water research 44 (2010) 3109–3120

Available at www.sciencedirect.com

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

Electrochemical degradation of b-blockers. Studies on single

and multicomponent synthetic aqueous solutions

Ignasi Sirés a,b,*, Nihal Oturan a, Mehmet A. Oturan a

a
Université Paris-Est, Laboratoire Géomatériaux et Géologie de l’Ingénieur, 5 Bd Descartes, 77454 Marne-la-Vallée Cedex 2, France
b
Laboratori d’Electroquı́mica dels Materials i del Medi Ambient, Departament de Quı́mica Fı́sica, Facultat de Quı́mica, Universitat de
Barcelona, Martı́ i Franquès 1-11, 08028 Barcelona, Spain

article info abstract

Article history: As far as we know, this is the first study reporting the electrochemical decontamination of
Received 26 October 2009 solutions containing b-blockers, which are pharmaceutical pollutants with a high occurrence
Received in revised form in natural waters. The oxidation ability of two pre-eminent, eco-friendly electrochemical
28 February 2010 advanced oxidation processes (EAOPs), namely anodic oxidation (AO) and electro-Fenton
Accepted 1 March 2010 (EF), has been compared at lab-scale by carrying out bulk electrolyses at pH 3.0 at constant
Available online 6 March 2010 current using a carbon-felt cathode able to electrogenerate H2O2 in situ. The studies of single
component aqueous solutions were focused on atenolol as a model b-blocker. The AO process
Keywords: was proven much more effective using a large surface area boron-doped diamond (BDD)
Atenolol anode than a Pt one, which was explained by the great amount of active hydroxyl radicals
b-blockers (BDD(OH)) and the minimization of their parasitic reactions. The EF process with a Pt anode
Electrochemical advanced oxidation and 0.2 mmol l1 Fe2þ showed even higher performance, with fast destruction of atenolol
processes (EAOPs) following pseudo-first order kinetics and fast mineralization because the oxidation process
Electro-Fenton in the bulk allows overcoming the mass transport limitations. The time course of the
Pharmaceutical pollutants concentration of the aromatic and short-chain carboxylic acid intermediates demonstrated
Water treatment technologies the progressive detoxification of the solutions. Almost 100% of the initial N content was
accumulated as NHþ
4 . Multicomponent solutions containing atenolol, metoprolol, and

propranolol, which usually occur together in the aquatic environment, were treated by
EF using the Pt/carbon felt cell. A high mineralization rate was observed up to the overall total
organic carbon (TOC) removal, which allowed reducing the energy consumption. The abso-
lute rate constant for the reaction of each b-blocker with OH was determined and the
reactivity was found to increase in the order: atenolol (1.42  109 l mol1 s1) < metoprolol
(2.07  109 l mol1 s1) < propranolol (3.36  109 l mol1 s1).
ª 2010 Elsevier Ltd. All rights reserved.

1. Introduction blockers, are b-adrenoceptors antagonists used to treat

cardiovascular disorders including hypertension, angina
Nowadays, the presence of pharmaceutical pollutants in the pectoris, and arrhythmia. They are ubiquitous in natural
aquatic environment is considered a relevant topic (Khetan waters since they are one of the pre-eminent groups of
and Collins, 2007). b-Blockers, also called b-adrenergic prescription drugs (Khetan and Collins, 2007). In this study, we

