Journal of Photochemistry and Photobiology C: Photochemistry Reviews

6 (2005) 264–273

Review

A review of synergistic effect of photocatalysis

and ozonation on wastewater treatment
T.E. Agustina ∗ , H.M. Ang, V.K. Vareek
Department of Chemical Engineering, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia
Received 18 August 2005; received in revised form 9 November 2005; accepted 28 December 2005

Abstract
For the treatment of wastewater that contain recalcitrant organic compounds, such as organo-halogens, organic pesticides, surfactants, and
colouring matters, wastewater engineers are now required to develop advanced treatment processes. A promising way to perform the mineralization
of this type of substance is the application of an advanced oxidation process (AOP).
Photocatalytic oxidation and ozonation appear to be the most popular treatment technologies compared with other advanced oxidation processes
(AOPs) as shown by the large amount of information available in the literature. The principal mechanism of AOPs function is the generation of
highly reactive free radicals. Consequently, combination of two or more AOPs expectedly enhances free radical generation, which eventually leads
to higher oxidation rates.
The use of combine photocatalysis and ozonation is an attractive route because of the enhancement of the performance for both agents by means
of the hydroxyl radical generation, a powerful oxidant agent that can oxidize completely the organic matter present in the aqueous system. The
scope of this paper is to review recently published work in the field of integrated photocatalysis and ozonation on wastewater treatment.
In this review the chemical effects of various variables on the rate of degradation of different pollutants are discussed. The mechanism and
kinetics has also been reported.
It can be concluded that photocatalytic oxidation in the presence of ozone is a process that is qualitatively and quantitatively different from the
well-known photocatalytic oxidation with oxygen and the ozonation without photocatalyst. The reason for the higher oxidation rate is probably a
photocatalytic induced decay of ozone, initiated by the combination of titanium dioxide and UV-A radiation.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Synergistic effect; Photocatalysis ozonation; Wastewater treatment

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
2. Photocatalysis and ozonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
3. Mechanism and kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
4. Intermediates detected . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
5. Effect of ozone dosage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
6. Effect of pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
7. Effect of crystal composition of the photocatalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8. Economics aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
9. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272

∗ Corresponding author. Tel.: +61 8 9266 3874; fax: +61 8 9266 2681.
E-mail address: agustint@mail.cage.curtin.edu.au (T.E. Agustina).

1389-5567/$20.00 © 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.jphotochemrev.2005.12.003
 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273 265

T.E. Agustina was born in Indonesia, in 1972. She

received a bachelor degree in chemical engineering
from the Department of Chemical Engineering at Sri-
wijaya University (South Sumatera, Indonesia) in 1996
and a master degree in chemical engineering from Gad-
jah Mada University (Yogyakarta, Indonesia) in 2001.
She joined the Faculty of Sriwijaya University as a
Lecturer in 2000. She is currently studying her PhD
at Curtin University of Technology, Western Australia,
under the direction of Ass. Prof. Ming Ang. During
this time she has been supported by a Technological
and Professional Skills Development Sector Project (TPSDP) scholarship. Her
research interest is in the area of photocatalytic on the wastewater treatment.

H.M. Ang is an Associate Professor in Chemical Engi- Fig. 1. Degradation of various organic compound (textile effluent in 1 h, aniline
neering and the Deputy Head of Department at Curtin in 2 h and 2,4-dichlorophenoxyaceticacid in 6 h) using photocatalysis, ozonation,
University of Technology. He is a fellow of both Engi- and combined photocatalysis ozonation ([15,9,25]).
neers Australia as well as the Institution of Chemical
Engineers, United Kingdom. Dr. Ang is also on the
Board of the Institution of Chemical Engineers in Aus-
of aniline from water and found that the decomposition rate of
tralia. He has supervised numerous PhD and Master
students over the years and his research areas are in aniline is larger than when individually treated by either of the
industrial crystallisation and environmental engineer- two methods. Wang et al. in 2002 studied the decomposition of
ing with special emphasis on wastewater treatment and formic acid in an aqueous solution using photocatalysis, ozona-
contaminated soil remediation. tion, and their combination. As a result, the decomposition rate
of formic acid by the combination of photocatalysis and ozona-
V.K. Vareek is currently a Senior Lecturer in the tion was found to be higher than the sum of the decomposition
Department of Chemical Engineering at Curtin Uni-
versity of Technology, Perth, Western Australia. Dr.
rates when formic acid was individually decomposed by a com-
Pareek completed his PhD in the area of photocatalysis bination of the two methods, which indicates the presence of a
from University of New South Wales. He is a Chemical synergistic effect of photocatalysis and ozonation [10].
Engineer and has received a master’s degree from IIT The scope of this paper is to review recently published work in
Delhi and a BE (Hons) from the University of Rajasthan the field of integrated photocatalysis and ozonation on wastewa-
(India). His expertise is in the area of computational
fluid dynamics (CFD) and reactor modelling. He is a
ter treatment. The summary of compounds degraded by various
member of IChemE. researchers using photocatalytic ozonation including informa-
tion about the UV source used, intermediates detected and results
is presented in Table 1. All results using different organic efflu-
1. Introduction ents always showed that the application of photocatalytic ozona-
tion gave the minimum total degradation time and produced
For the treatment of wastewater that contain trace recal- the highest removal of the organic effluent (see Figs. 1 and 2).
citrant organic compounds, such as organo-halogens, organic In particular, for the textile effluent, the decolourization was
pesticides, surfactants, and colouring matters, environmental almost complete with the combined method. The advantages of
engineers are now required to develop more advanced treatment photocatalysis giving decrease in TOC, and the advantages of
processes. In general, a combination of several methods gives ozonation giving no increase of high intermediate concentra-
high treatment efficiency compared with individual treatment. tions, were combined here.
For example, a certain organic compound can hardly be degraded
by ozonation or photolysis alone and the treated wastewater
may be more dangerous as a result of ozonation [1–4] but a
combination of several treatment methods, such as O3 /VUV
(ozone/vacuum ultraviolet), O3 /H2 O2 /UV (ozone/hydrogen per-
oxide/ultraviolet), and UV/H2 O2 (ultraviolet/hydrogen perox-
ide), improves the removal of pollutants from the wastewater
[5–7].
On the other hand, photocatalysis has a great potential for the
removal of organic pollutants from wastewater [8] although it is
still not in practical use because of its low oxidation rate. There-
fore, a combination of photocatalysis together with ozone, which
is a strong oxidizer, is reasonable for the treatment of difficult
to degrade organic compounds since the organic compounds are
then expected to decompose more quickly and thoroughly in the
presence of ozone, to carbon dioxide, water, or inorganic ions. Fig. 2. Degradation of nitrogen containing organic compounds by photocataly-
Sanchez et al. [9] combined these two methods for the removal sis, ozonation and combined photocatalysis and ozonation [30].
266 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273

