalysts. As seen in Figure 3i, the PL decay curves of these
samples were mathematically fitted by a biexponential model
(see the Supporting Information for details). In the inset
table in Figure 3i, the short-lifetime component (τ1) is attrib-
uted to nonradiative recombination of free-excitons with sur-
face defects, and the long-lifetime component (τ2) is assigned
to the radiative pathway from photogenerated electron and
hole recombination.[5,34] Relative to that of pristine BiVO4
(τ2 = 0.43 ns), the transient PL decay traces of BiVO4{010}–
Cu2O and BiVO4{010}–Au revealed the PL lifetimes (τ2)
increased to ≈0.65 ns and ≈0.87 ns, respectively, suggesting
that the photogenerated electrons in the CB of BiVO4 more
easily transferred to Au than directly transferring to Cu2O.
Additionally, the PL lifetime (τ2) of BiVO4{010}–Au–Cu2O fur-
ther increased to ≈1.0 ns, indicating that electrons from BiVO4
depleting Cu2O holes limits charge recombination opportuni-
ties and results in a longer electron survival time in BiVO4.
8 × number of CH4 molecules + 2× number of CO molecules
AQE =
× 100%
numeber of incident photons
(1)
The AQE of Z-scheme heterostructure BiVO4{010}–Au–
Cu2O, BiVO4{110}–Au–Cu2O, and BiVO4{010}–Cu2O was 0.43,
0.38, and 0.32, respectively, at 500 nm. The same trend of the
AQE and photocatalytic activities suggested that the CO2 reduc-
tion reaction was a result of the light-driven catalytic reaction.
BiVO4{010}–Au–Cu2O exhibited superior photoreduction activ-
ities, indicating that preferentially coating Au–Cu2O onto the
BiVO4 {010} crystal facet may have a strong positive association
with its excellent interparticle charge-transfer ability.
2.3. Experimental Evidence for Electron Transfer
by Schottky Junction
To investigate the interface-related dynamics of charge carriers,
a single-particle PL technique was employed to study PL inten-
sity changes on BiVO4 crystal facets coated with Au, Cu2O, or
Au–Cu2O. As shown in Figure 3a, a broad PL band (550–800 nm)
with a peak at 680 nm was observed on BiVO4 and can be
assigned to radiative recombination of photogenerated charges
trapped by crystal defects.[33] The similar, quenched character-
istic PL peaks suggested that all the BiVO4-based composite
©
Adv. Funct. Mater. 2018, 1801214
1801214 (5 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 3. a) Steady-state photoluminescence spectra of the as-prepared samples determined for individual BiVO4 crystals under 488 nm laser irradia-
tion. Single-particle fluorescence imaging on b) BiVO4 single-crystal particles, c) BiVO4 {010}–Au, d) BiVO4{110}–Au, e) Cu2O, f) BiVO4{010}–Au–Cu2O,
g) BiVO4{110}–Au–Cu2O, and h) BiVO4{010}–Cu2O. Scale bars are 1 µm. i) PL decay traces of the as-prepared samples and their corresponding
fluorescence lifetimes are shown in the inset of Figure 2i (BiVO4, Au and Cu2O are denoted as B, A, and C, respectively).
The lifetime component (τ2) for the BiVO4{010} interface-
related heterostructures modified with Au and Au–Cu2O is
0.87 and 1.0 ns, respectively. These lifetimes are both longer
than those for the corresponding configurations established on
the BiVO4{110} crystal facet, which implies that the fabricated
BiVO4{010}–Au interface may accelerate photoexcited electron
transfer from BiVO4 to Au, prolonging the carrier lifetime.
Electrochemical impedance spectroscopy (EIS) measurements
under irradiation showed an obvious decrease in the semicircle
diameters of BiVO4{010}–Au and BiVO4{010}–Au–Cu2O com-
pared with those of BiVO4{110}–Au and BiVO4{110}–Au–Cu2O
(Figure S11, Supporting Information), suggesting that the
BiVO4{010}–Au interface has better electrical conductivity.
