ww.sciencedirect.com

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7

Available online at w

ScienceDirect

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

Attenuation of trace organic compounds (TOrCs)in bioelectrochemical systems

Craig M. Werner a, Christiane Hoppe-Jones a, Pascal E. Saikaly a,Bruce E. Logan b, Gary L. Amy a,*

a Water Desalination and Reuse Center, King Abdullah University of Science and Technology (KAUST), Thuwal,

23955-6900, Saudi Arabiab Department of Civil and Environmental Engineering, Pennsylvania State University, University Park, PA 16802,

USA

a r t i c l e i n f o

Article history:

Received 20 September 2014

Received in revised form

6 January 2015

Accepted 7 January 2015

Available online 17 January 2015

Keywords:

Bioelectrochemical systems

Trace organic compounds

Biotransformation

Sorption

* Corresponding author. Tel.: þ966 12 808 23E-mail address: gary.amy@kaust.edu.sa (

http://dx.doi.org/10.1016/j.watres.2015.01.0130043-1354/© 2015 Published by Elsevier Ltd.

a b s t r a c t

Microbial fuel cells (MFCs) and microbial electrolysis cells (MECs) are two types of microbial

bioelectrochemical systems (BESs) that use microorganisms to convert chemical energy in

wastewaters into useful energy products such as (bio)electricity (MFC) or hydrogen gas

(MEC). These two systems were evaluated for their capacity to attenuate trace organic

compounds (TOrCs), commonly found in municipal wastewater, under closed circuit

(current generation) and open circuit (no current generation) conditions, using acetate as

the carbon source. A biocide was used to evaluate attenuation in terms of biotransfor-

mation versus sorption. The difference in attenuation observed before and after addition of

the biocide represented biotransformation, while attenuation after addition of a biocide

primarily indicated sorption. Attenuation of TOrCs was similar in MFCs and MECs for eight

different TOrCs, except for caffeine and trimethoprim where slightly higher attenuation

was observed in MECs. Electric current generation did not enhance attenuation of the

TOrCs except for caffeine, which showed slightly higher attenuation under closed circuit

conditions in both MFCs and MECs. Substantial sorption of the TOrCs occurred to the

biofilm-covered electrodes, but no consistent trend could be identified regarding the

physico-chemical properties of the TOrCs tested and the extent of sorption. The octanol

ewater distribution coefficient at pH 7.4 (log DpH 7.4) appeared to be a reasonable predictor

for sorption of some of the compounds (carbamazepine, atrazine, tris(2-chloroethyl)

phosphate and diphenhydramine) but not for others (N,N-Diethyl-meta-toluamide). Aten-

olol also showed high levels of sorption despite being the most hydrophilic in the suite of

compounds studied (log DpH 7.4 ¼ �1.99). Though BESs do not show any inherent advan-

tages over conventional wastewater treatment, with respect to TOrC removal, overall re-

movals in BESs are similar to that reported for conventional wastewater systems, implying

the possibility of using BESs for energy production in wastewater treatment without

adversely impacting TOrC attenuations.

© 2015 Published by Elsevier Ltd.

85.G.L. Amy).

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7 57

1. Introduction

A number of different trace organic compounds (TOrCs),

including pharmaceuticals, pesticides, insecticides, endocrine

disrupting compounds, personal care products and flame re-

tardants, have been detected in municipal wastewater efflu-

ents as well as in ground and surface waters around the world

(Daughton and Ternes, 1999; Hirsch et al., 1999; Snyder et al.,

2003; Benotti et al., 2008; Alidina et al., 2014; Fries and

Puttmann, 2003; Fatta-Kassinos et al., 2011). The most com-

mon pathway for these compounds into the aquatic envi-

ronment is through wastewater discharges due to incomplete

removal by conventional biological wastewater treatment

processes (Glassmeyer et al., 2005).

The increased attention given to the identification and

quantification of TOrCs in wastewaters and water sources is

due to the chronic toxicological risks they pose (Batt et al.,

2006; Quinn et al., 2008) and their persistence through con-

ventional wastewater and potable water treatment systems

(Daughton and Ternes, 1999; Stackelberg et al., 2004). The

incomplete removal of these compounds presents a challenge

to the practice of water reclamation and reuse of treated

municipal wastewater for non-potable (agricultural irrigation)

or direct and indirect potable reuse applications (Levine and

Asano, 2004; Snyder et al., 2007; Drewes et al., 2005).

Water reclamation and reuse provides a reliable and sig-

nificant water source particularly for regions with water

scarcity. The removal of these emerging compounds of

concern from wastewaters is thus increasingly important.

Removal of TOrCs from the aqueous phase ofwastewaters can

be attributed mainly to microbial transformation (biotrans-

formation), volatilization, and/or physical sorption to solids in

the liquid stream (Hyland et al., 2012). Volatilization is not

likely to play a significant role in the attenuation of the ma-

jority of water soluble TOrCs due to their low vapor pressures.

Effective removal of TOrCs requires advanced treatment

technologies such as ozonation or membrane filtration (Sui

et al., 2010; Drewes et al., 2002) which can be costly and en-

ergy intensive. For this reason, novel wastewater treatment

technologies to attenuate these emerging contaminants are

needed.

Bioelectrochemical systems (BESs), such as microbial fuel

cells (MFCs) and microbial electrolysis cells (MECs), have been

proposed as alternative wastewater treatment technologies as

energy production with these systems can significantly

reduce, or potentially offset, the specific energy consumption

(kWh/m3) associated with conventional wastewater treat-

ment processes (Logan and Regan, 2006; Heidrich et al., 2013).

Biological oxidation of organic compounds occurs at the

anode electrode by bacteria that form a biofilm, with transfer

of electrons from themicroorganisms to the electrode surface.

These electrons pass through an external electrical circuit to

the cathode where a reduction reaction occurs in order to

sustain anodic oxidation. In an MFC, oxygen is reduced at the

cathode (Logan et al., 2006), and in an MEC protons are

reduced to form hydrogen gas with the addition of electrical

energy (Logan et al., 2008).

There are limited reports on the use of BES for the atten-

uation or removal of TOrCs. Harnisch et al. (2013) reported the

complete removal of the antibiotic sulfamethoxazole, a TOrC

widely detected in wastewater effluents, using wastewater

derived anodic biofilms in a BES fed with acetate at a fixed

anode potential of 0.2 V (Ag/AgCl). Anodic biofilms are also

known to be robust and can tolerate high concentrations

(1000 mg/L) of potentially toxic compounds, such as sulfa-

methoxazole, without any noticeable deterioration in bio-

electrocatalytic performance (Patil et al., 2010). Apart from the

study by Harnisch et al. (2013), there is little information

available to evaluate the attenuation of various TOrCs in BES.

In this study, the attenuation of 10 TOrCs in MFCs and

MECs operated at fixed anode potential �0.4 V (vs Ag/AgCl)

was investigated using five pharmaceuticals, two antibiotics,

a flame retardant, a pesticide, and an insecticide, embodying a

range of physicochemical properties. The term attenuation, as

used here, includes both biotransformation (complete or

incomplete biodegradation) and sorption. While biotransfor-

mation represents elimination or transformation of the

parent compound, sorption can result in the retardation of its

transport through reversible or irreversible binding to the

biofilm or electrode material. The levels of attenuation were

compared to reports from the literature for TOrC removals in

conventional biological wastewater treatment systems.

