Rutile crystallites isolated from degussa (Evonik) P25 TiO2: Highly efficient photocatalyst for chemoselective hydrogenation of nitroaromatics

Article abstract of DOI:10.1021/cs400532p

We report that the rutile crystallites, isolated from Degussa (Evonik) P25 TiO₂ by a hydrofluoric acid treatment, behave as a highly efficient photocatalyst for hydrogenation of nitroaromatics. Photoirradiation (λ >300 nm) of the isolated rutil

Full text of DOI:10.1021/cs400532p

Subscriber access provided by MONASH UNIVERSITY
Article
Rutile Crystallites Isolated from Degussa (Evonik) P25 TiO2: Highly Efficient
Photocatalyst for Chemoselective Hydrogenation of Nitroaromatics
Yasuhiro Shiraishi, Hiroaki Hirakawa, Yoshiki Togawa,
Yoshitsune Sugano, Satoshi Ichikawa, and Takayuki Hirai
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/cs400532p • Publication Date (Web): 06 Sep 2013
Downloaded from http://pubs.acs.org on September 10, 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.
ACS Catalysis 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 10
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
Rutile Crystallites Isolated from Degussa (Evonik) P25 TiO2:
Highly Efficient Photocatalyst for Chemoselective Hydrogena-
tion of Nitroaromatics
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
†
Yasuhiro Shiraishi,*, Hiroaki Hirakawa,† Yoshiki Togawa,† Yoshitsune Sugano,† Satoshi Ichikawa,‡ and
Takayuki Hirai†
† Research Center for Solar Energy Chemistry, and Division of Chemical Engineering, Graduate School of Engineering Science,
Osaka University, Toyonaka 560-8531, Japan
‡ Institute for NanoScience Design, Osaka University, Toyonaka 560-8531, Japan
Supporting Information Placeholder
KEYWORDS: Photocatalysis; Titanium dioxide; Rutile; Oxygen vacancy; Structural defect
ABSTRACT: We report that the rutile crystallites, isolated from Degussa (Evonik) P25 TiO2 by a hydrofluoric acid treatment, behave
as a highly efficient photocatalyst for hydrogenation of nitroaromatics. Photoirradiation (λ >300 nm) of the isolated rutile particles
with alcohol as a hydrogen source successfully promotes chemoselective hydrogenation of nitroaromatics to anilines, with an activity
higher than that of commercially-available rutile TiO2. The high activity of the isolated rutile particles is due to the specific
distribution of structural defects (oxygen vacancy sites) on the particles. These particles contain a relatively small number of inner
defects behaving as recombination centers for photoformed electron (e–) and positive hole (h+) pairs, and a relatively large number of
surface defects behaving as reduction sites for nitroaromatics. Photoexcitation of the isolated particles therefore promotes efficient
charge separation between e– and h+, and facilitates rapid reduction of nitroaromatics adsorbed on the surface defects. This thus
results in very high hydrogenation activity on the rutile particles isolated from P25 TiO2.
INTRODUCTION
various functionalized anilines with >93% selectivity. The most
striking aspect of this system is that "rutile" TiO2 particles are
effective for efficient and selective nitro hydrogenation. It is
well known that TiO2 has two common polymorphic forms:
anatase and rutile. Anatase usually exhibits much higher
photocatalytic activity than rutile,11 and its activity is further
improved by coupling with ca. 20% rutile, which is available as
AEROXIDE P25 TiO2 by Degussa (Evonik).12 Both anatase and
P25 TiO2, however, show lower activity and selectivity than
rutile TiO2 for nitro hydrogenation.
Functionalized anilines are versatile intermediate for the syn-
thesis of pharmaceuticals, polymers, and fine chemicals.1 They
are generally produced by hydrogenation of nitroaromatics
using stoichiometric or excess amounts of reducing agents with
a concomitant formation of copious amount of wastes.2 Catalyt-
ic hydrogenation is therefore an ideal process; however, conven-
tional systems with platinum-based catalysts and molecular hy-
drogen (H2) also promote hydrogenation of other functionali-
ties such as vinyl and carbonyl groups, resulting in poor
chemoselectivity.3 There are only a few catalytic systems of
chemoselective nitro hydrogenation.4–9 In addition, these sys-
tems require high H2 pressure (>5 bar), high reaction tempera-
ture (>373 K), and noble metal catalyst such as gold and silver.
Alternative systems that promote chemoselective nitro hydro-
genation under milder reaction conditions (room temperature
and atmospheric pressure) without noble metal are necessary
for safe and clean synthesis of functionalized anilines.
X-ray photoelectron spectroscopy (XPS), diffuse-reflectance
infrared Fourier transform (DRIFT) spectroscopy, and DR UV-
vis spectroscopy of rutile TiO2 revealed that the active sites for
photocatalytic nitro hydrogenation on rutile TiO2 are the Ti3+
atoms located at the surface defects. As shown in Scheme 1a,
the rutile (110) surface is characterized by alternate rows of five-
fold coordinated Ti4+ atoms and bridging O2– atoms (Ob) that
run in the (001) direction.13 Surface defects are the Ob vacancies,
where two excess electrons associated with Ob are transferred to
the empty 3d orbitals of neighboring Ti4+ atoms, producing two
exposed Ti3+ atoms. As shown in Scheme 1b, these surface Ti
atoms behave as an adsorption site for nitroaromatics via an
electron donation to the oxygen atoms of nitro groups,14 and as
a trapping site for photoformed conduction band electrons.15
Earlier, we reported that photoexcitation (λ >300 nm) of
semiconductor TiO2 with alcohol as a hydrogen source pro-
motes chemoselective hydrogenation of nitroaromatics.10 The
reaction can be carried out at room temperature and atmos-
pheric pressure without noble metal, and successfully produces
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 10
1
2
3
4
5
6
These effects therefore facilitate rapid nitro-to-amine reduction
on the photoactivated rutile TiO2.
