Thin Solid Films 495 (2006) 327 – 332

www.elsevier.com/locate/tsf

Photocatalytically active TiO2 thin films produced by surfactant-assisted

sol–gel processing
Urh Černigoj a, Urxka Lavren*i* Štangar a,b,*, Polonca Trebxe a,
Urxa Opara Kraxovec c, Silvia Gross d
a
Laboratory for Environmental Research, Nova Gorica Polytechnic, P.O.B.301, 5001 -Nova Gorica, Slovenia
b
National Institute of Chemistry, Hajdrihova 19, 1000 -Ljubljana, Slovenia
c
University of Ljubljana, Faculty of Electrical Engineering, Traxka 25, 1000 -Ljubljana, Slovenia
d
CNR -ISTM, University of Padova, via Marzolo, I -35131 Padova, Italy
Available online 26 September 2005

Abstract

Thin TiO2 films were prepared from a titanium isopropoxide precursor by sol – gel processing with or without various nonionic surfactant
molecules (Brij 56, Triton X-100 or Pluronic F-127). The photocatalytic efficiency of the transparent films obtained by a dip-coating technique
was found to depend strongly on the use of and type of surfactant added. Titania/Pluronic sols resulted in homogeneous and crack-free TiO2
anatase films with a thickness as much as 300 nm after one dipping and heat-treatment (500 -C) cycle. Optical properties of the films were
characterized by UV-Vis spectroscopy and crystalline structures by X-ray diffraction. A surfactant-assisted sol – gel process retarded crystallization
of the anatase titania films, which resulted in smaller grain sizes (down to 10 nm) and presumably a larger active surface area. The morphology of
the film surfaces as obtained by SEM techniques could be also correlated with the results of our photodegradation studies. The photocatalytic
activity of the films was enhanced by first coating the glass substrate with a SiO2 protective layer prior to the deposition of the titania film. For our
in situ studies of photodegradation we chose the monoazo dye Plasmocorinth B as a model compound as it is stable under environmental
conditions and its degradation products are not coloured. The highest photobleaching rate was found for films deposited from the sol with addition
of the Pluronic surfactant and it was almost twice as high as that for films deposited from sols without the surfactant.
D 2005 Elsevier B.V. All rights reserved.

Keywords: Titania coatings; Sol – gel; Photocatalysis; Azo dye

1. Introduction ease and versatility of processing, homogeneity at a molecular

level, as well as mild and ambient atmosphere reaction
Titania in the form of thin films has attracted a great deal of conditions. As smaller grain sizes are usually associated with
interest as a low cost and very versatile material for higher surface areas [7– 9], the conventional alkoxide sol –gel
applications in photocatalysis, dye-sensitized photoelectro- route has been modified with the addition of a templating agent
chemical cells, electrochromic devices, sensors, and many to produce such smaller grains. A large surface area can be the
other uses. For many applications a nanostructured titania film determining factor in photodegradation reactions because a
is preferable to one which is dense, or compact. Numerous large contact surface exposed to organic pollutant molecules
reports have appeared on highly porous titania film structures promotes the reaction rate [7,9]. However, smaller particles can
either with large pores [1] or ordered mesopores [2– 5]. The also promote the recombination of photogenerated electrons
nanocrystalline structure of the films is usually produced and and holes which in turn leads to a poor photoactivity. Hence, a
controlled by surfactant-assisted sol – gel processing [e.g. 6]. balance between surface area and crystallinity must be found in
Sol –gel techniques offer some advantages compared to other order to obtain the optimum photocatalytic activity [7,10]. A
solution methods or gas phase deposition techniques, such as titania particle size of around 10 nm has been suggested to
produce the highest photocatalytic oxidation rates [11].
* Corresponding author. Nova Gorica Polytechnic, Vipavska 13, 5000 Nova
In this work we utilized several commercially available,
Gorica, Slovenia. Tel.: +386 5 3315 241; fax: +386 5 3315 296. non-ionic surfactants (Brij 56, Pluronic F-127, Triton X-100)
E-mail address: [email protected] (U.L. Štangar). with polyether chains of different lengths by applying them in
0040-6090/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.tsf.2005.08.240
328 U. Černigoj et al. / Thin Solid Films 495 (2006) 327 – 332