2. Materials and methods

2.1. Reactor design and operation

Duplicate cube-shaped single-chamber MFC and MEC re-

actors, having an internal cylindrical liquid chamber (28 mL,

7 cm2 cross section), were constructed as previously described

(Liu and Logan, 2004; Call and Logan, 2008). The anodes were

graphite fiber brushes (25 mm diameter � 25 mm length) with

a titanium core (Mill-Rose, OH, USA) that were pre-treated at

450 �C in an oven for 30 min before inoculation. These brush

anodes are estimated to have a very large surface area to

volume ratio of 18,200m2 perm3 of brush volume (Logan et al.,

2007).

Cathodes (projected surface area of 7 cm2) were prepared

from carbon cloth (30 wt% wet proofing polymer, Fuel Cell

Earth, MA, USA) with four PTFE diffusion layers and 0.5mg-Pt/

cm2 (Cheng et al., 2006). Reported current densities were

normalized to the surface area of the cathode. The reactors

were inoculated using a wastewater (5% by volume) from the

KAUST wastewater treatment plant, and fed sodium acetate

(1 g/L) in an 80 mM bicarbonate buffer solution (Ambler and

Logan, 2011) (NaHCO3, 6.71 g/L; NH4Cl, 0.31 g/L; Na2HPO4,

0.05 g/L; Na2H2PO4$H2O, 0.03 g/L) containing 10 mL/L trace

vitamins and minerals solutions (Balch et al., 1979). The so-

lutions were sparged for at least 1 h with a N2:CO2 (80:20) mix

to neutralize the pH (final pH of 7.3e7.5) and remove dissolved

oxygen.

The reactors were operated under a fixed anode potential

of �0.4 V (vs Ag/AgCl, or �0.2 versus a standard hydrogen

electrode, SHE) using a potentiostat (VMP3, BioLogic). This

potential was selected as it is close to the oxidation potential

of acetate (�0.28 V vs SHE under standard conditions and

neutral pH) (Rabaey and Rozendal, 2010). The anode was used

as the working electrode and the cathode as the counter

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 758

electrode. A reference electrode (Ag/AgCl, MF-2052, BASi) was

positioned between the two electrodes and closest to the

anode electrode. The reference electrode had a potential of

þ210 mV vs SHE. The potentiostat was operated using the EC-

Lab V10.19 software application (BioLogic, Claix, France). All

potentials are given here relative to Ag/AgCl. After the re-

actors had achieved reproducible performance with respect to

current generation, the TOrC attenuation test was initiated.

2.2. Preparation of TOrC solution

The chemical structures and compound-specific information

for the TOrCs analyzed in this study are shown in Table 1. The

octanolewater partition constant (Kow) was used to assess the

hydrophobicity of a particular compound (i) and is defined as

the ratio of the concentrations of the compound (i) in the

organic (n-octanol, l) and water (w) phases (Schwarzenbach

et al., 2003):

Kow ¼ Cil

Ciw(1)

where Cil and Ciw are the concentrations of compound i in the

organic and water phases at equilibrium, respectively. The

term Log D (distribution coefficient) shown in Table 1 is used

to assess the hydrophobicity of a compound at a selected pH

of interest. The Log D values reported in Table 1 are the

octanolewater partition coefficients at pH 7.4. For neutral

compounds Log Kow ¼ Log D (Carballa et al., 2008). The acid

dissociation constant for acidic compounds (Ka) is a conve-

nient way to assess the relative strengths of acids. If HA is

considered an acid, then its acid dissociation constant is

defined as the equilibrium constant for the reaction

(Benjamin, 2002):

HA4Hþ þA� (2)

Ka ¼ fHþgfA�gfHAg (3)

A stock solution of TOrCs was prepared by dissolving the

selected 10 compounds in puremethanol at a concentration of

1mg/L. This stock solutionwas then concentrated to 250mg/L

by drying under nitrogen. To prepare a medium (250 mL) with

a targeted final TOrC concentration of 50 mg/L for each com-

pound, 50 mL of the feed solution was replaced with 50 mL of

the stock TOrC solution. The stock TOrC solution was

concentrated to minimize the amount of methanol added

when preparing the medium and avoid introducing a second

carbon source. TOrCs have been detected at concentrations

higher than 70 mg/L in domestic wastewaters (Buerge et al.

2003) and therefore a TOrC concentration of 50 mg/L was

deemed appropriate for these experiments. This concentra-

tion also allowed for more accurate TOrCmeasurement in the

samples as samples were not concentrated before injection

into the LC-MS/MS.

2.3. Attenuation tests: feeding and sampling procedure

Attenuation tests were conducted under closed and open

circuit conditions using the same methodology. Chemicals

were added once the MFCs and MECs had achieved repro-

ducible performance, based on similar voltage profiles over at

least 3 cycles. The cycle time for each of the reactors was the

same (24 h). The feeding and sampling procedure was

designed to allow the measurement of TOrCs at the start and

end of a batch over successive cycles. A feed solution con-

taining the TOrCs (50 mg/L of each compound) was added into

the medium used in the first batch cycle, and a sample of this

solution was taken at the end of the cycle for analysis. After

the first attenuation test cycle, a 3 mL sample was taken from

each reactor for analysis, and 6 mL of a concentrated acetate

solution (5.5 g/L) that did not contain TOrCswas added to each

reactor. A glass tube fitted to the top of each reactor accom-

modated the extra volume during the sampling process.

Following a short equilibration period (30 min) a 3 mL sample

was taken to measure the initial TOrC concentration and pH

analysis. This procedure was repeated for the remaining four

cycles of the attenuation test. Samples were filtered through

0.45 mm sterile syringe filters prior to analysis. All results

shown indicate averages with error bars showing standard

deviations of measurements from the duplicate reactors.

2.4. Sorption tests

Sodium azide (NaN3, 6 mM) was used as a biocide to prevent

biotransformation, and allow measurement of TOrC adsorp-

tion to the biofilm-covered electrodes. NaN3 acts as a proton

conducting uncoupler by inhibiting the ATPase enzyme that is

responsible for ATP synthesis, (Kobayashi et al., 1977) and at

high concentrations it is toxic to bacteria. Tests using the

biocide were conducted after the attenuation tests were

completed. The same feeding and sampling procedure for the

attenuation test was used, except amedium containing 0.5 g/L

sodium acetate was used to replace sampled volume to pre-

vent the acetate concentration from increasing to a level that

would influence detection of TOrCs. The reactors were oper-

ated in open circuit mode for these experiments.

2.5. Analyses

Acetate concentrations were measured by high performance

liquid chromatography (HPLC; Accela system, Thermo Scien-

tific) using an Aminex HPX-87H Ion Exclusion Column (Bio-

Rad Laboratories, Hercules, CA). A 5 mM sulphuric acid solu-

tion was used as the mobile phase at a flow rate of 0.55 mL/

min. Acetate peaks were detected at 210 nm using a photo-

diode array detector. The pH of each sample was measured

using a pH meter (Cyberscan 6000, Eutech Instruments) at the

start and end of each cycle.