* Corresponding author. Université Paris-Est, Laboratoire Géomatériaux et Géologie de l’Ingénieur, 5 Bd Descartes, 77454 Marne-la-Vallée
Cedex 2, France. Tel.: þ34 93 402 12 23; fax: þ34 93 402 12 31.
E-mail address: i.sires@ub.edu (I. Sirés).
0043-1354/$ – see front matter ª 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2010.03.005
3110 water research 44 (2010) 3109–3120
have focused on the three most widely reported b-blockers in
the environment, namely atenolol, metoprolol, and propran-
olol, whose parent compound is 3-amino-1-phenoxy-2-prop-
anol (see Fig. 1). They are very polar, highly water soluble
compounds and are among the top selling pharmaceuticals.
As a result, they have been detected in sewage treatment
plants (STPs) effluents: in Europe, mean values of 0.33, 0.08,
and 0.01 mg l1 (Andreozzi et al., 2003; Vieno et al., 2007) and
maximum values of 1.2, 2.2, and 0.29 mg l1 (Hirsch et al., 1996;
Vieno et al., 2007) have been reported for atenolol, metoprolol,
and propranolol, respectively; in the USA, concentrations of
up to 1.20 mg l1 metoprolol and 1.90 mg l1 propranolol were
found (Huggett et al., 2003). The STPs are considered as ‘‘hot
spots’’ of aquatic contamination by pharmaceuticals and their
metabolites because the elimination is usually incomplete by
conventional water treatment technologies (Vieno et al.,
2007); chlorination is effective concerning the degradation of
the initial pollutant but harmful chloramines and trihalo-
methanes might be released into the environment (Bedner
and MacCrehan, 2006), whereas the oxidative degradation by
ozonation (Ikehata et al., 2006) and advanced oxidation
processes (AOPs) such as heterogeneous photocatalysis and
Fenton processes is usually safer (Ikehata et al., 2006;
Klavarioti et al., 2009).
Toxicological studies for b-blockers are rather scarce but
show that aquatic organisms are sensitive to them. The
three aforementioned b-blockers often occur together in
waters (Ternes et al., 2003; Maurer et al., 2007), demon-
strating a high degree of persistence (i.e., the supply rate is
greater than the transformation rate (Daughton and Ternes,
1999)) (Bendz et al., 2005), so it seems interesting to study
their behavior in mixed solutions. Their potential carcino-
genicity, mutagenicity, and teratogenicity have been care-
fully revised (Jackson and Fishbein, 1986). Studies with
plankton (Daphnia magna) and algae have revealed species-
specific toxic effects. In addition, evidence suggests that
fish contain b-adrenergic receptors that could interact with
b-blockers (Huggett et al., 2002). The toxicity of mixtures has
been predicted but hazard characterization is not suffi-
ciently clear (Cleuvers, 2005).
NH
NH2 NH
O O
OH
3-amino-1-phenoxy-2-propanol
NH
Fig. 1 – Chemical structures for the three b-blockers studied and their parent compound.
Despite their ubiquity, there exist few studies reporting the
use of technologies for the degradation of aqueous solutions
of these b-blockers. For example, the solar photodegradation
of dilute solutions showed half-lives of 16, 350, and 630 h for
propranolol, atenolol, and metoprolol, respectively (Liu and
Williams, 2007), whereas pulse and gamma radiolysis have
been used to assess the effect of the two major free radicals,
namely hydroxyl radical (OH) and hydrated electron (eaq), on
the three compounds (Slegers and Tilquin, 2006; Song et al.,
2008). Ozonation at pH 7.2 was effective to degrade the three
b-blockers in an STP effluent (Ternes et al., 2003). Also H2O2/
UV can lead to the degradation of propranolol, whereas TiO2
photocatalysis is less effective (Andreozzi et al., 2004).
However, most of the studies do not focus on the oxidation by-
products and the mineralization (conversion to CO2, H2O, and
inorganic ions) degree.
It is especially remarkable that the electrochemical degra-
dation of these three b-blockers has not been addressed yet,
because it has been shown to be more efficient than AOPs such
as the Fenton oxidation and ozonation for the treatment of
organic pollutants, favoring the destruction of refractory
compounds and reducing the operation costs (Cañizares et al.,
2009a, 2009b). Indeed, a large variety of electrochemical
advanced oxidation processes (EAOPs) have recently been
developed for the oxidative degradation of refractory organic
pollutants, and their fundamentals and applications are
detailed in some substantial reviews (Martı́nez-Huitle and
Brillas, 2008; Martı́nez-Huitle and Quiroz-Alfaro, 2008; Brillas
et al., 2009; Oturan et al., 2009; Panizza and Cerisola, 2009).
EAOPs are OH-mediated electrochemical treatments where the
destruction of pollutants takes place: (i) at the anode (i.e., anodic
oxidation (AO)), and/or (ii) in the bulk solution using the elec-
trogenerated Fenton’s reagent (H2O2 þ Fe2þ). Various classes of
pharmaceuticals such as analgesics/antipyretics (Sirés et al.,
2006; Guinea et al., 2008; Skoumal et al., 2009), antibiotics and
antimicrobials (Carlesi Jara et al., 2007; Sirés et al., 2007a; Li et al.,
2008), and blood lipid regulators metabolites (Sirés et al., 2007b)
have been successfully degraded by different EAOPs. The results
obtained by means of the electro-Fenton (EF) process using
a large surface area carbon-felt cathode are especially notorious
NH2
ATENOLOL
O
O
OH
O
METOPROLOL
OH
O PROPRANOLOL
OH
 water research 44 (2010) 3109–3120
due to its simplicity, low cost, and outstanding performance, as
confirmed for other organic pollutants (Panizza and Cerisola,
2001; Guivarch et al., 2003; Qiang et al., 2003; Hammami et al.,
2007). In such case, the EF mechanism involves the following
steps (Oturan et al., 2008; Brillas et al., 2009): (i) two main reac-
tions at the cathode, which are the continuous electro-
generation of H2O2 from the two-electron O2 reduction and the
regeneration of the Fe2þ catalyst, followed by (ii) the Fenton’s
reaction in the bulk between Fe2þ and H2O2 to produce the
powerful hydroxyl radical (OH).
Here we report the first study on the electrochemical degra-
dation of synthetic aqueous solutions of b-blockers. The
behavior of single component solutions was studied by using
atenolol as a model b-blocker. Bulk electrolyses were carried out
at constant current in a small, open cell with a carbon-felt
cathode and a Pt or boron-doped diamond (BDD) anode, in the
absence and presence of iron catalyst to compare the oxidation
ability of AO and EF processes. Reversed-phase and ion-
exclusion high performance liquid chromatography (HPLC)
allowed the quantification of the main intermediates. Synthetic
multicomponent solutions containing atenolol, metoprolol, and
propranolol were electrolyzed as well. The use of a gradient
HPLC method simplified the study because it allowed their
simultaneous determination. The absolute rate constant for the
reaction between each b-blocker and OH was also determined.
The results show that the EF process is a promising alternative
technology for the removal of b-blockers from waters.
2. Materials and methods
2.1. Chemicals
Atenolol (i.e., 4-[20 -hydroxy-30 -(isopropylamino)propoxy]ph-
enylacetamide; C14H22N2O3), metoprolol tartrate salt (i.e., 1-
(isopropylamino)-3-[p-(b-methoxyethyl)phenoxy]-2-propanol
tartrate salt; (C15H25NO3)2$C4H6O6), and propranolol hydro-
chloride (i.e., 1-isopropylamino-3-(1-naphthyloxy)-2-propanol
hydrochloride; C16H21NO2HCl) used as model b-blockers as
well as p-hydroxyphenylacetamide, p-hydroxyphenylacetic
acid, p-benzoquinone, and hydroquinone tested as possible
aromatic intermediates were reagent grade from Sigma–
Aldrich (98%). Reagent grade p-hydroxybenzoic acid from
Acros Organics was used as the competition substrate in some
kinetic experiments. Oxalic, phosphoric, and sulfuric acids,
anhydrous sodium sulfate (background electrolyte), heptahy-
drated ferrous sulfate used as Fe2þ source (catalyst) as well as
sodium nitrate, sodium nitrite, and ammonium oxalate
employed in ion chromatography were analytical grade from
Fluka, Merck, and Acros Organics. Organic solvents and the
other chemicals used were either HPLC or analytical grade
from VWR International (Prolabo), Fluka, and Panreac. All
aqueous solutions were prepared with ultra-pure water
obtained from a Millipore Milli–Q system with resistivi-
ty > 18 MU cm at room temperature.
2.2. Instruments and analytical procedures
Electrolyses were performed with a Hameg HM8040 triple
power supply at constant current. Solution pH was measured
3111
with a CyberScan pH 1500 pH-meter from Eutech Instruments.
The absorption spectra of atenolol were obtained in the range
200–500 nm by using a Perkin–Elmer (model Lambda 10) UV/
Vis spectrophotometer. The mineralization of the b-blockers
solutions was assessed from the decay of their dissolved
organic carbon, which can be considered as the total organic
carbon (TOC) when treating highly water soluble organic
compounds such as the ones studied. This analytical param-
eter was determined on a Shimadzu VCSH TOC analyzer.
Reproducible TOC values with 2% accuracy were found using
the non-purgeable organic carbon method. From these data,
the mineralization current efficiency (MCE) for the treated
solutions at a given time t (h) can be calculated from the
following equation (Flox et al., 2007):
ðDTOCÞexp n F Vs
MCEð%Þ ¼  100 (1)
4:32  107 m I t
where D(TOC)exp is the experimental TOC decay
(mg carbon l1), n is the number of electrons exchanged in the
mineralization process per molecule of organic compound, F
is the Faraday constant (¼96,487 C mol1), Vs is the volume of
the treated solution (l), 4.32  107 is the conversion factor for
units homogenization (¼3600 s h1  12,000 mg carbon mol1),
m is the number of carbon atoms in the molecule under study,
and I is the applied current (A). In particular, the degradation
of atenolol involves its transformation into CO2 and ammo-
nium ions, as explained below. Its mineralization reaction can
then be written as follows:
C14H22N2O3 þ 25 H2O / 14 CO2 þ 2 NHþ þ
4 þ 64 H þ 66 e

