ARTICLE IN PRESS

Available at www.sciencedirect.com

WAT E R R E S E A R C H 4 1 ( 2 0 0 7 ) 2 5 2 5 – 2 5 3 2

0043-1354/$ - see frodoi:10.1016/j.watres

�Corresponding auE-mail address:

journal homepage: www.elsevier.com/locate/watres

Bezafibrate removal by means of ozonation: Primaryintermediates, kinetics, and toxicity assessment

Renato F. Dantasa,�, Marisa Canterinob, Raffaele Marottab, Carme Sansa,Santiago Esplugasa, Roberto Andreozzib

aDepartament d’Enginyeria Quımica, Universitat de Barcelona, Martı i Franques 1, 08028 Barcelona, SpainbFacolta di Ingegneria, Dipartimento di Ingegneria Chimica, Universita degli Studi di Napoli ‘‘Federico II’’,

p.le V. Tecchio, 80-80125 Napoli, Italy

a r t i c l e i n f o

Article history:

Received 21 November 2006

Received in revised form

7 March 2007

Accepted 10 March 2007

Available online 27 April 2007

Keywords:

Bezafibrate

Pharmaceuticals

Ozonation

Biodegradability

Acute toxicity

Kinetic constants

nt matter & 2007 Elsevie.2007.03.011

thor. Tel.: +34 934021312;falcao@angel.qui.ub.es (R

a b s t r a c t

Bezafibrate (BZF) is a lipid regulator largely used for the treatment of hyperlipidaemia. As a

result of its wide use, unmetabolized BZF is released in the environment with potential

toxic effects for aquatic living organisms. The results obtained in this work show that

ozonation is an efficient method to degrade BZF: after 10 min of treatment (corresponding

to a dose of 0.73 mmol L�1 of ozone), the complete BZF abatement is achieved, starting from

an initial concentration of 0.5 mmol L�1. However, only a small part of the substrate is

mineralized.

Two different experimental approaches (absolute and competition method) are adopted

to estimate the second-order kinetic constants for the ozone attack at pH ¼ 6.0, 7.0 and 8.0.

A good agreement was observed between the two kinetic methods adopted.

The identification of main intermediates, attempted by high-performance liquid

chromatograph (HPLC)–MS technique, indicates that the oxidation of BZF develops through

both the hydroxylation of the aromatic ring and the attack of ozone on the unchlorinated

aromatic one.

The assessment of by-products biodegradability and acute toxicity demonstrates that

ozonation is a suitable technique to improve the biodegradability and reduce the toxicity of

waters containing BZF.

& 2007 Elsevier Ltd. All rights reserved.

1. Introduction

Nowadays, there is a great concern about the presence of new

(emerging) contaminants in the environment. Among these

compounds, the group of pharmaceuticals has received a

special attention in the last years and a large number of works

regarding their identification and treatment have been

published (Heberer, 2002; Stackelberg et al., 2004; Jjemba,

2006). Bezafibrate ((2-[4-2-[4-chlorobenzamido]ethylphenoxy]-

2-methylpropanoic acid) belongs to the group of fibrate drugs,

which is an important class of pharmaceuticals largely used

for the treatment of hyperlipidaemia, when raised cholesterol

r Ltd. All rights reserved.

fax: +34 934021291..F. Dantas).

levels are associated with increased levels of triglycerides.

Recently, bezafibrate (BZF) was included in the list of the most

used pharmaceuticals in the world (Cermola et al., 2005). BZF

concentrations up to 27 ng L�1 and 4.6mg L�1 have been found

in drinking waters and in sewage treatment plant (STP)

effluents in Germany (Stumpf et al., 1996; Ternes, 1998) and

up to 57.15 ng L�1 in Italian rivers (Calamari et al., 2003).

A recent work demonstrates that, at the concentration

detected in environment, BZF do not induce acute or chronic

toxic effects in non-target organisms. Nevertheless, it harm-

fulness to environment cannot be excluded due to the

possible mixture toxicity, synergistic and additives effects,

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 1 ( 2 0 0 7 ) 2 5 2 5 – 2 5 3 22526

bioaccumulation and biomagnification (Isidori et al., 2007).

Concerning BZF ozonation, no work aiming the toxicity

assessment of BZF ozonation byproducts was found in the

literature.

