Influence of activated carbons on the kinetics and mechanisms of aromatic molecules ozonation

Article abstract of DOI:10.1016/j.cattod.2010.03.044

Companies have been looking for new methods for treating toxic or refractory wastewaters; which can mainly be used prior to or after or in connexion with biological treatment processes. This paper compares conventional ozone oxidation with activated carbon (AC) promoted ozone oxidation, which helps developing a mechanism involving HO{radical dot} radical. For a compound which is quite easy to oxidise, like 2,4-dichlorophenol (2,4-DCP) conventional ozonation is efficient enough to remove the initial molecule. The mechanism involved mainly consists of an electrophilic attack on the aromatic ring, which is activated by the donor effect of the -OH group, then followed by a 1,3 dipolar cycloaddition (Criegee mechanism) that leads to aliphatic species, mainly carboxylic acids. Yet, the addition of AC, through the presence of HO{radical dot} radical, enhances the removal of these species which are more refractory. For a refractory compound like nitrobenzene (NB), with a de-activated aromatic ring because of the attractive effect of -NO₂, conventional ozonation is inefficient. On the contrary, this molecule can be quite easily removed with AC promoted oxidation and it is found that the mechanism (electrophilic attack followed by a 1,3 dipolar cycloaddition) is quite similar to the one corresponding to conventional ozonation, but with less selectivity. For both molecules, a mass balance has established that the by-products accounting for more than 75% of the remaining COD can be quantified. A significant part is composed of carboxylic acids (acetic, oxalic, etc.), which could afterwards be easily removed in an industrial wastewater treatment process followed by a final biological treatment step.

Full text of DOI:10.1016/j.cattod.2010.03.044

Catalysis Today 151 (2010) 166–172
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Influence of activated carbons on the kinetics and mechanisms of aromatic
molecules ozonation
T. Merlea, J.S. Pica,∗, M.H. Manerob, S. Mathéa, H. Debellefontainea
a Université de Toulouse, LISBP, UMR INSA/CNRS 5504 & UMR INSA/INRA 792, 135 Avenue de Rangueil, 31077 Toulouse Cedex 4, France
b Université de Toulouse, LGC, UMR CNRS 5503, IUT GCGP, 137 Avenue de Rangueil, 31077 Toulouse Cedex 4, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 24 April 2010
Companies have been looking for new methods for treating toxic or refractory wastewaters; which can
mainly be used prior to or after or in connexion with biological treatment processes.
This paper compares conventional ozone oxidation with activated carbon (AC) promoted ozone oxi-
dation, which helps developing a mechanism involving HO• radical. For a compound which is quite easy
to oxidise, like 2,4-dichlorophenol (2,4-DCP) conventional ozonation is efficient enough to remove the
initial molecule. The mechanism involved mainly consists of an electrophilic attack on the aromatic ring,
which is activated by the donor effect of the –OH group, then followed by a 1,3 dipolar cycloaddition
(Criegee mechanism) that leads to aliphatic species, mainly carboxylic acids. Yet, the addition of AC,
through the presence of HO• radical, enhances the removal of these species which are more refractory.
For a refractory compound like nitrobenzene (NB), with a de-activated aromatic ring because of the
attractive effect of –NO2, conventional ozonation is inefficient. On the contrary, this molecule can be
quite easily removed with AC promoted oxidation and it is found that the mechanism (electrophilic
attack followed by a 1,3 dipolar cycloaddition) is quite similar to the one corresponding to conventional
ozonation, but with less selectivity.
Keywords:
Ozone
Activated carbon
2,4-Dichlorophenol
Nitrobenzene, Oxidation by-products
For both molecules, a mass balance has established that the by-products accounting for more than 75%
of the remaining COD can be quantified. A significant part is composed of carboxylic acids (acetic, oxalic,
etc.), which could afterwards be easily removed in an industrial wastewater treatment process followed
by a final biological treatment step.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
Ozone decomposition in water is a well-known phenomenon.
