Environmental Technology

ISSN: 0959-3330 (Print) 1479-487X (Online) Journal homepage: http://www.tandfonline.com/loi/tent20

Evaluation of the simultaneous removal of

recalcitrant drugs (bezafibrate, gemfibrozil,
indomethacin and sulfamethoxazole) and
biodegradable organic matter from synthetic
wastewater by electro-oxidation coupled with a
biological system

Odín Rodríguez-Nava, Hugo Ramírez-Saad, Octavio Loera & Ignacio González

To cite this article: Odín Rodríguez-Nava, Hugo Ramírez-Saad, Octavio Loera & Ignacio
González (2016) Evaluation of the simultaneous removal of recalcitrant drugs (bezafibrate,
gemfibrozil, indomethacin and sulfamethoxazole) and biodegradable organic matter from
synthetic wastewater by electro-oxidation coupled with a biological system, Environmental
Technology, 37:23, 2964-2974, DOI: 10.1080/09593330.2016.1172669

To link to this article: http://dx.doi.org/10.1080/09593330.2016.1172669

Accepted author version posted online: 04

Apr 2016.
Published online: 28 Apr 2016.

Submit your article to this journal

Article views: 51

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tent20

Download by: [Universidad Autonoma Metropolitana] Date: 21 September 2016, At: 14:14
ENVIRONMENTAL TECHNOLOGY, 2016
VOL. 37, NO. 23, 2964–2974
http://dx.doi.org/10.1080/09593330.2016.1172669

Evaluation of the simultaneous removal of recalcitrant drugs (bezafibrate,

gemfibrozil, indomethacin and sulfamethoxazole) and biodegradable organic
matter from synthetic wastewater by electro-oxidation coupled with a biological
system
Odín Rodríguez-Navaa,b, Hugo Ramírez-Saadc, Octavio Loerab and Ignacio Gonzálezd
a
Departamento de Ingeniería en Sistemas Ambientales, Escuela Nacional de Ciencias Biológicas, Instituto Politécnico Nacional, Gustavo
A. Madero, México; bDepartamento de Biotecnología, Universidad Autónoma Metropolitana-Iztapalapa, Iztapalapa, México; cDepartamento de
Sistemas Biológicos, Universidad Autónoma Metropolitana – Xochimilco, Coyoacán, México; dDepartamento de Química, Universidad
Autónoma Metropolitana-Iztapalapa, Iztapalapa, México

ABSTRACT ARTICLE HISTORY

Pharmaceutical degradation in conventional wastewater treatment plants (WWTP) represents a Received 17 November 2015
challenge since municipal wastewater and hospital effluents contain pharmaceuticals in low Accepted 23 March 2016
concentrations (recalcitrant and persistent in WWTP) and biodegradable organic matter (BOM) is
KEYWORDS
the main pollutant. This work shows the feasibility of coupling electro-oxidation with a biological Wastewater; pharmaceutical
system for the simultaneous removal of recalcitrant drugs (bezafibrate, gemfibrozil, compounds; BDD oxidation;
indomethacin and sulfamethoxazole (BGIS)) and BOM from wastewater. High removal efficiencies activated sludge; treatment
were attained without affecting the performance of activated sludge. BGIS degradation was train
performed by advanced electrochemical oxidation and the activated sludge process for BOM
degradation in a continuous reactor. The selected electrochemical parameters from
microelectrolysis tests (1.2 L s−1 and 1.56 mA cm−2) were maintained to operate a filter press
laboratory reactor FM01-LC using boron-doped diamond as the anode. The low current density
was chosen in order to remove drugs without decreasing BOM and chlorine concentration
control, so as to avoid bulking formation in the biological process. The wastewater previously
treated by FM01-LC was fed directly (without chemical modification) to the activated sludge
reactor to remove 100% of BGIS and 83% of BOM; conversely, the BGIS contained in wastewater
without electrochemical pre-treatment were persistent in the biological process and promoted
bulking formation.

Introduction thousand times greater than drug concentration). BOM

is commonly removed by activated sludge due to the
Nowadays, pharmaceutical compounds end up in the
low cost of this process, WWTP does not remove drugs
environment and emerge as pollutants because neither
they nor their metabolites are considered in current reg- efficiently and the activated sludge exposed to drugs
ulations.[1,2] Pharmaceutical compounds are present in show different phenomena: Kraigher et al. [6] found
natural waters, drinking water and municipal wastewater the presence of a filamentous microorganism Chloroflexi,
(MWW). The concentrations in wastewater range from meanwhile Al-Ahmad et al. [7] found bacteria inhibition
nanograms to milligrams; therefore, they represent a of activated sludge. These phenomena could affect the
serious problem in natural ecosystems. The main issue sludge performance; thus, it is necessary to consider a
is related to toxicity and DNA damage, as well as oxida- process for removing them. The technology must be
tive stress caused to living organisms.[3,4] They are installed prior to the activated sludge stage to remove
also persistent in wastewater treatment plants (WWTP). drugs from MWW, thus preventing the pharmaceutical
[5] The removal of drugs from MWW is a challenge, compounds from entering the environment without suf-
because MWW is a complex mixture of biodegradable ficient degradation of the BOM.
organic matter (BOM), total suspended solid and Advanced oxidation processes (AOP) constitute a
various dissolved salts, particularly chloride and sulphate technology that produces a strong oxidant OH• radical,
ions. The BOM is the main pollutant in wastewater and in whose presence refractory organic matter is quickly
drugs are present in low concentration (hundred oxidized. Among the AOP that have been used for

