Energy-level matching of Fe(III) ions grafted at surface and doped in bulk for efficient visible-light photocatalysts

Article abstract of DOI:10.1021/ja401541k

Photocatalytic reaction rate (R) is determined by the multiplication of light absorption capability (α) and quantum efficiency (QE); however, these two parameters generally have trade-off relations. Thus, increasing α without decreasing QE remains a challenging issue for developing efficient photocatalysts with high R. Herein, using Fe(III) ions grafted Fe(III) doped TiO₂ as a model system, we present a novel method for developing visible-light photocatalysts with efficient R, utilizing the concept of energy level matching between surface-grafted Fe(III) ions as co-catalysts and bulk-doped Fe(III) ions as visible-light absorbers. Photogenerated electrons in the doped Fe(III) states under visible-light efficiently transfer to the surface grafted Fe(III) ions co-catalysts, as the doped Fe(III) ions in bulk produced energy levels below the conduction band of TiO₂, which match well with the potential of Fe³⁺/Fe²⁺ redox couple in the surface grafted Fe(III) ions. Electrons in the surface grafted Fe(III) ions efficiently cause multielectron reduction of adsorbed oxygen molecules to achieve high QE value. Consequently, the present Fe(III)-Fe_xTi _1-xO₂ nanocomposites exhibited the highest visible-light R among the previously reported photocatalysts for decomposition of gaseous organic compounds. The high R can proceed even under commercial white-light emission diode irradiation and is very stable for long-term use, making it practically useful. Further, this efficient method could be applied in other wide-band gap semiconductors, including ZnO or SrTiO₃, and may be potentially applicable for other photocatalysis systems, such as water splitting, CO₂ reduction, NO_x removal, and dye decomposition. Thus, this method represents a strategic approach to develop new visible-light active photocatalysts for practical uses.

Full text of DOI:10.1021/ja401541k

Subscriber access provided by MOSCOW STATE UNIV
Article
Energy-level matching of Fe(III) ions grafted at surface
and doped in bulk for efficient visible-light photocatalysts
Min Liu, Xiaoqing Qiu, Masahiro Miyauchi, and Kazuhito Hashimoto
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja401541k • Publication Date (Web): 17 Jun 2013
Downloaded from http://pubs.acs.org on June 20, 2013
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 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Energyꢀlevel matching of Fe(III) ions grafted at surface and
doped in bulk for efficient visibleꢀlight photocatalysts
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
‡
‡
,†,∥
,‡,§
Min Liu, Xiaoqing Qiu, Masahiro Miyauchi,* and Kazuhito Hashimoto*
†
Department of Metallurgy and Ceramics Science, Graduate School of Science and
Engineering, Tokyo Institute of Technology, 2ꢀ12ꢀ1 Ookayama, Meguroꢀku, Tokyo 152ꢀ8552,
Japan.
‡
§
∥
Research Center for Advanced Science and Technology, The University of Tokyo, 4ꢀ6ꢀ1
Komaba, Meguroꢀku, Tokyo 153ꢀ8904, Japan.
Graduate School of Engineering, The University of Tokyo, 7ꢀ3ꢀ1 Hongo, Bunkyoꢀku, Tokyo
1
13ꢀ8656, Japan.
Japan Science and Technology Agency (JST), PRESTO, 4ꢀ1ꢀ8 Honcho Kawaguchi, Saitama
32ꢀ0012, Japan.
3
Email: mmiyauchi@ceram.titech.ac.jp, hashimoto@light.t.uꢀtokyo.ac.jp
Abstract
Photocatalytic reaction rate (R) is determined by the multiplication of light absorption
capability (α) and quantum efficiency (QE); however, these two parameters generally have
tradeꢀoff relations. Thus, increasing α without decreasing QE remains a challenging issue for
developing efficient photocatalysts with high R. Herein, using Fe(III) ions grafted Fe(III)
2
doped TiO as a model system, we present a novel method for developing visibleꢀlight
photocatalysts with efficient R, utilizing the concept of energy level matching between
surfaceꢀgrafted Fe(III) ions as coꢀcatalysts and bulkꢀdoped Fe(III) ions as visible light
absorbers. Photogenerated electrons in the doped Fe(III) states under visible light efficiently
transfer to the surface grafted Fe(III) ions coꢀcatalysts, as the doped Fe(III) ions in bulk
produced energy levels below the conduction band of TiO , which match well with the
2
3
+
2+
potential of Fe /Fe redox couple in the surface grafted Fe(III) ions. Electrons in the surface
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
grafted Fe(III) ions efficiently cause multiꢀelectron reduction of adsorbed oxygen molecules
x 2
to achieve high QE value. Consequently, the present Fe(III)ꢀFe Ti1ꢀxO nanocomposites
exhibited the highest visible light R among the previously reported photocatalysts for
decomposition of gaseous organic compounds. The high R can proceed even under
commercial white light emission diode (LED) irradiation and is very stable for longꢀterm use,
making it practically useful. Further, this efficient method could be applied in other
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
wideꢀbandꢀgap semiconductors, including ZnO or SrTiO , and may be potentially applicable
for other photocatalysis systems, such as water splitting and, CO reduction, NO removal,
2
x
and dye decomposition. Thus, this method represents a strategic approach to develop new
visibleꢀlightꢀactive photocatalysts for practical uses.
1. Introduction
Heterogeneous photocatalysis using semiconductors has great potential for solving current
1ꢀ20
energy and environmental issues.
Efficient photocatalysts are typically wideꢀbandꢀgap
3
, ZnO, and SrTiO , owing to the high redox potential of
semiconductors, such as TiO
2
1
0
photogenerated charge carriers. Holes with high oxidation power in the valence band (VB)
and electrons with sufficient reduction power in the conduction band (CB) are generally
required for efficient photocatalytic reactions. However, wideꢀbandꢀgap semiconductors are
only activated under ultraviolet (UV) light irradiation, which limits their practical
applications.
The doping of various transition metal cations or anions into wideꢀbandꢀgap
semiconductors has been extensively studied to increase the visible light absorption of these
1ꢀ10
photocatalysts.
However, despite extensive research efforts, most systems remain
unsatisfactory for practical use. In particular, metalꢀion dopants introduce deep impurity
levels in the forbidden band of semiconductor photocatalysts, where they act as
1
ꢀ4
recombination centers and impair photocatalytic activity. In the case of anionꢀdoped
semiconductors, isolated states are formed above the VB and cause the quantum efficiency
(
QE) of the semiconductors to deteriorate, as the holes generated in these isolated states have
2
ACS Paragon Plus Environment