Table 1
Summary of compounds degraded by various researchers using photocatalytic ozonation including information about the UV used, intermediates detected and results
and comments
Compound Source of light Intermediates detected Results and comments Parameter monitored References
degraded/source of
effluent
Acid compounds
Acetic acid Six 6 W medium pressure Formic acid, The acetic acic degradation rate DOC and acetic acid [58]
Hg vapor lamp formaldehyde was seven times higher in the degradation
combined systems than that using
photocatalysis alone
Formic acid 6 W black light blue Not analysed The kinetic constant for the Decomposition of [10]
fluorescent lamp combination methods is more formic acid
than the sum of both the kinetics
constans for photocatalysis and
ozonation taken separately
Formic acid Six 6 W medium pressure None The degradation rate of formic DOC and formic acid [58]
Hg vapor lamp acid using combined method was degradation
nearly five times higher than
those using photocatalysis alone
Propionic acid Six 6 W medium pressure Acetic acid, formic acid, The degradation rate of propionic DOC and propionic [58]
Hg vapor lamp acetaldehyde, acid using photocatalysis acid degradation
formaldehyde ozonation system was almost 10
times higher than that in the
photocatalysis system
Monochloroacetic acid UV lamp Not analysed When using the combined Degradation of [32]
and pyridine method, the degradation rate of monochloroacetic
monochloroacetic acid and acid and pyridine
pyridine led to 4 and 24 times
higher degradation rate,
respectively
2,4- 20 W black light 2,4-Dichlorophenol Combined method for 4 h TOC removal [44]
Dichlorophenoxyacetic chlorohydroquinone resulted in 2/3 of TOC removal
acid 1,2,4-trichlorobenzene leading to about the same level
for photocatalysis alone after 20 h
Sulfosalicylic acid 14 W 254 nm lamp Not analysed The degradation efficiency at all COD removal rate [51]
(waters) pH values has the same order:
O3 /V − O/TiO2 > O3 /TiO2 /UV
> O3 /UV
Humic acids 125 W black white Not analysed The degradation of humic acids Organic matter [38]
fluorescent lamp in a sequential system including reduction and colour
preozonation and photocatalysis reduction
were found to be efficient in
organic matter and colour
removal
Nitrogen compounds
Aniline 125 W medium pressure Not analysed The degradation rate of aniline by TOC removal [9]
Hg lamp the combination of photocatalysis
and ozonation is larger than when
individually treated
Free cyanide ions 6 W black light lamp Cyanate, nitrate and The degradation time of cyanide Degradation of [37]
carbonate ions is almost complete in 90, 60, and cyanide
18 min run time when using
photocatalysis, ozonation, and
photocatalysis ozonation,
respectively
Nitrogen containing 450 W xenon short arc Ammonia, nitrite nitrate, A considerable increase in the TOC removal [57]
organic compound lamp ethylamine degradation efficiency of the
N-compounds is reached by a
combination of photocatalysis
and ozonation
Alcohols
Phenol, p-chlorophenol 700 W high pressure Hg Hydroquinone, Photoctalytic ozonation led to COD and TOC [40]
(CP), p-nitrophenol lamp resorcinol, catechol, lower degradation times for COD removal degradation
(NP) oxalic acid fumaric acid, and TOC removal. The of catechol
glyoxylic acid, phenol percentage of mineralization was
nearly 100, 90 and 75% in
phenol, CP and NP
 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273 267

Table 1 (Continued )
Compound Source of light Intermediates detected Results and comments Parameter monitored References
degraded/source of
effluent

Catechol 15 W low pressure UV Not analysed The combined photocatalysis TOC removal [43]
lamp ozonation process (TiO2 /UV/O3 )
considerably increased TOC
removal rate compared to
TiO2 /UV, O3 , and O3 /UV process
Other organic compounds
Textile effluent 125 W high pressure Hg Not analysed The decolorization was almost TOC removal [29]
lamp complete and the highest TOC decolorization
reduction was achieved when
using photocatalysis ozonation
mehods
Pesticides 6 W black light lamp Not analysed Photocatalytic ozonation process TOC reduction [36]
led to rapid decrease of the
concentration of the
biorecalcitrant pesticides and to a
strong TOC reduction
4-Chlorobenzaldehyde 10 W low pressure Hg Not analysed The highest efficiency for the Degradation of [41]
lamp substrate degradation was substrate
obtained by simultaneous
ozonation and photocatalysis
Acetic acid, 6 W low pressure Hg lamp Formic acid, acetic acic, The most rapid degradation of the Degradation of acetic [60]
monochloro acetic acid glycoxylic acid, glycolic intermediate compounds by the acid monochloro
and dimethyl-2,2,2- acid combined method explains the acetic acid, DEP
trichloro-1- greatest TOC elimination rate
hydroxyethylphospate produced by this combination
(DEP)
Dibutyl phthalate 15 W low pressure UV Not analysed The mineralization rate constant TOC removal [42]
lamp of dibutyl phthalate with
TiO2 /UV/O3 is 1.2–1.8 times
higher than that of O3 /UV with
the same ozone concentration
Oxalate ion 700 W medium pressure Not analysed The highest decrease of oxalate Degradation of [39]
Hg lamp concentration takes place in the oxalate concentration
photocatalysis ozonation system
Post harvest fungal 6 W UV lamp Not analysed TiO2 /UV/O3 process Fungicide [61]
spoilage of kiwi fruit synergystically degraded organic decomposition
compounds and inhibited
conidial germination when
compare to a single treatment of
O3 or TiO2 /UV
Organic pollutants in 15 W low pressure UV Not analysed TiO2 /UV/O3 -BAC (biological DOC removal PAEs [62]
raw water (phthalate lamp activated carbon) was more and POPs removal
esters (PAEs) and efficient than UV/O3 -BAC in
persistent organic DOC removal and its synergetic
pollutants (POPs)) effect occurred more significantly

