Accepted Manuscript

Bio-electrochemical Degradation of Paracetamol in a Microbial Fuel Cell-Fen-

ton System

Lijuan Zhang, Xuejiao Yin, Sam Fong Yau Li

PII: S1385-8947(15)00547-1
DOI: http://dx.doi.org/10.1016/j.cej.2015.04.065
Reference: CEJ 13554

To appear in: Chemical Engineering Journal

Received Date: 8 January 2015

Revised Date: 28 March 2015
Accepted Date: 11 April 2015

Please cite this article as: L. Zhang, X. Yin, S.F.Y. Li, Bio-electrochemical Degradation of Paracetamol in a
Microbial Fuel Cell-Fenton System, Chemical Engineering Journal (2015), doi: http://dx.doi.org/10.1016/j.cej.
2015.04.065

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Bio-electrochemical Degradation of Paracetamol in a Microbial Fuel
Cell-Fenton System

Lijuan Zhang a, Xuejiao Yin a, Sam Fong Yau Li a,b*

a
Department of Chemistry, Faculty of Science, National University of Singapore, Singapore

117543, Singapore

b
NUS Environmental Research Institute, National University of Singapore, Singapore 117411,

Singapore

*
Corresponding author: Tel.: +65 65162681; fax: +65 67791691; E-mail address:

chmlifys@nus.edu.sg (S.F.Y. Li).

2

Abstract
Paracetamol (PAM) has emerged as an important wastewater contaminant due to its wide usage. In

this study, Fenton reactions were introduced to microbial fuel cells (MFCs) for bio-electrochemical

degradation of PAM without external power supply. Dual-chamber MFC reactors were employed:

in the anode chambers, bio-electrons were released by oxidizing biodegradable pollutants in

low-strength real domestic wastewater; in the cathode chambers, the input fluxes of bio-electrons

from the anode could promote the yield of free radical ●OH by facilitating the regeneration of iron

source. Compared to conventional Fenton reactions in which no bio-electrons were transferred, our

approach exhibited the advantage that no continuous addition of Fenton reagents was required in

MFC-Fenton system. Significant changes in the UV-vis spectra of catholytes from an

MFC-powered process strongly indicated that bio-electricity input played an important role in PAM

degradation. The performance of electro-Fenton system was in good agreement with the

bio-electricity output-capacity from MFC reactor under expected optimal conditions. At total iron

concentration of 5 mg L-1, initial pH value of 2.0 and external resistance of 20 Ω, the highest PAM

degradation efficiency of 70% was attained within 9 h. A 25% PAM could be completely

mineralized while the majority was mainly converted to intermediate metabolites of p-nitrophenol

via p-aminophenol and to less hazardous dicarboxylic/carboxylic acids. These results suggested that

MFC-Fenton could be applied as an energy-saving and efficient approach to PAM-containing

wastewater treatment, and further to non-biocompatible pharmaceuticals degradation in aquatic

environment.

Key words: Paracetamol, Microbial fuel cell, Fenton, Bio-electrochemistry

 3

1. Introduction

Pharmaceutical-contaminated wastewater is newly emerging as a public concern in recent years.

Widespread survey and evaluation indicated that pharmaceutically active compounds are widely

found in natural surface streams, groundwater, effluent from wastewater treatment plants and

supplied drinking water [1-3]. Although usually low in strength (generally ng L-1 to µg L-1), these

pharmaceuticals are commonly regarded as potential hazards due to their accumulation in living

tissues, which in turn will trigger chain reactions and produce unwanted biological effects on

aquatic creatures and human beings [4, 5]. Paracetamol (PAM), clinical trade names of

acetaminophen (ACTP) or acetyl-para-aminophenol (APAP), is a common over-the-counter (OTC)

analgesic and antipyretic. While generally safe for use at the recommended dose, even small

overdoses may lead to fatal liver damage [6]. With its widespread application in pharmaceutical

industries and daily life, PAM is becoming a key environmental research topic.