Page 3 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1,2,6ꢀ9
lower oxidation power and mobility than those in the VB.
For example, the QE of
nitrogenꢀdoped TiO
2
under visible light is markedly lower than that of pure TiO under UV
2
6
,9
light. These previous studies indicate that it is difficult to improve the visibleꢀlight
absorption of semiconductors while maintaining a high QE value, because the reactivities of
photogenerated charge carriers in doped levels or narrowed bands are much less than those in
the VB and CB.
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
Very recently, our group demonstrated that the surface modification of TiO
2
with
2
1ꢀ30
coꢀcatalysts such as Cu(II) and Fe(III) ions
induces the efficient interfacial charge
31ꢀ35
transfer
of VB electrons upon visibleꢀlight irradiation and multiꢀelectron reduction
36ꢀ44
reactions of oxygen,
during which the excited electrons are consumed. Coꢀcatalysts also
2
5ꢀ30
3+
improve the visibleꢀlight activities of doped semiconductors.
For example, Ti selfꢀdoped
2
TiO , which is inactive even under UV light, functioned as an efficient visibleꢀlight
2
8
photocatalyst upon the surface grafting of Cu(II) or Fe(III) coꢀcatalysts. These results
indicate that a charge transfer occurs between photogenerated electrons from doped levels to
the surface Cu(II) or Fe(III) coꢀcatalysts. However, the reported QEs of these doped TiO2
2
photocatalysts were markedly lower than that of Cu(II)ꢀ or Fe(III)ꢀgrafted undoped TiO , as
the charge transfer is limited from doped levels to surface coꢀcatalysts. But, the visibleꢀlight
absorption capacity of Cu(II)ꢀ or Fe(III)ꢀgrafted undoped TiO is relatively weak, as
2
interfacial charge transfer occurs only at interfaces. Thus, such systems still exhibits low
photocatalytic activities, which implies that it is difficult to improve QE in systems with high
visibleꢀlight absorption. As a tradeꢀoff relation exists between visibleꢀlight absorption and QE,
it is difficult to obtain large reaction rates (Rs) under visibleꢀlight irradiation for
wideꢀbandꢀgap semiconductors.
Herein, we developed an efficient visibleꢀlight photocatalyst based on matching the energy
levels of surfaceꢀgrafted and bulkꢀdoped ions with similar energy states. We found that doping
2
Fe(III) ions in bulk introduces energy levels in the band gap of TiO that are similar to the
redox potential of surface Fe(III) ions. The bulkꢀdoped Fe(III) ions increase the visibleꢀlight
absorption of TiO , while the surfaceꢀgrafted Fe(III) ions help maintain the high QE of this
2
photocatalyst. Thus, efficient visibleꢀlight photocatalysts with high Rs were obtained using
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
this approach. A high R was even achieved under white light emitting diode (LED) irradiation,
2
and R values comparable to those of the best commercial TiO photocatalysts, such as Pꢀ25
(
Degussa), were obtained under UV light irradiation.
2. Experimental Section
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.1 Synthesis of Fe
x
Ti1ꢀx
O
2
Samples. The Fe(III)ꢀdoped TiO
2
(Fe
x
Ti1ꢀx
2
O ) nanocomposites
were prepared using a simple impregnation technique with commercial TiO
2
(rutile phase,
2
15ꢀnm grain size, 90 m /g specific surface area; MTꢀ150A, Tayca Co.) as the starting material.
Briefly, 1.5 g TiO
FeCl ꢁ6H O (Wako, 99.9%), acting as the source of Fe(III), was weighed to give a weight
fraction of Fe relative to TiO of 0.1% and was added to the TiO suspension, which was then
stirred for 0.5 h in a vial reactor. The suspension was dried under a room temperature and the
obtained residue was further heated at 950 °C for 3 h to form Fe . The calcined
Fe was treated with a 6 M HCl aqueous solution at 90 °C for 3 h under stirring. The
products were filtered twice through a membrane filter (0.025 ꢂm, Millipore) and then
washed with sufficient amounts of distilled water. Fe was obtained as a clear powder
and dried at 110 °C for 24 h. Subsequently, the obtained Fe was ground into a fine
powder using an agate mortar and pestle for the preparation of Fe(III)ꢀFe Ti O
2 2
powder was dispersed in 10 mL ethanol to form a TiO suspension.
3
2
2
2
x
Ti1ꢀxO
2
x
Ti1ꢀxO
2
x
Ti1ꢀxO
2
x
Ti1ꢀxO
2
x
1ꢀx
2
nanocomposites. Pure TiO
without adding FeCl solution, and was used to prepare Fe(III)ꢀTiO
.2 Modification of Fe
2
was obtained using the same annealing and acid treatment process
3
2
nanocomposites.
2
x
Ti1ꢀx
O
2
Samples with Fe(III) Ions. The grafting of Fe(III) ions
23
onto Fe Ti O was performed using an impregnation method, as reported previously.
x
1ꢀx
2
Briefly, 1 g Fe
9.9%) was weighed to give a weight fraction of Fe relative to Fe
then added to the aqueous Fe Ti O suspension. The suspension was heated at 90 °C, stirred
x
Ti1ꢀx
O
2
powder was first dispersed in 10 mL distilled water. FeCl
3 2
ꢁ6H O (Wako,
9
x 2
Ti1ꢀxO
of 0.1% and was
x
1ꢀx
2
for 1 h in a vial reactor, and then filtered twice with a membrane filter (0.025 ꢂm, Millipore)
and washed with sufficient amounts of distilled water. The resulting residue was dried at
110 °C for 24 h and subsequently ground into fine powder using an agate mortar and pestle.
Fe(III)ꢀTiO
2
was also prepared by the same impregnation method.
4
ACS Paragon Plus Environment

Page 5 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
2
.3 Sample Characterizations. The structural characteristics of the prepared samples were
measured by powder Xꢀray diffraction (XRD) at room temperature on a Rigaku
D/MAX25000 diffractometer with a copper target (λ=1.54056 Å). Electron spin resonance
(ESR) spectra were recorded on a Bruker ESP350E spectrometer. Elemental analyses of the
samples were performed using an inductively coupled plasmaꢀatomic emission spectrometer
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
5
7
(
ICPꢀAES, Pꢀ4010, Hitachi). Fe Mössbauer spectra were measured at room temperature in a
5
7
transmission geometry using a Co/Rh source. UV−visible absorption spectra were obtained
by the diffuse reflection method using a spectrometer (UVꢀ2550, Shimadzu). The
x 2
morphologies of the prepared Fe Ti1ꢀxO nanocomposites were investigated by scanning
electron microscopy (SEM) using a Hitachi SUꢀ8000 apparatus and transition electron
microscopy (TEM) on a Hitachi HFꢀ2000 instrument using an acceleration voltage of 200 kV.
Surface compositions were studied by Xꢀray photoelectron spectroscopy (XPS; model 5600,
PerkinꢀElmer). The binding energy data were calibrated with reference to the C 1s signal at
284.5 eV.
2
.4 Evaluation of Photocatalytic Properties. The photocatalytic activities of
photocatalysts were evaluated by the decomposition of gaseous 2ꢀpropanol (IPA) under
visible light, UV light, and white LED illumination. An Xe lamp (LAꢀ251Xe, Hayashi Tokei)
equipped with Lꢀ42, Bꢀ47, and Cꢀ40C glass filters (Asahi TechnoꢀGlass) or a Dꢀ36A glass
2
filter (Asahi TechnoꢀGlass) was used as a source of visible light (420−530 nm, 1 mW/cm )
2
and UV light (320ꢀ400 nm, 1 mW/cm ), respectively. A commercial white LED (LDA 7A, 7.2
W; Toshiba, Co. Ltd.) located 15 cm from the sample was used as an indoor light source (1
2
mW/cm , ~300 lux). The light intensity was measured using a light radiometer (USRꢀ45D,
2
Ushio Co.) and was then adjusted to 1 mW/cm . A 500ꢀmL cylindrical glass vessel was used
as the photocatalysis reactor. To perform the photocatalytic experiments, 300 mg
photocatalyst powder was first evenly spread on the bottom of a circular glass dish (area of
2
5
.5 cm ) that was mounted in the middle of the vessel reactor. The vessel was sealed with a
rubber Oꢀring and a quartz cover, evacuated, and filled with fresh synthetic air. To eliminate
organic contaminants on the sample surface, the vessel was illuminated with a Xe lamp
(
LAꢀ251Xe) until the CO
2
generation rate was less than 0.02 ꢃmol/day. The vessel was then
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
evacuated and refilled with fresh synthetic air. The pressure inside the vessel was kept at ~1
atm. Next, 300 ppmv (~6 ꢂmol) of gaseous IPA was injected into the vessel. Prior to light
irradiation, the vessel was kept in the dark for 12 h to achieve the absorption/desorption
equilibrium of IPA on the photocatalyst surfaces. The IPA concentration first decreased and
then remained constant, demonstrating the absorption/desorption equilibrium of IPA had been
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
reached. During the equilibration process, no acetone or CO were detected under dark
conditions, demonstrating that the IPA molecules were not decomposed by the photocatalysts
under dark conditions. The vessel was then irradiated with light, and 1 mL gaseous samples
were periodically extracted from the reaction vessel to measure the concentrations of IPA,
acetone, and CO
2.5 Density of states (DOS) calculation. The planeꢀwaveꢀbased density functional theory
PWꢀDFT) calculations were performed for TiO and Fe using the ab initio
totalꢀenergy and the Vienna Ab initio Simulation Package (VASP) molecular dynamics
2
using a gas chromatograph (model GCꢀ8A, Shimadzu Co., Ltd.).
(
2
x 2
Ti1ꢀxO
4
5,46
47,48
program.
The PerdewꢀBurkeꢀErnzerhof (PBE) functional
of generalized gradient
approximation (GGA) level was employed together with the pseudopotentials, which reduce
49
the number of PWs to lower the maximum kinetic energy of the PW. A 48ꢀatom 2 × 2 × 2
super cell was used for the calculation, in which a Ti atom was replaced with a Fe atom at the
center of the super cell. The crystal model is shown in Figure S1. The moderate rate of
substitution of Ti sites at the center of the super cell (6.25%) helped to reduce distortion of the
5
0
tetragonal lattice. The core orbitals were replaced by ultrasoft pseudopotentials with a
kinetic energy cutoff of 500 eV. The density of the MonkhorstꢀPack kꢀpoint sampling was 5 ×
5
2
× 3. The rutile phase of TiO was used for the DOS calculation.
3. Results and Discussion
The DOSs of rutile TiO and Fe Ti O were investigated using the planeꢀwave density
2
x
1ꢀx
2
functional theory program VASP. Figure 1a shows the DOS of pure TiO
2
. The VB edge of
51
TiO
2
consists mainly of O2p states, whereas the CB edge has predominantly Ti 3d character.
It should be noted that the calculated band gap is ~1.8 eV, which is markedly smaller than the
52
experimental band gap of 3.0 eV due to wellꢀknown GGA error. As illustrated in Figure 1b,
6
ACS Paragon Plus Environment

