Langmuir 2008, 24, 9907-9915 9907

Effect of Calcination Temperature on the Structure of a Pt/TiO2 (B)

Nanofiber and Its Photocatalytic Activity in Generating H2
Chiu-Hsun Lin,*,† Jiunn-Hsing Chao,‡ Chun-Hsuan Liu,† Jui-Chun Chang,† and
Feng-Chieh Wang†
Department of Chemistry, National Changhua UniVersity of Education, Changhua 500, Taiwan, and
Nuclear Science and Technology DeVelopment Center, National Tsing Hua UniVersity, Hsinchu 300,
Taiwan

ReceiVed February 22, 2008. ReVised Manuscript ReceiVed April 23, 2008

Hydrogen trititanate (H2Ti3O7) nanofibers were prepared by a hydrothermal method in 10 M NaOH at 403 K,
followed by acidic rinsing and drying at 383 K. Calcining H2Ti3O7 nanofibers at 573 K led to the formation of TiO2
(B) nanofibers. Calcination at 673 K improved the crystallinity of the TiO2 (B) nanofibers and did not cause any change
in the morphology and dimensions of the nanofibers. TiO2 (B) and H2Ti3O7 nanofibers are 10-20 nm in diameter
and several micrometers long, but FE-SEM reveals that several of these nanofibers tend to bind tightly to each other,
forming a fiber bundle. Calcination at 773 K transformed TiO2 (B) nanofibers into a TiO2 (B)/anatase bicrystalline
mixture with their fibrous morphology remaining intact. Upon increasing the calcination temperature to 873 K, most
of the TiO2 (B) nanofibers were converted into anatase nanofibers and small anatase particles with smoother surfaces.
In the photocatalytic dehydrogenation of neat ethanol, 1% Pt/TiO2 (B) nanofiber calcined at 673 K was the most active
catalyst and generated about the same amount of H2 as did 1% Pt/P-25. TPR indicated that the calcination of 1% Pt/TiO2
(B) nanofiber at 573 K produced a poor Pt dispersion and poor activity. Calcination at a temperature higher than 773
K (in ambient air) resulted in an SMSI effect similar to that observed over TiO2 in the reductive atmosphere. As
suggested by XPS, such an SMSI effect decreased the surface concentration of Pt metal and created Ptδs sites,
preventing Pt particles from functioning as a Schottky barrier and leading to a lower activity. Because of the synergetic
effect between TiO2 (B) and anatase phases, the bicrystalline mixture, produced by calcining at 773 K, was able to
counter negative effects such as the reduction in surface area and the SMSI effect and maintained its photocatalytic
activity.

1. Introduction captures the photogenerated electrons and reduces the rate of

Because global oil reserves are expected to be depleted in a
few decades, alternative energy sources must be found to replace
the current petroleum-based energy system. Hydrogen gas will
play a vital role in the future energy system because it is a
renewable, clean-burning energy source. Accordingly, many
researchers have reported on the photocatalytic production of H2
from neat or aqueous aliphatic alcohol solution over platinized
titanium oxide. Under deaerated conditions, both primary and
secondary alcohols can be effectively dehydrogenated to alde-
hydes or ketones with the evolution of an equimolar amount of
H2 over a semiconductor photocatalyst suspended in the alcohol
solution.1–5 The most frequently investigated semiconductor
photocatalyst of this reaction is TiO2,1–5 but other nontitanium-
based photocatalysts such as CdS,6 ZnS,7 and TiSe28 have been
used. The purpose of loading TiO2 with Pt is to generate at the
interface between Pt and TiO2 a Schottky barrier, which effectively
* Corresponding author. E-mail: chlin@cc.ncue.edu.tw. Tel: 886-4-
7232105 ext. 3541. Fax: 886-4-7292361.
†
National Changhua University of Education.
‡
National Tsing Hua University.
(1) Zou, J.-J.; Liu, C.-J.; Yu, K.-L.; Cheng, D.-G.; Zhang, Y.-P.; He, F.; Du,
H.-Y.; Gui, L. Chem. Phys. Lett. 2004, 400, 520.
(2) Enea, O.; Ali, A.; Duprez, D. Int. J. Hydrogen Energy 1988, 13, 569.
(3) Sakata, T.; Kawai, T. Chem. Phys. Lett. 1981, 80, 341.
(4) Bamwanda, G. R.; Tsubota, S.; Nakamaru, T.; Haruta, M. J. Photochem.
Photobiol. 1995, 89, 177.
(5) Teratani, S.; Nakamichi, J.; Taya, K.; Tanaka, K. Bull. Chem. Soc. Jpn.
1982, 55, 1688.
(6) Jin, Z.; Li, Q.; Zheng, X.; Xi, C.; Wang, C.; Zhang, H.; Feng, L.; Wang,
H.; Chen, Z; Jiang, Z. J. Photochem. Photobiol., A 1993, 71, 85.
(7) Chen, L.; Zhu, X.; Wang, F.; Gu, W. J. Photochem. Photobiol., A 1993,
73, 217.
(8) Iseda, K.; Osaki, T.; Taoda, H.; Yamakita, H. Bull. Chem. Soc. Jpn. 1993,
66, 1038.
electron-hole recombination9 and promotes the formation of H2
gas. Numerous transition metals had been investigated as the
promoter for this photocatalytic dehydrogenation reaction, and
1 wt % Pt was the most effective one.5,10
C2H5OH f CH3CHO + H2
Anatase, rutile, and brookite are the three well-known
crystallographic forms of TiO2 in photocatalysis. In contrast,
very few works have addressed the photocatalytic properties of
TiO2 (B).11–13 TiO2 (B) has a monoclinic unit cell with a )
1.21787 nm, b ) 0.37412 nm, c ) 0.65249 nm, and β ) 107.054°
and a crystal structure that consists of two edge-sharing octahedra
that are linked to the neighboring doublet octahedra unit at their
corners.14,15 We and others12,13,16 have reported the preparation
of a TiO2 (B) nanotube utilizing a hydrothermal method reported
by Kasuga.17 This TiO2 (B) nanotube was an excellent photo-
catalyst, producing 20% more H2 than P-25 in the photocatalytic
dehydrogenation of neat ethanol.12 However, calcining a TiO2
(9) Linsebigler, A. L.; Lu, G.; Yates, J. T., Jr Chem. ReV. 1995, 95, 735.
(10) Ohtani, B.; Iwai, K.; Nishimoto, S.-I.; Sato, S. J. Phys. Chem. B 1997,
101, 3349.
(11) Zhu, J.; Zhang, J.; Chen, F.; Anpo, M. Mater. Lett. 2005, 59, 3378.
(12) Kuo, H.-L.; Kuo, C.-Y.; Liu, C.-H.; Chao, J.-H.; Lin, C.-H. Catal. Lett.
2007, 113, 7.
(13) Lin, C.-H.; Lee, C.-H.; Chao, J.-H.; Kuo, C.-Y.; Cheng, Y.-C.; Huang,
W.-N.; Huang, Y.-M.; Shih, M.-K. Catal. Lett. 2004, 98, 61.
(14) Marchand, R.; Brohan, L.; Tournoux, M. Mater. Res. Bull. 1980, 15,
1129.
(15) Feist, T. P.; Mocarski, S. J.; Davies, P. K.; Jacobson, A. J.; Lawandowski,
A. J. Solid State Ionics 1988, 28-30, 1338.
(16) Armstrong, G.; Armstrong, A. R.; Canales, J.; Bruce, P. G. Chem. Commun.
2005, 2454.
(17) Kasuga, T.; Hiramatsu, M.; Hoson, A. Langmuir 1998, 14, 3160.

