Oxygen-vacancy-mediated exciton dissociation in biobr for boosting charge-carrier-involved molecular oxygen activation

Article abstract of DOI:10.1021/jacs.7b10997

Excitonic effects mediated by Coulomb interactions between photogenerated electrons and holes play crucial roles in photoinduced processes of semiconductors. In terms of photocatalysis, however, efforts have seldom been devoted to the relevant aspects. Fo

Full text of DOI:10.1021/jacs.7b10997

Subscriber access provided by READING UNIV
Article
Oxygen Vacancy Mediated Exciton Dissociation in BiOBr for
Boosting Charge-carrier-involved Molecular Oxygen Activation
Hui Wang, Dingyu Yong, Shichuan Chen, Shenlong Jiang, Xiaodong
Zhang, Wei Shao, Qun Zhang, Wensheng Yan, Bicai Pan, and Yi Xie
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b10997 • Publication Date (Web): 10 Jan 2018
Downloaded from http://pubs.acs.org on January 10, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Oxygen Vacancy Mediated Exciton Dissociation in BiOBr for
Boosting Charge-Carrier-Involved Molecular Oxygen Activation
Hui Wang,‡ Dingyu Yong,‡ Shichuan Chen, Shenlong Jiang, Xiaodong Zhang,* Wei Shao, Qun
Zhang,* Wensheng Yan, Bicai Pan, and Yi Xie*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hefei National Laboratory for Physical Science at the Microscale, CAS Center for Excellence in Nanoscience, iChEM,
Synergetic Innovation Center of Quantum Information and Quantum Physics, National Synchrotron Radiation
Laboratory, University of Science and Technology of China, Hefei, Anhui, 230026, P. R. China
ABSTRACT: Excitonic effects mediated by Coulomb interactions between photogenerated electrons and holes play
crucial roles in photoinduced processes of semiconductors. In term of photocatalysis, however, efforts have seldom been
devoted to the relevant aspects. As for the catalysts with giant excitonic effects, the co‐existing, competitive exciton
generation serves as a key obstacle to the yield of free charge carriers, and hence transformation of excitons into free
carriers would be beneficial for optimizing the charge‐carrier‐involved photocatalytic processes. Herein, by taking
bismuth oxybromide (BiOBr) as a prototypical model system, we demonstrate that excitons can be effectively dissociated
into charge carriers with the incorporation of oxygen vacancy, leading to excellent performances in charge‐carrier‐
involved photocatalytic reactions such as superoxide generation and selective organic syntheses under visible‐light
illumination. This work not only establishes an in‐depth understanding of defective structures in photocatalysts, but also
paves the way for excitonic regulation via defect engineering.
In view of the positive role of energetic disorder on
exciton dissociation, booming charge‐carrier‐involved
photocatalytic activity would be expected in catalysts
with disordered energy landscapes.[14–16] For instance, we
have recently demonstrated that enhanced electron
generation could be obtained in semicrystalline polymeric
photocatalysts, in which effective exciton dissociation at
order–disorder interfaces was identified, thus leading to
INTRODUCTION
For decades, semiconductor‐based
photocatalytic
reactions, such as water splitting, water pollution
treatments, and organic syntheses, have attracted
tremendous attention, due to their great potential in
solving energy crisis and environmental pollution.[1–4]
Among the numerous efforts, researches on the behaviors
of photogenerated excited species are of great significance
in pursuing advanced photocatalysts.[3–7] Nevertheless,
traditional viewpoints focused on free charge carriers are
somehow incomprehensive, since the Coulomb
interaction‐mediated many‐body effects have long been
ignored. Once these effects are considered, neutral
excitons (or bound electron–hole pairs) might be the
dominant photogenerated species, and the corresponding
exciton‐based resonance energy transfer could even
feasibly serve as an alternative photocatalytic mechanism
beyond the carrier‐based charge transfer.[8–10] Moreover,
the strong correlations among excitons would lead to
robust exciton–exciton annihilation that opens up an
important nonradiative depopulation pathway of
photoexcited species.[11–13] Undoubtedly, all these aspects
excellent
performance
in
charge‐carrier‐involved
photocatalytic reactions.[16] As is well known, defects can
significantly distort the distribution of electronic states in
the system, which inspires us to consider their potential
roles in exciton dissociation. In term of photocatalysis,
defect engineering has been widely accepted as one of the
most effective strategies that is closely related to
absorption range, charge separation, transport properties,
or active sites of catalysts,[17–21] whereas the relevant
impacts on excitonic processes are far from being
explored. To this end, we concentrate on bismuth
oxyhalides, whose robust excitonic effects, easy structural
modification, and promising photocatalytic activities[20,22–
26] enable us to evaluate the impact of defective structures
on the involved excitonic processes and the resulting
photocatalytic performance.
constitute
shackles
for
charge‐carrier‐involved
photocatalysis. Given the competitive relationship
between excitons and charge carriers in photoexcitation
processes, it would be reasonable to gain considerable
quantum yields of charge carriers by dissociating excitons
into free electrons and holes. In this case, it would be
meaningful to explore optimized strategies of exciton
Herein, by taking bismuth oxybromide (BiOBr) as a
prototypical model system, we demonstrate for the first
time that the incorporated oxygen vacancies (OV) can
significantly accelerate exciton dissociation, giving rise to
enhanced charge‐carrier generation. Benefiting from the
charge‐carrier harvesting, the defective BiOBr with
oxygen vacancies exhibits excellent performance in
dissociation
for
boosting
charge‐carrier‐involved
photocatalytic performance.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
1
2
3
4
5
6
7
8
9
charge‐carrier‐involved photocatalytic reactions like
ESR‐trapping measurements. 50 μL of aqueous
suspension of samples (4 g L‐1) was mixed with 500 μL of
2,2,6,6‐tetramethylpiperidine (TEMP, 10 mM) solution.