Advanced oxidation processes (AOPs) and, in particular,

ozonation have been applied to treat waters containing

antibiotics (Ernst and Jekel, 1999; Hua et al., 2006; Vogna

et al., 2004; Andreozzi et al., 2005) and as a pretreatment step

to improve the biodegradability of wastewaters containing

them (Balcioglu and Otker, 2003). Ozone is a strong oxidant

that can attack the organic molecules through a direct (ionic)

mechanism and a radical one which originates from its

decomposition in water to generate hydroxyl radicals (Sotelo

et al., 1987). Therefore, ozonation is an efficient method

capable to degrade several xenobiotic compounds and trans-

form refractory natural organic matter to biodegradable form,

i.e. biodegradable dissolved organic carbon (Goel et al., 1995;

Volk et al., 1993; Esplugas et al., 2002; Contreras et al., 2003),

although some authors have reported that the oxidized

intermediates can exhibit new toxicity to pure bacteria and

mixed microorganisms in the early stage of ozonation (Shang

et al., 2006).

In the present work, the results of a study on the ozonation

of BZF in aqueous solution are presented. An identification of

main intermediates is firstly attempted along with an

assessment of by-products biodegradability and acute toxi-

city. The second-order kinetic constants for ozone attack are

determined by means of absolute and competition method.

N

O

Cl

H

O

COOH

Ozonation

Ozonation

Fig. 1 – Possible centers of reaction for ozone attacks.

2. Materials and methods

Ozonation experiments were carried out in a 0.8 L reactor

with a continuous supply of ozone gas stream (Andreozzi

et al., 1992). Ozone was generated by means of a Fischer 502

Ozonator using oxygen as a feeding gas. The gaseous outlet

from the reactor is led to an ozone decomposer, for destroying

the residual ozone on a proper decomposition catalyst. The

aqueous solutions contaminated with an initial BZF concen-

tration ranging from 0.2 to 0.5 mmol L�1 were charged into the

reactor, where an ozonized gaseous mixture with an ozone

concentration of 1 mmol L�1 was bubbled in. The gas flow rate

was kept at 0.38 L min�1. Samples were withdrawn at fixed

reaction times and quickly analyzed. For the experiments

carried out at fixed pH, BZF solutions were buffered by the

addition of adequate quantities of Na2HPO4, Na3PO4 and

KH2PO4. For kinetic runs, the ionic strength (I) of the solutions

was adjusted at 0.1 mol L�1 by addition of phosphates salts.

The pH range chosen to perform the runs was from 6.0 to 8.0

(measured by an Orion 960 pH meter with a glass pH

electrode). No runs at pH lower than 6.0 could be carried out

because of the low solubility of BZF at acidic conditions

(21 mmol L�1 for pH 6.0; 0.27 mmol L�1 for pH 4.0 at 25 1C). The

data used to calculate the kinetic constants came from

samples withdrawn during the initial stages of reaction.

To prevent the intervention of radical mechanisms, some

experiments using t-butanol as scavenger were performed.

All the chemicals used were supplied by Aldrich (Italy). Total

organic carbon (TOC) measurements were performed by

means of a Shimadzu 5000 A TOC analyzer. The concentration

of the BZF was quantified by means of a high-performance

liquid chromatograph (HPLC) from Hewlett Packard (mod.

1090 II) equipped with a Synergy Polar 4RP column and a

UV–vis detector (l ¼ 200 nm). The eluent was a 60:40 buffered

solution (H2O:CH3OH:H3PO4 500:25:2):CH3CN with a flow rate

of 1.5 mL min�1. The LC–MS analyses were performed by

means of an HPLC–MS Agilent 1100 equipped with a Synergi

Polar 4RP column. Chromatographic conditions: a mixture of

75% formic acid (0.1% by volume) and 25% acetonitrile, flow

rate 0.7 mL min�1. The mass-spectrometric detection was

performed on MSD Quad VL (Agilent Mass Spectrometer)

equipped with electrospray ionization. MS data were acquired

in ESI mode (capillary temperature 350 1C; source voltage

3.5 kV, drying N2 gas flow 11 L min�1). The collision energy to

produce the desired quantity of [M+H]+ (positive mode) or

[M�H]� (negative mode) molecular ion was individually

optimized.