CONTACT Ignacio González [email protected] Departamento de Química, Universidad Autónoma Metropolitana-Iztapalapa, Av. San Rafael Atlixco
186, Col. Vicentina, Iztapalapa C.P. 09340, México
© 2016 Informa UK Limited, trading as Taylor & Francis Group
 ENVIRONMENTAL TECHNOLOGY 2965

removing drugs, there are photo-Fenton,[8,9] ozone oxi- Therefore, it is necessary to propose an oxidation pre-
dation,[10–12] photolytic and photocatalytic processes treatment prior to biological reactor treatment, with the
oxidation.[13] However, these processes take place aim of eliminating drugs present in the wastewater
within the bulk liquid non-specifically degrading without affecting the performance of activated sludge
organic matter,[14] and this nonspecific degradation in organic matter degradation. To our knowledge,
increases operating costs of BOM degradation. Light- there is no anodic oxidation system used as pre-treat-
based techniques decrease efficiency because waste- ment to simultaneously remove different drugs con-
water turbidity prevents oxidation.[15] tained in wastewater.
There are different studies where AOP have been used In this work, the following drugs were selected: beza-
as pre-treatment systems to increase biodegradability of bifrate, gemfibrozil,[22–24] indomethacin [25] and sulfa-
the initial recalcitrant pollutant in a specific chemical methoxazole (BGIS),[26] since they are commonly used
solution; the intermediates formed should, therefore, by humans, show recalcitrance and have toxic effects.
be more biodegradable than the target compounds, Anodic oxidation by BDD in the FM01-LC reactor is pro-
while the mineralization yield of organic carbon posed as a pre-treatment to simultaneously degrade
content should remain limited to ensure subsequent bio- drugs in synthetic wastewater (SWW) with minimal trans-
logical treatment and to minimize operational costs.[14– formation of BOM and produce chlorine at low concen-
16] In contrast, there are very few studies where electro- tration for an efficient performance of activated sludge
oxidation processes have been used as a pre-treatment in a continuous reactor. The SWW containing BGIS, pre-
system to increase the biodegradability of the pharma- viously treated in this reactor, was fed to activated
ceutical compounds.[17,18] The oxidation process is gen- sludge in a continuous reactor in order to remove
erally carried out with a single organic compound in a drugs and BOM contained in SWW.
specific electrolyte (i.e. sulphate aqueous solution) and
after, nutrients are added to the resulting solution to
Experimental
perform the biological process. Fontmorin et al. [17]
studied a coupled process for the removal of the chlori- All solutions were prepared with United States pharma-
nated pesticide 2,4-dichlorophenoxyacetic acid (500 mg copoeia grade reagents and deionized water with 18
L−1) in 0.1 M Na2SO4 solution and the electrolysed mΩ−1 cm−1 resistivity from the SymplicityTM system.
medium was enriched with minerals to feed activated SWW was prepared according to the ASTM standard.
sludge. Dissolved organic carbon measurements [27] The SWW has a composition similar to that of real
showed that a higher mineralization was observed in wastewater. The beer included in the preparation of
the oxidized solution during the first two days, as com- the SWW, as suggested by the ASTM standard, contri-
pared to the control biological reactor, which confirms butes to the BOM because of its sugar and protein
the relevance of the combined process. This same content, because these compounds are used as carbon
research group increased degradation of sulfamethazine and energy source by the microorganisms. Contents of
with activated sludge using the electro-Fenton as pre- Cl− and SO− −1
4 ions (490 and 85 mg L , respectively)
treatment system.[18] were relevant since these ions are electrolytes in
Meanwhile, the evaluation of the electro-oxidation by electro-oxidation processes. Meanwhile, the sulphate
boron-doped diamond (BDD) anode to degrade drugs ion promotes OH• formation on the BDD anode, increas-
present in wastewater as a pre-treatment system to the ing the removal of organic compounds, even of those
activated sludge process is interesting for several easily degraded by activated sludge. The SWW was mod-
reasons: first, it is an efficient indirect anodic oxidation ified so that the beer content was ∼580 mg SCOD L−1.
since the OH• radical is physisorbed on the BDD, and Flour was not added to avoid having suspended solids
the indirect oxidation reaction can be selectively con- in the biological reactor feed; NH4Cl and NaHCO3 were
trolled through the potential or current density unlike added so as to have 1 mg L−1 and 1 g L−1, respectively,
other AOPs, such as ozone, photo-Fenton and so on and the pH was adjusted to 7 with 1 M NaOH or HCl.
[19,20]; second, in the presence of chloride ion, the oxi- BGIS were purchased from Sigma-Aldrich. In total, 200
dizing power of the OH• radical decreases and active mg L−1 stock solution of BGIS was prepared to add 1
chlorine is produced as well,[21] modifying the oxidation mL L−1 of SWW to a final concentration of 200 µg L−1
power of the solutions (oxidation reduction potential). of each drug. Bezafibrate, gemfibrozil and indomethacin
[22] However, the continuously produced active chlorine were simultaneously dissolved in 10 mL of methanol to
could affect the biochemical system, since in the control prepare stock solution; sulfamethoxazole was previously
of pathogens in water, the chlorine concentration range dissolved in water and both solutions were mixed and
should be 0.05—0.1 mg L−1. 100 µL 30% NH4OH was added to remove the precipitate
2966 O. RODRÍGUEZ-NAVA ET AL.