TOrCs were analyzed by liquid chromatography with

coupled tandem mass spectrometry (LC-MS/MS) using an

Agilent Technology 1260 Infinity Liquid Chromatography

(LC) unit and an ABSciex 5500 Q-Trap MS/MS (Alidina et al.,

2014). A sample injection volume of 10 mL was used for all

samples. The mobile phase used in the positive electrospray

ionization (ESI) mode was 4 mM ammonium formate in

water containing 0.1% formic acid (A) and 4 mM ammonium

formate in methanol containing 0.1% formic acid (B). The

mobile phases were supplied as a binary gradient at a flow

rate of 800 mL/min as follows: 90% A kept for 0.5 min, then

Table 1eChemical structure and characteristics of the selected TOrCs used in this study and their respective charged statesestimated from the compound's pKa relative to the experimental pH (8e8.5).

Compound Structure Classification LogKow

aLog D

(pH 7.4)a,fpKa

f Chargestate

TCEP

(tris(2-chloroethyl)phosphate)

Flame retardant 1.44 0.48 NA Neutral

DEET

(N,N-Diethyl-meta-toluamide)

Insecticide 2.18 1.96 0.67b Neutral

Carbamazepine Anticonvulsant 2.45 2.67 14e Neutral

Atrazine Pesticide 2.61 1.54 1.7c Neutral

Caffeine Stimulant �0.07 �0.13 0.7g Neutral

Sulfamethoxazole Antibiotic 0.89 �0.32 5.6d Negative

Trimethoprim Antibiotic 0.91 0.67 7.12c Neutral

(continued on next page)

Table 1 e (continued )

Compound Structure Classification LogKow

aLog D

(pH 7.4)a,fpKa

f Chargestate

Atenolol Beta blocker 0.16 �1.99 9.6c Positive

Diphenhydramine Antihistamine 3.27 2.29 8.98c Positive

Primidone Anticonvulsant 0.91b 0.4 11.5 Neutral

a Chemspider.com.b Snyder et al., 2007.c SRC PhysProp Database.d Yangali-Quintanilla et al., 2009.e Scheytt et al., 2005.f Log Kow is the octanolewater partition coefficient which is a measure of the ratio of concentrations of a compound in octanol and water

(Schwarzenbach et al., 2003). Log D is the octanolewater partition coefficient for a compound at a specified pH (Carballa et al., 2008). pKa is

calculated as �log Ka.g Martınez-Hern�andez et al., 2015.

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 760

reduced to 50% at 0.51 min and decreased linearly to 5% at

8 min. It was kept at 5% for 6 min and then increased to 90%

A for an equilibration step of 4 min, giving a total run time

of 18 min.

In ESI negative mode, the mobile phase used was 2 mM

ammonium acetate in water (A), and 2 mM ammonium ace-

tate inmethanol (B). The binary gradient delivering themobile

phases at a flow rate of 800 mL/min was programmed as fol-

lows: 90% A kept for 0.5 min, then dropped to 60% at 0.51 min

and decreased linearly to 5% at 8 min. It was held at 5% for

11 min and then increased linearly to 90% A over a period of

3 min followed by a 4 min equilibration period at 90% A. The

total sample run time was also 18 min.

Mass spectrometry (MS) was performed using an AB SCIEX

QTRAP 5500 mass spectrometer (Applied Biosystems, Foster

City, CA). The ionization source used in the MS was limited to

ESI, while optimum polarity used for each compound was

selected based on previous work done by Alidina et al. (2014).

Samples for LC-MS/MS were prepared as follows: 900 mL

sample, 100 mL isotope solution and 1 mL sodium omidine so-

lution. Samples were stored at 4 �C if not measured

immediately.

3. Results and discussion

3.1. Reactor performance

The duplicate MEC reactors produced the same maximum

current densities of 6.72 ± 0.13 A/m2 and 6.73 ± 0.22 A/m2

(cathode surface area; 7 cm2), while those produced by MFC

reactors were reasonably similar (4.26 ± 0.45 A/m2, and

3.59 ± 0.29 A/m2) prior to the start of the TOrC attenuation

tests (average ± S.D. for 3 cycles) (Fig. 1). The maximum cur-

rent densities increased with each cycle in the TOrC attenu-

ation tests (Fig. 1) due to a build-up of salt in solution

associated with the addition of concentrated sodium acetate

at the start of each cycle. The improved current profiles over

successive cycles indicated that the presence of the trace

organic compounds did not adversely affect the anodic mi-

crobial community. Themaximum current densities returned

to the pre-attenuation test levels when the full contents of the

reactors were replaced with fresh medium (data not shown).

The acetate removals in MECs and MFCs were comparable,

97% and 98%, respectively. The cycle durationwas determined

Fig. 1 e Electric current generation in MFC and MEC

reactors operated at a set anode potential of ¡0.4 V (Ag/

AgCl). The blue arrow indicates the start of the TOrC

attenuation test. Blue and green curves (higher maximum

current densities) are MECs while red and orange curves

(lower maximum current densities) are MFCs. The

anomalous data shown in the third cycle for one of the

MFCs was due to a faulty anode connection that was

corrected. (For interpretation of the references to color in

this figure legend, the reader is referred to the web version

of this article.)

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7 61

using the minimum current density measured for each

reactor which allowed for similar cycle durations. The

coulombic efficiencies were 69 ± 1% (MECs) and 42 ± 10%

(MFCs) based on the change in acetate concentration over a

cycle and coulombs transferred. The higher CE observed for

MECs compared to MFCs was expected due to the lack of

substrate loss to oxygen that diffused into the medium

through the cathode. These coulombic efficiencies indicate

that a relatively large fraction of the COD was removed by the

non-exoelectrogenic microbial community. A recent study

showed that most of the COD removal in MFCs occurs in the

first 9 h of a batch cycle and when current densities are

highest (Ren et al., 2014a). The CE of MFCs increases at higher

current densities as exoelectrogens outcompete heterotrophic

bacteria for substrate thereby directing more substrate elec-

trons to current (Ren et al., 2014a). However, operating these

systems beyond the point where most COD is removed would

not result in improved energy recovery and highlights that a

trade-off exists between maximum energy recovery and

maximum COD removal. Due to the first-order relationship of

COD degradation rates relative to COD concentration in closed

or open circuit conditions, the rates of COD removal by non-

exoelectrogens decreases at higher CEs (Ren et al., 2014a).

The pH of the medium increased after the first cycle from

7.4 to approximately 8.2 (MFCs) and 8.5 (MECs) (Table 2). The

pH measurements of subsequent cycles were in the range of

Table 2 e Measured pH at the start and end of each batch cycle

Reactor Cycle 1 Cycle 2 Cy

Initial Final Initial Final Initial

MFC 7.39 8.21 ± 0.18 8.10 ± 0.4 7.85 ± 0.23 7.87 ± 0.02

MEC 7.39 8.49 ± 0.39 8.24 ± 0.25 8.47 ± 0.36 8.47 ± 0.03

7.8e8.2 (MFCs) and 8.2 to 8.8 (MECs) (Table 2). An increase in

pH in this range would not be expected to affect performance

(He et al., 2008).

3.2. Attenuation of TOrCs in MFCs and MECs

The selected TOrCs were classified according to several

different bins based on their attenuation behavior and trends.

The criteria used for this binning process were: 1) dependence

on mode of operation; 2) dependence on the presence of

electric current; and 3) effect of NaN3 inhibition. Under the

experimental conditions, the selected TOrCs were mostly

neutral or positively charged except for sulfamethoxazole

which was negatively charged (Table 1).