the sol – gel process with the titanium isopropoxide precursor of the TiO2 films was increased by repeating the dipping and
to induce nanocrystallinity in the resulting inorganic films in an heat-treatment cycle up to five times. Powder samples were
attempt to increase their photocatalytic activity for degradation prepared by casting the dip-coating solution into petri dishes.
of organic pollutants in waste water. In situ photodegradation After solvent evaporation the remaining thick film was calcined
of the azo dye Plasmocorinth B in aqueous solution in the at 500 -C for 2 h.
presence of various titania thin films was the model system in Silica sol was prepared from a tetraethoxysilane (TEOS,
our study. The reasons for the choice of this particular dye Aldrich)precursor, ethanol (Si/EtOH = 0.15), HNO3 (HNO3/
are:(i) it is stable under environmental conditions, (ii) its Si = 0.265)and water (H2O/Si = 8.1). After two hours of mixing,
degradation products are not colored and therefore easy to the resulting solution was used for silica film deposition by the
follow with UV-Vis spectroscopy, (iii) it does not adsorb dip-coating technique with pulling speeds up to 5 cm/min. The
extensively on the titania film so that the discoloration of the sol was unstable and therefore was always freshly prepared. All
dye solution could be directly related to its decomposition. silica films were produced by a single dipping. The silica films
The photocatalytic efficiency of titania films is also affected on soda – lime glass were used as substrates for the deposition
by additional factors such as the presence of cations (Na+, of titania thin films as described above.
Ca2+)which may diffuse from the glass substrate in the interior
of the film [12], and have been found to have a detrimental 2.2. Characterization techniques
effect on the photocatalytic activity [13,14]. To reduce the
migration of cationic impurities into the active film we covered The thickness of the films was measured on a Taylor-
the soda – lime glass substrates with a thin silica film made by a Hobson Talysurf profilometer. UV-Vis spectra of the films and
sol – gel route prior to depositing the titania film. the dye solutions were recorded on a Hewlett-Packard 8453
Crystallinity of the thin films and powders calcined at 500 UV-Vis spectrophotometer. X-ray diffraction measurements
-C was examined by XRD analysis. It is known that the anatase were made with a Philips PW1710 automated X-ray diffrac-
crystal structure shows greater photocatalytic activity than tometer using graphite monochromatized Cu Ka radiation in
rutile form for most reactions. The reason for this is rather the step-by-step mode. SEM images of the films were obtained
unclear. It has been ascribed to a slightly higher Fermi level for on a Supra 35 VP Carl Zeiss field-emission scanning electron
anatase, a lower capacity to adsorb oxygen and a higher degree microscope with an accelerating voltage of 1 kV. The surface
of surface hydroxylation [7]. Hydroxyl groups on the surface and in-depth composition of films were analysed by X-ray
have an important role in the photodegradation process since Photoelectron Spectroscopy (XPS). XPS spectra were run on a
they directly trap photogenerated holes to produce the very Perkin-Elmer A 5600 ci spectrometer.
reactive OH&radicals which subsequently oxidize the organic A continuous flow reactor for the experiments on the
pollutant. They also have an indirect role in producing oxygen photocatalytic activity of the films was constructed. The reactor
radicals and hindering the electron-hole recombination. could be purged with different gases, was cooled with tap water
and had on-line measuring capability of the UV-Vis spectra of
2. Experimental the solution. The as-prepared titania films were kept in
deionised water prior to the photocatalysis experiments. A 10
2.1. Materials preparation mm thick solution of NaBr (110 g) and Pb (NO3)2 (0.69 g) in
water (230 g) was used as a 335 nm cut-off filter in front of the
Titania sols were made from a modified Ti(OPri)4 precursor photocatalytic cell and a 125 W Xe lamp (Cermax xenon
(Fluka) by addition of acetic acid (CH3COOH/Ti = 1.0)(sols U, parabolic lamp) as a light source. A titania film on one side of a
R, T) or ethyl acetoacetate (EAA/Ti = 1.0)(sols D, C) in order to silica covered glass support (25  70 mm) was immersed in the
slow down the solvolysis – condensation of the titanium dye solution next to the wall of the photocatalytic cell and
alkoxide. The resulting exothermic reaction gave rise to a irradiated (surface 23  23 mm) along the normal direction.
stable solution into which the solvent 2-methoxyethanol was The dye solution (aqueous solution (4.8 mL) of Plasmocorinth
added [15] ((Ti/MeO(CH2)2OH = 0.075)without surfactant (sol B (40 mg/L)) was continuously purged with oxygen during the
U, D)or with surfactant, namely polyoxyethylene(10)cetyl irradiation. A peristaltic pump (Heidolph PD 5001) with a
ether (Brij 56, Aldrich)(15 wt.%, sol R), polyoxyethylene(10)i- silicon hose was used to drive the solution from the
sooctylphenyl ether (Triton X-100, Aldrich)(15 wt.%, sol T) or photocatalysis cell to the cell positioned in the UV-Vis
PEO100PPO65PEO100 triblock copolymer (Pluronic F-127, spectrophotometer for on-line absorbance measurements and
Sigma)(9.7 wt.%, sol C). Surfactant weight percentages are back to the photocatalytic cell at a flow rate of 10 mL/min.
given with respect to the total weight of solution. The resulting
solution was stirred at room temperature for at least three 3. Results and discussion
hours. The yellowish transparent sols were stable in a
refrigerator for several months. Coatings were made in the 3.1. Thickness and optical quality of the films
dip-coating unit with pulling speeds of up to 10 cm/min. The
supporting soda – lime glass plates (bare or coated with All titania films obtained after calcination were of a high
SiO2)had been cleaned with ethanol and dried. The xerogel optical quality (transparent, crack-free) and free of any
films were calcined at 500 -C for 30 to 90 min. The thickness organic residues as confirmed by IR spectroscopic measure-
 U. Černigoj et al. / Thin Solid Films 495 (2006) 327 – 332 329