Page 7 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
53
when Fe(III) ions were doped into TiO , the Fe 3d orbitals split into two bands. The upper
2
band was hybridized with the CB and the interband was located 0.3ꢀ 0.5 eV below the CB of
TiO
2
2
, a finding that is consistent with previous theoretical calculation results for Fe doped
53
TiO
. Considering the narrowing of the calculated value, the energy states of these interband
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
3
+
2+
0
54
levels are close to the redox potential of Fe /Fe (E =0.771 V, vs SHE, pH=0) in
surfaceꢀgrafted Fe(III) ions, demonstrating that surfaceꢀgrafted and bulkꢀdoped Fe(III) have
nearly identical energy levels.
Fe Ti O photocatalysts were prepared from commercial rutile TiO and FeCl ꢁ6H O using
x
1ꢀx
2
2
3
2
x 2
a simple impregnation method and heat treatment. The grafting of Fe(III) ions onto Fe Ti1ꢀxO
23
was performed by a similar impregnation method in a vial reactor. Using ICPꢀAES, the total
amount of Fe in the obtained samples was found to be nearly equal to the initial value used in
the preparation process (Table S1). Upon surface grafting and bulk doping of Fe(III), the color
of the obtained powders changed from white to yellow (Figure S2), indicating that the
prepared photocatalyst was visibleꢀlight sensitive. Diffuse reflectance spectroscopy
measurements clearly showed that surfaceꢀgrafted Fe(III) and bulkꢀdoped Fe(III) similarily
increased visibleꢀlight absorption from 400 to 600 nm (Figure 2a). The difference absorption
spectra of Fe(III)ꢀTiO
2 x 2 x 2 2
, Fe Ti1ꢀxO , and Fe(III)ꢀFe Ti1ꢀxO versus bare TiO indicated that the
light absorption of Fe(III)ꢀFe Ti O was mainly attributed to the bulkꢀdoped Fe(III) (Figure
x
1ꢀx
2
2
x 2 2
b). More importantly, the visibleꢀlight absorption peaks for Fe Ti1ꢀxO and Fe(III)ꢀTiO both
appeared around 420 nm. Taken together, these results indicate that the energy states of the
surfaceꢀgrafted and bulkꢀdoped Fe(III) ions have similar energy potential.
2
XRD patterns revealed that the prepared samples maintained a rutile TiO crystalline
structure (JCPDS card No. 21ꢀ1276) after Fe(III) doping or surface Fe(III) grafting (Figure 3
and Figure S3), indicating that neither surfaceꢀgrafted nor bulkꢀdoped Fe(III) ions affect the
crystal phase of TiO . The sharp XRD peaks indicate that the products were highly
2
55
crystallized with approximate particle sizes of 130 nm according to Scherrer’s equation .
ESR analysis revealed Fe
x
Ti1ꢀx
O
2
had strong Fe(III) signals (Figure S4a), and the calculated
19
density of Fe(III) ions was 1.1×10 per gram (/g), consistent with the amount of starting
3+
2
Fe(III) source material (0.1 wt%). When Fe(III) ions were doped into TiO , a very weak Ti
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3+
signal was observed in the ESR spectrum (Figure S4b). These Ti species originated in
56
oxygen vacancies and maintain the charge balance in Fe(III)ꢀdoped TiO
2
.
However, the
3
+
14
carrier density of Ti species was only 6.8×10 (/g), which was several orders of magnitude
19
x 2
lower than that of Fe(III) (1.1×10 (/g)). Therefore, the visibleꢀlight absorption by Fe Ti1ꢀxO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
is mainly due to the unoccupied Fe(III) species formed below the CB of TiO2.
x 2
SEM images (Figure S5) revealed that all of the prepared Fe(III)ꢀFe Ti1ꢀxO photocatalysts
were assemblies of uniformly distributed nanoparticles. Introduction of Fe(III) ions on the
TiO surface as well as into the TiO lattice did not change the morphology or particle size of
2
2
the obtained Fe(III)ꢀFe
x
Ti1ꢀx
O
2
samples. The average grain size of the nanoparticles was
approximately 200 nm, which is close to the estimated crystallite size, confirming that the
samples were highly crystallized.
XPS spectra were recorded to determine the surface composition and chemical states of the
surface elements (Figure 4 and Figure S6). In the Ti 2p coreꢀlevel spectra of the samples
57
(Figure 4a), no obvious differences could be seen in the chemical states of element Ti,
demonstrating that neither the surface grafting nor bulk doping of Fe(III) ions affected the
bonding structure between titanium and oxygen. In the Fe 2p coreꢀlevel spectra (Figure 4b),
Fe signals were only observed in Fe(III)ꢀgrafted samples, such as Fe(III)ꢀTiO
2
and
Fe(III)ꢀFe Ti O , confirming that Fe(III) was successfully grafted on the surface of TiO and
x
1ꢀx
2
2
Fe
x
Ti1ꢀx
O
2
. Fe signals were also detected in wideꢀscanned XPS spectra (Figure S6) of
and Fe(III)ꢀFe . According to our previous Xꢀray absorption fine structure
Fe(III)ꢀTiO
2
x 2
Ti1ꢀxO
2
3
(XAFS) measurements and XPS analysis of the Fe(III)ꢀTiO system, these signals can be
2
assigned to iron ions with an oxidation number of three. Based on these analyses of local
crystal structure, Fe(III) ions were grafted as distorted amorphous FeOOHꢀlike structures. In
addition, the chemical state and environment of surface Fe(III) ions in the present Fe(III)ꢀTiO
2
and Fe(III)ꢀFe Ti O samples are similar to those of the previous Fe(III)ꢀTiO system.
x
1ꢀx
2
2
Notably, no Fe(III) signal was detected for Fe
x
Ti1ꢀx
O
2
, as Fe(III) ions were doped in bulk and
the intensity of surfaceꢀdoped ions was below the detection limit of the XPS analysis.
To explore the states of surfaceꢀgrafted and bulkꢀdoped Fe(III) in more detail, Mössbauer
spectra and TEM images were recorded. The nuclear ground and excited levels involved in
8
ACS Paragon Plus Environment