2. Photocatalysis and ozonation
According to their recent paper, Serpone and Emiline [11]
defined that photocatalysis, in its most simplistic description,
is the acceleration of a photoreaction by action of a catalyst.
An earlier IUPAC document defined it as a catalytic reaction
involving light absorption by a catalyst or a substrate.
Photocatalytic degradation has proven to be a promising
technology for degrading various organic compounds includ-
ing refractory chlorinated aromatics [12–16] and more than
1700 references have been collected on this discipline [17].
Compared with other conventional chemical oxidation methods,
photocatalysis may be more effective because semiconductors
are inexpensive and capable of mineralizing various refractory
compounds [18].
Due to its capability of generating OH radicals, which is a
powerful oxidant species, photocatalysis can be considered as
an advanced oxidation process (AOP). However, photochem-
ical AOPs are light induced reactions, mainly oxidations that
rely on the generation of hydroxyl radicals by the combination
with added oxidants or semiconductors. The degradation effi-
ciency of photochemical AOPs is greatly enhanced using either
homogeneous or heterogeneous photocatalysis [19]. Heteroge-
neous processes employ semiconductor slurries (e.g. TiO2 /UV,
ZnO/UV) for catalysis, whereas homogeneous photochemistry
(e.g. H2 O2 /UV, Fe3+ /UV) is used in a single-phase system.
268 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
Fig. 3. Conduction and valence bands and electron–hole pair generation in semi-
conductor.
Photocatalytic reactions occur when charge separation are
induced in large bandgap semiconductor by excitation with ultra
bandgap radiations [20]. In this way, the absorption of light
by the photocatalyst greater than its bandgap energy excites
an electron from the valence band of the irradiated particle to
its conduction band, producing a positively charged hole in the
valence band and an electron in the conduction band [21] as
shown in Fig. 3. The hole in the valence band may react with
water absorbed at the surface to form hydroxyl radicals and
on the other hand, the conduction band electron can reduce
absorbed oxygen to form peroxide radical anion that can fur-
ther disproportionate to form OH• through various pathways
[22]. During the photocatalytic process, other oxygen contain-
ing radicals are also formed including superoxide radical anion
and the hydroperoxide radical [23]. In addition, the band elec-
tron may also react directly with the contaminant via reductive
processes [24]. It is not clear under which experimental condi-
tions, one reaction pathway is more important than the other.
It is generally accepted that the substrate adsorption in the sur-
face of semiconductors plays an important role in photocatalytic
oxidation [25].
Titanium dioxide (TiO2 ) is a semiconductor that absorbs light
at λ < 385 nm with the corresponding promotion of an electron
from the valence band to the conduction band. This excitation
process leaves behind a positively charged vacancy called a hole
(h+ ). The hole by itself is either a very powerful oxidizing agent
or it may generate hydroxyl radicals in presence of water and
molecular oxygen [26,27]. Unlike other semiconductors, TiO2
photocatalysts can function using sunlight as the energy source.
Titanium dioxide is extensively used as photocatalyst due to
its high chemical stability, optical and electronic properties
[28].
Photocatalysis is a potential new method capable of elimi-
nating relatively recalcitrant organic compounds [26,29]. This
methodology is based on the production of electron–hole pairs
by illumination with light of suitable energy, of a semicon-
ductor powder such as TiO2 dispersed in an aqueous medium.
These charges migrate to the particle surface and react with
adsorbed species of suitable redox potential. In aerated media,
adsorbed molecular oxygen accepts photogenerated electrons,
while water molecules can react with photogenerated holes to
produce hydroxyl radicals.
However, although photocatalysis has shown to be adequate
for the destruction of a wide variety of compounds, in some
cases the complete mineralization is slowly attained and the
efficiency of the processes, in terms of energy consumption,
Fig. 4. Laboratory-scale set-up for photocatalysis ozonation.
is only advantageous for very dilute effluents [26]. To overcome
this difficulty some additives such as H2 O2 , Fe2+ , Fe3+ , S2 O8 2− ,
Ag+ , etc., with different chemical roles have been added to the
photocatalytic systems [9].
Furthermore, ozone can also be used for degradation of
organic pollutants in water. Ozone is a powerful oxidant (electro-
chemical oxidation potential of 2.07 V versus 2.8 V for hydroxyl
radical), which is generally produced by an electric discharge
method in the presence of air or oxygen. Two reactions of ozone
with dissolved organic substance can be distinguished in water:
a highly selective attack of molecular ozone takes place on the
organic molecules at low pH, whereas free radicals from ozone
decomposition can also react non-selectively with the organic
compounds [30]. Additional free hydroxyl radicals can be pro-
duced in the aqueous media from ozone by pH modification
or can be introduced by combining ozone either with hydrogen
peroxide or UV-irradiation from a high pressure mercury lamp
[31].
On the other hand, the use of ozone for the destruction of
organics in water is also a well-known water treatment tech-
nique. Unlike photocatalysis, ozonation, due to its capability
for selective destruction of recalcitrant organics, is used as a