formed due to change of dielectric properties of the new calomel electrode (SCE) (RadiometerXR110) and the
solution, and finally the homogeneous solution was counter electrode was a graphite rod. The BDD electrode
gauged with water to 100 mL. This solution was pre- was regenerated with 5 ppm H2SO4 during 20 min and j
served at 4°C. Soluble Chemical Oxygen Demand 20 mA cm−2. The potential pulses were applied in SWW
(SCOD) of the SWW was 592 ± 78 mg L−1 and the SCOD with and without BGIS; pulses of 0.5 V ≤ E ≤ 3 V were
slightly increased when the BGIS were added at 663 ± applied with respect to the open circuit potential (OCP)
72 due to methanol addition. for 40 s, and a rotating cylinder was employed at 100
The analysis of BGIS in SWW was performed with a and 1300 rpm in order to evaluate the effect of turbu-
previous solid phase extraction according to the EPA lence on the oxidation process. The sampled current
1694 method [28] and quantified by the capillary zone density j versus (E-OCP) curves were constructed for
electrophoresis technique with 30 mM phosphate current sampling at different pulse times from current
buffer, pH 8. transients previously described.[30]
The performance of the activated sludge was deter-
mined through the measurement of efficiency removal
of SCOD, mixed liquor volatile suspended solids Electrochemical reactor operation
(MLVSS) and sludge volumetric index (SVI). The SCOD
was used as a unique parameter to evaluate organic The macroelectrolysis experiments were performed in an
matter consumption due to the negligible decrease of FM01-LC reactor whose dimensions and characteristics
this parameter in the electrochemical reactor, which indi- were described previously.[31] The recirculation system
cates that there was no change in the oxidation state of is schematically represented in Figure 1, along with
the organic matter. MLSS is usually used to calculate SVI; spacer geometry showing the inlet/outlet manifolds.
however, in this work, the organic suspended solid frac- The channel of FM01-LC reactor is 0.16 m long, 0.04 m
tion (MLVSS) has been employed as an indicator of the wide and 0.0195 m thick (gaskets included). The inlet/
state of the sludge. SVI was calculated using the follow- outlet manifold comprises five rectangular ducts of
ing equation: 0.0038 m × 0.0166 m. The reactor was installed in a recir-
culation system formed by a 4 L reservoir made of poly-
SVI = SV30/MLVSS, (1) carbonate, 0.25 HP centrifugal pump, rotameter-type
where SVI is the volume in mL occupied by 1 g of acti- flow meters of 0–20 L min−1, F-400 acrylic flow meter
vated sludge after settling the aerated liquor for 30 of 0.1–20 L min−1. Tubes, valves and accessories were
min (mL gVSS−1), SV30 is the sludge volume after settling made of PVC. The electrodes were connected to a BK
of the activated sludge for 30 min (mL L−1) and MLVSS is
the volatile suspended solids (g L−1)
SCOD, MLVSS and SVI were measured according to
the APHA methods.[29] Due to the high concentration
of the chloride ion, the corresponding SCOD evaluation
method was used (the AgNO3 concentration was
modified).
A Bio-Logic potentiostat/galvanostat, VMP3 model
and EC-Lab v. 10.18TM 20 were used for voltammetric
experiments and 20 A booster was used for FM01-LC
operation. For BGIS quantification, a capillary electro-
phoresis MDQTM (Beckman) was used at 18 kV and an
injection time of 5 s at 0.5 psi in a fused silica capillary
of 50 µm ID and 60 cm length. The Hach DR 5000 Spec-
trophotometer and Thermo Scientific Orion COD165
Thermoreactor were used for COD determination.

Microelectrolysis experiments
A 100 mL pyrex cell with a typical three-electrode config-
uration was used for microelectrolysis experiments. The
BDD anode (1 cm2) was used as the working electrode. Figure 1. The setup used for BGIS electro-oxidation in SWW; the
The potentials were measured versus the saturated elements are indicated in the figure.
 ENVIRONMENTAL TECHNOLOGY 2967

PRECISION power supply, model 1635.The working elec- for an efficient degradation of pharmaceutical com-
trode was made of BDD and stainless steel was used as a pounds (BGIS) and BOM in wastewater, the next strategy
counter electrode The reactor operating conditions were was followed: first, microelectrolysis studies established
the following: 1.2 L min−1 and SWW with BGIS was used the current density conditions for the oxidation process
as electrolyte at 25°C and pH 7, whereas current density in the SWW solutions. After that, an FM01-LC electroche-
was chosen from j versus (E-OCP) curve to get OH• radi- mical reactor was used to perform the degradation of
cals. During these experiments, free chlorine, SCOD and BGIS present in the SWW and chlorine production was
BGIS degradation were monitored. The effluent of this assessed to establish operating conditions. Finally, the
treatment (after 20 min electrolysis) was adjusted to pH continuous biological reactor was operated to determine
7 and fed to the continuous biological reactor. the efficiency of the anodic oxidation as a pre-treatment
for BGIS and BOM degradation in wastewater.
Figure 3 shows typical polarization curves constructed
Biological experiments
from the corresponding current transients at different
To evaluate the effect of the drug and electrochemical sampling times (20 ≤ ts ≤ 30 s) for BDD in SWW, with or
pre-treatment on the performance of activated sludge, without BGIS; the polarization values correspond to
three continuous Eckenfelder-type reactors (1 L) were imposed potential versus OCP. This strategy allowed the
operated without recirculation. The reactors were inocu- identification of phenomena (i.e. adsorption, mass trans-
lated with activated sludge obtained from an oil refin- fer control) occurring during the electrochemical
ery’s WWTP; the sludge was previously conditioned to process at different time constants, which is not possible
the SWW for 65 days in order to have a homogeneous using classical cyclic voltammetry.[30] In all experiments
microbial population. The first reactor was fed with within 0.5–2.2 V intervals, current density did not signifi-
SWW (R1), the second one (R2) was fed with SWW cantly increase, this potential zone corresponds to capaci-
+BGIS and the last reactor (R3) with previously electro- tive behaviour. For polarization more positive than 2.2 V
lysed SWW+BGIS (Figure 2). Aeration (under 1 L min−1), versus SCE, current density showed a sharp increase asso-
beer addition and influent wastewater flow were con- ciated with faradaic processes: water oxidation to form
trolled in continuous reactors to maintain dissolved hydroxyl radicals and oxidize chlorides into active chlor-
oxygen, SCOD and hydraulic retention time (HRT) ine.[33] In microelectrolysis, at 100 rpm of stirring
during experimental operation as shown in Table 1. (Figure 3(a)) the presence of BGIS slightly modifies the
Since differences of SCOD between the SWW and the current density trend of SWW in the potential range of
SWW+BGIS are small (71 mg SCOD L−1), the modification 2.1–2.4 V, but the modification vanishes at 1200 rpm
in SCOD during the reactor operation is roughly related (Figure 3(b)); such a behaviour suggests a process of
to the BOM; therefore, SCOD was evaluated every day. BGIS adsorption at low rpm. The analysis of the current
SVI was evaluated as an indicator of bulking [32] and (I ) transients (results not shown) confirms this assump-
BGIS were monitored in the effluents. tion, since log (I ) versus log (t) shows a slope of (0.77)
for the transients obtained at 2.1–2.4 V, meanwhile
those obtained for more positive potentials show a
Results and discussion slope of −0.5 (typical of diffusion process). The simulta-
In order to analyse the feasibility of the combined pro- neous adsorption of organic molecules and oxidant
cesses (anodic oxidation by BDD and biological reactor) species has been reported for organic compounds