Page 9 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
the Mössbauer transition were shifted or split because of the electrostatic interactions between
58
the nuclear charge and the surrounding electric charge. As shown in Figure 5a, clear signals
were seen in Fe(III)ꢀTiO
for pure TiO . Both Fe
shifts of around 0.38 mm/sec, indicating that both the bulkꢀdoped and surfaceꢀgrafted Fe(III)
2
, Fe
x
Ti1ꢀx
O
2
x 2
, and Fe(III)ꢀFe Ti1ꢀxO , whereas no signal was observed
2
x
Ti1ꢀx
O
2
and Fe(III)ꢀTiO exhibited quadrupole splitting and had isomer
2
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
5
9ꢀ63
ions are in paramagnetic states with a quadrupole doublet as an octahedral structure.
spectral peaks of Fe(III)ꢀFe could be perfectly fitted by the simple addition of spectra
for Fe Ti O and Fe(III)ꢀTiO in equal proportion. It is noteworthy that the quadrupole shifts
The
x
Ti1ꢀxO
2
x
1ꢀx
2
2
for the surfaceꢀgrafted and bulkꢀdoped Fe(III) ions were different. The quadrupole split for
Fe(III)ꢀdoped TiO was 0.51 mm/sec, whereas that for Fe(III)ꢀgrafted TiO was 0.77 mm/sec,
indicating that surface Fe(III) ions have a larger structural degree of freedom as compared to
2
2
5
9ꢀ63
4+
doped Fe(III) ions.
which is as large as that of Fe (0.064 nm).
are substitutionally introduced into TiO crystal at Ti sites,
Fe(III) ions exist as amorphous nanoclusters, consistent with our previous study of
Further, the ionic radius of Ti with six coodination is 0.061 nm,
3
+
2,3,53
These results indicate that doped Fe(III) ions
4
+
59ꢀ63
2
whereas surfaceꢀgrafted
23
Fe(III)ꢀTiO
nanoclusters of ~2 nm in size were well dispersed on the surface of Fe
observed good attachment of Fe(III) nanoclusters on the Fe Ti O surfaces. The clear lattice
2
.
TEM images (Figure 5b and Figure S7) confirmed that amorphous Fe(III)
x
Ti1ꢀxO
2
. In addition, we
x
1ꢀx
2
x 2
fringes of the nanoparticles demonstrated that the Fe Ti1ꢀxO was highly crystallized.
The photocatalytic activities of the prepared photocatalysts were evaluated by the
decomposition of gaseous IPA under visibleꢀlight irradiation. IPA was selected as a
representative volatile organic compound (VOC) as it is reported to be a serious pollutant of
6
4
2
indoor air. The oxidation of IPA to CO and its usefulness for the determination of QE have
6
4
been comprehensively established. The wavelength of the visible light source used in the
2
analysis was 400ꢀ530 nm, and the light intensity was 1 mW/cm (Figure S8), which
corresponds to an illuminance of only 300 lux and is comparable to the intensity of white
fluorescent lights and white LED lights. For the performance tests, the initial concentration of
IPA was 300 ppmv (~6 µmol), which is much higher than the VOC concentrations typically
encountered in indoor environments. Under these conditions, the complete decomposition of
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
IPA would result in a CO
2
concentration of 900 ppmv (~18 µmol), which is three times of
IPA concentration. A representative time course of the gas concentrations during the
decomposition of IPA by the Fe(III)ꢀFe
Figure 6a shows CO evolution from IPA decomposition by Fe(III)ꢀgrafted TiO
with various metal ions. Among the examined metal dopants, Fe(III)ꢀFe Ti O
2
x 2
Ti1ꢀxO sample is shown in Figure S9.
2
2
doped
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
x
1ꢀx
nanocomposites exhibited the best performance, completely decomposing gaseous IPA to
CO within 30 h. The CO generation rate achieved by Fe(III)ꢀFe was 0.69 ꢂmol/h
Figure S10), which is the highest R among reported visibleꢀlight photocatalysts. The high
2
2
x 2
Ti1ꢀxO
(
visibleꢀlight activity of Fe(III)ꢀFe
x
Ti1ꢀx
O
2
is due to the similar energy levels of surface Fe(III)
, which is recognized as one of the most efficient
and bulk Fe(III). Comparatively, TiO2ꢀx
N
x
6
visibleꢀlight photocatalysts, exhibited rather low activity, 0.16 ꢂmol CO /h, and required
2
over 300 h to completely decompose the gaseous IPA (Figure S11). The low activity of this
system is attributable to the markedly lower oxidation power of the photogenerated holes in
9
the nitrogen levels than those in the VB.
The importance of energy level matching between surfaceꢀgrafted and bulkꢀdoped Fe(III)
ions was further demonstrated by comparative studies of the photocatalytic activities of bare
x 2 2 x 2
Fe Ti1ꢀxO , Fe(III)ꢀTiO , and Fe(III)ꢀFe Ti1ꢀxO (Figure 6b). After doping Fe(III) ions in bulk,
Fe Ti O became sensitive to visible light, but exhibited relatively low R and QE values
x
1ꢀx
2
(
Table 1). Fe(III)ꢀTiO
2
also exhibited significant visibleꢀlight activity, as the surfaceꢀgrafted
Fe(III) ions acted as a coꢀcatalyst and effectively consumed the photogenerated electrons
3
1ꢀ44
through efficient interfacial charge transfer and multiꢀelectron reduction reactions,
resulting in a high QE (53.5%). However, the R of Fe(III)ꢀTiO remained low due to the
2
limited visibleꢀlight absorption by this photocatalyst (Figure 2 and Table 1). In contrast, after
the selected surface grafting and bulk doping of Fe(III) ions to achieve closely matched
energy levels, the Fe(III)ꢀFe Ti O nanocomposites demonstrated strong visibleꢀlight
x
1ꢀx
2
absorption with a QE of 47.3% as a result of efficient light absorption by bulkꢀdoped Fe(III)
and electron transfer between the doped and surfaceꢀgrafted Fe(III) ions, as well as efficient
23,36ꢀ44
multiꢀelectron reduction on the surface Fe(III) nanoclusters.
2
properties, the Fe(III)ꢀFe Ti1ꢀxO nanocomposites exhibited a high R.
Due to these excellent
x
10
ACS Paragon Plus Environment