10.1021/la800572g CCC: $40.75  2008 American Chemical Society

Published on Web 08/09/2008
9908 Langmuir, Vol. 24, No. 17, 2008
(B) nanotube at 673 K altered its morphology and phase
composition, which in turn affected the photocatalytic activity
of the nanotube. A solid fiber should be more resistant to thermal
calcination than a hollow nanotube and provided a broader
temperature range for investigation. The TiO2 (B) nanofiber can
be prepared using the same hydrothermal method to prepare the
TiO2 (B) nanotube, except that a higher hydrothermal temperature
should be used.18,19 Therefore, in this study, we prepared a
platinized TiO2 (B) nanofiber and examined, in more detail, the
effect of calcination temperature on the Pt promoter as well as
on the pore size and pore volume, morphology, crystalline phase
transformation, and band gap energy of the nanofiber.20,21 We
focused on the influence of the changes in physical and chemical
properties of the Pt/nanofiber caused by calcination on its ability
to produce H2 from neat ethanol.
2. Experimental Section
2.1. Photocatalyst Preparations. Anatase TiO2 powder (2.0 g,
Aldrich) was mixed with 600 mL of 10 M NaOH in a 1.0 L
perfluoroalkoxy container, and the mixture was kept at 403 K for
7 days. The hydrothermal temperature used here is 20 K higher than
that used in the preparation of the nanotube.12 The resulting slurry
was filtered to dryness by vacuum filtration to yield a paste. The
paste was then washed a few times in deionized water and then in
0.10 M HCl. The washing process was very thorough: 1.0 g of paste
was dispersed in 1000 mL of deionized water or 0.10 M HCl, and
the suspension solution was stirred at 300 rpm for 1.0 h using a
mechanical stirrer. Between washings, the suspension was always
filtered to dryness before the next washing. After it had been washed
with HCl, the paste was again washed in deionized water as described
above until no Cl- was present in the filtrate. The final paste was
then dried at 383 K for 24 h to produce the hydrogen trititanate
nanofiber. The titanate nanofiber was calcined between 573 and 873
K at a heating rate of 1 K min-1 for 3 h to yield the final TiO2 (B)
nanofiber.
An impregnation method was adopted to deposit Pt onto the TiO2
(B) nanofiber. The nanofiber was first mixed with 60 mL of deionized
water in a flask to generate a suspension solution. The calculated
volume of 1.0 mM H2PtCl6 (Showa) was added slowly to the
suspension and stirred for 2 h at room temperature. A rotary evaporator
with a water bath at 323 K was utilized to remove water from the
suspension solution. The resulting paste was dried at 383 K for 12 h
to yield a yellow powder, which was calcined between 573 and 873
K in air for 3 h to produce the catalyst precursors. The catalyst
precursor was then reduced in the flowing hydrogen at a rate of 30
mL min-1 at 423 K for 3 h to yield a dark-gray Pt metal-loaded
nanofiber. The reduction temperature of Pt was chosen on the basis
of the results of the temperature-programmed reduction experiments
(section 2.2). A physical mixture of Pt-loaded TiO2 (B) nanofiber
and the anatase phase was prepared to test the bicrystalline synergetic
effect. The nanofiber that was calcined at 873 K (pure anatase form)
was mixed thoroughly with that calcined at 673 K (pure TiO2 (B)
form) at different mole % values, and then Pt metal was deposited
onto this bicrystalline mixture following the same impregnation
procedure described above.
2.2. Photocatalyst Characterizations. The BET surface areas
of the photocatalysts were measured with Micromeritics ASAP 2010
using N2 gas at liquid-nitrogen temperature, and their pore size
distributions were determined by the BJH method. X-ray powder
diffraction patterns of calcined nanofibers were obtained using a
Shimadzu Laboratory-X XRD-6000 spectrometer with Fe KR
irradiation (λ ) 1.93604 Å). A JEOL JSM-6700F field-emission
scanning electron microscope (FE-SEM) was employed to observe
(18) Armstrong, A. R.; Armstrong, G.; Canales, J.; Bruce, P. G. Angew. Chem.,
Int. Ed. 2004, 43, 2286.
(19) Pavasupree, S.; Suzuki, Y.; Yoshikawa, S.; Kawahata, R. J. Solid State
Chem. 2005, 178, 3110.
(20) Yu, H.; Yu, J.; Cheng, B. Chemosphere 2007, 66, 2050.
(21) Yu, J.; Yu, H.; Cheng, B.; Trapalis, C. J. Mol. Catal. A 2006, 249, 135.
Lin et al.
the morphology of the photocatalyst at various calcination tem-
peratures. A Phillips Tecani 20 transmission electron microscope
coupled with an EDX detector (energy-dispersive X-ray) was adopted
to observe the fine structure of the prepared photocatalysts at an
accelerating voltage of 200 kV and to analyze their elemental
compositions.
Raman spectra were obtained using a 3D nanometer-scale Raman