The biodegradability and acute toxicity of raw and ozonated

BZF solutions was tested. To follow the biodegradability of the

samples, the biological oxygen demand (BOD5) (Standard

method, 5210 D) and the chemical oxygen demand (COD)

(Standard method, 5220 D) were carried out. The bacteria

used to perform the BOD test came from BOD-seed capsules

supplied by Cole-Parmer Instrument Company (USA) and the

ratio BOD5/COD has been used as a biodegradability indicator.

The acute toxicity test was carried out with a Microtoxs

M500 toxicity analyzer, using Vibrio fischeri strains, according

to the manufacturer procedure (Azur Environmental, Dela-

ware, USA). The basic Microtoxs test is based on measuring

light emission from the bacteria when exposed to a samples

dilutions series of 45.00%, 22.5%, 11.25% and 5.62%. The

diluting solution was supplied by Azur Environmental and the

dilutions were performed to a total volume of 1 mL.

Results are expressed as EC50;15 min, which represents the

percentage of initial solution dilution (% v/v) that causes 50%

reduction of bacteria bioluminescence in 15 min. The Micro-

toxs test was operated in duplicate.

3. Results and discussions

3.1. Bezafibrate removal

A simple analysis of the structure of BZF molecule indicates

the presence of different possible centers of reaction repre-

sented by the two aromatic rings (Fig. 1) (Bailey, 1982).

ARTICLE IN PRESS

Time (min)

0 20 40 60 80 100

C/C

0

0.0

0.2

0.4

0.6

0.8

1.0

TO

C r

em

ova

l (%

)

0

20

40

60

80

100

Fig. 2 – Ozonation bezafibrate, T ¼ 25 1C, pH ¼ 6.0; ~: BZF concentration; }: chloride release; n: TOC removal.

Fig. 3 – Chromatogram recorded during LC–MS analysis of the sample collected at 5 min of ozonation.

WAT E R R E S E A R C H 41 (2007) 2525– 2532 2527

At a first sight, the chlorine-containing ring can be

supposed to be characterized by a lower reactivity with

respect to the unchlorinated one due to the electron with-

drawing effect of chlorine itself. Fig. 2 shows the data

collected during an experiment in which a solution contain-

ing a 0.5 mmol L�1 of BZF was submitted to ozonation at

pH ¼ 6.0. For a reaction time of about 10 min a complete

disappearance of the substrate was recorded, with a degree of

mineralization of only 20% and about 73% of initial chlorine

content released into the solution as chloride ions. The

continuous ozone feeding to the reactor for reaction times

longer than 10 up to 105 min allowed achieving mineraliza-

tion degrees not higher than 30%. On the other hand, an

increase of the concentration of chloride ions in the solution

was observed for reaction times longer than 10 min with a

final value of about 0.5 mmol L�1, corresponding to 100% of

initial chlorine content in the molecule.

3.2. Identification of chemical intermediates

To throw light on the reaction mechanism through which

ozone attacks BZF, HPLC–MS analyses were carried out on the

reaction samples collected at different reaction times.

In Fig. 3, the chromatogram recorded during the HPLC–MS

analysis on the sample collected at 5 min of ozonation is

reported. Unfortunately the sole HPLC–MS technique does not

allow the identification of all the peaks recorded. The peaks

for which the identification procedure was successful are

indicated with a capital letters in the chromatogram and their

main fragmentations along with the proposed structures are

shown in Table 1.

The formation of the species whose structures are proposed

in Table 1 may be justified on the basis of the well-known

ozone chemistry. In particular, the formation of all the

compounds identified can be explained on the basis of

ARTICLE IN PRESS

Table 1 – Proposed structure for the peaks highlighted in Fig. 3

Peak Tr

(min)m/z Ionization

modeProposed structure (M)

A 3.2 228 (M+1), 183, 169, 139, 111 Positive

B 5.9 366 (M�1), 280, 264, 154 Negative

C 6.4 392 (M�1), 306,154 Negative

D 8.3 408 (M�1), 322 Negative

E 8.7 408 (M�1), 322 Negative

F 9.8 360 (M�1), 274, 154 Negative Bezafibrate

WAT E R R E S E A R C H 4 1 ( 2 0 0 7 ) 2 5 2 5 – 2 5 3 22528

a direct mechanism of ozonation and an anomalous one

(Yamamoto et al., 1979).

