ARTICLE IN PRESS
Y.-S. Jung et al. / Journal of Physics and Chemistry of Solids 69 (2008) 1464–1467
1465
solution, and the concentration of titanium vs. aluminum
was varied from 0 to 100 mol%. After stirring the prepared
solution at 75 1C for 24 h, the resultant sol was dried in
oven at 85 1C for 24 h. Synthesized samples were calcined at
various temperatures (600, 800, and 1000 1C) for 2 h.
Synthesized samples underwent more detailed analysis.
Crystal structure of synthesized samples was examined by
X-ray diffraction analysis (XRD, RIGAKU Miniflex), and
FT-Raman spectroscopy was measured with Bruker RFS
100/S laser Raman spectrometer at room temperature. An
Nd YAG laser with a wavelength of 1064 nm was used as
an excitation source. Scanning electron microscopy (SEM,
Hitachi Ltd., S-4300) was used to identify the surface
morphology, and sample composition was analyzed by
energy dispersive spectroscopy (EDS). The BET surface
areas of synthesized samples were determined by nitrogen
gas adsorption at liquid nitrogen temperature (Micromeri-
tics ASAP2000).
The decomposition of acetaldehyde was carried out in a
closed borosilicate reactor. A sample of 0.2 g was put in
reactor and the reactor was purged with N2 containing
1000 ppm acetaldehyde for 15 min. Proper amount of air
was then injected into reactor for the oxidation of
acetaldehyde. The reactor was illuminated from a light
source through optical fiber (LUMATEC, Series 2000),
which cuts off UV light allowing only visible light
(l4420 nm) to be transmitted. Xenon-arc lamp (SCIEN-
CETECH, 201–1 K) was used as a light source. After
photocatalytic reaction, the conversion of acetaldehyde
was determined by a gas chromatograph (Chrompack, CP-
9001) equipped with flame ionization detector.
content was from 30% to 70%, crystallized structures were
obtained after calcination at 800 1C. Samples with high
aluminum content (more than 90%) remained amorphous
even after calcination at 800 1C, and well-crystallized
structures were achieved after calcination at 1000 1C.
Second, the addition of aluminum into titania prevented
samples from phase transformation (anatase to rutile).
Phase transformation of pure titania from anatase to rutile
was observed at above 400 1C. When a small amount of
aluminum was added, however, anatase phase was main-
tained even after calcination at 800 1C. Third, for all mixed
oxides calcined at 800 1C, there is no evidence of super-
positional peaks of pure titania and pure alumina, which
indicates that the mixed oxides are not simply mixed phases
of the two pure oxides, but solid solutions with a single
phase. However, phase separation was observed after
calcination at 1000 1C, as shown in Fig. 1(c).
Raman spectra of synthesized samples after calcination
at 800 1C are shown in Fig. 1(d), and for comparison,
Raman spectrum of anatase-phase TiO2 is also shown.
Raman spectroscopy is a good technique for the elucida-
tion of structure of complex transition metal oxides,
because it probes structures and bonds by their vibrational
spectrum. For anatase titania, three broad peaks were
observed at 398.15, 517.72, and 639.22 cmÀ1, and rutile
titania exhibited two broad peaks at 448.2 and 610.2 cmÀ1
.
In the cases of mixed oxides, three broad peaks, which are
very similar to those of anatase TiO2, were observed. These
results indicate that mixed oxides maintain anatase phase
even after calcination at 800 1C, which is also evidenced
from XRD results. When aluminum content is very high
(more than 70%), no characteristic peak was observed due
to poor crystallinity. Several methods for the preparation
of titania nanocomposites with alumina, silica, and
zirconia, have been reported to increase the specific surface
area and the stability of the anatase structure [9]. It was
concluded from XRD and Raman spectroscopy that the
added aluminum seems to be highly dispersed and acts as a
stability promoter for anatase phase.
Fig. 2 shows BET surface area of the synthesized
aluminum–titanium mixed oxides after calcination at
800 1C with respect to aluminum content (%). The surface
areas of pure titania and alumina were 4.5 and 49.5 m2/g,
respectively. For mixed oxides, surface area was much
higher than either that of pure TiO2 or Al2O3, and surface
area increased with the increase of Al content. The
maximum surface area was observed for Ti(10%)–Al(90%)
mixed oxide with a value of 116.8 m2/g. It was confirmed
from pore size distribution analysis (data not shown) that
pore sizes of prepared samples were strongly dependent on
sample composition. With the increase of aluminum
content, pore size steadily decreased, resulting in the
increase of surface area. However, in the case of either
pure aluminum or pure titania, porous structure was not
well developed after calcination at 800 1C.
3. Results and discussion
The sample compositions were determined by EDS and
correlated remarkably well to the Ti:Al ratios in solution
within the error range of 5%, as shown in Table 1. All
samples were confirmed to be amorphous as-synthesized,
and XRD patterns after calcination at 600, 800, and
1000 1C are shown in Fig. 1(a–c). Several interesting
phenomena were observed from XRD analysis. First,
crystallization temperature was found to increase with the
increase of alumina content. When alumina content was
less than 30%, well-crystallized structure was observed
after calcination at 600 1C. In cases where aluminum
Table 1
The variation of composition as a function of Ti:Al ratio in solution,
measured by EDS analysis
Mole ratio in solution
Mole ratio in sample (%)
(Ti:Al)
9:1
7:3
(Ti:Al)
92.9:7.0
76.5:23.5
60.1:39.8
39.6:60.4
16.2:83.8
5:5
3:7
Before photocatalytic reaction, all samples were kept
without illumination (under dark condition) for 2 h and
1:9
Jung, Young-Soo
