Photocatalytic oxidation of toluene to benzaldehyde over anatase TiO 2 hollow spheres with exposed {001} facets

Article abstract of DOI:10.1016/j.catcom.2011.03.007

A new series of anatase TiO₂ hollow structures were prepared by a facile hydrothermal process. When the hydrothermal time was increased from 20 min to 72 h, the resulting TiO₂ solid spheres gradually transformed into TiO₂

Full text of DOI:10.1016/j.catcom.2011.03.007

Catalysis Communications 12 (2011) 946–950
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Photocatalytic oxidation of toluene to benzaldehyde over anatase TiO2 hollow
spheres with exposed {001} facets
⁎
Feng-Lei Cao, Jin-Guo Wang, Fu-Jian Lv, Die-Qing Zhang, Yu-Ning Huo, Gui-Sheng Li, He-Xing Li, Jian Zhu
Department of Chemistry, Shanghai Normal University, Shanghai 200234, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A new series of anatase TiO2 hollow structures were prepared by a facile hydrothermal process. When the
hydrothermal time was increased from 20 min to 72 h, the resulting TiO2 solid spheres gradually transformed
into TiO2 hollow spheres with higher surface crystallinity and exposed {001} facets. The as-prepared TiO2-
72 h sample exhibited the highest activity comparing to other TiO2-based samples and commercial product
Degussa P-25 towards the selective photocatalytic oxidation of toluene to benzaldehyde. Such great
photocatalytic performance was mainly attributed to enhanced UV-adsorption and better charge separation
efficiency due to higher surface crystallinity of TiO2-72 h.
Received 18 January 2011
Received in revised form 4 March 2011
Accepted 4 March 2011
Available online 14 March 2011
Keywords:
TiO2 hollow spheres
Exposed {001} facets
Photocatalysis
Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
Selective oxidation of toluene
1. Introduction
dominant {001} facets in selective photocatalysis was not investi-
gated so far.
Catalyst morphological and structural (both surface and bulk)
studies have attracted great interest since they are key to catalyst
design and performances [1–3]. Initiated by Lu group [4,5], many
recent investigations have been devoted to controlled synthesis
and potential applications of TiO2-based photocatalysts with
exposed {001} crystal facets. Such TiO2 catalysts with dominant
{001} facets can be divided into two major categories: micrometer
single crystals [6–11], and nanometer TiO2 crystals [12–14] that
can only be obtained under rigorous conditions. Both bulk crystals
and nano crystals are not ideal for heterogeneous catalysis since
micro crystals have relatively small surface area, while nano
crystals are difficult to handle. Microstructured particles consist
of nanostructured units have been expected to be a better model in
this sense.
In this work, we described a facile process to synthesize anatase
TiO2 hollow spheres based on a hydrothermal reaction between TiF4
and H2O. Structural characterizations showed that shells of such
hollow spheres were made of TiO2 nanocrystals with exposed {001}
facets. The catalytic performance of such TiO2 hollow spheres was
evaluated by using the selective photocatalytic oxidation of toluene to
benzaldehyde as a model reaction. The as-prepared TiO2 hollow
spheres exhibited higher photo-efficiency than commercial TiO2
Degussa P-25. Such high catalyst performance towards conversion
of toluene to benzaldehyde was attributed to higher UV absorbance
and lower recombination of free carriers of these hollow TiO2 spheres
with exposed {001} facets due to their unique high surface
crystallinity.
The TiO2 with dominant {001} facets exhibited superior
photoactivities in many reaction systems, such as water splitting
and pollutant removal [6,9,10]. Their intrinsic activity was mainly
focused on photocatalytic degradation of organic dyes [8,11,15].
Selective photocatalysis has developed into a promising area for
21st century organic chemistry[16] to synthesize organic com-
pounds by photocatalytic process, such as from ethylbenzene to
acetophenone, toluene to benzaldehyde, and cyclohexane to
cyclohexanol [13,16–19]. However, the performance of TiO2 with
2. Experimental
2.1. Catalysts preparations
In a typical run, 0.20 g of TiF4 was dissolved in 40 mL of distilled
water. After stirring for 10 min, a clear solution was obtained. The
solution was then transferred into a 50 mL of Teflon-lined stainless
steel autoclave. The autoclave was heated to 250 °C for a hydrother-
mal time ranging from 20 min to 72 h, then cooled to room
temperature. The samples were collected by centrifugation and
washed three times with alcohol. After being dried at 80 °C in air for
6 h, the resulting catalysts were labelled as TiO2-t where t denotes
different hydrothermal time.