Next is the exploration of highly active rutile TiO2 particles
for rapid nitro hydrogenation. It is well known that commer-
cially-available rutile TiO2 are synthesized by "wet process" (hy-
drolysis of Ti precursors followed by calcination).16 In contrast,
P25 TiO2 is synthesized by "dry process" (flame pyrolysis of gas-
eous TiCl4),17 and it is a mixture of anatase and rutile particles
(ca. 80/20 w/w).12 Herein, we report that the rutile particles,
isolated from P25 TiO2 by a hydrofluoric acid (HF) treatment,
promote photocatalytic nitro hydrogenation with an activity
higher than that of commercially-available rutile TiO2. DRIFT
and DR UV-vis analysis was carried out to clarify the reason for
high activity of the isolated rutile particles. Photocatalytic activi-
ty of rutile TiO2 depends strongly on the number of surface
defects as well as the number of inner defects on the particles
(Scheme 1b). The rutile particles isolated from P25 TiO2 have a
defect distribution that differs significantly from the commer-
cially-available rutile particles synthesized by wet process. The
isolated rutile particles contain a relatively small number of
inner defects behaving as recombination centers for
photoformed electron (e–) and positive hole (h+) pairs, and a
relatively large number of surface defects behaving as active sites
for nitro hydrogenation. Photoexcitation of the isolated rutile
particles therefore promotes efficient charge separation between
e– and h+, and facilitates rapid reduction of nitroaromatics ad-
sorbed on the surface defects, resulting in high activity for nitro
hydrogenation.
7
8
9
Scheme 1. (a) Surface and Inner Structure of Rutile TiO2
(110), and (b) Photocatalytic Cycle for Hydrogenation of Ni-
trobenzene on the Surface Defect.a
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
surface Ti–OH
a
surface defect
(Ob vacancy)
inner defect
b
a
The light blue and green spheres are Ob atoms that lie in the
[001] azimuth. The parallel red spheres are Ti atoms and small
yellow spheres are H atoms, respectively.
Table 1. Properties of TiO2 and the rate for photocatalytic hydrogenation of nitrobenzene.a
Recovery Crystalline
SBET
/
dp /
Ebg /
vf /
Nsurface
/
Ntotal /
sample
catalyst
m2 g–
57
nme
eV f
ꢀmol h–
ꢀmol g–
ꢀmol g–
/ % b
phase c
A 83/R 17
A 30/R 70
R
1 d
1 g
1 h
1 i
1
2
3
P25j
HF(5)/P25
HF(10)/P25
1st reusek
27, 340
107
3.15
3.00
2.97
9.9
11.2
17.3
17.5
16.7
7.1
16
8
36
32
71
40.2
53.8
2nd reuse k
HF(15)/P25
JRC-TIO-6 j
NS-51l
PT-101m
CR-ELm
4
5
6
R
R
R
R
R
R
R
38
100
7
61
21
2.95
3.00
2.97
2.98
2.95
2.96
2.98
10.4
168.0
15.9
81.6
11.1
1.9
49.4
253.1
32.9
15.6
11.9
11.9
11.4
<0.1
10.4
6
217
71
7
25
7
146.2
22.0
8
245
276
14
9
Wako rutile
HF(10)/JRC-TIO-6
15
114
19.9
10
a Photoirradiation was carried out with a 2 kW Xe lamp (light intensity at 300–450 nm is 27.3 W m–2). b Percent recovery of particles after
c
HF treatment. Determined by XRD analysis, where the ratio of anatase (A) and rutile (R) was determined with the equation; A (%) =
d
e
IA(101)/(IA(101) + 1.4IR(110)) × 100 (ref 19). Brunauer−Emmett−Teller surface area determined by N2 adsorption/desorption analysis. Particle
diameter determined by DLS analysis. f Bandgap energies determined by a plot of the Kubelka–Munk function versus the energy of light ab-
sorbed. g Initial rate for aniline formation determined by dividing the amount of aniline formed during 2 h photoreaction by 2 (h). h Number
of surface defects. i Number of total defects. j Japan Reference Catalyst supplied from Catalyst Society of Japan. k The catalyst was reused after
simple washing with solvent. l Supplied from Toho Titanium Co., Ltd. (Japan). m Supplied from Ishihara Sangyo, Ltd. (Japan).
ACS Paragon Plus Environment

Page 3 of 10
ACS Catalysis
1
2
3
4
5
6
7
8
RESULTS AND DISCUSSION
5
4
3
2
1
Preparation and properties of catalysts. HF solution dis-
solves anatase TiO2 more easily than rutile;18 therefore, rutile
particles can easily be isolated from P25 by simple HF treatment.
P25 TiO2 particles (1 g) were stirred in water (50 mL) with dif-
ferent concentrations of HF (x wt %) for 24 h at 298 K. The
resulting powders were washed thoroughly with water until the
pH of solution becomes ca. 7 to remove HF and F– and dried in
vacuo for 12 h, affording HF(x)/P25 particles. As summarized
in Table 1 (samples 2–4), the percent recovery of particles de-
creases with an increase in HF concentration, suggesting that
parts of P25 are dissolved in HF solution.
9
HF(10)/P25 (3)
P25 (1)
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
HF(15)/P25 (4)
HF(10)/P25 (3)
HF(5)/P25 (2)
P25 (1)
0
300
350
400
450
λ / nm
Figure 2. DR UV-vis spectra of HF(x)/P25. The spectra for other
rutile TiO2 particles are summarized in Figure S2 (Supporting In-
formation).