(2)
so that in this case n ¼ 66.
The time course of the concentration of all three b-blockers
and their aromatic intermediates was followed by reversed-
phase HPLC using a Merck Lachrom liquid chromatograph
equipped with an L-7100 pump, fitted with a Purospher RP-18,
5 mm, 25 cm  4.6 mm (i.d.) column at 40  C, and coupled with
an L-7455 photodiode array detector selected at the optimum
wavelength of 220 nm, as revealed by their absorption spectra.
The elution was performed isocratically with a methanol/
water (1% H3PO4) 8:92 (v/v) mixture as the mobile phase for
both, the single component studies with atenolol as a model
b-blocker, and the determination of the absolute rate
constant of atenolol (kabs,ATE) with p-hydroxybenzoic acid
(lmax ¼ 254 nm) as the competition substrate. The interme-
diates could be identified under such conditions; however,
a 15:85 (v/v) mixture allowed a significant reduction of the
monitoring time and prevented the peak overlapping, thus
facilitating the correct quantification of the intermediates. In
the multicomponent studies, the use of the following gradient
HPLC program based on an acetonitrile/water mixture allowed
the simultaneous identification and quantification of the
three b-blockers: acetonitrile was maintained at 4% from 0 to
10 min, then linearly increased to reach 18% at 15 min and 24%
at 25 min, kept at 24% for 5 min, and decreased to reach 4% at
40 min. The equilibration time between two consecutive
injections was set at 5 min. A flow rate of 0.8 ml min1 was
always used. The standard solutions to prepare the external
calibration curves of all aromatics were prepared in hydro-
organic medium to completely dissolve them; the iron catalyst
was added in some cases when it was convenient to simulate
3112 water research 44 (2010) 3109–3120
any possible complexation effect given during the EF treat-
ment. Oxalic acid was identified by ion-exclusion HPLC chro-
matography using a Merck Lachrom liquid chromatograph
equipped with an L-2130 pump, fitted with a C18 Acclaim OA
(i.e., organic acid), 4  250 mm (i.d.) column at 30  C, and
coupled with a L-2400 UV detector selected at l ¼ 210 nm.
A 100 mmol l1 Na2SO4 (adjusted at pH 2.65 with meth-
anesulfonic acid) at a flow rate of 0.2 ml min1 was used as the
mobile phase. Measurements were always controlled through
EZChrom Elite 3.1 software. Identification of intermediates
was made by comparison of retention times and UV spectra
with those of pure standards when they were available. The
nitrogenated inorganic ions released were determined by ion
chromatography upon injection of 25-ml aliquots into a Dionex
ICS-1000 Basic Ion Chromatography System. For the deter-
mination of anions, the system was fitted with an IonPac
AS4A-SC, 25 cm  4 mm (i.d.), anion-exchange column, linked
to an IonPac AG4A-SC, 5 cm  4 mm (i.d.), column guard. The
sensitivity of this detector was improved from electrolyte
suppression using an ASRS-ULTRA II self-regenerating
suppressor. Measurements were conducted with a solution
of 1.8 mmol l1 Na2CO3 and 1.7 mmol l1 NaHCO3 at
2.0 ml min1 as the mobile phase. For the determination of
cations, an IonPac CS12A, 25 cm  4 mm (i.d.), cation-
exchange column, was used, linked to an IonPac CG12A,
5 cm  4 mm (i.d.), column guard, and using a CSRS ULTRA II
suppressor. A 9.0 mmol l1 H2SO4 solution at 1.0 ml min1 was
used as the mobile phase. The system was equipped with
a DS6 conductivity detector containing a cell heated at 35  C
and was controlled through Chromeleon SE software.
2.3. Electrolytic systems for the degradation of
b-blockers
Bulk electrolyses of synthetic solutions were carried out in
a small, open, undivided glass cell of 6 cm diameter and
250 ml capacity. The AO process was performed using either
a 4.5 cm2 cylindrical Pt mesh, a 4.5 cm2 thin-film BDD elec-
trode deposited on a niobium substrate (from Magneto Chem),
or a similar 25 cm2 BDD electrode as the anode, and a large
surface area (14 cm  5 cm each side, 0.5 cm in width) carbon
felt from Carbone-Lorraine as the cathode. For the EF treat-
ment, only the Pt/carbon felt cell was tested. In all cases, the
anode was centered in the electrolytic cell and was sur-
rounded by the cathode, which covered the inner wall of the
cell. H2O2 was produced in all the electrolyses from reduction
of O2 dissolved in the solution. Continuous saturation of this
gas at atmospheric pressure was ensured by bubbling
compressed air having passed through a frit at about 1 l min1,
starting 10 min before electrolysis to reach a stationary O2
concentration. The cell voltage along the electrolyses was
displayed by the power supply.
Aqueous solutions of b-blockers were filtered prior to
electrolysis. The behavior of single component solutions was
assessed with 0.15 or 0.3 mmol l1 atenolol and 0.05 mol l1
Na2SO4, at pH 3.0 at room temperature (¼23  2  C), under the
application of a constant current in the range 30–300 mA. The
EF treatments were performed at pH 3.0, because it is the
optimum one to carry out the Fenton’s reaction (Brillas et al.,
2003; Sirés et al., 2006), and using 0.2 mmol l1 Fe2þ as the
catalyst, because it has been shown to be an optimum amount
to carry out the EF process with a reduced contribution of the
parasitic reactions (Brillas et al., 2009). Multicomponent solu-
tions containing atenolol, metoprolol, and propranolol, each
at a concentration of 0.15 mmol l1, were electrolyzed with the
Pt/carbon felt cell under EF conditions. All trials were
performed with solutions of 220 ml, with vigorous stirring by
a magnetic PTFE follower during the treatment to enhance
mass transport. In all cases, the treatment of high TOC
contents compared to the values usually reported in the
environment facilitated the comparison between the different
EAOPs and systems as well as the identification and moni-
toring of the oxidation intermediates.
3. Results and discussion
3.1. Degradation of atenolol by anodic oxidation and
electro-Fenton process
Atenolol has been selected as the model b-blocker to clarify
the behavior of single component solutions under the appli-
cation of several degradation treatments. Some preliminary
studies aimed at assessing the stability of atenolol against
phototransformation and hydrolysis in aqueous medium.
Solutions of 0.15 mmol l1 atenolol in 0.05 mol l1 Na2SO4 at
pH 3.0 were maintained under constant agitation at atmo-
spheric conditions for several hours, with or without the
incidence of indoor lighting. The absorption spectrum of the
b-blocker solution was recorded in the range 200–500 nm at
regular intervals using a spectrophotometer (not shown). The
experiments revealed the stability of the acidic aqueous
solutions of atenolol because the decrease of the absorbance
was insignificant and so, an open electrolytic cell could be
used for the application of EAOPs.
The AO process was the first electrochemical treatment
applied. A 0.15 mmol l1 (¼25 mg l1 TOC) acidic atenolol
solution was electrolyzed using a large surface area carbon-
felt cathode combined with different anodes. Since O2 satu-
ration was ensured by bubbling compressed air, H2O2 was
produced continuously; sometimes, this process is referred to
as AO-H2O2 (Guinea et al., 2008). However, the direct effect of
H2O2 on atenolol can be neglected, as confirmed from
preliminary tests in a divided cell (not shown); this agrees
with studies on other organic pollutants (Sirés et al., 2007a).
Fig. 2 shows the decay of the normalized TOC (i.e., TOCt/TOC0)
over electrolysis time. The solutions remained colorless along
all the electrolyses. The use of a Pt anode led to a very slow and
poor mineralization (curve a), with only 15% TOC removal at
360 min, whereupon the TOC remained almost unchanged.
The influence of the anode material can be seen in curve b; the
use of a BDD anode of the same size produced a remarkable
acceleration of the TOC removal and, at long term (>600 min),
almost total mineralization (>95%) could be reached. Such
enhancement is consistent with other studies reporting the
different oxidation power of both anodes (Skoumal et al.,
2009). Indeed, a high oxidation power anode such as BDD
has a much greater ability to produce active radicals (i.e.,
loosely adsorbed BDD(OH)) compared to a low oxidation
power anode such as Pt (Panizza and Cerisola, 2009; Martı́nez-
 water research 44 (2010) 3109–3120 3113

1.2 using the 25 cm2 BDD at 300 mA, and at 120 min using the
4.5 cm2 BDD at 60 mA. The contact with the initial pollutant
1 was then increasingly favored in the order: small BDD < large
a BDD < Pt. Note that the quicker oxidation kinetics of the initial
0.8 pollutant shown by Pt compared to BDD has also been
0
TOC / TOC

reported elsewhere for other organic compounds, and it arises

0.6 from the stronger adsorption properties of Pt (Brillas et al.,
t

2009). As stated, AO with a large BDD allows the fast elimi-

0.4 nation of atenolol; but, BDD is an expensive material and it is
b
interesting to study more cost-affordable processes. There-
0.2 fore, all the results discussed below were obtained by the EF
c
d process with a 4.5 cm2 Pt (i.e., Pt/carbon felt cell), in
0 0.05 mol l1 Na2SO4 medium with 0.2 mmol l1 Fe2þ (catalyst),
0 120 240 360 480 600
at pH 3.0 at room temperature.
Time / min
In all the EF treatments, the colorless atenolol solutions
Fig. 2 – Decay of the normalized TOC vs electrolysis time became pale yellow in the presence of iron, which suggests
during the degradation of 0.15 mmol lL1 atenolol the complexation of atenolol; however, such phenomenon did
(25 mg lL1 TOC) in 0.05 mol lL1 Na2SO4 medium, at pH 3.0 not prevent the degradation to take place, and the solutions
at room temperature, by anodic oxidation with a large became colorless over electrolysis time due to the disappear-
surface area carbon-felt cathode and a (a) 4.5 cm2 Pt, (b,c) ance of atenolol. Fig. 3a shows the effect of the applied current
4.5 cm2 BDD, or (d ) 25 cm2 BDD anode. Current: (a,b) 60, and and the atenolol concentration on the mineralization trend.
(c,d ) 300 mA.

