In situ Irradiated XPS Investigation on S-Scheme TiO2@ZnIn2S4 Photocatalyst for Efficient Photocatalytic CO2 Reduction

Article abstract of DOI:10.1002/smll.202103447

Reasonable design of efficient hierarchical photocatalysts has gained significant attention. Herein, a step-scheme (S-scheme) core-shell TiO₂@ZnIn₂S₄ heterojunction is designed for photocatalytic CO₂ reduction. The optimized sample exhibits much higher CO₂ photoreduction conversion rates (the sum yield of CO, CH₃OH, and CH₄) than the blank control, i.e., ZnIn₂S₄ and TiO₂. The improved photocatalytic performance can be attributed to the inhibited recombination of photogenerated charge carriers induced by S-scheme heterojunction. The improvement is also attributed to the large specific surface areas and abundant active sites. Meanwhile, S-scheme photogenerated charge transfer mechanism is testified by in situ irradiated X-ray photoelectron spectroscopy, work function calculation, and electron paramagnetic resonance measurements. This work provides an effective strategy for designing highly efficient heterojunction photocatalysts for conversion of solar fuels.

Full text of DOI:10.1002/smll.202103447

ReseaRch aRticle
www.small-journal.com
In situ Irradiated XPS Investigation on S-Scheme TiO2@ZnIn2S4
Photocatalyst for Efficient Photocatalytic CO2 Reduction
Libo Wang, Bei Cheng, Liuyang Zhang,* and Jiaguo Yu*
Recently, a novel step-scheme (S-scheme)
Reasonable design of efficient hierarchical photocatalysts has gained
significant attention. Herein, a step-scheme (S-scheme) core-shell
TiO2@ZnIn2S4 heterojunction is designed for photocatalytic CO2 reduction.
The optimized sample exhibits much higher CO2 photoreduction conversion
rates (the sum yield of CO, CH3OH, and CH4) than the blank control, i.e.,
ZnIn2S4 and TiO2. The improved photocatalytic performance can be attributed
to the inhibited recombination of photogenerated charge carriers induced
by S-scheme heterojunction. The improvement is also attributed to the large
specific surface areas and abundant active sites. Meanwhile, S-scheme
photogenerated charge transfer mechanism is testified by in situ irradiated
X-ray photoelectron spectroscopy, work function calculation, and electron
paramagnetic resonance measurements. This work provides an effective
strategy for designing highly efficient heterojunction photocatalysts for
conversion of solar fuels.
heterojunction was proposed and has
attracted researchers’ attention.[23–36] Usu-
ally, S-scheme heterojunction consists of
reduction photocatalyst (RP) and oxidation
photocatalyst (OP). Besides, the directional
migration of free electrons will lead to band
bending and internal electric field (IEF) at
their interface owing to the work function
difference. Notably, under the influence of
IEF, the photogenerated electrons of OP
with weak reduction ability can recombine
with the photogenerated holes of RP with
weak oxidation ability; while those with
strong redox abilities are preserved. There-
fore, reasonable construction of TiO2-based
S-scheme heterojunction is of great sig-
nificance to improve photocatalytic reaction
performance.
Apart from the charge carrier separation, the morphology
of photocatalysts is another important factor to influence the
photocatalytic performance.[15,17,19,37,38] Photocatalysts with
hollow structures have attracted great attention owing to mani-
fold advantages including larger specific surface area, abundant
active sites, shortened diffusion distance as well as improved
light reflection and scattering.[37,39–44] Therefore, the design of
hollow S-scheme heterojunction photocatalyst is of vital impor-
tance to enhance photocatalytic performance.
ZnIn2S4, as a typical reduction photocatalyst, stands out for
its layered structure, narrow bandgap, suitable redox potentials,
and good chemical stability. And it has been used for various
photocatalytic applications including hydrogen production, CO2
reduction, and organic degradation.[45–51] Unfortunately, pris-
tine ZnIn2S4 photocatalyst shows low photocatalytic efficiency
owing to the fast recombination of photogenerated charge
carriers.[45–49] Considering the suitable match of band gap of
ZnIn2S4 and TiO2 for S-scheme heterojunction,[27,52] we con-
struct the hollow TiO2@ZnIn2S4 core-shell structure. Up to
now, to the best of our knowledge, it has never been reported.
Herein, we grow ZnIn2S4 nanosheets on the outer surface of
TiO2 hollow spheres by in situ chemical bath deposition reaction.
This rational design is not only able to provide large specific surface
areas and abundant reaction sites for PCR reaction, but also can
effectively suppress the recombination of useful photogenerated
electrons and holes. As a result, the optimized TiO2@ZnIn2S4
heterojunction exhibits high PCR performance, and the total CO2
photoreduction conversion rates (the sum yield of CO, CH3OH and
CH4) are obviously higher than those of blank ZnIn2S4, TiO2, and
ex situ prepared TiO2-ZnIn2S4 composite. Finally, S-scheme mecha-
nism is also thoroughly analyzed and discussed in this work.
1. Introduction
Photocatalytic CO2 reduction (PCR) into chemical fuels under
mild conditions is recognized to be promising for alleviating
energy crisis and greenhouse effect.[1–11] Therefore, reasonable
design of efficient photocatalysts is necessary to improve PCR
performance. Among various semiconductor photocatalysts,
TiO2 has received great attention due to its non-toxicity, cost-
effectiveness, and excellent chemical stability.[12–16] However,
as a typical oxidation photocatalyst, unitary TiO2 suffers from
low activity owing to its poor photoreduction ability and fast
recombination of photogenerated electrons and holes.[12–15]
To overcome these drawbacks, various solutions have been
employed.[17–20] Among them, heterojunction construction is
proved to be an effective strategy due to the efficient transfer
and separation of photogenerated charge carriers.[21–23]
L. Wang, B. Cheng, L. Zhang, J. Yu
State Key Laboratory of Advanced Technology for Materials Synthesis
and Processing
Wuhan University of Technology
Wuhan 430070, P. R. China
E-mail: zly2017@whut.edu.cn; yujiaguo93@whut.edu.cn
J. Yu
Laboratory of Solar Fuel
Faculty of Materials Science and Chemistry
China University of Geosciences
388 Lumo Road, Wuhan 430074, P. R. China
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/smll.202103447.
