Photocatalytic degradation of several VOCs (n-hexane, n-butyl acetate and toluene) on TiO2 layer in a closed-loop reactor

Article abstract of DOI:10.1016/j.cattod.2012.10.017

This paper deals with photocatalytic degradation of three VOCs differing in chemical structure (n-hexane, n-butyl acetate and toluene) in a closed-loop reactor. Degradation tests were performed on particulate layers prepared from TiO₂ VLP7000 (Kronos). Concentration dependencies of VOCs were simulated and initial degradation rates calculated. Dependence of the latter on initial concentrations confirms single-site Langmuir-Hinshelwood kinetics with value of hexane adsorption constant 0.023 m³ mmol^-1. It was shown that butyl acetate is oxidized preferably to hexane when they react in a mixture, while their individual degradation seems to be kinetically similar. Toluene deactivates the photocatalyst and its degradation is thus very slow but the catalyst can be renewed by thermal treatment. The found selectivity or inhibitive ability of investigated VOCs in mixture is important for the proper choice of a suitable purification method of industrial exhaust fumes.

Full text of DOI:10.1016/j.cattod.2012.10.017

Catalysis Today 209 (2013) 153–158
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Photocatalytic degradation of several VOCs (n-hexane, n-butyl acetate and
toluene) on TiO2 layer in a closed-loop reactor
Frantisˇek Moulis∗, Josef Kry´sa∗∗
Institute of Chemical Technology, Prague, Technická 5, 166 28 Prague 6, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 June 2012
Received in revised form 3 October 2012
Accepted 16 October 2012
Available online 3 December 2012
This paper deals with photocatalytic degradation of three VOCs differing in chemical structure (n-hexane,
n-butyl acetate and toluene) in a closed-loop reactor. Degradation tests were performed on particulate
layers prepared from TiO2 VLP7000 (Kronos). Concentration dependencies of VOCs were simulated and
initial degradation rates calculated. Dependence of the latter on initial concentrations confirms single-
site Langmuir–Hinshelwood kinetics with value of hexane adsorption constant 0.023 m3 mmol−1. It was
shown that butyl acetate is oxidized preferably to hexane when they react in a mixture, while their
individual degradation seems to be kinetically similar. Toluene deactivates the photocatalyst and its
degradation is thus very slow but the catalyst can be renewed by thermal treatment. The found selectivity
or inhibitive ability of investigated VOCs in mixture is important for the proper choice of a suitable
purification method of industrial exhaust fumes.
Keywords:
Photocatalysis
Gas-phase
n-Hexane
n-Butyl acetate
Toluene
© 2012 Elsevier B.V. All rights reserved.
Closed loop
concentrations (␮g/m3, order of ppb). However, this technique
1. Introduction
brings numerous challenges as well, e.g. the maintenance of micro-
Harmful agents, e.g. volatile organic compounds (VOCs), are lib-
erated into the environment [1] during a large range of industrial
applications. There are two main domains of photocatalytic appli-
cations depending on the degree of air pollution: high-polluted
gas from industrial processes (agents in concentrations in order
of hundreds ppm) and low-contaminated air (order of units ppm).
Several conventional methods for VOC removal from industrial
exhausts exist differing in concentrations they can be applied on,
removal efficiency, costs, secondary risks or some other restraints
[2]. Thermal oxidation, catalytic oxidation, carbon adsorption or
absorption in a liquid solvent allow removal efficiency 90–99%,
condensation is less effective (50–90%). All these methods produce
secondary waste in the form of combustion products, spent carbon,
wastewater or condensate, which must be treated subsequently.
From the industrial point of view, carbon adsorption can han-
dle VOC concentrations from 20 to 5000 ppm and can be combined
with VOC recover; the organic solvent is commonly released by
heating the carbon with steam and then separated by means of
decantation or distillation [3]. Besides the techniques mentioned
above, biological treatment can be used alternatively at low VOC
bial growth in the polluted medium [4]. Comparing photocatalytic
degradation to contemporary (conventional) processes, the former
leads to the decomposition of pollutant towards non-harmful com-
pounds in one step at ambient conditions; no secondary waste is
produced. This makes from photocatalytic process a method with
good performance [5], especially at low concentrations [6–8].
In industrial domain of application, only continuous reactors
seem to be efficient regarding the large flow rates treated [9,10].