Figure 2. Continuous reactor flow chart of activated sludge to evaluate treatment train of anodic-activated sludge process to remove
drugs from SWW. Three reactors were operated and fed with different influent compositions: R1 with SWW; R2 with SWW+BGIS and R3
with electrochemical pre-treated SWW+BGIS.
2968 O. RODRÍGUEZ-NAVA ET AL.

Table 1. Operating conditions of activated sludge continuous reactors at pH 7.

No. Fed Flow (mL min−1) HRT/SRT (days) DO (mg L−1) SCOD (mg L−1) Organic loading (mg day−1)
R1 SWW 0.5 1.38 2.07 ± 0.6 592 ± 78 426.24
R2 SWW+BGIS 0.53 1.29 2.12 ± 0.6 663 ± 72 506.00
R3 SWW+BGIS electrolysed 0.52 1.33 2.25 ± 0.6 669 ± 64 500.68

oxidation on BDD.[34] This process could have a positive Different drugs (BGIS) were simultaneously added into
effect on BGIS degradation, and therefore the stirring con- SWW to get a final concentration of 200 μg L−1 and each
dition of 100 rpm (Re = 760) was considered in FM01-LC drug was detected during the electrolysis. Due to the
operation (an equivalent flow rate of 1.2 L min−1). complexity of SWW composition and low concentration
The current density indicated in the figure, 1.56 mA of the drugs, their quantification during electrolysis
cm−2 (associated with electrode potential of 2.6 V), was represents a challenge. Therefore, a solid phase extrac-
chosen in order to produce lower quantities of hydroxyl tion was performed to purify the sample prior to capillary
radicals and thus prevent BOM degradation. This current electrophoresis detection, achieving a good resolution
density is lower than those reported for organic mole- between peaks, Figure 4 shows an electropherogram,
cules oxidation on BDD.[35] The current density was where the migration times of each peak can be appre-
selected from several densities tested (data not shown), ciated: bezafibrate, 8.2 min; gemfibrozil, 8.4 min; indo-
and only one condition has been considered in the methacin, 8.6 min and sulfamethoxazole 9.8 min. The
text. When current densities were higher than 1.56 mA first peak corresponds to the contribution of the media.
cm−2, an excess of chlorine was formed, modifying the In the electrolysis of SWW+BGIS in the FM01-LC
BGIS degradation to a more complex mechanism reactor, SCOD was monitored for 180 min; the oxidation
without an increase in pharmaceutical degradation mediators formed during the oxidation process in BDD
rates and provoking BOM degradation. did not remove BOM over the entire electrolysis time
period when 1.56 mA cm−2 was applied (Figure 5). This
effect is associated with the presence of chlorides in was-
tewater. This is in agreement with Guzmán-Duque et al.
[21] who reported no significant crystal violet dye degra-
dation using BDD as the anode in the presence of chlor-
ides, which could be associated with the low hydroxyl
radical production in such presence, although low
current density was supplied compared with that used
in other works in the presence of the chloride ion.
[35,36] This is relevant for this study, since we were
able to remove drugs without degrading BOM through
the control of the potential (see below).

Figure 3. Curves of the current sampled vs. overpotential (Eimp-

OCP), constructed from the corresponding current transients
obtained on BDD plate (1 cm2) in SWW with/without BGIS (indi-
cated in the figure), at two solution stirring rates: (a) 100 rpm and Figure 4. Typical electropherogram of BGIS from SWW after a
(b) 1300 rpm. solid phase extraction using column C18.
 ENVIRONMENTAL TECHNOLOGY 2969

radicals formed during electrolysis. The advanced oxida-

tion of organic compounds (R) is carried out by free radi-
cals of OH• formed during water oxidation on BDD
substrate (Equation (2)), according to the following reac-
tions:

H2 O  OH†(ads) + H+ + e− , (2)

M(OH†(ads) ) + R  mCO2 + nH2 O + H+ + e− . (3)