1.2
Huitle and Quiroz-Alfaro, 2008). The effect of the applied
a 60

-1
(TOC)exp / mg l
1 50
current can be seen in curve c; when 300 mA were applied
40
instead of 60 mA using the same BDD anode, the mineraliza-
0.8 30
0

tion rate was much higher, with 50% TOC removal in 150 min
TOC / TOC

20
and almost total mineralization after 480 min of treatment. 10
0.6
This can be explained by the greater amount of BDD(OH) 0
t

0 120 240 360 480 600 720

produced from the anodic oxidation of water, which can Time / min
0.4
oxidize more quickly both, atenolol and its intermediates.
Finally, the effect of the anode surface area can be seen in
0.2
curve d; the use of a larger anode with 25 cm2 apparent surface
area instead of 4.5 cm2 used in b and c led to the fastest 0
process, with almost overall mineralization in about 300 mA,
35
b
even though the anodic current density (¼12 mA cm2) was
very similar to that in b (ca. 13 mA cm2) and less than one 30
fifth of that used in c (ca. 67 mA cm2). The remarkable
25
MCE / %

difference between the trends of curves c and d is due to the

fact that the AO treatment is an electrode process, which 20
means that the amount of area that is exposed to the organic
15
molecules plays a very important role because it determines
the amount of molecules that interact with the surface (e.g., 10
adsorption). In the case of a small electrode (curve c), only
5
a certain amount of atenolol molecules and oxidation inter-
mediates can interact with the surface; thus, there is a frac- 0
0 120 240 360 480 600 720
tion of BDD(OH) wasted in parasitic reactions that lead to
Time / min
their destruction (i.e., self-dimerization and reaction with H2O
or with H2O2). In contrast, a larger electrode area (curve d ) Fig. 3 – (a) Effect of the applied current and the atenolol
favors the contact between BDD(OH) and the organic mole- concentration on the decay of the normalized TOC vs
cules, which enhances the mineralization process; it also electrolysis time for the mineralization of atenolol in a cell
reduces the interaction between neighbor BDD(OH), and with a 4.5 cm2 Pt anode by EF process in 0.05 mol lL1 Na2SO4
hence the extent of their parasitic reactions, thus increasing medium with 0.2 mmol lL1 Fe2D, at pH 3.0 at room
the current efficiency. temperature. Current: (B) 30, (,) 60, (O) 120, and (>,A)
The importance of the interaction step in the AO process 300 mA; atenolol concentration: (B,,,O,>) 0.15 (25 mg lL1
was confirmed for the initial pollutant by means of reversed- TOC), and (A) 0.3 mmol lL1 (50 mg lL1 TOC). The inset shows
phase HPLC analysis for the previous treatments (not the evolution of D(TOC)exp for the same experiments.
shown). A well defined peak of atenolol appeared at 16.8 min. (b) Time course of the mineralization current efficiency
Atenolol disappeared at 60 min using Pt at 60 mA, at 80 min corresponding to the treatments shown in plot (a).
3114 water research 44 (2010) 3109–3120

A 0.15 mmol l1 acidic atenolol solution was electrolyzed at 30,
60, 120, and 300 mA and almost total TOC removal could be
achieved in all cases. The EF process is then much more
effective than AO with Pt. EF allows the Fenton’s reaction
between H2O2 and Fe2þ, which yields a high amount of OH in
the bulk; the fact that the oxidizing species are not confined to
the electrode surface or its vicinity allows the easier interac-
tion between the organic molecules and OH within the whole
bulk solution. Therefore, the EF process allows the significant
reduction of the mass transport limitations inherent to the
electrode processes at a given current value. As observed,
a higher current led to a faster mineralization. For example,
more than 720 min were required at 30 mA for almost total
mineralization, whereas only 360 min were needed at 300 mA.
The faster electrogeneration of H2O2 and regeneration of Fe2þ
at a higher current accounted for such acceleration. But, the
enhancement obtained when increasing from 120 to 300 mA is
not as remarkable as that produced when increasing from 60
to 120 mA, which suggests a greater limitation by mass
transport at high applied current (i.e., the oxidation rate
becomes mainly determined by the transport of oxygen and
iron ions toward the cathode). Also, note that the oxidation
mineralization (Flox et al., 2007). At 300 mA, taking into
account that the cell voltage was 3.3 V, 27 and 36 kWh m3 were
consumed for the EF treatment of 0.15 and 0.3 mmol l1 atenolol,
respectively. On the other hand, the energy consumption per
unit TOC mass was 1.08 and 0.72 kWh (g TOC)1, respectively;
the lower energy requirement when treating 0.3 mmol l1 agrees
with the higher amount of TOC removed, as shown in the inset
of Fig. 3a. Note that the use of a flow pilot plant can reduce the
energy consumption because of the enhanced mass transport.
The time course of the atenolol concentration was moni-
tored by reversed-phase HPLC during the previous EF treat-
ments of 0.15 mmol l1 atenolol (¼39.6 mg l1) at different
applied current. Fig. 4a shows the decay of the normalized
atenolol concentration (i.e., ct/c0) over electrolysis time. A fast
and complete decay was observed in all cases, although
a shorter time was needed for total destruction of the initial
pollutant at a higher applied current, as expected from the
1.2
a 3.5
3
1
2.5

ln (c / c )
t
ability of the EF process at 300 mA was even higher than that 2

0
0.8 1.5
of AO with the 4.5 cm2 BDD anode shown in Fig. 2. The effect of 1
0

the initial b-blocker concentration was studied by compara- 0.5

c /c

0.6 0
tively treating a 0.3 mmol l1 atenolol (¼50 mg l1 TOC) solu-
t

0 2 4 6 8 10 12 14 16
tion at 300 mA. In Fig. 3a, a similar mineralization rate (i.e., Time / min
0.4
percentage of TOC removal) can be observed at both concen-
trations. This is an interesting feature because it means that 0.2
the EF process can proceed when treating a high organic
content as well; at a higher pollutant concentration, a smaller 0
fraction of OH is wasted in parasitic reactions because the 0 5 10 15 20 25
oxidizing species interact with a larger number of organic Time / min
molecules that are available. This is shown in the inset, with 1.2
a larger amount of TOC being removed when treating b 3
0.3 mmol l1 atenolol. In the case of the EF process, the 1 2.5
ln (c / c )