It can be described thanks to a cyclic mechanism involving several
Wastewater is often treated by biological routes with good
removal efficiencies; however, toxic organic compounds cannot be
reliably eliminated by these bioprocesses. Additional treatments,
such as adsorption or chemical oxidation, must be used for non-
biodegradable molecules removal. Ozone is frequently used since
it is a powerful oxidizing agent. The reactivity of the ozone molecule
with unsaturated organics follows two main mechanisms [1,2]:
electrophilic substitution and 1,3 dipolar cycloaddition, known as
the “Criegee mechanism”. In the particular case of aromatic com-
pounds, the substituents of the aromatic ring play a major role in the
actual mechanism. Yet all these reaction pathways are selective and
unfortunately some organic compounds cannot be oxidised with a
single ozonation step. In these cases, the use of advanced oxidation
processes can be the best solution [3–6].
radical species and acido–basic equilibria [7]. Hydroxide ions and
hydrogen peroxide strongly participate in the initiation steps of this
cycle (Eqs. (1) and (2)):
•
•
O3 + HO− → O2 − + HO2
(1)
(2)
•
•
O3 + HO2 → HO + O2 − + O2
−
The kinetics of the first reaction is quite slow (second order
constant k1 = 70 L mol−1 s−1) whereas the kinetic constant for the
second reaction is much larger (k2 = 2.8 × 106 L mol−1 s−1), thus
hydrogen peroxide strongly enhances radicals production from
ozone [8]. The kinetics of ozone decomposition into HO• radical
are generally well described with a simple first order law according
to dissolved ozone concentration [9]. The value of the correspond-
ing kinetic constant (kC) depends, among other parameters, on the
pH of the solution, on hydrogen peroxide concentration or on the
presence of catalytic materials.
∗
In previous papers, we showed that ozone decomposition could
be obtained through activated carbons (AC) [10–12] or zeolite
Corresponding author. Tel.: +33 561 559 747; fax: +33 561 559 760.
E-mail address: jean-stephane.pic@insa-toulouse.fr (J.S. Pic).
0920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.03.044

T. Merle et al. / Catalysis Today 151 (2010) 166–172
167
Fig. 1. Experimental set-up.
Table 1
[13,14]. In particular, some authors assume that the formation of
Characteristics of the activated carbons.
HO• radicals from activated carbons is due to their surface basic-
ities [15]. Indeed, the basic-oxygen-containing functional groups
and/or delocalised ␲ electrons of the graphene layers of AC can act
as Lewis bases and form electron donor–acceptor complexes with
H2O molecules (Eq. (3)). The hydroxide ion which is released will
then initiate ozone decomposition
S-23
F-22
Specific area (m2 g−1
)
1374
0.56
0.05
17.00
1.46
1005
0.38
0.10
9.00
1.00
Microporous volume (cm3 g−1
)
Mesoporous volume(cm3 g−1
Mean pore diameter (Å)
Granular diameter (mm)
)
C
C(graphene) + H3O+ ↔ C+–CH(graphene) + H2O
(3)
2H2O ↔ H3O+ + HO−
matography using a HP 1050 series chromatograph equipped with
a Diode Array Detector (Agilent Technologies HP 1200) and a C-18
column (EQUISORB ODS2 5 ␮m). Samples were first passed through
a 0.45 ␮m filter before analyses in order to remove undissolved
compounds. Table 2 lists the different chemicals which were mon-
itored and gives their maximum absorbance wavelenghts (ꢀ) and
kinetic diameters (ꢁ). The chemical oxygen demand (COD) of the
solutions was measured by Hanna Instrument test tubes (COD MR
Reagent).
The present work studies the influence of AC on the oxidation by
ozone of two bio-recalcitrant compounds: nitrobenzene and 2,4-
dichlorophenol. The process efficiency is expressed as molecule
removal rate, reaction mechanisms, by-products quantification and
mass balance calculations.