However, in this paper we describe a batch (closed-loop) reactor,
which is more convenient for the evaluation of photocatalyst.
Heterogeneous photocatalytic reaction occurs in the presence
of photon at sufficient energy (wavelength), catalytic surface and
oxidizing agent (mostly molecular oxygen). The photon energy
must be higher than the photocatalyst band-gap one. Photocata-
lyst surface excitation leads to the formation of free electrons in
the conductivity band and positively charged holes in the valence
band. These electron–hole pairs are then used in oxidative mecha-
nisms to form hydroxyl radicals (from water) and superoxide ions
(from oxygen). Besides these reactions reverse recombination of
electron–holepairsexistsandis themainlimitingfactorthatlowers
the efficiency of photocatalytic processes [11].
Already in 1964 McLintock and Ritchie showed that organic
compounds ethylene and propylene can be completely oxidized
to CO2 and H2O [12]. Today, comparing the number of scientific
publications to the number of patents dealing with photocatalysis,
the majority of scientific publications deals with the degradation in
∗
Corresponding author. Tel.: +420 220 444 390; fax: +420 220 444 410.
Corresponding author. Tel.: +420 220 444 112; fax: +420 220 444 410.
E-mail addresses: frantisek.moulis@vscht.cz, moulisf@vscht.cz (F. Moulis),
∗∗
josef.krysa@vscht.cz (J. Kry´sa).
0920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2012.10.017

154
F. Moulis, J. Kry´sa / Catalysis Today 209 (2013) 153–158
Fig. 1. Apparatus scheme. IRA – mobile infrared analyser; DP – diaphragm pump.
water phase, while the patents are focused rather on the gaseous
phase [13]. This suggests that the application potential of photo-
catalysis is in the air purification domain. The real scientific interest
in photocatalysis in gaseous phase started in 1990s [14–18] and the
number of publications has grown sharply since 1995 [13].
This paper deals with some aspects of photocatalytic degrada-
tion of several model VOCs (n-hexane [19–22], n-butyl acetate [23]
and toluene [5,7,10,18,20,24–32]) in a closed-loop reactor. These
pollutants were chosen for their different chemical structure, which
results in different chemical properties, e.g. water solubility; these
are also abundantly used as industrial solvents. For instance, they
occur in automotive industry, namely in the process of plastics
treatment, where aromates, alkenes, alkyl esters or some other
hydrocarbons coexist. Their photocatalytic mineralization can be
summarized by overall chemical reactions (1)–(3).
its ambient conditions; its initial relative humidity was measured
ranging from 40 to 50%. The liquid pollutant is injected through the
feed septum into the reservoir. A period of several minutes pre-
cedes the start of degradation experiment in order to evaporate
the pollutant and to get a homogeneous mixture.
Concentration of individual pollutants and possible intermedi-
ates was determined by gas chromatograph Varian CP-3800 with
capillary column CP-Sil 5 CB 15 × 0.25 (0.25) and flame ioniza-
tion detector. Mobile infrared analyser (Aseko, Czech Republic)
is equipped with infrared gas cell and is calibrated for on-line
measurement of CO2 concentration, which enables the evalua-
tion of overall mineralization rate. Intermediates adsorbed on
the surface of photocatalyst were analysed directly with FT-IR
spectrometer Nicolet iS10. Gas chromatography (column DB-5MS
60 m × 0.32 mm × 1 ␮m) with mass detection was used for analysis
of desorbed species; this was done by headspace method, i.e. the
catalyst was heated for 15 min at 150 ◦C and the desorbed gases
were analysed.
TiO ,hꢀ
2
n-Hexane C6H14
+
19 O2 −→ 6CO2 + 7H2O
(1)
(2)
(3)
2
TiO ,hꢀ
2
n-Butyl acetate C6H12O2 + 8O2 −→ 6CO2 + 6H2O
TiO ,hꢀ
2
Toluene C7H8 + 9O2 −→ 7CO2 + 4H2O
2.2. Particulate layer preparation
The mentioned compounds were studied separately but also
at operational conditions similar to industrial exhaust fumes as
degradation in a mixture and subsequent pollutant injection.