The surface band alignment of a semiconductor–metal–
semiconductor hybrid structure is the dominant factor that
determines the interfacial dynamic behavior of photogenerated
carriers across the heterointerface. To acquire insight into this
issue, slab models of BiVO4 {010} and {110} surfaces were con-
structed, and their electronic structures were calculated by first-
principles simulations (Figure S12, Supporting Information).
As shown in Figure S13 (Supporting Information), the surface
potential analysis shows that the work functions of the {110}
and {010} surfaces are ≈5.51 and 4.24 eV, respectively, which
can be ascribed to their inherent variation in atomic coordi-
nation. Given that the Fermi level position of bulk BiVO4, an
n-type semiconductor, is near the CB, the surface band natu-
rally bends upward for thermodynamic equilibrium. The band
bending of the {110} facet is obviously higher than that of the
{010} facet owing to its high work function; i.e., a more pro-
nounced built-in electric field exists beneath the {110} facet and
drives the accumulation of spatial holes and electrons on the
{110} and {010} surfaces, respectively. This result agrees with
previous reports using surface photovoltage or photodeposition
investigations.[14,15] Upon understanding the facet-preferential
©
Adv. Funct. Mater. 2018, 1801214
1801214 (6 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 4. Electrostatic potentials for a) BiVO4{010}–Au and b) BiVO4{110}–Au interfaces obtained from first-principles calculations. Time course of
c) CH4 and d) CO evolution over BiVO4-based heterostructures (BiVO4, Au, and Cu2O are denoted as B, A, and C, respectively) under irradiation from
a 300 W Xe lamp with a cut-off filter (λ ≥ 420 nm). e) Cu 2p XPS spectra of BiVO4-based heterostructures after illumination for 20 h. f) Schematic
diagrams illustrating the surface potential positions of BiVO4 {010} and {110} and the Fermi level position of Au.
migration of electrons in BiVO4, i.e., from the bulk to the sur-
face, we further evaluated the effect of the BiVO4–Au contact
interface on the subsequent transfer of electrons from the
BiVO4 surface to a noble metal. According to the potentials
(Figure 4a,b), the potential of Au is ≈1.07 eV lower than that
of BiVO4 {010}, which means that a Schottky junction can
be constructed at the BiVO4{010}–Au interface to accelerate
photogenerated electron capture and effectively hinder the
back transfer of photoexcited, hot electrons trapped by metal.
In contrast, an electron diffusion layer rather than a Schottky
junction may form at the BiVO4{110}–Au interface because of
the similar work functions of Au and BiVO4 {110}, resulting in
reduced electron migration at this interface.
Eventually, we assessed the electron transport behavior
from Au to Cu2O. Considering that the work function of Cu2O
(4.8 eV) is close to that of Au (5.1 eV),[12,35] concentration dif-
fusion may be the driving force for electrons in Au recom-
bining with the holes in Cu2O because the band alignment is
relatively flat and does not have the effect of a Schottky junc-
tion (Figure S14, Supporting Information). In fact, when equal
amounts of Au particles were photodeposited on BiVO4{010}–
Au–Cu2O (denoted BiVO4{010}–Au–Cu2O–Au), the evolution
rates of the reduced products only slight increased compared to
those of BiVO4{010}–Au–Cu2O (entry 8 in Table 1), suggesting
that the Au–Cu2O interface does not have a significant impact
on the electron migration. Figure 4c,d shows the time course of
CO2 reduction over various samples for 20 h with an evacuation
after every 5 h of irradiation. The photocatalytic activity (CO2
to CH4 and CO) of BiVO4{010}–Au–Cu2O remains almost
unchanged after four cycles, whereas BiVO4{110}–Au–Cu2O
and BiVO4{010}–Cu2O only maintain 70% and 52%, respec-
tively, of their initial performance after the fourth cycle.