2. Materials and methods
Oxidation runs were performed in a 2 L well stirred semi-batch
reactor fed with ozonated oxygen (30 L h−1, 30 g m−3) through
a pipe placed below the stirrer. Ozone concentration was mea-
sured in the liquid phase (polarographic probe Orbisphere model
410/probe 31330.15xx) as well as in the gas inlet and outlet (UV
absorbance analyser Trailigaz UVOZON TLG 200). In most experi-
ments pH was not controlled.
Two microporous activated carbons (supplied by PICA Corp.,
France) called F-22 and S-23 were used with 1 g L−1 amounts.
Their specific areas and mean pore diameter are, respectively
1000 m2 g−1 – 0.9 nm and 1300 m2 g−1 – 1.7 nm.
3. Results and discussions
3.1. Ozone decomposition with or without activated carbon
The decomposition rate of ozone was studied according to
increasing concentrations of activated carbon F-22 and S-23. The
kinetic constant of ozone decomposition kC was calculated through
the mass balance for ozone dissolved in water (Eq. (4)). It was
compared to results without materials and plotted in Fig. 2. The
measurement method as well as the reference value obtained with-
out AC, 1.1 × 10 −3 s−1, are described by Lopez–Lopez et al. [16]
The experimental set up is shown in Fig. 1 and detailed physical
properties of the activated carbons are listed in Table 1.
QG([O3]G,in − [O]∞
)
G,out
kC
=
(4)
[O3]∞VL
Nitrobenzene and 2,4-dichlorophenol were chosen to prepare
synthetic polluted water. The initial concentrations of the two
pollutants were 250 mg L−1, i.e. 1879 ␮mol L−1 for nitrobenzene
and 1534 ␮mol L−1 for 2,4-dichlorophenol. The concentrations of
the pollutants and oxidation products (purchased from Acros
Chemicals) were determined by High Performance Liquid Chro-
L
with QG: volumetric gas flow rate (m3 s−1); [O3]G,in: ozone concen-
tration in inlet gas (g m−3); [O]G∞,out: ozone concentration in outlet
gas at steady state (g m−3); [O3]∞: ozone concentration in water at
L
steady state (g m−3); VL: liquid volume in the reactor (m3).

168
T. Merle et al. / Catalysis Today 151 (2010) 166–172
Table 2
Molecules characteristics of nitrobenzene, 2,4-dichlorophenol and ozonation by-products.
Molecule name
ꢀ (nm)
ꢁ (Å)
Stoichiometric ratio
COD/molecule (mg O2 mg−1
)
Nitrobenzene (NB)
267
274
274
317
5.9
8.1
8.1
6.7
1.951
1.612
1.612
1.612
o-Nitrophenol (o-NP)
m-Nitrophenol (m-NP)
p-Nitrophenol (p-NP)
2,4-Dichlorophenol (2,4-DCP)
3,5-Dichlorocatechol (3,5-DCC)
4,6-Dichlororesorcinol (4,6-DCR)
4-Chlorocatechol (4-CC)
2-Chlorohydroquinone (2-CHQ)
2-Chlorobenzoquinone (2-CBQ)
Catechol (C)
286
202
202
202
202
254
202
–
–
–
–
–
–
–
1.178
0.447
0.447
0.537
0.537
0.673
0.644
Oxalic acid (OA)
Acetic acid (AA)
202
202
–
–
0.177
1.066
S-23, which has a microporous structure, and as a consequence the
decomposition of ozone can be superior with F-22 than with S-23.