Particulate layers were prepared from TiO2 water suspension;
the suspension was placed on the substrate and was left to dry
out, then was thermally treated at 300 ◦C. Glass sheets of dimen-
sions 15 cm × 18 cm were used as a support. Amount of TiO2 was
0.5 mg cm−2, used TiO2 photocatalyst VLP7000 (Kronos). This pho-
tocatalyst was selected on the base of a previous study comparing
photoactivity of four different TiO2 photocatalysts in gaseous phase
(L.A. García, Diploma thesis: Photocatalytic degradation of pollut-
ants in liquid and gaseous phase, ICT Prague, 2010). For instance, we
give a comparison to standard P25 material in Fig. 2, from which is
seen that the degradation is faster on VLP7000 and this is confirmed
by faster generation of CO2. Some characteristics for both materi-
als are listed in Table 1, from which it can be seen that VLP7000
2. Experimental
2.1. Apparatus
The apparatus has recirculation loop geometry. It is composed
of diaphragm pump AL-6SB (Alita, Slovak Republic), photoreac-
tor and reservoir (in direction of circulation). Infrared analyser
and gas chromatograph are connected in parallel. Tubing is made
of polyamide tubes and push-in fittings. Apparatus volume is
2.85 dm3. The recirculation rate is held constant at value about
0.2 Nm3/h. The scheme of the entire apparatus is shown in Fig. 1.
The photoreactor is designed as a plate reactor, i.e. that the air
flows over the 15 by 18 cm photoactive layer in a thin film. The air
enters the reactor from jets in its shorter side and flows through
by an average velocity of ca. 5 cm s−1. Three 11 W lamps Sylvania
BLB with emission maximum at 367 nm were used as external UV
Table 1
Material characteristics of photocatalysts.
Material
(producer)
Specific surface
Crystal size
(nm)
Crystalline
structure
(m2 g−1
)
sources; the average light intensity was 2 mW cm−2
.
VLP7000 (Kronos)
P25 (Degussa)
267
56
7
Anatase
Anatase 71%
Rutile 29%
Anatase 22
Rutile 41
The reaction mixture is initially composed only of air and the
model pollutant(s). The air is taken directly from laboratory at

F. Moulis, J. Kry´sa / Catalysis Today 209 (2013) 153–158
155
Photocatalytic experiments in the gaseous phase are performed
rather in continuous reactors [5,6,9,10,14,15,17–25,28,30–34] than
in closed-loop (batch) reactors [16,26,27,29,35,36]. These two con-
figurations have some important differences from the kinetic point
of view. Firstly, the intermediates have not such importance in
continuous systems than in the batch ones, where they can easily
accumulate and thus make a kinetic study more difficult. Secondly,
the determination of degradation rate seems to be easier in the con-
tinuous systems where a conversion can be used almost directly.
On the other hand, this cannot be done for batch systems, where
time dependence must be taken into account and analysed; an ini-
tial degradation rate must be defined somehow or parameters of
an appropriate kinetic model must be determined.
In most cases the pollutant photocatalytic destruction on the
TiO2 is assumed to follow single-site Langmuir–Hinshelwood (L–H)
kinetics [34]. The reaction rate r(0) is defined as (in mol s−1):
Fig. 2. Hexane degradation on VLP7000 compared to P25 at initial concentration
830 ppm; data normalized by the maximal concentration.
KLc(0)
1 + KLc(0)
r(0) = kLH
(4)
where kLH is the apparent L–H rate constant in mol s−1, KL the
Langmuir adsorption constant in m3 mol−1 and c(0) is the initial
concentration in mol m−3. Mathematic fit using the form of Eq. (4)
was done for C5–C7 hydrocarbons degradation in gaseous phase
[19]. In that case instantaneous concentrations in a continuous
(flow-through) system were used and L–H kinetics was justified
by low initial concentrations (40–130 ppm) and thus negligible
by-products concentrations. L–H model was used also for ethanol
degradation and its intermediates (acetaldehyde, formaldehyde)
and transient model was established [35]. Determination of L–H
parameters was calculated from initial rates and initial concen-
trations issued from several experiments. As the experiment was
carried out in a closed recirculation loop, the complete min-
eralization was followed as well.For general photodegradation
experiment in a closed-loop system, the single-site L–H model [Eq.