After being illuminated for 2 min, the mixture was
superoxide generation and aerobic oxidative coupling of
amines. This study provides an in‐depth understanding
on the design of advanced photocatalysts via defect
engineering.
characterized using
a
Bruker EMX plus model
spectrometer operating at the X‐band frequency (9.4 GHz)
•−
at room temperature. The O2 and H2O2 trapping‐ESR
tests were performed in absolute methanol as described
above, except using 5,5‐dimethyl‐1‐pyrroline‐N‐oxide
(DMPO, 20 mM) as the spin‐trapping agent. A xenon
lamp with a 420‐nm cutoff filter was employed as the
light source.
EXPERIMENTAL SECTION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of samples. Firstly, BiOBr nanoplates
with (001) exposed facet were prepared by a modified
hydrothermal method according to previous reports.[26] In
detail, 2.425 g bismuth nitrate pentahydrate and 0.595 g
potassium bromide were successively added into the
mixture of 70 mL distilled water and 5 mL ethylene glycol
under continuous stirring for 30 min. The mixed solution
was then transferred into 100 mL Teflon lined autoclave
Characterizations. The X‐ray diffraction (XRD)
patterns were obtained on a Philips X’Pert Pro Super
diffractometer with Cu Kα radiation (λ = 1.54178 Å). The
scanning electron microscopy (SEM) images were
obtained on a JEOL JSM‐6700F SEM. The high‐angle
annular dark‐field scanning transmission electron
microscopy (HAADF–STEM) measurement was carried
out on a JEOL JEM‐ARM200F microscope with spherical
aberration correction. The X‐ray photoelectron spectra
(XPS) were collected on an ESCALAB MKII with Mg Kα
(ħν = 1253.6 eV) as the excitation source, using C 1s (284.6
eV) as a reference. The valance band XPS spectra were
collected at the catalysis and surface science endstation in
the National Synchrotron Radiation Laboratory (NSRL),
Hefei, China. The electron spin resonance (ESR)
measurements were carried out on a Bruker EMX plus
model spectrometer operating at the X‐band frequency
(9.4 GHz). The ultraviolet–visible (UV–vis) spectra were
collected on a Perkin Elmer Lambda 950 UV–vis–NIR
spectrophotometer. The Fourier transform infrared (FT‐
IR) spectra were recorded on a Magna‐IR750 FT‐IR
spectrometer in a KBr pellet at room temperature. The
NMR experiments were performed on a 400‐MHz Bruker
AVANCE AV III NMR spectrometer. The electrochemical
measurements were carried out on an electrochemical
workstation (CHI760E, Shanghai Chenhua Limited,
o
and heated in a sealed autoclave at 160 C for 24 hours.
Upon being cooled to room temperature naturally, the
precipitate was collected and washed with distilled water
for several times, and then the products were dried under
vacuum for 12 hours. After another heating treatment
under high‐purity O2 atmosphere (0.5 MPa, 2 hours at 300
oC, to remove the tiny amount of oxygen vacancies), the
powder sample (denoted BiOBr) was collected. The
defective sample (denoted BiOBr‐OV) was obtained
through a heating treatment under vacuum condition
(~10‐3 Pa, 3 hours at 300 oC).
Calculations. First‐principles calculations were
performed by using the Vienna ab initio simulation
package. The interactions between ions and valence
electrons were described using the projector augmented
wave potentials, and the generalized gradient
approximation in the Perdew–Burke–Ernzerhof form was
employed to correct the exchange correlations between
electrons. The one‐electron wave functions on a plane‐
wave basis were expanded with an energy cutoff of 500 eV.
The Brillouin zones (with the k‐point separation <0.05 Å−1)
for all the structures were sampled by using the Gamma‐
centered Monkhorst–Pack mesh. All of the atoms and
geometries were allowed to relax and fully optimized,
until the forces on atoms were less than 0.01 eV Å−1. The
energy cutoff for calculating the numbers of localized
valence and conduction charge density near band‐edge
was set to 0.3 eV.
China).
phosphorescence spectra as well as the time‐resolved
phosphorescence spectra were collected on
The
steady‐state
fluorescence
and
a
FLUOROLOG‐3‐TAU fluorescence spectrometer. The
nanosecond‐domain time‐resolved fluorescence spectra
were acquired on an FLS920 fluorescence spectrometer
(Edinburgh Instruments Ltd.). The ultrafast transient
absorption (TA) measurements were performed, under
ambient conditions, on an ExciPro pump–probe
spectrometer (CDP) coupled to an amplified femtosecond
(fs) laser system (3 mJ, 35 fs; Coherent) with a temporal
resolution of ~100 fs. The X‐ray absorption spectra (XAS)
were measured at the beamline BL12b of National
Synchrotron Radiation Laboratory (NSRL, China) in the
total electron yield mode by collecting the sample drain
current under a vacuum better than 5 × 10‐8 Pa.