3.2. NB and 2,4-DCP ozonation
Fig. 4 shows NB and 2,4-DCP ozonation removal rate, with and
without AC. First, it can be observed that none of the two media had
any beneficial effect on 2,4-DCP oxidation rate. Indeed, 2,4-DCP is
totally removed after only 1 h single or combined oxidation pro-
cess. On the other hand, the presence of activated carbon allows
a lower ozone consumption: it is lowered by 31% with F-22 AC
and by 11% with S-23 AC. Concerning NB, the presence of activated
carbon slightly enhances the oxidation rate: a total removal occurs
after 3.5 h single ozonation whereas only 2.5 h treatment are neces-
sary with AC. The performances of the two media are quite similar,
ozone consumption is reduced by 4% with F-22 and by 12% with
S-23.
The difference of reactivity with ozone itself can be easily
explained by the molecules mesomery. Indeed ozone direct path-
ways, Criegee addition or electrophilic attack, are improved when
the aromatic ring is strongly activated. The –NO2 substituent of NB
is a deactivating group whereas the –OH substituent of 2,4-DCP is
an activating group. Also the –Cl group is a deactivating group since
ozone reacts faster with phenol than with 4-chlorophenol (4-CP).
However, Benitez et al. [19] showed that the more phenol is sub-
stituted with –Cl groups, the faster ozone oxidation occurs. Thus
2,3,4,6-tetrachlorophenol is more quickly oxidised than phenol. A
strong substitution on the aromatic ring probably changes the reac-
tion mechanism, for example nucleophilic substitution onto –Cl
groups could take place [15]. These statements explain the faster
ozonation rate for 2,4-DCP than that of NB.
Fig. 2. Ozone decomposition constant versus activated carbons concentration
(pH 7, [O3]G,in = 30 g N m3, 25 ◦C) S-23 (ꢀ) and F-22 (᭹).
The kC constant evolution is proportional to the material con-
centration, in accordance with the results of Guiza et al. [17]. With
10 gAC L−1, the decomposition rate is multiplied by 5 for S-23 and
by 7.5 for F-22.
This phenomenon can be explained either by the AC physical
structure and also by the AC chemical composition which both
influence the reactivity of the catalyst. The experimental measure-
ments of surface groups according to Boehm method [18] showed
that both AC have basic sites on their surfaces: 157 ␮equiv. g−1
for F-22 and 306 ␮equiv. g−1 for S-23. These sites can enhance the
decomposition of ozone by the reaction with chromene and pyrone
groups, as shown in Fig. 3 [2]. Then, hydrogen peroxide provided
by this reaction can react with remaining ozone to give hydroxyl
radicals as explained in Section 1.
Moreover, material porosity is an important factor for the dif-
fusion of ozone molecules to the active sites. Even if F-22 has a
lower external specific area than that of S-23, it has a greater meso-
porous volume, so the diffusion in F-22 is probably easier than in
In the presence of activated carbon, NB oxidation kinetics is
enhanced whereas when 2,4-DCP is used the removal is not any
faster. Single adsorption cannot explain this behaviour since the
Fig. 3. Basic group (chromene up and pyrone down) reaction with ozone, leading to hydrogen peroxide.

T. Merle et al. / Catalysis Today 151 (2010) 166–172
169
Fig. 4. NB and 2,4-DCP removal by single ozonation ( ) and combined ozonation with S-23 (ꢀ) and with F-22 (᭹).
adsorption kinetics is relatively slow for both molecules [12]. Thus
the difference in NB removal rate is clearly attributed to a cat-
alytic effect of activated carbon. Indeed, NB is rather refractory to
direct ozonation, so radical oxidation with CA being less selective,
clearly enhances the oxidation kinetics. For 2,4-DCP, direct ozona-
tion kinetics is faster and thus is not improved through the radical
mechanism. Finally, for both molecules, the radical route probably
leads to a better yield for the oxidizing agent resulting in a reduced
ozone global consumption.
matic ring is meta directing and deactivating for an electrophilic
attack. The simultaneous presence of o-NP and p-NP in the solution
can be attributed either to another mechanism or to a low selectiv-
ity in the position of the electrophilic attack. This can occur when
the stability of the o, m and p-transition states are very close [15].