(4)] is valid only for the initial concentration, thus only for the
initial reaction rate when the by-products are produced in a neg-
ligible amount. For this purpose, we define initial degradation rate
Ri (mmol m−2 h−1) by Eq. (5):
Fig. 3. Hexane degradation at different initial concentrations; final volume concen-
trations of CO2 are indicated at the end of each curve.
has several times higher specific surface and in contrast to P25 is
composed of pure anatase.
ꢀ
ꢁ
3. Results and discussion
n0 dcrel
Ri =
(5)
A
dt
i
3.1. Influence of the initial concentration
where n0 is the initial molar amount of pollutant, A the irradiated
area, crel the relative pollutant concentration [actual concentration
c(t) divided by the initial one c(0), i.e. crel = c(t)/c(0)], t the time,
(dcrel/dt)i is the initial slope of concentration dependence on irra-
diation time. As a first step we need to fit the experimental data by
suitable function and then calculate initial slope of this function.
Fig. 4a–c shows some possibilities how to fit experimental data
from a degradation experiment in closed-loop photoreactor. Expo-
nential fit (assumption of first order kinetics) shows deviations for
short and also long irradiation times. Linear fit applied on several
first points is represented as well; however, the problem is how
The influence of initial hexane concentration was studied at
values of 70, 830, 4150 and 8300 ppm (2.9, 34.3, 171.4 and
342.7 mmol m−3 at 22 ◦C). The progress of all four experiments is
shown in Fig. 3 in relative concentrations of hexane. The lower the
initial concentration, the lower the time necessary for complete
pollutant degradation. Relative concentrations of CO2 (related to
the final concentrations) are also shown in Fig. 3, only for the lowest
initial concentration of hexane, signal of CO2 was not detectable. It
can be seen that for an order higher initial concentration of hexane
(8300 ppm) we reach an order higher CO2 concentration (3.28%).
Fig. 4. Exponential, linear and polynomial fit of a general degradation experiment; points used for the fit are represented by solid circles.

156
F. Moulis, J. Kry´sa / Catalysis Today 209 (2013) 153–158
Fig. 5. Dependence of initial degradation rate of hexane on initial concentration, fit
done by Eq. (6); initial mineralization rate is also indicated.
Fig. 6. Consecutive degradations of model pollutants at initial concentration
8300 ppm.
many points are taken into account. As seen from Fig. 4c, a poly-
nomial fit of 2nd order represents the concentration decay very
accurately. To calculate the initial slope (dcrel/dt)i we have chosen
a sufficiently low fixed time t = 9 min, so thus degradation rate Ri
defined by Eq. (5) can be considered as the initial one.
Initial degradation rates calculated from the experimental data
(Fig. 3) are shown in Fig. 5 as a dependence on the initial concen-
tration; the data are fitted by Eq. (6):
already reported decrease of KL issued from a degradation exper-
iment with increasing light intensity [33]. Our experiments thus
show that crucial parameters influencing the value of adsorption
constant KL are light wavelength and experimental configurations
(continuous versus closed-loop) arrangement.
3.2. Comparison of the model pollutants
0.0228c(0)
1 + 0.0228c(0)
Consecutive photocatalytic degradations of three air pollutants
are compared in Fig. 6 in the order hexane, butyl acetate and
toluene. The degradation kinetics of butyl acetate is similar to the
hexane one but degradation rate of toluene is considerably lower.
Furthermore, the photocatalyst changed the colour from white
towards yellow brown after toluene degradation experiment.
This deactivation was already reported for toluene concentrations
higher than 13 ppm [20,24,26–28,30]. Stable activity of TiO2 pho-
tocatalyst was observed at low toluene concentrations (35 ppb)
[7] or with special polycrystalline TiO2 (Merck) [31]. Using GC-MS
we detected adsorbed species on the photocatalyst, e.g. benzoic
acid, benzaldehyde and probably methylphenyl benzoate. Adsorp-
tion of benzaldehyde and benzoic acid on TiO2 was reported to be
irreversible [32]. Recently the presence of benzoate complexes on
TiO2 was confirmed by FT-IR [26]. In our case we observed bands
at 1604, 1450 and 1415 cm−1 in the FT-IR spectrum which corre-
Ri = 25.4
(6)
which compared to Eq. (4) would result in L–H parameters
kLH = 25 mmol m−2 h−1 and KL = 0.023 m3 mmol−1. As the kinetic
constant kLH is related to the immobilized catalyst surface, no
further normalization is needed; this would be necessary e.g. for
aqueous suspension of TiO2 [37]. A significant change of reaction
order follows from Fig. 5. We get first-order kinetics at low concen-
trations and zero-order kinetics at high concentrations. This change
of order was already reported in gaseous phase for heptane [36],
butyric acid [33] or some other alkanes [19].