PL microspectrometer (Tokyo Instruments, Inc.) to monitor the
formation of the TiO2 (B) phase during calcination. The diffuse-
reflectance UV-vis spectra of these photocatalysts were recorded
using a Shimazru 3560 UV-vis spectrometer that was equipped
with an integrating sphere to estimate the band gap energies of the
photocatalysts. The sodium ion content of these photocatalysts was
determined by neutron activation analysis using the neutron source
from the THOR nuclear reactor locating at Hsinchu, Taiwan, and
Na2CO3 was the calibration standard.
Temperature-programmed reduction experiments were performed
using a Micromeritics Autochem 2910. The instrument has a
cryocooler device that can lower the reactor temperature to 203 K
using cold N2 gas that is generated from liquid N2 and is used to
conduct subambient temperature TPR experiments (SAT-TPR). The
SAT-TPR procedure is as follows: After 100 mg of catalyst was
placed in a U-shaped quartz reactor, the instrument was purged with
Ar (99.999%) at a flow rate of 30 mL min-1 for 30 min at room
temperature. The reactor was then cooled to 223 K in the flowing
Ar. Once the baseline was stabilized, the SAT-TPR experiments
were initiated by switching the gas to 10% H2/Ar and simultaneously
ramping the temperature from 223 to 673 K at a heating rate of 5
K min-1.
2.3. Photocatalytic Activity Tests. An 80 mL quartz tube that
was sealed with a rubber septum was used as the photoreactor to
study the production of H2 gas from neat ethanol. The ethanol was
dried over molecular sieves 5A before use. Ten milligrams of
photocatalyst and 15 mL of neat ethanol were placed in the
photoreactor in an ice bath and purged with Ar for 20 min to
remove O2 that was dissolved in the ethanol. In these experiments,
the photoreactor was carefully placed at a fixed distance in front of
the UV lamps. The suspension solution was stirred with a magnetic
stirrer and irradiated with two 15 W UV lamps (λmax ) 352 nm,
Sankyo Denki) for various periods of time at ambient temperature.
Ten microliters of the gaseous products was sampled using a gastight
syringe and analyzed using a Varian 3300 GC that was equipped
with a TCD detector and a CP-Carbon PLOT P7 capillary column
with Ar as the carrier gas. The liquid products (0.4 µL) were analyzed
using an HP 6890 GC that was equipped with an FID detector and
an HP-5 5% phenyl methyl siloxane capillary column with N2 as
the carrier gas.
3. Results and Discussion
3.1. Characterization of Morphology and Microstructure.
Figure 1a displays the FE-SEM micrograph of the anatase TiO2
powder that was used to prepare the nanofiber. It indicates that
the starting material comprises particles of sizes 50-150 nm.
Figure 1b is the FE-SEM micrograph of the hydrothermal product
after washing in deionized water and HCl and drying at 383 K.
The micrograph in Figure 1b demonstrates that the hydrothermal
product consists of many fiberlike materials with diameters of
20-200 nm and lengths of several micrometers. The higher-
magnification FE-SEM micrograph in the inset of Figure 1b
reveals that these large-diameter fibers are not single fibers but
fiber bundles. The TEM micrograph in Figure 2a confirms that
the fiber bundle is actually composed of many smaller fibers
with outer diameters of 10-20 nm that are tightly bound to each
other. The lack of bright and dark contrast in the fibrous material
in the micrograph in Figure 2a verifies that it is a solid fiber
rather than a hollow tube. The HRTEM micrograph in Figure
2b shows that the HCl-washed nanofibers that are dried at 383
K have a lattice fringe of 0.90 nm; this value is close to the
Effect of Temperature on Pt/TiO2 (B) Structure Langmuir, Vol. 24, No. 17, 2008 9909

Figure 1. FE-SEM micrograph of (a) the anatase particles (the starting material) and (b) the nanofibers after washing with 0.10 M HCl and drying
at 383 K. The inset in b is a higher-magnification micrograph showing the fiber bundle.

Figure 2. TEM micrograph of (a) the fiber bundle after acidic washing and drying at 383 K, (b) the HRTEM micrograph showing the lattice fringe
of the nanofiber, and (c) the EDX spectrum of the nanofiber.

interplanar distance of the (001) crystal plane of hydrogen are consistent with EDX analysis: <100 ppm Na+ was present
trititanate (H2Ti3O7).22 The EDX spectrum in Figure 2c reveals in these acid-washed nanofibers.
that the material has only Ti and O elements and no Na ions are The FE-SEM micrographs in Figure 3a-d show the morphol-
present in the material. The results of neutron activation analysis ogy of the nanofibers calcined between 573 and 873 K. Increasing
9910 Langmuir, Vol. 24, No. 17, 2008 Lin et al.

Figure 3. FE-SEM micrographs of the nanofibers after acidic washing and calcining at (a) 573, (b) 673, (c) 773, and (d) 873 K.

the calcination temperature of these nanofibers from 383 to 673

K did not cause any noticeable change in morphology (cf. Figures
1b and 3a,b). However, as the calcination temperature was
increased to 773 K, the surface of the nanofiber became smoother;
this change was observed more clearly at the closed end of the
nanofiber (inset of Figure 3c). For some of the shorter nanofibers,
this smoothing effect became so significant that they looked like
round particles. This phenomenon was even more obvious in the
inset of Figure 3d as the calcination temperature was increased
to 873 K.
3.2. Characterization of Surface Area and Pore Structure.
Figure 4 presents the pore size distributions determined by the
BJH method of this fibrous material that was calcined between
383 and 873 K. The Figure reveals a wide range of pore
diameterssfrom a few nanometers to a maximum of between
40 to 90 nm for this material. Because this material is a solid
fiber, it has no pores. Therefore, the smaller pores (d < 10 nm)
observed in Figure 4 are the secondary pores that is formed by
the space between the nanofibers in the fiber bundle, and the Figure 4. Pore size distribution of the nanofibers after acidic washing
larger pores (d ) 40-90 nm) are those formed in the spaces and calcining at (a) 383, (b) 573, (c) 673, (d) 773, and (e) 873 K.
between the randomly arranged fiber bundles (Figures 1b and Table 1 summarizes the physical properties of these nanofibers
Figures 3 and their insets). When these nanofibers were calcined at different calcination temperatures, including BET surface area,
at 873 K, the volume of the smaller pores (d < 10 nm) shrunk total pore volume, and band gap energy. It is noted that loading
significantly, but that of the larger pores (d ) 40-90 nm) these nanofibers with Pt metal by impregnation does not reduce
decreased only slightly. The diminishing of the smaller pores is their surface areas and pore volumes. The results seem reasonable
probably related to the surface smoothing and formation of smaller because loading very small Pt particles (2 nm, section 3.5) onto
particles during high-temperature calcination at 873 K as revealed these solid nanofibers, which have only mesopores and
by FE-SEM. The surface area and pore size of the nanofiber that macropores, should not affect these physical properties.
was dried at 383 K were 99 m2 g-1 and 0.39 cm3 g-1, respectively, 3.3. Characterization of Phase Composition. Figure 5
falling to 43 m2 g-1 and 0.26 cm3g-1 after calcining at 873 K. displays the XRD spectra of the nanofiber after washing in
Effect of Temperature on Pt/TiO2 (B) Structure Langmuir, Vol. 24, No. 17, 2008 9911

Table 1. Physical Properties of Pure and 1% Pt/TiO2 (B) Nanofibers at Different Calcination Temperatures and Their H2 Production
Rates
catalysta surface area (m2 g-1) pore volume (cm3 g-1) H2 production rate (µmol h-1)d areal rate (µmol h-1 m-2)d band gap (eV)
HTNF383K 99 0.39 3.189
HTNF573K 84 (84)b 0.34 (0.46)b 147 1.75 3.190
HTNF673K 74 (78) 0.34 (0.36) 238 3.06 3.195
HTNF773K 59 (59) 0.31 (0.37) 225 4.03 3.195
HTNF873K 43 (45) 0.26 (0.35) 116 2.58 3.194
1% Pt/P-25c 55 0.19 220 4.00
a
HTNF383K means that the hydrogen titanate nanofiber was calcined at 383 K for 3 h. b The values in parentheses are for the photocatalysts doped with
1% Pt. c H2PtCl6/P-25 was calcined at 673 K for 3 h. d The H2 production rate is obtained over 10 mg of catalyst, and the areal rate is the H2 production
rate per unit surface area of catalyst.

Raman results suggested that the material in Figure 5a was a

Figure 5. XRD spectra of the nanofibers after acidic washing and calcining
at (a) 383, (b) 573, (c) 673, (d) 773, and (e) 873 K.
deionized water and 0.10 M HCl and calcining between 383 and
873 K. The weak, broad spectrum in Figure 5a was obtained
from the material that was dried at 383 K and is identical to those
presented elsewhere.21–24 Because the spectrum is broad,
identifying its crystallographic structure by XRD is rather difficult,
and the structure of the nanofiber is still a matter of dispute. The
crystallographic structures of this material as described in the
literature include dititanate (H2Ti2O4(OH)2),23,24 trititanate (H2Ti3-
O7),22,25 tetratitante (H2Ti4O8(OH)2), 26 and lepidocrocite
(HxTi2 - x/40x/4O4H2O where x ) 0.7 and 0 ) vacancy).27 Our
(22) Chen, Q.; Du, G. H.; Zhang, S.; Peng, L. M. Acta Crystallogr. B 2002,
58, 587.
(23) Zhang, M.; Jin, Z.; Zhang, J.; Gao, X.; Yang, J.; Li, W.; Wang, X.; Zhang,
Z. J. Mol. Catal. A 2004, 217, 203.
(24) Tsai, C.-C.; Teng, H. Chem. Mater. 2006, 18, 367.
(25) Sun, X.; Li, Y. Chem.sEur. J. 2003, 9, 2229.
(26) Nakahira, A.; Kato, W.; Tamai, M.; Isshiki, T.; Nishio, K.; Aritani, H.
J. Mater. Sci. 2004, 39, 4239.
(27) Ma, R.; Bando, Y.; Sasaki, T. Chem. Phys. Lett. 2003, 380, 577.
hydrogen trititanate (H2Ti3O7). (See below in this section.) This
fact is also consistent with the observed lattice fringe of the
material in HRTEM experiments (Figure 2b).
As the calcination temperature was increased from 383 to 573
K, the XRD spectrum changed (Figure 5b), indicating that a
crystallographic structural transformation had occurred. Notably,
the diffraction peaks in Figure 5a,b are quite broad, suggesting
the poor crystallinity of the materials in this temperature range.
Increasing the temperature to 673 K significantly improves the
XRD peak intensity (Figure 5c) and therefore the crystallinity
of the material, enabling its crystalline phase to be identified.
Comparing the XRD spectrum in Figure 5c to the JCPDS
diffraction database, the new crystalline phase that was formed
at 673 K was TiO2 (B) (JCPDS 35-0088). This fact was verified
by the observation of the lattice fringe of these nanofibers that
were calcined at 673 K. (See section 3.5 and the HRTEM
micrograph in Figure 9.) The phase transformation from the
H2Ti3O7 nanotube to the TiO2 (B) nanotube occurred at 573 K.12
When the calcination temperature was increased further to
773 K, another crystallographic structural transformation occurred
in which the anatase phase of TiO2 began to form in the calcined
mixture. This fact is inferred by comparing XRD spectra in Figure
5c,d and is verified by the Raman spectra in Figure S1 in
Supporting Information. The theoretical XRD peak intensity ratio
of the (110) plane to the (002) plane in the TiO2 (B) phase should
be close to 100/90 according to JCPDS 35-0088, but for the
mixture that was calcined at 773 K, this ratio in Figure 5d markedly
exceeds the theoretical value. Therefore, the material that was
calcined at 773 K was actually a bicrystalline mixture of TiO2
(B) and anatase phases. The formation of the anatase phase at
773 K is revealed more clearly by the Raman spectra described
in Supporting Information. Again, we noticed that the appearance
of anatase particles in the calcined TiO2 (B) nanotubes was
observed previously at a lower temperature of 673 K.12
Increasing the calcination temperature to 873 K caused the
anatase phase to become the dominating phase in the calcined
mixture, with a small amount of remaining TiO2 (B) phase. We
need to point out the synthesis conditions and the starting material
used to prepare TiO2 (B) nanofibers exert an influence on the
nature of TiO2. Bruce prepared his TiO2 (B) nanofiber using
anatase TiO2 and 15 M NaOH at a hydrothermal temperature of
443 K. The TiO2 (B) nanofibers obtained could withstand the
calcination at 873 K without undergoing a phase transformation.18
Pavasupree reported the preparation of the TiO2 (B) nanofiber
using a natural rutile sand and 10 M NaOH at 423 K. Such a TiO2
(B) nanofiber had a much lower surface area than ours (20 vs
74 m2 g-1at 673 K) and could be converted into a crystalline
anatase nanofiber only by calcining at 973 K.19
(28) Debeila, M. A.; Raphulu, M. C.; Mokoena, E.; Avalos, M.; Petranovskii,
V.; Coville, N. J.; Scurrell, M. S. Mater. Sci. Eng., A 2005, 396, 70.
9912 Langmuir, Vol. 24, No. 17, 2008 Lin et al.

Figure 6. Diffuse-reflectance UV-vis spectra of the nanofibers after Figure 7. SAT-TPR spectra of the 1% Pt-promoted nanofibers calcined
acidic washing and calcining at (a) 383, (b) 573, (c) 673, (d) 773, and at (a) 573, (b) 673, (c) 773, and (d) 873 K.
(e) 873 K and (f) UV-vis spectrum of P-25 TiO2. The inset is used to
measure the band gap energy for the nanofibers calcined at 673 K.

The variations of phase composition of Pt/TiO2 (B) nanofibers

with calcination temperature were similar to those of pure TiO2
(B) nanofibers stated above, and their XRD spectra are included
in Figure S2 in Supporting Information. Because Pt particles
were highly disperse (see the HRTEM micrograph in Figure 9a
for a sample calcined at 673 K), no Pt signal was observed in
the XRD spectra until the calcination temperature reached 873
K. At this temperature, the diffraction peak due to the Pt (111)
crystal plane was observed at 2Θ ) 51°.
3.4. Band Gap Energy. Because our laboratory has an interest
in using the TiO2 (B) nanofiber as a photocatalyst, the diffuse-
reflectance UV-vis spectra of the material were obtained. Figure
6 presents these spectra in which the absorption band edge is red
shifted as the calcination temperature is increased from 383 to
673 K and then blue shifted as the temperature is increased
further to 873 K. On the basis of the assumption that the
fundamental absorption of these TiO2 (B) nanofibers involves Figure 8. Pt signals in the XPS spectra of the 1% Pt-promoted nanofibers
calcined at (a) 573, (b) 673, (c) 773, and (d) 873 K.
an indirect transition, their band gap energies (Eg) are evaluated
by plotting (F(R)E)1/2 against E (where F(R) is the Kubelka-Munk
TPR spectra have three peaks: two positive peaks between 273
function and E is the photon energy) and extrapolating the straight-
and 323 K (the shoulder peak is marked with an arrow in Figure
line part of the curve to (F(R)E)1/2 ) 0 (inset in Figure 6).28 Table
8b,c) are H2 consumption peaks that are associated with the
1 presents the band gap energies obtained using this method. The
reduction of PtOx and an inverted peak between 331 and 337 K
variation in the band gap energies of these nanofibers with the
that is associated with H2 desorption by reverse H2 spillover on
calcination temperature is small. The value of Eg increases slightly
these TiO2 (B) nanofiber surfaces. Interestingly, the first two
from 3.189 eV at 383 K to 3.195 eV at 673 K and is relatively
PtOx reduction peaks were shifted to a lower temperature as the
constant at around 3.195 eV at higher calcination temperatures.
calcination temperature of the Pt catalyst precursor was increased.
Therefore, the band gap energies of the TiO2 (B) nanofiber
Furthermore, the PtOx reduction peaks from the material calcined
(calcined at 673 K) and the anatase nanofiber (calcined at 873
at 873 K are so broad that they are barely visible.
K) are almost identical and are similar to that obtained for the
The calcination of the Pt chloride precursor at a higher
anatase powder (3.2 eV).
temperature produces larger PtOx particles on the oxide sur-
3.5. Characterization of Pt Metal Particles. As explained
face.29–31 The first step in the reduction of PtOx particles involves
above, loading TiO2 with Pt metal creates a Schottky barrier that
the dissociation of hydrogen, which is facilitated by the proximity
captures the photogenerated electrons and thus enhances the
of the two hydrogen atom adsorption sites. Such adsorption sites
photocatalytic activity. Therefore, the particle size, the distribu-
are more easily found on a larger PtOx particle so that larger PtOx
tion, and the oxidation state of the Pt particles that are supported
particles are more easily reduced. Therefore, the PtOx reduction
on the TiO2 (B) nanofiber are expected to affect the photocatalytic
peaks shift to a lower reduction temperature as the calcination
activity. Temperature-programmed reduction (TPR) is an effective
temperature of the Pt precursor/nanofiber increases. However,
tool for characterizing the chemical environment of transition
PtOx on TiO2 begins to decompose into Pt metal and oxygen
metals on the solid surface.29 Figure 7 depicts the SAT-TPR
molecules by calcining at temperatures above 673 K.31 This fact
spectra of the catalyst precursors that contain 1% Pt. These SAT-
(30) Hwang, C.-P.; Yeh, C.-T. J. Mol. Catal. A 1996, 112, 295.
(29) Mcnicol, B. D. Catal. ReV. Sci. Eng. 1982, 24, 233. (31) Huizinga, T.; Grondelle, J.; van Prins, R. Appl. Catal. 1984, 10, 199.
Effect of Temperature on Pt/TiO2 (B) Structure
Figure 9. (a) HRTEM micrograph of the 3% Pt-promoted TiO2 (B) nanofibers calcined at 673 K and (b) the Pt particle size distribution.
may explain the smaller reduction peak in Figure 7d for the Pt
catalyst precursor that is calcined at 873 K, at which temperature
most of the PtOx is already decomposed into Pt metal. The broad
nature of the TPR peak observed at 873 K may be associated
with the inhomogeneous composition of the catalyst precursor
because it contains a mixture of Pt metal and PtOx. The area ratio
under these Pt reduction peaks from nanofibers that are calcined
between 573 and 873 K is 0.26:1.0:0.63:0.36. These TPR results
suggested that 673 K might be the optimal calcination temperature
for the preparation of a better-dispersed Pt/TiO2 (B) nanofiber
catalyst.
Pt metal is well known to promote the H2 spillover phenomenon
on the TiO2 surface at a temperature of less than 473 K.32,33 As
soon as PtOx was reduced to Pt metal, H2 molecules dissociated
into hydrogen atoms over the Pt metal surface and migrated onto
the TiO2 support. This process is reversible. As the temperature
was increased further, H atoms migrated from the TiO2 surface
back to the Pt metal surface to replenish the desorbed H2 molecules
in the reverse spillover process. The negative desorption peak
at 331-337 K is attributed to such a reverse H2 spillover process
over the Pt/TiO2(B) nanofiber surface.
Figure 8 depicts the Pt signal in the XPS spectra of the Pt/TiO2
(B) nanofibers, which were calcined in air at between 573 and
873 K and then in flowing H2 at 423 K to reduce Pt. The peaks
at 70.9 and 74.3 eV were 4f7/2 and 4f5/2 peaks of the Pt0 metallic
state reported on titanium oxide, respectively.34,35 The reverse
in the intensity ratio of 4f7/2 to 4f5/2 peaks implies that PtO species
may also be present on the TiO2 (B) nanofiber.36 As the calcination
temperature of the Pt/TiO2 (B) nanofibers was increased to 773
and 873 K, the intensities of the Pt signals were decreased, and
their binding energies were shifted downward systematically.
Interestingly, the binding energy of the 4f7/2 peak for the sample
calcined at 873 K was shifted to 69.8 eV, which was 1.1 eV
lower than that of the Pt0 state. The shift in Pt binding energy
is not due to the transformation from TiO2 (B) to anatase. Yu
et al. reported a similar low-binding-energy phenomenon in his
mesoporous Pt/TiO2 and Au/TiO2 nanoarchitectured materi-
(32) Li, X. S.; Li, W. Z.; Chen, Y. X.; Wang, H. L. Catal. Lett. 1995, 32, 31.
(33) Ebitani, K.; Hattori, H. Bull. Chem. Soc. Jpn. 1991, 64, 2422.
(34) Silvestre-Albero, J.; Sepulveda-Escribano, A.; Rodriguez-Reinoso, F.;
Anderson, J. A. J. Catal. 2004, 223, 179.
(35) Kim, K. S.; Winograd, N.; Davis, R. E. J. Am. Chem. Soc. 1971, 93, 6296.
(36) Wang, X.; Yu, J. C.; Yip, H. Y.; Wu, L.; Wong, P. K.; Lai, S. Y.
Chem.sEur. J. 2005, 11, 2997.
Langmuir, Vol. 24, No. 17, 2008 9913
als.36,37 They attributed the low binding energy of Pt and Au to
the existence of some surface metal atoms with a low coordination
number in their Ultrafine metal particles or to the well-known
SMSI effect between transition metals and TiO2.
In this case, the SMSI effect seems to be more likely. The
traditional strong interaction at the Pt and TiO2 interface occurred
only in a reducing gas atmosphere (H2), but Jin et al. recently
observed the SMSI effect in an oxidative atmosphere (O2).38,39
Their XPS results indicated that in an oxidizing environment Pt
atoms diffused into the TiO2 lattice to be oxidized to Pt2+, which
is substituted for Ti4+ or forms interstitial ions at a temperature
as low as 473 K. The TPR results herein indicate that when the
catalyst precursor (H2PtCl6/TiO2(B) nanofiber) was calcined at
a temperature that exceeded 773 K the surface PtOx decomposes
to produce metallic Pt, which diffuses into the TiO2 (B) lattice
in a manner similar to that reported by Jin. (Therefore, this may
be the another reason that a smaller PtOx reduction is observed
in Figure 7c,d.) Furthermore, our FE-SEM results revealed that
the surface of the TiO2 (B) nanofiber became smoother after it
was calcined at 773 K. This fact suggested that above this
temperature the TiO2 (B) surface became very mobile and was
reconstructed. During the reconstruction, some of the Pt atoms
might be buried underneath the TiO2 (B) surface similar to that
happening in the SMSI phenomenon. These experimental results
seem to be consistent with the observations of Yu and Jin and
explain the reduction in both the Pt intensity and the binding
energy of the Pt0 signal intensity in our XPS spectra for the
Pt/TiO2 (B) nanofiber calcined at temperatures higher than 773
K. These results also suggest 673 K is the optimum calcination
temperature for the preparation of the platinized TiO2 (B)
nanofiber photocatalyst.
Figure 9a presents the HRTEM micrograph of 3% Pt supported
on a TiO2 (B) nanofiber that was calcined at 673 K and reduced
at 423 K. The micrograph shows a TEM image of a nanofiber
decorated homogeneously with Pt nanoparticles. It can be seen
from the micrograph that the lattice fringe of the nanofiber has
an interlayer distance of 0.62 nm, which is close to the interplanar
distance of a (001) crystal plane in the TiO2 (B) phase. This fact
further verifies the XRD and Raman results that indicate that
(37) Yu, J.; Wang, C. X.; Wu, L.; Ho, W.; Zhang, L.; Zhou, G. AdV. Funct.
Mater. 2004, 14, 1178.
(38) Li, Q.; Wang, K.; Zhang, S.; Zhang, M.; Jin, Z. J. Mol. Catal. A 2006,
258, 83.
(39) Zhang, M.; Jin, Z.; Zhang, Z.; Dang, H. Appl. Surf. Sci. 2005, 250, 29.
9914 Langmuir, Vol. 24, No. 17, 2008
Figure 10. H2 yields at various reaction times over 10 mg of 1% Pt-
promoted TiO2 (B) nanofibers calcined at (a) 573, (b) 673, (c) 773, and
(d) 873 K (e) the H2 yields of 10 mg of 1% Pt/P-25.
these nanofibers indeed have a TiO2 (B) crystalline phase when
the material is calcined at 673 K. The size distribution of Pt
particles is shown in Figure 9b, and it shows that most of the
Pt nanoparticles on these TiO2 (B) nanofibers have a size between
1.5 and 2 nm.
3.6. Photocatalytic Activity of the Pt/TiO2 (B) Nanofibers.
The activity of the 1% Pt/nanofiber in generating H2 gas from
the ethanol in the photocatalytic dehydrogenation reaction was
investigated. H2 was the main gaseous product with trace amounts
of CO2 and CH4 as byproducts. In the liquid phase, acetaldehyde
was the principal product with a small amount of 1,1′-
diethoxyethane as the byproduct. The acetal was generated by
a secondary reaction between acetaldehyde and ethanol. This
photocatalytic reaction is summarized by the following reaction
paths, where h+ and e- denote a photogenerated hole and electron,
respectively.13
TiO2 + hν f h+ + e-
C2H5OH + 2h+ f CH3CHO + 2H+
2H++2 e- f H2
CH3CHO + 2C2H5OH f CH3CH(OC2H5)2 + H2O
CH3CHO f CH3COOH f CH4 + CO2
Figure 10 depicts the H2 yields at various reaction times for
1% Pt/nanofiber calcined at different temperatures. The H2 yields
for 1% Pt/P-25 are also included to compare the performance
of the photocatalysts herein. The slope of the linear time profile
in Figure 10 is the rate of H2 production, which is used to compare
the activity of the photocatalysts. Table 1 summarizes the H2
production rate of 10 mg of catalyst (in µmol h-1) and the areal
rate (in µmol h-1 m-2) for the Pt/nanofiber calcined at different
temperatures. The nanofiber calcined at 673 K has a higher H2
production rate (238 µmol h-1) than at other calcination
temperatures, and the rate also slightly exceeds that of P-25 (220
µmol h-1). These data demonstrate that a crystalline TiO2 (B)
nanofiber is a very active photocatalyst in the photocatalytic
dehydrogenation of ethanol. The H2 production rate for calcination
at 673 K (238 µmol h-1) significantly exceeds that at 573 K (147
µmol h-1) because of the improvement in crystallinity by
calcination at a higher temperature. However, lower H2 production
rates were also obtained at 773 K (225 µmol h-1) and at 873 K
(116 µmol h-1). This result is interpreted as the reduction in
Lin et al.
surface area of these nanofibers caused by calcination, leading
to lower photocatalytic activities. Nanofibers calcined at 873 K
have a pure anatase crystalline phase. Therefore, these H2
production rates also suggest that the TiO2 (B) nanofiber is more
active than the anatase nanofiber in the photocatalytic dehy-
drogenation of ethanol. Calcination has long been known to affect
markedly the crystalline phase composition and the crystallinity
and surface area of TiO2, altering the photocatalytic activities of
TiO2 photocatalysts.9,40,41 The change in the band gap energy
of the TiO2 (B) nanofiber with calcination temperature was rather
small, and it does not seem to play a role in this photocatalytic
reaction.
In addition to the change in the photophysical properties of
TiO2 nanofibers that was stated above, Pt metal particles also
played a role in influencing the photocatalytic activity of the
TiO2 (B) nanofiber. As evidenced in the TPR results in Figure
7, calcination at 673 K could provide a better dispersion of Pt
particles than calcination at 573 K. However, calcination at 873
K resulted in an SMSI effect, leading to a reduction in the exposed
Pt metal particles on the nanofiber surface and a lower binging
energy for Pt metal particles, as suggested by the XPS spectra
in Figure 8. The former means a smaller number of Schottky
barriers to reduce the rate of electron-hole recombination during
photocatalysis. The latter implies that Pt metal particles carry a
negative charge, which may repel electrons and reduces its
effectiveness as a Schottky barrier. Therefore, calcination at a
temperature higher than 673 K has a negative effect on Pt as a
promoter in this photocatalytic reaction.
Furthermore, it is interesting that raising the calcination
temperature from 673 to 773 K reduced the surface area of the
nanofiber by 24% but reduced the H2 production rate by only
5.5%. XRD and Raman data show that in addition to reducing
the surface area, calcination at 773 K led to the formation of an
anatase phase in the calcined TiO2 (B) nanofiber. At this
temperature, the nanofiber was a bicrystalline mixture of TiO2
(B) and the anatase phase. The formation of the bicrystalline
mixture seems to be able to offset the reduction in the H2
production rate. On the basis of the areal rates in Table 1, of all
the calcined nanofibers, that calcined at 773 K was the most
active photocatalyst and had an areal rate of 4.03 µmol h-1 m-2.
This value equals that of P-25 (4.00 µmol h-1 m-2). A
photocatalyst that has two crystalline phases promotes the
separation of photogenerated holes and electrons, enhancing the
photocatalytic activity, which has been referred to as the
bicrystalline synergetic effect.42–46 In fact, the benchmark P-25
photocatalyst is a mixture of anatase and rutile forms of TiO2,
and its high photocatalytic activity is the result of such a
bicrystalline synergetic effect.
We herein propose that the synergetic effect similar to that
observed in P-25 is also operating in our TiO2 (B)/anatase fibrous
mixture. The existence of the bicrystalline synergetic effect was
demonstrated by physically mixing two crystalline components
as described in the Experimental Section. Figure 11 presents the
amount of H2 generated after 2 h over 10 mg of photocatalysts
with different mole ratios of catalyst calcined at 673 K (pure
(40) Ohtani, B.; Ogawa, Y.; Nishimoto, S.-I. J. Phys. Chem. B 1997, 101,
3746.
(41) Zhang, Q.; Gao, L.; Guo, J. Appl. Catal. B: EnViron. 2000, 26, 207.
(42) Ohno, T.; Tokieda, K.; Higashida, S.; Masumura, M. Appl. Catal., A
2003, 244, 383.
(43) Yu, J. C.; Zhang, L.; Yu, J. Chem. Mater. 2002, 14, 4647.
(44) Wu, C.; Yue, Y.; Deng, X.; Hua, W.; Gao, Z. Catal. Today 2004, 93-95,
863.
(45) Yu, J.; Zhang, L.; Cheng, B.; Su, Y. J. Phys. Chem. C 2007, 111, 10582.
(46) Yu, J.; Xiong, J.; Cheng, B.; Liu, S. Appl. Catal. B: EnViron. 2005, 60,
211.
Effect of Temperature on Pt/TiO2 (B) Structure
Figure 11. Variation of H2 yield with anatase content in a physically
mixed TiO2 (B)/anatase bicrystalline mixture after 2 h of reaction.
TiO2(B) phase) to that calcined at 873 K (mostly anatase phase).
When the proportion of anatase in the mixture is less than 30%,
increasing the anatase phase content (and reducing the surface
area) reduced the H2 yield. The synergetic effect was absent
from these crystalline compositions, probably because the amount
of anatase in ethanol was not enough to produce proper contact
between the two phases.44 Interestingly, further increasing the
proportion of anatase to 40-60% clearly increased the H2 yield
although the surface area of the bicrystalline mixture in these
crystalline compositions should continue to decrease. At an even
higher anatase content (>70%), the effect of low anatase surface
area dominated the bicrystalline synergetic effect, resulting in
a continuous decrease in photocatalytic activity.
4. Conclusions
Calcination temperature has a profound effect on the mor-
phology, surface area, pore volume, and crystalline phase
Langmuir, Vol. 24, No. 17, 2008 9915
composition of the prepared nanofiber, and these factors in turn
affect its photocatalytic activity. The H2Ti3O7 nanofiber was
prepared by a hydrothermal method in 10 M NaOH solution,
followed by acidic washing and drying at 383 K. Calcining the
H2Ti3O7 fiber at 673 K produced a crystalline TiO2 (B) nanofiber,
which, after calcination at 773 K, transformed into a bicrystalline
mixture containing TiO2(B) and anatase phases with fibrous
morphology. Upon increasing the calcination temperature to 873
K, most of the TiO2 (B) nanofibers were converted into anatase
nanofibers and small anatase particles. Nanofibers indeed have
a higher temperature for phase transformation and morphology
deformation than do nanotubes. Therefore, they are more stable
against thermal calcination than nanotubes. In the photocatalytic
dehydrogenation of ethanol, a TiO2 (B) nanofiber impregnated
with 1% Pt and calcined at 673 K generated more H2 than those
calcined at other temperatures. Its activity with respect to the
catalytic dehydrogenation of ethanol is equal to that of 1% Pt/
P-25. TPR indicated that calcination at 573 K provided a poor
Pt dispersion on the TiO2 (B) nanofiber and a lower activity.
However, calcination at 773 and 873 K produced an SMSI effect
that decreased the surface concentration of Pt metal and created
Ptδs sites as suggested by XPS, preventing Pt particles from
functioning like a Schottky barrier. Because of a synergetic effect
between TiO2 (B) and the anatase phase, the bicrystalline mixture
produced by calcining at 773 K was able to counter negative
effects such as the reduction in surface area and the SMSI effect
and maintained its photocatalytic activity.
Acknowledgment. C-H.L. is grateful for a grant from the
National Science Council of Taiwan (NSC-93-2113-M-018-006)
in support of this research.
Supporting Information Available: Sample preparation pro-
cedures for electron microscopy, Raman spectra of H2Ti3O7 nanofibers,
and XRD spectra of Pt/TiO2 (B) nanofibers at different calcination
temperatures. This material is available free of charge via the Internet
at http://pubs.acs.org.
LA800572G