3,3',5,5'‐tetramethylbenzidine (TMB) tests. In
a
typical test, 1 mg catalyst was dispersed into 2 mL TMB
solution (500 μM, HAc/NaAc buffer solution), the
mixture was illuminated after the saturation of adsorption
by using a xenon lamp (PLS‐SXE300/300UV, Trusttech
Co., Ltd., Beijing) with a 420‐nm cutoff filter as the light
source. The oxidation of TMB molecules was evaluated by
monitoring the absorbance change of the solution with a
UV–vis spectrophotometer. The scavenger tests were
performed as the above procedure with certain amounts
of different scavengers, that is, carotene, 4 mg; mannite,
50 mM, 300 μL; catalase, 4000 unit/mL, 300 μL;
superoxide dismutase (SOD), 4000 unit/mL, 300 μL.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
1
2
3
obvious 0.4‐eV red‐shift, which can be assigned to the
4
5
6
7
8
9
breakage of Bi–O bond caused by oxygen vacancies. As
one of the most effective techniques for probing defective
structures, ESR spectroscopy was employed to glean more
direct evidence. As displayed in Figure 1d, the BiOBr‐OV
sample turned out to possess an typical ESR signal
centered at g = 2.002, owing to the electrons trapped on
the oxygen vacancies. No obvious ESR signal was detected
for the BiOBr case. As an effective technique for
charactering local atomic environments, XAS was
employed to verify the obvious change of hybridization
strengths of Bi–O bonds induced by oxygen vacancies
(Figure S6). The UV–vis spectra (Figure S7) revealed the
similar absorption range and band gaps (~2.85 eV) for the
two samples. Moreover, excitonic absorption in BiOBr
system was observed in the high‐resolution UV–vis
spectra，showing weak absorption bands at 470 nm for
both cases. It is worth noting that the oxygen vacancies in
BiOBr‐OV are mainly the bulk defects, rather than surface
defects, in light of the similar nitrogen adsorption
isotherms for the two samples (Figure S8).[23] According to
the above characterizations and discussions, we can safely
conclude that the BiOBr samples with oxygen vacancies
have been successfully prepared.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. (a) XRD pattern and (b) atomic‐resolution
HAADF–STEM image of BiOBr‐OV. (c) O 1s spectra and
(d) room‐temperature ESR spectra of both BiOBr‐OV
and BiOBr samples.
In this study, the BiOBr samples with and without
oxygen vacancies (denoted BiOBr‐OV and BiOBr,
respectively) were prepared by an annealing treatment on
the BiOBr nanoplates under vacuum and O2 atmosphere,
respectively. The structural investigation of both samples
with the XRD patterns (Figure 1a and Figure S1)
demonstrated the high purity with orientation along the
[001] direction (JCPDS card No. 73‐2061). The typical
vibration modes revealed by Raman spectra (Figure S2)
also confirmed the structural information of the samples,
while the defective BiOBr‐OV sample exhibits red‐shifted
bands with respect to the BiOBr sample, which can be
assigned to the slight structural distortion induced by
oxygen vacancies.[24] The SEM measurements (Figure S3)
suggested a similar morphology for the two samples.
Further atomic‐resolution HAADF–STEM (Figure 1b)
image revealed the single crystal feature of the sample, in
which the atomic distance of ~0.277 nm was in
accordance with the Bi atom arrangement of (001) facet.
Given that the intensity of HAADF–STEM image is
roughly proportional to the square of atomic number
(namely, Z‐contrast), the extremely large specific value of
It is widely accepted that defective structures would
inevitably influence the electronic structure of materials.
Herein, we conducted first‐principles calculations based
on density functional theory to investigate the impact of
oxygen vacancy on the electronic structure of the BiOBr
system. As shown in Figure S9, the calculated charge
2
ZBi2/ ZO (~107.6) makes it impossible to observe oxygen
vacancy directly. The XPS results gave detailed
information on chemical bonds of the samples. The XPS
survey spectra revealed the chemical compositions (i.e.,
bismuth, oxygen, and bromine), indicating the high
purity of the samples (Figure S4). As compared to BiOBr
with two main components from lattice oxygen (530.3 eV)
and hydroxyl groups (531.9 eV), the broadened O 1s
spectrum (Figure 1c) with BiOBr‐OV can be ascribed to
the promoted contribution (531.2 eV) from the adsorbed
oxygen species at the vacancy sites.[27,28] The Bi 4f
spectrum of BiOBr (Figure S5) presents a typical spin–
orbit doublet splitting centered at 159.7 and 165.0 eV. By
comparison, the Bi 4f spectrum of BiOBr‐OV exhibits an
Figure 2. (a) Valence charge density and (b) conduction
charge density near the band edges of BiOBr model with
oxygen vacancy. (c) The (NcNv)1/2 value of atomic sites at
different distances around oxygen vacancy.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
1
2
3
4
5
6
7
8
9
density of the defect energy level in BiOBr with oxygen
On the basis of the above theoretical simulations, we
propose that the event of exciton dissociation can be
realized in the presence of oxygen vacancy. To verify this,
we performed photoluminescence measurements under
both ambient and low‐temperature conditions. The
prompt fluorescence spectra at 300 K recorded on the two
samples are shown in Figure 3a. As for BiOBr‐OV, the
fluorescence spectrum exhibits two main peaks centered
at 430 and 610 nm: the former can be assigned to the
intrinsic band‐edge emission (corresponding to the band
gap of ~2.85 eV), while the latter is a characteristic
emission from the defect state induced by oxygen
vacancies.[23] In comparison, no obvious emissions were
observed for BiOBr due to its indirect band feature. To
further understand the photoexcitation processes
involved in BiOBr‐OV, we performed time‐resolved
prompt fluorescence measurements by monitoring the
corresponding emission peaks. The results shown in
Figure 3b indicated that the lifetimes of such emissions
are both on a nanosecond timescale. Strikingly, the
emission induced by oxygen vacancy exhibits a much
longer lifetime (~19 ns) with respect to the intrinsic band‐
edge emission (~5 ns), suggesting distinctly different
relaxation processes of the residential charge carriers in
the two states.