2
t

parasitic reactions involved are those mentioned for the

AO, as well as the reaction between OH and Fe2þ, with 1.5
0

0.8
1
k ¼ 3.2  108 l mol1 s1 (Sun and Pignatello, 1993). As a result,
0.5
0
c /c

a more efficient process was obtained when treating a higher 0.6

0
t

amount of organic matter, as depicted in Fig. 3b. The MCE for 0 5 10 15 20 25

Time / min
the treatment of 0.15 mmol l1 atenolol at 300 mA was lower 0.4
than that for 0.3 mmol l1 during all the electrolysis time. Also,
a lower applied current led to a higher MCE at any time, which 0.2
means that the OH were used more effectively in the degra-
dation of the organic molecules, even though the OH 0
0 5 10 15 20 25 30 35
concentration was lower. In all cases, an MCE < 100% was
Time / min
observed; it must be reminded that OH is not directly gener-
ated in the medium by applying the current to the electrodes Fig. 4 – (a) Decay of the normalized atenolol concentration
as in the case of the AO process, but it requires mass transport vs electrolysis time for the EF treatments of 0.15 mmol lL1
of reactants, thus affecting the efficiency of the mineraliza- atenolol at different current values shown in Fig. 3a.
tion. In addition, H2 evolution in acidic medium from water (b) Decay of the normalized concentrations of (C,-)
reduction contributes to such current waste, which is espe- atenolol and (B,,) p-hydroxybenzoic acid during the EF
cially relevant at high current density. The MCE always treatments of solutions containing 0.15 mmol lL1 of each
underwent a progressive decay over time due to two organic compound, 0.05 mol lL1 Na2SO4 and 0.2 mmol lL1
phenomena (Sirés et al., 2006; Oturan et al., 2008): (i) the Fe2D, at pH 3.0, room temperature, and current of (C,B) 30,
increasing mass transport limitations caused by the decay of and (-,,) 60 mA. The corresponding kinetic analyses
the organic matter concentration, and (ii) the formation of assuming that p-hydroxybenzoic acid and/or atenolol are
more refractory by-products. The energy consumption per destroyed following a pseudo-first order reaction with OH
unit volume could then be calculated for the total are given in the inset panels.
 water research 44 (2010) 3109–3120
higher OH production rate from the Fenton’s reaction. Thus,
atenolol disappeared in 24, 17, 8, and 6 min working at 30, 60,
120, and 300 mA, respectively. The corresponding kinetic
analysis is shown in the inset of Fig. 4a. The decay of atenolol
was in good agreement with pseudo-first order reaction
between the b-blocker and OH, with excellent linear correla-
tions (R2 > 0.99). This suggests that a constant OH concen-
tration reacts with atenolol in the bulk. From this analysis,
gradually increasing apparent rate constant (kapp) values of
0.17, 0.24, 0.45, and 0.57 min1 were obtained with the current
rising from 30 to 300 mA. It must also be noted that the rate
constants did not increase proportionally, which confirms the
progressive enhancement of the parasitic reactions; it is then
expected that the destruction rate will reach a maximum once
a certain current value is applied.
The second-order or absolute rate constant for the reaction
between atenolol and OH (kabs,ATE) was determined by the
competition kinetics method, with p-hydroxybenzoic acid as
the standard competition substrate whose absolute rate
constant is well known (¼2.19  109 l mol1 s1) (Beltran De
Heredia et al., 2001). Solutions containing equimolar concen-
trations of atenolol and p-hydroxybenzoic acid were electro-
lyzed at 30 and 60 mA for a short time, and reversed-phase HPLC
analyses showed the peaks at 16.8 and 29.5 min, respectively.
Both compounds always decreased exponentially, as illustrated
in Fig. 4b, in agreement with pseudo-first order kinetics. It was
possible to simultaneously determine their kapp (R2 > 0.99) from
the slopes in the inset of Fig. 4b: 0.052 and 0.083 min1 at 30 mA,
and 0.12 and 0.18 min1 at 60 mA for atenolol and p-hydrox-
ybenzoic acid, respectively. From these data and using the
appropriate relationship (Sirés et al., 2008), kabs,ATE values of
1.37  109 and 1.46  109 l mol1 s1 were obtained at 30 and
60 mA, respectively, which yielded a mean value of
1.42  109 l mol1 s1. This value is slightly lower than
(7.05  0.27)  109 l mol1 s1 reported from radiolysis experi-
ments (Song et al., 2008); however, it lies within the range 108–
1010 l mol1 s1 usually found for the reaction between aromatic
compounds and OH (Buxton et al., 1988). A steady state OH
concentration value of 2.1  1012 and 2.7  1012 mol l1 was
calculated for the EF electrolyses at 30 and 60 mA respectively,
which are similar to values reported by ourselves (Oturan
et al., 2009).
3.2. Time course of the oxidation by-products during the
degradation of atenolol by electro-Fenton
The previous mineralization and kinetic studies have been
accompanied by the identification of aromatic intermediates,
ultimate carboxylic acids, and inorganic ions. Since OH is
a non-selective oxidizing species, various by-products were
formed at low concentration levels. b-Blockers contain two
kind of reactive sites, i.e., the aromatic ring and the side chains.
Fig. 5 shows the time course of the main by-products accu-
mulated at different applied current. It can be seen that the by-
products are formed since the very beginning of the electro-
lyses in concomitance with the disappearance of the primary
compound. A general oxidation pathway for the complete
mineralization of atenolol by the EF process is also proposed.
From reversed-phase HPLC measurements with a methanol/
water (1% H3PO4) 8:92 (v/v) mixture as the mobile phase, it has
3115
been possible to detect three important aromatic by-products
that exhibited peaks at 5.2, 11.1, and 13.1 min. The absorption
spectrum of these peaks was very similar with a lmax ¼ 220 nm,
and it was not very different from atenolol’s spectrum. First,
their identity was assessed from comparison with the chro-
matograms and UV spectra of available standards. Neither p-
benzoquinone nor hydroquinone, typically formed upon
degradation of aromatic pollutants, corresponded to these
peaks. Also p-hydroxyphenylacetic acid was discarded because
it showed a longer retention time. The only remaining available
standard with a similar structure to that of atenolol was
p-hydroxyphenylacetamide; indeed, the peak at 11.1 min was
related to this compound. The structures of peaks at 5.2 (called
peak B) and 13.1 min (called peak A) have been hypothesized
since no analytical standards were available. From the fact that
the absorption spectrum of atenolol, p-hydroxyphenyl-
acetamide, and p-hydroxyphenylacetic acid is very similar, it
can be inferred that the 4-oxyphenylacetyl structure is respon-
sible for such absorption. Thus, peaks A and B, which show the
same absorption pattern, can be related to intermediates that
exhibit such structure but having undergone hydroxylation on
the benzenic ring or the side chain. This kind of hydroxylation
confers a higher polarity to the compounds, thus leading to
a shorter retention time. More precisely, compounds A and B
can be expected to maintain the 4-oxyphenylacetamide struc-
ture from atenolol because the substitution of the –NH2 group by
an –OH group causes a much slower elution, as observed for the
p-hydroxyphenylacetic acid compared to the p-hydroxy-
phenylacetamide. As a result, compound A was ascribed to
a hydroxylated atenolol (i.e., hydroxylation on the side
chain), whereas compound B was related to a hydroxylated
p-hydroxyphenylacetamide, as shown in Fig. 5. These
compounds were more polar and eluted more quickly than
atenolol and p-hydroxyphenylacetamide, respectively. Alter-
natively, a different isomer could be proposed for the hydroxy-
lated atenolol, with OH attack on the benzenic ring to form
a resonance-stabilized carbon-centered radical that finally
yields hydroxylated atenolol (Song et al., 2008). If its existence
was accepted, it could directly yield the hydroxylated
p-hydroxyphenylacetamide upon OH attack.
As a reasonable approach, the structural similarity between
compounds A and B and their corresponding non-hydroxylated
primary compounds was thought to yield close molar extinction
coefficients; that is to say, the calibration curves of atenolol and
p-hydroxyphenylacetamide were also used for compounds
A and B, respectively, in order to quantify them and provide
their time course, which is presented in Fig. 5a–c. The three
aromatic by-products exhibited similar features: their concen-
tration increased during the first minutes, until it reached
a maximum and then fell until the total destruction of the
compounds. A higher current led to a faster cycle of
accumulation-destruction due to the action of a greater amount
of OH. Their disappearance in less than 20 min was in good
agreement with the time required for the destruction of atenolol
shown in Fig. 4, and hence no aromatic compounds remained in
the solution after a few minutes of the EF treatment.
As observed, the maximum concentration of aromatic by-
products was always low. The equivalent carbon content,
plus that corresponding to the existing atenolol, was far from
justifying the total TOC during the first minutes. This means
3116 water research 44 (2010) 3109–3120

NH2 NH2
NH •OH NH OH
O O
O O
OH OH

ATENOLOL 0.0025 Hydroxylated atenolol

a
0.002
•OH

-1
c / mmol l
0.0015
- NH4+
0.001

A
0.0005

0
0 5 10 15 20
Time / min
•OH

- NH4+
NH2 NH2
•OH
O O
HO OH
HO

Hydroxylated p-hydroxyphenylacetamide p-Hydroxyphenylacetamide

0.005
-1

0.005
/ mmol l

c b
0.004 0.004
-1
c / mmol l

p-hydroxyphenylacetamide

0.003 0.003

0.002 0.002
B

0.001 0.001
c

0 0
0 5 10 15 20 25 0 5 10 15 20
Time / min Time / min

• OH

- NH4+

COOH
COOH
Oxalic acid
0.25
d
/ mmol l-1

0.2

0.15
•OH
CO2
oxalic acid

0.1

0.05
c

0
0 120 240 360 480 600
Time / min

Fig. 5 – General reaction scheme proposed for the complete mineralization of acidic aqueous solutions of atenolol by EF
process using a Pt/carbon felt cell. The time course of the concentration of the main aromatic and short-chain carboxylic
acid intermediates accumulated during the treatment of 0.15 mmol lL1 atenolol in 0.05 mol lL1 N2SO4 with 0.2 mmol lL1
Fe2D, at pH 3.0, room temperature, and current of (B) 30, (,) 60, (O) 120, and (>) 300 mA is also shown.