However, after 1 h experiment, the amount of nitrophenols only
accounts for 10% of the oxidised NB. Thus the electrophilic attack
is not the only pathway in NB ozonation and the Criegee reaction
occurs too. Indeed carboxylic acids such as acetic (Fig. 5) or oxalic
acids are produced. Unfortunately the quantification of oxalic acid
was unsuccessful because of interferences with nitrates, also pro-
duced. Finally, a pathway for NB direct ozonation is suggested in
Fig. 6.
The addition of activated carbons slightly improved the removal
rates for the four by-products. Hence, the radical route increases
reaction rates without much changing the nature of the by-
products. The reason why radical efficiency is higher is probably
due to less selectivity.
A global mass balance was performed through two COD estima-
tions. The actual COD was measured, and the part of COD due to
NB was calculated from NB concentration. The difference between
3.2.1. Oxidation mechanisms of NB
Ortho-nitrophenol (o-NP), meta-nitrophenol (m-NP), para-
nitrophenol (p-NP) and acetic acid, which are four NB ozonation
by-products, were quantified using HPLC. o-NP and m-NP
were equally produced with maximum concentrations reach-
ing 14–22 ␮mol L−1 after 30 min experiment. The corresponding
curves are not showed here for lack of space. p-NP concentration
was twice higher with 40 ␮mol L−1 after 1 h experiment. At last
about 1330 ␮mol L−1 acetic acid were formed after 1.5 h experi-
ment (Fig. 5). The presence of meta-nitrophenol can be explained
by NB resonance. Indeed, the –NO2 substituent linked to the aro-
Fig. 5. NB ozonation by-products versus time: single ozonation ( ), ozonation with S-23 (ꢀ) and with F-22 (᭹).
Fig. 6. Suggested reaction pathway for nitrobenzene direct ozonation.

170
T. Merle et al. / Catalysis Today 151 (2010) 166–172
Fig. 7. Evolution of nitrobenzene by-products COD, estimated from (a) actual COD and NB concentration and (b) by-products concentration: single ozonation ( ), ozonation
with S-23 (ꢀ), ozonation with F-22 (᭹).
Fig. 8. 2,4-DCP oxidation by-products versus time: single ozonation ( ), ozonation with S-23 (ꢀ), with F-22 (᭹).
the two values stands for the actual by-products COD, plotted in
Fig. 7a [3]. Alongside this experiment, a theoretical COD value for
the by-products was calculated from the measured by-products
concentrations, as showed in Fig. 7b. The actual COD was higher
than the theoretical COD, which indicates that about 25% of the by-
products were not quantified. Again it should be pointed out that
pared to those in single ozonation. This also explains the gain in
global ozone consumption and clearly justifies the interest of a
radical route for NB oxidation.
3.2.2. Oxidation mechanisms of 2,4-DCP
Eight 2,4-DCP ozonation by-products, listed in Table 2, were
monitored. Fig. 8 shows the results for two of them: an intermedi-
ate aromatic species (3,5-DCC) and a final product (oxalic acid). The
aromatic intermediates reached their maximum concentrations
between 15 and 30 min experiment. The resultant values were:
3,5-DCC (56 ␮mol L−1), 4,6-DCR (2 ␮mol L−1), 4-CC (14 ␮mol L−1),
the monitoring of oxalic acid was unsuccessful.
−1
For a single ozonation, 130 mg
L
of by-products were
COD
detected after 4 h experiment whereas only 30 mg L−1 were
observed in the AC/ozone combined process. Thus, the addition of
AC greatly enhances the oxidation rates of the by-products com-
Fig. 9. Suggested reaction pathway for 2,4-dichlorophenol direct ozonation.

T. Merle et al. / Catalysis Today 151 (2010) 166–172
171
Fig. 10. 2,4-DCP by-products COD removal, estimated from (a) actual COD and 2,4-DCP concentration and (b) by-products concentration: single ozonation ( ), ozonation
with S-23 (ꢀ),with F-22 (᭹).