DOI: 10.1002/smll.202103447
2103447 (1 of 9)
© 2021 Wiley-VCH GmbH
Small 2021, 2103447

www.advancedsciencenews.com
www.small-journal.com
solid spheres are etched with HF solution to remove the SiO2
templates, and then TiO2 hollow spheres are obtained. Field-
emission scanning electron microscopy (FESEM) images
clearly show that the sphere structures are intact after HF
solution treatment (Figure 1b, Figure S1c, Supporting Infor-
mation). Afterwards, hierarchical TiO2@ZnIn2S4 CSHS are
successfully acquired by in situ chemical bath deposition.
During the reaction processes, due to the positively charged
surface of TiO2 (pH = 2.5) (Figure S2, Supporting Informa-
tion), the negative-charged S2− ions released from the decom-
position of thioacetamide can be adsorbed on the outer surface
2. Results and Discussion
The preparation processes of hierarchical TiO2@ZnIn2S4 core-
shell hollow spheres (CSHS) are illustrated in Figure 1a. Firstly,
SiO2 spheres with average diameters of 350 nm as the hard-tem-
plates are prepared by a modified Stöber method (Figure S1a,
Supporting Information).[53] Afterwards, TiO2 nanoparticles
are coated on the surface of SiO2 spheres to form TiO2@SiO2
solid spheres with average diameters of 500 nm via versatile
kinetics-controlled coating and thermal annealing (Figure S1b,
Supporting Information).[54] Subsequently, these TiO2@SiO2
Figure 1. a) Illustration of the preparation processes of TiO2@ZnIn2S4 core-shell hollow spheres (CSHS). Field-emission scanning electron micros-
copy (FESEM) images of b) TiO2 hollow spheres and c) TiO2@ZnIn2S4 CSHS. d) Transmission electron microscopy (TEM) and its corresponding
e) high-resolution TEM (HRTEM) image of TiO2@ZnIn2S4 CSHS. f) High-angle annular dark-field (HAADF) image of TiO2@ZnIn2S4 CSHS. g–k) EDS
mapping images of Ti, O, Zn, In, and S elements.
Small 2021, 2103447
2103447 (2 of 9)
© 2021 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
of TiO2 hollow spheres through electrostatic interactions. Fur-
ther reactions between positive metal ions (Zn2+ and In3+) and
negative S2− ions attached on TiO2 promote the formation of
the ZnIn2S4 nanosheets on the outer surface of TiO2 hollow
spheres with tight contact (Figure 1c). Transmission electron
microscopy (TEM) image clearly validates the formation of
hierarchical TiO2@ZnIn2S4 CSHS (Figure 1d). Meanwhile,
its corresponding high-resolution TEM (HRTEM) image indi-
cates that the spacing of distinct lattice fringes with are 0.33
and 0.35 nm, which can be respectively indexed into the (101)
planes of ZnIn2S4 and TiO2 (Figure 1e). Furthermore, high-
angle annular dark-field (HAADF) image (Figure 1f) and cor-
responding elemental mapping images (Figure 1g–k) clearly
demonstrate the distribution of Ti, O, Zn, In, and S elements in
TiO2@ZnIn2S4 composite. Most notably, the innermost region
of TiO2 hollow sphere core and the outer ZnIn2S4 nanosheet
shell can be distinguished. These results directly indicate that
hierarchical TiO2@ZnIn2S4 CSHS are successfully constructed.
By contrast, in the absence of TiO2 hollow sphere core, random
aggregation of blank ZnIn2S4 is observed (Figure S1d, Sup-
porting Information).
The X-ray powder diffraction (XRD) patterns were measured
to identify the phase structures of samples. As presented in
Figure 2a, the peaks of TiO2 and ZnIn2S4 correspond to the typical
diffraction peaks of anatase TiO2 (JCPDS No. 84–1286) and hexag-
onal ZnIn2S4 (JCPDS No. 72–0773),[45,47] respectively. Furthermore,
the presence of TiO2 and ZnIn2S4 in the TiO2@ZnIn2S4 can be
observed. Besides, compared with the pristine TiO2 and ZnIn2S4,
the positions of diffraction peaks in TiO2@ZnIn2S4 composite are
almost unchanged, indicating that the growth of ZnIn2S4 does not
alter their respective crystal structures.
To identify the surface element composition and electronic
chemical states of TiO2, ZnIn2S4, and TiO2@ZnIn2S4, X-ray
photoelectron spectroscopy (XPS) was employed. As presented
in Figure S3 (Supporting Information), XPS survey spectra
indicate that Ti, O, Zn, In, and S elements can be detected in
the TiO2@ZnIn2S4 composite, which are in according with the
EDS results. Besides, as for Ti, the peaks centered at 458.5 and
464.2 eV can be assigned to the 2p3/2 and 2p1/2 of Ti, respectively
(Figure 2b).[24,55] The peaks centered at 529.4 and 531.4 eV can
be ascribed to the Ti-O bond and surface hydroxyl (−OH) group,
respectively (Figure 2c).[23,55] While for Zn, the peaks located at
1022.0 and 1045.1 eV are attributed to its 2p3/2 and 2p1/2 orbitals,
respectively (Figure 2d).[45–47] The binding energy difference
is about 23.1 eV, which is typical of Zn2+ in ZnIn2S4.[45–47] In
Figure 2e, the peaks located at 444.9 and 452.5 eV can be attrib-
uted to the 3d5/2 and 3d3/2 of In, respectively, suggesting that
its valence state is +3.[45–47] In Figure 2f, the peaks centered at
161.6 and 162.8 eV, assigning to the 2p3/2 and 2p1/2 of S, respec-
tively, is the characteristic of S2− and in agreement with the lit-
erature.[45–47] Notably, the electron binding energy of element
originates from the strong Coulomb attraction between the
outer electrons of atom and atomic nucleus.[20,56,57] Meanwhile,
the changes of element binding energy can directly reflect the
changes of electron density.[20,56,57] This means that the decrease
of electron density can lead to the increase of binding energy.