that other intermediates were formed. Ion-exclusion chro-
matograms showed the presence of oxalic acid (retention
time ¼ 11.8 min) from the first stages of the electrolyses at all
current values studied. A cycle of accumulation-destruction
can be seen as well in Fig. 5d; but, the maximum concentra-
tion of oxalic acid was ca. 100 times higher compared to the
aromatics, which can be explained by the fast transformation
of the primary aromatics into short-chain aliphatic carboxylic
acids, and a much longer time was required to achieve the
total destruction. It is then confirmed that more refractory by-
products are formed over electrolysis time, as mentioned
before. At 300 mA, oxalic acid disappeared at 360 min, which is
the time needed for almost total TOC removal (Fig. 3a). At
30 mA, the slow degradation of oxalic acid accounted for the
much lower mineralization rate (Fig. 3a). In fact, the absolute
rate constant for oxalic acid is 1.6  106 l mol1 s1 (Buxton
et al., 1988), which is smaller than that of the aromatics.
The possible loss of the initial nitrogen content of atenolol
in the form of inorganic ions such as NHþ 4 , NO2, and NO3
was investigated during its mineralization. No nitrite ions
were detected. Fig. 6 shows the evolution of the main nitro-
genated inorganic ions, namely NO3 (retention
time ¼ 3.1 min) and NHþ 4 (retention time ¼ 5.4 min), for the
treatment of atenolol solutions at 300 mA. The atenolol
concentration was 0.3 mmol l1 in order to help the identifi-
cation of the ions. The NHþ 4 ions were predominant and
reaction (2) could then be proposed. A rapid accumulation of
NHþ 4 was observed during the early stages of the treatment,
which agrees with the fast destruction of the aromatic
compounds. The percentage of NHþ 4 already was w 100% (i.e.,
the concentration was very close to 0.6 mmol l1, which was
the maximum quantity that could be formed) at 360 min. The
NO3 concentration was always very low accordingly,
becoming almost negligible after 180 min. The same was
found at 60 mA (not shown). On the other hand, solutions with
0.6 mmol l1 sodium nitrate were treated (not shown) to
clarify if NHþ4 is formed from NO3 reduction or directly from
the release of N-containing groups. At 60 mA, NO3 was
completely stable, and so the NHþ 4 observed during the
0.7
-1
0.6
/ mmol l
0.5
nitrogenated inorganic ions
0.4
0.3
0.2
c
0.1
0 very large, especially at 300 mA, which again can be explained
0 120 240
Time / min
Fig. 6 – Time course of the nitrogenated inorganic ions (B)
NHD4 and (C) NO3L released upon EF degradation of
0.3 mmol lL1 atenolol solutions in 0.05 mol lL1 N2SO4 with
0.2 mmol lL1 Fe2D, at pH 3.0, room temperature, and
300 mA using a Pt/carbon felt cell.
water research 44 (2010) 3109–3120 3117
degradation of atenolol at 60 mA was generated from the
release of N-containing groups; at 300 mA, the NO3 from
sodium nitrate could be progressively reduced to NHþ 4 . It is
then confirmed that the primary and secondary amines of
atenolol are released directly as NHþ 4 , with some NO3 only
appearing at high current and being quickly reduced to NHþ 4.
Similar results were found for the electrochemical treatment
of paracetamol, whose structure presents a secondary amine,
by different EAOPs: 93% of the nitrogen content ended up as
NHþ 4 (Sirés et al., 2006).
The oxidation pathway shown in Fig. 5 involves the
primary attack of OH on atenolol to form a hydroxylated
atenolol (compound A). Further attack on this compound
leads to p-hydroxyphenylacetamide with release of NHþ 4 . This
aromatic by-product can also be formed from the hydroxyl-
ation of atenolol, and generates a hydroxylated p-hydrox-
yphenylacetamide (compound B) upon the attack of OH. Next,
the cleavage of the aromatic intermediates leads to the
formation of oxalic acid as the ultimate carboxylic acid, with
loss of NHþ 4 . Finally, oxalic acid is totally mineralized to CO2 at
long term. Note that both, oxalic acid and its complexes with
Fe2þ, can be oxidized by OH, contrary to other systems where
hardly oxidizable Fe3þ complexes are formed (Oturan et al.,
2008). The time course of the intermediates, with a progres-
sive transformation into more biodegradable compounds, can
be related to the decrease of the toxicity of the solutions. Note
that some of the proposed EAOPs could be also employed as
a tertiary treatment aiming at the effluent disinfection
(Martı́nez-Huitle and Brillas, 2008).
3.3. Electro-Fenton treatment of b-blockers mixtures
Multicomponent aqueous solutions containing three b-blockers
including atenolol, metoprolol, and propranolol, each one at
a concentration of 0.15 mmol l1, were electrolyzed at 60 and
300 mA by the EF process with 0.2 mmol l1 Fe2þ using the
previous Pt/carbon felt cell. This is an interesting task because
organic pollutants usually occur simultaneously in waste-
waters, but such studies are rarely reported (Sirés et al., 2008).
Fig. 7 shows the decay of the normalized TOC of the
mixtures over the electrolysis time. The solutions contained
85 mg l1 TOC (as a result of the 25, 31, and 29 mg l1 TOC from
atenolol, metoprolol tartrate, and propranolol hydrochloride,
respectively). The effect of the applied current on TOC
removal was similar to that discussed for atenolol solutions,
with a much greater mineralization rate at a higher current
due to the increasing production rate of Fenton’s reagent.
More than 600 min were required at 60 mA, whereas only
480 min were needed at 300 mA. The time required to achieve
the almost overall TOC removal was longer than that observed
for atenolol alone in Fig. 3a, as expected from the higher
content of organic matter; however, the difference was not
360 480
by the higher efficiency of the EF treatment when OH can
encounter a higher number of organic molecules (Fig. 3b). In
fact, the main difference was obtained during the first
minutes, before a high concentration of refractory carboxylic
acids could be accumulated. At 300 mA, for example, 60% and
50% TOC removal was achieved after 60 min for single and
multicomponent solutions, respectively; at 180 min, 95% TOC
3118 water research 44 (2010) 3109–3120

1.2 100 100

a PROPRANOLOL
(32.9 min)
1

m AU

mAU
0.8 50 50
0

METOPROLOL
TOC / TOC

(23.9 min)
ATENOLOL
0.6 (11.1 min)
t

0.4 0 0
5 10 15 20 25 30 35 40
Minutes
0.2
1.2
0 b 4
0 120 240 360 480 600 1 3.5
3

ln (c / c )
t
Time / min 2.5
2

0
0.8 1.5
Fig. 7 – Effect of the applied current on the decay of the 1

0
0.5

c /c
normalized TOC vs electrolysis time for the mineralization 0.6 0

t
of multicomponent aqueous solutions containing 0 10 20 30 40
Time / min
0.15 mmol lL1 of three b-blockers including atenolol 0.4
(25 mg lL1 TOC), metoprolol (31 mg lL1 TOC as metoprolol
tartrate), and propranolol (29 mg lL1 TOC) in 0.05 mol lL1 0.2
N2SO4 with 0.2 mmol lL1 Fe2D, at pH 3.0, room
0
temperature, and current of (,) 60 and (>) 300 mA, by EF 0 10 20 30 40 50 60 70
treatment using a Pt/carbon felt cell. Time / min