For example, it has been reported (Bailey, 1982) that an

ozone attack on the aromatic ring (Fig. 4, step 1) may result

also in a hydroxylation reaction (species D).

The formation of the species A may be explained through

an anomalous ozonolysis mechanism (Fig. 4, steps 2a and 2b)

of the species C.

The presence of the species B and C may be explained on

the basis of a direct ozone attack to the unchlorinated

aromatic ring. For the species D and E the signals in the

mass spectra indicate that these structures differ from that of

BZF for three oxygen atoms ([M�1] ¼ 408 m/z for D and E and

[M�1] ¼ 360 m/z for BZF). On the basis of the known ozone

chemistry it can be put forward that the three oxygen atoms

are introduced in the molecules as hydroxyl groups. No

conclusive indications have been collected to fix the exact

position of these groups. The unavailability of the chemical

standards prevented the determination of the concentrations

of all the identified species.

3.3. Biodegradability and toxicity of intermediates

Since it is well known that oxidative processes can generate

intermediates which are more toxic than the starting species,

mainly in the early stage of oxidation, the biodegradability

and toxicity of the ozonation intermediates of BZF have been

investigated. Thus BOD5 and COD were measured during an

ozonation run up to 105 min. First of all, BOD5 measurements

indicated that a 0.5 mmol L�1 aqueous solution of BZF is

not readily biodegradable (BOD5 ¼ 0). In Fig. 5, the results

obtained on the samples collected during the course of the

ozonation are reported along with the ratio BOD5/COD which

continuously increases from 0 up to 0.26 with increasing

ozone dose. An increase in the BOD5/COD ratio indicates an

improvement in the biodegradability of the pollutant due to

the formation of biochemically more degradable reaction

products. A biodegradability ratio over 0.2 indicates a partially

biodegradable effluent being 0.4 (Sarria et al., 2002) a thresh-

old value above which the stream can be considered as

easily biodegradable. When all BZF was eliminated from the

ARTICLE IN PRESS

HO+

O OO-

H

+O3

O2 (step 1)

O3

HO C OCH

HC

HO C

OO

OH

+

HO C

CCHHO C(step 2a)

H+HC

HO C

OO

OH

+COOH+

OCHO

+

C

OH

OH2 OOHCH

+H- (step 2b)

Fig. 4 – Direct and anomalous mechanisms of ozonation.

Time (min)

CO

D/1

0, B

OD

5 (

mg/L

)

0

0

10

20

20

30

40

40 60 80 100

50

BO

D5

/CO

D0.0

0.2

0.4

0.6

0.8

1.0

Fig. 5 – Biodegradability increment during the ozonation of BZF. K: BOD5; J: COD; n: BOD5/COD.

WAT E R R E S E A R C H 41 (2007) 2525– 2532 2529

solution (after having dosed 0.73 mmol of ozone), the

biodegradability ratio was only 0.15, suggesting the presence

in the reacting solution of some intermediates not biodegrad-

able. Therefore it can be stated that longer ozonation times

are requested to increase the biodegradability of BZF-contain-

ing aqueous solutions.

To assess the acute toxicity of the samples collected during

the ozonation runs, the inhibition of the light emission of

V. fischeri bacteria caused by the presence of toxic compounds

was measured by Microtoxs test. In Fig. 6, the values of

1=EC50;15 min calculated for different preozonated samples are

presented. From this figure it can be observed that the acute

toxicity of BZF starting solution is relatively low (EC50 near

80%). Acute toxicity response from Microtoxs considerably

changes when analyzing the ozonation products. Since a

higher value of 1/EC50 implies a higher inhibition to the

bacteria, a significant increase in toxicity was observed in the

early stage of BZF ozonation. Nevertheless, the subsequent

intermediates degradation promotes a toxicity reduction

which after 105 min of ozonation was lower than that of the

initial BZF solution.

The Microtoxs toxicity trend of preozonated samples is in

agreement with the BOD5/COD results calculated for the same

ozonation time. It can be thus concluded that ozonation is a

suitable technique to improve the biodegradability and reduce

the toxicity of waters containing BZF, although optimal ozone

dose needs to be determined according to the effluent

composition.