The attenuation observed for each compound is dis-

cussed in the subsequent sections in terms of biotransfor-

mation and sorption, the two dominant processes for

elimination of TOrCs in wastewater treatment processes

(Ternes et al., 2004a) and compared to conventional acti-

vated sludge (CAS). The term sorption is used as a collective

term for hydrophobic and electrostatic interactions of TOrCs

with the biofilm covered electrodes. Hydrophobic in-

teractions occur between the compound and the lipophilic

cell membrane of microorganisms and electrostatic in-

teractions result from the difference in the charged states of

compounds and the surface of the biofilm (Ternes et al.,

2004a). BESs have lower amounts of biosolids compared to

CAS due to anaerobic growth and most of the biosolids in

BES occur as biofilms attached to the electrode. CAS systems

have higher amounts of biosolids that are predominantly

suspended.

3.2.1. Bin 1: caffeineCaffeine was binned independently as this was the only

compound which showed that electric current increased

levels of attenuation in both MFC and MEC reactors (Fig. 2).

Attenuation in MFCs was 57 ± 6% (current flowing) and

33 ± 0.2% (no current), and 72 ± 1% (current flowing) and

53 ± 8% (no current) in MECs (Fig. 2). Attenuation in the

presence of NaN3 was 22 ± 1% (MFCs) and 17 ± 8% (MECs)

(Fig. 2). These observations suggest that the biotransformation

of caffeine is linked to anodic oxidation of acetate but addi-

tional studies are needed for conclusive confirmation. As

discussed in Section 3.1, the difference in attenuation of

TOrCs in MECs comparedwith MFCs is likely due to the higher

current density and CE in MECs compared to MFCs and a

smaller fraction of COD removed by non-exoelectrogens.

Caffeine attenuation attributed to biotransformation was

higher thanmost of the other TOrCs tested. Caffeine is usually

well removed (80e100 %) in aerobic wastewater treatment

plants (Buerge et al., 2003; P�erez et al., 2005; Okuda et al., 2008).

Caffeine is hydrophilic and has relatively low tendency to sorb

over the course of the attenuation test.

cle 3 Cycle 4 Cycle 5

Final Initial Final Initial Final

7.99 ± 0.09 7.82 ± 0.14 8.20 ± 0.23 7.88 ± 0.02 8.24 ± 0.51

8.64 ± 0.32 8.44 ± 0.2 8.82 ± 0.14 8.50 ± 0.28 8.44 ± 0.11

Fig. 2 e Attenuation behavior of caffeine in MFC and MEC

reactors (Error bars indicate ±S.D. for duplicate reactors).

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 762

to organic matter, and therefore biotransformation is

considered to be the principal mechanism for removal in

other systems (Buerge et al., 2003). The attenuation of caffeine

measured following NaN3 addition in this study demonstrated

its relatively low tendency for sorption. It has been reported

that caffeine can be removed using anaerobic treatment pro-

cesses with similar efficiency as an activated sludge reactor

(Froehner et al., 2011). They found that the caffeine concen-

tration in a domestic wastewater decreased from 9.3 mg/L to

below detection limits following treatment in an upflow

Fig. 3 e Attenuation of (A) carbamazepine, (B) atrazine, (C) TCEP

represents attenuation in closed circuit operation, the second r

attenuation after addition of sodium azide.

anaerobic sludge blanket (UASB), and from 8.2 to 0.02 mg/L in

an activated sludge aeration tank (Froehner et al. 2011).

Caffeine also has a positive effect on biogas production in

anaerobic digestion. Prabhudessai et al. reported an increase

in biogas production to 408 mL/g total solids (TS) compared to

only 182 mL/g TS for the control reactor with caffeine at a

concentration of 100 ppm (Prabhudessai et al. 2009).

3.2.2. Bin 2: carbamazepine, atrazine, TCEP, DEETThis second bin of compounds showed that attenuation was

not dependent on mode of operation (MEC or MFC), the pres-

ence of electric current, and the NaN3 sorption test showed

similar levels of attenuation in both MFC and MEC reactors.

This bin was comprised of carbamazepine, atrazine, TCEP and

DEET (Fig. 3).

Carbamazepine, an anticonvulsant drug that is ubiqui-

tously present in wastewater influents and effluents, was well

removed in both MFCs and MECs with overall attenuation of

the parent compound >80% in closed or open circuit condi-

tions (Fig. 3A). Up to 60% of the initial concentration was still

removed after addition of NaN3. By difference, this indicated

that ~20% was biologically degraded in the reactors with

actively respiring biofilms (Fig. 3A), and that sorption was

responsible for the bulk of the removal. Carbamazepine is

resistant to biodegradation in most conventional biological

treatment systems, with removals rarely exceeding 10% of the

initial concentration (Zhang et al., 2008; Joss et al., 2005). It is

also considered to have a low tendency for sorption to sludge

(Joss et al., 2005; Ternes et al., 2004b; Wick et al., 2009). Car-

bamazepine also showed no elimination from sewage sludge

and (D) DEET in MFC and MEC reactors. The first bar

epresents open circuit conditions and the third represents

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7 63

after treatment by anaerobic digestion (Carballa et al., 2007).

For these reasons, carbamazepine has been proposed as an

anthropogenic marker compound for the aquatic environ-

ment (Clara et al., 2004).

The observed attenuation of carbamazepine in this study is

higher than reports for conventional wastewater treatment in

literature particularly regarding sorption. Carbamazepine is

hydrophobic and has a high log D of 2.67 (octanolewater

partition coefficient at pH 7.4) (Table 1). Considering its high

pKa of 14, hydrophobic sorption is likely and has previously

been reported to be the dominant mechanism for attenuation

of carbamazepine in sandy sediments (Scheytt et al., 2005).

Hydrophobic sorption is proposed as the dominant mecha-

nism for the sorption of carbamazepine observed in this

study. The higher sorption of carbamazepine shown in this

study as compared to sorption to activated sludge may be due

to the relatively large surface area of the anode electrode

where most of the biofilm is attached. Sorption of TOrCs to

available binding sites of the exposed electrode may also

occur.

Atrazine showed similar attenuation (74e85 %) in MFCs

and MECs under closed and open circuit conditions (Fig. 3B).

Attenuation in the presence of NaN3 was approximately 60%

for MFCs and MECs, so biotransformation was around 15e25

%. Previous reports have demonstrated that atrazine was not

well removed in an activated sludge pilot plant (Monteith

et al., 1995). However, enhanced removals of atrazine (>90%)

have been reported using genetically modified organisms in

an MBR treating domestic wastewater spiked with atrazine

(Liu et al., 2008). Atrazine has been reported to be biodegraded

under anaerobic conditions. Ghosh et al. reported up to 45%

biodegradation of atrazine in a co-metabolic process in which

dextrose served as the primary carbon source (Ghosh and

Philip, 2004).

The flame retardant TCEP, a chlorinated organo-

sphosphate, was attenuated by 51 ± 4% (current flowing) and

76 ± 3% (no current) in MFCs (Fig. 3C). In MECs, this compound

was attenuated by 74 ± 3% (current flowing) and 67 ± 3% (no

current) (Fig. 3C). An opposing trend of attenuation under

closed and open circuit conditions can be observed in MFCs

and MECs which could be due to the activity of different mi-

crobial communities (Sayess et al., 2013; Shehab et al., 2013).