Previous studies on removal of PAM from wastewater mainly focused on advanced oxidation

processes (AOPs). Hydroxyl radicals (
OH) are responsible for the major
OH-PAM reaction [7].

One commonly applied AOP method is photo-degradation catalyzed by TiO2. Various active

species besides hydroxyl radicals, including valence band holes (h+), conductive band electrons (e-),

superoxide ions (O2
-), hydrogen peroxide radicals (HO2
) and hydrogen peroxide (H2O2), can be

generated to mineralize PAM when an irradiation energy is higher than the band gap energy of TiO2

[6]. Another extensively discussed method is Fenton oxidation process where
OH can be generated

in-situ from the classical reaction between iron (II) (Fe2+) and H2O2. Recent advances made in the

improvement of conventional Fenton technology have led to various Fenton-like systems with high
4

efficiency, such as electron-Fenton by applying external voltage, ultraviolet (UV)/photo-Fenton by

exploiting light irradiation and ultrasonication (US)/sono-Fenton by introducing

ultrasonication/sonolysis [8-10]. However, external energy (e.g. electricity, light and ultrasonic

wave) input causes extra operating cost.

Microbial fuel cell (MFC) technology has been well developed in harvesting electric energy and

powering miniature devices such as sediment-based systems and a data collector in remote ocean

[11, 12]. It was first introduced to Fenton reactions by Zhu and Ni [13] for simultaneous electricity

generation and p-nitrophenol (PNP) degradation with no external power supply. Then MFC was

integrated with Fenton reactions to dispose the bio-refractory and/or toxic compounds, such as

various organic dyes and arsenite [14-20]. These findings shed light on the feasibility of PAM

removal by MFC-based electro-Fenton without additional electricity requirements. In the past

decade, the power output of MFCs has increased to as high as several watt per square meter of

surface area of electrode [21], further making it practical to use MFCs as power sources for

electro-Fenton systems. In an MFC, the electrochemically active microorganisms (electricigens)

grown on the anode act as biological catalysts to lower the electrode overpotential effectively,

promote anodic reactions and then facilitate electron transfer to the cathode on which oxidized

pollutants are reduced [22]. The implementation of the separator, most commonly a membrane

between the two electrodes, prevents the bio-activity of electricigens in the anode from being

inhibited by the pollutants in the cathode. The performance of an MFC was affected by various

factors including the degradation of substrate by electricigens grown on the anode, transfer of

electron in the electrical circuit, mass transfer in bulk solution, performance of the separator,

electrode and electrolyte etc. [23, 24]. Among which, cathodic process, such as oxygen reduction
 5

(ORR) catalyzed by precious Pt, has the most significant roles and accounts for more than 50% of

MFC’s capital cost [25, 26]. Therefore, there is a great demand for exploring cheaper and

high-performance cathodic reactions to achieve high-efficiency degradation and removal of target

species.

Electro-catalytic degradation of PAM in MFCs has rarely been studied and reported to date. Thus

this project aims at exploring the feasibility of enhanced PAM degradation by an MFC-Fenton

system. Some key factors influencing the PAM degradation efficiency, such as iron dosage, initial

pH and external resistance, were investigated. The possible catalytic roles of iron source and

bio-electrons as contributors to the bio-electrochemically reactions were discussed. To better

understand the stepwise degradation of PAM, the pathway of bio-electrochemical reaction was

investigated and kinetic mechanisms were proposed based on the intermediates identified.

2. Materials and Methods

2.1. MFC-Fenton system configuration

A dual-chamber MFC reactor was constructed with a cathodic working volume of 216.0 mL and

an anodic working volume of 108.0 mL (Fig. 1). The anode consisted of three highly porous

graphite felts (6.0 cm length × 5.5 cm width × 1.0 cm thickness), which could provide sufficient

surface area for the attachment and growth of electricigens. The cathode was a single graphite plate

with a smaller projected surface area of 15 cm2 and lower porosity and specific surface area as

compared to the anode. Proton exchange membrane (PEM, 6 cm × 10 cm cross-sectional area,

Nafion-117, DuPont, USA) was installed as a separator to prevent potential diffusion of dissolved
6

oxygen (DO, aerated with gaseous oxygen from air) as well as the transfer of iron (III)/iron (II)

(Fe3+/Fe2+) ions (sourced from FeSO4·7H2O added directly) from cathode to anode chamber. A data

acquisition system (Adam-4017, Advantech Co., Ltd., China) was coupled to a reference electrode

(Ag/AgCl, CHI 111, USA) to monitor the voltage output and anode and/or cathode potentials. The

whole circuit was connected via titanium wires (> 99.9%). All established MFC-Fenton systems

were maintained in air-conditioned rooms controlled at 25 ℃.

Fig. 1. Schematic diagram of the MFC-Fenton system (PEM = Proton Exchange Membrane).

2.2. Start up and Experiments

 7

A successful start-up, indicated by constant and stable power output, could be accomplished by

the following step-by-step operations. Firstly, the anode chamber of a dual-chamber MFC was

inoculated with a selectively enriched mixed microbial culture from the effluent of a

single-chamber air cathode MFC that has been operated for more than two years. Secondly, the

MFC was sealed and then kept at open circuit for 48 h to achieve an anaerobic condition and

facilitate the growth of electricigens onto the anode. Subsequently, these electricigens were fed with

50 mM phosphate buffer saline (PBS)-buffered medium (pH = 7.0) at batch mode with a hydraulic

retention time of 48 h. The 50 mM PBS buffer contained (mg L-1): Na2HPO4·2H2O, 5736 and

NaH2PO4·H2O, 2452. Dissolved oxygen was purged by nitrogen sparging from the PBS-buffered

anodic medium containing (mg L-1): CH3COONa, 500 (approximately COD 390, as an electron

donor); NaCl, 5850; NH4Cl, 310; KCl, 130; mineral salts medium (MSM) and vitamins [27]. The

cathode chamber was filled with the same PBS buffer as that used in anode chamber. Air was

bubbled into the cathode chamber at a flow rate of 1.0 L h-1 to supply dissolved oxygen as an

electron acceptor. Then the system was loaded with different external resistances (ERs) decreased

from +∞ (open circuit), 1000 Ω, 800 Ω, 600 Ω, 390 Ω to 180 Ω and maintained for 48 h at each

resistance value. The ER value was fixed at 180 Ω in the remaining period of experiment unless

otherwise specified. A rapid and effective enrichment of electrochemically active bacteria could be

obtained within two weeks.

After start-up, the microbial fuel (electron donor) was switched from acetate in synthetic medium

to multiple organic pollutants (equivalent to five-day biochemical oxygen demand, BOD5 of 220 ±

34 mg L-1) in real domestic wastewater (pH = 7.02 ± 0.03) collected from Ulu Pandan Reclamation

Plant (Singapore). After a 48-h acclimation of electrigens to real domestic wastewater, the PBS in
8

cathode chamber was replaced with deionized (DI) water. Both the anolyte and catholyte were

modified by adding 5850 mg L-1 NaCl to enhance the electric conductivity of solution before

spiking into the MFC system. Dissolved oxygen, an essential substance to produce free hydroxyl

radicals, was continuously provided by air bubbling at a flow rate of 1.0 L h-1.

Three experiments were carried out to examine the effects of different operating parameters on

PAM degradation corresponding to total iron dosage, solution pH and external resistance at a fixed

initial PAM concentration of 10 mg L-1. In the first experiment, different iron dosage, i.e. total iron

concentration of [Fe]Tot = 0 mg L-1, 1 mg L-1, 2 mg L-1, 5 mg L-1, 10 mg L-1 and 15 mg L-1, were

tested by directly applying FeSO4·7H2O into the cathode chamber of an MFC. The solution pH in

the cathode chambers was adjusted to a typical acidic value of 2.5 by the addition of HCl. An

external resistance of 180 Ω, whose value was close to the internal resistance of the system

determined by polarization curves (Fig. S1), was loaded to ensure a maximum power output from

the MFC. In the second experiment, acidic pH values, ranging from 3.5 to 1.5 at 0.5 intervals, were

studied at optimal iron dosage determined in the previous test. The external resistance remained

unchanged. In the third experiment, enhanced degradation of PAM was proved by the electron

fluxes introduced from the bio-anode to the cathode where electrochemically reactions occurred.

The external resistance was reduced from high level (1000 Ω) to medium level (180 Ω), and

eventually lowered to 20 Ω. Iron source and initial solution pH were operated at respective optimal

conditions.

All chemicals were purchased from Sigma-Aldrich (USA) and prepared in DI water (Milli-Q,

Academic system, Millipore Co., USA). Each experiment was carried out in parallel and samples

were taken in triplicate.

 9

2.3. Analyses and calculations

The output voltage (U in V) across the external resistance was recorded at 5 s intervals. The

current density (I/Scat in mA m-2) and power density (P/Scat in mW m-2) based on the surface area of

cathode plate (Scat in m2) were calculated as I/Scat = U/(RScat) and P/Scat = UI/Scat (R = external

resistance value in Ω). Polarization and power density curves as a function of current density were

obtained by changing the external resistance in 18 steps from +∞ (open circuit) to 20 Ω and

stabilizing the voltage for at least 10 min at each step. The internal resistance (IR) could then be

determined by linear or polynomial fit via these curves.

All solution samples were filtered through 0.20 µm cellulose acetate membrane filters (Minisart,

Sartorius Stedim Biotech GmbH, Germany). The general evolution of the PAM degradation was

monitored by UV-Visible spectrum on a Hach DR 5000 UV-Vis spectroscopy, with scanning

wavelength from 190 nm to 400 nm. An Agilent ZORBAX Eclipse Plus C18 column (2.1 mm ×

100 mm, 3.5 µm, 25 ℃) on a HPLC (Ultimate 3000, Dionex, CA, USA) coupled with

quadrupole-ion trap or time-of-flight mass spectrometers (QTRAP 5500 or QTOF 5600, AB Sciex,

DC, USA), was used for the separation and quantitation of PAM or identification of the possible

metabolites in solution. Mobile phase A was 0.1% formic acid in DI water. Mobile phase B was 0.1%

formic acid in acetonitrile. A typical gradient was: 0 min, 2% B; 4 min, 2% B; 6 min, 90% B; 8 min,

90% B; 8.1 min, 2% B; and 13 min, 2% B. Flow rate was 0.2 mL min-1. To quantify PAM, samples

were ionized in positive mode using electrospray ionization (ESI+) and detected with two

multi-reaction monitoring (MRM) transitions (m/z): 152-110 and 152-65 by QTRAP mass

spectrometer. The concentrations of unknown samples were quantified using an external calibration
10

curve plotted from seven standard PAM solutions (0 µg L-1, 1 µg L-1, 5 µg L-1, 10 µg L-1, 25 µg L-1,

50 µg L-1 and 100 µg L-1). Intermediate samples (catholyte mixtures) were qualitatively scanned in

both positive (ESI+) and negative (ESI-) modes to detect the possible metabolites by QTRAP mass

spectrometer. Apparent rate constant (kapp) for PAM was determined as

Ct
− ln = k app t (1)
C0

where Ct and C0 are the concentrations of PAM at time t and time zero, respectively. t is the

reaction time.

A Shimadzu TOC-VWS Total Organic Carbon Analyzer (Japan) was used to analyze TOC

concentration in catholyte. The concentration of ammonia was measured as NH4+-N by Standard

Lachat Method from US environmental protection agency (EPA). Biodegradable organic

compounds in real domestic wastewater were measured as BOD5 according to EPA Standard

Method 5210 B.

Quantification of total and respective ion species of Fe3+/Fe2+ was achieved by the

1,10-phenanthroline colorimetric methods, according to ISO 6332. ● OH was monitored by a spin

trap using 100 µM N,N-dimethyl-p-nitrosoaniline (RNO) and the bleaching of the yellow color was

observed at 440 nm [28]. H2O2 was analyzed by a hydrogen peroxide test kit (Model HYP-1, Hach,

USA) using ammonium molybdate and sulfate as reagents to develop a blue color, and thence

titrating with sodium thiosulfate as a stabilizer. Computational chemical equilibrium was modelled

by environmental research software MINEQL (Version 4.6, USA) to investigate the speciation and

distribution patterns of Fe2+/Fe3+ in aquatic solutions.

 11

3. Results and Discussion

3.1. Performance of MFC-Fenton systems

The degradation of PAM was monitored by two peaks centered at 199 nm and 243 nm which

corresponded to the π-π* transition and n-π* transition of the C=O group respectively using UV-vis

spectrometer for a period of 12 h. Input fluxes of bio-electrons are efficient in the enhanced

degradation of PAM as shown in Fig. 2. As electro-Fenton reactions proceeded, the UV-vis spectra

of catholytes from a closed-circuit MFC, where an average power density of 217.27 ± 23.24 mW m-2

was obtained at a current density of 757.41 ± 65.47 mA m-2, presented significant changes with

reaction time. The intensity of the two peaks dropped gradually from 0.65 to 0.42 as electro-Fenton

reations proceeded within 12 h, indicating that the concentration of PAM was reduced in the

MFC-Fenton system. Nevertheless, there was only a slight difference between those absorption peak

intensities of catholytes from an open-circuit MFC. Upon extended electro-Fenton degradation, there

was a newly emerged absorption peak intensified at 320 nm. This absorption peak, which was also

reported in the degradation course of PAM such as photocatalysis process [29], was most likely due

to the presence of intermediate product of PNP. The absorption characteristic (n-π* transition) of

this phenolic compound was strongly influenced by a batochromic shift due to the NO2 substituent

on the benzene ring. These observed growing peaks at 320 nm, representing the possible

intermediates with different molecular structures from PAM, indicate the undergoing

bio-electrochemically process of PAM degradation in an MFC-Fenton system.

12

Fig. 2. Evolution of UV-vis spectra for PAM degradation in MFC-Fenton systems at (A) close

circuit with external resistance of 180 Ω and (B) open circuit ([PAM]Initial = 10 mg L-1, [Fe]Total = 1

mg L-1, initial pH = 3.0 and aerated).

 13

3.2. Effects of operation parameters on PAM degradation

3.2.1 Effect of iron dosage

The efficiency of PAM degradation is affected by the initial iron dosage as illustrated in Fig. 3.

When the total iron concentration increased from 0 mg L-1 to 5 mg L-1, PAM degradation efficiency

at the end of 9 h was enhanced by around 33%. In typical Fenton reaction mechanism, both Fe2+

and H2O2 are responsible for the generation of free hydroxyl radical ● OH [Equation (2) and (3)].

Since the catholyte was saturated with dissolved oxygen due to continuous aeration (e.g. fixed H2O2

supply), the only variability in the current Fenton’s performance was the iron dosage. During the

course of free radical generation, Fe2+ was oxidized to Fe3+ and was consumed more rapidly than

being regenerated [10, 30]. In addition, Fe2+ was prone to be consumed by the dissolved oxygen or

other oxidants in an aqueous solution under aerobic conditions (Equation 4), hence effective

concentration of Fe2+ was a limiting factor for the generation of ●OH. In MFC-Fenton process, Fe2+

regeneration could be enhanced by introducing bio-electrons into the Fe3+/Fe2+ redox cycle

(Equation 5) and thus promote the Fenton chain reaction efficiency. The initially dosed iron source

thus acted as a catalyst and remained at a relatively constant, quite low but highly efficient level

throughout the experimental period. For this reason, continuous addition of iron source could be

avoided in an MFC-Fenton system. A lower operation cost would then be achieved as compared to

the conventional chemical Fenton approach.

O2 + 2H+ +2e- → H2O2 (2)

Fe2+ + H2O2 + H+ → Fe3+ + ●OH + H2O (3)

4Fe2+ + O2 + 4H+ → 4Fe3+ + 2H2O (4)

14

Fe3+ + e- → Fe2+ (5)

Fig. 3. Effect of total iron concentration [Fe]Total on PAM degradation in MFC-Fenton systems

([PAM]Initial = 10 mg L-1, initial pH = 2.5, external resistance = 180 Ω and aerated).

When further increasing iron dosage to 10 mg L-1 and 15 mg L-1, the degradation efficiency

began to decrease and eventually fell back to 25%. This is probably due to the higher concentration

of iron source would trigger the side reactions between Fe2+, H2O2 and ●OH (Equations 6 and 7),

which lead to the loss of active species accounting for the PAM removal. Therefore, the optimal

stoichiometric molar ratio of Fe2+ to H2O2 occurred at an iron dosage of 5 mg L-1.

Fe2+ + ● OH + H+ → Fe3+ + H2O (6)

H2O2 + ●OH → H2O + HO2● (7)

 15

It is worth noting that only dominant reactants are considered herein, although there are other

parasitic reactions involved in the consumption of effective species in PAM degradation (more

information is available in Supplementary Data) [6, 10, 21, 31].

3.2.2 Effect of initial pH

The influence of initial pH on the Fenton reaction was studied using five various pH values.

When the pH was lowered from 3.5 to 2.0, the relative PAM concentration decreased from 0.94 ±

0.01 to 0.47 ± 0.04 after 9-h reaction (Fig. 4). Higher pH values were not tested because there may

not be sufficient protons to participate the formation of H2O2 or it was not acidic enough to prevent

the precipitation of ferric ions as oxide-hydroxides of iron at pH > 4.0 (calculated by MINEQL).

However, the PAM degradation efficiency did not always increase with the step-by-step

acidification of catholyte. When pH was lowered to 1.5, only less than 40% of PAM was removed

from the catholyte. This phenomenon may due to the diffusion of protons from cathode chamber to

anode chamber, resulting in a strong acidic environment in anode chamber. This strong acidic

environment may inhibit or even terminate the bioactivity of anodic bacteria, which suppressed the

generation of bio-electricity and consequently affected the electron-Fenton reactions. Therefore, the

optimal pH condition was determined as 2.0.

16

Fig. 4. (A) Effect of initial pH on PAM degradation and (B) average hydroxyl radical concentration

generated at different pH in MFC-Fenton systems ([PAM]Initial = 10 mg L-1, [Fe]Total = 5 mg L-1,

external resistance = 180 Ω and aerated).

 17

The optimal pH of 2.0 was related to the concentration of ●OH. Referring to Equations (2) and

(3), hydroxyl radical ●OH was generated from dissolved O2 via H2O2 catalyzed by Fe2+. At standard

ambient conditions (25 ℃ and 1 atm), the theoretical maximum concentration of dissolved O2 in

catholyte is 260 µM (based on saturated value of 8.24 mg L-1), from which identical moles of H2O2

and ●OH can be generated [32]. However, though being continuously aerated (to ensure a dissolved

O2 concentration of 7.89 ± 0.21 mg L-1 in the catholyte), the measured concentrations of ●OH and

H2O2 in the cathode chamber were much lower than the theoretical values due to the fast reactions

(e.g. a rate constant between 53 M-1 s-1 and 76 M-1 s-1 for Equation 3) [30]. The maximum ●OH

concentration on average was 85 ± 11 µM in the catholyte at pH 2.0 across the 9-h reaction period.

A higher or lower pH than 2.0 may lead to reduced ● OH yield. The changing trend in the average

concentration of free radical was in accordance with its performance in PAM degradation as a

function of pH.