3.4. Determination of kinetic constants: absolute method

The ozonation runs were carried out by adding 0.1 mL of a

radical scavenger (t-butanol) to the solution (VL ¼ 800 mL) in

order to avoid the influence of radical mechanism (Hoigne

and Bader, 1983). BZF solutions were buffered at pH 6.0, 7.0

and 8.0 and the ionic strength adjusted to 0.1 mol L�1. BZF

solutions with initial concentrations ranging from 0.2 to

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 1 ( 2 0 0 7 ) 2 5 2 5 – 2 5 3 22530

0.5 mmol L�1 were ozonated following the BZF abatement and

ozone concentration in the freeboard of the reactor.

To determine the second-order kinetic constants for the

ozonation of BZF, two methods were carried out. In the first

one (absolute method) the data were analyzed by a proper

model, which had been developed by coupling fluid-dynamic

and chemical sub-models.

By assuming that the ozonation of BZF developed in a

quasi-diffusional regime of absorption with reaction, a set

of differential mass-balance equations for ozone in the

two phases (bubble, [O3]B and freeboard, [O3]F), each consid-

ered as a well-mixed stirred reactor and the substrate can be

written:

d½O3�B

dt¼

QVBð½O3�in � ½O3�BÞ �

koLa½O3�BaE

VBVL, (1)

d½O3�F

dt¼

QVFð½O3� � ½O3�FÞ, (2)

d½BZF�dt

¼ �ko

La½O3�BaEz

, (3)

where VB and VF are bubble and freeboard volumes, respec-

tively, and Q the volumetric gas flow rate.

The Enhancement factor E was calculated through the

following formulas:

E ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1þ Ha2

p, (4)

Table 2 – BZF kinetic parameters obtained from experimental

pH 6.0

kBZF (L mol�1 s�1) 2.66� 10370.29� 103

A 1.9671.6� 10�2

B 0.9670.20

sS (%) 6.08

sO3(%) 3.76

T ¼ 25 1C, I ¼ 0.1 M.

sS ¼ Percentage standard deviation for the substrate (BZF).

sO3 ¼ Percentage standard deviation for the ozone in freeboard.

Time (min)

0 20 40 60 80 100

1/E

C50

0.00

0.02

0.04

0.06

0.08

0.10

Fig. 6 – 1=EC50;15 min from Microtoxs test versus time for the

BZF ozonation.

where Ha is the dimensionless Hatta’s number:

Ha ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiDO3 zkBZF½BZF�o

ðkoLÞ

2

s, (5)

being known for the adopted experimental conditions the

mass-transfer coefficient koL, the overall volumetric physical

mass-transfer coefficient koLa in absence of chemical reaction

(Andreozzi et al., 2003), the ozone diffusivity DO3(Reid et al.,

1983) and Ostwald coefficient a (Andreozzi et al., 1996a).

As previously reported (Andreozzi and Marotta, 1999;

Andreozzi et al., 1996b), an overall reaction has been used to

describe the oxidation of the substrate which in the present

case is

BZFþ ZO3�!kBZF

Products;

where z is the stoichiometric coefficient represented as a

linear function of the reaction time (z ¼ a+bt).

This model allowed to estimate the best values of kinetic

constant kO3 and the stoichiometric coefficients ‘‘a’’ and ‘‘b’’

by means of a suitable optimizing procedure in which the

minimum of the objective function is found

f ¼XNm

m¼1

XNj

j¼1

ðYm;j � Cm;jÞ2,

where Nm and Nj represent the set of measured species and

experimental reaction times, respectively, Ym,j, Cm,j indicate

calculated and measured mth component concentrations at

the jth reaction time.

In Table 2, the values calculated for these parameters are

shown. The values of kinetic constants show a small increase

by raising the pH, although all are of the same order of

magnitude.

With the aim of validating the results reported above, a

second method was used for estimating the kinetic constants

of the ozonation of BZF at the three investigated pH values.

3.5. Determination of kinetic constants: competitionmethod

The competition method is based on the comparison between

the degradation rate of BZF and that of a reference compound.