1600
over the whole visible light region. Bands from the interference
C on SiO2-glass C
1400 R on SiO2-glass
color of the films appeared in the wavelength range 350– 800
R on bare glass nm and have higher amplitudes in the case of thinner films
1200 U on bare glass made without surfactant. Similar but much stronger effect was
Film thickness (nm)

U on SiO2-glass
1000 D on SiO2-glass observed for TiO2/PDMS hybrid films, where the interference
peaks were even eliminated by the addition of PDMS, which
800
increased surface roughness and porosity of titania [16].
R
600 Accordingly, such phenomena in UV-VIS spectra can be
ascribed to the change in density and hence in porosity of
400 U,D
the film structure. When our samples are compared, the films
200 made by using templates possess higher porosity (which is also
discussed in the next section) and therefore their UV-VIS
0
1 2 3 4 5 spectra are characterized by less pronounced interference peaks
Number of dippings (Fig. 2, C and R spectra). In addition it can be observed that
interference fringes differ even inside one particular spectrum
Fig. 1. Relationship between film thickness and number of dippings from
various sols.
(e.g. spectrum C) suggesting that films are multilayers with
various refractive indexes.This may be due to repeated
ments of the films deposited on Si-wafers. Fig. 1 shows the coating –heating cycles, where the first layer underwent longer
dip-coating efficiency of the titania films made from different heat-treatment time than the second one and so on. It should be
sols. It is apparent that the addition of surfactant in the sol also noted that titania films are deposited on glass with a thin
results in a thicker film obtained by the same number of SiO2 protective layer, which may also contribute to the
dipping and heat-treatment cycles. This is a consequence of observed phenomenon.
increased viscosity of the sol upon addition of a polymeric Inset in Fig. 2 shows the UV absorption spectra of films
templating agent. The highest efficiency was achieved by deposited from sol C which contained Pluronic surfactant. It is
using Pluronic (Brij and Triton gave similar results), which clearly seen that an increase in the number of dippings
has the longest polymeric chains. Film thickness obtained by produced a red-shift of the absorption edge of the film, which
single dipping could be as high as 300 nm without was also observed for all other films in agreement with other
deterioration of the optical quality. A linear increase in film preparation methods [12,17,18]. The shift has been ascribed to
thickness with the number of dippings was observed in all differences in the size of the crystallites. Thicker films that
cases. It is also evident from Fig. 1 that type of the substrate underwent longer heat treatment by repeated coating – heating
(bare glass or SiO2-coated glass) has no influence on the cycles have relatively larger anatase crystallites and this causes
resulting thickness of the titania film. the onset of absorption to shift to the red part of the spectrum
UV-Vis transmittance spectra of various films prepared with [4]. Absorption of light below 370 nm is due to the excitation
four dippings are presented on Fig. 2. The transparency of all of electrons from the valence band to the conduction band of
films is high and the film with greatest thickness (ca.1200 nm TiO2. Pure anatase crystalline powder has an intense absorption
on both sides, curve C) still has a transmittance of around 80% band at 335 nm [19]. A red-shift of the absorption edge also