Contrarily, the decrease of binding energy is owing to the fact
that the atom obtains electrons.[20,56,57] Therefore, the binding
energy change can be used to examine the direction of charge
carrier transfer in heterojunction photocatalysts.[23,27,55,58] As
shown in Figure 2b,c, compared with TiO2, the binding ener-
gies of O 1s and Ti 2p in TiO2@ZnIn2S4 composite obviously
Figure 2. a) X-ray powder diffraction (XRD) patterns of TiO2, ZnIn2S4, and TiO2@ZnIn2S4. High-resolution X-ray photoelectron spectroscopy (XPS)
spectra of b) Ti 2p and c) O 1s of TiO2 and TiO2@ZnIn2S4. High-resolution XPS spectra of d) Zn 2p, e) In 3d, and f) S 2p of ZnIn2S4 and TiO2@ZnIn2S4.
2103447 (3 of 9)
© 2021 Wiley-VCH GmbH
Small 2021, 2103447

www.advancedsciencenews.com
www.small-journal.com
shift toward lower energy levels in darkness, indicating an
increase in the electron density of TiO2.[23,55] In contrast, com-
pared with ZnIn2S4, the binding energies of Zn 2p, In 3d, and S
2p in TiO2@ZnIn2S4 significantly shift to higher energy levels
in darkness (Figure 2d–f), implying that the electrons migrate
from ZnIn2S4 to TiO2. Furthermore, when in situ irradiated
XPS measurements are performed,[20,23,27,55–58] compared with
TiO2@ZnIn2S4 in darkness, the binding energies of O 1s and
Ti 2p in the TiO2@ZnIn2S4 composite under light irradiation
markedly move to higher energy levels (Figure 2b,c); while the
binding energies of Zn 2p, In 3d and S 2p in the nanocomposite
obviously shift to lower energy levels (Figure 2d–f), suggesting
the photogenerated electrons transfer from TiO2 to ZnIn2S4
under light irradiation.[27,58] Therefore, these XPS results pro-
vide important evidences of the charge carrier transfer pathway
across the TiO2@ZnIn2S4 heterojunction interface under light
irradiation. In detail, the migration of photogenerated elec-
trons from TiO2 to ZnIn2S4, matching well with the S-scheme
mechanism.
Most importantly, the difference in work functions (W)
or Fermi levels (Ef) for semiconductor photocatalysts is the
internal driving force of charge transfer in heterojunction
nanocomposite. W of TiO2, ZnIn2S4, and TiO2@ZnIn2S4 were
obtained by Kevin probe measurements.[58,59] As a result, W
of TiO2, ZnIn2S4, and TiO2@ZnIn2S4 are determined to be
4.69, 3.92 and 4.31 eV, respectively (Figure 3a, Figure S4 and
Table S1, Supporting Information). Ef of samples is calculated
on the basis of the following equation (1):
pointing from ZnIn2S4 to TiO2 (Figure 3b).[23,24,27,55,61] When
TiO2@ZnIn2S4 heterojunction is exposed to light, electrons
are excited from VB to CB of ZnIn2S4 and TiO2, respectively.
Under the combined effect of IEF, interface band bending
and Coulomb interaction, the photogenerated electrons in CB
of TiO2 easily transfer to VB of ZnIn2S4 and recombine with
the photogenerated holes of ZnIn2S4 (Figure 3c).[23,24,27,55,61]
This interfacial charge transfer behavior of TiO2@ZnIn2S4 het-
erojunction accords with the XPS results and is in agreement
with the S-scheme mechanism. Meanwhile, the photogenerated
holes on VB of TiO2 and the photogenerated electrons on CB of
ZnIn2S4 are maintained, which participate in the photocatalytic
redox reaction, respectively (Figure 3c). Besides, due to the inti-
mate interface contact between ZnIn2S4 and TiO2, the intensity
of IEF at interface is greatly enhanced, which is favorable for
the interfacial photogenerated charge transfer as well as effec-
tively improve the PCR performance.
Meanwhile, the band structures of samples were measured
by Mott–Schottky (M–S) and UV-vis diffuse reflectance spectra
(UV-vis DRS spectra) measurements. M-S plots of TiO2 and
ZnIn2S4 exhibit positive slopes, suggesting that both samples
are n-type semiconductors (Figure 4a).[55,58,62,63] Moreover, the
flat-band potentials (Efb) of TiO2 and ZnIn2S4 are −0.22 and
−0.86 V (versus normal hydrogen electrode (NHE), at pH = 0.