On BDD, OH• is physisorbed and the chemical oxida-

tion of R takes place (Equation (3)). The chloride ions
Figure 5. SCOD removal in FM01-LC reactor with BDD plate (64 are chemisorbed on BDD and oxidized by OH• producing
cm2) during electrolysis of SWW with an initial concentration of chorine and degrading the BDD surface. Chlorine shows
200 µg L−1 of BGIS and 1.2 L min flow rate at current density oxidation properties distinct from those of OH•, and the
1.56 mA cm−2. interaction with drugs and BOM is also different. Further
kinetic study is necessary to understand this effect, but is
Once no degradation of BOM was observed in the outside of the scope of this article. More details about
FM01-LC reactor using BDD as the anode, both BGIS this mechanism are reported elsewhere.[22,37]
decrease and active chlorine production were monitored To our knowledge, there are no studies using a BDD
during electrolysis at 1.56 mA cm−2 (Figure 6). Due to the anode to oxidize bezafibrate, gemfibrozil or indometha-
difficulty of quantifying BGIS, six points were obtained cin to compare the percentage of degradation. With
only to provide tendency curves. Drug adsorption on respect to sulfamethoxazole, de Amorin et al. [38]
the electrochemical reactor was negligible in comparison obtained highly efficient removals at different current
with drug quantities evaluated during electrolysis. Drug densities (7.2, 21.7 and 36 mA cm−2) and at 3 h of elec-
removal greater than 50% was achieved after 20 min of trolysis. The above results show that the SWW+BGIS
electrolysis (Figure 6(a)–(d)), meanwhile, active chlorine pre-treated in the FM01-LC electrochemical reactor at
was not detected (Figure 6(e)), after that kinetics of 1.56 mA cm−2, 1.2 L s−1 of flow, during 20 min allowed
BGIS degradation changed, chlorine appeared and higher efficiency in drug removal in liquid phase with
increased with electrolysis time (30 min of electrolysis, no chlorine detected. Since BGIS is partially removed in
with a linear slope accumulating 0.65 mg L−1 in 50 min; this pre-treatment, the typical evaluation of energy con-
Figure 6(e)). This behaviour could be associated with sumption is not possible. However, a rough estimation of
the simultaneous effect between chorine and OH• the energy consumption in the pre-treatment of the spe-
cific SWW volume was calculated, as in previous works,
[39,40] using the following equation:

ItEcell 1
Ec = × ,
V 3600

where Ec is the energy (kW-h m3), Ecell is the cell potential

(V), I is the current applied in the electrolysis (A), t is the
time (s) and V is the solution volume (L).
The energy consumption required to transform the
BGIS by electro-oxidation is 0.055 kW-h m−3, while the
cost is 0.004 USD m−3. After this pre-treatment, the influ-
ent is more friendly to the biological process; this low
cost should be added to that of the overall process to
evaluate the total cost of WWTP operation. In spite of
Figure 6. Drugs removal, BGIS removal and (e) active chlorine this cost, actually photovoltaic cells could be implement
production during electrolysis (at 1.56 mA cm−2) of SWW with
in electro-oxidation systems.[41] Meanwhile, WWTP
an initial concentration of 200 µg L−1 of ABGIS and flow rate
1.2 L min, in FM01-LC reactor. Each drug was evaluated by capil- operation cost estimation without pre-treatment
lary electrophoresis: (a) bezafibrate, (b) gemfibrozil, (c) indo- should take into account the sludge management, as a
methacin and (d) sulfamethoxazole. result of the bulking effect provoked by BGIS stress in
2970 O. RODRÍGUEZ-NAVA ET AL.

Figure 7. SCOD variation of the influents (full peaks) and effluents (empty peaks) during the operation of aerobic activated sludge in
Eckenfelder reactors with 1.3 days of retention time removal. The reactors were fed with: (a, b) SWW; (c, e) SWW+BGIS and (f, g) elec-
trochemically pre-treated SWW+BGIS.

the sludge and chlorine reagent saving. This evaluation is (Table 2). Influent SCOD average values in aerobic reac-
beyond the scope of this study. tors did not present any significant difference when
When generator anodes of OH• are used in the pre- BGIS were present in SWW, R2 and R3; however, differ-
sence of chlorine and organic compounds, the risk ences were observed in the absence of BGIS, R1
involved is the formation of chlorinated compounds, (Table 2). Higher SCOD values in R2 and R3 could be
which have been commonly found in the presence of detected since methanol was used to dissolve bezafi-
high concentrations of organic matter, high current den- brate, gemfibrozil and indomethacin, whereas the efflu-
sities and high times of electrolysis.[42,43] The low ent showed a better SCOD removal efficiency in the
current density employed in the electrochemical pre- SWW, R1 (86%) and pre-treated SWW+BGIS, R3 (83%)
treatment implies that the production of these com- in relation to SWW+BGIS, R2 (78%) (Table 2). These
pounds is negligible. This is confirmed since the acti- results were verified by the analysis of significant differ-
vated sludge system, fed with the electrolysed solution, ences in SCOD removal efficiencies, which found that
does not show the presence of a specific stress (see R1 and R3 did not show significant differences in SCOD
below). removal, exhibiting higher BOM removal efficiency with
The effluent (electrolysed wastewater) was collected respect to the reactor R2, which suggests that drugs
for the subsequent aerobic activated sludge treatment. have a slightly negative effect on BOM removal in R2.
Three aerobic activated sludge continuous reactors pre-
conditioned to SWW (Figure 2) were operated to evalu-
ate the performance due to the presence of BGIS and
electrochemical pre-treatment: the first reactor was
only fed with SWW (R1), the second reactor was fed
with SWW+BGIS (R2) and the last one was fed with elec-
trolysed wastewater containing BGIS (R3). It is important
to note that the electrolysed wastewater was directly fed
to the biological reactor without further composition
modification. The three continuous reactors were moni-
tored over 51 days, during which samples were regularly
taken to measure pH, SCOD, SVI and BGIS. The continu-
ous operating conditions are shown in Table 1.
Figure 7 shows SCOD values of influents and effluents
of these three biological reactors and SCOD removal effi-
ciency is shown in Figure 8. A Turkey–Kramer multiple
comparison test with a 0.05 significance level was con- Figure 8. SCOD removal efficiency during the operation of
ducted to analyse SCOD removal efficiency, SCOD influ- aerobic activated sludge in Eckenfelder reactor: (a) SWW; (b)
ents and SCOD effluents during biological processes SWW+BGIS and (c) electrochemically pre-treated SWW+BGIS.
 ENVIRONMENTAL TECHNOLOGY 2971

Table 2. Summary of biological continuous reactor results.