Fig. 8 – (a) HPLC chromatogram for the multicomponent

removal was already reached for the mixture. The energy
consumption per unit volume and per unit TOC mass for the
total mineralization at 300 mA were 36 kWh m3 and
0.42 kWh (g TOC)1, respectively. The latter value is almost
one third of that found for 0.15 mmol l1 atenolol solutions.
The decay of each b-blocker during the treatment of the
mixtures was analyzed by gradient HPLC. The use of such
technique allowed the simultaneous determination of the
three compounds. As shown in Fig. 8a, the three peaks were
eluted showing a good separation between them, from the
most to the least polar; thus, atenolol, metoprolol, and
propranolol appeared at 11.1, 23.9, and 32.9 min, respectively.
Fig. 8b shows the decay of the normalized concentration of
each b-blocker during the EF treatment of Fig. 7 at 60 mA.
Atenolol and metoprolol disappeared at 60 min, whereas
propranolol only required 40 min of treatment. The order of
disappearance was also confirmed at 300 mA (not shown):
atenolol and metoprolol were destroyed in 15 min while only
10 min were required for propranolol. The kinetic analysis for
each b-blocker is illustrated in the inset considering that
pseudo-first order reaction is accomplished; at 60 mA, the kapp
values (R2 > 0.99) were 0.055, 0.080, and 0.13 min1 for ate-
nolol, metoprolol, and propranolol, respectively. The highest
reactivity was then exhibited by propranolol, which was also
found by other authors for the reaction with OH (Song et al.,
2008). Note that the EF treatment of atenolol alone at 60 mA,
shown in Fig. 4a, revealed a higher kapp of 0.24 min1, which is
four times greater. A much quicker disappearance (¼17 min)
was observed accordingly. This makes it evident the compe-
tition effect arisen from the presence of the two other
b-blockers in the mixture, which react more quickly with OH
due to their higher kapp.
Since kabs,ATE has been calculated in this work, atenolol can
be taken as a standard competition substrate in the EF
aqueous solutions mentioned in Fig. 7 before starting the
electrolyses, obtained under the application of gradient
HPLC. (b) Decay of the normalized concentration of (B)
atenolol, (,) metoprolol, and (O) propranolol vs
electrolysis time for the EF treatment at 60 mA shown in
Fig. 7. The inset panel shows the kinetic analysis for the
decay of each b-blocker considering a pseudo-first order
reaction.
treatment (Fig. 8b) in order to determine the absolute rate
constant of metoprolol (kabs,MET) and propranolol (kabs,PROP) as
well. Following the same procedure commented in Section 3.1,
values of 2.07  109 and 3.36  109 l mol1 s1 were calculated
for metoprolol and propranolol, respectively. In all cases, the
values were slightly smaller than those reported by Song and
co-workers for radiolysis experiments; they used a different
method to calculate the rate constants, based on the absorp-
tion change with concentration at the lmax of the transient
hydroxycyclohexadienyl radicals formed, which yielded
(8.39  0.06)  109 and (1.07  0.02)  1010 l mol1 s1 for
metoprolol and propranolol, respectively (Song et al., 2008). In
any case, all the values lie within the range expected for
aromatic compounds (Buxton et al., 1988), and the order of
reactivity agrees perfectly. The difference in structure
accounts for the difference in reactivity: the slight increase for
metoprolol compared to atenolol is consistent with the
weaker electron-withdrawing effect of the CH3OCH2CH2–
lateral chain compared to that of the NH2COCH2–chain; the
highest value, found for propranolol, reflects the much higher
electron density and the larger number of sites prone to
oxidation in the naphthalene group compared to the phenol
groups of atenolol and metoprolol.
 water research 44 (2010) 3109–3120
4. Conclusions
A pioneering study reporting the electrochemical degradation
of aqueous solutions of b-blockers has been presented. Single
component solutions containing atenolol have been treated by
the AO process with Pt and BDD anodes. The greatest miner-
alization ability found for the AO with a large surface area BDD
anode has been accounted for by the enhanced interaction
between very active BDD(OH) and the organic molecules near
the electrode surface. The EF process using a Pt/carbon felt cell
has then been thoroughly studied as a more cost-affordable
alternative. Fast and almost overall TOC removal and
atenolol destruction could be achieved. The time course of the
oxidation by-products has been discussed in order to propose
a reaction pathway up to the total mineralization of atenolol.
Also, multicomponent solutions containing atenolol, meto-
prolol, and propranolol have been treated by the EF treatment.
The mineralization current efficiency, the energy consump-
tion for the complete TOC removal, and the absolute rate
constants for the three b-blockers have been determined. The
order of reactivity of the three pharmaceuticals increased in
the order: atenolol < metoprolol < propranolol.
The time course of the intermediates, with a progressive
transformation into more biodegradable compounds, can be
related to the decrease of the toxicity of the solutions. In
conclusion, EF is proven a very effective detoxification method
for the treatment of pharmaceutical pollutants such as
b-blockers in aqueous effluents.
Acknowledgments
I. Sirés acknowledges support from the Laboratoire Géo-
matériaux et Géologie de l’Ingénieur (G2I) at the Université
Paris-Est Marne-la-Vallée.
references
Andreozzi, R., Marotta, R., Paxéus, N., 2003. Pharmaceuticals in
STP effluents and their solar photodegradation in aquatic
environment. Chemosphere 50 (10), 1319–1330.
Andreozzi, R., Campanella, L., Fraysse, B., Garric, J., Gonnella, A., Lo
Giudice, R., Marotta, R., Pinto, G., Pollio, A., 2004. Effects of
advanced oxidation processes (AOPs) on the toxicity of
a mixture of pharmaceuticals. Water Sci. Technol. 50 (5), 23–28.
Bedner, M., MacCrehan, W.A., 2006. Reactions of the amine-
containing drugs fluoxetine and metoprolol during
chlorination and dechlorination processes used in wastewater
treatment. Chemosphere 65 (11), 2130–2137.
Beltran De Heredia, J., Torregrosa, J., Dominguez, J.R., Peres, J.A.,
2001. Kinetic model for phenolic compound oxidation by
Fenton’s reagent. Chemosphere 45, 85–90.
Bendz, D., Paxéus, N.A., Ginn, T.R., Loge, F.J., 2005. Occurrence
and fate of pharmaceutically active compounds in the
environment, a case study: Höje River in Sweden. J. Hazard.
Mater. 122 (3), 195–204.
Brillas, E., Boye, B., Dieng, M.M., 2003. General and UV-assisted
cathodic Fenton treatments for the mineralization of
herbicide MCPA. J. Electrochem. Soc. 150 (11), E583–E589.
3119
Brillas, E., Sirés, I., Oturan, M.A., 2009. Electro-Fenton process and
related electrochemical technologies based on Fenton’s
reaction chemistry. Chem. Rev. 109 (12), 6570–6631.
Buxton, G.U., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988.
Critical review of rate constants for reactions of hydrated
electrons, hydrogen atoms and hydroxyl radicals (OH/O) in