Maleic acid (MAL) was chosen as a reference compound since

during a literature survey it has been found that it shows

kinetic constant values for the ozonation which are very close

runs at different pH values

7.0 8.0

4.24�10370.66� 103 1.0� 10471.07� 103

2.40733�10�2 2.02719� 10�2

0.5370.36 1.5370.24

7.60 5.81

8.98 8.67

ARTICLE IN PRESS

WAT E R R E S E A R C H 41 (2007) 2525– 2532 2531

to those calculated for the BZF by means of the absolute

method. From the literature review, four different kinetic

constants for MAL ozonation were found (Hoigne and Bader,

1983; Pryor et al., 1984; Caprio et al., 1987; Benbelkacem et al.,

2003).

For each experiment a solution containing both MAL and

BZF in concentration of 0.4 mmol L�1 was ozonized and their

concentrations were followed with the time. In Fig. 7, the

results of one of these runs are shown.

If one considers that the following equations can be written

to describe the degradation of the two species:

d½BZF�dt

¼ �kBZF½O3�½BZF�, (6)

d½MAL�dt

¼ �kMAL½O3�½MAL�. (7)

Dividing Eq. (7) by Eq. (6) and integrating between t ¼ 0 and tR

the following relationship is obtained:

ln½MAL�½MAL�0

¼ �kMAL

kBZFln½BZF�½BZF�0

. (8)

It is possible to calculate the ratio kMAL/kBZF by plotting the

neperian logarithm of the normalized concentration of both

compounds one against the other thus obtaining a straight

line. As the kMAL is known from the literature, the kBZF can be

easily calculated.

Because previously stated different values can be adopted

for the kinetic constant kMAL according the literature (Table 3),

calculated second-order kinetic constants for BZF ozonation

are plotted versus pH (Fig. 8). All the values for kBZF calculated

according to literature data reported for MAL are inserted in

Fig. 8 and connected by dotted lines. It is interesting to note

Table 3 – Value adopted for kMAL (L mol�1 s�1) according to lite

pH Ratio kMAL/kBZF Caprio et al. (1987) Hoigne and Bad

6.0 3.4 1.60� 104 5.00�10

7.0 4.2 2.82� 104 —

8.0 2.3 3.12� 104 —

Time (min)

0.0

C/C

0

0.0

0.2

0.4

0.6

0.8

1.0

2.52.01.51.00.5

Fig. 7 – Normalized concentration of BZF (J) and MAL (K)

versus time. Buffered at pH ¼ 6.0, T ¼ 25 1C, ionic

strength ¼ 0.1 mol L�1.

that the values calculated at each pH (Deborde et al., 2005)

employing the data for MAL given by Caprio et al., are about

2–3 times higher then those obtained by other researchers.

The values obtained by means of the absolute method (full

circles) are within the area delimited by dotted lines thus

showing a good agreement between the two methods,

although it was not possible to carry out a careful statistical

analysis, no uncertanties on the kinetic constant values being

given in the original quoted papers.

A slight increase of the kinetic constants values presented

for both methods when the pH passes from to 6.0 to 8.0 is

evident from the diagrams of Fig. 8. Considering that in the

studied range of pH all the BZF molecules are dissociated

(pKa ¼ 3.29, ACD Software, 2006), the observed increase in the

reactivity cannot be ascribed to a change in the dissociation

degree when the pH varied from 6.0 to 8.0. A possible

explanation may be found in the intervention of the radical

mechanism whose occurrence could not be completely

prevented by the scavenger addition at increasing the pH.

4. Conclusions

Ozonation has been proven to be an effective method to

remove BZF from aqueous solution. Nevertheless, low miner-

alization degrees have been achieved during the ozonation

runs performed in the present study although a chloride

release corresponding to the 100% of initial chlorine content

in BZF molecule was observed. The LC–MS analyses indicated

that ozone react mainly with the unchlorinated ring, with

rature to evaluate kBZF

er (1983) Benbelkacem et al. (2003) Pryor et al. (1984)

3 — —

1.20� 104 —

— 2.40�104

pH

5.5

kB

ZF

(L m

ol-1

s-1

)

0

2000

4000

6000

8000

10000

12000

14000

16000

8.58.07.57.06.56.0

Fig. 8 – Kinetic constant values for the BZF ozonation versus

pH obtained by means of absolute method (K) and

calculated using the literature data for maleic acid reported

in Table 3 (J: Caprio et al., 1987; }: Hoigne and Bader, 1983;

,: Benbelkacem et al., 2003; &: Pryor et al., 1984).