In the presence of NaN3 the attenuation of TCEP was

approximately 25% in both MFCs and MECs (Fig. 3C). TCEP has

been reported to be difficult to attenuate in wastewater

treatment plants with reports of zero (Meyer and Bester, 2004)

to 42% elimination (Bernhard et al., 2006) in activated sludge

systems. MFCs and MECs showed higher attenuation of this

compound than has been reported for conventional

treatment.

The insecticide DEET was attenuated by 28 ± 0.4% in MFCs,

and 39 ± 0.8% in MECs with current generation, but there was

slightly higher attenuation of 43 ± 6% (MFCs) and 51 ± 7%

(MECs) under open circuit conditions, (Fig. 3D). Higher

removal without current was somewhat surprising, suggest-

ing that the non-exoelectrogenic community present in these

systems contributed to the attenuation of DEET. Based on

NaN3 tests, 30% (MFCs) and 33% (MECs) of the attenuation can

be attributed to sorption to the biofilm (Fig. 3D), which shows

that sorption was responsible for the bulk of the overall

attenuation. The limited biotransformation of DEET in both

MFCs and MECs is unexpected since previous studies have

shown DEET to be biodegradable in conventional wastewater

treatment processes (Sui et al., 2010; Okuda et al., 2008) with

removals reaching 80% (Okuda et al., 2008). The low levels of

biotransformation are likely due to growth of microorganisms

that are unable to degrade DEET versus those that develop in

CAS. In an earlier study, the removal of DEET fromwastewater

did not exceed 30% after the anaerobic treatment step of a full

scale anaerobic/anoxic/aerobic process (Xue et al., 2010).

Anaerobic digestion of sewage sludge showed no removal of

DEET (Narumiya et al., 2013).

There is general agreement between the observed sorption

and log D values for carbamazepine, atrazine and TCEP but

not DEET (Figure S1, Supporting information). DEET has a

higher log D (1.96) than TCEP (0.48) but the sorption trends

were similar. In CAS, sorption can be considered as a removal

mechanism since the biosolids are continuously wasted. In an

MFC or MEC the solids retention time (SRT) is essentially

infinite since the bacteria form a biofilm at the anode. This

implies that sorption is not a sustainablemechanism for TOrC

attenuation since the available binding sites will be exhausted

with time. An increased reliance on biotransformation will be

required for TOrC attenuation in BES.

3.2.3. Bin 3: trimethoprim and sulfamethoxazoleTrimethoprim and sulfamethoxazole were binned together as

their attenuation was dependent on mode of operation (MFC

or MEC) but independent of current generation, and NaN3 was

only effective in inactivating cells and preventing biotrans-

formation of TOrCs under aerobic conditions. Trimethoprim

showed relatively good attenuation in MFCs under closed

(70 ± 9%) and open circuit (84 ± 5%) conditions (Fig. 4A).

Attenuation in MECs was similar under both operating con-

ditions (approx. 90%) and was higher than observed in the

MFCs (Fig. 4A). However, a statistical analysis (t-test) of this

data indicates that this difference in removal was not signif-

icantly different. Reports on the biodegradation of trimetho-

prim show that this compound is not easily removed in

conventional biological systems. Perez et al. (2005) and Gobel

et al. (2007) showed no removal of trimethoprim in CAS, but

Perez et al. (2005) found complete removal under nitrifying

conditions.

The attenuation of trimethoprim due to sorption, based on

the procedures used here, was 60% inMFCs and almost 90% in

MECs (Fig. 4A). A similar trend was observed for sulfameth-

oxazole, but 100% attenuationwas observed in the presence of

NaN3 (Fig. 4B). The reason for the complete attenuation in the

presence of the biocide is not clear. It may be that NaN3 was

simply not effective at inhibiting the metabolism responsible

for attenuation of these two compounds. It is unlikely that

sorption was responsible for all of the observed attenuation,

particularly for sulfamethoxazole, since this compound is

relatively hydrophilic under the experimental conditions, has

a low tendency for sorption, and is known to be biodegradable.

Anaerobic digestion of sewage sludge has shown to signifi-

cantly reduce the concentrations of sulfamethoxazole and

trimethoprim (Narumiya et al., 2013;Martin et al., 2015). In one

study, biodegradation of sulfamethoxazole and trimethoprim

during anaerobic digestion of sewage sludge reduced their

Fig. 4 e Attenuation of (A) trimethoprim and (B) sulfamethoxazole was higher in MECs compared to MFCs.

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 764

concentrations to below quantifiable concentrations (0.46 and

0.14 mg/kg for sulfamethoxazole and trimethoprim respec-

tively) (Martın et al., 2015).

Sulfamethoxazole was completely attenuated in MECs

with and without current and in MFCs with current (Fig. 4B).

Attenuation in MFCs without current reached 86 ± 0.3%

(Fig. 4B). Sulfamethoxazole is considered to bemore amenable

to biodegradation than trimethoprim in activated sludge

systems (P�erez et al., 2005; G€obel et al., 2007) and trace con-

centrations (6 mg/L) have been shown to be completely

removed in BESs (Harnisch et al., 2013). Harnisch et al. (2013)

also showed, by analyzing the sulfamethoxazole N4-acetyl

metabolite, that the removal of sulfamethoxazole in BES did

not proceed via reformation of the parent compound. Sulfa-

methoxazole has been shown to be biodegraded under

anaerobic conditions (Carballa et al., 2007; Narumiya et al.,

2013; Martın et al., 2015). Rates of removals greater than 85%

were observed for sulfamethoxazole after anaerobic treat-

ment of sewage sludge (Carballa et al., 2007).

Although the anode potential can be fixed to the same

value in MFCs and MECs, the redox conditions within these

two types of systems can be different. Due to oxygen diffusion

through the open air cathode, dissolved oxygen in single

chamberMFCs can reach near saturation levels of (6mg/L) (Qu

et al., 2012). Oxygen is minimized in an MEC by sealing the

cathode from the atmosphere, and therefore, the redox con-

ditions are expected to be more anaerobic than MFCs. This

may be the reason for the higher attenuation of trimethoprim

Fig. 5 e Attenuation of (A) atenolol and (B) diphenh

in MECs compared to MFCs, similar to that observed for

caffeine. The current density was higher under MEC condi-

tions and resulted in a higher CE that may have enhanced the

attenuation of trimethoprim.

3.2.4. Bin 4: atenolol and diphenhydramineThe beta blocker atenolol and the antihistamine diphenhy-

dramine were binned together since the presence of NaN3

did not affect the attenuation of TOrCs in MFCs and MECs.

The attenuation of these two compounds was similar in both

MFCs and MECs under closed and open circuit conditions and

exceeded 80% in all cases (Fig. 5A and B). Atenolol has been

reported to be poorly removed with removal efficiencies of

10e55 % reported in activated sludge treatment systems

(Castiglioni et al., 2006). However, in a study that compared

the occurrence of pharmaceutically active compounds in the

sludge of different sewage sludge stabilization processes,

anaerobic digestion showed higher removals of some PhACs,

including atenolol, compared to aerobic treatments (Martın

et al., 2015). Diphenhydramine has been shown to be

removed by up to 68% in conventional municipal treatment

but with significant variability (±40%) (Du et al., 2014).

Treatment of wastewater sludge by mesophilic anaerobic

digestion showed negligible removal of diphenhydramine

(CCME, 2010).

A high attenuation when NaN3 was added was unexpected

for atenolol, as it is highly hydrophilic (log D ¼ �1.99, Table 1).