aqueous solution. J. Phys. Chem. Ref. Data 17 (2), 513–886.
Cañizares, P., Hernández-Ortega, M., Rodrigo, M.A., Barrera-
Dı́az, C.E., Roa-Morales, G., Sáez, C., 2009a. A comparison
between conductive-diamond electrochemical oxidation and
other advanced oxidation processes for the treatment of
synthetic melanoidins. J. Hazard. Mater. 164 (1), 120–125.
Cañizares, P., Paz, R., Sáez, C., Rodrigo, M.A., 2009b. Costs of the
electrochemical oxidation of wastewaters: a comparison with
ozonation and Fenton oxidation processes. J. Environ. Manage.
90, 410–420.
Carlesi Jara, C., Fino, D., Specchia, V., Saracco, G., Spinelli, P., 2007.
Electrochemical removal of antibiotics from wastewaters.
Appl. Catal. B: Environ. 70 (1–4), 479–487.
Cleuvers, M., 2005. Initial risk assessment for three b-blockers found
in the aquatic environment. Chemosphere 59 (2), 199–205.
Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal
care products in the environment: agents of subtle change?
Environ. Health Perspect. 107 (6), 907–938.
Flox, C., Cabot, P.L., Centellas, F., Garrido, J.A., Rodrı́guez, R.M.,
Arias, C., Brillas, E., 2007. Solar photoelectro-Fenton
degradation of cresols using a flow reactor with a boron-doped
diamond anode. Appl. Catal. B: Environ. 75 (1–2), 17–28.
Guinea, E., Arias, C., Cabot, P.L., Garrido, J.A., Rodrı́guez, R.M.,
Centellas, F., Brillas, E., 2008. Mineralization of salicylic acid in
acidic aqueous medium by electrochemical advanced
oxidation processes using platinum and boron-doped
diamond as anode and cathodically generated hydrogen
peroxide. Water Res. 42 (1–2), 499–511.
Guivarch, E., Oturan, N., Oturan, M.A., 2003. Removal of
organophosphorous pesticides from water by electrogenerated
Fenton’s reagent. Environ. Chem. Lett. 1 (3), 165–168.
Hammami, S., Oturan, N., Bellakhal, N., Dachraoui, M., Oturan, M.A.,
2007. Oxidative degradation of direct orange 61 by electro-
Fenton process using a carbon felt electrode: application of
the experimental design methodology. J. Electroanal. Chem.
610, 75–84.
Hirsch, R., Ternes, T.A., Haberer, K., Kratz, K.L., 1996.
Determination of beta-blockers and b-sympathomimetics in
the aquatic environment. Vom Wasser 87, 263–274.
Huggett, D.B., Brooks, B.W., Peterson, B., Foran, C.M., Schlenk, D.,
2002. Toxicity of select beta adrenergic receptor-blocking
pharmaceuticals (B-blockers) on aquatic organisms. Arch.
Environ. Contam. Toxicol. 43 (2), 229–235.
Huggett, D.B., Khan, I.A., Foran, C.M., Schlenk, D., 2003.
Determination of beta-adrenergic receptor blocking
pharmaceuticals in United States wastewater effluent.
Environ. Pollut. 121 (2), 199–205.
Ikehata, K., Naghashkar, N.J., El-Din, M.G., 2006. Degradation of
aqueous pharmaceuticals by ozonation and advanced
oxidation processes: a review. Ozone Sci. Eng. 28 (6), 353–414.
Jackson, C.D., Fishbein, L., 1986. A toxicological review of beta-
adrenergic blockers. Fundam. Appl. Toxicol. 6 (3), 395–422.
Khetan, S.K., Collins, T.J., 2007. Human pharmaceuticals in the
aquatic environment: a challenge to green chemistry. Chem.
Rev. 107 (6), 2319–2364.
Klavarioti, M., Mantzavinos, D., Kassinos, D., 2009. Removal of
residual pharmaceuticals from aqueous systems by advanced
oxidation processes. Environ. Int. 35 (2), 402–417.
Li, S., Bejan, D., McDowell, M.S., Bunce, N.J., 2008. Mixed first and
zero order kinetics in the electrooxidation of
sulfamethoxazole at a boron-doped diamond (BDD) anode.
J. Appl. Electrochem. 38 (2), 151–159.
3120 water research 44 (2010) 3109–3120
Liu, Q.T., Williams, H.E., 2007. Kinetics and degradation products
for direct photolysis of b-blockers in water. Environ. Sci.
Technol. 41 (3), 803–810.
Martı́nez-Huitle, C.A., Brillas, E., 2008. Electrochemical
alternatives for drinking water disinfection. Angew. Chem.
Int. Ed. 47 (11), 1998–2005.
Martı́nez-Huitle, C.A., Quiroz-Alfaro, M., 2008. Recent
environmental applications of diamond electrode: critical
review. J. Environ. Eng. Manage. 18 (3), 155–172.
Maurer, M., Escher, B.I., Richle, P., Schaffner, C., Alder, A.C., 2007.
Elimination of b-blockers in sewage treatment plants. Water
Res. 41 (7), 1614–1622.
Oturan, M.A., Pimentel, M., Oturan, N., Sirés, I., 2008. Reaction
sequence for the mineralization of the short-chain carboxylic
acids usually formed upon cleavage of aromatics during
electrochemical Fenton treatment. Electrochim. Acta 54 (2),
173–182.
Oturan, N., Sirés, I., Oturan, M.A., Brillas, E., 2009. Degradation of
pesticides in aqueous medium by electro-Fenton and related
methods. A review. J. Environ. Eng. Manage. 19 (5), 235–255.
Panizza, M., Cerisola, G., 2001. Removal of organic pollutants from
industrial wastewater by electrogenerated Fenton’s reagent.
Water Res. 35 (16), 3987–3992.
Panizza, M., Cerisola, G., 2009. Direct and mediated anodic
oxidation of organic pollutants. Chem. Rev. 109 (12), 6541–6569.
Qiang, Z., Chang, J.-H., Huang, C.-P., 2003. Electrochemical
regeneration of Fe2þ in Fenton oxidation processes. Water Res.
37 (6), 1308–1319.
Sirés, I., Garrido, J.A., Rodrı́guez, R.M., Cabot, P.L., Centellas, F.,
Arias, C., Brillas, E., 2006. Electrochemical degradation of
paracetamol from water by catalytic action of Fe2þ, Cu2þ, and
UVA light on electrogenerated hydrogen peroxide. J.
Electrochem. Soc. 153, D1–D9.
Sirés, I., Oturan, N., Oturan, M.A., Rodrı́guez, R.M., Garrido, J.A.,
Brillas, E., 2007a. Electro-Fenton degradation of antimicrobials
triclosan and triclocarban. Electrochim. Acta 52 (17), 5493–5503.
Sirés, I., Arias, C., Cabot, P.L., Centellas, F., Garrido, J.A.,
Rodrı́guez, R.M., Brillas, E., 2007b. Degradation of clofibric acid
in acidic aqueous medium by electro-Fenton and
photoelectro-Fenton. Chemosphere 66 (9), 1660–1669.
Sirés, I., Guivarch, E., Oturan, N., Oturan, M.A., 2008. Efficient
removal of triphenylmethane dyes from aqueous medium by
in situ electrogenerated Fenton’s reagent at carbon-felt
cathode. Chemosphere 72 (4), 592–600.
Skoumal, M., Rodrı́guez, R.M., Cabot, P.L., Centellas, F., Garrido, J.A.,
Arias, C., Brillas, E., 2009. Electro-Fenton, UVA photoelectro-
Fenton and solar photoelectro-Fenton degradation of the
drug ibuprofen in acid aqueous medium using platinum
and boron-doped diamond anodes. Electrochim. Acta 54 (7),
2077–2085.
Slegers, C., Tilquin, B., 2006. Final product analysis in the e-beam
and gamma radiolysis of aqueous solutions of metoprolol
tartrate. Rad. Phys. Chem. 75 (9), 1006–1017.
Song, W., Cooper, W.J., Mezyk, S.P., Greaves, J., Peake, B.M., 2008.
Free radical destruction of b-blockers in aqueous solution.
Environ. Sci. Technol. 42 (4), 1256–1261.
Sun, Y., Pignatello, J.J., 1993. Photochemical reactions involved in
the total mineralization of 2,4-D by Fe3þ/H202/UV. Environ. Sci.
Technol. 27 (2), 304–310.
Ternes, T.A., Stüber, J., Herrmann, N., McDowell, D., Ried, A.,
Kampmann, M., Teiser, B., 2003. Ozonation: a tool for removal
of pharmaceuticals, contrast media and musk fragranzes
from wastewater? Water Res. 37 (8), 1976–1982.
Vieno, N., Tuhkanen, T., Kronberg, L., 2007. Elimination of
pharmaceuticals in sewage treatment plants in Finland. Water
Res. 41 (5), 1001–1012.