ARTICLE IN PRESS

WAT E R R E S E A R C H 4 1 ( 2 0 0 7 ) 2 5 2 5 – 2 5 3 22532

respect to chlorinated one, leading to aldehydic and ketonic

intermediates. Moreover the presence of some species with

three additional oxygen atoms suggests that hydroxylation

mechanisms of both the aromatic rings may also occur.

A good agreement was observed between the absolute and

the competition kinetic method, adopted for the kinetic

investigations. Second-order kinetic constants for the ozone

attack to the substrate were found in the range between

2.7�103 and 1.0�104 L mol�1 s�1 with absolute method and

1.5�103 and 1.3�104 L mol�1 s�1 with competition one. The

ozonation of BZF-containing aqueous solutions showed a

continuous increase of the ratio BOD5/COD up to a value of

0.26 thus indicating an enhancement of the biodegradability

of the preozonated samples. The toxicity of these treated

solutions determined by means of the Microtoxs test

changed in a more complex way with an initial increase (in

the early ozonation stages) and a successive reduction to

values lower than that measured for the starting solution.

R E F E R E N C E S

Advanced Chemistry Development (ACD/Labs) Software V8.14 forSolaris (& 1994–2006 ACD/Labs).

Andreozzi, R., Marotta, R., 1999. Ozonation of p-chlorophenol inaqueous solution. J. Hazard. Mater. B69, 303–317.

Andreozzi, R., Insola, A., Caprio, V., D’Amore, M.G., 1992. Quino-line ozonation in aqueous solutions. Water Res. 26 (5), 639–643.

Andreozzi, R., Caprio, V., Ermellino, I., Insola, A., Tufano, V., 1996a.Ozone solubility in phosphate buffered aqueous solutions: theeffect of temperature, t-butyl alcohol, pH. Ind. Eng. Chem. Res.35, 1467–1471.

Andreozzi, R., Caprio, V., Insola, A., Tufano, V., 1996b. Measuringozonation rate constants in gas–liquid reaction under thekinetic–diffusional transition regime. Chem. Eng. Commun.143, 195–197.

Andreozzi, R., Caprio, V., Marotta, R., Radovnikovic, A., 2003.Ozonation and H2O2/UV treatment of clofibric acid in water: akinetic investigation. J. Hazard. Mater. B103, 233–246.

Andreozzi, R., Canterino, M., Marotta, R., Paxeus, N., 2005.Antibiotic removal from wastewaters: the ozonation ofamoxicillin. J. Hazard. Mater. 122 (3), 243–250.

Bailey, P.S., 1982. Ozonation in Organic Chemistry, vol. 39-II.Academic Press, New York.

Balcioglu, I.A., Otker, M., 2003. Treatment of pharmaceuticalwastewater containing antibiotics by O3 and O3/H2O2 pro-cesses. Chemosphere 50, 85–95.

Benbelkacem, H., Cano, H., Mathe, S., Debellefontaine, H., 2003.Maleic acid ozonation: reactor modeling and rate constantsdetermination. Ozone-Sci. Eng. 25 (1), 13–24.

Calamari, D., Zuccato, E., Castiglioni, S., Bagnati, R., Fanelli, R.,2003. Strategic survey of therapeutic drugs in the riversPoand Lambro in Northen Italy. Environ. Sci. Technol. 37,1241–1248.

Caprio, V., D’Amore, M.G., Insola, A., 1987. The kinetics ofozonation of maleic acid in aqueous solution. Oxid. Com-mun. 10 (1–2), 105–113.

Cermola, M., Della Greca, M., Iesce, M.R., Previtera, L., Rubino, M.,Temussi, F., Brigante, M., 2005. Phototransformation of fibratedrugs in aqueous media. Environ. Chem. Lett. 3, 43–47.

Contreras, S., Rodrıguez, M., Momani, F.A., Sans, C., Espulgas, S.,2003. Contribution of the ozonation pre-treatment to the

biodegradation of aqueous solutions of 2,4-dichlorophenol.Water Res. 37, 3164–3171.

Deborde, M., Rabouan, S., Duguet, J.P., Legube, B., 2005. Kinetics ofaqueous ozone-induced oxidation of some endocrine disrup-tors. Environ. Sci. Technol. 39 (16), 6086–6092.