P25 (1)
a) P25 (1)
HF(5)/P25 (2)
HF(10)/P25 (3)
HF(15)/P25 (4)
[101]
50 nm
b) HF(5)/P25 (2)
[200]
[004]
anatase
[110]
[211]
[101]
rutile
20
30
40
50
60
2θ
/ deg
50 nm
Figure 1. XRD patterns of respective TiO2 particles and standard
patterns for anatase (JCPDS 21-1272) and rutile (JCPDS 21-1276).
The text in the parenthesis denotes the sample number listed in
Table 1. The XRD patterns for other rutile particles are summa-
rized in Figure S1 (Supporting Information).
c) HF(10)/P25 (3)
Figure 1 shows the X-ray diffraction (XRD) patterns of P25
and HF(x)/P25. Pure P25 exhibits distinctive diffractions as-
signed to anatase and rutile crystallites. As shown in Table 1
(sample 1), the ratio of anatase and rutile phases for P25, de-
termined by the intensity ratio,19 is 83 : 17. HF treatment of
P25 decreases the anatase diffractions due to the dissolution of
anatase phase. HF(5)/P25 (sample 2) still contains anatase
(30%), but HF(10)/P25 (sample 3) and HF(15)/P25 (sample 4)
exhibit solely rutile diffractions.
50 nm
d) HF(15)/P25 (4)
Figure 2 shows the DR UV-vis spectra of HF(x)/P25. The ab-
sorption edge of pure P25 is ca. 400 nm, and its bandgap ener-
gy is 3.15 eV (Table 1). HF treatment of P25 leads to a shift of
the absorption edge to longer wavelength (ca. 420 nm), along
with a decrease in the bandgap energy. This suggests that
anatase crystallites with bandgap energy larger than rutile crys-
tallites20 are selectively dissolved in HF solution.
50 nm
Figure 3. Typical (left) TEM and (right) high-resolution TEM imag-
es of HF(x)/P25. More images of HF(10)/P25 and HF(15)/P25 are
summarized in Figures S3 and S4, respectively (Supporting Infor-
mation).
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 10
1
2
3
4
P25 (1)
5
6
7
8
Figure 4 shows the distribution of particle size determined by
dynamic laser scattering (DLS) analysis. As shown by the black
line, pure P25 exhibits bimodal distribution with average diam-
eters 27 and 340 nm, respectively. The smaller particles are
assigned to the dispersed anatase particles,23 and the larger par-
ticles are the interwoven aggregates of anatase and rutile parti-
cles.24 HF treatment of P25 leads to a disappearance of the dis-
persed anatase particles, along with a decrease in the size of
aggregates. HF(10)/P25 shows monodispersed particle distribu-
tion with average diameter 71 nm, indicating that
monodispersed rutile particles are isolated from P25 by the HF
treatment.
40
20
0
26.5 nm
HF(15)/P25 (4)
61.1 nm
HF(10)/P25 (3)
71.1 nm
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
HF(5)/P25 (2)
107 nm
340 nm
101
102
103
Particle size / nm
Photocatalytic activity. The activity of respective catalysts was
tested by hydrogenation of nitrobenzene to aniline with 2-
PrOH. Each respective catalyst (5 mg) was added to 2-PrOH (5
mL) containing nitrobenzene (10 mM), and the suspension was
photoirradiated by a Xe lamp (λ >300 nm) with stirring under
N2 atmosphere (1 atm). The time-course profiles for aniline
formation on the respective catalysts are shown in Figure S6
(Supporting Information). It is noted that the absence of cata-
lyst promotes almost no reaction. Table 1 summarizes the initial
rate for aniline formation determined by 2 h photoreaction on
Figure 4. Size distribution of HF(x)/P25 particles determined by
DLS analysis. The distributions of other rutile TiO2 are summa-
rized in Figure S5 (Supporting Information).
Figure 3 shows the typical transmission electron microscopy
(TEM) images of TiO2 particles. Pure P25 (Figure 3a) is a mix-
ture of small particles with 20–40 nm diameters and large parti-
cles with 60–100 nm diameters, which are assigned to anatase
and rutile particles, respectively.18 HF treatment of P25 (Figure
3b–d) leads to a disappearance of small particles, suggesting that
anatase particles are indeed dissolved in HF solution. Most of
the HF(10)/P25 particles are surrounded by low indexed plane.
In contrast, the shape of HF(15)/P25 is almost round, and the
size of particles is smaller than that of HF(10)/P25. This means
that parts of the rutile particles are decomposed by the treat-
ment with higher concentration of HF.21,22 These data suggest
that the treatment of P25 with ca. 10% HF successfully isolates
rutile particles while maintaining high crystallinity.
–
1
the respective catalyst (vf / ꢀmol h ). The vf of P25 (sample 1) is
–
1
9.9 ꢀmol h . HF treatment of P25 increases the reaction rate; vf
–
1
of HF(5)/P25 (sample 2) is 11.2 ꢀmol h , indicating that rutile
TiO2 is indeed effective for nitro hydrogenation.10 HF(10)/P25
–
1
(sample 3) exhibits much higher activity (17.3 ꢀmol h ),
whereas HF(15)/P25 (sample 4) shows decreased activity (7.1
–
1
ꢀmol h ). In the dark condition, HF(10)/P25 promotes almost
no reaction, suggesting that photoexcitation of the catalyst ini-
tiates nitro hydrogenation.
a) 20
c) 20
b) 20
3
3
3
5
5
5
15
10
5
15
10
5
15
10
5
7
6
8
6
7
6
7
8
9
8
4
4
4
9
9
0
0
0
0
50
SBET / m2 g−1
100
0
100
surface / ꢀmol g−1
200
0
100
200
300
N
Ntotal / ꢀmol g−1
d)
e)
20
15
10
5
20
15
10
5
3
3
5
5
6
6
7
7
8
8
4
4
9
9
0
0
0
50
Ninner / ꢀmol g−1
100
0
0.5
Nsurface / Ntotal
1
Figure 5. Relationship between vf versus (a) SBET, (b) Nsurface, (c) Ntotal, (d) Ninner, and (e) Nsurface/Ntotal for respective rutile TiO2 (sample 3–
9). The sample numbers correspond to those listed in Table 1.