pre-treatment step before ordinary biological techniques, thus
being more efficient for highly contaminated wastewater. The
simultaneous application of ozonation and photocatalysis has
the capability for efficient treatment of organically contaminated
waters over a wide range of concentrations [9]. A schematic dia-
gram of a laboratory-scale set-up for photocatalysis ozonation
is illustrated in Fig. 4.
3. Mechanism and kinetics
Photocatalytic oxidation and ozonation appear to be the
most promising pre-treatment technologies compared with other
advanced oxidation processes (AOPs) as shown by the large
amount of information available in the literature. The principal
mechanism of AOPs function is the generation of highly reactive
free radicals.
 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
Consequently, a combination of two or more AOPs enhances
free radical generation, which eventually leads to higher oxida-
tion rates.
Sanchez et al. [9] in their investigation, found that although
the strategy of ozonation pre-treatment followed by photo-
catalysis would be a satisfactory route for aniline degradation,
the simultaneous ozonation and photocatalysis methodology
resulted in much higher total organic carbon (TOC) removal
which, in spite of more energy and material demanding, could
be preferred from an application point of view (see Table 1).
In fact, besides the direct ozonation of the intermediate com-
pounds, in the presence of TiO2 under illumination, ozone can
generate OH• radicals through the formation of an ozonide rad-
ical (O3 •− ) in the adsorption layer:
TiO2 + hν → e− + h+
O3 + e− → O3 •−
The generated O3 •− species rapidly reacts with H+ in the
solution to give HO3 • radical, which evolves to give O2 and
OH• as shown in Eqs. (3) and (4) below:
O3 •− + H+ → HO3 • (3)
•
HO3 → O2 + OH•
In photocatalysis, the hydroxyl radical is generally consid-
ered to be mainly responsible for the organic attack. It must be
considered that the OH• radical can react with O3
OH• + O3 → O2 + HO2 •
in competition with aniline and the corresponding organic inter-
mediates. Further, the O2 •− species participates in a closed loop
reaction scheme, which leads to a continuous consumption of
ozone. In the absence of O3 , dissolved O2 itself can accept TiO2
conduction band electron and generate O2 •−
− •−
O2 + e → O 2
Which in turn can be protonated to form HO2 •:
O2 •− + H+ → HO2 •
In contrast with HO3 • (see Eq. (3)), this species cannot give
OH• radicals in a single step and an alternative reaction pathway
has been proposed to account for OH radical’s generation from
HO2 • :
2HO2 • → H2 O2 + O2
H2 O2 + O2 •− → OH• + HO + O2 −
This mechanism requires a total of three electrons for the
generation of a single OH• species, which is a less favoured
situation if compared with the one electron needed through the
O3 •− reaction pathway [9].
To get more information about the photocatalytic ozonation
processes, Kopf et al. [32] measured the ozone decomposition
by illumination in aqueous TiO2 suspension in the absence of
organic compounds. The solutions (pH 3) were saturated with
ozone. After stopping the gas supply illumination was started.
269
It is obvious that the decomposition is accelerated by the appli-
cation of titanium dioxide and appropriate light. The reaction
depends on light intensity. If the light intensity is decreased,
the ozone decomposition is also decreased. These results show
that a real photocatalytic reaction occurs in the suspension, not
the radiation alone but the combination of radiation and pho-
tocatalyst initiates a decomposition reaction of ozone. Another
investigation by Kopf et al. [32] indicated that oxygen should
have an influence on the photocatalytic ozonation. Besides other
reaction paths, such as direct ozone attack (Eq. (29)), or direct
electron transfer from TiO2 to the ozone molecule (Eqs. (1)–(4)),
a further possible reaction path is proposed [32].
Charge separation:
TiO2 + hν → h+ + e− (10)
(1)
Charge transfer of positive charge:
(2)
h+ + H2 O → OH• + H+ (11)
Electron transfer and further reactions:
e− + O2 → O2 •− (12)
O2 •− + H+ ↔ HO2 • (13)
(4)
O3 + O2 •− → O3 •− + O2 (14)
O3 •− + H+ ↔ HO3 • (15)
HO3 • → OH• + O2 (16)
(5)
Oxidation of the organic compound R-H:
OH• + R-H → R• + H2 O (17)
or
OH• + R → R• -OH (18)
(6)
In reactions (13)–(16) the ozone decomposition according to
the mechanism of Weiss [33] with the extension of Buhler et al.
is described [34].
(7) The difference between photocatalytic ozone decomposition
and ozone decomposition in aqueous solution is the initiation
of the reaction. The starting radical is formed photochemically
by an electron transfer from titanium dioxide to oxygen and
not, like in the Weiss-mechanism, by the reaction of OH− ion
with ozone. In both cases primarily O2 •− is formed. O2 •− reacts
(8) with ozone forming the ozonide ion (14). After protonation OH-
(9) radicals are formed ((15) and (16)). The formation of OH radicals
in acidic solutions via direct electron transfer from the catalyst
to the ozone molecule or via O2 •− explains the degradation
of organic compounds by photocatalytic ozonation which react
very slowly with the ozone molecule or with HO2 • , one of the
primary oxidising species in photocatalytic processes [32].
Furthermore, Wang et al. [10] indicated in their investiga-
tion that a synergistic effect occurred when the photocatalytic
ozonation treatments are carried out simultaneously. Two possi-
ble mechanisms are considered for the synergistic effect. Firstly,
it is considered that in the photocatalytic reactor, the dissolved
270 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