The mineralization rate is calculated from the initial slope of CO2
generation in Fig. 3, Eq. (1), known apparatus volume and equation
of state for ideal gas. The results are represented in Fig. 5 and are in
good agreement with the initial degradation rates. Carbon balance
results in degree of mineralization around 0.7 for all three initial
concentrations where CO2 was measured.
spond to those in benzoate spectrum (1604, 1454 and 1419 cm−1
)
[26]. The only intermediate in the gaseous phase detectable by GC-
FID was benzaldehyde in very low concentrations hundred times
lower than toluene. It can be thus concluded that the irreversible
adsorption of benzoic acid and benzaldehyde is responsible for
photocatalyst deactivation.
Toluene degradation influenced negatively consecutive degra-
dations carried out on the same layer. This is shown in Fig. 6,
which compares hexane degradation ability of VLP7000 material
in the form of fresh photocatalyst, toluene deactivated one and
regenerated one. Although the rates of butyl acetate and hexane
degradation are very similar (see Fig. 6), the deactivation impact of
toluene on the consecutive degradations is stronger in hexane case
than in butyl acetate case. Without any treatment photocatalyst
remains deactivated but after its regeneration by thermal treat-
ment (for 2 h at 300 ◦C) the degradation ability was completely
renewed.
Our parameters from Eq. (6) differ significantly from those
reported previously for hexane degradation in continuous system
(kLH = 0.78 mmol m−2 h−1 and KL = 13 m3 mmol−1) [19]. Further-
more even in very reduced range of concentrations (up to
4 mmol m−3) a zero-order region was reached already above con-
centration 1 mmol m−3 [19], which at our experimental conditions
did not occur bellow 170 mmol m−3 (see Fig. 5). The main rea-
son for the deviation could be different experimental conditions.
Boulamanti [19] used continuous system, different UV light wave-
length (254 nm) and different photocatalyst with higher coverage
(P25, 3.5 mg cm−2), UV light intensity was not reported. Zhang
[22] investigated hexane degradation in a continuous system with
a 254 nm UV lamp and got similar value of adsorption constant
(KL = 4.4 m3 mmol−1) as Boulamanti [19].
We found adsorption constant KL on TiO2 for similar compound
– n-heptane – as 0.31 m3 mmol−1 issued from pure adsorption
isotherm measurement [29] or as 0.030 m3 mmol−1 issued from
degradation experiments in batch reactor with UV light wavelength
365 nm [36]. Our value (0.023 m3 mmol−1) for hexane is in very
good agreement with that for heptane (0.030 m3 mmol−1) [36].
It is most probably due to the similar experimental conditions:
closed-loop photoreactor, wavelength (365 nm) and light intensity
(∼5 mW cm−2). The smaller value of our KL can be explained by
3.3. Degradation of mixture of pollutants
In real life, industrial exhausts contain more than one pollutant.
Thus we studied how a mixture of model pollutants is degraded,
especially considering the different elimination of constituent pol-
lutants. The initial concentration was set to 4150 ppm for both

F. Moulis, J. Kry´sa / Catalysis Today 209 (2013) 153–158
157
Fig. 7. Hexane degradation (8300 ppm) on the same layer before and after toluene
Fig. 9. Selective degradation of progressively injected pollutants: hexane (1 mmol,
degradation (8300 ppm).