Ernst, M., Jekel, M., 1999. Advanced treatment combination forgroundwater recharge of municipal wastewater by nanofil-tration and ozonation. Water Sci. Technol. 40 (4–5), 277–284.

Esplugas, S., Gımenez, J., Contreras, S., Pascual, E., Rodrıguez, M.,2002. Comparison of different advanced oxidation process forphenol degradation. Water Res. 36, 1034–1042.

Goel, S., Raymond, M.H., Bouwer, E.J., 1995. Biodegradation ofNOM: effect of NOM source and ozone dose. J. Am. WaterWorks Assoc. 87 (1), 90–105.

Heberer, T., 2002. Occurrence, fate and removal of pharmaceuticalresidues in the aquatic environment: a review of recentresearch data. Toxicol. Lett. 131, 5–17.

Hoigne, J., Bader, H., 1983. Rate constants of reactions of ozonewith organic and inorganic compounds in water. II Dissociat-ing organic compounds. Water Res. 17, 185–194.

Hua, W., Bennett, E.R., Letcher, R.J., 2006. Ozone treatment and thedepletion of detectable pharmaceuticals and atrazine herbi-cide in drinking water sourced from the upper Detroit River,Ontario, Canada. Water Res. 40 (12), 2259–2266.

Isidori, M., Nardelli, A., Pascarella, L., Rubino, M., Parrella, A., 2007.Toxic and genotoxic impact of fibrates and their photopro-ducts on non-target organisms. Environ. Int., in press.

Jjemba, P.K., 2006. Excretion and ecotoxicity of pharmaceuticaland personal care products in the environment. Ecotoxicol.Environ. Safety 63 (1), 113–130.

Pryor, W.A., Giamalva, D.H., Church, D.F., 1984. Kinetics ofozonation. 2. Amino acids and model compounds in waterand comparisons to rates in nonpolar solvents. J. Am. Chem.Soc. 106, 7094–7100.

Reid, R.C., Prausnitz, J.M., Poling, B.E., 1983. The Properties ofGases and Liquids. Wiley, New York.

Sarria, V., Parra, S., Adler, N., Peringer, P., Benitez, N., Pulgarin, C.,2002. Recent developments in the coupling of photoassistedand aerobic biological processes for the treatment of bior-ecalcitrant compounds. Catal. Today 76 (2–4), 301–315.

Shang, N.-C., Yu, Y.-H., Ma, H.-W., Chang, C.-H., Liou, M.-L., 2006.Toxicity measurements in aqueous solution during ozonationof mono-chlorophenols. J. Environ. Manage. 78, 216–222.

Sotelo, J.L., Benitez, F.J., Beltran-Heredia, J., 1987. Ozone decom-position in water: kinetics study. Ind. Eng. Chem. Res. 26,39–43.

Stackelberg, P.E., Furlong, E.T., Meyer, M.T., Zaugg, S.D., Hender-son, A.K., Reissman, D.B., 2004. Persistence of pharmaceuticalcompounds and other organic wastewater contaminantsin a conventional drinking-water-treatment plant. Sci. TotalEnviron. 329 (1/3), 99–113.

Stumpf, M., Ternes, T.A., Haberer, K., Seel, P., Baumann, W., 1996.Determination of drugs in sewage treatment plants and riverwater. Vom Wasser 86, 291–303 (in German).

Ternes, T.A., 1998. Occurrence of drugs in German sewagetreatment plants and rivers. Water Res. 32, 3245–3260.

Vogna, D., Marotta, R., Napolitano, A., Andreozzi, R., d’Ischia, M.,2004. Advanced oxidation of the pharmaceutical drug diclo-fenac with UV/H2O2 and ozone. Water Res. 38 (2), 414–422.

Volk, C., Renner, P., Paillard, H., Joret, J.C., 1993. Effects of ozone onthe production of biodegradable dissolved organic carbon(BDOC) during water treatment. Ozone Sci. Eng. 15 (5), 389–404.

Yamamoto, Y., Niki, E., Shiokawa, H., Kamiya, Y., 1979. Ozonationof organic compounds 2. Ozonation of phenol in water. J. Org.Chem. 44, 2137.