ozone was easy to get electrons produced on the surface of tita-
nium dioxide according to the mechanism:
O3(ads) + e− → O3(ads) −
The recombination of electrons and positive holes could be
interferred by the reaction between ozone and electrons on the
surface of titanium dioxide. Consequently, a larger number of
radicals are produced, thereby accelerating the photocatalytic
reaction. Secondly, it is considered that a larger number of
hydrogen peroxide and hydroxyl radicals were produced from
the dissolved ozone as a result of UV-irradiation [35] In the
ozonation under UV-irradiation, the hydroxyl radical is pro-
duced according to the following mechanism [10]:
hν
O3 + H2 O−→H2 O2 + O2
hν
H2 O2 −→2OH•
H2 O2 ↔ HO2 − + H+
−
O3 + HO2 → O3 •− + HO2 •
HO2 • → O2 •− + H+
O3 + O2 •− → O3 •− + O2
•− + H+ ↔ • formed can be classified into three categories. These categories
O3 HO3
Recently, Farre et al. [36] suggested that for the photocatalytic
ozonation system, the reaction also proceeded by radical attack
to organic molecule. However, the OH radical is produced by (i)
reaction of absorbed H2 O molecule with photogenerated holes
at the illuminated TiO2 particle:
TiO2 + hν → h+ + e−
+ +
H2 O + h → OH• +H
and (ii) by reaction of adsorbed O3 and photogenerated electrons
at the TiO2 particles, namely reaction (1)–(4) above by Sanchez
et al. [9] The presence of dissolved ozone in the irradiated TiO2
aqueous suspension increases the OH radical production and
decreases the electron–hole recombination, increasing the effi-
ciency of the photocatalytic process.
To date, a kinetic model based on Langmuir–Hinshelwood
(L–H) equations adequately describes the reactivity results and
provides the values of kinetic constant and equilibrium adsorp-
tion constant in the degradation of organic compound using
photocatalyc ozonation methods [9,10,37–39].
Farre et al. found that the degradation processes follow zero
order kinetics when photocatalytic ozonation is applied [36].
On the other hand, Beltran et al. [40] discovered that phenol
removal obeyed first order kinetics. Other researchers, Balcioglu
et al. [41] reported in their investigation which also employed the
combined method, the decomposition modes follow pseudo-first
order kinetics. And Li et al. [42] described in their experimen-
tation on photocatalysis ozonation process, the mineralization
of dibuthyl phthalate follows pseudo-zero-order kinetics depen-
dence on ozone concentration. The photocatalytic ozonation of
catechol also followed pseudo-zero-order-kinetics, as reported
by Li et al. [43].
4. Intermediates detected
Intermediates are formed during the process of degradation.
(19)
Table 1 shows that most of the researchers did not analyse for
intermediates and focused mainly on the degradation rate of
organics. Some of them analysed TOC with only a few men-
tioning the intermediates formed.
Regarding the intermediates, Muller et al. [44] got quite
different results in the photocatalysis, ozonolysis and the
combination of both methods in the treatment of 2,4-
dichlorophenoxyacetic acid (2,4-D). They could identify 2,4-
dichlorophenol (2,4-DCP), chlorohydroquinone (CHQ) and
1,2,4-trichlorobenzene (THB). While CHQ and THB are minor
intermediates, 2-4-DCP is the main and long living intermediate
(20) in all three methods. In photocatalysis, the 2,4-DCP disappeared
after a total running time of about 40 h. In ozonolysis, 2,4-DCP
(21) formed only in negligible amounts. In the combination approach,
(22) the maximum concentration of 2,4-DCP was only about 10%
of the maximal level in photocatalysis and it vanished within
(23) 10–12 h, which is in about the same amount of time with the
degradation of 2,4-dichlorophenoxyacetic acids.
(24)
Beltran et al. [40] reported the intermediates found in
(25) photocatalytic ozonation of phenols. Intermediate compounds
(26)
are polyphenols (mainly hidroquinone), unsaturated carboxylic
acids (mainly fumaric acid) and saturated carboxylic acids
(mainly oxalic acid). Oxalic acid was the main end product
that remained in water, especially during the oxidation of p-
nitrophenol. Also, nitrogen as nitrate from p-nitrophenol and
chloride from p-chlorophenol were detected in solution.
Other researchers did not analyse the intermediates without
(27) giving any particular reasons. However, the researchers used the
(28) advantage of ozonolysis, which results in no build up of high
intermediates concentration.
5. Effect of ozone dosage
Li et al. [42] accounted the effect of ozone dosage in their
research on photocatalytic ozonation of dibutyl phthalate over
TiO2 film. The addition of ozone considerably improved TOC
removal compared to photocatalysis and ozonolysis near UV
(UV/O3 ). After 30 min, TOC removal rate was over 75% when
ozone dosage was over 25 mg/h, while only 73% using UV/O3
when ozone dosage was 50 mg/h. These results showed the
synergistic effect of UV/O3 , TiO2 /UV and TiO2 /O3 [45–47].
When the semiconductor is in contact with water, the surface
of TiO2 is readily hydroxylated. Under near-UV illumination,
electron–hole pairs are formed in the semiconductor, the oxida-
tion potential of hydroxylated TiO2 must lie above the position
of the semiconductor valence band, and the oxidation of surface-
bound OH− and H2 O by TiO2 valence band holes to form OH is
thermodynamically possible and expected [48]. Thus the min-
eralization of dibuthyl phthalate was accelerated.
Effect of ozone dosage was also examined by Li et al. [48]
in their study on removing organic pollutants in secondary
effluents using photocatalysis ozonation and subsequent biolog-
ical activated carbon. They found that DOC removal efficiency
 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
increased with ozone dosage. The total DOC removal efficiency
increased from about 52 to 59% in the process when ozone
dosage increased from 3 to 9 mg/L. However, DOC removed
per unit ozone consumption decreased with ozone dosage. It
was 2.11 mg DOC/mg O3 at 3 mg/L ozone dosage, while it
decreased to 0.95 mg DOC/mg O3 when ozone dosage increased
to 9 mg/L. At three ozone dosages (3–9 mg/L), the amount of
ozone consumption is 2.38, 4.39, and 6.01 mg/L, respectively.
And it indicated that low ozone dosage was much more econom-
ically efficient.
Kerc et al. [38] investigated the effect of ozone dosage of
between 1.47 and 7.35 mg/L on the degradation of humic acids
by photocatalysis and ozonation. They found that for higher
ozone application rates the maximum reduction achieved was
80% for organic matter degradation and 90% for colour reduc-
tion. A slight decrease in pH was observed with increasing ozone
dosage rates. The pH decrease may be attributed to the forma-
tion of carboxylic groups with more acidic characteristic. Li et
al. [43] also studied the effect of ozone dosage on photocat-
alytic ozonation of catechol. The concentration of ozone ranged
between 0 and 50 mg/h. The increase of TOC removal by apply-
ing TiO2 /O3 /UV was evident compared with UV/O3 . These
results showed the synergistic effect of the UV/O3 , TiO2 /UV
and TiO2 /O3 [45–47]. The TOC removal rate was over 94.5%
after 30 min when the flowrate of ozone was over 37.5 mg/h.
6. Effect of pH
Ozone has many desirable oxidizing properties for use in
wastewater treatment. Unlike other oxidants, there are two
mechanisms through which ozone can degrade organic pollu-
tants, namely (i) direct electrophilic attack (Eq. (29)) and (ii)
indirect attack through the formation of hydroxyl radicals (Eqs.
(30)–(31)) [30,49]. Direct attack by molecular ozone (commonly
known as ozonolysis) occurs at acidic or neutral conditions and
is a selective reaction resulting in the formation of carboxylic
acids as end products that cannot be oxidized further by molec-
ular ozone:
O3 + R → Rox
where R is the organic solutes and Rox is the oxidised organic
products. Compounds susceptible to ozonolysis are those con-
taining C C double bonds, specific functional groups (e.g. OH,
CH3 , OCH3 ) and atoms carrying negative charge (N, P, O, S) [49]
At high pH value, ozone decomposes to non-selective hydroxyl
radicals, which, in turn, attack the organic pollutants:
O3 + OH− → OH• + (O2 • ↔ HO2 • )
OH• + R• → R ox
where R ox is the oxidised organic products for radical-type
reactions.
Therefore, the pH of the effluent is a major factor determin-
ing the efficiency of ozonation since it can alter degradation
pathways as well as kinetics. Ozonation is usually coupled with
another oxidant such as hydrogen peroxide or UV-irradiation to
271
enhance the formation of hydroxyl radicals in aqueous phase
[50].
Many researches [51,52] found that increasing pH will
accelerate ozone decomposition to generate OH radical, which
destroys the organic compounds more effectively than ozone
does.
For the photocatalytic process, generally pH has a varied
effect on the process and it is expected that increase or decrease in
the pH should affect the rate of degradation as adsorption varies
with pH. In the case of substances which are weakly acidic, the
photocatalytic degradation increases at lower pH because of an
increase in the adsorption. Some substances undergo hydrolysis
at alkaline pH, which is one of the reasons for the increase in the
photocatalytic degradation at alkaline pH values [53] At alka-
line pH values, the concentration of OH• radicals is relatively
higher and this may also be another reason for the increase in
the photocatalytic degradation rate. It can be concluded that the
effect of pH is negligible so pH adjustment is not required in the
photocatalytic degradation.
For the photocatalysis and ozonolysis, Muller et al. [44] found
that neither the preference for low pH values of the photocatal-
ysis nor that for high pH values for the ozonolysis could be
better than the degradation rate obtained by simultaneous appli-
cation of both methods at neutral pH values. The kinetic results
showed that degradation at pH 7 using the combination approach
is more than 1.5 times faster than the best result of ozonolysis at
pH 11, and more than three times faster than the fastest degra-
dation applying photocatalysis using the unbuffered set-up (pH
3.4–3.5).
Tong et al. [51] reported that the pH values for the degrada-
tion of sulfosalicylic acid solution by photocatalysis ozonation
decreased firstly within a period of time and increased subse-
quently with the mineralization of intermediates. They found
that the pH value in the system increased an initial pH of 2.2–2.8
for ozonation of 45 min.
7. Effect of crystal composition of the photocatalyst
Taking into account the impact of different possible photo-
(29) catalyst on the: (1) chemical activity, (2) stability under different
operating conditions, (3) availability and handiness in many
physical forms, (4) cost, and (5) lack of toxicity, the most widely
used photocatalyst is titanium dioxide [54].
In this review, TiO2 photocatalyst is used in all experiments.
TiO2 has three crystal structures, namely rutile, anatase and
brukite. Only rutile and anatase are stable enough and can be
used as a photocatalyst. It was shown that photocatalytical prop-
erties of different TiO2 materials might differ considerably, with
(30)
rutiles exhibiting the lowest photoactivity [55]. Anatase has bet-
(31) ter activity than rutile TiO2 . The combination of anatase and
rutile decrease the recombination of photogenerated electrons
and holes and hence it shows more activity than anatase, as in
the case of Degussa P-25.
Thevenet et al. compared the efficiency of P-25 Degussa
and PC500-Millennium titanium dioxide photocatalysts. P-25
Degussa is made of two allotropic phases of the titanium diox-
ide; 80% of it is crystallized into the anatase phase and 20%
272 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
into the rutile phase. Its specific area is 50 m2 g−1 with average
particle size of 30 nm. The PC500-Millennium photocatalyst is
made of only one crystallographic phase, anatase, with a specific
area of 300 m2 g−1 which is much higher than P25 but with lower
particle size of 5–10 nm. The mixed anatase/rutile P-25 Degussa
is reported as the most efficient in their experimental condition.
Crystallographic phase contact between anatase and rutile in
P-25 Degussa, as well as the crystallite size, was suggested to
explain its efficiency [56]. Therefore Degussa P-25 is the most
suitable photocatalyst. All the investigations reported here used
Degussa P-25 TiO2 for their study [9,29,32,36–40,57,58].
The photocatalyst derives its activity from the fact that when
photons of certain wavelength strike its surface, electrons are
promoted from the valence band and transferred to the conduc-
tance band. This leaves positive holes in the valence band, which
react with the hydroxylated surface to produce OH• radicals. In
Degussa P-25 the conduction band electron of the anatase part
jumps to the less positive rutile part, reducing the rate of recom-
bination of electrons and positive holes in the anatase part [59].
Hernandez-Alonso et al. [37] tested two photocatalysts with
known activity for cyanide photodegradation using photocatal-
ysis ozonation method. Two commercial TiO2 samples were
used, that is Degussa P-25 (805 anatase, 20% rutile) and BDH
TiO2 (100% anatase). The synergetic effect of ozone and TiO2
under irradiation is present in both TiO2 samples; however, it
is more pertinent with the BDH specimen than with the P-25.
The enhancement of reactivity in the presence of ozone with
respect to oxygen is related to the ability of ozone of trap-
ping electrons and, therefore, it also depends on the availability
of electrons in the photocatalyst. The used photocatalyst dif-
fer in their crystal composition: BDH specimen is pure anatase
while P-25 is mixture of anatase and rutile. The higher BDH
photoactivity is probably related to the lower band gap energy
of the anatase sample that facilitates the electron trapping by
ozone.
8. Economics aspects
The economic aspects of the different methods must be con-
sidered in light of the energy consumption for all of them.
Photocatalytic ozonation needs additional electrical energy for
the ozone generator. Therefore energy consumption of the three
processes was compared according to consumption of electrical
power (kW h) of the laboratory device under the experimental
conditions.
Kopf et al. [32] calculated the energy consumption for the
UV-lamp and ozone generator in their investigation on the photo-
catalytic ozonation of monochloroacetic acid and pyridine. They
found that the energy consumption of the UV lamp and ozone
generator was 0.046 and 0.018 kW h in 60 min, respectively.
The energy consumption for DOC (dissolved organic carbon)
decrease was compared for three methods, i.e. TiO2 /O2 (pho-
tocatalysis), H2 O/O3 (ozonation) and TiO2 /O3 (photocatalytic
ozonation) and is shown in Table 2. The photocatalytic ozona-
tion method resulted in much lower specific energy consumption
than the other two alternatives.
Table 2
Photocatalytic ozonation of monochloroacetic acid; specific energy consump-
tion (kW h/g DOC-reduction)
TiO2 /O2 H2 O/O3 TiO2 /O3
Monochloroacetic acid 19 110 5.4
Pyridine 120 73 11
Reprinted from Ref. [32] with permission from Elsevier.
9. Conclusion
It was shown that the photocatalytic oxidation in the presence
of ozone is a process, which is qualitatively and quantitatively
different from either the well-known photocatalytic oxidation
with oxygen or ozonation without photocatalyst.
Five possible mechanisms are considered for the synergis-
tic effect: (i) direct electron transfer from TiO2 to the ozone
molecule which generate OH• radicals through the formation of
an ozonide radical (Eqs. (1)–(4)), (ii) electron transfer from TiO2
to the oxygen molecule to form O2 •− radical, then O2 •− radical
reacts with ozone to give OH• radicals through the formation
of an ozonide radical (Eqs. (10)–(16)), (iii) a recombination of
electrons and positive holes interfered by the reaction between
ozone and electrons on the surface of titanium oxide, so a large
number of radicals produced, thereby accelerating the photocat-
alytic reaction (Eq. (19)), (iv) production of hydrogen peroxide
and OH• radicals increased by UV-irradiation of ozone (Eqs.
(20)–(26)), and (v) OH• radicals produced by the reaction of
absorbed H2 O molecule with photogenerated holes at the illu-
minated TiO2 particle (Eqs. (27) and (28)).
The main problem of AOPs lies in the high cost of reagents
such as ozone, hydrogen peroxide or energy source such as ultra-
violet light. For the photocatalytic ozonation methods the energy
demand of the O3 /TiO2 /UV could be considerably decreased by
the use of solar-irradiation and in situ electrochemical O3 gen-
eration. However, the use of solar radiation as an energy source
can reduce costs.
The synergistic effect summarized in Table 1 showed the
advantages of the application of photocatalytic ozonation
method. Photocatalytic ozonation could be an effective alter-