Hydrophobic interactions, adsorption of the aliphatic and

ydramine was similar in both MECs and MFCs.

Fig. 6 e Attenuation of primidone was limited under all

conditions tested.

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7 65

aromatic groups of a compound within the lipophilic cell

membrane of themicroorganisms, (Ternes et al., 2004a) of this

chemical within the biofilm would likely be limited and the

bulk of the compound would be expected to remain in solu-

tion. Diphenhydramine is themost hydrophobic compound of

all the selected TOrCs in this study, and thus hydrophobic

interactions would be expected to dominate. Du et al. (2014)

showed that removal of diphenhydramine increased from

20% in the effluent of an aerobic treatment system to 99%

when this system was linked to a constructed wetland.

3.2.5. Bin 5: primidoneThe behavior of primidone was unique among all of the

compounds as it showed limited attenuation under all

experimental conditions (Fig. 6). The low attenuation of pri-

midone in MFCs and MECs demonstrated that they have little

capacity to remove this compound. The attenuation attrib-

uted to sorption was also low due to its relatively low log D of

0.4. The limited attenuation of this chemical is not unexpected

as primidone is often used as a tracer compound to follow

transport of wastewater components to potable water sup-

plies (Krasner et al., 2006; Chen et al., 2009) and subsurface

systems (Hoppe-Jones et al., 2010). However, 30e50 % of pri-

midone was reported to be removed in an activated sludge

treatment process (Wick et al., 2009). The removal was

attributed to biodegradation since there was no significant

difference between the sorbed fraction in the sludge and sol-

uble fraction. In contrast, batch experiments using sludge

from the same plant showed no significant removal of pri-

midone over 48 h. The authors concluded that the sludge may

have lost its ability to transform primidone. Less than 2%

removal efficiency was obtained for primidone at an initial

concentration of 1.7 mg/L from a wastewater treated in an

anaerobic membrane bioreactor (Monsalvo et al., 2014).

4. Conclusions

There was considerable attenuation of several TOrCs due to

sorption in the biofilm-based systems, with enhanced

biotransformation observed for some of the compounds.

However, no clear trends between the physico-chemical

properties of the compounds and their attenuation in BESs

were observed. Caffeine and sulfamethoxazole showed high

levels of biotransformation relative to the other TOrCs tested,

consistent with reports in the literature for other types of

biological wastewater treatment systems. However, TCEP

was also relatively well biotransformed, which has not been

observed in these other treatment systems. Based on the

redox conditions of MFCs and MECs, the order for bio-

transformability of the compounds as determined in these

tests is as follows: caffeine > sulfamethoxazole > TCEP >carbamazepine > atrazine > trimethoprim > DEET for MFCs;

and sulfamethoxazole > caffeine > TCEP > trimethoprim >carbamazepine > atrazine > DEET for MECs. Atenolol,

diphenhydramine and primidone showed no measurable

biotransformation. Primidone was not well attenuated under

any of the conditions tested, which supports its use as a

tracer compound to assess the water quality of wastewater

impacted sources. BESs showed comparable rates of attenu-

ation as compared to conventional wastewater treatment,

but a multi-barrier approach, such as BES followed by

membrane treatment, ozonation (could also be used as a pre-

treatment) or other advanced treatment would be needed for

improved removals of TOrCs prior to water reuse applica-

tions. Considering the trade-off that exists between

maximum energy recovery and overall COD removal it would

be reasonable to operate BESs to optimize the efficiency of

energy recovery rather than to achieve complete COD

removal. High current densities would enable more efficient

energy recovery, decrease the rate of COD removal by non-

exoelectrogenic microorganisms, and shorten the HRT. The

decreased rate of removal by heterotrophs and shorter HRTs

may likely have an impact on overall TOrC attenuation as the

results from these findings indicate that the non-

exoelectrogenic community plays a role in the removal of

these compounds. In this case a second treatment step, such

as incorporating membrane separation within the process or

an anaerobic fluidized membrane bioreactor (Ren et al.,

2014b) would be required as an additional barrier for the

attenuation of TOrCs. Overall, BESs provide TOrC attenuation

comparable to CAS, with the added advantage of energy

production.

Acknowledgments

This work was supported by a PhD fellowship award (C.W.),

discretionary investigator funds (P.S.) from the King Abdullah

University of Science and Technology (KAUST), and Award

KUS-I1-003-13 from KAUST.

Appendix A. Supplementary data

Supplementary data related to this article can be found at

http://dx.doi.org/10.1016/j.watres.2015.01.013.

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 766

r e f e r e n c e s

Alidina, M., Hoppe-Jones, C., Yoon, M., Hamadeh, A.F., Li, D.,Drewes, J.E., 2014. The occurrence of emerging trace organicchemicals in wastewater effluents in Saudi Arabia. Sci. TotalEnviron. 478, 152e162.

Ambler, J.R., Logan, B.E., 2011. Evaluation of stainless steelcathodes and a bicarbonate buffer for hydrogen production inmicrobial electrolysis cells using a new method for measuringgas production. Int. J. Hydrog. Energy 36 (1), 160e166.

Balch, W., Fox, G., Magrum, L., Woese, C., Wolfe, R., 1979.Methanogens: reevaluation of a unique biological group.Microbiol. Mol. Bio. R. 43 (2), 260.

Batt, A.L., Bruce, I.B., Aga, D.S., 2006. Evaluating the vulnerabilityof surface waters to antibiotic contamination from varyingwastewater treatment plant discharges. Environ. Pollut. 142(2), 295e302.

Benjamin, M.M. (Ed.), 2002. Water Chemistry. McGraw-Hill Seriesin Water Resources and Environmental Engineering.

Benotti, M., Trenholm, R., Vanderford, B., Holady, J., Stanford, B.,Snyder, S., 2008. Pharmaceuticals and endocrine disruptingcompounds in US drinking water. Environ. Sci. Technol. 43 (3),597e603.

Bernhard, M., Muller, J., Knepper, T.P., 2006. Biodegradation ofpersistent polar pollutants in wastewater: Comparison of anoptimised lab-scale membrane bioreactor and activatedsludge treatment. Water Res. 40 (18), 3419e3428.

Buerge, I.J., Poiger, T., Muller, M.D., Buser, H.-R., 2003. Caffeine, ananthropogenic marker for wastewater contamination ofsurface waters. Environ. Sci. Technol. 37 (4), 691e700.

Call, D., Logan, B.E., 2008. Hydrogen production in a singlechamber microbial electrolysis cell lacking a membrane.Environ. Sci. Technol. 42 (9), 3401e3406.

Carballa, M., Fink, G., Omil, F., Lema, J.M., Ternes, T., 2008.Determination of the solidewater distribution coefficient (Kd)for pharmaceuticals, estrogens and musk fragrances indigested sludge. Water Res. 42 (1e2), 287e295.

Carballa, M., Omil, F., Ternes, T., Lema, J.M., 2007. Fate ofpharmaceutical and personal care products (PPCPs) duringanaerobic digestion of sewage sludge. Water Res. 41 (10),2139e2150.

Castiglioni, S., Bagnati, R., Fanelli, R., Pomati, F., Calamari, D.,Zuccato, E., 2006. Removal of pharmaceuticals in sewagetreatment plants in Italy. Environ. Sci. Technol. 40 (1),357e363.