ACS Paragon Plus Environment

Page 5 of 10
ACS Catalysis
1
2
3
4
5
6
Table 2. Photocatalytic hydrogenation of nitroaromatics on HF(10)/P25 (sample 3) or JRC-TIO-6 (sample 5) catalyst.a
substrate
product
7
8
9
solventb
2-PrOH
t / h c
entry
substrate
catalyst
product
conv. / % d
yield / %d
1e
2e
3e
4e
5
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
HF(10)/P25
JRC-TIO-6
97
94
81
98
76
96
71
94
83
98
78
98
78
94
85
3
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
84
>99
78
2-PrOH/toluene
(1/9 w/w)
3
4
6
5
5
5
>99
78
2-PrOH/toluene
(1/9 w/w)
6
7
>99
87
2-PrOH/toluene
(9/1 w/w)
8
9
>99
80
2-PrOH/THF
(5/5 w/w)
10
11
12
13 f
14 f
>99
82
2-PrOH/toluene
(1/9 w/w)
>99
91
2-PrOH/toluene
(1/9 w/w)
a Reaction conditions: substrate (50 ꢀmol), catalyst (10 mg), solvent (5 mL), temperature (303 K), N2 (1 atm), Xe lamp (λ >300 nm). b Sol-
vents were selected in respect to the solubility of substrates. Photoirradiation time. d Determined by GC. catalyst (5 mg). substrate (25
c
e
f
ꢀmol).
sample 10 (Table 1), JRC-TIO-6, treated with 10% HF solution,
shows activity (10.4 ꢀmol h ) much lower than that of pure
150
100
50
100
80
60
40
20
0
–
1
–
1
JRC-TIO-6 (sample 5; 15.6 ꢀmol h ). This clearly indicates that
HF treatment does not enhance photocatalytic activity, but the
rutile particles contained in P25 inherently possess high activity.
It is also note that, as shown in Table 1 (sample 3), the
HF(10)/P25 catalyst recovered after photoreaction, when re-
used for further reaction, shows almost the same activity as the
virgin catalyst. This suggests that the catalyst is reusable without
the loss of activity.
acetone
aniline
nitrobenzene
0
Chemoselective nitro hydrogenation is successfully promoted
on HF(10)/P25. Table 2 summarizes the results for photoreac-
tion of nitroaromatics with various substituents. All of the
nitroaromatics are converted to the corresponding anilines with
>94% yields even in the presence of reducible substituents.
These activities are higher than those of JRC-TIO-6 rutile TiO2
with larger surface area.10 These results suggest that HF(10)/P25
promotes rapid and chemoselective nitro hydrogenation.
0
1
2
3
4
5
6
Time / h
Figure 6. Time-dependent change in the amounts of substrate and
products during photoreaction of nitrobenzene in 2-PrOH with
HF(10)/P25 (sample 3). The reaction conditions are identical to
those in Table 1.
It must be noted that the activity of HF(10)/P25 (sample 3) is
higher than that of commercially-available rutile TiO2 (samples
5–9). It is well known that photocatalytic activity of TiO2 de-
pends on their surface area; the catalyst with larger surface area
shows enhanced activity.25 The HF(10)/P25 catalyst (sample 3)
Figure 6 shows the time-dependent change in the amounts of
substrate and products during photocatalytic reaction of nitro-
benzene with the HF(10)/P25 catalyst. The substrate nitroben-
zene (50 ꢀmol) is quantitatively transformed to aniline by 3 h
photoreaction, along with a formation of almost three equiva-
lents of acetone (150 ꢀmol). This indicates that the H atoms of
alcohol, removed by oxidation with photoformed h+, are con-
sumed quantitatively by the nitro-to-amine hydrogenation. Stoi-
chiometric oxidation and reduction reactions therefore occur
on the rutile particles isolated from P25, as is the case for
commercially-available rutile TiO2,10 as shown in Scheme 1b
and as follows.
–
1
has a surface area (32 m2 g ) much smaller than JRC-TIO-6
–
1
rutile TiO2 (100 m2 g ; sample 5), but shows higher activity. In
addition, as shown in Figure 5a, the plot of vf versus the surface
area for respective TiO2 (samples 3–9) shows almost no rela-
tionship. These data suggest that the high photocatalytic activity
of the rutile particles isolated from P25 cannot be explained by
the surface area. It must also be noted that, as shown by the
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 10
1
2
3
4
5
6
7
3(CH3)2CHOH + 6h+ → 3(CH3)2CO + 6H+
Ar–NO2 + 6e– + 6H+ → Ar–NH2 + 2H2O
(1)
(2)
active site for nitro reduction. The number of nitrobenzene
adsorbed onto the surface Ti3+ atoms can be determined by the
intensity ratio of the respective bands. The number of surface
Ti3+ atoms (surface defects) per gram catalyst (Nsurface / ꢀmol g–1)
is equal to twice the amount of adsorbed nitrobenzene.
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
The Nsurface values for respective rutile TiO2 are summarized in
Table 1. Figure 5b plots Nsurface versus the initial rate for aniline
formation (vf) during photocatalytic reduction of nitrobenzene
for respective rutile TiO2. The plotting data indicate that there
is almost no relationship between these factors. In that, the
HF(10)/P25 catalyst (sample 3) with relatively smaller Nsurface
value shows the highest activity. This suggests that the Nsurface
value is not the sole factor affecting the phocatalytic activity of
rutile TiO2.