Page 11 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
We next examined the visibleꢀlight activity of Fe(III)ꢀFe Ti O nanocomposites under
x
1ꢀx
2
conditions mimicking actual indoor environments. White LED, which is widely used as an
indoor light source, does not emit UV light (Figure S12) can therefore be used to examine the
true visibleꢀlight activity of photocatalysts. Under white LED irradiation with a light intensity
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
2
of 1 mW/cm , the Fe(III)ꢀFe Ti O nanocomposites completely decomposed IPA to CO
x
1ꢀx
2
2
(
Figure S13). We also evaluated the photocatalytic activity of commercial Pꢀ25 TiO
2
, which is
10
considered to be an efficient UVꢀlight photocatalyst, under UV light irradiation with the
2
same light intensity as white LED (1 mW/cm ). The R of Fe(III)ꢀFe Ti O under white LED
x
1ꢀx
2
was 0.29 ꢂmol/h, which is 16% of the R value of Pꢀ25 under UVꢀlight irradiation (1.78
ꢂmol/h). This visibleꢀlight activity of Fe(III)ꢀFe is markedly higher than those of
previously reported visibleꢀlight photocatalysts, which were two orders of magnitude lower
x
Ti1ꢀxO
2
9
,65
than that of pure TiO
Fe(III)ꢀFe under white LED irradiation in the present experimental conditions was
approximately 16% of that of Pꢀ25 under UV irradiation. However, the QE of
Fe(III)ꢀFe Ti O under indoor light irradiation was as high as that of commercial Pꢀ25 under
2
under UV light.
Notably, the absorbed photon number of
x
Ti1ꢀxO
2
x
1ꢀx
2
x 2
UV irradiation. Therefore, Fe(III)ꢀFe Ti1ꢀxO is expected to be usable for various indoor
applications, as its photocatalytic performance under indoor light is comparable to that of
Pꢀ25 under sunlight.
In addition to performance, the stability of photocatalysts is critical for practical
applications. Previous studies have reported that doped elements are unstable under light
66ꢀ70
irradiation, as they are often oxidized or reduced during excitation.
Thus, the
photocatalytic activities of doped metal oxides tend to decrease under longꢀterm light
6
7
illumination. For example, Chen et al. reported that the photocatalytic capability of Nꢀdoped
TiO decreases gradually due to the oxidation of lattice nitrogen by photogenerated holes
2
during the degradation reaction, leading to photocatalytic instability. In our present system,
photogenerated electrons are efficiently transferred to surface Fe(III) and consumed in
efficient reduction reaction processes due to the good energy level matching of
surfaceꢀgrafted and bulkꢀdoped Fe(III) ions. We demonstrated that the high performance of
Fe(III)ꢀFe
x
Ti1ꢀx
O
2
could be maintained under repeated light irradiation in air for 1 year, as
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
shown in Figure 6c, and estimated that the turnover number of this system exceeded 80. In
addition, we optimized the various experimental conditions and found that 0.1 wt% was the
optimal amount of both doped and surfaceꢀgrafted Fe(III) ions (Figures S14 and S15). The
x 2
Fe(III)ꢀFe Ti1ꢀxO sample was also very active under UVꢀlight irradiation. Thus, the high
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
photocatalytic activity and stability of Fe(III)ꢀFe Ti O indicates that this photocatalyst has
x
1ꢀx
2
great potential for practical applications.
To further investigate the influence of energy level matching of surfaceꢀgrafted and
bulkꢀdoped Fe(III) ions, we changed the surface Fe(III) to Cu(II), which is also reported to be
21,22
an efficient coꢀcatalyst.
x 2
However, the activity of the prepared Cu(II)ꢀFe Ti1ꢀxO sample
was clearly decreased under the same light irradiation conditions used for Fe(III)ꢀFe
x
Ti1ꢀxO
2
2+
+
54
(
Figure S16). As the reported redox potential of Cu /Cu is 0.16 V (vs. SHE, pH=0), the
energy level of surface Cu(II) is not matched with that of the bulkꢀdoped Fe(III) ions. Further,
we prepared photocatalysts with Cu(II) ions as dopants in place of Fe(III). However, the
photocatalytic activities of Fe(III)ꢀCu
x 2 x 2
Ti1ꢀxO and Cu(II)ꢀCu Ti1ꢀxO were very low, because
the states of the doped Cu(II) ions appeared at much deeper levels in the band gap, locating at
7
1
the upside of VB. Particularly, the activity of Fe(III)ꢀCu
Cu(II)ꢀCu , further proved the deeper energy levels of doped Cu(II) ions, because the
redox potential of Fe /Fe , 0.771 V, is more positive than that of Cu /Cu , 0.16 V. Thus, the
activity of Cu(II)ꢀCu was negligible (Figure S17), owing to the large energy level
difference between the surfaceꢀgrafted and bulkꢀdoped Cu(II) ions.
the photocatalytic mechanisms operating in the Fe(III)ꢀFe Ti O system are proposed, as
x 2
Ti1ꢀxO is higher than that of
x
Ti1ꢀxO
2
3+
2+
2+
+
x
Ti1ꢀxO
2
5
4,71
Based on these results,
x
1ꢀx
2
illustrated in Figure 7. Pure TiO
After the selected surface grafting and bulk doping of Fe(III) ions, which have closely
matched energy levels, the visibleꢀlight absorption of TiO was drastically enhanced by the
2
is inactive under visible light owing to its wide band gap.
2
bulkꢀdoped Fe(III) ions and the QE was unaffected because of the efficient transfer of
electrons between doped Fe(III) and surface Fe(III), which acts as an efficient coꢀcatalyst for
multiꢀelectron reduction reactions. Moreover, a good junction between surfaceꢀgrafted and
bulkꢀdoped Fe(III) ions was also critical for the efficient charge transfer. Notably, if a thin
12
ACS Paragon Plus Environment

Page 13 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
layer was introduced between the surface Fe(III) ions and doped TiO (Figure S18), the
2
visibleꢀlight activity was markedly reduced (Figure S19).
2
In the present paper, using Fe(III) ions grafted Fe(III) doped TiO as a model system, we
comprehensively studied energy level matching between surfaceꢀgrafted Fe(III) ions as
coꢀcatalysts and bulkꢀdoped Fe(III) ions as visible light absorbers. It has been confirmed that
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
our concept is applicable for not only TiO
2
as a metal oxide semiconductors, but also other
), as
semiconductor photocatalysts, including zinc oxide (ZnO) and strontium titanate (SrTiO
3
described in the Supporting Information (Figure S20). Further, the energy level matching of
surfaceꢀgrafted and bulkꢀdoped Fe(III) ions is just one example of a photocatalytic reaction
involving oxygen reduction. Although Fe(III) ion nanoclusters are reported to be effective
23
cocatalysts for oxygen reduction, a number of other cocatalysts, including Cu(II), Pt, Rh,
NiO , RuO , and IrO , have been have been reported to enhance the charge separation
efficiencies for various photocatalytic reactions, including water splitting, CO reduction, NO
For example, we have previously
shown that the coupling of cerium (Ce) doping and copper (Cu) grafting on zinc oxide (ZnO)
x
2
2
2
x
2
1ꢀ24,27ꢀ30,71ꢀ76
removal, and organic and dye decomposition.
2
9
is also very effective for constructing visibleꢀlightꢀsensitive photocatalysts. Since the energy
level of doped Ce is similar to that of surfaceꢀgrafted Cu(II) ions, Ceꢀdoped ZnO grafted with
Cu(II) ions exhibits visibleꢀlight activity. Our concept of matching the energy levels of
dopants with the redox potential of surfaceꢀgrafted ions may help researchers identify suitable
systems for target photocatalytic reactions. We believe that our concept of energy matching
between surfaceꢀgrafted and bulkꢀdoped metal ions is not limited to reaction systems of
specific semiconductor materials, but represents a general strategy for developing new
visibleꢀlightꢀsensitive photocatalysts.
4. Conclusions
We have developed a novel visibleꢀlightꢀdriven photocatalyst based on the concept of
energy level matching between surfaceꢀgrafted and bulkꢀdoped Fe(III) ions. Our findings
show that bulkꢀdoped Fe(III) ions produce energy levels below the CB of TiO , which match
2
well with the redox potential of the surfaceꢀgrafted Fe(III) ions. The doping of Fe(III) ions
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
increase the visibleꢀlight absorption of TiO , while the surface grafting of Fe(III) ions
2
maintains the high QE of this oxide. Due to the close energy level matching of bulk and
surface Fe(III), electrons on the former were transferred to the latter, thereby efficiently
x 2
driving oxygen reduction. Thus, the present Fe(III)ꢀFe Ti1ꢀxO nanocomposites exhibited
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
strong visibleꢀlight absorption and maintained a high QE, leading to the highest visible light R
among reported photocatalysts for the decomposition of gaseous organic compounds. A high
R was achieved even under commercial white LED irradiation and was stable during
longꢀterm use, demonstrating the practical utility of this photocatalyst. Further, the developed
nanocomposites are composed of ubiquitous and safe elements. Notably, the energy level
matching method described here can potentially be applied to other wideꢀbandꢀgap
semiconductors and photocatalysis systems. Thus, our present findings open a new avenue for
constructing advanced visibleꢀlight photocatalysts for practical use.
Acknowledgment
This work was performed under the management of the Project to Create Photocatalysts
Industry for RecyclingꢀOriented Society supported by the New Energy and Industrial
Technology Development Organization (NEDO) in Japan. This research is also supported by
JST, PRESTO. We also thank Prof. Shinichi Ohkoshi, Dr. Asuka Namai, Ms. Marie Yoshikiyo,
and Ms. Kana Yamada of the University of Tokyo for their assistance with Mössbauer spectra
measurements, and Dr. Xingfu He of the Chinese Academy of Sciences and Dr. Daisuke
Satoh of the University of Tokyo for performing theoretical calculations.
Supporting Information Available: Table S1, the amount of Fe(III) in the sample derived
from ICPꢀAES spectrophotometry. Figure S1, Crystal models. Figures S2, photos of samples.
x 2 x 2
Figure S3, XRD patterns of Fe Ti1ꢀxO . Figure S4, ESR spectra of Fe(III)ꢀFe Ti1ꢀxO . Figure
S5, SEM images of samples. Figure S6, fullꢀscale XPS spectra of samples. Figures S7, TEM
images of samples. Figure S8, light source for the visibleꢀlight irradiation. Figure S9, time
course of gas concentrations during the decomposition of IPA by Fe(III)ꢀFe
x
Ti1ꢀx
O2. Figure
14
ACS Paragon Plus Environment