Note: The potentials in this paper are relative to that of NHE
unless otherwise mentioned.), respectively. Usually, the Efb of
n-type semiconductor roughly equals to the conduction band
(CB) potential.[55,58,62,64] Besides, the band gap (Eg) of TiO2
(3.20 eV) and ZnIn2S4 (2.43 eV) are obtained by UV-vis DRS
spectra and Tauc plots (Figure 4b,c), respectively. Therefore,
considering the CB positions and Eg of TiO2 and ZnIn2S4, the
valence band (VB) potentials of TiO2 and ZnIn2S4 are 2.98 and
1.57 V, respectively. Meanwhile, according to the above results
and analysis, the band structures of TiO2 and ZnIn2S4 are pre-
sented in Figure 4d. Noticeably, the reduction potentials of CO,
CH3OH and CH4 are -0.12, 0.03 and 0.17 V, respectively,[27,58,65]
implying that TiO2 and ZnIn2S4 are thermodynamically feasible
for converting CO2 into CO, CH3OH and CH4 (Figure 4d).
Additionally, to further validate the S-scheme mechanism of
TiO2@ZnIn2S4 heterojunction, the electron paramagnetic res-
Ef =Vac −W
(1)
where Vac is the vacuum level (assumed as 0 eV).[60] Thus, Ef of
TiO2, ZnIn2S4 and TiO2@ZnIn2S4 are determined to be −4.69,
−3.92 and −4.31 eV, respectively. Obviously, TiO2 has a larger
W (4.61 eV) and lower Ef (−4.61 eV), ZnIn2S4 has a smaller W
(3.92 eV) and higher Ef (−3.92 eV) (Figure 3a). Moreover, when
TiO2@ZnIn2S4 heterojunction is formed, the electrons will
transfer from ZnIn2S4 to TiO2 easily until the Ef at the interface
tends to equilibrate (Figure 3b). Concurrently, ZnIn2S4 shows
upward interface band bending and is positively charged at the
interface due to loss of electrons; while TiO2 exhibits down-
ward interface band bending and is negatively charged at the
interface due to accumulation of electrons (Figure 3b). In con-
sequence, the electrons transfer creates an IEF at the interfaces
−
onance (EPR) spectra of hydroxyl (·OH) and superoxide (·O2 )
radicals via 5, 5-dimethyl-1-pyrroline N-oxide (DMPO) were
measured.[24,27,32] The redox potentials of H2O/·OH (2.37 V) or
−
OH−/·OH (2.69 V) and O2/·O2 couples (−0.33 V) are shown
Figure 3. Schematic illustration of the S-scheme transfer mechanism between TiO2 and ZnIn2S4. Vac: Vacuum level; W: Work function; Ef: Fermi level;
CB: conduction band; VB: valence band; IEF: internal electric field. a) Before contact, b) after contact, and c) after contact in light.
Small 2021, 2103447
2103447 (4 of 9)
© 2021 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Figure 4. a) M–S plots, b) UV-vis diffuse reflectance spectra (UV-vis DRS), c) Tauc plots, and d) band structures of samples. Electron paramagnetic
resonance (EPR) spectra of e) DMPO-·OH adducts (in aqueous solution) and f) DMPO-·O2− adducts (in methanol solution) over TiO2, ZnIn2S4, and
TiO2@ZnIn2S4 under illumination for 2 min. Schematic diagrams of g) conventional type-II and h) S-scheme heterojunction.
in Figure 4d.[24,27,32] Meanwhile, EPR spectra of DMPO-·OH
and DMPO- O2 adducts of TiO2, ZnIn2S4 and TiO2@ZnIn2S4
CB of TiO2 (-0.22 V) is positive than the redox potential of
−
−
O2/·O2 couples (-0.33 V) (Figure 4d).[24,27,32] But for ZnIn2S4,
are detected under irradiation, respectively (Figure 4e,f). It
can be seen that the DMPO-·OH signal of TiO2@ZnIn2S4 is
obviously stronger than that of pristine TiO2; while the DMPO-
·OH signal of pristine ZnIn2S4 is hardly observed (Figure 4e).
the trend is reversed (Figure 4d). Notably, if the charge transfer
mechanism of TiO2@ZnIn2S4 heterojunction conforms to
traditional type-II heterojunction, the generation of ·OH and
−
·O2 radicals are not facilitated (Figure 4g). However, it is
−
−
Moreover, the DMPO-·O2 intensity of TiO2@ZnIn2S4 is evi-
notable that for the DMPO-·OH or DMPO-·O2 adducts, the
dently higher than that of pristine ZnIn2S4; while the DMPO-
·O2 intensity of pristine TiO2 is very weak (Figure 4f). The
main reason is that the oxidation potential of photogenerated
holes on VB of TiO2 (2.98 V) is more positive than the redox
potential of H2O/·OH (2.37 V) or OH−/·OH (2.69 V) couples;
while the reduction potential of photogenerated electrons on
signal intensity of TiO2@ZnIn2S4 are obviously stronger than
those of pristine TiO2 and ZnIn2S4 (Figure 4e,f). These results
indicate that the photogenerated charge transfer mechanism
of TiO2@ZnIn2S4 heterojunction is in agreement with the
S-scheme mechanism rather than traditional type-II hetero-
junction (Figure 4h).[24,27,32]
−
2103447 (5 of 9)
© 2021 Wiley-VCH GmbH
Small 2021, 2103447

www.advancedsciencenews.com
www.small-journal.com
Figure 5. a) Yield of products in photocatalytic CO2 reduction (PCR) reaction of different samples. b) Stability tests of TiO2@ZnIn2S4. c) X-ray powder
diffraction (XRD) patterns before and after the stability tests of TiO2@ZnIn2S4.
PCR reaction activities of different samples were measured sources by gas chromatography-mass spectrometry under the
under full-spectrum irradiation. As shown in Figure 5a, CO, same conditions (Figure S6, Supporting Information). These
CH3OH and CH4 are the major photoreduction products of results indicate that the detected photoreduction products
samples. Obviously, total CO2 photoreduction conversion rates originate from CO2 instead of impurities.