CCME, 2010. Canadian Council of Ministers of the Environment.Emerging Substances of Concern in Biosolids: Concentrationsand Effects of Treatment Processes. Final report e Fieldsampling program CCME Project #447e2009. Submitted byHydromantis, Inc., University of Waterloo and TrentUniversity.

Chen, B., Nam, S.-N., Westerhoff, P.K., Krasner, S.W., Amy, G.,2009. Fate of effluent organic matter and DBP precursors in aneffluent-dominated river: a case study of wastewater impacton downstream water quality. Water Res. 43 (6), 1755e1765.

Cheng, S., Liu, H., Logan, B.E., 2006. Power densities usingdifferent cathode catalysts (Pt and CoTMPP) and polymerbinders (Nafion and PTFE) in single chamber microbial fuelcells. Environ. Sci. Technol. 40, 364e369.

Clara, M., Strenn, B., Kreuzinger, N., 2004. Carbamazepine as apossible anthropogenic marker in the aquatic environment:investigations on the behaviour of Carbamazepine inwastewater treatment and during groundwater infiltration.Water Res. 38 (4), 947e954.

Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personalcare products in the environment: agents of subtle change?Environ. Health Persp 907e938.

Drewes, J., Heberer, T., Reddersen, K., 2002. Fate ofpharmaceuticals during indirect potable reuse. Water Sci.Technol. 46 (3), 73e80.

Drewes, J.E., Bellona, C., Oedekoven, M., Xu, P., Kim, T.-U., Amy, G.,2005. Rejection of wastewater-derivedmicropollutants in high-pressure membrane applications leading to indirect potablereuse. Environ. Prog. 24 (4), 400e409.

Du, B., Price, A.E., Scott, W.C., Kristofco, L.A., Ramirez, A.J.,Chambliss, C.K., Yelderman, J.C., Brooks, B.W., 2014. Comparisonof contaminants of emerging concern removal, discharge, andwater quality hazards among centralized and on-sitewastewatertreatment system effluents receiving commonwastewaterinfluent. Sci. Total Environ. 466e467, 976e984.

Fatta-Kassinos, D., Meric, S., Nikolaou, A., 2011. Pharmaceuticalresidues in environmental waters and wastewater: currentstate of knowledge and future research. Anal. Bioanal. Chem.399 (1), 251e275.

Fries, E., Puttmann, W., 2003. Monitoring of the threeorganophosphate esters TBP, TCEP and TBEP in river water andgroundwater (Oder, Germany). J. Environ. Monit. 5 (2), 346e352.

Froehner, S., Piccioni, W., Machado, K., Aisse, M., 2011. Removalcapacity of caffeine, hormones, and bisphenol by aerobic andanaerobic sewage treatment. Water Air Soil Poll. 216 (1e4),463e471.

Ghosh, P.K., Philip, L., 2004. Atrazine degradation in anaerobicenvironment by a mixed microbial consortium. Water Res. 38(9), 2277e2284.

Glassmeyer, S.T., Furlong, E.T., Kolpin, D.W., Cahill, J.D.,Zaugg, S.D., Werner, S.L., Meyer, M.T., Kryak, D.D., 2005.Transport of chemical and microbial compounds from knownwastewater discharges: potential for use as indicators ofhuman fecal contamination. Environ. Sci. Technol. 39 (14),5157e5169.

G€obel, A., McArdell, C.S., Joss, A., Siegrist, H., Giger, W., 2007. Fateof sulfonamides, macrolides, and trimethoprim in differentwastewater treatment technologies. Sci. Total Environ. 372(2e3), 361e371.

Harnisch, F., Gimkiewicz, C., Bogunovic, B., Kreuzig, R., Schr€oder, U.,2013. On the removal of sulfonamides usingmicrobialbioelectrochemical systems. Electrochem. Commun. 26, 77e80.

He, Z., Huang, Y., Manohar, A.K., Mansfeld, F., 2008. Effect ofelectrolyte pH on the rate of the anodic and cathodic reactionsin an air-cathode microbial fuel cell. Bioelectrochemistry 74(1), 78e82.

Heidrich, E.S., Dolfing, J., Scott, K., Edwards, S.R., Jones, C.,Curtis, T.P., 2013. Production of hydrogen from domesticwastewater in a pilot-scale microbial electrolysis cell. Appl.Microbiol. Biotechnol. 97 (15), 6979e6989.

Hirsch, R., Ternes, T., Haberer, K., Kratz, K.-L., 1999. Occurrence ofantibiotics in the aquatic environment. Sci. Total Environ. 225(1e2), 109e118.

Hoppe-Jones, C., Oldham, G., Drewes, J.E., 2010. Attenuation of totalorganic carbon and unregulated trace organic chemicals in U.S.riverbank filtration systems. Water Res. 44 (15), 4643e4659.

Hyland, K.C., Dickenson, E.R.V., Drewes, J.E., Higgins, C.P., 2012.Sorption of ionized and neutral emerging trace organiccompounds onto activated sludge from different wastewatertreatment configurations. Water Res. 46 (6), 1958e1968.

Joss, A., Keller, E., Alder, A.C., G€obel, A., McArdell, C.S., Ternes, T.,Siegrist, H., 2005. Removal of pharmaceuticals and fragrancesin biological wastewater treatment. Water Res. 39 (14),3139e3152.

Kobayashi, H., Maeda, M., Anraku, Y., 1977. Membrane-boundadenosine triphosphatase of Escherichia coli III. Effects ofsodium azide on the enzyme functions. J. Biochem. 81 (4),1071e1077.

Krasner, S.W., Pastor, S.J., Garcia, E.A., 2006. Measurement of thePharmaceutical Primidone as a Conservative Tracer of

wat e r r e s e a r c h 7 3 ( 2 0 1 5 ) 5 6e6 7 67

Wastewater Influences in Drinking Water Supplies. AWWA,Denver, CO, USA.

Levine, A.D., Asano, T., 2004. Peer reviewed: recoveringsustainable water from wastewater. Environ. Sci. Technol. 38(11), 201Ae208A.

Liu, C., Huang, X., Wang, H., 2008. Start-up of a membranebioreactor bioaugmented with genetically engineeredmicroorganism for enhanced treatment of atrazine containingwastewater. Desalination 231 (1e3), 12e19.

Liu, H., Logan, B.E., 2004. Electricity generation using an air-cathode single chamber microbial fuel cell in the presence andabsence of a proton exchange membrane. Environ. Sci.Technol. 38 (14), 4040e4046.

Logan, B.E., Regan, J.M., 2006. Microbial fuel cellsdchallenges andapplications. Environ. Sci. Technol. 40 (17), 5172e5180.

Logan, B.E., Aelterman, P., Hamelers, B., Rozendal, R.,Schr€oder, U., Keller, J., Freguia, S., Verstraete, W., Rabaey, K.,2006. Microbial fuel cells: methodology and technology.Environ. Sci. Technol. 40 (17), 5181e5192.

Logan, B.E., Call, D., Cheng, S., Hamelers, H.V.M., Sleutels, T.H.J.A.,Jeremiasse, A.W., Rozendal, R.A., 2008. Microbial electrolysiscells for high yield hydrogen gas production from organicmatter. Environ. Sci. Technol. 42 (23), 8630e8640.

Logan, B.E., Cheng, S., Watson, V., Estadt, G., 2007. Graphite fiberbrush anodes for increased power production in air-cathodemicrobial fuel cells. Environ. Sci. Technol. 41 (9), 3341e3346.