2-CHQ (76 ␮mol L−1) and catechol (118 ␮mol L−1). The first two
molecules are generated through the electrophilic attack of ozone
on 2,4-DCP. The amount of 3,5-DCC is higher because the –OH group
activates the aromatic ring in position 6 while the –Cl group has a
smaller influence on position 5. Then, the two mono-chlorinated
compounds are probably formed through the electrophilic substi-
tution (called the “ipso effect”) of ozone on the –Cl group position.
The difference in the concentration of these two species was
not explained. At last the presence of catechol cannot be eas-
ily explained, but however, it is certainly obtained after several
elementary steps. Finally, the aromatic molecules are opened fol-
lowing the Criegee reaction, and carboxylic acids are formed. Up to
1000 ␮mol L−1 acetic acid was generated in the first 30 min and was
subsequently totally removed after about 2 h. On the other hand,
oxalic acid was continuously formed during the 2 h and reached
889 ␮mol L−1 before slowly being removed from the solution until
the level of 222 ␮mol L−1 recorded after 8 h experiment. Finally a
pathway for 2,4-DCP ozonation is proposed in Fig. 9.
mechanisms of the various molecules. Classical mechanisms of
electrophilic substitution and dipolar cycloaddition explained most
of the sub-products formations. The same by-products have been
found with or without AC, but the reaction rates are enhanced by
the presence of activated carbon.
Finally, considering the ability of HO• radicals to remove an
extremely wide range of refractory organics, including trace pol-
lutants and emerging compounds, it appears that this AOP process,
coupling activated carbon and ozonation, could be a good candi-
date for an additional treatment after biological treatment, above
all for strongly refractory to ozone molecules oxidation.
Acknowledgement
The authors thank French National Agency (ANR) for financial
support – Program Precodd, project “PHARE”.
The addition of activated carbons had no noticeable effect on
the removal of 2,4-DCP and its by-products. Yet, there seems to be
a slight increase in oxalic acid production in the presence of S-23
media. This may be due to the production of dissolved organic
species (either oxalic acid or other ones such as formic, acetic
or glyoxylic acids) resulting from the oxidation of the AC itself
but this hypothesis still has not been verified. Nevertheless, this
difference is not really significant, as oxalic acid amounts to very
little in the global mass balance (Fig. 10). Whatever the oxidation
route, about 150 mg L−1 COD was generated after 1 h experiment,
followed by a slow decrease during the 7 remaining hours. Thus,
the radical route does not improve 2,4-DCP oxidation, the direct
ozonation kinetics is too fast to be enhanced by activated carbons,
at least for the AC concentrations (1 g L−1) used in this study.
References
[1] P.S. Bailey, Ozonation in Organic Chemistry, Academic Press, New York, 1978.
[2] P.M. Alvarez, J.F. Garcia-Araya, F.J. Beltran, I. Giraldez, J. Jaramillo, G. Serrano,
The influence of various factors on aqueous ozone decomposition by granular
activated carbons and the development of a mechanistic approach, Carbon 44
(2006) 3102–3112.
[3] R. Andreozzi, V. Caprio, A. Insola, R. Marotta, Advanced oxidation pro-
cesses (AOP) for water purification and recovery, Catal. Today 53 (1) (1999)
51–59.
[4] B. Legube, N. Karpel Vel Leitner, Catalytic ozonation: a promising advanced
oxidation technology for water treatment, Catal. Today 53 (1) (1999) 61–72.
[5] J. Lin, A. Kawai, T. Nakajima, Effective catalysts for decomposition of aqueous
ozone, Appl. Catal. B 39 (2) (2002) 157–165.
[6] F.J. Beltran, J.F. Garcia-Araya, I. Giraldez, Gallic acid water ozonation using acti-
vated carbon, Appl. Catal. B 63 (2005) 249–259.