(the sum yield of CO, CH3OH and CH4) of TiO2 hollow spheres
Additionally, it is widely known that larger specific surface
(4.13 µmol h−1 g−1) are higher than those of bulk TiO2 non- area of photocatalysts is beneficial to PCR reaction owing
hollow structure sample (Figure S5, Supporting Information) to the fact that it can provide the abundant active sites.[19,37]
(2.04 µmol h−1 g−1), showing the advantage of hollow struc- Therefore, the surface area was evaluated via nitrogen adsorp-
ture in the PCR reaction. Besides, total CO2 photoreduction tion-desorption isotherms. As shown in Figure 6a, all the
conversion rates of 6.65 µmol h−1 g−1 are achieved by pristine samples display typical type IV isotherms with the type H3
ZnIn2S4 nanosheets. Gratifyingly, the construction of hetero- hysteresis loop, signifying their mesoporous structures.[45,47,54]
junction obviously elevates the PCR activities. In particular, Meanwhile, the pore-size distribution curves further testify
the total CO2 photoreduction conversion rates have increased their mesoporous structures (Figure 6a (inset)). Moreover,
by 2.75 and 4.43 times compared with the pristine ZnIn2S4 and Brunauer–Emmett–Teller surface area (SBET
) of TiO2@
TiO2, respectively, attaining 18.32 µmol h−1 g−1. Moreover, this ZnIn2S4 (107 m2 g−1) is larger than those of TiO2 (78 m2 g−1)
PCR reaction performance is higher than that of many other and ZnIn2S4 (102 m2 g−1) (Table S4, Supporting Informa-
photocatalysts under similar conditions (Table S2, Supporting tion). Besides, large specific surface area of the composite can
Information). Interestingly, undetected hydrogen is owing supply abundant surface active and adsorption sites, which
to the fact that the gas (CO2+H2O)-solid (photocatalyst) reac- is favorable for the capture and activation of CO2. Mean-
tion system is more favorable for adsorption and activation of while, as shown in Figure 6b, the CO2 adsorption capacity
CO2, thus the hydrogen evolution reaction is inhibited.[66,67] of TiO2@ZnIn2S4 (0.12 mmol g−1) is higher than those of
Besides, the amount of photooxidation product O2 was also TiO2 (0.09 mmol g−1) and ZnIn2S4 (0.10 mmol g−1). Besides,
close to its theoretical value (Table S3, Supporting Informa- the adsorption energy (Ead) of CO2 on ZnIn2S4 (−0.09 eV) is
tion). Meanwhile, the oxygen evolution rate of TiO2@ZnIn2S4 more negative than that on TiO2 (−0.06 eV) (Figure S7, Sup-
is obviously higher than those of pristine ZnIn2S4 nanosheets porting Information), which indicates that the adsorbed CO2
and TiO2 hollow spheres (Figure 5a). Based on the above anal- on ZnIn2S4 is more stable than that on TiO2. Together, all of
ysis, the improvement of PCR performance can be ascribed to these results confirm the superiority of TiO2@ZnIn2S4 over
construction of S-scheme heterojunction. The TiO2@ZnIn2S4 pristine TiO2 and ZnIn2S4 in PCR.
S-scheme heterojunction can efficiently facilitate the separa-
To further clarify the excellent CO2 photoreduction activity
tion and transfer of photogenerated carriers. Meanwhile, a of TiO2@ZnIn2S4, photoluminescence (PL) and photoelectro-
large number of useful photogenerated electrons with strong chemical characterizations were performed. As displayed in
reduction ability can be retained in the CB of ZnIn2S4 to the Figure 6c, the steady-state PL intensity of TiO2@ZnIn2S4 is
CO2 photoreduction reaction, while the photogenerated holes obviously lower than those of pristine TiO2 and ZnIn2S4, indi-
with strong oxidation ability can be retained in the VB of TiO2 cating that construction of S-scheme heterojunction photocat-
to the H2O photooxidation reaction. Besides, the PCR per- alyst can largely inhibit the recombination of photogenerated
formance of TiO2@ZnIn2S4 nearly retains after running for charge carriers.[55,58] Moreover, as shown in Figure 6d, the
three cycles (Figure 5b). XRD of TiO2@ZnIn2S4 shows no average PL lifetime (τave) of TiO2@ZnIn2S4 (2.43 ns) is obvi-
noticeable change before and after PCR reaction (Figure 5c), ously shorter than those of pristine TiO2 (3.35 ns) and ZnIn2S4
suggesting the good stability of the TiO2@ZnIn2S4 hetero- (4.81 ns), confirming a more efficient separation and transfer
junction. Moreover, to investigate the carbon origin of pho- of photogenerated carriers in the S-scheme heterojunc-
toreduction products, 13CH4 (m/z = 17) and 13CO (m/z = 29) tion.[23,45,47,55,58] Additionally, TiO2@ZnIn2S4 heterojunction
characteristic peaks are detected in 13C-labeled CO2, and 12CH4 shows the smaller electrochemical impedance spectroscopy
(m/z = 16) and 12CO (m/z = 28) are detected in 12C-labeled CO2 (EIS) radius and higher photocurrent density than those of
Small 2021, 2103447
2103447 (6 of 9)
© 2021 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Figure 6. a) Nitrogen adsorption-desorption isotherms and pore-size distribution curves (inset), b) CO2 adsorption isotherms, c) steady-state pho-
toluminescence (PL) spectra, d) time-resolved PL spectra, e) electrochemical impedance spectroscopy (EIS) spectra, and (f) transient photocurrent
spectra of TiO2, ZnIn2S4 and TiO2@ZnIn2S4.
pristine TiO2 and ZnIn2S4 (Figure 6e,f), reaffirming the highly
efficient photogenerated charge carriers separation and transfer
over TiO2@ZnIn2S4 heterojunction.[45,47,58,62,64] For comparison,
control experiments indicate that the PCR performance of a
physical mixture of TiO2 and ZnIn2S4 (TiO2-ZnIn2S4) is much
inferior to that of TiO2@ZnIn2S4 under the same condition
(Figure S8a, Supporting Information). Besides, the EIS radius
of TiO2-ZnIn2S4 is much larger than that of TiO2@ZnIn2S4
(Figure S8b, Supporting Information). Therefore, these results
imply that the crucial roles of intimate interfacial contact and
hierarchical core-shell hollow structures in TiO2@ZnIn2S4 het-
erojunction for PCR reaction.