XPS analysis was employed to further investigate the sur-
face chemical states of BiVO4{010}–Cu2O, BiVO4{010}–Au–
Cu2O, and BiVO4{110}–Au–Cu2O after the photoreduction
reaction. As shown in Figure 4e, the Cu 2p core-level spec-
trum exhibited two main peaks, 2p3/2 at 932.6 eV and 2p1/2
at 952.3 eV, suggesting the characteristic +1 oxidation state.
Two additional peaks with higher binding energies at 934.6
and 954.5 eV can be ascribed to a higher valence state, i.e.,
Cu2+.[27,36] Notably, the Cu2+ 2p3/2 and 2p1/2 peaks observed
for BiVO4{010}–Au–Cu2O are obviously weaker than those
of BiVO4{110}–Au–Cu2O and especially BiVO4{010}–Cu2O,
implying that the self-oxidization of Cu+ to Cu2+ was effectively
suppressed in BiVO4{010}–Au–Cu2O during the photoreduc-
tion process. Based on the above results, we concluded that
the difficulty of the photoexcited electrons on BiVO4 crossing
the BiVO4–Au interfaces determines the charge-separation effi-
ciency in Cu2O and controls the photoreduction activity of the
entire Z-scheme system. In BiVO4{010}–Au–Cu2O, an efficient
electron-transfer channel at the BiVO4{010}–Au interface was
constructed to expedite the electron migration from BiVO4 to
Cu2O, resulting in a large number of holes in Cu2O that can be
consumed by BiVO4 electrons. Consequently, the self-oxidation
©
Adv. Funct. Mater. 2018, 1801214
1801214 (7 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Scheme 1. Scheme to describe the carrier migration behavior from BiVO4 to Cu2O with and without Au particles deposited on the different facets of
BiVO4.
process (Cu+ + h+ → Cu2+) and photogenerated charge recombi-
or Cu2O on the surface of BiVO4. To synthesize BiVO4{010}–Au, the
as-prepared BiVO4 powder (0.1 g) and Au precursor (HAuCl4) with a
3 wt% weight ratio of products to BiVO4 were mixed in deionized water to
obtain 100 mL of a suspension. The suspension was irradiated by a 300 W
nation are simultaneously suppressed, producing an enhanced
stability and better photocatalytic activity. The inferior perfor-
mances of both BiVO4{110}–Au–Cu2O and BiVO4{010}–Cu2O
illustrate that an available Schottky junction plays a critical role
in achieving fast electron migration across an interface in a
Z-scheme system (Figure 4f and Scheme 1).
Xe lamp (λ ≥ 420 nm) with magnetic stirring for 7 h. The obtained products
were filtered, washed with deionized water several times, and dried at
60 °C overnight. Using a similar irradiation procedure, BiVO4{010}–
Cu2O was obtained using an aqueous solution containing BiVO4
(0.1 g, 100 mL) and CuCl2 (1.5 mL, 0.1 m). To synthesize BiVO4{110}–Au,
polyvinyl pyrrolidone (PVP) (0.1 g) and BiVO4 powder (0.1 g) were mixed
in deionized water (100 mL). Then, an HAuCl4 aqueous solution (3 wt%,
weight ratio of Au to BiVO4) was added. The resulting suspension was
irradiated by a 300 W Xe lamp with a 420 nm filter for 7 h. The obtained
product was filtered, washed with deionized water several times, and
dried at 60 °C overnight.