Page 15 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
S10, details of the quantum efficiency calculations. Figure S11, photocatalytic activities of
TiO
2
, TiO2ꢀx
N
x
, and Fe(III)ꢀFe
x
Ti1ꢀx
O
2
. Figure S12, UV and LED indoor light source. Figure
S13, comparative studies of CO
2
generation over Fe(III)ꢀFe under white LED
x 2
Ti1ꢀxO
irradiation and Pꢀ25 under UV light irradiation with the same light intensity. Figures S14 and
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
S15 comparative evaluation of CO generation by Fe(III)ꢀgrafted and bulkꢀdoped samples
2
under visibleꢀlight irradiation. Figure S16, comparative studies of CO generation by
2
Fe(III)ꢀFe 2, Cu(II)ꢀFe , and Fe(III)ꢀCu ,. Figures S17, CO generation by
x
Ti1ꢀx
O
x
Ti1ꢀx
O
2
x
Ti1ꢀx
O
2
2
Fe(III)ꢀFe Ti O and Cu(II)ꢀCu Ti O . Figure S18 and S19, Fe(III)ꢀFe Ti O with a thin
x
1ꢀx
2
x
1ꢀx
2
x
1ꢀx
2
layer between the surfaceꢀgrafted and and bulkꢀdoped Fe(III) ions. Figure S20, UVꢀVis
spectra and comparative studies of CO generation by other semiconductors. This material is
2
available free of charge via the internet at http://pubs.acs.org.
References
(
(
(
(
(
(
1) Thompson, T. L.; Yates, J. T.; Chem. Rev. 2006, 106, 4428ꢀ4453.
2) Chen, X.; Mao, S. S. Chem. Rev. 2007, 107, 2891ꢀ2959.
3) Hoffmann, S. T. Martin, W. Y. Choi, D. W. Bahnemann, Chem. Rev. 1995, 95, 69ꢀ96.
4) Highfield, J. G.; Pichat, P. New. J. Chem. 1989, 13, 61–66.
5) Chen, X.; Liu, L.; Yu, P. Y.; Mao, S. S. Science 2011, 331, 746ꢀ750.
6) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science 2001, 293, 269ꢀ271.
(7) Khan, S. U. M.; AlꢀShahry, M.; Ingler Jr., W. B. Science 2002, 297, 2243ꢀ2245.
(8) Choi, W. Y.; Termin, A.; Hoffmann, M. R. J. Phys. Chem. 1994, 98, 13669ꢀ13679.
(9) Irie, H.; Watanabe, Y.; Hashimoto, K. J. Phys. Chem. B 2003, 107, 5483ꢀ5486.
(
(
10) Ohtani, B.; Chem. Lett. 2008, 37, 216ꢀ229.
11) Liu, M.; Piao, L. Y.; Zhao, L.; Ju, S. T.; Yan, Z. J.; He, T.; Zhou, C. L.; Wang, W. J.
Chem. Commun. 2010, 46, 1664ꢀ1666.
(
12) Liu, M.; Piao, L. Y.; Lu, W. M.; Ju, S. T.; Zhao, L.; Zhou, C. L.; Li, H. L.; Wang, W. J.
Nanoscale 2010, 2, 1115ꢀ1117.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
(
13) Li, H. M.; Zeng, Y. S.; Huang, T. C.; Piao, L. Y.; Yan, Z. J.; Liu, M. Chem. Eur. J.
2012, 18, 7525ꢀ7532.
14) Li, H. M.; Zeng, Y. S.; Huang, T. C.; Piao, L. Y.; Liu, M. ChemPlusChem 2012, 77,
1017ꢀ1021.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(15) Kisch, H.; Zhang, L.; Lange, C.; Maier, W. F.; Antonius, C.; Meissner, D. Angew.
Chem. Int. Ed.1998, 37, 3034ꢀ3036.
(
16) Abe, R.; Takami, H.; Murakami, N.; Ohtani, B. J. Am. Chem. Soc. 2008, 130,
780ꢀ7781.
7
(
17) Tada, H.; Jin, Q.; Nishijima, H.; Yamamoto, H.; Fujishima, M.; Okuoka, S.; Hattori, T.;
Sumida, Y.; Kobayashi, H. Angew. Chem. Int. Ed. 2011, 50, 3501ꢀ3505.
18) Liu, M.; Piao, L. Y.; Wang, W. J. J. Nanosci. Nanotech. 2010, 10, 7469ꢀ7472.
19) Liu, M.; Piao, L. Y.; Ju, S. T.; Lu, W. M.; Zhao, L.; Zhou, C. L.; Wang, W. J. Mater.
Lett. 2010, 64, 1204ꢀ1207.
(
(
(20) Wu, Q.; Liu, M.; Wu, Z. J.; Li, Y. L.; Piao, L. Y. J. Phys. Chem. C. 2012, 116,
6800ꢀ26804.
2
(
(
21) Irie, H.; Miura, S.; Kamiya, K.; Hashimoto, K. Chem. Phys. Lett. 2008, 457, 202ꢀ205.
22) Irie, H.; Kamiya, K.; Shibanuma, T.; Miura, S.; Tryk, D. A.; Yokoyama, T.; Hashimoto,
K. J. Phys. Chem. C 2009, 113, 10761ꢀ10766.
(
23) Yu, H. G.; Irie, H.; Shimodaira, Y.; Hosogi, Y.; Kuroda, Y.; Miyauchi, M.; Hashimoto,
K. J. Phys. Chem. C 2010, 114, 16481ꢀ16487.
(24) Qiu, X. Q.; Miyauchi, M.; Sunada, K.; Minoshima, M.; Liu, M.; Lu, Y.; Li, D.;
Shimodaira, Y.; Hosogi, Y.; Kuroda, Y.; Hashimoto, K. ACS Nano 2012, 6, 1609ꢀ1618.
(25) Yu, H. G.; Irie, H.; Hashimoto, K. J. Am. Chem. Soc. 2010, 132, 6898ꢀ6899.
(
26) Qiu, X. Q.; Miyauchi, M.; Yu, H. G.; Irie, H.; Hashimoto, K. J. Am. Chem. Soc. 2010,
32, 15259ꢀ15267.
1
(
(
(
(
27) Anandan, S.; Ohashi, N.; Miyauchi, M. Appl. Catal. B 2010, 100, 502ꢀ509.
28) Liu, M.; Qiu, X. Q.; Miyauchi, M.; Hashimoto, K. Chem. Mater. 2011, 23, 5282−5286.
29) Anandan, S.; Miyauchi, M. Phys. Chem. Chem. Phys. 2011, 13, 14937ꢀ14945.
30) Anandan, S.; Miyauchi, M. Electrochem. 2011, 79, 842ꢀ844.
16
ACS Paragon Plus Environment