SCOD (mg L−1) % SCOD Sludge discharged MLVSS SVI % Drug removal
No. effluent removal (mg day−1) (mg L−1) MLSS/MLVSS (mL gVSS−1) B G I S
R1 85 ± 49 86 93.62 130.04 ± 3 1.5 ± 0.38 117 ± 56 NA NA NA NA
R2 146 ± 56 78 164.85 221. ± 62 1.51 ± 0.27 467 ± 30 35 ± 1 41 ± 18 35 ± 12 48 ± 23
R3 111 ± 53 83 99.46 132.83 ± 3 1.48 ± 0.48 88 ± 40 ≈100
Note: B = bezafibrate, G = gemfibrozil, I = indomethacin, S = sulfamethoxazole, R1 = SWW, R2 = SWW+BGIS, R3 = SWW+BGIS electrolysed.

The results obtained during 51 days of operation of the values, where WWTP showed persistence of these com-
continuous reactors show that drug concentration in pounds [5]; they obtained average percentage removal
wastewater did not significantly affect the removal of of these drugs from various reports of WWTP, and the
organic matter. Likewise, electrochemical pre-treatment average removal percentage obtained were as follows:
using the BDD as the anode did not affect the removal bezafibrate 66%, 50% indomethacin, gemfibrozil 65%
of organic matter by activated sludge either, indicating and 60% sulfamethoxazole.[5]
that the residual synthetic electrolysed water has no During the first stage of reactor operation, SVI was
toxic effects (the presence of organochlorides in SWW– nearly 280 mL gVSS−1 and similar in the three reactors
BGIS electrolysed R3 could be discarded) on the acti- (Figure 10); after 12 days, SVI were different: 57.42 ±
vated sludge. 16.99 and 87.68 ± 41.56 mL gVSS−1; these were
Furthermore, drug degradation was evaluated in obtained when the SWW and electrolysed SWW+BGIS
effluent biological reactors. The continuous reactor fed were, respectively, fed to the biological reactor, while
with electrochemically pre-treated SWW+BGIS (R3) 466.51 ± 30.02 mL gVSS−1 was obtained when the
showed total BGIS degradation (Table 2). In contrast, reactor was fed with SWW+BGIS. For effective settling,
reactor R2 (SWW+BGIS) showed the presence of drugs sludge SVI values must be between 50 and 185 mL
in the effluent (Figure 9); during the first 16 days, the g−1.[32] The higher value of SVI in R2 is associated
BGIS removal diminished (in all four drugs) suggesting with filamentous microorganism growth, where the
an adsorption process in bioflocs. Afterwards, removal presence to mycelium is clear (Figure 11(b) and 11(c)),
efficiency increased to values below 40% in all cases which increases the costs for sludge settling. This phe-
(Figure 9). These results are consistent with the reported nomenon is related to stress provoked by BGIS despite
their lower concentration. Levantesi et al. [44] con-
firmed the presence of filamentous Alphaproteobacteria
sp. in industrial WWTP, where compounds are usually
recalcitrant and may be related with bulking. The elec-
trolysed influent (R3) did not cause this phenomenon;
Figure 11(d) shows bacillus and not mycelium. In R3,
the BGIS were absent and residual active chlorine

Figure 10. Variation of SVI during operation of aerobic sludge

activated Eckenfelder reactor with 2 days of retention time
Figure 9. BGIS removal in biological aerobic reactor fed with removal. The reactors were fed with: (a) SWW; (b) SWW+BGIS
SWW+BGIS (R2). and (c) electrochemically pre-treated SWW+ABGIS.
2972 O. RODRÍGUEZ-NAVA ET AL.

Figure 11. Scanning electron microscope images of activated sludge produced after 51 days of aerobic activated sludge Eckenfelder
reactor operation with 1.3 days of retention time removal. The reactors were fed with: (a) SWW; (b, c) SWW+BGIS and (d) electroche-
mically pre-treated SWW+BGIS.