In situ diffuse reflectance infrared Fourier transform spec-
troscopy (in situ DRIFTS) can be used to directly detect the
carbon-based groups formed on the surface of photocatalyst
during the PCR reaction, which is helpful to analyze the forma-
tion mechanism of reaction intermediates.[68] Therefore, to get
further insight into the evolution process of PCR over the sur-
face of TiO2@ZnIn2S4, in situ DRIFTS was analyzed. As shown
in Figure 7a, without introducing CO2 and H2O gas in the dark,
no absorption peaks can be detected (Figure 7a(i)). When intro-
ducing CO2 and H2O gas without light irradiation (Figure 7a(ii)),
−
bicarbonate species (HCO3 , 1223, 1239, 1395 and 1449 cm−1),
monodentate carbonate species (m-CO32−, 1303, 1422, 1488, 1541
Figure 7. a) In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) spectra of TiO2@ZnIn2S4. b) Possible photocatalytic CO2
reduction (PCR) reaction pathways on the surface of TiO2@ZnIn2S4. The blue dashes boxes highlight the key intermediate species during photocatalytic
CO2 reduction (PCR) reaction.
2103447 (7 of 9)
© 2021 Wiley-VCH GmbH
Small 2021, 2103447

www.advancedsciencenews.com
www.small-journal.com
and 1557 cm−1) and bidentate carbonate species (b-CO32−, 1317,
1519 and 1653 cm−1) can be detected.[39,55,58] These peaks mean
that the CO2 and H2O are effectively adsorbed and activated on
the surface of TiO2@ZnIn2S4. Furthermore, some new peaks
can be observed under irradiation condition (Figure 7a(iii)). The
emerging peaks can be attributed to carboxylate species (COO−,
1342 and 1363 cm−1), formaldehyde species (HCHO−, 1507 and
1773 cm−1), formic acid species (HCOO−, 1573 cm−1) and methoxy
groups (CH3O−, 1682, 1702 and 1745 cm−1).[39,55,58] Meanwhile,
these peaks of intermediates reveal that the effective photo-
catalytic activation and reaction of CO2 occur on the surface of
TiO2@ZnIn2S4. Based on the previous reports,[2,3,39,55,58] the
COO−, HCHO−, HCOO− and CH3O− species are assigned to the
important intermediates of PCR reaction. Therefore, according
to analysis on the above of intermediate species, a possible CO2
photoreduction conversion pathway over TiO2@ZnIn2S4 sur-
face can be proposed in Figure 7b. Under light irradiation, CO2
molecules are adsorbed and activated on TiO2@ZnIn2S4 surface.
First, the photogenerated electrons react with the activated CO2
molecules to form COO− intermediate species. Partial COO−
recombine with protons to form ·COOH radials. Afterwards,
·COOH radials react with protons and photogenerated electrons
to form H2O and CO. The rest of COO− intermediate radials
undergo further reaction with protons to form HCOO− interme-
diate species. Subsequently, HCOO− further accepts protons and
photogenerated electrons to form HCHO− species. Subsequently,
HCHO− species can be further converted into CH3O− interme-
diate species by reacting with the protons and photogenerated
electrons. Finally, CH3OH and CH4 are generated. Therefore, the
PCR conversion pathway over the surface of S-scheme TiO2@
ZnIn2S4 CSHS follows the formaldehyde pathway.[2,3,58,69]
Acknowledgements
This work was supported by the National Natural Science Foundation
of China (NSFC) (Nos. 51932007, 51872220, 51961135303, 21871217,
52073223, U1905215, and U1705251), the Fundamental Research Funds
for the Central Universities (No. WUT: 2021IVA137), and the Innovative
Research Funds of Foshan Xianhu Laboratory of the Advanced Energy
Science and Technology Guangdong Laboratory (XHD2020-001).
Conflict of Interest
The authors declare no conflict of interest.
Data Availability Statement
Research data are not shared.
Keywords
charge transfer, core-shell, hollow structures, S-scheme mechanism,
step-scheme heterojunction
Received: June 13, 2021
Revised: July 18, 2021
Published online:
[1] C. Bie, H. Yu, B. Cheng, W. Ho, J. Fan, J. Yu, Adv. Mater. 2021, 33,
2003521.
[2] T. Kong, Y. Jiang, Y. Xiong, Chem. Soc. Rev. 2020, 49, 6579.
[3] Z. Sun, N. Talreja, H. Tao, J. Texter, M. Muhler, J. Strunk, J. Chen,
Angew. Chem., Int. Ed. 2018, 57, 7610.
[4] S. Das, J. Pérez-Ramírez, J. Gong, N. Dewangan, K. Hidajat,
B. Gates, S. Kawi, Chem. Soc. Rev. 2020, 49, 2937.
[5] Y. Xia, Z. Tian, T. Heil, A. Meng, B. Cheng, S. Cao, J. Yu,
M. Antonietti, Joule 2019, 3, 2792.
[6] Z. Pan, M. Liu, P. Niu, F. Guo, X. Fu, X. Wang, Acta Phys.-Chim. Sin.