vacancy suggested a strong localized charge density
around the vacancy sites, leading to energetic disorder
therein. Moreover, it turned out that oxygen vacancy
tends to pose more impacts on the charge density
distributions at the bromine sites, as seen from the
obvious localized charge densities on the second‐ and
third‐nearest neighbor bromine sites. Given that the
interacting holes and electrons (related to the excitonic
effects) mainly originate from the states close to the
valence band maximum and conduction band minimum,
it is necessary to investigate the charge localization near
the band edges in defective BiOBr.[26,29,30] As displayed in
Figure 2a, the valence‐band‐edge charge density is mainly
contributed by oxygen and bromine atoms, while oxygen
vacancy shows an evident impact on the charge density
localization of the nearby bromine atoms (within a
distance of ~7 Å). In comparison, Figure 2b illustrates the
disorder distribution of the conduction‐band‐edge charge
density with a dominant contribution from the bismuth
atoms. Note that the partial contributions from the
nearby bromine atoms and the nearest oxygen atoms
have also been verified. Given that the simultaneous
localizations of conduction and valence band‐edge states
on Bi atomic sites are responsible for the significant
electron–hole interaction, the disorder‐redistributed
band‐edge states would greatly influence the stability of
excitons near the vacancy sites. Herein, the products of
the numbers of localized valence and conduction states
near band edges for different atoms (denoted NcNv, where
Nc and Nv are the numbers of localized conduction and
valence states near band edges, respectively, within a
cutoff energy) were calculated to elucidate the electron–
hole interaction at certain atoms (Figure 2c). For the
three kinds of atoms far away from the oxygen vacancies,
the corresponding (NcNv)1/2 values were found to keep
constant, among which the bismuth atoms possess the
largest (NcNv)1/2 values, which revealed the significant,
simultaneous localizations of conduction and valence
band‐edge states, in accordance with our previous
report.[26] Nevertheless, the (NcNv)1/2 values of the three
kinds of atoms close to the vacancy sites exhibit distinctly
different tendencies. As for the nearby bromine atoms
(~10 Å), the (NcNv)1/2 values obviously increase with
decreasing the distances, suggesting enhanced electron–
hole interaction at these atoms. By contrast, the (NcNv)1/2
values for the oxygen and bismuth atoms show a
negligible increase aside from the nearest neighboring
sites. By virtue of the comparable (NcNv)1/2 values for the
three kinds of atoms, the extended (that is, delocalized)
band‐edge states can be expected, thus leading to the
formation of much less compact excitons, which is less
stable than the compact ones (Bi‐case). In this case, it is
rational to infer that excitons are unstable once they
approach to the oxygen vacancies. It should be noted that
that the DFT results only gave a qualitative prediction of
the trend in excitonic properties, rather than the kinetics
of excitons.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The dark feature and the robust nonradiative relaxation
of excitons urged us to pursue excitonic emissions under
low temperatures.[31–33] As displayed in Figure 3c, the
prompt fluorescence spectrum of BiOBr recorded at 8 K
exhibits a strong sub‐bandgap emission at around 482 nm,
which can be attributed to the radiative decay of excitons.
Figure 3. (a) Prompt fluorescence of BiOBr and BiOBr‐
OV samples under room temperature. (b) Time‐resolved
photoluminescence spectra of BiOBr‐OV sample
monitored at 430 and 610 nm. (c) Low temperature (8 K)
prompt fluorescence and (d) low temperature steady‐
state phosphorescence recorded at a delayed time of 10
ms for both BiOBr and BiOBr‐OV samples. All the
photoluminescence spectra are recorded with a 380‐nm
excitation.
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The slight red‐shift in emission with respect to our
oxygen vacancy on excitonic processes, we further
interrogated the BiOBr and BiOBr‐OV samples with
femtosecond time‐resolved transient absorption (TA)
spectroscopy, a robust tool for tracking in real time the
relaxation kinetics of photogenerated charge carriers in
semiconductor systems.[36–38] Herein, we chose a pump–
probe configuration with a 380‐nm pump and a white‐
light probe (430–610 nm) to trace the photogenerated
electron kinetics. As shown in Figure 4a, similar TA
spectral profiles (taken at a probe delay of 3 ps)
manifested as photoinduced absorption (in positive
values) were observed for both BiOBr and BiOBr‐OV, in
which the intensity difference can serve as an indicator of
previous report[26] is attributed to the different synthetic
methods in the two works. However, apart from the
excitonic emission, the BiOBr‐OV sample exhibits
another distinct emission peak at around 570 nm that can
be associated with the oxygen vacancy states. Note that
the blue‐shift of oxygen‐vacancy‐induced emission from
610 to 570 nm arises as a result of the reduced electron–
phonon interaction under low temperature.[34,35] In view
of the spin‐forbidden character of the excitonic emission,
we further performed phosphorescence measurements on
the samples (Figure 3d). With a long delayed time of 10
ms, the BiOBr sample exhibits a similar emission to the
prompt one, expect for the red‐shift from 482 to 495 nm.
In contrast, the BiOBr‐OV sample exhibits a dominant
emission at around 570 nm, with negligible excitonic
emission. The huge difference between the vacancy‐state
lifetime (~19 ns) and the delayed time (10 ms) suggested
the strong correlation between the excitonic and defect
emissions; in other words, the exciton dissociation at the
oxygen‐vacancy sites is responsible for the lasting defect
emission. The time‐resolved phosphorescence spectra
(Figure S10) recorded at 495 and 570 nm for BiOBr and
BiOBr‐OV suggested the sub‐second lifetimes for the
emissions (~90.2 and 54.0 ms, respectively), of which the
faster decay in BiOBr‐OV also verified the promoted
exciton relaxation induced by oxygen vacancy. Also, the
high photocurrent response of BiOBr‐OV echoed well to
its promoted exciton dissociation and increased carrier
concentrations (Figure S11), whereas the relatively low
photocurrent response of BiOBr was in accordance with
the robust excitonic effects of the system. The above
results clearly demonstrated the oxygen‐vacancy‐induced
exciton dissociation in BiOBr‐OV.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
carrier concentration. Nevertheless,
a
pronounced
difference in relaxation kinetics for the two cases was
observed. Since it turned out that the TA kinetics were
insensitive to the probing wavelength, we present in
Figure 4b a representative set of data taken 550 nm.