Martın, J., Santos, J.L., Aparicio, I., Alonso, E., 2015.Pharmaceutically active compounds in sludge stabilizationtreatments: anaerobic and aerobic digestion, wastewaterstabilization ponds and composting. Sci. Total Environ.503e504, 97e104.

Martınez-Hern�andez, V., Meffe, R., Herrera, S., Arranz, E., deBustamante, I., 2015. Corrigendum to “Sorption/desorption ofnon-hydrophobic and ionisable pharmaceutical and personalcare products from reclaimed water onto/from a naturalsediment” Sci Total Environ 472 (2014) 273e281. Sci. TotalEnviron. 505 (0), 1232e1233.

Meyer, J., Bester, K., 2004. Organophosphate flame retardants andplasticisers in wastewater treatment plants. J. Environ. Monit.6 (7), 599e605.

Monsalvo, V.M., McDonald, J.A., Khan, S.J., Le-Clech, P., 2014.Removal of trace organics by anaerobic membranebioreactors. Water Res. 49, 103e112.

Monteith, H.D., Parker, W.J., Bell, J.P., Melcer, H., 1995. Modelingthe fate of pesticides in municipal wastewater treatment.Water Environ. Res. 67 (6), 964e970.

Narumiya, M., Nakada, N., Yamashita, N., Tanaka, H., 2013. Phasedistribution and removal of pharmaceuticals and personalcare products during anaerobic sludge digestion. J. HazardMater. 260, 305e312.

Okuda, T., Kobayashi, Y., Nagao, R., Yamashita, N., Tanaka, H.,Tanaka, S., Fujii, S., Konishi, C., Houwa, I., 2008. Removalefficiency of 66 pharmaceuticals during wastewater treatmentprocess in Japan. Water Sci. Technol. 57 (1), 65e71.

Patil, S., Harnisch, F., Schr€oder, U., 2010. Toxicity response ofelectroactive microbial biofilmsdA decisive feature forpotential biosensor and power source applications.ChemPhysChem 11, 2834e2837.

P�erez, S., Eichhorn, P., Aga, D.S., 2005. Evaluating thebiodegradability of sulfamethazine, sulfamethoxazole,sulfathiazole, and trimethoprim at different stages of sewagetreatment. Environ. Toxicol. Chem. 24 (6), 1361e1367.

Prabhudessai, V., Ganguly, A., Mutnuri, S., 2009. Effect of caffeineand saponin on anaerobic digestion of food waste. Ann.Microbiol. 59 (4), 643e648.

Qu, Y., Feng, Y., Wang, X., Logan, B.E., 2012. Use of a coculture toenable current production by Geobacter sulfurreducens. Appl.Environ. Microbiol. 78 (9), 3484e3487.

Quinn, B., Gagn�e, F., Blaise, C., 2008. An investigation into theacute and chronic toxicity of eleven pharmaceuticals (andtheir solvents) found in wastewater effluent on the cnidarian,Hydra attenuata. Sci. Total Environ. 389 (2e3), 306e314.

Rabaey, K., Rozendal, R.A., 2010. Microbialelectrosynthesisdrevisiting the electrical route for microbialproduction. Nat. Rev. Microbiol. 8 (10), 706e716.

Ren, L., Ahn, Y., Logan, B.E., 2014b. A two-stage microbial fuel celland anaerobic fluidized bed membrane bioreactor (MFC-AFMBR) system for effective domestic wastewater treatment.Environ. Sci. Technol. 48 (7), 4199e4206.

Ren, L., Zhang, X., He, W., Logan, B.E., 2014a. High currentdensities enable exoelectrogens to outcompete aerobicheterotrophs for substrate. Biotechnol. Bioeng. http://dx.doi.org/10.1002/bit.25290.

Sayess, R.R., Saikaly, P.E., El-Fadel, M., Li, D., Semerjian, L., 2013.Reactor performance in terms of COD and nitrogen removaland bacterial community structure of a three-stage rotatingbioelectrochemical contactor. Water Res. 47 (2), 881e894.

Scheytt, T., Mersmann, P., Lindst€adt, R., Heberer, T., 2005.Determination of sorption coefficients of pharmaceuticallyactive substances carbamazepine, diclofenac, and ibuprofen,in sandy sediments. Chemosphere 60 (2), 245e253.

Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M., Wiley, J.,2003. Environmental Organic Chemistry. Wiley Online Library.

Shehab, N., Li, D., Amy, G.L., Logan, B.E., Saikaly, P.E., 2013.Characterization of bacterial and archaeal communities in air-cathode microbial fuel cells, open circuit and sealed-offreactors. Appl. Microbiol. Biotechnol. 97 (22), 9885e9895.

Snyder, S.A., Adham, S., Redding, A.M., Cannon, F.S., DeCarolis, J.,Oppenheimer, J., Wert, E.C., Yoon, Y., 2007. Role ofmembranes and activated carbon in the removal of endocrinedisruptors and pharmaceuticals. Desalination 202 (1),156e181.

Snyder, S.A., Westerhoff, P., Yoon, Y., Sedlak, D.L., 2003.Pharmaceuticals, personal care products, and endocrinedisruptors in water: implications for the water industry.Environ. Eng. Sci. 20 (5), 449e469.

Stackelberg, P.E., Furlong, E.T., Meyer, M.T., Zaugg, S.D.,Henderson, A.K., Reissman, D.B., 2004. Persistence ofpharmaceutical compounds and other organic wastewatercontaminants in a conventional drinking-water-treatmentplant. Sci. Total Environ. 329 (1e3), 99e113.

Sui, Q., Huang, J., Deng, S., Yu, G., Fan, Q., 2010. Occurrence andremoval of pharmaceuticals, caffeine and DEET in wastewatertreatment plants of Beijing, China. Water Res. 44 (2), 417e426.

Ternes, T., Joss, A., Siegrist, H., 2004a. Peer reviewed: scrutinizingpharmaceuticals and personal care products in wastewatertreatment. Environ. Sci. Technol. 38 (20), 392Ae399A.

Ternes, T.A., Herrmann, N., Bonerz, M., Knacker, T., Siegrist, H.,Joss, A., 2004b. A rapid method to measure the solidewaterdistribution coefficient (Kd) for pharmaceuticals and muskfragrances in sewage sludge. Water Res. 38 (19), 4075e4084.

Wick, A., Fink, G., Joss, A., Siegrist, H., Ternes, T.A., 2009. Fate ofbeta blockers and psycho-active drugs in conventionalwastewater treatment. Water Res. 43 (4), 1060e1074.

Xue, W., Wu, C., Xiao, K., Huang, X., Zhou, H., Tsuno, H.,Tanaka, H., 2010. Elimination and fate of selected micro-organic pollutants in a full-scale anaerobic/anoxic/aerobicprocess combined with membrane bioreactor for municipalwastewater reclamation. Water Res. 44 (20), 5999e6010.

Yangali-Quintanilla, V., Sadmani, A., McConville, M., Kennedy, M.,Amy, G., 2009. Rejection of pharmaceutically active compoundsand endocrine disrupting compounds by clean and foulednanofiltration membranes. Water Res. 43 (9), 2349e2362.

Zhang, Y., Geißen, S.-U., Gal, C., 2008. Carbamazepine anddiclofenac: removal in wastewater treatment plants andoccurrence in water bodies. Chemosphere 73 (8), 1151e1161.