8500 ppm), butyl acetate (0.38 mmol, 3200 ppm) and toluene (0.1 mmol, 860 ppm).
pollutants in the binary mixture of hexane and butyl acetate (Fig. 8)
in order to balance the photocatalyst utilization in the same man-
ner as for the individual pollutants (8300 ppm; Fig. 6). There is an
evident selectivity towards butyl acetate, which is degraded much
faster as seen from Fig. 7. It takes approximately 2 h to eliminate
the significant content of butyl acetate and to start the degradation
of hexane in a more apparent way. The possible reason is the fifty
times higher water solubility of butyl acetate (0.7% at 20 ◦C) than
hexane (0.014% at 20 ◦C). Due to the presence of water microfilm on
the photocatalyst surface [29], butyl acetate is easily adsorbed than
hexane and thus easily degradable. As can be seen from the CO2 con-
centration curve, the slope is fairly constant (it means very similar
mineralization rate of both pollutants) and decreases significantly
only when both pollutants are oxidized and thus not present in
the system (3.75 h). The slope changes only slightly when all butyl
acetate is degraded from the system (at about 2.5 h). This increase is
consistent with higher mass of carbon in unit mass of hexane (com-
paring to butyl acetate) and with almost the same degradation rates
of hexane and butyl acetate (Fig. 6).
Another possibility how to compare individual pollutants degra-
dation in mixture is their progressive injection. This is represented
in Fig. 9. Degradation of hexane (1 mmol, 8500 ppm) was inter-
rupted by injecting butyl acetate (0.38 mmol, 3200 ppm). According
to the supposition that butyl acetate is degraded preferentially (as
in Fig. 6) the degradation of hexane was significantly slowed down
in favour of the butyl acetate one. Nevertheless, CO2 concentration
grows nearly linearly with time and shows that the mineraliza-
tion of both pollutants is uniform. However, toluene injection
(0.1 mmol, 860 ppm) causes slowing down of butyl acetate degra-
dation and also overall mineralization. Toluene itself is oxidized
quickly immediately after injection but after about 1 h its degra-
dation rate significantly decreases. Overall mineralization rate
gradually decreases and after 3 h it is negligible as confirmed by
the CO2 concentration staying constant.
4. Conclusion
We have shown that hexane and butyl acetate can be photocat-
alytically degraded from air. Butyl acetate is oxidized preferably to
hexane when they react in a mixture, while their individual degra-
dation seems to be kinetically similar. Solubility of butyl acetate
in water is about an order higher that of hexane, which suggests
that the oxidation reaction proceeds in the aqueous microfilm
present on the surface of the TiO2 photocatalyst. Dependence of ini-
tial degradation rates on initial concentration confirms single-site
Langmuir–Hinshelwood kinetics with value of hexane adsorption
constant 0.023 m3 mmol−1
.
Oxidation products of toluene deactivate the photocatalyst and
its degradation is thus very slow at our operational conditions (TiO2
photocatalyst VLP7000 and initial concentration 8300 ppm). We
detected benzaldehyde, benzoic acid and methylphenyl benzoate
as intermediates. The deactivation is fully reversible and the cata-
lyst can be renewed by thermal treatment. The found selectivity or
inhibitive ability of studied organic pollutants could play an impor-
tant role in the purification of industrial exhaust fumes containing
always more than one pollutant, e.g. can even block elimination of
some pollutants if the preferentially degraded ones are present in
an important concentration.
Acknowledgements
Authors acknowledge the financial support from the Ministry
of Industry and Trade of the Czech Republic (project FR-TI3/242)
and from specific university research (MSMT no. 21/2012). Authors
would like also to thank Tomásˇ Florisˇ for his substantial contribu-
tion to the experimental part of the work.
References
[1] J.J. Shah, H.B. Singh, Environmental Science and Technology 22 (1988)
1381–1388.
[2] E.N. Ruddy, L.A. Carroll, Chemical Engineering Progress 89 (1993) 28–35.
[3] M.J. Ruhl, Chemical Engineering Progress 89 (1993) 37–41.
[4] B. Guieysse, C. Hort, V. Platel, R. Munoz, M. Ondarts, S. Revah, Biotechnology
Advances 26 (2008) 398–410.
Fig. 8. Degradation of binary mixture of hexane and butyl acetate at initial concen-
trations 4150 ppm.

158
F. Moulis, J. Kry´sa / Catalysis Today 209 (2013) 153–158
[5] P. Pichat, J. Disdier, C. Hoang-Van, D. Mas, G. Goutailler, C. Gaysse, Catalysis
Today 63 (2000) 363–369.
[22] P.Y. Zhang, J. Liu, Journal of Photochemistry and Photobiology A 167 (2004)
87–94.
[6] W.K. Jo, K.H. Park, Chemosphere 57 (2004) 555–565.