could avoid the bulking formation. The bulking could in wastewater without significant production of chlorine
not be attributed to the active chlorine because in R1 and no BOM removal. At 20 min of electrolysis, drug
this compound was absent and the phenomenon was removal in the liquid phase was higher than 50% with
not present (Figure 11 (a)). Biological process results no BOM being removed. Furthermore, chlorine produc-
are summarized in Table 2. tion could be controlled to prevent bulking formation,
These results show the feasibility of coupling an which could otherwise affect microorganisms in the
anodic electro-oxidation with a biological system for sludge. The effluent of the electrochemical reactor was
the removal of the recalcitrant drug and BOM, in conven- fed into the activated sludge reactor, where 100% of
tional WWTP with high removal efficiencies without the drug was removed in addition to 83% removal of
affecting the performance of the activated sludge. Addi- SCOD. The energy consumption to transform the SWW
tionally, unlike other works,[17,18,45] this study used +BGIS to a feed stream more friendly to biological pro-
SWW with similar characteristics to those found in cessing is 0.055 kW-h m−3, the cost of which should be
MWW, with very small concentrations of drugs and BOM. added to that of the overall process. WWT operation
cost estimation without pre-treatment should take into
account the sludge management, as a result of bulking
Conclusions effect provoked by BGIS stress in the sludge and chlorine
To our knowledge, this is the first time that different reagent saving. This evaluation is beyond the scope of
drugs contained in wastewater were simultaneously this study.
removed by a sequential anodic oxidation-active
sludge process without modifying the chemical compo-
sition of the solution resulting from the oxidation
References
process. The sequential process was able to remove
BGIS (200 µg L−1) and BOM present in SWW (SCOD [1] Boxal A. The environmental side effects. EMBO Rep.
2004;5:1110–1112.
669.5 mg L−1). Operating conditions of the FM01-LC
[2] Owen R, Jobling S. The hidden cost of flexible fertility.
reactor using BDD (1.2 L s−1 and a current density of Nature. 2012;485:441.
1.52 mA cm−2) determined through microelectrolysis [3] Santos LH, Araújo AN, Fachini A, Pena A, Delerue-Matos C,
experiments were chosen to remove drugs contained Montenegro MCBSM. Ecotoxicological aspects related to
 ENVIRONMENTAL TECHNOLOGY 2973

the presence of pharmaceuticals in aquatic environment. [19] Sirés I, Enric B. Remediation of water pollution caused by
J Hazard Mater. 2010;175:45–95. pharmaceutical residues based on electrochemical
[4] Gómez-Oliván LM, Galar-Martínez M, Islas-Flores H, García- separation and degradation technologies: a review.
Medina S, SanJuan-Reyes N. DNA damage and oxidative Environ Int. 2012;40:212–229.
stress induced by acetylsalicylic acid in Daphnia magna. [20] Feng L, van Hullebusch ED, Rodrigo MA, Esposito G,
Comp Biochem Physiol C Pharmacol Toxicol. Oturan MA. Removal of residual anti-inflammatory and
2014;164:21–26. analgesic pharmaceuticals from aqueous systems by elec-
[5] Miège C, Choubert J, Ribeiro L, Eusèbe M, Coquery M. Fate trochemical advanced oxidation processes. Chem Eng J.
of pharmaceuticals and personal care products in waste- 2013;228:944–964.
water treatment plants – Conception of a database and [21] Guzmán-Duque FL, Palma-Goyes RE, González I, Peñuela
first results. Environ Pollut. 2009;157:1721–1726. G, Torres-Palma RA. Relationship between anode material,
[6] Kraigher B, Kosjek T, Heath E, Kompare B, Mandic-Mulec I. supporting electrolyte and current density during electro-
Influence of pharmaceutical residues on the structure of chemical degradation of organic compounds in water. J
activated sludge bacterial communities in wastewater Hazard Mater. 2014;278:221–226.
treatment bioreactors. Water Res. 2008;42:4578–4588. [22] Martínez-Huitle CA, Ferro S. Electrochemical oxidation of
[7] Al-Ahmad A, Daschner F, Kümmerer K. Biodegradability of organic pollutants for the wastewater treatment: direct
cefotiam, ciprofloxacin, meropenem, penicillin G, and sul- and indirect processes. Chem Soc Rev. 2006;35:1324–
famethoxazole and inhibition of wastewater bacteria. 1340.
Arch Environ Contam Toxicol. 1999;37(2):158–163. [23] Quinn B, Gagné F, Blaise C. An investigation into the acute
[8] Gónzalez O, Sans C, Esplugas S. Sulfamethoxazole abate- and chronic toxicity of eleven pharmaceuticals (and their
ment by photo-Fenton toxicity, inhibition and biodegrad- solvents) found in wastewater effluent on the cnidarian,
ability assessment of intermediates. J Hazard Mater. Hydra attenuata. Sci Total Environ. 2008;389:306–314.
2007;146:459–464. [24] Raldúa D, André M, Babin PJ. Clofibrate and gemfibrozil
[9] Trovó AG, Nogueira RF, Agüera A, Fernandez-Alba AR, induce an embryonic malabsorption syndrome in zebra-
Sirtori C, Malato S. Degradation of sulfamethoxazole in fish. Toxicol Appl Pharmacol. 2008;228:301–314.
water by solar photo-Fenton. Chemical and toxicological [25] López-Doval JC, Kukkonen JV, Rodrigo P, Muñoz I. Effects
evaluation. Water Res. 2009;43:3922–3931. of indomethacin and propanolol on Chironomus riparius
[10] Gao S, Zhao Z, Xu Y, et al. Oxidation of sulfamethoxazole and Physella (Costatella) acuta. Ecotoxicol Environ Saf.
(SMX) by chlorine, ozone and permanganate – a compara- 2012;78:110–115.
tive study. J Hazard Mater. 2014;274:258–269. [26] Kosma CI, Lambropoulou DA, Albanis TA. Investigation of
[11] Rodayan A, Roy R, Yargeau V. Oxidation products of sulfa- PPCPs in wastewater treatment plants in Greece: occur-
methoxazole in ozonated secondary effluent. J Hazard rence, removal and environmental risk assessment. Sci
Mater. 2010;177:237–243. Total Environ. 2014;466–467:421–438.
[12] Beltrán FJ, Aguinaco A, García-Araya JF, Oropesa A. Ozone [27] ASTM International. D 5905–98. Standard practice for the
and photocatalytic processes to remove the antibiotic preparation of substitute wastewater; 1998.
sulfamethoxazole from water. Water Res. 2008;42:3799– [28] U.S. Environmental Protection Agency. Method 1694:
3808. pharmaceuticals and personal care products in water,
[13] Nasuhoglu D, Yargeau V, Berk D. Photo-removal of sulfa- soil, sediment, and biosolids by HPLC/MS/MS; 2007.
methoxazole (SMX) by photolytic and photocatalytic pro- [29] APHA, Standard methods for examination of water & was-
cesses in a batch reactor under UV-C radiation (λmax=254 tewater [Greenberg AE; Eaton A D, Clesceri LS, editors];
nm). J Hazard Mater. 2011;186:67–75. 1998.
[14] Ganzenko O, Huguenot D, van Hullebusch ED, Esposito G, [30] Rodríguez FA, Mateo MN, Aceves JM, Rivero EP, González I.
Oturan MA. Electrochemical advanced oxidation and bio- Electrochemical oxidation of bio-refractory dye in a simu-
logical processes for wastewater treatment: a review of lated textile industry effluent using DS electrodes in a
the combined approaches. Environ Sci Pollut Res. filter-press type FM01-LC reactor. Environ Technol.
2014;21:8493–8524. 2013;34:573–583.
[15] Oller I, Malato S, Sánchez-Pérez JA. Combination of [31] Nava J, Núñez F, González I. Electrochemical incineration
advanced oxidation processes and biological treatments of p-cresol and o-cresol in the filter-press-type FM01-LC
for wastewater decontamination – a review. Sci Total electrochemical cell using BDD electrodes in sulfate
Environ. 2011;409:4141–4166. media at pH 0. Electrochem Acta. 2007;52:3229–3235.
[16] Szpyrkowicz L, Kaul SN, Neti RN. Tannery wastewater [32] Palm JC, Jenkins D, Parker DS. Relationship between organic
treatment by electro-oxidation coupled with a biological loading, dissolved oxygen concentration and sludge settle-
process. J Appl Electrochem. 2005;35:381–390. ability in the completely-mixed activated sludge process. J
[17] Fontmorin JM, Fourcade F, Geneste F, Floner D, Huguet S, Water Pollut Control Fed. 1980;52:2484–2506.
Amrane A. Combined process for 2,4-dichlorophenoxya- [33] Fabianka A, Bialk-Bielinska A, Stepnowski P, Stolte S,
cetic acid treatment – coupling of an electrochemical Siedlecka EM. Electrochemical degradation of sulfona-
system with a biological treatment. Biochem Eng J. mides at BDD electrode: kinetics, reaction pathway and
2013;70:17–22. eco-toxicity evaluation. J Hazard Mater. 2014;280:579–
[18] Mansour D, Fourcade F, Huguet S, et al. Improvement of 587.
the activated sludge treatment by its combination with [34] Panizza M, Cerisola G. Direct and mediated anodic oxida-
electro Fenton for the mineralization of sulfamethazine. tion of organic pollutants. Chem Rev. 2009;109:6541–
Int Biodeterior Biodegrad. 2014;88:29–36. 6569.
2974 O. RODRÍGUEZ-NAVA ET AL.