νsym
νasym
1346
1522
HF(10)/P25 ( 3)
HF(15)/P25 ( 4)
F(R∞)
0.03
Surface and inner defects as deactivation sites. As shown in
Scheme 1a, oxygen vacancies exist inside the TiO2 particles as
well as on the particle surface. These inner and surface defects
create some energy levels around the mid-gap and behave as
recombination centers for photoformed e– and h+, resulting in
deactivation of photoexcited TiO2.27 This means that the inner
defects behave as recombination centers, whereas the surface
defects behave as active sites for reduction as well as
recombination centers. This implies that the rutile particles
with smaller number of defects would allow efficient charge
separation between e– and h+. The number of inner defects
JRC−TIO−6 (5)
NS−51 (6)
PT−101 (7)
CR−EL(8)
Wako rutile (9)
ATR
(Ninner) or total defects (Ntotal) may therefore affect the
1700 1600 1500 1400 1300 1200
Wavenumber / cm−1
photocatalytic activity.
As reported,28 the total number of defects on TiO2 can be de-
termined by photocatalytic reduction of methyl viologen (MV2+).
TiO2 particles (5 mg) were added to water (10 mL) containing
triethanolamine (TEOA, 10%) as a h+ scavenger, and the sus-
pension was photoirradiated with stirring under Ar atmosphere
at 298 K. After the photoreaction, a chloride salt of MV2+ (125
ꢀmol) was added to the resulting blue suspension and stirred
for 1 h in the dark at 298 K under Ar. The resulting solution
was recovered by centrifugation and subjected to UV-vis absorp-
tion analysis. Figure 8 shows the absorption spectra of the solu-
tion recovered after photoreaction with PT-101 catalyst (sample
7) followed by reaction with MV2+ as an example. A distinctive
absorption band appears at 606 nm, which is assigned to the
Figure 7. DRIFT spectra of nitrobenzene adsorbed onto the respec-
tive TiO2 particles in the gas phase at 303 K. Attenuated total re-
flection (ATR) spectrum for nitrobenzene obtained at 303 K is also
shown in the figure.
Surface defects as active reduction site. The active site for
photocatalytic reduction of nitroaromatics on rutile TiO2 is the
exposed Ti3+ atoms located at the surface defect (Scheme 1).10
The Ti3+ atoms behave as adsorption sites for nitroaromatics
and trapping sites for photoformed e−, thus promoting rapid
nitro-to-amine reduction. This suggests that the number of
surface defect is probably an important factor affecting the
photocatalytic activity of rutile TiO2. The number of surface
defect can be determined by DRIFT analysis of nitrobenzene
adsorbed onto the catalysts in the gas phase.10 Respective rutile
particles (50 mg) were placed in a DR cell and evacuated (0.9
Pa) at 423 K for 3 h. Excess amount of nitrobenzene (16.5
ꢀmol) was introduced to the cell at 303 K and left for 1 h. The
cell was then evacuated (0.9 Pa) for 1 h to remove the physical-
ly-adsorbed nitrobenzene. Figure 7 shows the DRIFT spectra of
the nitrobenzene adsorbed onto the respective rutile TiO2. Two
●
+
radical cation of methyl viologen (MV ), and its intensity in-
creases with the photoirradiation time. During the photoreac-
tion, the photoformed h+ oxidizes TEOA, while the e– reduces
the Ti4+ atoms located at the surface or inner defects, producing
Ti3+ atoms.
Ti4+ (defect site) + e– → Ti3+
(3)
These accumulated e– (Ti3+ sites) reduces MV2+ and produces
●
+ 29
MV .
–
1
–
Ti3+ + MV2+ → Ti4+ + MV●
(4)
+
absorption bands appear at 1522 and 1346 cm . The 1522 cm
band is assigned to the asymmetric stretching vibration (νasym
of nitro group adsorbed onto the surface Ti–OH group, which
is inactive for nitro reduction.10 In contrast, the 1346 cm–1
band is assigned to the symmetric stretching vibration (νsym) of
nitro group adsorbed onto the surface Ti3+ atoms,26 which is the
1
)
The amount of MV● produced, therefore, reflects the total
number of Ti4+ atoms located at the defects on the particles. As
shown in Figure 8, the increase in 606 nm band is saturated by
+
>36 h photoirradiation. The saturated amount of MV● is iden-
+
ACS Paragon Plus Environment

Page 7 of 10
ACS Catalysis
1
2
3
4
5
6
7
8
9
the recombination with h+. This suggests that the smaller num-
tical to the total number of defects on the catalyst (Ntotal). Table
ber of inner defects promotes efficient charge separation be-
tween e– and h+; in contrast, the larger number of surface de-
fects provides a larger number of active sites for reduction. As a
result of this, the catalyst with a smaller number of inner de-
fects and a larger number of surface defects shows higher activi-
ty (eq. 5), exhibiting linear relationship between vf and
Nsurface/Ntotal (Figure 5e). The higher photocatalytic activity of
HF(10)/P25 particles is therefore due to their higher
Nsurface/Ntotal value.
1 summarizes Ntotal of respective rutile TiO2. The Ntotal values for
the samples 5 and 8 are 253 and 22 ꢀmol g–1, which are very
similar to the reported values (242 and 21 ꢀmol g–1),30 indicat-
ing that the obtained Ntotal values are reliable. Figure 5c and d
plot vf versus Ntotal or the number of inner defects (Ninner = Ntotal
– Nsurface). There is almost no relationship between these factors,
suggesting that the number of neither total defects nor inner
defects is the sole factor affecting the photocatalytic activity of
rutile TiO2.