3. Conclusions
In summary, a fantastic crystal-facet-dependent Z-scheme
system with BiVO4{010}–Au–Cu2O was successfully con-
structed by directionally anchoring an Au–Cu2O cell-shell layer
onto the {010} facet of a BiVO4 truncated octahedron for the
first time, and the system exhibited a significant improve-
ment in CO2 photoreduction with long-lived activity. The
single-particle PL analysis results and theoretical simula-
tions suggested that the formation of a Schottky junction
in the BiVO4{010}–Au interface facilitates the extraction of
photogenerated electrons from BiVO4 to Cu2O, which is ben-
eficial for prolonging the survival times of electrons in the CB
of Cu2O and holes in the VB of BiVO4. This work combines
the advantages of surface-induced charge directional distribu-
tion and an optimal metal–semiconductor contacted interface
to provide a fast electron-transfer channel with activated photo-
catalytic activity and may provide a new strategy for developing
more efficient Z-scheme photocatalysts.
Preparation of BiVO4{010}–Au–Cu2O and BiVO4{110}–Au–Cu2O
Samples: The synthesis of BiVO4{010}–Au–Cu2O and BiVO4{110}–Au–
Cu2O follows the same procedure. Typically, a CuCl2 aqueous solution
(1.5 mL, 0.1 m), sodium dodecyl sulfate (SDS) surfactant (1 g), and
BiVO4{010}–Au or BiVO4{110}–Au powder (0.05 g) were mixed in
deionized water (90 mL). After magnetic stirring for 1 h, a hydroxylamine
hydrochloride (NH2OH·HCl) solution (2.5 mL, 0.2 m) was slowly
dropped into the above suspension. After 4 h of stirring, the color of the
suspension changed from blue to light brown. The light-brown product
was collected by suction filtration, washed with distilled water, and dried
at 50 °C under vacuum.
Sample Characterizations: The crystallographic structure of the
as-synthesized samples was confirmed with XRD (Rigaku Ultima III)
using Cu Kα radiation at 40 kV and 40 mA. The morphology and
distribution of elements in the products were observed by SEM (Nova
NanoSEM 230FEI Co.), TEM (JEM-200CX) and energy-dispersive X-ray
spectroscopy (EDS) techniques. Ultraviolet–visible (UV–vis) diffuse
reflectance spectra were recorded by a UV–vis spectrophotometer
(UV-2550, Shimadzu), and the corresponding absorption spectra were
calculated by a Kubelka–Munk equation. The BET surface area was
obtained by using a surface area analyzer (Micromeritics Tristar-3000,
USA) at the temperature of liquid nitrogen (77 K). The amounts of
CO2 physisorbed on the synthesized samples were evaluated by BET
method at the temperature of ice water (273 K). The constituents of the
samples were characterized by XPS (Thermo ESCALAB 250). For the
single-particle PL measurements, the photoluminescence spectra were
recorded on an FLS 980 spectrometer (Edinburgh Instruments, UK).
A 488 nm CW laser (Coherent, OBIS 488LS) was employed to excite
BiVO4 and Cu2O. The emission images were recorded on an electron-
multiplying charge-coupled device (EMCCD) camera (Ultra897, Andor
Technology) at a rate of 20 frames s−1. A dichroic mirror (Di03-R488,
Semrock) and long-pass filter (BLP01−514R, Semrock) were used to
4. Experimental Section
Preparation of BiVO4 Truncated Octahedron: In a typical synthesis
procedure,[21,37] BiVO4 truncated octahedron with coexposed {010} and
{110} crystal facets was prepared via a simple solid–liquid state reaction.
Typically, Bi2O3 powder (5 mmol) and V2O5 powder (5 mmol) were
added to a HNO3 solution (50 mL, 0.5 m). After ultrasonic dispersion,
the mixture was stirred for 96 h at room temperature. The obtained vivid
yellow powder was rinsed with deionized water thoroughly and dried at
60 °C for 12 h.