Clearly, the BiOBr‐OV sample exhibits a much longer
recovery lifetime of photoexcited electrons as compared
to the BiOBr sample, which would facilitate the migration
of charge carriers for photocatalytic reactions. As for
BiOBr, the observed bi‐exponential decay processes (τ1 =
3.9 ± 0.5 ps, 74%; τ2 = 81.8 ± 10.9 ps, 26%) suggested a
consecutive, trap‐state‐mediated relaxation of the
photogenerated electrons,[39,40] with an average lifetime of
~24 ps. Given the robust excitonic effects in the BiOBr
system, the two consecutive relaxation pathways can be
understood according to the scheme in Figure 4c (black
dotted lines), where the excitonic state acts as an
intermediate trap state (i.e., exciton‐mediated trap state).
Markedly, it turned out that the TA kinetics of BiOBr‐OV
can only be fitted well by a tri‐exponential decay function
(τ1 = 4.4 ± 0.2 ps, 51%; τ2 = 53.0 ± 4.2 ps, 21%; τ3 = 944.7 ±
50.1 ps, 28%; average lifetime ~278 ps), rather than a bi‐
exponential decay function. By combining the TA
analyses with the information from the above
photoluminescence results, we schematically illustrate
the three‐step relaxation pathways in Figure 4c by the
olive solid lines. After the photoexcited electrons undergo
To gain in‐depth understanding on the influence of
a
fast relaxation (τ1, a few picoseconds) from the
conduction band minimum to the exciton‐mediated trap
state, the resulting excitons would then relax to the
oxygen‐vacancy‐mediated trap state and meanwhile
dissociate into free carriers (τ2, a few tens of picoseconds),
and eventually decay on a nanosecond timescale (τ3). The
above assignments were based on the following facts: the
first component of BiOBr‐OV (~4.4 ps) nicely matches the
corresponding value of BiOBr (~3.9 ps); the second
relaxation in BiOBr‐OV (~53 ps) is faster than the exciton
decay in BiOBr (~81.8 ps); the third component (~0.9 ns)
in the TA probing time window is in agreement with the
nanosecond‐level lifetime of the oxygen‐vacancy‐induced
fluorescence. It is worth noting that the photophysical
processes in Figure 4c only illustrated the relaxation of
electrons (with the energy position of the trap states
being schematically depicted; see detailed information of
the band structures in Figure S12). Therefore, the ultrafast
TA analyses further highlighted the role of the oxygen‐
Figure 4. (a) Representative TA spectra taken at the
probe delay of 3 ps for both samples. (b) TA kinetic
traces probed at 550 nm for both samples. (c) Schematic
illustration of the photophysical processes involving
exciton‐ and oxygen‐vacancy‐mediated trap states.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
1
2
3
which can provide one of the most convictive evidences
4
5
to identify the ROS products, was further performed to
verify the different photocatalytic ROS generation in the
two cases. Here, 5,5‐dimethyl‐1‐pyrroline‐N‐oxide (DMPO)
was selected as the trapping agent for the detection of
O2•− in methanol. The strong, sextet ESR signal for BiOBr‐
OV case displayed in Figure 5c verified the formation of
DMPO–OOH, which is a spin adduct derived from
DMPO–O2•−, thus offering conclusive evidence for the
promoted generation of O2•−. In contrast, the BiOBr
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
•−
sample led to negligible O2 production. On the other
1
hand, the O2 generation was verified by using 2,2,6,6‐
tetramethylpiperidine (TEMP) as the trapping reagent. As
shown in Figure 5d, the typical 1:1:1 triplet signal with g‐
value of 2.0056 for BiOBr is in good agreement with those
of
2,2,6,6‐tetramethylpiperidine‐N‐oxyl
(TEMPO),
1
Figure 5. (a) Time‐dependent absorption spectra of
TMB oxidation with BiOBr and BiOBr‐OV in air. (b) The
absorbance evolution of TMB oxidation with BiOBr and
BiOBr‐OV monitored at the peak around 380 nm in the
presence of different scavengers. (c) ESR spectra of
different samples in the presence of DMPO in methanol.
(d) ESR spectra of different samples in the presence of
TEMP in water.
confirming its high O2 generation. In comparison, a tiny
TEMPO signal was observed in the BiOBr‐OV case,
demonstrating the significant suppression in 1O2
generation. Moreover, negligible H2O2 generation was
detected under visible‐light illumination in both cases
•−
(Figure S13). Besides, the O2
generation was
demonstrated to be dependent on the concentration of
oxygen vacancies, verifying the defect‐mediated exciton
dissociation in BiOBr system (Figure S14). The influence
of different light absorption and potential hydrogen
incorporation in the two samples were also investigated,
showing negligible impacts on the ROS generation
(Figure S15 and S16). On the basis of the above
measurements, we conclude that the existence of oxygen
vacancies greatly facilitates the charge‐carrier‐involved
photocatalytic ROS generation in the BiOBr system.
vacancy‐induced exciton dissociation and the involved
photoinduced processes in the BiOBr system.