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
180
Scheme 2. Proposed Reaction and Deactivation Processes
Occurring on the Photoactivated Rutile TiO2.
120
36 h
1
60
0
0
12 24 36 48
Time / h
0
recombination
surface defect
500
600
λ / nm
700
800
e-
recombination
Figure 8. Time-dependent change in the absorption spectra of an
aqueous solution obtained after photoreaction with TEOA and PT-
101 catalyst (sample 7) followed by reaction of MV2+ (1 h). (Inset)
h+
inner defect
change in the amount of formed MV● with the photoirradiation
+
time.
Distribution of defects on TiO2 particles. The larger
Nsurface/Ntotal value for the HF(10)/P25 particles is due to their
structural specificity originated from the synthesis method, "dry
process" (flame pyrolysis of gaseous TiCl4). These particles con-
tain a much smaller number of inner defects than commercial-
ly-available rutile particles, while containing a relatively large
number of surface defects. As reported,31 the number of defects
on the commercially-available rutile particles synthesized by "wet
process" (hydrolysis of Ti precursors followed by calcination)
depends on the particle size. Figure 9a shows the relationship
between the particle size and Ntotal for respective rutile TiO2. For
commercially-available TiO2 (samples 5–9), Ntotal decreases with
an increase in the particle size. This means that the defect den-
sity in the particles decreases with the particle size increase.
Larger TiO2 particles are produced by calcination at higher
temperature. During calcination, coalescence of small particles
creates larger particles, along with a disappearance of defects by
sintering. This thus creates larger particles with lower defect
density (Figure 9a).31 In contrast, as shown by the red symbol,
Ntotal for HF(10)/P25 (sample 3) does not follow this relation-
ship, and the value is much lower than that expected from the
particle size. In addition, as shown by the blue symbol, Ntotal for
HF(15)/P25 (sample 4) is also very low. These findings suggest
that the rutile particles contained in P25 contain a much small-
er number of defects than commercially-available rutile TiO2
synthesized by wet process; in other words, these particles con-
tain a much smaller number of deactivation sites for recombi-
nation of photoformed e– and h+.
Effect of inner and surface defects on the activity. The
above results suggest that the high photocatalytic activity of
HF(10)/P25 cannot be explained by the number of surface,
inner, or total defects. The surface defects behave as the active
sites for reduction, whereas both surface and inner defects be-
have as the deactivation sites that promote recombination of
photoformed e– and h+. The photocatalytic activity of rutile
TiO2 can therefore be explained by the number of surface de-
fects relative to the number of total defects (Nsurface/Ntotal). As
shown in Figure 5e, the plot of vf versus Nsurface/Ntotal clearly
shows linear relationship, meaning that the catalyst with a
smaller number of inner defects and a larger number of surface
defects exhibits higher photocatalytic activity. This suggests that
photocatalytic activity of rutile TiO2 depends on the number of
active sites relative to the number of deactivation sites on the
catalyst, as follows:
Nsurface
Ntotal
the number of active site
(5)
vf ∝
≈
the number of deactivation site
As depicted in Scheme 2, the e– and h+ photoformed within
TiO2 migrate through the materials and appear on the surface.27
During the migration, the e– is trapped by the inner defect and
deactivated by the recombination with h+.27 In contrast, the e–
successfully trapped by the surface defect promotes nitro reduc-
tion, whereas the parts of these Ti3+ sites are also deactivated by
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 10
1
2
3
4
5
6
7
P25 with higher concentration of HF decomposes surface struc-
ture and decreases the number of surface defects. The result is
consistent with the activity data (Figure 5e).
Figure 9b shows the relationship between Nsurface and Ninner for
respective rutile TiO2. As reported,31 for commercially-available
rutile particles (samples 5–9), there is a positive correlation
8
9
between these factors; Nsurface increases with an increase in Ninner
.
The above data suggest that the rutile particles isolated from
P25 possess a defect distribution different from that of rutile
TiO2 synthesized by wet process. The large Nsurface/Ntotal charac-
ter specific for the isolated particles facilitates efficient nitro-to-
amine reduction. The dry process (flame pyrolysis of gaseous
TiCl4) definitely creates such a specific defect distribution and
produces rutile particles with high Nsurface/Ntotal value. It is, how-
ever, unclear yet how the specific defect distribution is created
on the particles by the dry process. Nevertheless, the obtained
findings clearly suggest that photocatalytic hydrogenation of
nitroaromatics occurs very efficiently on the rutile particles with
larger Nsurface/Ntotal value, and the dry process successfully creates
rutile particles with a desirable defect distribution.
The larger increase in Nsurface than Ninner is because the number
of surface defect increases significantly with an increase in the
surface area, associated with a particle size decrease. The posi-
tively correlated data suggest that both Nsurface and Ninner for
commercially-available rutile TiO2 increase with a particle size
decrease.31 This means that smaller particles contain a larger
number of surface defects but also contain a larger number of
inner defects. This trade-off relation therefore results in lower
Nsurface/Ntotal values and exhibits insufficient photocatalytic activ-
ity for commercially-available TiO2 (Figure 5e).
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
a)
300
5
CONCLUSION
200
We found that rutile TiO2 particles, isolated from P25 by HF
treatment, promote highly efficient photocatalytic hydrogena-
tion of nitroaromatics, with an activity higher than that of
commercially-available rutile TiO2 synthesized by the wet pro-
cess. The activity of rutile particles depends on the number of
defects on the particles. The inner defect behaves as the deacti-
vation site for recombination of photoformed e– and h+ pairs,
whereas the surface defect behaves as the active reduction site
for nitroaromatics as well as the deactivation site. The catalyst
with a larger number of surface defects relative to the number
of total defects (Nsurface/Ntotal) therefore exhibits higher activity.