[7] J. Staehelin, R.E. Bühler, J. Hoigné, Ozone decomposition in water studied by
•
pulse radiolysis. II. HO• and HO4 as chain intermediate, J. Phys. Chem. 88 (24)
(1984) 5999–6004.
[8] J. Staehelin, J. Hoigné, Decomposition of ozone in water: rate of inititation
by hydroxide ions and hydrogen peroxide, Environ. Sci. Technol. 16 (1982)
676–681.
4. Conclusions
[9] W. Stumm, Der zerfall von ozon in wässriger Lösung, Helv. Chim. Acta 37 (1954)
773–778.
[10] J. Reungoat, J.S. Pic, M.H. Manero, H. Debellefontaine, Adsorption of nitroben-
zene from water onto high silica zeolites and regeneration by ozone, Sep. Sci.
Technol. 42 (2008) 1447–1463.
[11] T. Merle, J.S. Pic, M.H. Manero, H. Debellefontaine, Enhanced bio-recalcitrant
organics removal by combined adsorption and ozonation, Water Sci. Technol.
60 (11) (2009) 2921–2928.
[12] T. Merle, J.S. Pic, M.H. Manero, H. Debellefontaine, Comparison of activated
carbon and hydrophobic zeolite efficiencies in 2,4-dichlorophenol advanced
ozonation, Ozone: Sci. Eng., in press.
[13] P. Monneyron, S. Mathé, M.H. Manero, J.N. Foussard, Regeneration of high sil-
ica zeolites via A.O.P.—a preliminary study about adsorbent reactivity toward
ozone, Chem. Eng. Res. Des. 81 (A9) (2003) 1193–1198.
[14] P. Monneyron, M.H. Manero, S. Mathe, A combined selective adsorption and
ozonation process for VOCs removal from air, Can. J. Chem. Eng. 85 (2007)
326–332.
[15] J. Rivera-Utrilla, M. Sánchez-Polo, Ozonation of 1,3,6-naphatalenetrisulfonic
acid catalysed by activated carbon in aqueous phase, Appl. Catal. B 39 (2002)
319–329.
The feasibility and interest of adding activated carbon to ozona-
tion process was studied in this paper, in order to develop a hybrid
process (combining at the same time and in the same device adsorp-
tion and oxidation) for refractory molecules treatment. The results
showed the ability of the activated carbons to decompose molec-
ular ozone and to create radical species, which were much more
efficient for oxidation as they are non-selective. For organic pollu-
tants which are easy to oxidise by ozone, like 2,4-dichlorophenol,
the removal rate was not significantly enhanced, but a smaller
ozone consumption was noticed. For organic pollutants much more
refractory to ozone oxidation, like nitrobenzene, the development
of a radical route dramatically increased the removal rate and the
ozonation kinetics became faster.
In addition, the analysis of by-products obtained during oxida-
tion enabled an understanding of the major steps in the oxidation

172
T. Merle et al. / Catalysis Today 151 (2010) 166–172
[16] A. Lopez-Lopez, J.S. Pic, H. Benbelkacem, H. Debellefontaine, Influence of t-
butanol and pH on hydrodynamic and mass transfer parameters in an ozonation
process, Chem. Eng. Process. 46 (2007) 649–655.
[17] M. Guiza, A. Ouedemi, A. Ratel, Decomposition of dissolved ozone in the pres-
ence of activated carbon: an experimental study, Ozone: Sci. Eng. 26 (3) (2004)
299–307.
[18] H.P. Boehm, Chemical identification of surface groups, Adv. Catal. 16 (1966)
179–274.
[19] F. Benitez, J. Beltrán-Heredia, J.L. Acero, F.J. Rubio, Rate constants for the reac-
tions of ozone with chlorophenols in aqueous solutions, J. Hazard. Mater. B79
(2000) 271–285.