Preparation of BiVO4{010}–Au, BiVO4{110}–Au, and BiVO4{010}–
Cu2O Samples: A photoreduction method was employed to deposit Au
©
Adv. Funct. Mater. 2018, 1801214
1801214 (8 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
improve the signal-to-noise ratio. All experimental operations were
carried out at room temperature. The data were analyzed using the open
source image software ImagePro. All the photoluminescence spectra
were measured in air unless otherwise noted. The time-resolved
fluorescence decay curves of the as-fabricated products were estimated
on an FLS 980 spectrometer (Edinburgh Instruments, UK) using a
picosecond Ti:Sapphire laser (Mira HP, from Coherent) as the excitation
source. The decay profiles of the synthesized products were fitted by a
biexponential function of I(t) = A1 exp(−t/τ1) + A2 exp(−t/τ2), where I(t)
is the fluorescence intensity at time t, A1 and A2 are the preexponential
factors, and τ1 and τ2 are the lifetimes.
Fundamental Research Funds for the Central Universities (021314380133
and 021314380084). The authors are grateful to the High Performance
Computing Center (HPCC) of Nanjing University for doing the numerical
calculations in this paper on its IBM Blade cluster system.
Conflict of Interest
The authors declare no conflict of interest.
Photocatalytic CO2 Conversion Test: The photocatalytic activity of the
as-synthesized samples was estimated based on the photocatalytic
reduction of CO2 under irradiation from a 300 W Xe lamp with a
cut-off filter (λ ≥ 420 nm). Typically, the prepared powders (0.1 g) were
homogeneously dispersed on a glass reactor with an area of 4.2 cm2.
The volume of the reaction system was ≈230 mL. High-purity CO2 gas
(99.999%) was introduced into the reaction system to reach ambient
pressure after the system was vacuum-treated several times. Then,
0.4 mL of deionized water was injected into the reaction system as the
reducing agent. During irradiation, ≈1 mL of gas was withdrawn from
the reaction tank at a given interval and used for the CH4 concentration
analysis by gas chromatography (GC-2014, Shimadzu Corp, Japan). The
AQE for CO2 reduction was measured using the same experimental
procedure except a monochromatic bandpass filter (Y50, 500 nm) was
employed. The number of photons reaching the catalyst was measured
using a spectroradiometer (LS-100, EKO Instruments Co., LTD.).
Assuming that the incident photons were completely absorbed by the
catalyst, the total number of incident photons at 500 nm was estimated
to be 2.36 × 1021 photon h−1.
Theoretical Calculations: All calculations were performed using the
Cambridge Series Total Energy Pack (CASTEP) code[38] employing the
ultrasoft pseudopotential, and the exchange and correlation effects were
described by PBEsol in a generalized gradient approximation (GGA).[39]
The Monkhorst–Pack scheme K-points grid sampling was set as 3 × 3 × 1
or 5 × 2 × 1 for the irreducible Brillouin zone. The fast Fourier transform
grid was set as 36 × 30 × 288 or 20 × 45 × 540. The Kohn–Sham wave
function was extended using an energy cutoff value of 380 eV. The
convergence criterion was as follows: maximum force on the atom is
0.01 eV Å−1, maximum stress on the atom is 0.02 GPa, maximum atomic
displacement is 0.0004 Å, and the maximum energy change per atom
is 5 × 10−6 eV. In order to improve the computational accuracy, dipole
calibration was applied to all models. Finally, the electronic structure was
calculated based on the optimized models. The heterostructure models
possess 8 layers of BiVO4 (96 atoms) and 12 layers of Au (96 atoms).
Between the BiVO4/Au heterostructure slabs, a 20 Å thickness vacuum
layer is added to eliminate the mirror-image interaction of 3D periodic
boundary conditions. Based on the optimized models, the electron
density could be estimated. On the other hand, the energy eigenvalue
also could be obtained in the self-consistent field calculation. According
to the definition, the Fermi energy is further calculated in the present
work.
