10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
IMPROVED VISIBLE LIGHT RESPONSE OF METAL/NON-METAL DOPED TiO2
NANOPARTICLES UTILIZATION IN ORGANIC POLLUTANTS DEGRADATION
Maria E. Pappa1, Petros M. Sakkas1, Tatjana Kosanovic Milickovic1,2,
Alexandros Zoikis-Karathanasis2 and Ioanna Deligkiozi*1
1
Center for Research and Technology Hellas, Institute for Research and Technology of Thessaly
2
Artia Nano Engineering & Consulting
ABSTRACT
Non-metal and metal modified TiO2 photocatalysts as well as pure TiO2 powders were synthesized via sol-gel
method, using titanium precursors with non-metal [N or S,N] and metal [Ag] doping elements. The synthesized
powders have been imposed to different heat treatment cycles at several temperatures above 400oC. Structural
and morphological characterizations have been accomplished through X-ray powder diffractometry (XRD),
scanning electron microscopy (SEM), DR-UV-Vis and IR spectroscopy. Band gap energy (E g) measurements
have been performed via UV-Vis spectroscopy and accordingly calculated from a plot of modified Kubelka–
Munk function versus the energy of exciting light by converting the scanning wavelength (λ) into photon band
gap energies (Eg). According to XRD analysis, doped TiO2 powders consisted mainly of the TiO2 anatase phase,
especially in the case of calcination temperatures less than 600oC and the crystallite size ranges between 8–30
nm. Furthermore, the band gap energy was estimated to be less than 2.5 eV, therefore the photocatalytic
performance of doped TiO2 powders is extended in the visible range. Concerning the photocatalytic activity
determination, several experiments have been conducted in a photo-reactor, where the activity of the doped
nanostructured TiO2 catalysts were evaluated through the degradation of two model pollutants, namely
Methylene Blue (MB) and Methyl Orange (MO). The new powders exhibited significantly improved
photocatalytic performance in MO and MΒ in comparison to the non-doped TiO2 and reference compounds by
reducing up to 90% the model pollutants tested.
Keywords: photocatalysis, doped titania, pollutants degradation, building materials
INTRODUCTION
Titania (TiO2) based materials are well known semiconductors able to be activated under light irradiation and
therefore catalyze several chemical processes. This photocatalytic performance in combination with their
chemical stability, low toxicity and low cost of precursors, render such materials as potential candidates for
environmental friendly processes such as wastewater treatment and air purification [1,2,3]. In the present
research, these performance capabilities aspire to find application in novel multifunctional paintable systems for
façade applications acting as external surface coatings that will exhibit organic pollutants degradation through its
photocatalytic activity. Since TiO2 absorbs only a minor part of the active solar spectrum that constitutes only
the 5% of solar light spectrum, primary objective of this study is to improve light harvesting of these materials
making them absorb in the visible range of the electromagnetic spectrum and thereby improving their efficiency.
For this purpose TiO2 was modified via doping with metal (silver, Ag) and/or non-metal elements (nitrogen, N
and sulfur, S) that have the potential to stimulate the visible light response of the synthesized particles. The
origin of the shifted absorption and the therefrom derived photocatalytic activity mechanism of the [N]-doped
TiO2 remains a matter of controversial discussions [4] that is attributed among others to a delocalized mixing
between the O 2p and N 2p orbitals created from nitrogen doping [5,6] or to an isolated mid-gap level above the
valence gap generated from the dopant atom orbitals [5,7,8]. Similarly, in the case of the [S]-doped TiO2,
additional electronic states above the valence band edge of pure TiO 2 have been reported [9]. On the other hand,
noble metals such as Ag, known for exhibiting unique electronic and catalytic properties, significantly promote
the electron-hole separation and therefore enhance the photocatalytic activity of doped TiO2 [10]. Furthermore,
in order the modified TiO2 particles to exhibit an improved photocatalytic activity, they need to meet some extra
criteria regarding their crystal structure as well as the crystallite and particle size. Anatase titania usually exhibits
higher photocatalytic activity than rutile and brookite, owing to its higher density of localized states, the surfaceadsorbed hydroxyl radicals and slower charge carrier recombination [11], while the small crystallite size
generally improve the phototoactivity of TiO2 as has been reported by several researches [12,13] For the
evaluation of the photocatalytic activity of the doped TiO2 powders several chemical compounds are used as
model pollutants in experimental protocols such as textile dyes, phenols, acetaldehyde etc. In the present
research two textile dyes, methyl orange and methylene blue have been chosen as model pollutants for the
photocatalytic experiments.