On the basis of the boosted exciton dissociation and
charge‐carrier generation, one can expect a promoted
charge‐carrier‐involved photocatalytic activity with the
defective BiOBr sample. Herein, we carried out molecular
oxygen activation measurements to evaluate the involved
•−
1
The conversely generated O2 and O2 yields suggested
the totally different photoexcitation processes induced by
photocatalytic
performance
under
visible‐light
illumination ( ≥ 420 nm). 3,3',5,5'‐tetramethylbenzidine
(TMB) was selected as the probe molecule, by monitoring
the absorbance change of the solution. The absorption
evolution of TMB solutions (Figure 5a) clearly suggested
the generation of reactive oxygen species (ROS) for both
samples. Obviously, the BiOBr‐OV sample exhibits
promoted molecular oxygen activation ability. The
scavenger measurements were further performed to
identify the functional ROS in the two cases, where
superoxide dismutase (SOD), mannitol, catalase, and
carotene were used to selectively eliminate superoxide
(O2•−), hydroxyl radicals (•OH), hydrogen peroxide (H2O2),
and singlet oxygen (1O2), respectively. As shown in Figure
5b, significant suppression of TMB oxidation was
observed in the presence of SOD, whereas the other
scavengers were found to execute much less influence on
1
oxygen vacancy. Generally, O2 generation is believed to
undergo an energy transfer process between ground‐state
molecular oxygen and long‐lived excitons,[41,42] while O2
•−
generation mainly involves photogenerated electron
transfer between molecular oxygen and catalysts. The
1
possible charge transfer process for O2 generation in
BiOBr was excluded by scavenger tests (Figure S17). In
this case, it is reasonable for the detection of high 1O2 but
•−
negligible O2 in the presence of BiOBr, whose confined
layered structure sustains a considerable exciton binding
energy. The large exciton binding energy implies strong
correlations between photogenerated holes and electrons,
thus giving rise to high exciton concentration but low
1
electron yield. As for the defective case, the tiny O2
generation suggested the significantly suppressed
•−
excitonic processes, while a much higher O2 yield
•−
TMB oxidation. Therefore, O2 can be determined to be
indicated the promoted electron concentration in BiOBr‐
the dominating ROS in the photocatalytic molecular
OV with respect to BiOBr. The intriguing, complementary
1
oxygen activation of BiOBr‐OV. On the other hand, O2
1
•−
evolutions of O2 and O2 in BiOBr‐OV further confirmed
the positive role of oxygen vacancy on exciton
dissociation into free charge carriers.
mainly accounts for TMB oxidation in the BiOBr case,
pointing to its giant exciton effects. ESR measurement,
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
1
2
3
reaction. The results suggested that the defective sample
4
5
6
7
8
9
possesses much higher yields than the BiOBr sample for
both electron‐withdraw (para‐F, entry 3; para‐Cl, entry 4)
and electron‐donating (para‐CH3O, entry 5; para‐CH3,
entry 6) cases, which confirmed the dominating
advantage arising from the oxygen‐vacancy‐induced
exciton dissociation in BiOBr‐OV. Moreover, as compared
with the electron‐donating cases, the electron‐withdraw
groups exhibited positive roles in these oxidative coupling
reactions (refer to the much higher reaction rates),
echoing well to the O2•−‐involved mechanism in the
BiOBr‐OV case. The factor of substitution position in
benzylamine derivatives was also investigated, which
exhibited negligible impact on the yields of corresponding
imines (ortho‐Cl, entry 7; meta‐Cl, entry 8; ortho‐CH3,
entry 9; meta‐CH3, entry 10). The cycling tests of oxidative
coupling of benzylamines (Figure S20) were further
performed, which exhibited negligible reduction within
five consecutive cycles, verifying the excellent
photocatalytic stability of the BiOBr‐OV samples. The
above results confirmed the outstanding performance of
the BiOBr‐OV system for selective oxidation, opening an
intriguing prospective in further applications.
Table 1. Aerobic oxidative coupling of amines.[a]
BiOBr
BiOBr‐OV
Entry
R
Time / h
Conv.[b] Select.[c] Conv. Select.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2[d]
3[e]
4
H
H
12
12
12
8
24
9
95
trace
27
96
5
99
trace
trace
99
H
35
31
17
p‐F
90
95
99
99
94
56
99
99
99
99
5
p‐Cl
p‐Me
p‐OMe
o‐Cl
m‐Cl
o‐Me
m‐Me
8
24
49
34
27
36
56
42
99
6
15
24
8
94
95
7
97
90
8
99
97
99
9
8
99
10
11
15
15
95
96
95
95
[a] Reaction conditions: substrate (0.2 mmol), catalyst (10 mg),
acetonitrile (1 ml), air, 20 oC, xenon lamp with a cut‐off filter ( ≥
420 nm). [b] Determined by NMR analyses, using s‐trioxane as the
internal standard, mol%. [c] Selectivity
=
yieldimine*2
/
conversionamine
,
mol%. [d] Argon atmosphere. [e] Additional
CONCLUSION
methanol 100 μL.