Keywords
charge separation, CO2 reduction, photocatalysis, Schottky junctions,
Z-scheme photocatalysts
Received: February 14, 2018
Revised: May 15, 2018
Published online:
[1] M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi,
E. A. Santori, N. S. Lewis, Chem. Rev. 2010, 110, 6446.
[2] J. L. White, M. F. Baruch, J. E. PanderIII, Y. Hu, I. C. Fortmeyer,
J. E. Park, T. Zhang, K. Liao, J. Gu, Y. Yan, T. W. Shaw, E. Abelev,
A. B. Bocarsly, Chem. Rev. 2015, 115, 12888.
[3] P. Zhou, J. G. Yu, M. Jaroniec, Adv. Mater. 2014, 26, 4920.
[4] H. J. Li, Y. Zhou, W. G. Tu, J. H. Ye, Z. G. Zou, Adv. Funct. Mater.
2015, 25, 998.
[5] H. J. Li, Y. Y. Gao, Y. Zhou, F. T. Fan, Q. T. Han, Q. F. Xu, X. Y. Wang,
M. Xiao, C. Li, Z. G. Zou, Nano Lett. 2016, 16, 5547.
[6] A. Iwase, Y. H. Ng, Y. Ishiguro, A. Kudo, R. Amal, J. Am. Chem. Soc.
2011, 133, 11054.
[7] H. Tada, T. Mitsui, T. Kiyonaga, T. Akita, K. Tanaka, Nat. Mater.
2006, 5, 782.
[8] Q. Wang, T. Hisatomi, Q. X. Jia, H. Tokudome, M. Zhong,
C. Z. Wang, Z. H. Pan, T. Takata, M. Nakabayashi, N. Shibata,
Y. B. Li, I. D. Sharp, A. Kudo, T. Yamada, K. Domen, Nat. Mater.
2016, 15, 611.
[9] Q. Wang, T. Hisatomi, Y. Suzuki, Z. H. Pan, J. Seo, M. Katayama,
T. Minegishi, H. Nishiyama, T. Takata, K. Seki, A. Kudo, T. Yamada,
K. Domen, J. Am. Chem. Soc. 2017, 139, 1675.
[10] S. J. Moniz, S. A. Shevlin, D. J. Martin, Z. X. Guo, J. W. Tang, Energy
Environ. Sci. 2015, 8, 731.
[11] M. R. Khan, T. W. Chuan, A. Yousuf, M. N. Chowdhury, C. K. Cheng,
Catal. Sci. Technol. 2015, 5, 2522.
[12] L. L. Wang, J. Ge, A. L. Wang, M. S. Deng, X. J. Wang, S. Bai, R. Li,
J. Jiang, Q. Zhang, Y. Luo, Y. J. Xiong, Angew. Chem., Int. Ed. 2014,
53, 5107.
[13] M. G. Kibria, S. Zhao, F. A. Chowdhury, Q. Wang, H. P. T. Nguyen,
M. L. Trudeau, H. Gu, Z. Mi, Nat. Commun. 2014, 5, 3825.
[14] R. G. Li, F. X. Zhang, D. Wang, J. X. Yang, M. R. Li, J. Zhu, X. Zhou,
H. X. Han, C. Li, Nat. Commun. 2013, 4, 1432.
[15] J. Zhu, F. T. Fan, R. T. Chen, H. Y. An, Z. C. Feng, C. Li, Angew.
Chem., Int. Ed. 2015, 54, 9111.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
[16] L. C. Mu, Y. Zhao, A. L. Li, S. Y. Wang, Z. L. Wang, J. X. Yang,
Y. Wang, T. F. Liu, R. T. Chen, J. Zhu, F. T. Fan, R. G. Li, C. Li, Energy
Environ. Sci. 2016, 9, 2463.
[17] P. Li, Y. Zhou, Z. Y. Zhao, Q. F. Xu, X. Y. Wang, M. Xiao, Z. G. Zou,
J. Am. Chem. Soc. 2015, 137, 9547.
[18] J. Yu, J. Low, W. Xiao, P. Zhou, M. Jaroniec, J. Am. Chem. Soc. 2014,
136, 8839.