[7] C.H. Ao, S.C. Lee, C.L. Mak, L.Y. Chan, Applied Catalysis B: Environmental 42
(2003) 119–129.
[23] V. Keller, P. Bernhardt, F. Garin, Journal of Catalysis 215 (2003) 129–138.
[24] H. Einaga, S. Futamura, T. Ibusuki, Applied Catalysis B: Environmental 38 (2002)
215–225.
[8] C.H. Ao, S.C. Lee, Journal of Photochemistry and Photobiology A 161 (2004)
131–140.
[25] A. Bouzaza, C. Vallet, A. Laplanche, Journal of Photochemistry and Photobiology
A: Chemistry 177 (2006) 212–217.
[9] R. Portela, S. Suarez, R.F. Tessinari, M.D. Hernandez-Alonso, M.C. Canela, B.
Sanchez, Applied Catalysis B: Environmental 105 (2011) 95–102.
[10] A.J. Maira, J.M. Coronado, V. Augugliaro, K.L. Yeung, J.C. Conesa, J. Soria, Journal
of Catalysis 202 (2001) 413–420.
[11] A. Fujishima, X.T. Zhang, D.A. Tryk, Surface Science Reports 63 (2008) 515–582.
[12] I.S. McLintoc, M. Ritchie, Transactions of the Faraday Society 61 (1965)
1007–1016.
[13] Y. Paz, Applied Catalysis B: Environmental 99 (2010) 448–460.
[14] L.A. Dibble, G.B. Raupp, Environmental Science and Technology 26 (1992)
492–495.
[26] M.D. Hernandez-Alonso, I. Tejedor-Tejedor, J.M. Coronado, M.A. Anderson,
Applied Catalysis B: Environmental 101 (2011) 283–293.
[27] S.B. Kim, H.T. Hwang, S.C. Hong, Chemosphere 48 (2002) 437–444.
[28] Y. Luo, D.F. Ollis, Journal of Catalysis 163 (1996) 1–11.
[29] A. Maudhuit, C. Raillard, V. Hequet, L. Le Coq, J. Sablayrolles, L. Molins, Chemical
Engineering Journal 170 (2011) 464–470.
[30] R. Mendez-Roman, N. Cardona-Martinez, Catalysis Today 40 (1998) 353–365.
[31] G. Marci, M. Addamo, V. Augugliaro, S. Coluccia, E. Garcia-Lopez, V. Loddo, G.
Martra, L. Palmisano, M. Schiavello, Journal of Photochemistry and Photobio-
logy A 160 (2003) 105–114.
[15] J. Peral, D.F. Ollis, Journal of Catalysis 136 (1992) 554–565.
[16] L.A. Phillips, G.B. Raupp, Journal of Molecular Catalysis 77 (1992) 297–311.
[17] G.B. Raupp, C.T. Junio, Applied Surface Science 72 (1993) 321–327.
[18] T.N. Obee, R.T. Brown, Environmental Science and Technology 29 (1995)
1223–1231.
[19] A.K. Boulamanti, C.J. Philippopoulos, Atmospheric Environment 43 (2009)
3168–3174.
[20] O. d’Hennezel, D.F. Ollis, Journal of Catalysis 167 (1997) 118–126.
[21] C.T. Brigden, S. Poulston, M.V. Twigg, A.P. Walker, A.J.J. Wilkins, Applied Catal-
ysis B: Environmental 32 (2001) 63–71.
[32] L.X. Cao, Z. Gao, S.L. Suib, Abstracts of Papers of the American Chemical Society
220 (2000) 269.
[33] S. Brosillon, L. Lhomme, C. Vallet, A. Bouzaza, D. Wolbert, Applied Catalysis B:
Environmental 78 (2008) 232–241.
[34] L.A. Dibble, G.B. Raupp, Catalysis Letters 4 (1990) 345–354.
[35] M.L. Sauer, D.F. Ollis, Journal of Catalysis 158 (1996) 570–582.
[36] J. Shang, Y.G. Du, Z.L. Xu, Chemosphere 46 (2002) 93–99.
[37] V. Augugliaro, A. Bianco Prevot, J. Cáceres Vázquez, E. García-López, A. Irico, V.
Loddo, S. Malato Rodríguez, G. Marcí, L. Palmisano, E. Pramauro, Advances in
Environmental Research 8 (2004) 329–335.