Fig. 2. UV-Vis transmittance spectra of TiO2 films deposited from various sols on SiO2-glass substrates with 4 dipping – heating cycles. Inset: UV-Vis absorbance
spectra of TiO2 films deposited from sol C with 1 to 5 dipping – heating cycles.
330 U. Černigoj et al. / Thin Solid Films 495 (2006) 327 – 332

180
bare glass support was thereby estimated for films prepared

(101)
160
with surfactant. The slightly larger anatase grains of films
deposited on bare glass substrates may be a consequence of the
140 higher concentration of Na+ ions diffused from the substrate.
Indeed, the first results obtained by X-ray photoelectron
Intensity (a.u.)

120 spectroscopy showed that a relevant atomic percentage of

(004)
C sodium (4 –5%) was detected along the film thickness of titania
100
deposited on bare glass support, while the concentration of
80 D alkali metal ions in titania film deposited on SiO2-glass was
under detection limit. Na+ ions have been found to act as a flux
S material for crystal growth [21]. It has been also suggested that
60
diffusion of sodium ions from the substrate during thermal
40 treatment could stimulate the recrystallization of the anatase to
20 30 40 50 60
rutile [22]. Our studies suggest that crystallization is slower for
2θ (degree)
the surfactant-assisted process compared with that in surfac-
Fig. 3. XRD patterns of TiO2 films deposited from sol C (with Pluronic) and D tant-free systems. Smaller grains favour a larger inner surface
(without Pluronic) on SiO2-glass substrate with 4 dipping – heating cycles (S: area, which we achieved effectively with the use of surfactant.
XRD pattern of the substrate). BET surface area measurements carried out on powder samples
gave preliminary values of 122 m2/g for TiO2 prepared in the
indicates a decrease in the band gap of TiO2 with increasing presence of Triton and 18 m2/g when prepared without
number of coating –heating cycles. surfactant.
Larger crystallites were found for powder samples (Table 1),
3.2. Crystalline structure and morphology where the estimated grain size was 17– 23 nm for surfactant-
assisted process and 29 nm for surfactant-free process of
Typical XRD diffraction patterns of films deposited on powder preparation. The diffraction patterns of the calcined
SiO2-coated glass substrates are shown in Fig. 3. The only one powder samples also showed the presence of the rutile phase in
major peak at 2u = 25.4- corresponds to (101) reflections of the agreement with the results obtained by Martyanov and
anatase phase, which is present in all our thin film samples. The Klabunde [18]. These authors made a comparative study of
broad background originates from the glass substrate. The TiO2 particles in powder form and as a thin nanostructured film
diffraction pattern of a film produced without surfactant added on quartz. They pointed out that TiO2 particles immobilized on
to the titania sol revealed a peak at 2u = 37.9-, which the quartz support remained in the anatase form even at 800 -C,
corresponds to (004) reflections of anatase. These observations whereas self -supported TiO2 particles in powders were more
show that both the orientation and the growth of anatase easily converted into the rutile form starting already at 500 -C.
crystallites in the solid film are influenced by the addition of The percentage content of rutile phase relative to anatase
surfactant in the initial sol. The use of surfactant contributes to (calculated from the intensity ratio I rutile(110) / I anatase(101)) in
the formation of highly oriented anatase phase along the (101) our powder samples decreased from the highest value of
crystallographic plane and to the smaller mean grain size for 31%for titania made without surfactant to 16% (with Triton),
films and powders (Table 1). The grain size, i.e. the effective 9% (with Brij)and to the lowest value of 5% for titania made
size of coherently scattering domains, was calculated from the with Pluronic surfactant.Pluronic was therefore found to have
(101) peak of anatase using the Scherrer’s equation: L = K k / the biggest impact by the greatest retardation of the formation
bcos u [20]. Here k is the wavelength of X-rays (1.54 A for Cu of the rutile phase and by producing the smallest size of anatase
K a radiation), K is a constant taken to be 0.9 (assuming no
crystal distortion in the lattice), b is the full width at half
maximum of the signal in radians corrected for the instrumental
broadening (0.07-), and u is the diffraction angle in degrees. A
crystallite size of 10 –11 nm on SiO2-glass and 13 –15 nm on