Page 17 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(
(
(
31) Creutz, C.; Brunschwig, B. S.; Suntin, N. J. Phys. Chem. B 2005, 109, 10251ꢀ10260.
32) Creutz, C.; Brunschwig, B. S.; Suntin, N. J. Phys.Chem. B 2006, 110, 25181ꢀ25190.
33) Hush, N. S. Electrochim. Acta 1968, 13, 1005ꢀ1023.
(34) Hush, N. S. J. Electroanal. Chem. 1999, 460,5ꢀ29.
(35) Hush, N. S. J. Electroanal. Chem. 1999, 470, 170ꢀ195.
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
(
36) Himo, F.; Eriksson, L. A.; Maseras, F.; Siegbahn, P. E. M. J. Am. Chem. Soc. 2000,
22, 8031ꢀ8036.
1
(
37) Goldstein, S.; Czapski, G.; Eldik, R.; Cohen, H.; Meyerstein, D. J. Phys. Chem. 1991,
5, 1282ꢀ1285.
9
(
(
38) Kitajima, N.; Morooka, Y. Chem. Rev. 1994, 94, 737ꢀ757.
39) Cole, A.P.; Root, D. E.; Mukherjee, P.; Solomon, E. I.; Stack, T. D. P. Science 1996,
273, 1848ꢀ1850.
(40) Abe, R.; Takami, H.; Murakami, N.; Ohtani, B. J. Am. Chem. Soc.2008, 130,
7780ꢀ7781.
(
(
41) Zhao, Z. G.; Miyauchi, M. Angew. Chem., Int. Ed. 2008, 47, 7051ꢀ7055.
42) Arai, T.; Horiguchi, M.; Yanagida, M.; Gunji, T.; Sugihara, H.; Sayama, K. Chem.
Commun. 2008, 5565ꢀ5567.
(
(
43) Kim,Y. H.; Irie, H.; Hashimoto, K. Appl. Phys. Lett. 2008, 92, 182107.
44) Arai, T.; Yanagida, M.; Konishi, Y.; Iwasaki, Y.; Sugihara, H.; Sayama, K. Catal.
Commun. 2008, 9, 1254ꢀ1258.
(45) Kresse, G.; Hafner, J. Phys. Rev. B 1993, 48, 13115ꢀ13118.
(46) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169ꢀ11186.
(47) Perdew, J. P.; Burke, K.; and Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865ꢀ3868.
(
(
(
48) Perdew, J. P.; Burke, K.; and Ernzerhof, M. Phys. Rev. Lett. 1997, 78, 1396ꢀ1396.
49) Vanderbilt, D. Phys. Rev. B 1990, 41, 7892ꢀ7895.
50) Yu, X. H.; Li, C. S.; Ling, Y.; Tang, T. A.; Wu, Q.; Kong, J. J.; J. Alloy. Compd. 2010,
507, 33ꢀ37.
(
51) Henderson, M. A. A. Surf. Sci. Rep. 2011, 66, 185ꢀ297.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
(
52) Gai, Y. Q.; Li, J. B.; Li, S. S.; Xia, J. B.; and Wei, S. H.; Phys. Rev. Lett. 2009, 102,
0
36402.
53) Wen, L. P.; Liu, B. S.; Zhao, X. J.; Nakata, K.; Murakami, T.; and Fujishima, A. Int. J.
Photoenergy 2012, 2012, 368750.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(54) Bard, A.J.; Parsons, R.; Jordan, J. Standard Potentials in Aqueous Solution, Marcel
Dekker, New York, 1985.
(
55) Cullity, B.D.; Stock, S.R. Elements of XꢀRay Diffraction, 3rd Ed., PrenticeꢀHall Inc.,
New Jersey, 2001.
(
56) Bapna, K.; Choudhary, R. J.; Pandey, S. K.; Phase, D. M.; Sharma, S. K.; Knobel, M.
Appl. Phys. Lett. 2011, 99, 112502.
(
(
57) Oku, M.; Wagatsuma, K.; Kohiki, S. Phys. Chem. Chem. Phys. 1999, 1, 5327ꢀ5331.
58) Millet, J. M. Adv. Catal. 2007, 51, 309–350.
(59) Pan, X.; Jiang, D.; Lin, Y.; Ma, X. J. Magn. Magn. Mater. 2006, 305, 388ꢀ391.
(60) Lee, H. M.; Kim, C. S. J. Magn. Magn. Mater. 2007, 310, 2099ꢀ2101
(
61) Balcells, L.; Frontera, C.; Sandiumenge, F.; Roig, A.; Martinez, B.; Kouam, J.; Monty,
C. Appl. Phys. Lett. 2006, 89, 122501.
(
62) Lin, F.; Jiang, D. M.; Lin, Y.; Ma, X. M.; Physica B 2008, 403, 2193ꢀ2196.
63) Cabrera, A. F.; Rodriguez Torres, C. E.; Errico, L.; Sanchez, F. H. Physica B 2006, 384,
(
345ꢀ347.
(
64) Ohko, Y.; Hashimoto, K.; Fujishima, A. J. Phys. Chem. A 1997, 101, 8057ꢀ8062.
(65) Miyauchi, M.; Ikezawa, A.; Tobimatsu, H.; Irie, H.; Hashimoto, K. Phys. Chem. Chem.
Phys. 2004, 6, 865ꢀ870.
(66) Koudriachova, M. V.; Harrison, N. M. J. Mater. Chem. 2006, 16, 1973ꢀ1977.
(
(
(
67) Chen, X. F.; Wang, X. C.; Hou, Y. D.; Huang, J. H.; Wu, L.; Fu, X. Z. J. Catal. 2008,
55, 59ꢀ67.
2
68) Mowbray, D. J.; Martinez, J. I.; Garcia Lastra, J. M.; Thygesen, K. S.; Jacobsen, K. W.
J. Phys. Chem. C 2009, 113, 12301ꢀ12308.
69) Braconnier, B.; Paez, C. A.; Lambert, S.; Alie, C.; Henrist, C.; Poelman, D.; Pirard, J.
P.; Cloots, R.; Heinrichs, B. Microporous Mesoporous Mater. 2009, 122, 247ꢀ254.
18
ACS Paragon Plus Environment

