Dependence of photocatalytic activity on particle size of a shape-controlled anatase titanium(IV) oxide nanocrystal

Article abstract of DOI:10.1016/j.molcata.2012.03.003

Decahedral anatase titanium(IV) oxide (TiO₂) with {1 0 1} and {0 0 1} exposed crystal faces was prepared by hydrothermal treatment of peroxo titanic acid (PTA) solution with polyvinyl alcohol (PVA) as a shape-control reagent. pH of the PTA solution and amounts of PVA and amorphous titania included in the PTA solution had a large influence on size and shape of the prepared particles, and particle width of the decahedral anatase TiO₂ was controllable between 25 and 60 nm. Photocatalytic activity of the decahedral anatase TiO₂ was examined in terms of the relationship between particle size and photocatalytic activity. Decahedral anatase TiO ₂ with particle width of ca. 40 nm showed excellent activity because of the optimized balance between efficient separation of redox sites and large specific surface area.

Full text of DOI:10.1016/j.molcata.2012.03.003

Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Dependence of photocatalytic activity on particle size of a shape-controlled
anatase titanium(IV) oxide nanocrystal
a
a
a
a,b,∗
Naoya Murakami , Shota Kawakami , Toshiki Tsubota , Teruhisa Ohno
a
Department of Applied Chemistry, Faculty of Engineering, Kyushu Institute of Technology, 1-1 Sensuicho, Tobata, Kitakyushu 804-8550, Japan
JST, PRESTO, 4-1-8 Honcho Kawaguchi, Saitama 332-0012, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Decahedral anatase titanium(IV) oxide (TiO ) with {1 0 1} and {0 0 1} exposed crystal faces was prepared
2
Received 2 November 2011
Received in revised form 22 February 2012
Accepted 4 March 2012
by hydrothermal treatment of peroxo titanic acid (PTA) solution with polyvinyl alcohol (PVA) as a shape-
control reagent. pH of the PTA solution and amounts of PVA and amorphous titania included in the PTA
solution had a large influence on size and shape of the prepared particles, and particle width of the dec-
ahedral anatase TiO2 was controllable between 25 and 60 nm. Photocatalytic activity of the decahedral
anatase TiO2 was examined in terms of the relationship between particle size and photocatalytic activ-
ity. Decahedral anatase TiO2 with particle width of ca. 40 nm showed excellent activity because of the
optimized balance between efficient separation of redox sites and large specific surface area.
Available online 13 March 2012
Keywords:
Photocatalyst
Titanium(IV) oxide
Decahedral anatase
Separation of redox sites
Acetaldehyde decomposition
©
2012 Elsevier B.V. All rights reserved.
1
. Introduction
has been intensively studied [3–10]. Our group has also proposed
that anatase TiO with specific exposed crystal faces exhibits higher
2
Crystal size is one of the most important properties influencing
photocatalytic activity because of spatial separation of redox sites
induced by different kinds of exposed crystal faces [11,12]. In our
performance of material function. Nano-sized particles have been
expected to exhibit excellent performance due to their large spe-
cific surface area and quantum size effect. For example, nano-sized
particles having a large numbers of adsorption sites have frequently
been used in photocatalytic reaction, which has been a promising
function for environmental remediation [1,2]. Photocatalytic reac-
tion is induced by excited electrons and positive holes, followed by
reduction and oxidation with surface-adsorbed species. However,
back reaction easily occurs on the nanoparticle surface because
reduction and oxidation proceeds in neighboring sites, and this
back reaction as well as recombination between excited electrons
and positive holes greatly decrease the efficiency of photocatalytic
reaction. Therefore, it is important to optimize the balance between
adsorption ability and utilization efficiency of photogenerated elec-
trons and positive holes by enlargement of the specific surface
area, an improvement of crystallinity by thermal treatment and
combined use with other materials.
previous study, we prepared decahedral anatase TiO particles with
2
{1 0 1} and {0 0 1} exposed crystal faces by hydrothermal treat-
ment of peroxo titanic acid (PTA) with polyvinyl alcohol (PVA) as a
shape-control reagent [12]. PTA species are known as a water-base
precursor for TiO2 with relatively high stability against hydroly-
sis over a wide pH range under ambient condition [13]. Moreover,
shape-control reagent, such as organic acid and hydrophilic poly-
mer, easily coordinate on the peroxio titanic acid species [12,14].
These properties are suitable for formation of specific exposed crys-
tal faces. The anatase TiO2 prepared in previous study showed
higher activity than that of P-25, which is a well-known commercial
TiO2 [12]. This is presumably because of separation of redox sites
induced by predominant reduction and oxidation over {1 0 1} and
{0 0 1} exposed crystal faces, which were determined by photode-
position method [11,12]. However, the prepared shape-controlled
anatase TiO2 had a primary particle size of ca. 25 nm, which is
thought to be insufficient for efficient separation of redox sites.
Thus, an increase in the particle size of a shape-controlled nanocrys-
tal will possibly enhance photocatalytic activity as a result of more
effective separation of redox sites (Fig. 1).
Recently, the dependence of photocatalytic activity on exposed
crystal faces of an anatase titanium(IV) oxide (TiO ) nanocrystal
2
Here, we report a method for preparation of a decahedral
anatase TiO2 nanocrystal with larger particle size than that of
TiO2 prepared in our previous study [12]. Precursor conditions,
∗ Corresponding author at: Department of Applied Chemistry, Faculty of Engineer-
ing, Kyushu Institute of Technology, 1-1 Sensuicho, Tobata, Kitakyushu 804-8550,
Japan. Tel.: +81 93 884 3318; fax: +81 93 884 3318.
E-mail address: tohno@che.kyutech.ac.jp (T. Ohno).
i.e., pH of the PTA solution (pHPTA) and amounts of PVA (MPVA)
1
381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2012.03.003

N. Murakami et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
107
Hydrothermal treatment with magnetic stirring was carried out
using a hydrothermal reactor with a magnetic stirrer (Sanaikagaku
Co., RDV-TMS-100 and HM-19G-U). After hydrothermal treatment,
the residue in the Teflon bottle was washed with milli-Q water
−
1
until ionic conductivity of the supernatant was <10 ␮S cm . The
◦
particles were dried under reduced pressure at 60 C for 12 h. The
◦
obtained powders were thermally treated at 350 C for 2 h in order
to remove residual organic compounds on the TiO2 particles.
2.2. Characterization
The crystal structures of the TiO2 powders were characterized
with an X-ray diffractometer (Rigaku, MiniFlex II) with Cu K␣ radi-
ation (␭ = 1.5405 A˚ ). Primary particle size for the prepared samples
(
dhkl) were estimated from peaks attributed to anatase hkl in XRD
0.9ꢀ
patterns using the Scherrer equation:dhkl = ˇ
,where ꢀ is
cos ꢁ
hkl
B,hkl
Fig. 1. Schematic image of photocatalytic reaction over octahedral anatase with
different particle sizes.
the wavelength of X-rays, ꢁB,hkl is the Bragg angle and ˇhkl is the
full width at half maximum corrected by the full width at half
maximum of silicon powders. The specific surface areas (SBET) of
the particles were determined with a surface area analyzer (Quan-
tachrome, Nova 4200e) by using the Brunauer–Emmett–Teller
equation (Table 1). The morphology of prepared TiO2 particles was
observed by using scanning electron microscopy (SEM; JEOL, JSM-
6701FONO). Average particles of width (Wave), length (Lave) and
aspect ratio (rasp) were estimated by SEM observation. The crystal
size values of Wave and Lave showed linear relationship with d1 0 1
and d2 0 0, and d0 0 4, which were determined by the XRD peaks using
Scherrer equation. Due to the difficulty in estimation of particle size
by SEM observation, average particle sizes for AMT-100 (TAYCA Co.)
and ST-01 (Ishihara Sangyo Co.) were estimated to be 6 nm and
and amorphous titania (Mam) included in the PTA solution, had a
large influence on size and shape of the prepared particles. More-
over, stirring during hydrothermal treatment was carried out for
improvement of particle size distribution and surface structure. In
the present study, photocatalytic activity of the prepared shape-
controlled anatase TiO2 was examined in terms of the relationship
between particle size and photocatalytic activity.
2
. Experimental
2.1. Sample preparation
8
nm from peaks attributed to anatase 1 0 1 in XRD patterns using
3
Milli-Q water (3.0 cm ) was added to 10 cm3 of titanium(IV)
the Scherrer equation.
ethoxide and 50 cm3 of ethanol with vigorous stirring, and the
mixture was stirred for 1 h at room temperature. The resulting pre-
cipitate was centrifugally separated from the solution and dried
under reduced pressure. One or three grams of obtained amor-
2.3. Photocatalytic decomposition of acetaldehyde
Photocatalytic activities of TiO samples were evaluated by pho-
2
3
phous titania was dispersed in 100 cm of milli-Q water and then
tocatalytic decomposition of acetaldehyde (AcH). Two hundred
milligrams of TiO2 powders, which have complete extinction of
the incident radiation, were spread on a glass dish, and the glass
irradiated by ultrasonication. 32 cm3 of 30% hydrogen peroxide
was added, and yellow PTA solution was obtained. After 3 h, pH
2
3
of the solution was adjusted to 7 by adding ammonium (NH ) solu-
dish (8.55 cm ) was placed in a 125 cm Tedlar bag (As one). Five
hundred parts per million of gaseous AcH was injected into the
Tedlar bag, and photoirradiation was performed at room tempera-
ture after the AcH had reached adsorption equilibrium. The gaseous
composition in the Tedlar bag was 79% N , 21% O , <0.1 ppm of
3
tion, and a certain amount of PVA (MW = 22,000) as a shape-control
reagent was added to the solution. Excess amount of solution was
◦
evaporated by stirring the solution at ca. 60 C, and then the solu-
tion was cooled down. Finally, pH and volume of the solution were
2
2
3
adjusted to 7, 9, 10 or 11 and to 70 cm , respectively, by addition
carbon dioxide (CO ) and 500 ppm of AcH, and relative humid-
2
of NH3 solution.
ity was ca. 30%. A light-emitting diode (Nichia, NCCU033), which
emitted light at a wavelength of ca. 365 nm and an intensity of
100 ␮W cm , was used as the light source. The light intensity was
The solution in a Teflon bottle sealed with a stainless jacket
◦
−2
(
Sanaikagaku Co., HU-100) was heated at 200 C for 48 h in an oven.
Table 1
Summary of the preparation condition and average width (Wave), average length (Lave), aspect ratio (rasp) and specific surface area (SBET) of prepared samples.
2
Sample
Mam/g
MPVA/mg
pHPTA
Stirring
Wave/nm
Lave/nm
rave/nm
SBET/m /g
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
3
3
3
3
3
3
3
3
3
3
3
3
3
1
3
10
10
10
20
20
20
30
30
30
20
20
30
25
30
0
9
10
11
9
10
11
9
10
11
10
11
11
10
7
No
No
No
No
No
No
No
No
No
Yes
Yes
Yes
Yes
No
No
57.8
46.6
34.4
38.2
40.6
59.0
36.2
27.7
76.3
44.1
62.6
60.2
54.3
26.2
42.9
84.9
82.6
55.7
51.9
57.6
87.2
50.0
38.8
107.1
60.7
88.6
83.4
79.1
33.8
68.5
1.45
1.75
1.68
1.35
1.42
1.46
1.37
1.40
1.41
1.36
1.41
1.39
1.45
1.29
1.56
32
26
37
34
31
14
35
48
18
35
21
23
23
–
10
–

1
08
N. Murakami et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
Fig. 2. SEM images and histograms of particle width (red) and length (blue) obtained from SEM observation of the prepared anatase TiO2. (a, e) sample-n, (b, f) sample-e, (c,
g) sample-f and (d, h) sample-i. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

N. Murakami et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
109
Fig. 3. SEM images and histograms of particle width (red) and length (blue) obtained from SEM observation of the prepared anatase TiO2. (a, c) sample-b and (b, d) sample-o.
measured by an optical power meter (HIOKI, 3664). The concen-
trations of AcH and CO2 were estimated by gas chromatography
hydroxide in other studies [15,16]. These results indicate that PVA
in PTA solution induces formation of {1 0 1} exposed crystal faces
and retardation of growth in the [0 0 1] direction, which was pro-
moted under a high pH condition. Some reports suggests that
alcohol show more strongly adsorption to the {0 0 1} faces, which
depressed the growth rate along the [0 0 1] direction [17]. Thus, PVA
is thought to act in a similar way.
(
Agilent Technologies 3000, micro TCD detector) with PLOT-Q and
OV1 columns. AMT-100, AMT-600, TKP-102, TITANIX JA-1 (TAYCA
Co.) and ST-01, ST-21 and ST-41 (Ishihara Sangyo Co.) as commer-
cial anatase TiO powders and P-25 (Japan Aerosil Co.) as a standard
2
TiO2 powder were used.
On the other hand, a larger amount of PVA induced a nar-
rower distribution of particle widths and lengths though increase
of particle width and length was retarded to some extent. These
results indicate that PVA works not only as a regulator of crystal
growth in the [0 0 1] direction but also as a separator for prevention
of particle aggregation in a hydrothermal growth process. There-
fore, formation of well-defined faces requires appropriate MPVA and
pHPTA. Although higher pHPTA increased particle size up to ca. 80 nm
(Fig. 2(d, h)), the shapes and sizes of particles were not uniform.
This is because rapid hydrolysis of PTA to titania under a high pH
condition presumably induced formation of non-uniform particles
having no well-defined faces and a wide distribution of sizes.
Fig. 4 shows SEM images and histograms of particle width
and length obtained from SEM observation of TiO2 prepared by
hydrothermal treatment with magnetic stirring. Stirring during the
hydrothermal treatment did not induce a change in crystal struc-
ture but induced a decrease in particle size distribution. This was
clearly observed in TiO2 prepared under a high pHPTA condition and
large MPVA, and surface structure, i.e., well-defined exposed crystal
faces, was improved.
3
. Results and discussion
All of the prepared samples were attributed to single-phase
anatase TiO2 structure from results of XRD analysis. This coincided
with results of our previous study indicating that hydrothermal
treatment of PTA solution under a high pHPTA condition induced
preferential formation of anatase TiO2 structure [12]. However,
primary particle size and shape of the prepared TiO2 strongly
depended on pHPTA and MPVA. Fig. 2 shows SEM images and his-
tograms of widths and lengths of the particles obtained from SEM
observation of representative prepared samples. TiO2 prepared in
most conditions showed a decahedral shape with large {1 0 1} and
small {0 0 1} exposed crystal faces, judging from angles between
exposed crystal faces in SEM images. This shape was similar to that
of TiO2 prepared by hydrothermal treatment of PTA solution at pH
7
, which showed particle width of ca. 25 nm (Fig. 2(a, e)) [12]. How-
ever, anatase TiO2 particles prepared by hydrothermal treatment
of PTA solution with a lower ratio of MPVA to Mam under a higher
pHPTA condition had widths of more than 30 nm (Fig. 2(b, f), (c, g),
(
d, h)). Addition of a small amount of PVA (10 mg of PVA) under
Photocatalytic activity was evaluated by decomposition of
acetaldehyde under ultraviolet irradiation. Fig. 5 shows photocat-
alytic activity as an initial rate of CO2 generation against particle
widths of commercial and prepared anatase TiO2 samples. Photo-
catalytic activity for commercial anatase TiO2 samples showed a
monotonic increase with increase in particle size and saturation
a high pHPTA condition induced rod-like growth (Fig. 3(a, c)), sim-
ilar to that of the sample prepared without PVA addition (Fig. 3
(
b, d)). Side faces of the rod were presumably attributed to {1 0 0}
faces or zigzag surface of {1 0 1} faces [12]. Similar anatase rods
were prepared by PTA solution and aqueous solution of sodium

1
10
N. Murakami et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
Fig. 4. SEM images and histograms of particle width (red) and length (blue) obtained from SEM observation of the prepared anatase TiO2 under the condition of stirring
during hydrothermal treatment. (a, e) Sample-j, (b, f) sample-m, (c, g) sample-k and (d, h) sample-l.

N. Murakami et al. / Journal of Molecular Catalysis A: Chemical 358 (2012) 106–111
111
h)). However, even among anatase TiO samples with well-defined
2
faces, photocatalytic activity decreased with increase in particle
size from 40 nm to 60 nm. These results also indicate that well-
defined faces as well as specific surface area are important factors
for higher photocatalytic activity. Difference in surface area ratio
of {0 0 1} to {1 0 1} for decahedral anatase TiO2 might have an
influence on efficiency of photocatalytic reaction. However, slight
differences in the aspect ratio evaluated from SEM observation
(1.35–1.46) and the integrated intensity ratio of anatase 004 peak to
that of anatase 1 0 1 peak in XRD patterns (0.185–0.220) for decahe-
dral anatase TiO indicate that the difference in surface area ratio of
2
{
0 0 1} to {1 0 1} was not large enough to change the photocatalytic
efficiency.
4. Conclusions
In this paper, we discussed the relationship between photocat-
alytic activity and particle size of decahedral anatase TiO2 with
1 0 1} and {0 0 1} exposed crystal faces. Anatase TiO2 with well-
{
defined faces showed higher activity, and decahedral anatase with
particle width of ca. 40 nm showed excellent activity because of
the optimized balance between efficient separation of redox sites
and large specific surface area. The results of present study indicate
that change in particle size as well as change in surface area ratio of
reduction to oxidation sites of a shape-controlled nanocrystal has
the potential to control reaction properties.
Fig. 5. Initial rates of CO2 generation as photocatalytic activity (rCO2 ) against Wave
of (square) commercial TiO2, (blue open-triangle) particles with rasp of more than
1
.5, (red open-circle) particles with rasp of less than 1.5, (red filled-circle) samples
prepared with stirring, (green open-diamond) sample-n.
tendency with particle width of more than 40 nm. This result
can be explained by a trade-off relationship between particle size
(
crystallinity) and specific surface area (adsorption sites), though
Acknowledgements
photocatalytic activity depends also on other physical and chemi-
cal properties. Higher photocatalytic activity for decahedral anatase
with particle width of ca. 25 nm than that for commercial anatase
indicates that two kinds of exposed crystal faces, i.e., {1 0 1} and
This work was supported by the Knowledge Cluster Initiative,
Grant-in-Aid of Young Scientist (B) (22750139) implemented by
the Ministry of Education, Culture, Sports, Science and Technology
{
0 0 1} exposed crystal faces, induced efficient photocatalytic reac-
(MEXT), and the JST PRESTO program.
tion as a result of separation of redox sites.
Photocatalytic activity of decahedral anatase TiO2 increased
with increase in particle width up to ca. 40 nm. This is because redox
reaction proceeded more efficiently due to longer spatial separa-
tion of redox sites on larger particles. Electron diffusion length of
References
[1] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995)
69–96.
[
2] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C: Photochem. Rev.
1 (2000) 1–21.
more than a few micrometers in a polycrystalline anatase TiO film
2
[
18] indicates that electrons generated near oxidation sites ({0 0 1}
[3] F. Amano, T. Yasumoto, O.O.P. Mahaney, S. Uchida, T. Shibayama, B. Ohtani,
Chem. Commun. (2009) 2311–2313.
exposed crystal faces) can sufficiently migrate to reduction sites
(
particle size of ca. 40 nm.
[
4] F. Amano, O.O.P. Mahaney, Y. Terada, T. Yasumoto, T. Shibayama, B. Ohtani,
Chem. Mater. 21 (2009) 2601–2603.
{1 0 1} exposed crystal faces) in an anatase TiO2 nanocrystal with
[5] D. Zhang, G. Li, H. Wang, K.M. Chan, J.C. Yu, Cryst. Growth Des. 10 (2010)
130–1137.
6] F. Amano, T. Yasumoto, O.O.P. Mahaney, S. Uchida, T. Shibayama, Top. Catal. 53
2010) 455–461.
1
In contrast, rod-like anatase TiO2 prepared by hydrothermal
treatment of PTA solution with less MPVA showed lower activity
with increase in particle width. This might be due to lower reduc-
tion performance of side-exposed crystal faces ({1 0 0} faces or
zigzag surfaces of {1 0 1} faces). However, our group and other
groups have shown that anatase TiO2 with {1 0 0} faces or zigzag
faces of {1 0 1} faces and moderate {0 0 1} faces shows suffi-
ciently high photocatalytic activity [8,12,16]. Therefore, the lower
photocatalytic activity of rod-like anatase TiO2 is attributable to
not-optimized surface area ratio of reduction to oxidation sites [19].
Further increase in particle size (>ca. 40 nm) of decahe-
dral anatase deteriorated photocatalytic activity because of the
decrease in specific surface area as well as worse-defined faces,
resulting in less efficient separation of redox sites. Actually, anatase
TiO2 prepared under the condition of large MPVA and high pHPTA
with stirring showed a higher photocatalytic activity than that of
anatase TiO2 prepared without stirring because stirring during the
hydrothermal growth process improved surface structure (Fig. 2(b,
f) → Fig. 4(a, e), Fig. 2(c, g) → Fig. 4(c, g), Fig. 2(d, h) → Fig. 4(d,
[
(
[7] W.Q. Fang, X. Gong, H.G. Yang, J. Phys. Chem. Lett. 2 (2011) 725–734.
[8] J. Pan, G. Liu, G.Q. Lu, H. Cheng, Angew. Chem. Int. Ed. 50 (2011) 2133–2137.
[
9] J. Li, Y. Yu, Q. Chen, J. Li, D. Xu, Cryst. Growth Des. 10 (2010) 2111–2115.
[
10] T. Tachikawa, S. Yamashita, T. Majima, J. Am. Chem. Soc. 133 (2011)
7197–7204.
[11] T. Ohno, K. Sarukawa, M. Matsumura, New J. Chem. 26 (2002) 1167–1170.
12] N. Murakami, Y. Kurihara, T. Tsubota, T. Ohno, J. Phys. Chem. C 113 (2009)
062–3069.
[
3
[13] M. Kakihana, M. Tada, M. Shiro, V. Petrykin, M. Osada, Y. Nakamura, Inorg. Chem.
40 (2001) 891–894.
14] K. Tomita, V. Petrykin, M. Kobayashi, M. Shiro, M. Yoshimura, M. Kakihana,
Angew. Chem. Int. Ed. 45 (2006) 2378–2381.
15] Y. Gao, H. Luo, S. Mizusugi, M. Nagai, Cryst. Growth Des. 8 (2008) 1804–1807.
[16] J. Li, D. Xu, Chem. Commun. 46 (2010) 2301–2303.
[
[
[
[
[
17] J.H. Liao, L.Y. Shi, S. Yuan, Y. Zhao, J.H. Fang, J. Phys. Chem. C 113 (2009)
8778–18783.
1
18] S. Nakade, M. Matsuda, S. Kambe, Y. Saito, T. Kitamura, T. Sakata, Y. Wada, H.
Mori, S. Yanagida, J. Phys. Chem. B 106 (2002) 10004–10010.
19] N. Murakami, S. Katayama, M. Nakamura, T. Tsubota, T. Ohno, J. Phys. Chem. C
115 (2011) 419–424.