Table 1
The effective size of coherently scattering domains (=grain size)of TiO2
powder (p) and film (f) samples (calculated using Scherrer’s equation)
Calcination time, Average grain size (nm) for samples from
sample form different sols:
U T R C
(no surfactant) (Triton) (Brij) (Pluronic)
2h, p. 29 22 23 17
4  30V, f. on SiO2-glass 18 11 10
Fig. 4. SEM image of TiO2 film deposited from sol C on SiO2-glass substrate
4  30V, f. on glass 17 15 13 with 4 dipping – heating cycles.
 U. Černigoj et al. / Thin Solid Films 495 (2006) 327 – 332 331

nanocrystals. However, no ordered mesoporous structure was photocatalyst under the irradiation conditions used for all
obtained as would have been evident by the appearance of a photocatalysis experiments.
low -angle XRD peak. Fig. 5 shows photobleaching curves (absorbance at 527 nm
Scanning electron microscopy (SEM)was employed to as a function of irradiation time) in aqueous solution of
characterize the surface morphology of the manufactured films. Plasmocorinth B in the presence of various titania films made
TiO2 films made without surfactant addition in the sol exhibit a by four coating – heating cycles. Only one side of the TiO2/
flat and ill-defined surface, whereas TiO2 films made with SiO2-glass slide was in contact with the solution. It is clearly
surfactant addition in the sol (Fig. 4 with Pluronic) exhibit a seen that the film made with Pluronic surfactant added showed
rougher but still moderately flat texture and well-structured the highest photodegradation rates compared with the other
granular nanosurface with clear interstices between the films with the same number of layers. The dye solution with
particles/aggregates. Crystallite sizes obtained from XRD Pluronic –titania based films was almost completely discolored
measurements (10 nm) are only roughly comparable to the in less than 1 h of irradiation. All photobleaching curves
particle size obtained from SEM photographs. The well- indicate first-order kinetics. The highest photodegradation rate
structured morphology displayed in the SEM image of constant (7.3  10 4 s 1 for films made from sol C with 4
templated films is likely a consequence of template removal dippings) was almost twice that for films deposited from sols
during the thermal treatment of thin films. It creates a higher without the surfactant (3.8  10 4 s 1 for films made from sol
degree of porosity, which should have a beneficial influence on D with 4 dippings). The reasons for the Pluronic – titania based
the photocatalytic activity of the templated films provided that films having the highest efficiency are that (i) the films are
the pores are accessible for an organic pollutant. The difference thicker than the others, which results in an increase of the UV
in porosity among the various films studied was indirectly absorbance, and that (ii) the grains are smaller and exhibit a
estimated from the densities of the films obtained by taking well-structured surface, which gives in higher photochemical
into account the film thickness and its weight. It was found that reaction rates because of the increased active surface area. An
Pluronic-templated film (2200 g/dm3)was only half as dense as appropriate balance between the two factors should be
the film prepared in the absence of surfactant (4300 g/dm3). achieved to obtain highest photoefficiency. First, while films
The former one could be therefore considered as much more have to be sufficiently thick to achieve sufficient catalyst
porous than the latter one. All coatings were crack-free and loading, opacity and internal mass transfer resistance is also
uniform on a micrometer scale. A rougher and more porous increased with increased thickness. This could lead to higher
surface can play an important role in the photocatalytic activity electron/hole recombination rates. Aggregation and growth of
since the photocatalysis reaction occurs at the surface, and a TiO2 particles in the interior region of thick films, which
larger surface area provides for more photocatalysis reactions underwent longer calcinations times, may also cause a decrease
to take place. in the number of surface active sites [12]. Second, as far as the
optimal grain size is concerned, smaller grains result in a higher
3.3. Photodegradation active surface area, but electron-hole recombination rates are
also higher for the smaller particles.
An aqueous solution of Plasmocorinth azo dye (chemical
structure is shown in Fig. 5 as an inset) was used for 4. Summary
photodegradation studies in the continuous flow cell described
above. The dye solution was stable in the absence of the TiO2 Nanostructured and transparent titania films with anatase
grain sizes of as small as 10 nm were deposited on SiO2 pre-
coated soda – lime glass substrates in order to enhance their
activity for the photocatalytic degradation of an azo-dye
(Plasmocorinth B). The use of the Pluronic F-127 templating
agent in a surfactant-assisted sol –gel process reduced the
number of dip coatings necessary to obtain a sufficient titania
loading with photocatalytically active surface area. The films
were immersed in the colored solution and photobleaching was
followed in situ to near completion in one hour. The titania
catalyst can be easily removed from the solution, which is one
of the principal advantages of using the immobilized films as
catalysts rather than titania powders.