[35] Murugananthan M, Latha S, Bhaskar Raju G, Yoshihara S. [40] Dominguez-Ramos A, Aldaco R, Irabien A. Electrochemical
Role of electrolyte on anodic mineralization of atenolol oxidation of lignosulfonate: total organic carbon oxidation
at boron doped diamond and Pt electrodes. Sep Purif kinetics. Ind Eng Chem Res. 2008;47:9848–9853.
Technol. 2011;79:56–62. [41] Dominguez-Ramos A, Aldaco R, Iraben A. Photovoltaic
[36] Mascia M, Vacca A, Polcaro AM, Palmas S, Ruiz JR, Da solar electrochemical oxidation (PSEO) for treatment of
Pozzo A. Electrochemical treatment of phenolic waters lignosulfonate wastwater. J Chem Technol Biotechnol.
in presence of chloride with boron-doped diamond 2010;85:821–830.
(BDD) anodes: experimental study and mathematical [42] Garcia-Segura S, Keller J, Brillas E, Radjenovic J. Removal of
model. J Hazard Mater. 2010;174:314–322. organic contaminants from secondary effluent by anode
[37] Comninellis Ch. The electrochemical oxidation (or com- oxidation with a boron-doped diamond anode as tertiary
bustion) of organics with simultaneous oxygen evolution. treatment. J Hazard Mater. 2015;283:551–557.
Electrochim Acta. 1994;39:1857–1862. [43] Anglada Á, Urtiaga A, Ortiz I, Mantzavinos D, Diamadopoulos
[38] de Amorin KP, Romualdo LL, Andrade LS. Electrochemical E. Boron-doped diamond anodic treatment of landfill lea-
degradation of sulfamethoxazole and trimethoprim at chate: evaluation of operating variables and formation of
boron-doped diamond electrode: performance, kinetics oxidation by-products. Water Res. 2011;45:828–838.
and reaction pathway. Sep Purif Technol. 2013;120:319– [44] Levantesi C, Beimfohr C, Geurkink B, et al. Filamentous
327. alphaproteobacteria associated with bulking in industrial
[39] Panizza M, Michaud PA, Cerisola G, Comninellis Ch. wastewater treatment plants. System Appl Microbiol.
Electrochemical treatment of wastewaters containing 2004;27:716–727.
organic pollutants on boron-doped diamond electro- [45] Han WQ, Wang LJ, Sun XY, Li JS. Treatment of bactericide
des: prediction of specific energy consumption and wastewater by combined process chemical coagulation,
required electrode area. Electrochem Commun. electrochemical oxidation and membrane bioreactor. J
2001;3:336–339. Hazard Mater. 2008;151:306–315.