In conclusion, we have systematically investigated the
influence of oxygen vacancy on photoinduced processes
in a prototypical model of BiOBr, where the oxygen‐
vacancy‐mediated exciton dissociation in the catalyst was
highlighted for the first time. Theoretical simulations
suggested that oxygen vacancy (OV) can significantly
distort the localization of the band‐edge states around the
defective sites, leading to instability of excitons. By the
joint observations from photoluminescence spectroscopy
and photoelectrochemical experiments, we demonstrated
the oxygen‐vacancy‐mediated exciton dissociation that
results in promoted charge‐carrier generation in the
system. Such an effect was further verified by ultrafast
transient absorption spectroscopy, establishing an in‐
depth understanding on the pertinent photophysical
processes. Benefiting from the robust exciton dissociation,
the defective BiOBr‐OV system was demonstrated to
exhibit excellent performance in charge‐carrier‐involved
photocatalytic reactions such as superoxide radical
generation and selective oxidative‐coupling reaction. This
work not only sheds new light on the photoinduced
processes of the well‐established catalysts, but also
•−
By virtue of its high‐efficiency O2 generation, the
defective BiOBr‐OV sample can be expected to hold great
potential in photocatalytic organic syntheses. Herein,
aerobic oxidative coupling of amines to imines, which is
an extensively studied reaction due to its critical role in
the fields of chemistry and biology,[43,44] was employed to
evaluate the photocatalytic performance of the defective
sample (Figure S18). Herein, a series of amines were
selected as the oxidation substrates, of which the
conversion rates and selectivities under the optimized
conditions are listed in Table 1. In detail, the defective
sample can convert benzylamine (entry 1; Figure S19) to
N‐benzylidenebenzylamine with high efficiency and
selectivity, while the yield of corresponding imine was
much lower in the BiOBr case. The high selectivity in the
1
BiOBr case is ascribed to the O2‐involved mechanism for
the oxidative coupling of benzylamines,[45] whereas the
low conversion rate may originate from the exciton–
exciton annihilation induced, relatively inefficient
quantum yields. To verify the vital role of ROS on the
reaction, control groups under inert atmosphere were
examined, where the conversion yields were significantly
suppressed for both BiOBr and BiOBr‐OV (entry 2).
Besides, the additional methanol (an effective hole‐
scavenger, entry 3) significantly suppressed the
conversion rate and selectivity for BiOBr‐OV case, in
accordance with the formation of cation radical complex
through the hole‐oxidation within the O2•−‐involved
mechanism (as illustrated in Figure S18),[46] whereas the
incomplete suppression for BiOBr case could be related to
provides
a defect‐engineering strategy on excitonic
regulation.
ASSOCIATED CONTENT
Supporting Information. This material is available free of
charge via the Internet at http://pubs.acs.org.
Additional structural characterizations; theoretical
calculations;
photoluminescence
spectra;
1
the O2‐involved mechanism. We further looked into the
influence of substituent groups on the oxidative‐coupling
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
1
2
3
4
5
6
7
8
9
(15) Paquin, F.; Rivnay, J.; Salleo, A.; Stingelin, N.; Silva‐Acuña,
C. J. Mater. Chem. C 2015, 3, 10715.
photoelectrochemical tests; and photocatalytic
measurements (PDF)
(16) Wang, H.; Sun, X.; Li, D.; Zhang, X.; Chen, S.; Shao, W.;
Tian, Y.; Xie, Y. J. Am. Chem. Soc. 2017, 139, 2468.
(17) Chen, X.; Liu, L.; Yu, P.; Mao, S. Science 2011, 33, 746.
(18) Nowotny, M.; Sheppard, L.; Bak, T.; Nowotny, J. J. Phys.
Chem. C 2008, 112, 5275.
AUTHOR INFORMATION
(19) Ida, S.; Kim, N.; Ertekin, E.; Takenaka, S.; Ishihara, T. J.
Am. Chem. Soc. 2015, 137, 239.
Corresponding Author
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
zhxid@ustc.edu.cn; qunzh@ustc.edu.cn; yxie@ustc.edu.cn
(20) Guan, M.; Xiao, C.; Zhang, J.; Fan, S.; An, R.; Cheng, Q.;
Xie, J.; Zhou, M.; Ye, B.; Xie, Y. J. Am. Chem. Soc. 2013, 135, 10411.
(21) Wang, S.; Pan, L.; Song, J.; Mi, W.; Zou, J. J.; Wang, L.;
Zhang, X. J. Am. Chem. Soc. 2015, 137, 2975.
(22) Li, J.; Yu, Y.; Zhang, L. Nanoscale 2014, 6, 8473.
(23) Li, H.; Shang, J.; Ai, Z.; Zhang, L. J. Am. Chem. Soc. 2015,
137, 6393.
(24) Wang, X.; Zhao, Y.; Li, F.; Dou, L.; Li, Y.; Zhao, J.; Hao, Y.
Sci. Rep. 2016, 6, 24918.
(25) Li, H.; Li, J.; Ai, Z.; Jia, F.; Zhang, L. Angew. Chem., Int. Ed
2017, doi:10.1002/anie.201705628
(26) Wang, H.; Chen, S.; Yong, D.; Zhang, X.; Li, S.; Shao, W.;
Sun, X.; Pan, B.; Xie, Y. J. Am. Chem. Soc. 2017, 139, 4737.
(27) Uhlenbrock, S.; Scharfschwerdt, C.; Neumann, M.; Illing,
G.; Freund, H. J. Phys.: Condens. Matter 1992, 4, 7973.
(28) Wang, L.; Ma, Y.; Wang, Y.; Liu, S.; Denng, Y. Catal.
Commun. 2011, 12, 1458.
(29) Dvorak, M.; Wei, S. H.; Wu, Z. Phys. Rev. Lett. 2013, 110,
016402.
(30) Su, Y. Q.; Zhu, Y.; Yong, D. Y.; Chen, M. M.; Su, L. X.;
Chen, A. Q.; Wu, Y. Y.; Pan, B. C.; Tang, Z. K. J. Phys. Chem. Lett.
2016, 7, 1484.
Author Contributions
‡ These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
The work was supported by the National Key R & D Program
on Nano Science
&
Technology of the MOST
(2017YFA0207301, 2016YFA0200602), the National Natural
Science Foundation of China (U1532265, U1632149, 11621063,
21401181, 21573211, 21633007), the Youth Innovation Promotion
Association of CAS (2017493), Anhui Provincial Natural
Science Foundation (1708085QB24), Key Research Program
of Frontier Sciences (QYZDY‐SSW‐SLH011), the National
Postdoctoral Program for Innovative Talents (BX201700219),
and the Fundamental Research Funds for the Central
Universities (WK2340000063). The authors thank Dr.
Huanxin Ju (National Synchrotron Radiation Laboratory,
USTC) for his helpful suggestions on the valance band XPS
analyses.
(31) Mortimer, I. B.; Nicholas, R. J. Phys. Rev. Lett. 2007, 98,
27404.
(32) Yersin, H. Top. Curr. Chem. 2004, 241, 1.
(33) Zhao, H.; Mazumdar, S. Phys. Rev. Lett. 2004, 93, 157402.
(34) Karaiskaj, D.; Engtrakul, C.; McDonald, T.; Heben, M. J.;
Mascarenhas, A. Phys. Rev. Lett. 2006, 96, 106805.
(35) Karaiskaj, D.; Mascarenhas, A. Phys. Rev. B 2007, 75,
115426.
(36) Hong, X. P.; Kim, J.; Shi, S. F.; Zhang, Y.; Jin, C. H.; Sun, Y.
H.; Tongay, S.; Wu, J. Q.; Zhang, Y. F.; Wang, F. Nat. Nanotech‐
nol. 2014, 9, 682.
REFERENCES
(1) Congreve, D.; Lee, J.; Thompson, N.; Hontz, E.; Yost, S.;
Reusswig, P.; Bahlke, M.; Reineke, S.; Van Voorhis, T.; Baldo, M.
Science 2013, 340, 334. 1. Hoffmann, M.; Martin, S.; Choi, W.;
Bahnemann, D. Chem. Rev. 1995, 95, 69.
(2) Chen, C.; Ma, W.; Zhao, J. Chem. Soc. Rev. 2010, 39, 4206.
(3) Chen, X.; Shen, S.; Guo, L.; Mao, S. Chem. Rev. 2010, 110,
6503.
(4) Serpone, N.; Emeline, A. J. Phys. Chem. Lett. 2012, 3, 673.
(5) Tian, Y.; Tatsuma, T. J. Am. Chem. Soc. 2005, 127, 7632.
(6) Mangold, M.; Weiss, C.; Calame, M.; Holleitner, A. Appl.
Phys. Lett. 2009, 94, 161104.
(7) Marschall, R. Adv. Funct. Mater. 2014, 24, 2421
(8) Mongin, C.; Garakyaraghi, S.; Razgoniaeva, N.; Zamkov, M.;
Castellano, F. Science 2016, 351, 369.
(9) Linsebigler, A.; Lu, G.; Yates, J. Chem. Rev. 1995, 95, 735.
(10) Wang, H.; Jiang, S.; Chen, S.; Li, D.; Zhang, X.; Shao, W.;
Sun, X.; Xie, J.; Zhao, Z.; Zhang, Q.; Tian, Y.; Xie, Y. Adv. Mater.
2016, 28, 6940.
(11) Ma, Y.; Valkunas, L.; Dexheimer, S.; Bachilo, S.; Fleming, G.
Phys. Rev. Lett. 2005, 94, 157402.
(12) Hendry, E.; Schins, J.; Candeias, L.; Siebbeles, L.; Bonn, M.
Phys. Rev. Lett. 2004, 92, 196601.
(13) Wang, H.; Jiang, S.; Chen, S.; Zhang, X.; Shao, W.; Sun, X.;
Zhao, Z.; Zhang, Q.; Luo, Y.; Xie, Y. Chem. Sci. 2017, 8, 4087.
(14) Paquin, F.; Latini, G.; Sakowicz, M.; Karsenti, P.; Wang, L.;
Beljonne, D.; Stingelin, N.; Silva, C. Phys. Rev. Lett. 2011, 106,
197401.
(37) Zhu, H.; Yang, Y.; Wu, K.; Lian, T. Annu. Rev. Phys. Chem.
2016, 67, 259.
(38) Zhang, Q.; Luo, Y. High Power Laser Sci. Eng. 2016, 4, e22.
(39) Bi, W.; Li, X.; Zhang, L.; Jin, T.; Zhang, L.; Zhang, Q.; Luo,
Y.; Wu, C.; Xie, Y. Nat. Commun. 2015, 6, 8647.
(40) Yang, W.; Zhang, L.; Xie, J.; Zhang, X.; Liu, Q.; Yao, T.;
Wei, S.; Zhang, Q.; Xie, Y. Angew. Chem., Int. Ed. 2016, 55, 6716.
(41) Kovalev, D.; Fujii, M. Adv. Mater. 2005, 17, 2531.
(42) Schweitzer, C.; Schmidt, R. Chem. Rev. 2003, 103, 1685.
(43) Jiang, G.; Chen, J.; Huang, J.; Che, C. Org. Lett. 2009, 11,
4568.
(44) Zhang, N.; Li, X.; Ye, H.; Chen, S.; Ju, H.; Liu, D.; Lin, Y.;
Ye, W.; Wang, C.; Xu, Q.; Zhu, J.; Song, L.; Jiang, J.; Xiong, Y. J.
Am. Chem. Soc. 2016, 138, 8928.
(45) Ushakov, D. B.; Gilmore, K.; Kopetzki, D.; McQuade, D.
T.; Seeberger, P. H. Angew. Chem., Int. Ed. 2014, 53, 557.
(46) Su, C.; Acik, M.; Takai, K.; Lu, J.; Hao, S.‐j.; Zheng, Y.; Wu,
P.; Bao, Q.; Enoki, T.; Chabal, Y. J.; Ping Loh, K. Nat. Commun.
2012, 3, 1298.
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
1
2
3
4
5
Table of Contents
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment