Dependence of nitrogen doping on TiO2 precursor annealed under NH3 flow

Article abstract of DOI:10.1016/j.jssc.2007.02.010

N-doped TiO₂ photocatalysts were prepared by annealing two different precursors, P25 and a TiO₂ xerogel powder under NH₃/Ar flow at 500, 550, and 600 °C. The xerogel powder prepared by peptizing Ti(OH)₄ with HNO₃ was composed of nanoparticles and had large specific surface area. During the annealing process, the xerogel powder underwent increase in crystallinity, grain growth and phase transformation, whereas P25 did not show obvious changes. Compared with the N-doped TiO₂ photocatalysts from P25, the N-doped TiO₂ photocatalysts from the xerogel powder possessed higher concentrations of the substitutional nitrogen and exhibited more obvious absorption in the visible light region. The N-doped TiO₂ photocatalysts from the xerogel powder exhibited obvious visible-light activities for photodegrading methylene blue and the sample prepared at 500 °C achieved the best performance with a rate constant (k) about 0.44 h^-1, whereas those from P25 did not exhibit improved visible-light activities.

Full text of DOI:10.1016/j.jssc.2007.02.010

ARTICLE IN PRESS
Journal of Solid State Chemistry 180 (2007) 1325–1332
www.elsevier.com/locate/jssc
Dependence of nitrogen doping on TiO precursor annealed
2
under NH flow
3
Ã
Xiaoming Fang , Zhengguo Zhang, Qinglin Chen, Hongbing Ji, Xuenong Gao
The Key Laboratory of Enhanced Heat Transfer and Energy Conservation, Ministry of Education, School of Chemical and Energy Engineering,
South China University of Technology, Guangzhou 510640, China
Received 27 November 2006; received in revised form 11 February 2007; accepted 12 February 2007
Available online 22 February 2007
Abstract
N-doped TiO
2 2 3
photocatalysts were prepared by annealing two different precursors, P25 and a TiO xerogel powder under NH /Ar
flow at 500, 550, and 600 1C. The xerogel powder prepared by peptizing Ti(OH) with HNO was composed of nanoparticles and had
4
3
large specific surface area. During the annealing process, the xerogel powder underwent increase in crystallinity, grain growth and phase
transformation, whereas P25 did not show obvious changes. Compared with the N-doped TiO photocatalysts from P25, the N-doped
TiO photocatalysts from the xerogel powder possessed higher concentrations of the substitutional nitrogen and exhibited more obvious
2
2
absorption in the visible light region. The N-doped TiO photocatalysts from the xerogel powder exhibited obvious visible-light activities
2
for photodegrading methylene blue and the sample prepared at 500 1C achieved the best performance with a rate constant (k) about
ꢀ
1
.44 h , whereas those from P25 did not exhibit improved visible-light activities.
0
r 2007 Elsevier Inc. All rights reserved.
2
Keywords: Photocatalysis; TiO ; Visible-light active; Nitrogen doping
1
. Introduction
Photocatalysis is one of the most interesting technologies
in elimination of toxic organic compounds either in gas or
generation photocatalytic materials [11], in which nitrogen-
doped TiO is the representative.
In the past 5 years, nitrogen-doped TiO photocatalysts
2
have been under extensive investigation, which mainly
focus on the following aspects: the preparation, structure
characterization and visible-light activity evaluation
[12–16], the theoretical calculations of the electronic
properties of N-doped TiO2 to elucidate the origin of
visible-light responses [17,18], and the investigation on the
photocatalytic degradation process of some organic
2
in liquid phase [1]. TiO is the most important inorganic
2
photocatalyst due to its low cost, outstanding stability and
highly efficient oxidative power [2,3]. However, the wide
use of TiO in photocatalysis is limited by its wide band
gap (3.2 eV for anatase), which facilitates response to UV
light. Therefore, numerous studies have been attempted to
extend the photoresponse of TiO further into the visible
2
compounds over N-doped TiO under visible light irradia-
2
2
light region by coupling with organic dye sensitizers [4], by
doping with transition metals [5], or by reducing with
hydrogen [6]. Recently, since the report by Asahi et al. in
tion [19,20]. The methods for preparing nitrogen-doped
TiO powders are generally as follows: the hydrolysis of
2
organic and inorganic titanium compounds with ammonia
water followed by heating the resultant precipitates or their
mixtures with urea [21,22], the nitridation of TiO colloidal
2
001 [7], anionic non-metal (such as carbon, sulfur,
fluorine, and nitrogen) doping has been recognized as a
more appropriate approach [8–10]. Moreover, anion-doped
TiO2 photocatalysts are considered to be the second-
2
nanoparticles with alkylammonium salts at room tempera-
ture [23], the mechanochemical reaction of titania with
hexamethylenetetramine or urea [24], the spray pyrolysis
from a mixed aqueous solution containing TiCl and an N-
4
Ã
precursor [25], and the heat treatment of a TiO precursor
2
in an ammonia atmosphere [7,26,27]. Among the above
Corresponding author. Fax: +86 20 8711 3870.
E-mail address: cexmfang@scut.edu.cn (X. Fang).
0022-4596/$ - see front matter r 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2007.02.010

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1326
methods, the annealing of TiO powder under NH flow is
2
2.2. Characterization
3
the simplest and most applicable process. However,
controversial ideas exist on the properties of the N-doped
TiO photocatalysts prepared by the process. On the one
The TEM image of the xerogel powder was obtained
using a transmission electron microscope (JEM-2010HR).
The Brunnauer–Emmet–Teller (BET) surface areas of the
precursors and their corresponding products were deter-
mined using nitrogen adsorption (Model ASAP 2010,
Micromeritics, USA). XRD patterns of the precursors and
their corresponding products were obtained using an X-ray
diffractometer (Model X’pert PRO, PANalytical, Holand).
The XPS data were obtained with an X-ray photoelectron
spectrometer (Kratos AXis Ultra) using Al (monochro-
mated) Ka radiation. The binding energies were referenced
to the C1s line at 284.9 eV from adventitious carbon.
Optical absorbance spectra of the precursors and their
corresponding products were measured using the diffuse
reflection method on a UV–vis spectrometer equipped with
an integrating sphere setup (U-3010, Hitachi, Japan).
2
hand, Asahi et al. [7] and Irie et al. [26] reported that the N-
doped TiO photocatalysts prepared by treating the ST-01
2
anatase powders (Ishihara Sangyo Kaisha, Japan) in the
NH (67%)/Ar atmosphere, revealed optical absorption
3
and photocatalytic activity under visible light irradiation
and on the other hand, Mrowetz et al. [28] found that the
N-doped TiO catalyst prepared by treating a commercial
2
anatase TiO powder (Aldrich) under a NH (80%)/Ar gas
2
3
ꢀ
ꢀ
flow, failed to catalyze the oxidation of HCOO into CO
+
or of NH –OH into NO under visible light illumination.
2
ꢀ
3
3
In the current work, two different TiO powders were
2
used as the precursors, and two series of N-doped TiO2
photocatalysts were prepared by annealing the two
precursors under NH /Ar flow at 500, 550, and 600 1C.
3
The microstructures of the two series of N-doped TiO2
photocatalysts were characterized by X-ray diffractometer
2.3. Photocatalytic activity evaluation
(
XRD) and X-ray photoemission spectroscopy (XPS), and
were brought into comparison. The optical absorption and
photocatalytic activities of the two series of N-doped TiO2
photocatalysts in the visible light region were investigated,
and intriguing results were obtained.
The photocatalytic activities of the samples were
evaluated by measuring decomposition rates of methylene
blue under visible light irradiation. The visible light was
obtained by a 500 W Xenon lamp with a 420 nm cut-off
filter, and its intensity was 100 mW/cm . Firstly, the
powdered samples (about 200 mg) were dispersed into
2
200 mL 15 mg/L (C ) of methylene blue aqueous solution
i
2
. Experimental
stirring magnetically for 30 min to establish absorption/
desorption equilibrium. Subsequently, 5 mL of the suspen-
sions were sampled. After the suspensions were centrifuged
2
.1. Preparation of N-doped TiO photocatalysts
2
at 2000 rpm, the concentrations (C ) of the solutions were
0
The two TiO precursors were the P25 TiO powder
2
Nippon Aerosil Co., Ltd.) and a xerogel powder. The
measured at 662 nm on a UV–vis spectrometer. Then, the
Xenon lamp was switched on, and the aqueous suspensions
containing methylene blue and powdered catalysts were
irradiated under the Xenon lamp. At the given time
intervals, 5 mL of the suspensions was taken from the
suspensions and their concentrations were measured.
2
(
procedure for preparing the xerogel powder was as follows.
Ti(SO ) , ammonia, and HNO were used as start
4
2
3
materials, and were of analytical reagent grade. Firstly,
% (wt%) of ammonia aqueous solution was added
dropwise to Ti(SO ) aqueous solution to prepare Ti(OH)
4
7
4
2
precipitate. Then, after being filtrated and washed with
distilled water five times, the precipitate was dispersed into
an HNO aqueous solution (HNO /TiO (mole) ¼ 0.6) at
3. Results and discussion
3.1. Characterization of the TiO xerogel powder
2
3
3
2
6
0 1C while stirring, and the precipitate was peptized and
transformed to a transparent sol with pale blue tint. After
being aged at 50 1C, the sol transformed to a gel. Finally,
the xerogel powder was obtained by drying the TiO gel at
2
When the ammonia aqueous solution was added to the
Ti(SO ) aqueous solution, an amorphous Ti(OH) white
precipitate was prepared. After the washed precipitate was
4
2
4
5
TiO photocatalysts were prepared by annealing the two
precursors under NH (67%)/Ar flow at 500, 550, and
3
0 1C followed by grinding, which was named FT. N-doped
dispersed into the HNO aqueous solution stirring at 60 1C,
3
the precipitate transformed to a transparent TiO sol by
peptization [29]. The TiO xerogel was obtained by aging
2
2
2
6
P25 were called N-P25-500, N-P25-550, and N-P25-600,
respectively. By analogy, the N-doped TiO products from
2
the FT xerogel powder were called N-FT-500,
N-FT-550, and N-FT-600, respectively. As references,
P25 and FT were annealed in the air at 550 1C for 3 h,
and the products obtained were named O-P25-550 and
O-FT-550.
00 1C for 3 h. The N-doped TiO2 products from
and drying the sol at 50 1C. The TEM image of the xerogel
powder is shown in Fig. 1. It can be seen that the xerogel
powder consisted of nanoparticles with good dispersibility,
most of which were less than 10 nm in size. Furthermore,
the BET specific surface area of the xerogel powder was
measured to be as large as 236 m /g. It has been reported in
our previous paper that the mesoporous nanocrystalline
TiO2 powders with a pore distribution centered at ca.
2

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1327
Fig. 1. TEM image of the FT xerogel powder.
4
5
0 nm were obtained by calcining the xerogel powders at
50 and 650 1C, and the mesopores disappeared after
calcination at 750 1C [30]. It was revealed that, during the
annealing process at the temperatures lower than 750 1C,
the xerogel powder underwent the conglomeration of the
nanoparticles to form mesoporous nanocrystalline TiO2
powders.
Fig. 2. XRD patterns of the two precursors and their corresponding
products: (a) FT and (b) P25.
3.2. Comparison on microstructures of N-doped TiO2
powders from different precursors
crystallinity were obtained accompanying the growth of the
anatase grain, as shown in Table 1. The average grain sizes
of anatase in N-FT-500, N-FT-550, and N-FT-600 were
10.4, 12.6, and 16.3 nm, respectively, indicating the average
3
.2.1. XRD analysis
XRD patterns of the two precursors and their corre-
sponding products are displayed in Fig. 2. The average
grain sizes of anatase in the products were determined from
the broadening of corresponding X-ray diffraction peaks
by Scherrer’s formula. The rutile fractions in the products
were calculated based on the respective intensities of
anatase and rutile peaks [31]. The values of the average
grain sizes of anatase and rutile fractions in all the products
are displayed in Table 1. It can be seen from Fig. 2(a) that
the FT xerogel powder had anatase form with poor
crystallinity. Since the temperatures for aging the sol and
drying the gel were as low as 50 1C, it could be inferred that
the anatase embryos were formed during the peptizing
grain size of anatase in the N-doped TiO powders from
2
FT increased with the annealing temperature. The BET
surface areas of N-FT-500, N-FT-550, and N-FT-600 were
2
measured to be 70, 59, and 52 m /g, respectively. Com-
pared with the precursor, the significant reduction in the
BET specific surface areas of the products was due to the
conglomeration of the TiO2 nanoparticles during the
annealing process to produce mesoporous nanocrystalline
TiO powders [30]. Furthermore, N-FT-500, N-FT-550,
2
and N-FT-600 consisted of anatase in majority and rutile
in minority, and their rutile fractions were 15%, 17%, and
39%, respectively. It was indicated that the rutile fraction
process at 60 1C with HNO as the peptizer. It suggested
3
that the amorphous Ti(OH) precipitate conglomerations
in the N-doped TiO powders from FT increased with the
2
4
underwent directional arrangement to form the ordered
structure of anatase during the peptization process. It can
also be seen from Fig. 2(a) that, compared with FT, sharp
diffractive peaks were observed for N-FT-500, N-FT-550,
and N-FT-600, suggesting that the products with high
annealing temperature, especially at 600 1C. From the
results it was revealed that, during the annealing process
under NH /Ar flow, FT underwent the increase in crystal-
3
linity, grain growth, phase transformation from anatase to
rutile and formation of mesoporous structure.

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1328
Table 1
Properties of the products from the two different precursors
Precursors
Products
FT
P25
O-FT-550
N-FT-500
N-FT-550
N-FT-600
O-P25-550
N-P25-500
N-P25-550
N-P25-600
Color
Average grain size of anatase
White
15.9
Yellow
10.4
Yellow green
12.6
Dark green
16.3
White
24.5
Ivory
24.5
Light yellow
24.3
Dark gray
24.3
Phase composition
Anatase (%)
Rutile (%)
80
20
85
15
83
17
61
39
75
25
74
26
75
25
73
27
2
BET specific surface area (cm /g)
53
70
59
52
60
65
62
58
N1s
b-N (%)
0
100
10
90
44
56
70
30
0
100
6
94
39
61
53
47
g-N
2
(%)
Substitutional N (at%)
0
0.081
0.28
0.81
0
0.012
0.035
0.63
Ti2P
4
Ti (%)
+
96
4
91
9
83
17
55
45
100
0
100
0
97
3
23
77
3+
Ti (%)
Relative coefficient (R)
Rate constant (k) (h
0.997
0.087
0.993
0.44
0.994
0.33
0.991
0.18
0.989
0.24
0.995
0.23
0.990
0.20
0.988
0.12
ꢀ1
)
P25 is a commercial form of TiO and is frequently used
2
spectra. It can be seen from Fig. 3(a) that the N1s spectra
for the N-doped TiO powders from FT consisted of two
in photocatalytic studies [32]. The P25 TiO powder that
2
2
consists of anatase (75%) and rutile (25%) had ca. 25 nm in
the average grain size of anatase, and its BET specific
peaks: one was centered at 396–397 eV and the other was
centered at 400–401 eV; with the increase in the annealing
temperature, the intensity of the peak at 396–397 eV
obviously increased, whereas the intensity of the peak at
400–401 eV gradually decreased. The N1s XPS spectrum
for O-FT-550 consisted of two peaks: one was centered at
400 eV and the other was centered at 402 eV. It can be seen
from Fig. 3(b) that the N1s spectra for the N-doped TiO2
powders from P25 also consisted of two peaks centered at
396–397 and 400–401 eV; the intensities of the former
peaks for N-P25-500 and N-P25-550 were very low,
whereas a significant increase in the intensity of the former
peak was observed for N-P25-600. One peak centered at
400–401 eV was observed for O-P25-550.
2
surface area was about 67 m /g. It can be seen from
Fig. 2(b) and Table 1 that the intensities of the diffractive
peaks for N-P25-500, N-P25-550, N-P25-600, and O-P25-550
were close to those of the diffractive peaks for P25, and the
grain sizes of anatase and the rutile fractions in the
products were almost the same. It suggested that P25 did
not show obvious changes in the grain size of anatase and
the rutile fraction during the annealing process at the
temperature between 500 and 600 1C under NH flow or in
3
the air. Furthermore, the BET specific surface areas of N-
P25-500, N-P25-550, N-P25-600, and O-P25-550 were
2
measured to be 65, 62, 58, and 60 m /g, respectively,
indicating that P25 also did not show significant change in
the BET specific surface area after being annealed at the
temperature between 500 and 600 1C.
In the reported papers concerning N-doped TiO2,
controversies exist on the assignment of the peak feature
in N1s XPS for the N-doped TiO . Asahi et al. [7] reported
2
From the above analyses, it was revealed that, different
precursors underwent different changes during the anneal-
ing process under NH /Ar flow the temperature between
3
that three peak structures at the binding energies of 402,
400, and 396 eV were observed for the TiO2 N films and
the peaks were assigned as atomic b-N (396 eV) and
ꢀx
x
5
TiO product was dependent not only on the annealing
conditions but also on the property of the precursor used.
00 and 600 1C. Thus, the microstructure of the N-doped
molecularly chemisorbed g-N (400 and 402 eV). Diawald
2
et al. [33] assigned the N1s peak at 399.6 eV as the peak
feature for the N-doped rutile single crystals and presented
that the doping state of nitrogen is chemically bound to
hydrogen and oxygen, and this form of nitrogen is most
likely located in an interstitial site. In the N1s XPS
2
3
.2.2. XPS analysis
Presently, XPS is generally used to characterize the
substitution of the oxygen sites with nitrogen atoms in the
titania structure. The global XPS profiles for the two series
of products from different precursors are displayed in
Fig. 3, and the insets in Fig. 3 show the corresponding N1s
spectrum of the N-doped TiO nanomaterial with high N
2
doping concentration (up to 8%), a broad binding energy
peak centered at 401.3 eV was observed, which was
attributed to O–Ti–N [34]; however, Gopinath [35] argued

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1329
estimated by calculating the 396 eV peak areas and the
00 eV peak areas, and the values obtained are summarized
4
in Table 1. Furthermore, the concentrations of the
substitutional nitrogen were calculated by multiplying the
percentages of the atomic b-N by the total nitrogen
concentrations in all the N-doped TiO products and are
summarized in Table 1. It can be seen that the concentra-
tions of the substitutional nitrogen in N-FT-500, N-FT-
2
550, and N-FT-600 were 0.081, 0.28, and 0.81 at%,
respectively, indicating that the concentrations of the
substitutional nitrogen in the N-doped TiO powders from
2
FT increased with the annealing temperature, whereas only
molecularly chemisorbed g-N2 existed in the product
prepared in the air. Few substitutional nitrogen existed in
the products prepared by annealing P25 under NH /Ar
3
flow at temperatures 500 and 550 1C, and the remarkable
increase in the concentration of the substitutional nitrogen
occurred at the annealing temperature of 600 1C. Com-
pared with the products from P25, the N-doped TiO2
powders from FT at the same annealing temperatures
possessed higher concentrations of the substitutional
nitrogen. This result might be attributed to that, during
the annealing process under NH /Ar flow, FT experienced
3
the increase in crystallinity, grain growth, phase transfor-
mation from anatase to rutile and formation of mesopor-
ous structure, whereas P25 did not.
Oxygen vacancy state in anatase TiO , which was
2
3
+
generally correlative with the amount of Ti , was below
the lower end of the conduction band and acted as a
recombination center for holes and electrons [26]. In order
to investigate the oxygen vacancies existing in the N-doped
TiO products from the two precursors, the Ti P peaks
2
2 3/2
in all the products were fitted using the XPSPEAK
4
software to separate the Ti peaks and Ti peaks. The
+
3+
4
+
3+
percentages of Ti and Ti were estimated by comparing
their peak areas, and the values obtained are summarized
Fig. 3. N 1s XPS spectra for the corresponding products from the two
precursors: (a) FT and (b) P25.
3
+
in Table 1. It can be seen that the Ti percentages in N-
FT-500, N-FT-550, and N-FT-600 were 9%, 17%, and
3
+
4
5%, respectively, indicating that the amount of Ti
in
the N-doped TiO products from FT increased with the
2
that the N1s peak at 398.2 eV should be attributed to the N
state in O–Ti–N. Kisch and co-worker [36] prepared N-
doped TiO by calcining the hydrolysate of TiCl with an
annealing temperature, especially at 600 1C. It could be
inferred that the amount of oxygen vacancy in the N-doped
TiO2 products from FT increased with the annealing
2
4
3
+
ammonia aqueous solution. They found that no peak at
96–397 eV was observed in the N1s spectrum of the N-
temperature. The amount of Ti
P25-550 was very low whereas the percentage of Ti in N-
in N-P25-500 and N-
3
+
3
4
+
doped product, whereas a broad and weak peak at 404 eV
was observed, which was attributed to the presence of
hyponitrite. On the basis of the reports mentioned above, it
could be presumed that the difference in the preparation
methods may lead to the change in the N states in the N-
P25-600 was as high as 77.26%, larger than that of Ti
3
.
in the products prepared at
+
The high amount of Ti
600 1C resulted from the fact that the precursors were
reduced by H , which was produced from the decomposi-
tion of NH at over 550 1C [26]. It could be inferred that
2
3
doped TiO products, thus to the distinction in the active
2
the amount of oxygen vacancy in N-P25-600 was very high,
as compared with N-P25-500 and N-P25-550.
sites responsible for their visible-light activities. It was
obvious that the peak structures in the N1s XPS spectra for
the N-doped TiO powders from FT and P25 were similar
The above XPS analyses revealed that the concentration
of substitutional nitrogen and the amount of oxygen
vacancy in the N-doped TiO powder were dependent not
2
to those reported by Asahi et al. [7]. According to their
assignment, the percentages of the atomic b-N and
chemisorbed g-N2 in the N-doped TiO2 products were
2
only on the annealing temperature but also on the
precursor used.

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1330
3
from different precursors
.3. Comparison on properties of N-doped TiO powders
2
compared with FT, remarkable shifts of the absorption
shoulders into the visible light region were observed for N-
FT-500, N-FT-550, and N-FT-600, while little shift was
observed for O-FT-550. It indicated that compared with O-
FT-550 prepared by annealing FT in the air, the N-doped
TiO powders prepared from FT under NH /Ar flow were
3
.3.1. Optical absorption in the visible light region
The colors of the two precursors and their corresponding
products are also shown in Table 1. It was found that, after
2
3
being annealed under NH /Ar atmosphere at different
3
visible-light active. In more detail, the absorption edges of
N-FT-500, N-FT-550, and N-FT-600 in the UV region
(o400 nm) were almost the same as that of O-FT-550; in
the range of wavelengths between 400 and 500 nm,
significant absorption occurred for N-FT-500, N-FT-550,
and N-FT-600, and their absorbance gradually increased
with the annealing temperature; in the region of wave-
lengths longer than 500 nm, slight absorption was observed
for N-FT-500, obvious absorption for N-FT-550, and
remarkable absorption was observed for N-FT-600,
indicating that the absorbance dramatically increased with
the annealing temperature. It can be seen from Fig. 4(b)
that, compared with P25, little shift of the absorption
shoulder into the visible light region was observed for O-
P25-550, slight shift for N-P25-500 and N-P25-550, and
noticeable shift was observed for N-P25-600, especially at
wavelengths longer than 500 nm. Furthermore, it was
temperatures, the FT powder changed its color from white
to yellow at 500 1C, to yellow green at 550 1C and to dark
green at 600 1C, whereas O-FT-550 remained white. As for
P25, O-P25-550 was as white as P25, N-P25-500 was ivory,
N-P25-550 was light yellow, and N-P25-600 was dark gray.
It indicated that P25 did not show obvious color change
after being annealed at 500 and 550 1C, which revealed that
the color of the N-doped TiO powder was dependent on
its precursor, and was distinct with the annealing
temperature.
2
Fig. 4 shows the optical absorbance spectra obtained by
diffuse reflection of the two precursors and their corre-
sponding products. It could be seen in Fig. 4(a) that
revealed that the N-doped TiO powders prepared from
2
the xerogel at the annealing temperature of 500
and 550 1C exhibited more obvious absorption in the
visible light region, as compared with the N-doped TiO2
powders prepared from P25 at the same annealing
conditions; both significant absorptions were observed for
N-FT-600 and N-P25-600, especially at wavelengths longer
than 500 nm.
More recently, several researchers focused their attention
on the theoretical calculations of the electronic properties
of N-doped TiO to elucidate the origin of visible-light
2
responses. Lin et al. [37] presented the results of spin-
polarized density functional theory (DFT) calculations and
optical absorption spectra that arise for a range of
concentrations of substitutional nitrogen and oxygen
vacancies in anatase TiO . Their research results showed
2
that absorption below 500 nm was mainly due to nitrogen
states located above the valence bands, whereas absorption
above 500 nm was mainly caused by oxygen vacancies.
Based on their research results and our XPS analyses, it
could be concluded that the increase in the absorbances of
the N-doped TiO2 powders from FT in the range of
wavelengths between 400 and 500 nm was attributed to the
increase in the concentration of the substitutional nitrogen
with the annealing temperature. The dramatic increase in
the absorbances of the N-doped TiO powders from FT at
2
wavelengths greater than 500 nm was due to the remark-
able increase in the amount of oxygen vacancy with the
annealing temperature. The slight absorption for N-P25-
500 and N-P25-550 in the visible light region was attributed
to that few substitutional nitrogen existing in them.
Noticeable absorption for N-P25-600 and N-FT-600 at
wavelengths longer than 500 nm might be due to their high
amount of oxygen vacancy.
Fig. 4. Optical absorbance spectra obtained by diffuse reflection of the
two precursors and their corresponding products: (a) FT and (b) P25.

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1331
3
.3.2. Photocatalytic activities under visible light irradiation
The photocatalytic degrading curves for the methylene
550, indicating that N-FT-500, N-FT-550, and N-FT-600
exhibited obvious visible-light activities for photodegrad-
ing methylene blue, whereas the visible-light activity of O-
FT-550 was poor. It was revealed that annealing the FT
blue aqueous solution (15 mg/L) in the presence of the
products from FT and P25 under visible light irradiation
are displayed in Fig. 5. It was found that after the
powdered samples have been dispersed into the methylene
blue aqueous solution for 30 min under no light irradiation,
a decrease in the concentration of methylene blue solution
occurred, which was attributed to the sorption of the TiO2-
based powders for methylene blue molecules. The relative
coefficients (R) and the slopes of the lines obtained by
xerogel powder under NH /Ar flow produced the visible-
3
light-active photocatalysts. It could be inferred that the
visible-light activities of N-FT-500, N-FT-550, and N-FT-
600 might be attributed to that the remarkable shifts of the
absorption shoulders for their optical absorbance spectra
at wavelengths between 400 and 500 nm, which was due to
the substitutional nitrogen existing in the products. The
photocatalytic activities of the N-doped TiO2 powders
from FT gradually decreased with increase in annealing
temperature, and N-FT-500 achieved the best perfor-
mance. The decrease in the photocatalytic activities of the
fitting linear ln(C /C) via irradiating time (t) are also shown
0
in Table 1. It can be seen from Table 1 that the relative
coefficients for all the products were close to 1, suggesting
that the degradation of methylene blue in the presence of
the photocatalysts was first-order kinetics reaction, and the
slopes of the lines could be considered as the photode-
gradation reaction rate constants. It was obtained that the
N-doped TiO powders from FT with annealing tempera-
2
ture might be owing to the dramatical increase in the
amount of oxygen vacancy with the annealing temperature,
because the oxygen vacancies acted as recombination
centers for holes and electrons [26].
ꢀ
1
ꢀ1
rate constants were 0.44 h for N-FT-500, 0.33 h for N-
ꢀ
1
FT-550, 0.18 h for N-FT-600, and 0.087 h for O-FT-
ꢀ1
As for the N-doped TiO powders from P25, intriguing
2
ꢀ
1
results were obtained. The rate constants were 0.24 h for
ꢀ
1
O-P25-550, 0.23 h for N-P25-500, 0.20 h for N-P25-
5
ꢀ1
ꢀ
1
50, and 0.12 h for N-P25-550, respectively, indicating
that the photocatalytic activities of N-P25-500 and N-P25-
50 were comparable to that of O-P25-550, and N-P25-600
5
exhibited the worst photocatalytic activity. It was revealed
that annealing P25 under NH /Ar flow did not produce the
3
photocatalysts with improved visible-light activity, as
compared with the product prepared in the air. Recently,
there are several papers reporting that P25 exhibited
photocatalytic activity under visible light irradiation
[
38,39]. It had been revealed from the above XRD analyses
that P25 did not show obvious changes in the grain size and
rutile fraction during the annealing process at the
temperature between 500 and 600 1C under NH flow or
3
in the air. It was reasonable that O-P25-550 was also
visible-light active. Since only slight shifts of the absorption
shoulders into the visible light region at wavelength
between 400 and 500 nm were observed for N-P25-500
and N-P25-550, and few oxygen vacancies existed in them,
it could be inferred that the visible-light activities of N-P25-
5
5
00 and N-P25-550 were comparable to these of O-P25-
50. The worst visible-light activity for N-P25-600 was
owing to its large amount of oxygen vacancy.
It was found that compared with the N-doped TiO2
powders from P25, the N-doped TiO powders from FT
2
exhibited excellent photocatalytic activities under visible
light irradiation, especially at the annealing temperature of
500 and 550 1C. The photocatalytic activity of the N-doped
TiO powder was relative to its specific surface area, grain
2
size, phase composition, substitutional nitrogen concentra-
tion and oxygen vacancy amount. Based on the above
XRD and XPS analyses, it was inferred that compared
with N-P25-500 and N-P25-550, N-FT-500 and N-FT-550
possessed comparative specific surface area, smaller grain
size, less rutile fraction, higher concentration of the
Fig. 5. Photocatalytic degradation curves of methylene blue under visible
light irradiation in presence of the products from the two precursors: (a)
FT and (b) P25.

ARTICLE IN PRESS
X. Fang et al. / Journal of Solid State Chemistry 180 (2007) 1325–1332
1332
substitutional nitrogen and comparative amount of oxygen
vacancy, which probably resulted in their more excellent
photocatalytic activities. It was concluded from the above
[6] I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihara, K.
Takeuchi, J. Mol. Catal. A: Chem. 161 (2000) 205.
[
7] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293
2001) 269.
(
results that when N-doped TiO was prepared by annealing
2
[8] J.C. Yu, M. Zhou, B. Cheng, X. Zhao, J. Mol. Catal. A: Chem. 246
(2006) 176.
a TiO precursor under NH flow, the visible-light activity
2
3
of the N-doped TiO photocatalyst was not only dependent
2
[9] J.C. Yu, J.G. Yu, W.K. Ho, Z.T. Jiang, L.Z. Zhang, Chem. Mater. 14
2002) 3808.
10] N. Venkatachalam, A. Vinu, S. Anandan, B. Arabindoo, V.
(
on the annealing conditions but also relative to the
precursor used. It is important to select a suitable precursor
for preparing the N-doped TiO photocatalysts with high
[
Murugesan, J. Nanosci. Nanotechnol. 6 (2006) 2499.
11] H. Wang, J.P. Lewis, J. Phys. Condens. Matter 18 (2006) 421.
12] T. Lindgren, J.M. Mwabora, E. Avendano, J. Jonsson, A. Hoel, C.G.
2
[
[
visible-light activity.
Granqvist, S.E. Lindquist, J. Phys. Chem.
709.
B
107 (2003)
5
4
. Conclusions
[
[
13] O. Diwald, T.L. Thompson, T. Zubkov, E.G. Goralski, S.D. Walck,
J.T. Yates Jr., J. Phys. Chem. B 108 (2004) 6004.
14] S. Livraghi, A. Votta, M.C. Paganini, E. Giamello, Chem. Commun.
4 (2005) 498.
The xerogel was composed of nanoparticles and had
large specific surface area. During the annealing process at
the temperature between 500 and 600 1C, FT experienced
the increase in crystallinity, grain growth and phase
transformation from anatase to rutile, whereas P25 did
not show obvious change in the grain size and rutile
fraction. Compared with the products from P25, the N-
doped TiO2 powders from FT at the same annealing
temperatures possessed higher concentrations of the sub-
stitutional nitrogen and exhibited more obvious absorption
in the visible-light region. Compared with the product
prepared by annealing the xerogel in the air, the N-doped
TiO powders prepared from the xerogel under NH /Ar
flow exhibited obvious visible-light activities for photode-
grading methylene blue, and the sample prepared at 500 1C
achieved the best performance with a rate constant (k)
about 0.44 h . It was found that annealing P25 under
NH /Ar flow did not produce the photocatalysts with
[15] S. Sato, R. Nakamura, S. Abe, Appl. Catal. A 284 (2005) 131.
[16] A. Oriov, M.S. Tikhov, R.M. Lambert, Co. R. Chim. 9 (2006)
7
94.
[
[
17] Y. Wang, D.J. Doren, Solid State Commun. 136 (2005) 186.
18] J.Y. Lee, J. Park, J.H. Cho, Appl. Phys. Lett. 87 (2005) (Art No.
011904).
[19] Y. Irokawa, T. Morikawa, K. Aoki, S. Kosaka, T. Ohwaka, Y. Taga,
Phys. Chem. Chem. Phys. 8 (2006) 1116.
[
20] T. Tachikawa, Y. Takai, S. Yojo, M. Fujitsuka, H. Irie, K.
Hashimoto, T. Majima, J. Phys. Chem. B 110 (2006) 13158.
21] T. Ihara, M. Miyoshi, Y. Iriyama, O. Matsumoto, S. Sugihara, Appl.
Catal. B: Environ. 42 (2003) 403.
[
[22] K. Kobayakawa, Y. Murakami, Y. Sato, J. Photochem. Photobiol.
A: Chem. 170 (2005) 177.
2
3
[
23] C. Burda, Y.B. Lou, X.B. Chen, A.C.S. Samia, J. Stout, J.L. Gole,
Nano Lett. 3 (2003) 1049.
[
24] S. Yin, K. Ihara, M. Komatsu, Q.W. Zhang, F. Saito, T. Kyotani, T.
Sato, Solid State Commun. 137 (2006) 132.
ꢀ
1
[25] D. Li, H. Haneda, S. Hishita, N. Ohashi, Mater. Sci. Eng. B 117
2005) 67.
26] H. Irie, Y. Watanabe, K. Hashimoto, J. Phys. Chem. B 107 (2003)
483.
3
(
improved visible-light activity. It is important to select a
suitable precursor for preparing the N-doped TiO photo-
catalysts with high visible-light activity.
[
[
[
[
2
5
27] B. Kosowska, S. Mozia, A.W. Morwski, B. Grzmil, M. Janus, K.
Kalucki, Sol. Energy Mater. Sol. Cell 88 (2005) 269.
28] M. Mrowetz, W. Balcerski, A.J. Colussi, M.R. Hoffmann, J. Phys.
Chem. B 108 (2004) 17269.
29] N. Yao, G.X. Xiong, Y.H. Zhang, W.S. Yang, Sci. China B 31 (2001)
Acknowledgments
This work was supported by the National Natural
Science Foundation of China (Project no. 20573038) and
the project (2004CCA03100) from Ministry of Science and
Technology of the People’s Republic of China.
3
55.
[30] X.M. Fang, Z.G. Zheng, Q.L. Chen, Rare Metal Mater. Eng.,
in press.
[
[
31] R.A. Spurr, W. Mayers, Anal. Chem. 29 (1957) 760.
32] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D.
Tilley, J. Solid State Chem. 92 (1991) 178.
References
[33] O. Diawald, T.L. Tompson, T. Zubkov, E.G. Goralski, S.D. Walck,
J.T. Yates, J. Phys. Chem. B 108 (2004) 6004.
[
[
1] M. Hoffman, S. Martin, W. Choi, Chem. Rev. 95 (1995) 69.
2] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1
[34] X.B. Chen, C. Burda, J. Phys. Chem. B 108 (2004) 15446.
[35] C.S. Gopinath, J. Phys. Chem. B 110 (2006) 7079.
[36] S. Sakthivel, H. Kisch, Chemphyschem 4 (2003) 487.
[37] Z.S. Lin, A. Orlov, R.M. Lambert, M.C. Payne, J. Phys. Chem. B 109
(2005) 20948.
(2000) 1.
[
3] S.A. Bilmes, P. Mandelbaum, F. Alvarez, N.M. Victoria, J. Phys.
Chem. B 104 (2000) 9851.
4] T. Samuel, C. Cheury, Chemosphere 36 (1998) 2461.
[
[
[38] S.W. Yang, L. Gao, J. Am. Ceram. Soc. 87 (2004) 1803.
[39] R. Silveyra, L.D.T. Saenz, W.A. Flores, V.C. Martinez, A.A.
Elguezabal, Catal. Today 107/108 (2005) 602.
5] D. Dvoranova, V. Brezova, M. Maz u´ r, M.A. Malati, Appl. Catal. B:
´ ´
Environ. 37 (2002) 91.