MATERIALS AND METHODS
Synthesis
Pure and doped TiO2 powders were synthesized using the sol-gel method. 29.4mmol of Titanim(IV) butoxide
were hydrolysed in 50mL of acidified aqueous solution. After 3 hours of vigorous stirring, 109mmol of alcohol
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
were added and the mixture was mildly stirred for 24 hours. In the case of non-metal doped TiO2 powders ([N]doped and [S,N]-doped), various amounts of urea (16.7-666mmol) or thiourea (13.1-197mmol), respectively,
were added and the mixture was stirred for additional 2 hours followed by a sonication step around 15min. In the
case of the metal co-doped powders ([Ag-S,N]) various amounts of silver nitrate (0.382-1.531mmol) were
investigated and the mixture was sonicated and stirred under dark conditions for 24h. After the evaporation of
the solvents the crude amorphous powders were calcinated at 400 oC for 2h and then purified using an acidified
aqueous solution or distilled water. For all the samples there have been followed additional calcination steps for
1h each at 500-800oC. Titanium(IV) butoxide, reagent grade, 97% was purchased from Aldrich, 1-butanol, Urea,
Thiourea, Silver nitrate, Methylene blue hydrate and Nitric acid 65% were all analytical grade (a.g.), purchased
from PENTA and Methyl orange from Panreac. All chemicals were used as received without further purification.
CHARACTERIZATIONS
X-ray powders diffraction (XRD)
The crystal structure of the synthesized powders was analyzed by X-ray diffraction using a Siemens D5000 Xray diffractometer (XRD) operated at 40 kV and 35 mA with a Cu-Kα radiation λ of 1.54 Ǻ. All the XRD data
were collected using a θ-2θ configuration and recording the diffraction intensity in the range of 2θ between 20–
100 degrees. The Difrac®Plus Eva software, version 2.0 (Siemens Energy and Automatization, Inc.) assisted to
identify the crystalline peaks. The crystalline phase of all samples has been identified by comparison with the
ICDD (International Centre for Diffraction Data) database and Joint Committee on Powder Diffraction Standards
(JCPDS) files. Anatase/rutile percentages have been calculated from the resulting diffracrograms using the Spurr
–Myers equation, while the average diameter of the crystallites has been estimated by using the Scherrer
equation.
Scanning Electron Microscopy (SEM)
The morphology of the powders has been investigated applying Scanning Electron Microscopy (SEM). Energydispersive X-ray (EDX) spectroscopy analyses were obtained using a FEI Quanta 200 electron microscope.
Fourier transform infrared spectroscopy (FTIR)
This technique has been performed with a Perkin Elmer Spectrum 1 FTIR spectrophotometer using a KBr pellet
for sample preparation. All the samples were measured by using the attenuated total reflectance technique (ATR)
in the region 400-4000 cm-1. The total numbers of scans for each measurement were 100 and the applied
resolution was about 4cm−1. Under these specific test conditions, it has been feasible to evaluate the most of the
visible light sensitizing groups of the modified TiO2 nano particles.
Diffuse reflectance spectra
The optical reflectivity of the produced powders has been measured using a UV-Vis spectrometer (Agilent Cary
60) using a diffuse reflectance accessorize including remote diffused reflectance probe Harrick Barrelino, optical
fiber, optical coupler and computer control using the “Scan” software. The light source provided the light in the
range 250nm – 1100nm. UV-Vis scan rate was 300nm/min, with a data interval of 0.50 nm and an average time
0.1s. The baseline was corrected by using Agilent Halon plate as 100% reflectivity standard. The evaluation of
the band-gap (Eg) was performed by using Origin applications after the received R% vs. wavelength spectra were
modified to [F(R)hv]1/2 vs. hv spectra.
Photocatalytic activity
The experiments for the evaluation of the photocatalytic activity of the powders have been conducted in a home
made photo-reactor, where all the samples were irradiated under visible light. The photo-reactor consisted of a
sealed and properly ventilated box to avoid any influence of the external radiation, equipped with a multiple
stirrer and a light system. Temperature was kept at 27±1ºC (room temperature). The irradiation system was
equipped by four parallel lamps, 380-790 nm, Daylight Sylvania, F18/54-765-T8 1050 Lumen with the average
irradiance measured at samples surface being 40±6W/m2 (plus 0.31±0.06 W/m2 of UV-light irradiance). A series
of experiments has been accomplished under the presence of Methylene Blue (MB) [3,7-bis (Dimethylamino)phenothiazin-5-ium
chloride]
and/or
Methyl
orange
(MO)
(Sodium
4-[(4dimethylamino)phenyldiazenyl]benzenesulfonate) that have been used as model pollutants. The experimental
procedure includes the preparation of a fresh aqueous solution of MB or MO (2mg/L), kept in the dark before
use. Each powder (333mg/L) was dispersed into this solution in a 250mL beaker and stirred for 40 min in the
dark (t=0) before the experiment starts. The pollutant/catalyst system was magnetically stirred under dark
conditions in order to reach the adsorption/desorption equilibrium. The suspensions obtained were irradiated
under visible light for five or more hours and aliquots have been periodically collected. After being centrifuged
for 20 min in 13000 rpm in order the nanoparticles to be removed, the absorbance spectrum of each aliquot is
measured in a UV-Vis spectrometer.
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
RESULTS AND DISCUSSION
Different synthetic route combinations and conditions have been investigated by altering various parameters,
namely pH, water to Ti-precursor ratio, duration of hydrolysis, loading of doping sources, aging time and
calcination temperature. The non-metal elements that have been investigated were N, S and C and in the case of
metal doped TiO2 various amounts of Ag have been evaluated. Based on the best performed proportions of the
non-metal and metal doped powders, co doping with Ag and N or S,N took place. In Table 1 the compositions
as well as the characterizations of the TiO2 powders that have been chosen to be presented in this work are
shown.
X-ray powders diffraction (XRD)
In order to evaluate the morphology, crystallinity and crystal size of the as prepared powders, XRD
measurements have been implemented. These types of measurements provide qualitative and morphological
information about the doped TiO2 powders mostly based on the ratio percentage of anatase to rutile phases. In all
cases the patterns of the powders (doped and non-doped) prepared by the sol-gel method and calcinated at 400
o
C for 2 h with a heating rate of 5oC/min have been monitored in order to elucidate the accompanied structural
characteristics followed by 500, 600, 700, and 800 oC. It should be noted that depending on the synthesis
parameters the powders showed an almost amorphous material at low temperatures that indicates a dominant
single anatase phase, while TiO2 that has been formed by a rapid hydrolysis shows a high rutile phase
transformation that has been indicated as a second phase. Figure 1 illustrates the X-ray diffraction spectra of
some [N]-doped, [S,N]-doped and [Ag-S,N]-co-doped TiO2 powders calcinated at 500oC. The spectra reveal
anatase as the predominant crystalline phase (2θ=25.25ο), that ranges from 66 to 84% in the case of [N]-doped
TiO2 powders, from 75 to 81% in the case of the [S,N]-doped powders and from 70 to 79% for the [Ag-S,N]-codoped powders. The calcination temperature increase results in the phase transformation from anatase to rutile,
where in the case of 800oC calcination the rutile phase dominates absolutely in all cases. The average crystallite
size of the synthesized powders (calculated by Scherrer’s equation), ranges from 8 to 12 nm both in the case of
the un-doped and doped TiO2 particles, regardless of the type of doping elements. Due to their small crystallite
size, all particles can be characterized as nanocrystalline. Moreover, as expected, the calcination resulted in
crystallite size increase. In Figure 1 it is clearly depicted that the doped TiO2 nanopowders are primarily in the
amorphous phase until 400 oC. Annealing at higher temperatures 500, 600, 700 oC leads to the formation of the
rutile and in most of the cases at 800 oC the anatase completely disappears, and rutile has been observed as the
dominate phase.
Table 1: Synthesized powders, synthesis conditions and characterizations.
Sample
Urea
(mmol)
Thiourea
(mmol)
AgNO3
(mmol)
Eg (eV)
Anatase %
Rutile %
Crystallite
size (nm)
1
-
-
-
3.0
82
18
11
2
16.7
-
-
3.0
66
34
9
3
83.3
-
-
2.3
82
18
-
4
166.5
-
-
2.2
78
22
11
5
249.8
-
-
2.3
73
27
9
6
333.0
-
-
2.3
77
23
10
7
499.5
-
-
2.2
84
16
9
8
666.0
-
-
2.3
64
36
8
9
-
13.1
-
3.0
-
-
12
10
-
65.7
-
2.4
75
25
-
11
-
131.4
-
2.2
78
22
10
12
-
197.0
-
2.3
81
19
12
13
-
131.4
0.382
3.0
79
21
8
14
-
131.4
0.765
2.3
75
25
12
15
-
131.4
1.531
3.0
70
30
-
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
Figure 1: XRD spectra of the doped TiO2 powders. A and R referred to the characteristic peaks of anatase and
rutile, respectively.
Scanning electron microscopy (SEM)
SEM and EDAX techniques were used to study the morphology and composition of the doped materials. SEM
images revealed in all cases a relative homogeneity regarding the shape and size, which is attributed to the
milling process applied. The EDAX spectrum was recorded in the binding energy region of 0-10keV. EDAX
analysis of the synthesized powders confirmed the presence of high Titanium and Oxygen content and at the
same time some relatively smaller amounts of Nitrogen and Sulphur (0.6-0.9%, atomic percentage) attributed to
the doping sources (urea and thiourea). The atomic percentage of Silver ranges from 0.14% to 0.30%. In Figure
2 and
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
Figure 3 the SEM images and EDAX spectra of a [S,N]-doped and [S,N-Ag]-co-doped TiO2 powders are
indicatively presented.
Figure 2: SEM image (mix mode image) (left) and EDAX spectrum (right) of the [S,N]-doped TiO2 (sample
11).
Figure 3: SEM image (mix mode image) (left) and EDAX spectrum (right) of the [Ag-S,N]-co-doped TiO2
(sample 14).
Fourier transform infrared spectroscopy (FTIR)
Σhe FTIR spectra revealed in all cases a broad peak in the range of 3700-2900cm-1 corresponding to the strongly
absorbing hydroxyl stretching vibrations and a peak at around 1630cm-1 , attributed to the bending vibration of
hydroxyl groups. In the case of the doped TiO2 powders several bands are detected in the 1500-1000cm-1 region.
These bands can be partly ascribed to the vibrations of N-Ti bond, as well as to Ti-O-N and Ti-O-S bonds. The
strong band located in the range of 400-800cm-1 is attributed to Ti-O stretching and Ti-O-Ti bridging stretching
modes. In the case of the [Ag-S,N]-co-doped TiO2 powders, a small shift of the characteristic bands towards
lower wavenumbers has been observed.
Band gap (Eg)
Considering that the synthesized TiO2 powders have an indirect energy band gap, the Eg was obtained by using
the modified Kubelka–Munk function, following by a Tauc plot (F(R).hν)1/2 vs. E(=hc/λ) [14].The KubelkaMunk function F(R), enables the optical absorbance of the samples to be estimated from its reflectance. The
band gap for optical absorption is defined as the minimum energy required to excite an electron from the ground
state (HOMO, at the top of the valence band) to the exciting state (LUMO, at the bottom of the conduction
band). As shown in Figure 4, in the case of the undoped TiO2 powder (sample 1), the energy band gap (E g) was
found 3.0eV, whereas in most of the cases of the [N]- and [S,N]-doped TiO2 powders the band gap has been
significantly reduced beyond 2.5eV in comparison to the undoped TiO2. These results demonstrate that the
doping with urea and thiourea causes a narrowing of the band gap suggesting that the obtained powders could be
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
active in the visible light range. In the case of the [Ag-S,N]-co-doped TiO2 powder (sample 14) there is also a
significantly reduced Eg, but this stands only for a certain thiourea/silver nitrate amount combination. As it can
Figure 4: Plots of [F(R)hv]1/2 vs hv for indirect transition of an undoped (sample 1), [N]-doped (sample 4),
[S,N]-doped (sample 11) and [Ag-S,N]-co-doped (sample 14) powder.
be seen from Table 1, the low dopant amounts of urea and thiourea (16.7 and 13.1mmols, respectively, per
29.4mmol of Ti-precursor) do not cause any improvement in the E g, which in both cases has been estimated at
3.0eV (samples 2 and 9) like in the case of the undoped TiO2. By increasing the dopant amounts, the Eg has been
significantly improved, with the best E g value accomplished at 166.5mmols of urea and 131.4 mmols of thiourea
(sample 4 and 11). Further increase of dopant amounts resulted in good though not further improved E g . On the
other hand, in the case of the [Ag-S,N]-co-doped TiO2 powders only a certain amount of silver nitrate resulted in
Eg narrowing and the decrease or increase of this certain amount not only does not reduce or increase the E g of
the new particles but also prevents the E g reduction due to thiourea doping that has been well confirmed until
now.
Photocatalytic activity
The powders exhibiting the best morphological, structural as well as functional properties have been further
evaluated concerning their photocatalytic activity under visible light irradiation.
Figure 5 presents the photocatalytic performance of a [S,N]-doped and a [Ag-S,N]-co-doped TiO2 powder
(sample 11 and 14, respectively) under visible light irradiation, using Methyl orange and Methylene blue as
model pollutants (Graphs of normalized degradation C and D, respectively). In both experiments the
photocatalytic performance of the two powders was quite satisfactory after 5 hours of constant visible light
irradiation. More specifically, the photocatalytic degradation activity of the [S,N]-doped and [Ag-S,N]-co-doped
powders, expressed as the removal% of methylene blue in the aqueous solution, was 84% and 78%, respectively,
after five hours of visible light irradiation. In a similar experiment using methyl orange as model pollutant, the
photocatalytic activity of the two powders was also significant, with the [S,N]-doped and [Ag-S,N]-co-doped
powders exhibiting 64 and 67% removal of methyl orange in the aqueous solution, respectively. In order to
estimate the effectiveness of the photocatalytic performance a benchmark material, Degussa P25 was used where
the degradation has been estimated to be around 23 %. The characteristic absorbance spectra (Images A and B)
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
show the photocatalytic activity of Degussa P25 and sample 14 during five hours of constant visible light
irradiation.
Figure 5: Photocatalytic activity of [S,N]-doped (sample 11) and [Ag-S,N]-co-doped (sample 14) powders in
methylene blue and methyl orange degradation under visible light irradiation.
CONCLUSIONS
In the present experimental study an attempt has been made to improve the photocatalytic activity of TiO2
particles by doping with metal and non-metal elements, namely nitrogen, sulfur and silver. All the compositions
were characterized using XRD, SEM, EDX and DR-UV-Vis spectrophotometry. The most representative results
were presented and discussed. The improvement of TiO2 photocatalytic activity was focused in producing doped
TiO2 particles with specific structural and morphological characteristics as well as functional properties. The
synthesized doped TiO2 particles were anatase based nanocrystalline materials and exhibited significantly
reduced bang gap energy (E g) beyond 2.5eV. Moreover, the new doped TiO2 particles demonstrated a quite
10ο ΠΑΝΔΛΛΗΝΙΟ ΔΠΙ΢ΣΗΜΟΝΙΚΟ ΢ΤΝΔΓΡΙΟ ΥΗΜΙΚΗ΢ ΜΗΥΑΝΙΚΗ΢, ΠΑΣΡΑ, 4-6 ΙΟΤΝΙΟΤ, 2015.
satisfactory photocatalytic activity in the model pollutants tested, by removing the pollutants up to 84%. This
first evaluation renders these catalysts as promising doped TiO2 based photocatalytic materials ready to be
exploited in various applications, among others in the environmental pollutants degradation.
ACKNOWLEDGEMENTS
This work has partially received funding from the European Union’s Seventh Framework Programme for
research, technological development and demonstration under grant agreement No. 609345 (MF-Retrofit:
“Multifunctional facades of reduced thickness for fast and cost-effective retrofitting”).
REFERENCES
[1]. Zaleska A., Recent Patents on Engineering, 2, 157-164, (2008).
[2]. Ohama Y., D. Van Gemert, ISBN 978-94-007-1296-6, (2011).
[3]. Guarino M., Costa A., Porro M., Bioresource Technology, doi:10.1016/j.biortech.2007.04.25, (2007).
[4]. Mitoraj D., PhD, (2009).
[5]. Rockafellow E. M., Stewart L. K., Jenks W. S., Applied Catalysis B: Environmental 91 554–562, (2009).
[6]. Asahi R., Morikawa T., Ohwaki T., Aoki K., Taga Y., Science (Washington, DC,U. S.), 293, 269-271,
(2001).
[7]. Liu H., Imanishi A., Nakato Y., J. Phys. Chem. C 111 8603–8610, (2007).
[8]. Tachikawa T., Takai Y., Tojo S., Fujitsuka M., Irie H., Hashimoto K., Majima T., The Journal of Physical
Chemistry, B 110 13158–13165, (2006).
[9]. Chen X. and Burda C., Journal of the American Chemical Society, Vol.130, No.15, pp 5018-5019, (2008).
[10]. Hamal D.B., Klabunde K. J., Journal of Colloid and Interface Science 311 514–522, (2007).
[11]. Li Y., Qin Z., Guo H., Yang H., Zhang G., Ji S., Zeng T., PloS One., 9(12):e114638, (2014).
[12]. Carrera-Lopez R., Castillo-Cervantes S., Superficies y Vacio 25(2) 82-87, (2012).
[13]. Kim D. S., Han S. J., Kwak S. Y., Journal of Colloid and Interface Science, Vol.316, Issue 1, pages 85-91,
(2007).
[14]. J.Tauc, Plenum Press, London, (1974).