2020, 36, 1906014.
[7] U. Ulmer, T. Dingle, P. Duchesne, R. Morris, A. Tavasoli, T. Wood,
G. Ozin, Nat. Commun. 2019, 10, 3169.
[8] H. Wang, Y. Wang, L. Guo, X. Zhang, C. Ribeiro, T. He,
Chin. J. Catal. 2020, 41, 131.
[9] Y. Zhou, Z. Wang, L. Huang, S. Zaman, K. Lei, T. Yue, Z. Li, B. You,
B. Xia, Adv. Energy Mater. 2021, 11, 2003159.
[10] A. Crake, K. Christoforidis, A. Gregg, B. Moss, A. Kafizas, C. Petit,
Small 2019, 15, 1805473.
[11] J. He, C. Janáky, ACS Energy Lett. 2020, 5, 1996.
[12] V. Rodríguez-González, S. Obregón, O. Patrón-Soberano,
C. Terashima, A. Fujishima, Appl. Catal., B 2020, 270, 118853.
[13] R. Wang, M. Shi, F. Xu, Y. Qiu, P. Zhang, K. Shen, Q. Zhao, J. Yu,
Y. Zhang, Nat. Commun. 2020, 11, 4465.
3. Conclusions
In summary, hierarchical S-scheme TiO2@ZnIn2S4 CSHS
heterojunction is successfully fabricated by growing ZnIn2S4
nanosheets on the outer surface of TiO2 hollow spheres for
efficient PCR reaction. The S-scheme charge transfer mecha-
nism is revealed by by in situ irradiated XPS, work function
calculation and EPR measurements. The PCR performance
of S-scheme TiO2@ZnIn2S4 heterojunction photocatalyst is
significantly improved. The total CO2 photoreduction con-
version rates (the sum yield of CO, CH3OH and CH4) of
S-scheme TiO2@ZnIn2S4 heterojunction have increased by
2.75 and 4.43 times compared with the pristine ZnIn2S4 and
TiO2, respectively. Experimental results demonstrate that this
rational design can not only effectively facilitate the separation
of useful photogenerated electrons and holes, but also provide
large surface areas and abundant active sites for multi-electron
PCR reaction. Therefore, this work provides a feasible approach
to address the formation and photogenerated charge transfer
mechanism of heterojunction photocatalysts without cocatalysts
and sacrificial agents for the gas-solid PCR reaction system.
[14] A. Meng, B. Cheng, H. Tan, J. Fan, C. Su, J. Yu, Appl. Catal., B 2021,
289, 120039.
[15] J. Li, X. Wu, S. Liu, Acta Phys.-Chim. Sin. 2021, 37, 2009038.
[16] X. Zhou, Acta Phys.-Chim. Sin. 2021, 37, 2008064.
[17] A. Mamaghani, F. Haghighat, C. Lee, Appl. Catal., B 2020, 269,
118735.
[18] H. Tang, Y. Ren, S. Wei, G. Liu, X. Xu, Rare Met. 2019, 38, 453.
[19] C. Xu, P. Anusuyadevi, C. Aymonier, R. Luque, S. Marre, Chem. Soc.
Rev. 2019, 48, 3868.
Supporting Information
[20] L. Wang, B. Zhu, B. Cheng, J. Zhang, L. Zhang, J. Yu, Chin. J. Catal.
2021, 42, 1648.
Supporting Information is available from the Wiley Online Library or
from the author.
Small 2021, 2103447
2103447 (8 of 9)
© 2021 Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
[21] J. Tan, S. Gavrielides, M. Belekoukia, W. Thompson, L. Negahdar,
F. Xia, M. Maroto-Valer, A. Beale, Chem. Commun. 2020, 56, 12150.
[22] J. Wang, G. Wang, B. Cheng, J. Yu, J. Fan, Chin. J. Catal. 2021, 42, 56.
[23] F. Xu, K. Meng, B. Cheng, S. Wang, J. Xu, J. Yu, Nat. Commun. 2020,
11, 4613.
[24] Z. Wang, Y. Chen, L. Zhang, B. Cheng, J. Yu, J. Fan, J. Mater. Sci.
Technol. 2020, 56, 143.
[25] X. Lian, W. Xue, S. Dong, E. Liu, H. Li, K. Xu, J. Hazard. Mater. 2021,
412, 125217.
[26] S. Wageh, A. Al-Ghamdi, R. Jafer, X. Li, P. Zhang, Chin. J. Catal.
2021, 42, 667.
[27] Q. Xu, L. Zhang, B. Cheng, J. Fan, J. Yu, Chem 2020, 6, 1543.
[28] Z. Li, Z. Wu, R. He, L. Wan, S. Zhang, J. Mater. Sci. Technol. 2020,
56, 151.
[29] X. Fei, H. Tan, B. Cheng, B. Zhu, L. Zhang, Acta Phys.-Chim. Sin.
2021, 37, 2010027.
[30] R. He, H. Liu, H. Liu, D. Xu, L. Zhang, J. Mater. Sci. Technol. 2020,
52, 145.
[47] S. Wang, Y. Wang, S. Zhang, S. Zang, X. Lou, Adv. Mater. 2019, 31,
1903404.
[48] D. Majhi, K. Das, R. Bariki, S. Padhan, A. Mishra, R. Dhiman,
P. Dash, B. Nayak, B. Mishra, J. Mater. Chem. A 2020, 8, 21729.
[49] A. Chachvalvutikul, T. Luangwanta, S. Pattisson, G. Hutchings,
S. Kaowphong, Appl. Surf. Sci. 2020, 544, 148885.
[50] Q. Han, L. Li, W. Gao, Y. Shen, L. Wang, Y. Zhang, X. Wang, Q. Shen,
Y. Xiong, Y. Zhou, ACS Appl. Mater. Interfaces 2021, 13, 15092.
[51] G. Zhang, D. Chen, N. Li, Q. Xu, H. Li, J. He, J. Lu, Angew. Chem.,
Int. Ed. 2020, 132, 8332.