The P25 TiO2 created by the dry process contains rutile parti-
cles with a high Nsurface/Ntotal ratio. This thus promotes efficient
charge separation of the photoformed e– and h+ pairs and facili-
tates rapid nitro-to-amine reduction on the rutile particles iso-
lated from P25 TiO2. The recovery of rutile particles after HF
treatment of P25 is only ca. 8%, which is one of the drawbacks
for practical applications. Nevertheless, the results presented
here may contribute to the design of more active photocatalyst
for reduction reactions and may help open a new strategy to-
wards green organic synthesis based on photocatalysis.
7
100
3
6
4
8
9
0
0
100
200
300
Particle size / nm
b)
200
5
100
7
3
6
4
8
9
0
0
50
100
Ninner / ꢀmol g−1
Figure 9. Relationship between (a) particle size and Ntotal and (b)
Ninner and Nsurface for respective rutile TiO2 (sample 3–9). The sam-
ple numbers correspond to those listed in Table 1.
EXPERIMENTAL SECTION
Materials. All regents used were purchased from Wako, To-
kyo Kasei, and Sigma-Aldrich, and used without further purifi-
cation. Water was purified by the Milli Q system. P25 TiO2
(JRC-TIO-4) was kindly supplied from the Catalyst Society of
Japan. HF(x)/P25 was prepared as follows: P25 particles (1 g)
were stirred in water (50 mL) with different concentrations of
HF (x wt %) for 24 h at 298 K. The resulting powders were
washed with water until the pH of solution becomes ca. 7 and
dried in vacuo for 12 h.
In contrast, as shown by the red symbol (Figure 9b), the data
for HF(10)/P25 (sample 3) does not follow the tendency for
commercially-available TiO2. Its Nsurface is very large although the
Ninner is very small. This means that these particles contain a
relatively large number of surface defects despite a small num-
ber of inner defects, thus resulting in larger Nsurface/Ntotal value
(Figure 5e). The smaller number of inner defects suppresses
recombination of e– and h+, and the relatively large number of
surface defects facilitates efficient nitro reduction, promoting
highly efficient nitro reduction on the HF(10)/P25 catalyst. As
shown by the blue symbol (Figure 9b), Nsurface of HF(15)/P25
(sample 4) is lower than HF(10)/P25, because the treatment of
Photoreaction. Each respective substrate was dissolved in 2-
PrOH solution. The solution and catalyst were added to a Pyrex
glass tube (ϕ 12 mm; capacity, 20 mL), and the tube was sealed
with a rubber septum cap. The catalyst was dispersed well by
ultrasonication for 5 min, and N2 was bubbled through the
solution for 5 min. The tube was photoirradiated with magnetic
ACS Paragon Plus Environment

Page 9 of 10
ACS Catalysis
1
2
3
4
5
6
7
8
9
stirring using a 2 kW Xe lamp (USHIO Inc.).32 The wavelengths
(7) Corma, A.; Concepción, P.; Serna, P. Angew. Chem., Int. Ed. 2007,
46, 7266–7269.
of the light irradiated to the solution are λ >300 nm due to the
light absorption by the Pyrex grass tube. The intensity of light at
300–450 nm, which may mainly promote photoexcitation of
TiO2, was determined to be 27.3 W m–2. The reaction was per-
formed without the control of temperature; the temperature of
solution during photoirradiation was ca. 303 K. After the reac-
tion, the catalyst was recovered by centrifugation, and the re-
sulting solution was analyzed by GC-FID. The substrate and
product concentrations were calibrated with authentic samples.
Analysis was performed at least three times and the errors were
±0.2%. It is noted that the solubility of substrate and products
in solution is sufficiently high, and the amount of them ad-
sorbed onto the catalyst is negligible.
(8) Boronat, M.; Concepcion, P.; Corma, A.; González, S.; Illas, F.;
Serna, P. J. Am. Chem. Soc. 2007, 129, 16230–16237.
(9) Mitsudome, T.; Mikami, Y.; Matoba, M.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. Angew. Chem., Int. Ed. 2012, 51, 136–139.
(10) Shiraishi, Y.; Togawa, Y.; Tsukamoto, D.; Tanaka, S.; Hirai, T.
ACS Catal. 2012, 2, 2475–2481
(11) Tanaka, K.; Capule, M. F. V.; Hisanaga, T. Chem. Phys. Lett.
1991, 187, 73–76.
(12) Ohtani, B.; Prieto-Mahaney, O. O.; Li, D.; Abe, R. J. Photochem.
Photobiol. A 2010, 216, 179–182.
(13) Papageorgiou, A. C.; Beglitis, N. S.; Pang, C. L.; Teobaldi, G.;
Cabailh, G.; Chen, Q.; Fisher, A. J.; Hofer, W. A.; Thornton, G. Proc.
Natl. Acad. Sci. USA 2010, 107, 2391–2396.
(14) Rodriguez, J. A.; Jirsak, T.; Liu, G.; Hrbek, J.; Dvorak, J.; Maiti,
A. J. Am. Chem. Soc. 2001, 123 9597–9605.
(15) Dohnálek, Z.; Lyubinetsky, I.; Rousseau, R.; Prog. Surf. Sci. 2010,
85, 161–205.
(16) Kirchnerova, J.; Herrera Cohen, M.-L.; Guy, C.; Klvana, D.
Appl. Catal. A: General 2005, 282, 321–332
(17) Sachtleben Chemie GmbH, June 24, 1994.
(18) Ohno, T.; Sarukawa, K.; Matsumura, M. J. Phys. Chem. B 2001,
105, 2417–2420.
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
Analysis. XRD analysis was carried out on a Philips X’Pert-
MPD spectrometer.33 TEM observations were carried out using
an FEI Tecnai G2 20ST analytical electron microscope operated
at 200 kV.34 DLS analysis was performed on a Horiba LB-500
dynamic light-scattering particle size analyzer.35 The measure-
ment was carried out with water containing respective TiO2
particles (20 mg/L), after ultrasonication for 5 min. DRIFT
spectra were measured on a FT/IR 610 system equipped with a
DR-600B in situ cell (JASCO Corp.). DR UV−vis spectra were
measured on an UV−vis spectrophotometer (JASCO Corp.; V-
550 equipped with Integrated Sphere Apparatus ISV-469) with
BaSO4 as a reference.
(19) Ramis, G.; Busca, G.; Cristiani, C.; Lietti, L.; Forzatti, P.;
Bregani, F. Langmuir 1992, 8, 1744–1749.
(20) Pillai, S. C.; Periyat, P.; George, R.; McCormack, D. E.; Seery,
M. K.; Hayden, H.; Colreavy, J.; Corr, D.; Hinder, S. J. J. Phys. Chem. C
2007, 111, 1605–1611.
(21) Yu, J. C.; Yu, J.; Zhao, J. Appl. Catal. B 2002, 36, 31–43.
(22) Kitano, M.; Iyatani, K.; Tsujimaru, K.; Matsuoka, M.; Takeuchi,
M.; Ueshima, M.; Thomas, J. M.; Anpo, M. Top. Catal. 2008, 49, 24–
31.
(23) Ohno, T.; Sarukawa, K.; Tokieda, K.; Matsumura, M. J. Catal.
2001, 203, 82–86.
(24) Bickley, R. I.; Gonzalez-carreno, T.; Lees, J. S.; Palmisano, L.;
Tilley, R. J. D. J. Solid State Chem. 1991, 92, 178–190.
(25) Kuang, Q.; Yang, S. ACS Appl. Mater. Interfaces 2013, 5, 3683–
3690.
(26) Ahmad, I.; Dines, T. J.; Rochester, C. H.; Anderson, J. A. J.
Chem. Soc., Faraday Trans. 1996, 92, 3225–3231.
(27) Kong, M.; Li, Y.; Chen, X.; Tian, T.; Fang, P.; Zheng, F.; Zhao,
X. J. Am. Chem. Soc. 2011, 133, 16414–16417.
(28) Ikeda, S.; Sugiyama, N.; Murakami, S.; Kominami, H.; Kera, Y.;
Noguchi, H.; Uosaki, K.; Torimoto, T.; Ohtani, B. Phys. Chem. Chem.
Phys. 2003, 5, 778–783.
(29) Watanabe, T.; Honda, K. J. Phys. Chem. 1982, 86, 2617–2619.
(30) Murakami, N.; Prieto Mahaney, O. O.; Abe, R.; Torimoto, T.;
Ohtani, B. J. Phys. Chem. C 2007, 111, 11927–11935.
(31) Yan, J.; Wu, G.; Guan, N.; Li, L.; Li, Z.; Cao, X. Phys. Chem.
Chem. Phys. 2013, 15, 10978–10988.
ASSOCIATED CONTENT
XRD patterns for TiO2 (Figure S1), DR UV-vis spectra for TiO2
(Figure S2), TEM images of HF(10)/P25 (Figure S3) and
HF(15)/P25 (Figure S4), size distribution of TiO2 particles (Figure
S5), Time-dependent change in the aniline formation during pho-
toreaction of nitrobenzene on respective TiO2 (Figure S6). This
material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
shiraish@cheng.es.osaka-u.ac.jp
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This work was supported by the Grant-in-Aid for Scientific
Research (No. 23360349) from the Ministry of Education,
Culture, Sports, Science and Technology, Japan (MEXT).
(32) Shiraishi, Y.; Sugano, Y.; Tanaka, S.; Hirai, T. Angew. Chem. Int.
Ed. 2010, 49, 1656–1660.
REFERENCES
(1) Downing, R. S.; Kunkeler, P. J.; van Bekkum, H. Catal. Today
1997, 37, 121–136.
(2) Burawoy, A.; Critchley, J. P. Tetrahedron 1959, 5, 340–351.
(3) Rylander, P. N. Catalytic Hydrogenation in Organic Synthesis, Aca-
demic Press, New York, 1979; p 122.
(4) Corma, A.; Serna, P.; Concepción, P.; Calvino, J. J. J. Am. Chem.
Soc. 2008, 130, 8748–8753.
(33) Tsukamoto, D.; Shiraishi, Y.; Sugano, Y.; Ichikawa, S.; Tanaka,
S.; Hirai, T. J. Am. Chem. Soc. 2012, 134, 6309–6315.
(34) Sugano, Y.; Shiraishi, Y.; Tsukamoto, D.; Ichikawa, S.; Tanaka,
S.; Hirai, T. Angew. Chem. Int. Ed. 2013, 52, 5295–5299.
(35) Shiraishi, Y; Tanaka, K.; Shirakawa, E.; Sugano, Y.; Ichikawa,
S.; Tanaka, S.; Hirai, T, Angew. Chem., Int. Ed. 2013, 52, 8304–8308.
(5) Reis, P. M.; Royo, B. Tetrahedron Lett. 2009, 50, 949–952.
(6) Corma, A.; Serna, P. Science 2006, 313, 332–334.
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 10
1
2
3
4
5
6
7
8
SYNOPSIS TOC
Rutile Crystallites Isolated from Degussa (Evonik) P25 TiO2: Highly Efficient Photocatalyst for Chemoselective Hydrogenation
of Nitroaromatics
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
Yasuhiro Shiraishi,* Hiroaki Hirakawa, Yoshiki Togawa, Yoshitsune Sugano, Satoshi Ichikawa, and Takayuki Hirai
ACS Paragon Plus Environment