Acknowledgements

Authors wish to thank Prof. B. Orel and Prof. V. Kau*i* for

Fig. 5. Photobleaching of Plasmocorinth B solution (absorbance at 527 nm vs. valuable discussions and XRD measurements, respectively,
irradiation time) in the absence and presence of TiO2 films deposited from Dr. M. Čekada for thickness measurements, D. Strm*nik and
various sols on SiO2-glass substrates with 4 dipping – heating cycles. M. Zorko for SEM and Dr. K. Dahmouche for BET surface
332 U. Černigoj et al. / Thin Solid Films 495 (2006) 327 – 332

area measurements. We are very grateful to Prof. J. Eckert for [10] K. Mogyorósi, I. Dékány, J.H. Fendler, Langmuir 19 (2003) 2938.
revising the paper. This work was financially supported by the [11] Z.B. Zhang, C.C. Wang, R. Zakaria, J.Y. Ying, J. Phys. Chem., B 102
(1998) 10871.
Ministry of Higher Education, Science and Technology of the [12] J. Yu, X. Zhao, Q. Zhao, Mater. Chem. Phys. 69 (2001) 25.
Republic of Slovenia. [13] C. Guillard, B. Beaugiraud, C. Dutriez, J.-M. Herrmann, H. Jaffrezic, N.J.
-Renault, M. Lacroix, Appl. Catal., B Environ. 39 (2002) 331.
References [14] Y. Yang, X. Li, J. Chen, L. Wang, J. Photochem. Photobiol., A Chem. 163
(2004) 517.
[1] S. Bu, Z. Jin, X. Liu, H. Du, Z. Cheng, J. Sol – Gel Sci. Technol. 30 (2004) [15] U.L. Štangar, B. Orel, A. Régis, Ph. Colomban, J. Sol – Gel Sci. Technol.
239. 8 (1997) 965.
[2] E.L. Crepaldi, G.J. de A.A Soler-Illia, D.D. Grosso, F. Cagnol, F. Ribot, [16] K. Iketani, R.-D. Sun, M. Toki, K. Hirota, O. Yamaguchi, J. Phys. Chem.
Solids 64 (2003) 507.
C. Sanchez, J. Am. Chem. Soc. 125 (2003) 9770.
[3] P.C.A. Alberius, K.L. Frindell, R.C. Hayward, E.J. Kramer, G.D Stucky, [17] C. Guillard, D. Debayle, A. Gagnaire, H. Jaffrezic, J.-M. Herrmann,
B.F. Chmelka, Chem. Mater. 14 (2002) 3284. Mater. Res. Bull. 39 (2004) 1445.
[4] J.C. Yu, X. Wang, X. Fu, Chem. Mater. 16 (2004) 1523. [18] I.N. Martyanov, K.J. Klabunde, J. Catal. 225 (2004) 408.
[19] H. Yoshitake, T. Sugihara, T. Tatsumi, Chem. Mater. 14 (2002) 1023.
[5] H. Yun, K. Miyazawa, I. Honma, H. Zhou, M. Kuwabara, Mater. Sci.
Eng., C 23 (2003) 487. [20] H.P. Klug, L.E. Alexander, X-ray Diffraction Procedures, 2nd edR, John
[6] E. Stathatos, P. Lianos, C. Tsakiroglou, Langmuir 20 (2004) 9103. Wiley and Sons, 1973, p. 687.
[21] H.-J. Nam, T. Amemiya, M. Murabayashi, K. Itoh, J. Phys. Chem., B 108
[7] O. Carp, C.L. Huisman, A. Reller, Prog. Solid State Chem. 32 (2004) 33.
[8] P. Bouras, E. Stathatos, P. Lianos, C. Tsakiroglou, Appl. Catal., B (2004) 8254.
Environ. 51 (2004) 275. [22] C.C. Trapalis, P. Keivanidis, G. Kordas, M. Zaharescu, M. Crisan, A.
[9] K. Mogyorósi, J. Németh, I. Dékány, J.H. Fendler, Prog. Colloid & Szatvanyi, M. Gartner, Thin Solid Films 433 (2003) 186.
Polym. Sci. 117 (2001) 88.