[52] Z. Mei, G. Wang, S. Yan, J. Wang, Acta Phys.-Chim. Sin. 2021, 37,
2009097.
[53] W. Stöber, A. Fink, E. Bohn, J. Colloid Interface Sci. 1968, 26, 62.
[54] H. Liu, W. Li, D. Shen, D. Zhao, G. Wang, J. Am. Chem. Soc. 2015,
137, 13161.
[55] F. He, B. Zhu, B. Cheng, J. Yu, W. Ho, W. Macyk, Appl. Catal., B
2020, 272, 119006.
[56] P. Zhang, Y. Li, Y. Zhang, R. Hou, X. Zhang, C. Xue, S. Wang,
B. Zhu, N. Li, G. Shao, Small Methods 2020, 4, 2000214.
[57] J. Low, B. Dai, T. Tong, C. Jiang, J. Yu, Adv. Mater. 2019, 31, 1802981.
[58] L. Wang, H. Tan, L. Zhang, B. Cheng, J. Yu, Chem. Eng. J. 2021, 411,
128501.
[31] C. Cheng, B. He, J. Fan, B. Cheng, S. Cao, J. Yu, Adv. Mater. 2021,
33, 2100317.
[32] P. Xia, S. Cao, B. Zhu, M. Liu, M. Shi, J. Yu, Y. Zhang, Angew. Chem.,
Int. Ed. 2020, 59, 5218.
[33] S. Wageh, A. Al-Ghamdi, L. Liu, Acta Phys.-Chim. Sin. 2021, 37,
2010024.
[59] Z. Xing, J. Hu, M. Ma, H. Lin, Y. An, Z. Liu, Y. Zhang, J. Li, S. Yang,
J. Am. Chem. Soc. 2019, 141, 19715.
[34] L. Liu, T. Hu, K. Dai, J. Zhang, C. Liang, Chin. J. Catal. 2021, 42, 46.
[35] Y. Ren, Y. Li, X. Wu, J. Wang, G. Zhang, Chin. J. Catal. 2020, 42, 69.
[36] W. Wang, W. Zhao, H. Zhang, X. Dou, H. Shi, Chin. J. Catal. 2021,
42, 97.
[37] Y. Wei, N. Yang, K. Huang, J. Wan, F. You, R. Yu, S. Feng, D. Wang,
Adv. Mater. 2020, 32, 2002556.
[38] C. Nguyen, N. Vu, T. Do, J. Mater. Chem. A 2015, 3, 18345.
[39] C. Bie, B. Zhu, F. Xu, L. Zhang, J. Yu, Adv. Mater. 2019, 31, 1902868.
[40] Z. Hu, X. Li, S. Zhang, Q. Li, J. Fan, X. Qu, K. Lv, Small 2020, 16,
2004583.
[60] Y. Xia, B. Cheng, J. Fan, J. Yu, G. Liu, Small 2019, 15, 1902459.
[61] H. Deng, X. Fei, Y. Yang, J. Fan, J. Yu, B. Cheng, L. Zhang, Chem.
Eng. J. 2020, 409, 127377.
[62] T. He, L. Wang, F. Fabregat-Santiago, G. Liu, Y. Li, C. Wang,
R. Guan, J. Mater. Chem. A 2017, 5, 6455.
[63] Y. Chiu, K. Chang, Y. Hsu, J. Mater. Chem. A 2018, 6, 4286.
[64] Z. He, Y. Han, S. Liu, W. Cui, Y. Qiao, T. He, Q. Wang,
ChemElectroChem 2020, 7, 4390.
[65] Z. Yang, Y. Qi, F. Wang, Z. Han, Y. Jiang, H. Han, J. Liu, X. Zhang,
W. Ong, J. Mater. Chem. A 2020, 8, 24868.
[41] M. Xiao, Z. Wang, M. Lyu, B. Luo, S. Wang, G. Liu, H. M. Cheng,
L. Wang, Adv. Mater. 2019, 31, 1801369.
[66] L. Pei, Y. Yuan, W. Bai, T. Li, H. Zhu, Z. Ma, J. Zhong, S. Yan, Z. Zou,
ACS Catal. 2020, 10, 15083.
[42] S. Wang, Y. Wang, S. Zang, X. Lou, Small Methods 2020, 4, 1900586.
[43] Y. Wang, S. Wang, L. Zhang, X. Lou, Angew. Chem., Int. Ed. 2020,
59, 11918.
[44] P. Zhang, X. Lou, Adv. Mater. 2019, 31, 1900281.
[45] S. Wang, B. Guan, X. Lou, J. Am. Chem. Soc. 2018, 140, 5037.
[46] Y. Xia, B. Cheng, J. Fan, J. Yu, G. Liu, Sci. China Mater. 2020, 63, 552.
[67] F. Chen, H. Huang, L. Ye, T. Zhang, Y. Zhang, X. Han, T. Ma,
Adv. Funct. Mater. 2018, 28, 1804284.
[68] X. Li, S. Wang, L. Li, Y. Sun, Y. Xie, J. Am. Chem. Soc. 2020, 142,
9567.
[69] S. Habisreutinger, L. Schmidt-Mende, J. Stolarczyk, Angew. Chem.,
Int. Ed. 2013, 52, 7372.
2103447 (9 of 9)
© 2021 Wiley-VCH GmbH
Small 2021, 2103447