[19] X. Liu, G. Dong, S. Li, G. Lu, Y. Bi, J. Am. Chem. Soc. 2016, 138,
2917.
Acknowledgements
This work was supported primarily by the National Basic Research Program
of China (2013CB632404), the National Natural Science Foundation of China
(51572121, 21603098, and 21633004), the Natural Science Foundation
of Jiangsu Province (BK20151265, BK20151383, and BK20150580), the
Postdoctoral Science Foundation of China (2017M611784), and the
©
Adv. Funct. Mater. 2018, 1801214
1801214 (9 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
[20] A. G. Scheuermann, J. P. Lawrence, K. W. Kemp, T. Ito, A. Walsh,
C. E. D. Chidsey, P. K. Hurley, P. C. McIntyre. Nat. Mater. 2016, 15,
99
[21] H. L. Tan, X. M. Wen, R. Amal, Y. H. Ng, J. Phys. Chem. Lett. 2016,
7, 1400.
[29] K. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, Electrochem.
Commun. 2000, 2, 207.
[30] S. H. Yoon, J. Lee, Environ. Sci. Technol. 2005, 39, 9695.
[31] G. G. Liu, X. Z. Li, J. C. Zhao, S. Horikoshi, H. Hidaka, J. Mol. Catal.
A 2000, 153, 221.
[22] C. H. Kuo, T. E. Hua, M. H. Huang, J. Am. Chem. Soc. 2009, 131,
17871.
[32] A. J. Bard, R. Parsons, J. Jordan, Standard Potentials in Aqueous
Solution, Marcel Dekker, Inc., New York 1985.
[23] J. W. Hong, D. H. Wi, S. U. Lee, S. W. Han, J. Am. Chem. Soc. 2016,
138, 15766.
[24] H. Q. Li, W. S. Hong, Y. M. Cui, X. Y. Hu, S. H. Fan, L. J. Zhu, Mater.
Sci. Eng., B 2014, 181, 1.
[25] S. Y. Bao, Q. F. Wu, S. Z. Chang, B. Z. Tian, J. L. Zhang, Catal. Sci.
Technol. 2017, 7, 124.
[26] Q. B. Zhang, K. L. Zhang, D. G. Xu, G. C. Yang, H. Huang, F. Nie,
C. M. Liu, S. H. Yang, Prog. Mater. Sci. 2014, 60, 208.
[27] J. C. Wang, L. Zhang, W. X. Fang, J. Ren, Y. Y. Li, H. C. Yao,
J. S. Wang, Z. J. Li, ACS Appl. Mater. Interfaces 2015, 7,
8631.
[33] T. Tachikawa, T. Ochi, Y. Kobori, ACS Catal. 2016, 6, 2250.
[34] A. Tanaka, K. Hashimoto, H. Kominami, J. Am. Chem. Soc. 2014,
136, 586.
[35] G. B. Murdoch, M. Greiner, M. G. Helander, Z. B. Wang, Z. H. Lu,
Appl. Phys. Lett. 2008, 93, 083309.
[36] F. Du, Q. Y. Chen, Y. H. Wang, J. Phys. Chem. Solids. 2017, 104, 139.
[37] A. Iwase, H. Kato, A. Kudo, J. Sol. Energy Eng. 2010, 132, 021106.
[38] S. J. Clark, M. D. Segall, C. J. Pickard, P. J. Hasnip, M. I. J. Probert,
K. Refson, M. C. Payne, Z. Kristallogr. 2005, 220, 567.
[39] J. P. Perdew, A. Ruzsinszky, G. I. Csonka, O. A. Vydrov,
G. E. Scuseria, L. A. Constantin, X. L. Zhou, K. Burke, Phys. Rev.
Lett. 2008, 100, 136406.
[28] S. F. Chen, Y. G. Yang, W. Liu, J. Hazard. Mater. 2011, 186, 1560.
©
Adv. Funct. Mater. 2018, 1801214
1801214 (10 of 10)
2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim