Exploring the reactivity of multicomponent photocatalysts: Insight into the complex valence band of BiOBr

Article abstract of DOI:10.1002/chem.201202602

The band structure of multicomponent semiconductor photocatalysts, as well as their reactivity distinction under different wavelengths of light, is still unclear. BiOBr, which is a typical multicomponent semiconductor, may have two possible valence-band structures, that is, two discrete valence bands constructed respectively from O 2p and Br 4p orbitals, or one valence band derived from the hybridization of these orbitals. In this work, aqueous photocatalytic hydroxylation is applied as the probe reaction to investigate the nature and reactions of photogenerated holes in BiOBr. Three organic compounds (microcystin-LR, aniline, and benzoic acid) with different oxidation potentials were selected as substrates. Isotope labeling (H₂¹⁸O as the solvent) was used to determine the source of the O atom in the hydroxyl group of the products, which distinguishes the contribution of different hydroxylation pathways. Furthermore, a spin-trapping ESR method was used to quantify the reactive oxygen species (^.OH and ^.OOH) formed in the reaction system. The different isotope abundances of the hydroxyl O atom of the products formed, as well as the reverse trend of the ^.OH/^.OOH ratio with the oxidative resistance of the substrate under UV and visible irradiation, reveal that BiOBr has two separate valence bands, which have different oxidation ability and respond to UV and visible light, respectively. This study shows that the band structure of semiconductor photocatalysts can be reliably analyzed with an isotope labeling method. Which isotope won? Hydroxylation in the BiOBr photocatalytic system with H₂¹⁸O solvent and ¹⁶O₂ oxidant gives products with different isotope profiles under visible-light versus UV irradiation (see scheme). The isotope abundance is also affected by the oxidative resistance of the substrate. Together with the ^.OH/ ^.OOH ratio, the results show that BiOBr has two valence bands that have different oxidation abilities and responses to UV and visible light. Copyright

Full text of DOI:10.1002/chem.201202602

DOI: 10.1002/chem.201202602
Exploring the Reactivity of Multicomponent Photocatalysts: Insight into the
Complex Valence Band of BiOBr
[
a]
[a]
[a]
[b]
Yan-Fen Fang, Wan-Hong Ma, Ying-Ping Huang,* and Gen-Wei Cheng
Abstract: The band structure of multi-
component semiconductor photocata-
lysts, as well as their reactivity distinc-
tion under different wavelengths of
light, is still unclear. BiOBr, which is a
typical multicomponent semiconductor,
may have two possible valence-band
structures, that is, two discrete valence
bands constructed respectively from
O 2p and Br 4p orbitals, or one valence
band derived from the hybridization of
these orbitals. In this work, aqueous
photocatalytic hydroxylation is applied
as the probe reaction to investigate the
nature and reactions of photogenerated
holes in BiOBr. Three organic com-
pounds (microcystin-LR, aniline, and
benzoic acid) with different oxidation
potentials were selected as substrates.
COOH) formed in the reaction system.
The different isotope abundances of
the hydroxyl O atom of the products
formed, as well as the reverse trend of
the COH/COOH ratio with the oxidative
resistance of the substrate under UV
and visible irradiation, reveal that
BiOBr has two separate valence bands,
which have different oxidation ability
and respond to UV and visible light, re-
spectively. This study shows that the
band structure of semiconductor photo-
catalysts can be reliably analyzed with
an isotope labeling method.
1
8
Isotope labeling (H2 O as the solvent)
was used to determine the source of
the O atom in the hydroxyl group of
the products, which distinguishes the
contribution of different hydroxylation
pathways. Furthermore, a spin-trapping
ESR method was used to quantify the
reactive oxygen species (COH and
Keywords: ESR spectroscopy
hydroxylation · isotopic labeling ·
photocatalysis · semiconductors
·
AHCTUNGTRENNUNG
Introduction
by different wavelengths of light (O 2p to Bi 6p and Br 4p
+
to Bi 6p) and have h
with different oxidation potentials
vb
+
+
[4]
Since Asahi and co-workers reported in 2001 that N-doped
(hO2p and hBi6p ). Even though the catalyst has a band
structure of the second kind, the interband relaxation of the
[1]
TiO responds to visible light, the difference in oxidative
2
+
+
+
+
reactivity between visible- and UV-induced activation of
multicomponent semiconductor photocatalysts has remained
hvb (h
to hBr4p ) would also lead to a single hvb type
O2p
+
(hBr4p ) under both UV and visible irradiation (Figure 1),
which makes the determination of the band structure very
difficult.
[2]
unclear. Bismuth oxybromide (BiOBr), which has two
2
ꢀ
ꢀ
kinds of anions (O and Br ) and responds to both UV and
[3]
visible irradiation, is an ideal system to investigate the
band structure of multicomponent semiconductors and their
reactivity under different wavelength regions of irradiation.
BiOBr may have two plausible valence-band structures:
In photocatalytic reactions, COH is the most important
radical intermediate. It is formed by the oxidation of H O
2
1) one valence band derived from the hybridization of O 2p
and Br 4p orbitals, the activation of which by different
+
wavelengths of light would lead to a hole (hvb ) with the
same reactivity; and 2) two discrete valence bands construct-
ed respectively by O 2p and Br 4p orbitals, which are excited
[
a] Dr. Y.-F. Fang, Prof. W.-H. Ma, Prof. Y.-P. Huang
Engineering Research Centre of Eco-environment in
Three Gorges Reservoir Region, Ministry of Education
China Three Gorges University, Hubei 443002 (P.R. China)
Fax : (+86)717-639-7488
E-mail: huangyp@ctgu.edu.cn
[
b] Prof. G.-W. Cheng
Figure 1. Potentials (versus normal hydrogen electrode, NHE) for the
single-electron transfer of H O and O and substrates used in this investi-
2 2
gation, with corresponding reactions on the conduction band (CB) and
valence band (VB) of activated BiOBr. No assumptions are made con-
cerning the type of charge transfer involved. ROS=reactive oxygen spe-
cies, MC-LR=microcystin-LR.
Institute of Mountain Hazards and Environment of
Chinese Academy of Sciences
Chengdu 610041 (P.R. China)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202602.
3224
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3224 – 3229

FULL PAPER
+
by hvb and the reduction of O by the conduction-band
Results and Discussion
2
ꢀ
electron (ecb ; see Figure 1). The COH formed via these two
pathways attacks organic substrates identically. The COH
[5]
In our previous study on the photocatalytic oxidation of
formed in reaction systems is typically quantified with the
technique of spin-trapping electron spin resonance (ESR)
spectroscopy. However, this method cannot distinguish the
source of COH. Isotopic labeling provides a reliable method
of analysis of the source and formation pathway of COH,
MC-LR, epimeric products a and b (see Table 1), which are
6
7
formed from the 1,2-dihydroxylation of the C =C bond,
[8]
were found to be the main intermediates. Because two
O atoms are added, three combinations of the origin of
O atoms (both O atoms come from H O, both O atoms
2
that is, if the photocatalytic reaction is carried out in an
come from O , one O atom comes from H O and the other
2
2
1
8
16
16
18
oxygen isotope-labeled system (H2 O/ O or H O/ O ),
comes from O ) are possible for each product. These combi-
2
2
2
2
the analysis of the isotope abundance in the hydroxylated
product formed can determine the proportions of these two
pathways of COH formation.
It has been proposed that, in aqueous photocatalytic sys-
tems, the hydroxylation of CꢀH moieties follows three pos-
sible pathways: 1) direct COH addition to form an OH-
nations reflect the pathway of the reaction. We used the iso-
tope labeling method to analyze the origin of added
O atoms. The photocatalytic oxidation of MC-LR was car-
1
8
16
ried out in H2 O with O in air as oxidant. The isotope
2
abundance of the formed hydroxylated product was ana-
lyzed by HPLC–ESI-MS (Figure S1 in the Supporting Infor-
mation). Because the possibility of isotopic exchange be-
adduct radical, which is further oxidized to the product
[6a–c]
+
18
[Eq. (1)];
2) direct oxidation of substrate by hvb to form
tween the products and H2 O was eliminated by control ex-
a cationic radical, which undergoes hydrolysis to yield the
periments, the oxygen isotope abundance of hydroxylated
products clearly indicates the oxygen source. The measured
isotope abundance was corrected with the natural isotope
abundance to determine the origin of the hydroxyl O atoms
(details are shown in the Supporting Information), and the
results are displayed in Table 1. The reaction was also car-
[6b,c]
products [Eq. (2)];
and 3) reaction of molecular O with
2
the cationic radical of the substrate to form an O adduct,
2
which further decomposes to the hydroxylated product
[6a,d]
[Eq. (3)].
RH þ COH ! CRHOH !! ROH
ð1Þ
ð2Þ
ð3Þ
ried out in a TiO /UV system for comparison. When BiOBr
2
is activated by UV light, about half of the hydroxyl O atoms
þ
H2O
RH þ h
! RHCƒƒ! ! ROH
vb
come from O (Table 1, entry 2, 49.9% for product a and
2
+
ꢀ
and ecb
O2
59.0% for product b). These data indicate that hvb
þ
RHC ƒƒ!ROOC !! ROH
þ
ꢀ
H
contribute equally to the reaction. The isotope abundance
of the products formed in the TiO /UV system are nearly
2
Pathways (1) and (2) have been verified in various reac-
tion systems; on the other hand, no experimental evidence
corroborates that pathway (3) really works in photocatalysis.
The products formed through pathway (2) have the hydroxyl
identical to those formed by BiOBr/UV (Table 1, entry 3,
54.1% for product a and 47.7% for product b). With TiO2/
UV photocatalysis, it is known that the COH initiating path-
+
way (1) derives primarily from H O oxidation by h
(in
2
vb
O atom derived from H O. In contrast, the oxygen source of
products formed through pathway (1) was uncertain, be-
the valence band constructed with O 2p orbitals). However,
this is not necessarily the case with BiOBr/UV because
pathway (2), initiated by the direct oxidation of the sub-
strate, can also generate products with hydroxyl O atoms de-
2
cause COH can originate from either H O oxidation or O
2
2
(
H O ) reduction.
2 2
In this work, aqueous photocatalytic hydroxylation is ap-
rived from H O. In sharp contrast to UV irradiation, the ac-
2
plied as the probe reaction to investigate the nature and re-
actions of photogenerated hvb in BiOBr. Three organic
tivation of BiOBr by visible light yields products containing
+
primarily O -derived O atoms (Table 1, entry 1, 81.4% for
2
compounds (microcystin-LR (MC-LR), aniline, and benzoic
acid) with different redox potentials (Figure 1) were selected
as substrates. Isotopic labeling (with H2 O as the solvent)
was used to determine the source of the O atom in the hy-
droxyl group of the products. Reactive oxygen species (COH
and COOH) formed in the system were quantified by using
an in situ spin-trapping ESR method. Our results reveal that
BiOBr has two separate valence bands, the absorption edges
of which are in the UV and visible region, respectively. The
two valence-band holes generated have different oxidizing
ability and induce different reactions.
product a and 84.6% for product b). The difference between
UV and visible-light irradiation indicates separate valence
bands of BiOBr. Otherwise, the reactions induced by UV
[7]
18
+
and visible light would be the same; that is, the hvb gener-
ated by irradiation with different wavelengths of light would
all locate at the hybridized valence band and have the same
reactivity. The different isotope abundance of the products
formed in BiOBr/UV and BiOBr/Vis systems can be inter-
preted in two ways: 1) the production of COH from H O oxi-
2
dation or 2) the direct oxidation of substrate is significantly
inhibited when BiOBr is activated with visible instead of
UV light.
The difference in oxygen source for MC-LR hydroxyla-
tion with UV versus visible excitation should be correlated
+
with the oxidative reactivity of hvb in BiOBr. This assump-
tion is confirmed by the hydroxylation of other substrates
Chem. Eur. J. 2013, 19, 3224 – 3229
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3225

Y.-P. Huang et al.
6
7
18
16
2 2
Table 1. Isotopic abundances of the two oxygen atoms added to the C =C bond in the hydroxylation of MC-LR in H O and O photocatalytic sys-
ACHTUNGTRENNUNGt ems.
[
a]
Time
[min]
Conv.
[%]
Hydroxylation product a [%]
Hydroxylation product b [%]
16 16 16 18 18 18
Entry
Conditions
1
6
16
16 18
18 18
16
16
A
H
U
G
R
N
U
G
O O
O O
O O
Total
O
O O
O O
O O
Total O
[
b]
1
2
3
BiOBr/Vis
BiOBr/UV
30
5
5
19.4
52.9
79.6
63.7
3.5
8.4
35.4
92.7
91.4
0.95
3.8
0.15
81.4
49.9
54.1
71.1
12.8
5.3
26.9
82.4
84.7
2.0
4.8
9.9
84.6
59.0
47.7
[
b]
[
c]
TiO
2
/UV
ꢀ1
18
2
16
ꢀ1
ꢀ1
[
a] cMC-LR =5 mgL , H
O (1 mL), under aerated ( O
2
2
) conditions. [b] BiOBr, 2 gL . [c] P25 TiO , 2 gL .
+
that are more resistant to oxidation than MC-LR (aniline
and benzoic acid, see Table 2). According to Marcusꢁs
theory, the rate of electron transfer is related to the stand-
ard free-energy change of the reaction, that is, the higher
the oxidation potential of the substrate, the slower the oxi-
dation. When aniline and benzoic acid were used as sub-
(hO2p ) band with strong oxidation ability converts to the
+
Br 4p band and forms less oxidative h
relaxation of hvb would significantly reduce the difference
between the UV and visible irradiation systems, since in
. The interband
Br4p
[9]
+
+
both systems, hBr4p , formed from direct excitation (visible
system) or relaxation (UV system), initiates the oxidation of
substrates. Experimental data confirmed this ex-
pectation; under both UV and visible-light irradia-
Table 2. Isotopic abundances of the hydroxyl O atoms in the monohydroxylated prod-
1
8
[a]
tion, most of the hydroxyl O atoms (>90%) in the
2
ucts formed in H O-labeled photocatalytic oxidation of aromatic compounds.
hydroxylated products of aniline and benzoic acid
1
6
come from O (Table 2, entries 1, 2, 4, and 5). In
2
contrast, in the TiO /UV system, most of the hy-
2
droxyl O atoms in the hydroxylated products come
1
8
1
6
from H2 O (Table 2, entries 3 and 6). These results
indicate that in BiOBr systems under both UV and
Entry
Substrate
Conditions
Time
h]
Conv.
[%]
Abundance of O [%]
[
para
meta
ortho
+
visible-light irradiation, the hvb -induced oxidation
is of the same kind, and the oxidation ability is re-
1
2
3
4
5
6
7
8
9
BiOBr/Vis
BiOBr/UV
2.0
1.3
3.0
3.0
0.3
2.0
2.0
1.5
2.0
8.7
4.4
61.0
4.4
5.9
17.0
–
98.3
99.4
48.0
98.0
93.0
21.0
–
97.9
98.9
48.0
98.9
98.2
10.3
–
98.5
97.6
70.1
90.7
91.4
20.1
–
[
b]
aniline
+
TiO
2
/UV
markably different from that of the h
of TiO2.
vb
BiOBr/Vis
BiOBr/UV
TiO /UV
2
BiOBr/Vis
BiOBr/UV
16
The O abundances of the hydroxylated products
of aniline and benzoic acid are considerably higher
than that of MC-LR (Table 1). The close isotope
abundances of the hydroxylated products for the
BiOBr/UV and BiOBr/Vis systems when aniline or
[
c]
benzoic acid
[
[
[
e]
e]
f]
[d]
nitrobenzene
–
–
–
–
–
TiO
2
/UV
53.4
50.3
62.0
1
8
ꢀ1
[
a] H
aerated ( O
initial concentration of BA was 20 mm. [d] The initial concentration of NB was 50 mm.
e] No substrate conversion or product formation was observed. [f] The conversion of
2 2
O suspension (1 mL) containing 2 gL photocatalyst (BiOBr or TiO
), under benzoic acid was used as substrate are also in con-
1
6
2
) conditions. [b] The initial concentration of aniline was 50 mm. [c] The
trast to the oxidation of MC-LR. These differences
among MC-LR, aniline, and benzoic acid systems il-
[
+
lustrate that the hvb initiate the hydroxylation of
substrate was too low to be quantified.
substrates primarily through the direct oxidation of
strate, their direct oxidation was much slower than that of
MC-LR. If the photogenerated hvb oxidizes primarily the
the substrate [pathway (2)], rather than the oxidation of
+
H O to COH [initiating pathway (1)], otherwise the oxidative
2
substrate rather than H O (to COH), its lifetime in the ani-
line and benzoic acid systems would be much longer than
that in the MC-LR system. If BiOBr has separate valence
susceptibility of substrates cannot affect the oxygen source
for the hydroxylated product. If the oxidation potential of
2
the substrate is much higher than that of the valence band
+
bands, constructed with O 2p and Br 4p orbitals, respective-
of O 2p, the substrate cannot be oxidized by h . The oxi-
vb
+
+
+
ly, the extending of the lifetime of h
would make possible
dation of H O is also limited by the rapid h
!h
re-
vb
2
O2p
Br4p
+
+
+
the interband relaxation of hvb , that is, the hvb in the O 2p
laxation. Therefore, the reaction of hvb is completely im-
3226
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3224 – 3229

Reactivity of Multicomponent Photocatalysts
FULL PAPER
peded, and the catalytic cycle, including the reduction of O2
ꢀ
by ecb to COH, would shut down. Indeed, when nitroben-
0
zene (E =2.9 V vs. NHE) was used as the substrate
(
Table 2, entries 7 and 8), neither visible nor UV irradiation
induced the oxidation of the substrate. In contrast, even
+
though direct oxidation of nitrobenzene by the h
of TiO2
vb
[6b]
is also not viable (E =2.7 V), nitrobenzene can be oxi-
vb
dized in the TiO /UV system, forming hydroxylated prod-
2
1
8
ucts with approximately 50% O. This can be attributed to
the oxidation of H O by h
+
(cannot relax as in the case
2
O2p
of BiOBr) to produce COH, which induces the hydroxylation
of nitrobenzene (Table 2, entry 9).
To further explore the activation of H O and O in the
2
2
BiOBr system, spin-trapping ESR spectroscopy was used to
detect the radicals (specifically COH and COOH) formed
during the hydroxylation of MC-LR, aniline, and benzoic
acid (Figure 2). The typical signals of the trapped COH and
COOH were measured in these systems, but their intensity
altered with the light source, substrate, and irradiation time.
Aniline and benzoic acid, which are more difficult to oxi-
dize, gave weaker ESR intensities for both COH and COOH
relative to those for MC-LR. The COH is produced from two
sources, the reduction of O (H O ) and the oxidation of
2
2
2
H O, whereas COOH is formed exclusively from molecular
2
[
5a]
O2.
The ratio of COH to COOH indicates the relative
amounts of H O and O activated. Although the absolute
2
2
quantification of COH and COOH by ESR spectroscopy is
difficult, a comparison of the COH/COOH ratio for each sub-
strate reaction under both UV and visible-light irradiation is
realistic and significant for understanding the initial activa-
tion step on BiOBr (Figure 3).
During MC-LR hydroxylation by BiOBr under visible
light, the signal of trapped COOH was very weak and the
COH signal relatively strong (COH/COOH=8). However, a
significant decrease of the COH/COOH ratio (to 0.9 and 0.7,
respectively) occurred for both aniline and benzoic acid
under similar conditions, that is, the easier the oxidation of
the substrate, the larger the COH/COOH ratio. Considering
that H O is a key intermediate in O reduction, and that
2
2
2
COH derives primarily from the reduction of O (via H O ),
2
2
2
we attribute this phenomenon to the reduction of H O by
2
2
ꢀ
ecb . As noted above, a substrate more easily oxidized would
increase the rate of hvb capture, inhibit the hvb –ecb re-
+
+
ꢀ
combination, and then facilitate the reduction of H O to
2
2
ꢀ
COH by ecb . Therefore, the COH/COOH ratio would increase
with the oxidative susceptibility of the substrate.
Under UV irradiation, the distribution of radicals is con-
verse to that of the BiOBr/Vis system, that is, the COH/
COOH ratio increases dramatically with the substrate resist-
Figure 2. ESR signals of the dimethylpyrrolidine 1-oxide (DMPO)–COH
and DMPO–COOH adducts in aqueous solution (panels 1 and 3) and
method solution (panels 2 and 4). Panels 1 and 2 were under visible-light
irradiation and panels 3 and 4 were under UV-light irradiation in the sub-
strate/BiOBr system. A) MC-LR; B) aniline; C) benzoic acid. Reaction
ꢀ
1
ꢀ1
conditions: MC-LR, 3.0 mgmL
0
; BiOBr, 10 mgmL ; benzoic acid,
ꢀ1
.1 mm; aniline, 0.1 mm; DMPO, 0.4 molL
.
Chem. Eur. J. 2013, 19, 3224 – 3229
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3227

Y.-P. Huang et al.
positioned inside a cylindrical Pyrex reactor and surrounded by a circu-
lating-water jacket for cooling. To ensure illumination by only visible
light, a cutoff filter was placed outside the Pyrex jacket to completely
eliminate any radiation at wavelengths below 420 nm. Similarly, a UV-
3
65 filter was used to avoid direct photolysis from ultraviolet B and C ra-
diation. Prior to the irradiation, the suspensions were stirred magnetically
in the dark for approximately 2 h to ensure the establishment of an ad-
ꢀ
1
sorption/desorption equilibrium. Microcystin-LR (3.0 mgL , 10 mL; Ex-
press Technology Co., China) and catalyst powders (2 mg) were placed in
a Pyrex vessel. At a given time, the solution (300 mL) was collected, cen-
trifuged, and then filtered through a Millipore filter (pore size 0.22 mm)
to remove the solid catalyst particles. The solution pH was 6.25.
Analysis: The oxygen isotope abundance of the hydroxylated product
was analyzed by HPLC–ESI-MS (Agilent LC 1200/Ion Trap 6310) with a
C-18 column (250 mꢂ2.1 mm). Each measurement was repeated at least
three times to ensure accuracy. Figure S1 in the Supporting Information
gives the typical ESI-MS spectra (acquired by UV and MS detectors) of
the substrate. The measured isotope abundance of the product was cor-
Figure 3. Intensity ratio of COH/COOH versus redox potential of the sub-
strate, obtained by using ESR spectroscopy with DMPO spin trapping.
Reaction conditions are the same as in Figure 2.
rected with the oxygen isotope abundance of solvent H
2
O and the natural
isotope abundance of the product by use of Equations (4) and (5), in
ance to oxidation (COH/COOH is 0.18 for MC-LR, 0.37 for
aniline, and 0.99 for benzoic acid, Figure 3). The very small
COH/COOH ratio for MC-LR, relative to that of the visible-
1
8
which C
p n w
, C , and C are the O percentages of the measured isotope
abundance of the product, natural isotope abundance of the product, and
measured isotope abundance of solvent H O, respectively.
2
light system, is explained by the production of high-energy
+
+
hO2p in the BiOBr/UV system. As shown in Figure 1, hO2p
can oxidize H O (from O reduction) to COOH. This pro-
C ꢀ C
p
n
H
2
O% ¼
ꢁ 100
ð4Þ
ð5Þ
C
w
ꢀ C
n
2
2
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
cess would facilitate the formation of COOH, and inhibit the
formation of COH (mainly from the reduction of H O )
C
C
w
w
ꢀ C
ꢀ C
p
n
2
2
2
O % ¼
ꢁ 100
through competitive consumption of H O . As the substrate
2
2
becomes more difficult to oxidize, due to the relaxation of
EPR spin-trap analysis: A Brucker EPR spectrometer (model E500)
+
+
+
hO2p to hBr4p , the effect of hO2p on COH and COOH would
diminish. As a result, a higher COH/COOH ratio was meas-
ured in the systems of aniline and benzoic acid.
equipped with a Quanta-Ray Nd:YAG laser (355 and 532 nm) was used
ꢀ
for measuring the electron spin resonance (ESR) signals of
COH and O2C
spin-trapped by dimethylpyrrolidine 1-oxide (DMPO). To minimize ex-
perimental errors, the same quartz capillary tube was used for all ESR
measurements.
Conclusion
Hydroxylation reactions, photocatalyzed by BiOBr under
both UV and visible light, were investigated by using
O isotopic labeling. The correlation of COH/COOH ratio
Acknowledgements
18
We thank Dr. David M. Johnson of Ferrum College (Ferrum, VA) and
Dr. Li Yue of Nankai University (Tianjin, P.R. China) for their assistance
in revising the manuscript. This work was funded by the National Natural
Science Foundation of China (No. 21177072 and 21207079), the National
Basic Research Program of China (No. 2008CB417206) and the Youth
Scientific Research Funds (No. 65121003).
with substrate redox potential in both systems was measured
and analyzed. Our results show that BiOBr has two separate
valence bands, which are constructed with O 2p and Br 4p
orbitals, respectively, instead of a hybridized valence band.
These two valence bands respond respectively to UV and
visible irradiation and display complex reactivity. To the
best of our knowledge, this is the first example of the exis-
tence of separate valence bands in a semiconductor that has
been determined experimentally.
[
1] R. Asahi, T. Morikawa, T. Waki, K. Aoki, Y. Taga, Science 2001, 293,
269–271.
[2] a) C. Gomes Silva, R. Juꢃrez, T. Marino, R. Molinari, H. Garcia, J.
Am. Chem. Soc. 2011, 133, 595–602; b) H. Park, W. Y. Choi, J. Phys.
Chem. B 2005, 109, 11667–11674; c) X. H. Wang, J. G. Li, H. Ka-
miyama, Y. Moriyoshi, T. Ishigaki, J. Phys. Chem. B 2006, 110, 6804–
6
1
809; d) J. M. Herrmann, J. Disdier, P. Pichat, Chem. Phys. Lett. 1984,
08, 618–622.
Experimental Section
[
3] a) X. Chang, J. Huang, C. Cheng, Q. Sui, W. Sha, G. Ji, S. Deng, G.
Yu, Catal. Commun. 2010, 11, 460–464; b) J. Henle, P. Simon, A.
Frenzel, S. Scholz, S. Kaskel, Chem. Mater. 2007, 19, 366–373; c) J.
Zhang, F. J. Shi, J. Lin, D. F. Chen, J. M. Gao, Z. X. Huang, X. X.
Ding, C. Tang, Chem. Mater. 2008, 20, 2937–2941; d) Z. H. Ai, W. K.
Ho, S. C. Lee, L. Z. Zhang, Environ. Sci. Technol. 2009, 43, 4143–
2
ꢀ1
Materials: BiOBr (surface area, ca. 13 m g , JCPDS card no. 78-0348)
was prepared. Microcystin-LR (MC-LR) was purchased from Express
1
8
2
Technology Company. H O, from Jiangsu Changshu Chemical Limited,
had an initial isotopic enrichment of 85.6% as determined by mass spec-
trometry. Other chemicals were of reagent grade and used without fur-
ther purification. Deionized and doubly distilled water was used through-
out this study.
4
2
150; e) X. Zhang, Z. H. Zhi, F. Jia, L. Z. Zhang, J. Phys. Chem. C
008, 112, 747–753.
[
4] W. L. Huang, Q. Zhu, J. Comput. Chem. 2009, 30, 183–190.
Photocatalysis: The light sources used were a 500 W halogen lamp and a
[5] a) M. R. Hoffmann, S. T. Martin, W. Choi, D. W. Bahnemann, Chem.
Rev. 1995, 95, 69–96; b) M. A. Fox, M. T. Dulay, Chem. Rev. 1993,
1
00 W mercury lamp (Institute of Electric Light Source, Beijing, China)
3228
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3224 – 3229

Reactivity of Multicomponent Photocatalysts
FULL PAPER
9
3, 341–357; c) C. Chen, W. Ma, J. Zhao, Chem. Soc. Rev. 2010, 39,
sen, G. Merenyi, J. Phys. Chem. 1993, 97, 11278–11282; d) C.
Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
[8] F. Yanfen, H. Yingping, Y. Jing, W. Pan, C. Genwei, Environ. Sci.
Technol. 2011, 45, 1593–1600.
4
206–4219.
[
6] a) T. D. Bui, A. Kimura, S. Ikeda, M. Matsumura, J. Am. Chem. Soc.
010, 132, 8453–8458; b) Y. Li, B. Wen, C. Yu, C. Chen, H. Ji, W.
2
Ma, J. Zhao, Chem. Eur. J. 2012, 18, 2030–2039; c) S. Goldstein, G.
Czapski, J. Rabani, J. Phys. Chem. 1994, 98, 6586–6591; d) G. A. Rus-
sell, J. Am. Chem. Soc. 1957, 79, 3871–3877.
[9] a) R. A. Marcus, J. Phys. Chem. C 2009, 113, 14598–14608; b) P. Ke-
barle, S. Chowdhury, Chem. Rev. 1987, 87, 513–534.
[
7] a) T. Tachikawa, M. Fujitsuka, T. Majima, J. Phys. Chem. C 2007, 111,
Received: July 24, 2012
Revised: November 8, 2012
Published online: January 16, 2013
5
259–5275; b) A. V. Humble, G. M. Gadd, G. A. Codd, Water Res.
997, 31, 1679–1686; c) M. Jonsson, J. Lind, T. Reitberger, T. E. Erik-
1
Chem. Eur. J. 2013, 19, 3224 – 3229
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3229