Page 19 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(
(
70) Hanaor, D. A. H.; Assadi, M. H. N.; Li, S.; Yu, A. B.; Sorrell, C. C. Comput. Mech.
2012, 50,185–194.
71) Errico, L. A.; Rentería, M.; Weissmann, M. Phys. Rev. B 2005, 72, 184425.
(72) Moon, S. C.; Mametsuka, H.; Suzuki, S.; and Anpo, M.; Chem. Lett. 1998, 27,
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
117ꢀ118.
(
(
73) Yamaguti, K.; Sato, S.; J. Chem. Soc., Faraday Trans. 1 1985, 81, 1237ꢀ1246.
74) Shimizu, K.; Itoh, S.; Hatamachi, T,; Kodama, T., Sato, M.; Toda, K.; Chem. Mater.
2005, 17, 5161ꢀ5166.
(
(
75) Yuan, Y.; Zheng, J.; Zhang, X.; Li, Z.; Yu, T.; Ye. J.; Zou, Z.; Solid Sate Ionics 2008,
78, 1711ꢀ1713.
1
76) Abe, R.; Higashi, M.; Domen, K.; J. Am. Chem. Soc. 2010, 132, 11828ꢀ11829.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Performances of the indicated photocatalysts.
Sample
TiO2ꢀx
N
x
Fe
x
Ti1ꢀx
O
2
Fe(III)ꢀTiO
2
Fe(III)ꢀFe
x
Ti1ꢀxO
2
i
R
p
16
16
16
16
1
.30×10
1.30×10
1.39×10
0.09
1.30×10
1.30×10
1.46×10
0.69
(quanta/sec)
a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R
p
15
15
14
15
4
.10×10
7.48×10
(
quanta/sec)
Rco2
0.16
0.40
53.5
(
ꢀmol/h)
QE
3
.9
6.5
47.3
(
%)
i
R
p
, rate of incident photons.
a
p
R , absorbed photon number.
2
Rco2, CO generation rate.
20
ACS Paragon Plus Environment

Page 21 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure captions.
2 x 2
Figure 1. DOSs for (a) rutile TiO and (b) Fe Ti1ꢀxO (x=0.0625). The tops of the VBs are
assigned as 0 eV.
2 2 x 2
Figure 2. (a) UVꢀvisible reflectance spectra of TiO , Fe(III)ꢀTiO Fe Ti1ꢀxO and Fe(III)ꢀ
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
Fe Ti O nanocomposites. (b) Difference UVꢀvis spectra versus bare TiO for
x
1ꢀx
2
2
Fe(III)ꢀTiO
Figure 3. XRD patterns of bare TiO
nanocomposites at x=0.1 wt%.
Figure 4. (a) Ti 2p coreꢀlevel spectra and (b) Fe 2p coreꢀlevel spectra of bare TiO
Fe(III)ꢀTiO , Fe , and Fe(III)ꢀFe nanocomposites at x=0.1 wt%.
Figure 5. (a) Mössbauer spectra of bare TiO , Fe(III)ꢀTiO , Fe , and Fe(III)ꢀ
Fe nanocomposites at x=0.1 wt%. (b) HRTEM images of Fe(III)ꢀFe
2
, Fe
x
Ti1ꢀx
O
2
and Fe(III)ꢀFe
x
Ti1ꢀx
O
2
nanocomposites at x=0.1 wt%.
2
, Fe(III)ꢀTiO
2
, Fe
x
Ti1ꢀx
O
2
x 2
, and Fe(III)ꢀFe Ti1ꢀxO
2
,
2
x
Ti1ꢀx
O
2
x 2
Ti1ꢀxO
2
2
x 2
Ti1ꢀxO
x
Ti1ꢀx
O
2
x
2
Ti1ꢀxO .
Fe(III) nanoclusters are indicated by the dotted lines. The good attachment of
Fe(III) nanoclusters to Fe nanoparticles can be clearly observed.
Figure 6. (a) CO generation curves for Fe(III)ꢀM Ti O (M=Fe, Ce, Cu, and Ni, x=0.1 wt%)
x
Ti1ꢀxO
2
2
x
1ꢀx
2
samples under visibleꢀlight irradiation. (b) Comparative studies of CO
by bare Fe , Fe(III)ꢀTiO , and Fe(III)ꢀFe at x=0.1 wt%, respectively.
c) Cycling measurements of the IPA decomposition over Fe(III)ꢀFe Ti O at
2
generation
x
Ti1ꢀx
O
2
2
x 2
Ti1ꢀxO
(
x
1ꢀx
2
x=0.1 wt% under visibleꢀlight irradiation. Experiments were conducted over a
2ꢀmonth period.
Figure 7. Proposed photocatalysis processes. Pure TiO
to its wide band gap. After selected surface grafting and bulk doping of Fe(III) ions,
1
2
is inactive for the visible light owing
which have closely matched energy levels, the visibleꢀlight absorption of TiO
2
is
drastically enhanced by the bulkꢀdoped Fe(III) ions and the photogenerated
electrons can be effectively transferred to the surface Fe(III), which acts as an
efficient coꢀcatalyst for multiꢀelectron reduction reactions. Holes with high
oxidation power are kept in the deep level of VB and decompose the organic
compounds efficiently. Thus, efficient visible light photocatalysts with high R are
achieved.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 1.
3
2
1
0
a
b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
total
Ti
O
total
Ti
O
Fe
ꢀ
1
2
3
ꢀ
ꢀ
DOS (a.u.)
DOS (a.u.)
22
ACS Paragon Plus Environment

Page 23 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 2.
100
18
a
b
pure TiO2
Fe(III)ꢀTiO2
15
8
6
4
2
0
0
0
0
0
Fe(III)ꢀTiO2
Fe Ti1ꢀxO2
x
Fe Ti1ꢀxO2
Fe(III)ꢀFe Ti1ꢀxO2
x
12
9
x
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
Fe(III)ꢀFe Ti1ꢀxO2
x
6
3
0
200
200
300
400
500
600
700
800
300
400
500
600
700
800
Wavelength (nm)
Wavelength (nm)
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 3.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pure TiO2
Fe(III)ꢀTiO2
Fe Ti1ꢀxO2
x
Fe(III)ꢀFe Ti1ꢀxO2
x
10
20
30
40
50
60
70
80
2θ
(degree)
24
ACS Paragon Plus Environment

Page 25 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 4
a
b
pure TiO2
Ti 2p3/2
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
Fe Ti1ꢀxO2
x
Fe(III)ꢀTiO2
Ti 2p1/2
Fe(III)ꢀFe Ti1ꢀxO2
x
Fe(III)ꢀFex
Fe(III)ꢀTiO2
Fex i1ꢀxO2
O2
Ti1ꢀxO2
T
Pure
Ti
470
465
460
455
450
445
740
730
720
710
Binding enery (eV)
Binding Energy (eV)
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5.
a
b
Pure TiO2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fe(III)ꢀTiO2
Fe Ti1ꢀxO2
x
Fe(III)ꢀFe Ti1ꢀxO
x
2
ꢀ4
ꢀ2
0
2
4
Velocity (mm/s)
26
ACS Paragon Plus Environment

Page 27 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 6.
a
b
18
Fe
1
1
1
8
5
2
9
6
3
0
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
Ce
Mo
Cu
Ni
W
Cr
15
1
2
9
6
3
0
Fe Ti1ꢀxO
x
2
Fe(III)ꢀTiO
2
Fe(III)ꢀFe Ti1ꢀxO
x
2
0
30
60
90
120
4
150
0
30
60
90
120
150
Irradiation time (h)
Irradiation time (h)
Time (month)
6
0
2
8
10
12
c
1
8
1
5
12
9
6
3
0
0
100
200
300
400
500
600
700
Irradiation time (h)
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 29
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 7.
CB
CB
0.771V
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O /HO =
ꢀ0.046V
2
2
eꢀ
O2
eꢀ
ꢀ
e
ꢀ
O /H O = 0. 68V
e
2
2
2
Reduction
O /H O = 1.23V
2
2
IPA
Oxidation
3
h+
VB
VB
Pure TiO2
Fe(III)ꢀFe Ti O
2
x
1ꢀx
28
ACS Paragon Plus Environment

Page 29 of 29
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table of contents
Visible light
CB
Fe(III)
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
eꢀ
ꢀ
eꢀ eꢀ
e
e
Fe(III)
TiO2
multiꢀelectron
reduction
eꢀ
ꢀ
h+
VB
oxidation
29
ACS Paragon Plus Environment