native way for oxidation and elimination of recalcitrant organic
compounds under specific conditions for the treatment of indus-
trial wastewater.
References
[1] S.J. Kim, K.H. Oh, S.H. Lee, S.S. Choi, K.C. Lee, Water Sci. Technol.
36 (1997) 325.
[2] S. Echigo, H. Yamada, S. Matsui, S. Kawanishi, K. Shishida, Water Sci.
Technol. 34 (1996) 81.
[3] S.D. Lambert, N.J.D. Graham, Water Res. 29 (1995) 2427.
[4] V. Camel, A. Bermond, Water Res. 32 (1998) 3208–3222.
[5] J.Y. Hu, S. Wang, W.J. Ng, S.L. Ong, Water Res. 33 (1999) 2587–2592.
[6] J.D. Plummer, J.K. Edzwald, Water Sci. Technol. 37 (1998) 49.
[7] A. Hirvonen, T. Tuhkanen, P. Kalliokoski, Water Sci. Technol. 33 (1996)
67.
[8] R.P.S. Suri, J. Liu, D.W. Hand, J.C. Crittenden, D.L. Perram, M.E.
Mullins, Water Environ. Res. 65 (1993) 665.
 T.E. Agustina et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 6 (2005) 264–273
[9] L. Sanchez, J. Peral, X. Domenech, Appl. Catal. B: Environ. 19 (1998)
59–65.
[10] S. Wang, F. Shiraishi, K. Nakano, Chem. Eng. J. 87 (2002) 261–271.
[11] N. Serpone, A.V. Emeline, Int. J. Photoenergy 4 (2002) 91–131.
[12] P.C. Calza, C. Minero, E. Pelizetti, Environ. Sci. Technol. 31 (1997)
2198–2203.
[13] W. Chu, Environ. Sci. Technol. 33 (1999) 421–425.
[14] A.P. Davis, D.L. Green, Environ. Sci. Technol. 33 (1999) 609–617.
[15] D.C. Schmelling, K.A. Gray, P.V. Kamat, Water Res. 31 (1997)
1439–1447.
[16] A. Topalov, D. Molnar-Gabor, J. Csanadi, Water Res. 33 (1999)
1371–1376.
[17] D.M. Blake, Bibliography of work on the photocatalytic removal of haz-
ardous compound from water and air, in: NREL/TP-430-22197, National
Renewable Energy Laboratory Golden Co., 1999.
[18] Y. Ku, R.M. Leu, K.C. Lee, Water Res. 30 (1996) 2569–2578.
[19] O. Legrini, E. Oliveros, A.M. Braun, Chem. Rev. 93 (1993) 671–698.
[20] L. Rideh, A. Wehrer, D. Ronze, A. Zoulalian, Ind. Eng. Chem. Res. 36
(1997) 4712–4718.
[21] I. Izumi, W.W. Dunn, K.O. Wilbourn, F.F. Fan, A.J. Bard, J. Phys. Chem.
84 (1980) 3207–3210.
[22] H. Al-Ekabi, G. Edwards, W. Holden, A. Safarzadeh-Amiri, J. Story, in:
W.W. Eckenfelder, A.R. Bowers, J.A. Roth (Eds.), Chemical Oxidation,
Technomic Publishing Co., Lancaster, 1992, pp. 254–261.
[23] D.D. Dionysiou, A. Khodadoust, A.M. Kern, M.T. Suidan, I. Baudin,
J.M. Laine, Appl. Catal. B 24 (2000) 139–155.
[24] A.P. Jones, R.J. Watts, J. Environ. Eng. 123 (1997) 974–981.
[25] D. Duonghong, J. Ramsden, M. Gratzel, J. Am. Chem. Soc. 104 (1982)
2977–2985.
[26] M.R. Hoffman, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95
(1995) 69–75.
[27] M.R. Prairie, L.R. Evans, B.M. Stange, S.L. Martinez, Environ. Sci.
Technol. 27 (1993) 1776–1795.
[28] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735.
[29] S.G. Moraes, R.S. Freire, N. Duran, Chemosphere 40 (2000) 369–373.
[30] J. Hoigne, H. Bader, Water Res. 10 (1976) 377–386.
[31] S. Chiron, A. Fernandez-Alba, A. Rodriguez, E. Garcia-Calvo, Water
Res. 34 (2000) 366–377.
[32] P. Kopf, E. Gilbert, S.H. Eberle, J. Photochem. Photobiol. A 136 (2000)
163–168.
[33] J. Weiss, Trans. Faraday Soc. 31 (1935) 668.
[34] R.E. Buhler, J. Staehlin, J. Hoigne, J. Phys. Chem. 88 (1984) 2560.
[35] V. Beschkkov, G. Bardarska, H. Gulyas, I. Sekoulov, Water Sci. Technol.
36 (1997) 131.
[36] M.J. Farre, M.I. Franch, S. Malato, J. Ayllon, J. Peral, X. Domenech,
Chemosphere (2004).
273
[37] M.D. Hernandez-Alonso, J.M. Coronado, A.J. Maira, J. Soria, V. Loddo,
V. Augugliaro, Appl. Catal. B: Environ. 39 (2002) 257–267.
[38] A. Kerc, M. Bekbolet, A.M. Saatci, Ozone Sci. Eng. 25 (2003) 497–504.
[39] M. Addamo, V. Augugliaro, E.G. Lopez, V. Loddo, G. Marci, L.
Palmisano, Catal. Today 107/108 (2005) 612–618.
[40] F.J. Beltran, F.J. Rivas, O. Gimeno, J. Chem. Technol. Biotechnol. 80
(2005) 973–984.
[41] I.A. Balcioglu, G. Nicola, B. Miray, J. Photochem. Photobiol. A 135
(2000) 229–233.
[42] L. Li, W. Zhu, L. Chen, P. Zhang, Z. Chen, J. Photocem. Photobiol. A:
Chem. 175 (2005) 172–177.
[43] L. Li, W. Zhu, P. Zhang, L. Chen, W. Han, Water Res. 37 (2003)
3646–3651.
[44] T.S. Muller, Z. Sun, G.M.P. Kumar, K. Itoh, M. Murabayashi, Chemo-
sphere 36 (1998) 2043–2055.
[45] R. Gracia, S. Cortes, J. Sarasa, P. Ormad, J.L. Ovelleiro, Water Res.
(2000) 1525–1532.
[46] H. Allemane, B. Delouane, B. Legube, Ozone Sci. Eng. 15 (1999)
419–432.
[47] B. Ohtani, S.W. Zhang, S. Nishimoto, J. Chem. Soc. Faraday Trans. 88
(1992) 1049–1053.
[48] L. Li, P. Zhang, W. Zhu, W. Hand, Z. Zhang, J. Photocem. Photobiol.
A: Chem. 171 (2005) 145–151.
[49] A. Alvares, C. Diaper, S. Parsons, Environ. Technol. 22 (2001) 409–427.
[50] D. Mantzavinos, E. Psillakis, J. Chem. Technol. Biotechnol. 79 (2004)
431–454.
[51] S. Tong, D. Xie, H. Wei, W. Liu, Ozone Sci. Eng. 27 (2005) 233–238.
[52] B. Kasprzyk-Hordern, M. Ziolek, J. Nawrocki, Appl. Catal. B 46 (2003)
639–669.
[53] D. Bhatkande, V. Pangarkar, A. Beenackers, J. Chem. Technol. Biotech-
nol. 77 (2001) 102–116.
[54] A.E. Cassano, O.M. Alfano, Catal. Today 58 (2000) 167–197.
[55] M. Lindner, J. Theurich, D.W. Bahnemann, Water Sci. Technol. 35
(1997) 79–86.
[56] F. Thevenet, O. Guaitella, J.M. Hermann, A. Rousseau, C. Guillard,
Appl. Catal. B: Environ. 61 (2005) 58–68.
[57] M. Klare, G. Waldner, R. Bauer, H. Jacobs, J.A.C. Broekaert, Chemo-
sphere 38 (1999) 2013–2027.
[58] I. Ilisz, A. Bokros, A. Dombi, Ozone Sci. Eng. 26 (2004) 585–594.
[59] K.M. Schindler, M. Kunst, J. Phys. Chem. 94 (1990) 8222–8226.
[60] K. Tanaka, K. Abe, T. Hisanaga, J. Photocem. Photobiol. A: Chem. 101
(1996) 85–87.
[61] J. Hur, S. Oh, K. Lim, J.S. Jung, J. Kim, Y.J. Koh, Postharvest Biol.
Technol. 35 (2005) 109–113.
[62] L. Li, W. Zhu, P. Zhang, Q. Zhang, Z. Zhang, J. Hazard. Mater., in
press.