⁎
Corresponding author. Tel./fax: +86 21 64322272.
E-mail address: jianzhu@shnu.edu.cn (J. Zhu).
1566-7367/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.03.007

F.-L. Cao et al. / Catalysis Communications 12 (2011) 946–950
947
Fig. 1. XRD patterns of TiO2 samples with different hydrothermal time: (a) 20 min;
Fig. 2. Raman spectra of TiO2 samples with different hydrothermal time: (a) 20 min;
(b) 40 min; (c) 6 h; (d) 12 h; and (e) 72 h.
(b) 40 min; (c) 6 h; (d) 12 h; and (e) 72 h.
2.2. Catalysts characterizations
ionization detector (FID). The GC-MAS (Agilent 6890 N/5973) was
used to confirm the products. Only benzyl alcohol and benzaldehyde
have been detected. The saturated Ca(OH)2 solution was used as
absorbance to collect carbon dioxide. NO precipitates were found after
reaction. The selectivity was calculated from the yield of benzyl
alcohol and benzaldehyde. Control experiments showed no measur-
able amounts of oxidation of toluene. Reproducibility of the experi-
ments was checked by repeating the runs at least three times and was
found to be within acceptable limits ( 3%).
X-ray diffractions of the catalysts were determined using XRD
(D/MAX-2000 with Cu Kα radiation). The crystal sizes were calculated
by applying the Scherrer equation. The morphologies were observed
by field emission scanning electron microscopy (FESEM, HITACHI
S4800), transmission electron microscopy (TEM), and selected area
electron diffraction (SAED) on a JEM-2100. Nitrogen adsorption–
desorption isotherms were measured at 77 K (Quantachrome NOVA
4000e), and the surface area (SBET) were calculated using the BET
method. The light absorption and photoluminescence spectra were
recorded by photoluminescence spectrum (PLS, Varian Cary-Eclipse
500) and UV–visible diffuse reflectance spectrum (DRS, MC-2530).
The Raman spectra were conducted on a Dilor Super LabRam II (Light
source, 650 nm). The diffuse reflectance FTIR spectroscopy (DRIFTS)
was performed on MAGNA IR 760.
3. Results and discussion
3.1. Catalyst characterization
Fig. 1 showed XRD patterns of various TiO2 samples with different
hydrothermal time. All TiO2 samples displayed well-defined charac-
teristic diffraction peaks of anatase phase corresponding to 2θ of 25.3°
(101), 37.8° (004), and 48.0° (200) [20]. Anatase crystals were easily
obtained after hydrothermal treatment at 250 °C for 20 min. With
increased hydrothermal time, the intensities of the diffraction peaks
were gradually enhanced, indicating better crystallinity of anatase
phase. As summarized in Table 1 [21,22], there was a clear trend
between crystal size and crystallinity.
2.3. Catalysts evaluations
Selective photocatalytic oxidations of toluene to benzaldehyde
were carried out in self-designed quartz tube-like reactor with length,
diameter and volume of 200 mm, 31 mm and 150 mL. The initial
solution contained 0.10 g of TiO2, 10 mL of deionized H2O, and
0.010 mL of toluene. The concentration of toluene was 9.4 mmol/L and
the total volume of the reaction solution was 10 mL. Six 6 W UV lamps
(310 nm) were used as light source. The distance between light and
reactor was 60 mm. After reacting for 2 h at 20 °C, the mixtures were
extracted with diethyl ether, then analysed on a gas chromatograph
(Shimadzu, GC-17A) equipped with a TC-1 column and a flame
The qualities of crystals were also investigated by Raman spectra. As
shown in Fig. 2, all TiO2 samples exhibited peak around 144 cm−1
,
which was ascribed to the Eg mode of anatase. This result was consistent
with the XRD data. Increasing the hydrothermal time from 20 min to 6 h
resulted in a negative shift toward the lower wave number, which
indicated a decrease of surface O to Ti atoms ratio possibly due to the
decreased surface oxygen vacancies [23]. Meanwhile, extending the
time from 6 h to 72 h showed no obvious shifts. To describe the relative
crystallinity of different samples, the Raman adsorption intensities
around 144 cm−1 were semi-qualitatively calculated as shown in
Table 1. The relative intensity of Raman showed similar trend to that of
the XRD data. However, small differences were noticed between
TiO2-20 min and TiO2-40 min samples. As observed in the XRD patterns,
the relative intensity of TiO2-40 min (50%) was much higher than that of
TiO2-20 min (29%). However, both samples had a similar Raman
intensity (6% and 7% for TiO2-20 min and TiO2-40 min, respectively). It
should be noted that Raman spectrum is surface sensitive, while XRD is a
representation of bulk material. Such discrepancy between the XRD and
Table 1
Physicochemical characteristics of various TiO2 samples.
Samples
SBET (m2/g)
VP (cm3/g)
Crystal size (nm)
Relative
intensitya (%)
XRD
Raman
TiO2-20 min
TiO2-40 min
TiO2-6 h
TiO2-12 h
TiO2-72 h
19
17
7.3
6.8
6.6
0.061
0.054
0.037
0.033
0.032
31
38
50
52
56
29
50
60
71
100
6
7
62
81
100
a
Relative intensity of the main peak in XRD and Raman spectra was calculated by
comparing different samples with TiO2-72 h.

948
F.-L. Cao et al. / Catalysis Communications 12 (2011) 946–950
Fig. 3. FESEM and TEM images of TiO2 samples with different hydrothermal time: (a) 20 min; (b) 40 min; (c) 6 h; (d) 12 h; and (e, f) 72 h. The inset of (f) is the texture of a shell
fragment and the corresponding top and side view of SAED patterns.
Raman data may imply an intrinsic crystallization process from the core
to the outside. Similar results related to the transformation from anatase
to rutile have been reported by Li group [24].
The exposed crystal facets of shell fragments were also investigated.
As shown in the inset of Fig. 3f, a square shape, which is indicative of
{001} facet, was found from the top view. The atomic planes of (200)
or (020) crystal faces with the d-spacing around 0.189 nm were
marked in SAED pattern, which was in good agreement with reported
data [13]. The side view TEM image showed that the individual
elongated fragments were single-crystalline with the long axes
parallel to the [001] face of the crystal, further confirming the
existence of exposed {001} facets on the external surfaces.
Such a structure evolution was also supported by the nitrogen
sorption experiment (Fig. S1). The N2 adsorption/desorption iso-
thermals of TiO2-20 min and TiO2-40 min ascribed to type IV
indicated a mesoporous structure, which was also clearly seen by
TEM. With extended growth time, the hysteresis loop (type H1)
gradually shrunk due to the loss of mesopores, which resulted in a
sharp decrease of the specific surface area (SBET). As shown in Table 1,
the SBET decreased from 17 to 7.3 m2/g when the reaction time was
increased from 40 min to 6 h. The significant decrease in SBET and pore
Fig. 3 showed the morphology evolution of TiO2 samples over time.
Uniform and smooth solid spherical particles (a) were obtained after
20 min. Partial quadrate pieces and (b) were formed on the external
surfaces after 40 min. With prolonged reaction time, the whole
external surfaces were covered by quadrate pieces, which then
gradually crystallized into decahedron blocks with enhanced degree
of crystallinity. Meanwhile, solid spheres transformed into hollow
structures(c–f). Interestingly, the size of these particles increased very
little during the transformation (from 1.1 μm to 1.3 μm when the
reaction time was extended from 20 min to 72 h). The SEM images
(Fig. 3b) also showed the crystalline pieces firstly formed on outside.
After forming the first layer of crystal pieces, the Ostwald reopening
process will drive the dissolution of inner TiO2 with low crystallinity
and formation of high crystalline TiO2 on outside. Similar Ostwald
ripening process has been reported previously by Zeng group [25].

F.-L. Cao et al. / Catalysis Communications 12 (2011) 946–950
949
a
b
72 h
12 h
λex= 280 nm
20 min
40 min
6.0 h
6.0 h
40 min
20 min
12 h
72 h
200
300
400
500
530 540 550 560 570 580 590
Wavelength (nm)
Wavelength (nm)
Fig. 5. Photocatalytic performances of TiO2 samples with different hydrothermal time.
Fig. 4. UV–Vis DRS (a) and PL (b) spectra of TiO2 samples with different hydrothermal
time.
catalyst. The tendency of surface crystallinity was well consistent with
the change of activity. By comparing with commercial sample, the
TiO2-72 h showed activity two times higher than P-25. Because that P-25
has higher surface area (45 m2/g) and crystallinity of {101} facets than
TiO2-72 h, the higher activity of TiO2-72 h may be induced by higher
crystallinity of {001} facets. We proposed that the crystallinity, especially
the surface crystallinity of TiO2, was more important than surface area in
determining its final activity in photocatalysis. To our best knowledge,
this was the first time that surface crystallinity of anatase crystals with
exposed {001} facets showed a profound influence on its photocatalytic
activities. It is also mentioned that TiO2-72 h can easily precipitate out of
solution under gravity in 10 min providing easier separation than other
reported nanocrystals.
volume was most likely due to the destruction of porous structure
initiated by the crystallization of small crystals.
UV–Vis DRS spectra (Fig. 4a) demonstrated that all TiO2 samples
were inactive in the visible range. While significant differences were
found in the UV region (b400 nm). The UV absorbance increased
significantly when the hydrothermal time was prolonged from 20 min
to 72 h. Considering the similar SBET and morphology of TiO2-6 h,
TiO2-12 h and TiO2-72 h, the difference was mainly attributed to
different degrees of anatase crystallinity. The low absorbance of
TiO2-20 min and TiO2-40 min further confirmed that surface crystal-
linity was possibly more important than bulk crystallinity in
determining UV absorbance, which was in agreement with the
above XRD and Raman analysis.
4. Conclusion
PL spectra of various TiO2 samples were also examined to
investigate the efficiencies of charge carrier trapping, migration and
transfer (Fig. 4b). As TiO2 crystals evolved over time, PL intensity
gradually decreased. TiO2-72 h sample displayed much weaker PL
than that of the others, suggesting a lower recombination rate
between photoelectrons and holes. Taking into account the fact that
TiO2-72 h has the highest surface and bulk crystallinity, we concluded
that the increased quantum efficiency of TiO2-72 h was due to a high
crystalline single-crystal structure on external surfaces with exposed
{001} facets. Similar results have shown that high crystallinity of
anatase accelerates the transfer of free carriers and limits the
recombination of free carriers [22,26].
In conclusion, a simple hydrothermal method has been developed
for the synthesis of TiO2 hollow spheres with exposed {001} facets.
The resulting TiO2-72 h was more active than other TiO2 samples and
commercial P-25 in photocatalytic conversion of toluene to benzal-
dehyde, which could be attributed to enhanced UV-adsorption and
charge separation efficiency due to its higher surface crystallinity of
anatase with exposed {001} facets.
Acknowledgments
This work was supported by the National Science Foundation of
China (Grant Nos. 20907032 and 21007040) and Shanghai Govern-
ment (Grant Nos. 09520715300, 07dz22303, 10QA1405300, S30406,
10YZ69 and 10160503200).
3.2. Efficiency evaluation
Appendix A. Supplementary data
The selective oxidation of toluene to benzaldehyde was selected as a
model reaction to examine the photocatalytic performances of TiO2
samples. The diffuse reflectance FTIR spectroscopy (DRIFTS) showed
obvious IR absorption in the spectral range of 1660–1760 cm−1, which
can be assigned to the C=O bonds, confirming the existence of
benzaldehyde on the surface of photocatalyst after reaction. As shown
in Fig. 5, the conversion of toluene gradually increased from 9.0% to 21%
when the hydrothermal time was increased from 20 min to 6 h. And all
samples showed 90+% selectivity for benzaldehyde. With the increase of
surface crystallinity, the obtained TiO2 samples showed activities in the
order of TiO2-20 minbTiO2-40 minbbTiO2-6 hbTiO2-12 hbTiO2-72 h.
From XRD and Raman spectra, it seemed that the crystallinity of catalyst
have been increased by extending solvothermal time. In addition, crystal
piece with exposed {001} facet were gradually formed on the surface of
Supplementary data to this article can be found online at
doi:10.1016/j.catcom.2011.03.007.
References
[1] H.X. Li, Z.F. Bian, J. Zhu, D.Q. Zhang, G.S. Li, Y.N. Huo, H. Li, Y.F. Lu, J. Am. Chem. Soc.
129 (2007) 8406.
[2] S.E. Skrabalak, J.Y. Chen, L. Au, X.M. Lu, X.D. Li, Y.N. Xia, Adv. Mater. 19 (2007)
3173.
[3] J.G. Yu, S.W. Liu, H.G. Yu, J. Catal. 249 (2007) 59.
[4] H.G. Yang, C.H. Sun, S.Z. Qiao, J. Zou, G. Liu, S.C. Smith, H.M. Cheng, G.Q. Lu, Nature
453 (2008) 638.
[5] H.G. Yang, G. Liu, S.Z. Qiao, C.H. Sun, Y.G. Jin, S.C. Smith, J. Zou, H.M. Cheng, G.Q. Lu,
J. Am. Chem. Soc. 131 (2009) 4078.
[6] D.Q. Zhang, G.S. Li, X.F. Yang, J.C. Yu, Chem. Commun. 45 (2009) 4381.

950
F.-L. Cao et al. / Catalysis Communications 12 (2011) 946–950
[7] X.Y. Ma, Z.G. Chen, S.B. Hartono, H.B. Jiang, J. Zou, S.Z. Qiao, H.G. Yang, Chem.
Commun. 46 (2010) 6608.
[16] G. Palmisano, V. Augugliaro, M. Pagliarob, L. Palmisano, Chem. Commun. 43
(2007) 3425.
[8] M. Liu, L.Y. Piao, L. Zhao, S.T. Ju, Z.J. Yan, T. He, C.L. Zhou, W.J. Wang, Chem.
Commun. 46 (2010) 1664.
[9] G. Liu, H.G. Yang, X.W. Wang, L.N. Cheng, J. Pan, G.Q. Lu, H.M. Cheng, J. Am. Chem.
Soc. 131 (2009) 12868.
[10] G. Liu, C.H. Sun, H.G. Yang, S.C. Smith, L.Z. Wang, G.Q. Lu, H.M. Cheng, Chem.
Commun. 46 (2010) 755.
[11] F. Amano, O.O. Prieto-Mahaney, Y. Terada, T. Yasumoto, T. Shibayama, B. Ohtani.
Chem. Mater. 21 (2009) 2601.
[17] C.B. Almquist, P. Biswas, Appl. Catal. A 214 (2001) 259.
[18] M.W. Xue, J.Z. Ge, H.L. Zhang, J.Y. Shen, Appl. Catal. A 330 (2007) 117.
[19] J.G. Lv, Y. Shen, L.M. Peng, X.F. Guo, W.P. Ding, Chem. Commun. 46 (2010) 5909.
[20] P.C.A. Alberius, K.L. Frindell, R.C. Hayward, E.J. Kramer, G.D. Stucky, B.F. Chmelka,
Chem. Mater. 14 (2002) 3284.
[21] B. Ohtani, Y. Ogawa, S.I. Nishimoto, J. Phys. Chem. B 101 (1997) 3746.
[22] J. Zhu, Z.F. Bian, J. Ren, Y.M. Liu, Y. Cao, H.X. Li, W.L. Dai, H.Y. He, K.N. Fan, Catal.
Commun. 8 (2007) 971.
[12] J.M. Li, D.S. Xu, Chem. Commun. 46 (2010) 2301.
[13] J. Zhu, S.H. Wang, Z.F. Bian, S.H. Xie, C.L. Cai, J.G. Wang, H.G. Yang, H.X. Li,
CrystEngComm 12 (2010) 2219.
[14] C.T. Dinh, T.D. Nguyen, F. Kleitz, T.O. Do, ACS Nano 3 (2009) 3737.
[15] N. Murakami, Y. Kurihara, T. Tsubota, T. Ohno, J. Phys. Chem. C 113 (2009) 3062.
[23] J.C. Parker, R.W. Siegei, Appl. Phys. Lett. 57 (1990) 943.
[24] J. Zhang, M.J. Li, Z.C. Feng, J. Chen, C. Li, J. Phys. Chem. B 110 (2006) 927.
[25] H.G. Yang, H.C. Zeng, J. Phys. Chem. B 108 (2004) 3492.
[26] H.X. Li, G.S. Li, J. Zhu, Y. Wan, J. Mol. Catal. A 226 (2005) 93.