3.2.3 Effect of external resistance

The input fluxes of bio-electrons, measured as the electrical current in this study, played an

important role in MFC-Fenton system. As was discussed earlier, the input fluxes of bio-electrons

could directly influence both the generation of H2O2 and the regeneration of Fe2+, and eventually

the in-situ production of free ●OH in the cathodes. According to the Ohm’s Law, a higher current

density could be achieved by decreasing the external resistance. When external resistance was

lowered from 1000 Ω to 180 Ω, the average current density increased correspondingly threefold

from 327 ± 18 mA m-2 to 1094 ± 29 mA m-2 during the testing period (Fig. 5A). By setting external

resistance to as low as 20 Ω, the PAM degradation efficiency was observed to be enhanced further.
18

Up to 70% of PAM could be removed from the catholyte at an average current density of 2383.33 ±

212.13 mA m-2 within 9 h (Fig. 5B). Even smaller external resistance value would benefit more

efficient removal of PAM, but it may not be effective enough to collect the voltage output. It is

worth noting that the relative PAM concentration decreased by 53% when external resistance was

lowered from 1000 Ω to 20 Ω. This observation indicates that increased input of bio-electron fluxes

contributed to the enhanced degradation of PAM.

 19

Fig. 5. (A) Effect of external resistance on PAM degradation and (B) average current density in

MFC-Fenton systems ([PAM]Initial = 10 mg L-1, [Fe]Total = 5 mg L-1, initial pH = 2.0 and aerated).

Reproducible performance could be attained after optimization procedures when operating two

sets of MFC-Fenton systems for three batch cycles. The difference in average PAM degradation

efficiency between these two systems running in parallel was less than 8%. These repeatable results

proved that the MFC-Fenton technology could serve as a reliable tool for complicated organic

pollutants removal from wastewater. In natural water bodies, iron-reducing bacteria are common in

metal-leaching environments, while gaseous oxygen is abundant in the atmosphere. Therefore,

natural Fenton’s reagents of ferrous ion and dissolved oxygen are both available in an MFC testing

site, e.g. in a scaling-up sediment MFC field. MFC-Fenton technology can also be applied to a

wider range of pharmaceutical compound other than PAM.

3.3. Mechanism of PAM degradation by electro-Fenton

20

3.3.1 Possible pathways

Compared to the iron-free and open-circuit systems in which PAM was degraded slowly, our

results showed that the PAM concentration dropped significantly in a Fenton reactor bio-electrically

powered by an MFC. There was a rapid decrease in the relative PAM concentration from 0.95 ±

0.02 (around 5% loss due to adsorption onto electrode, filter and tube wall etc.) to 0.30 ± 0.03 in the

first 9 h (fast stage), followed by a slow decline to as low as 0.11 ± 0.01 when the reaction time was

extended to 24 h (slow stage) (Fig. S2). The cathodic degradation of PAM included the

electrochemical reduction of PAM in the early stage and the chemical reactions associated with

Fenton processes in the following stage (Fig. 6). The first PAM reduction reaction was coupled to

the bio-electrochemical reactions on the anode. By receiving the bio-electrons and protons under an

acidic condition, PAM was preliminary reduced to p-aminophenol (PAP), giving a by-product of

acetic acid which could be detected at a reaction time of 2 h (Table S1). PAP has previously

reported to be easily oxidized to PNP in the presence of ●OH [29], which could explain the fact that

no significant level of PAP was detected in this study. The second chemical Fenton process started

with electrophilic attack by ●OH on the benzene ring of PAM, which was evidenced by the

detectable PNP in catholyte mixtures after 3 h. The proposed structure of PNP also matched well

with the newly emerged absorption peak caused by NO2 substituent on its benzene ring in the

evolution of UV-vis spectra (inset of Fig. 2A). The degradation of PAM was then proposed to

undergo breakdown and hydroxylation of the benzene ring via ●OH addition and subsequent H2O

elimination, generating three dicarboxylic acids of maleic acid, malonic acid and malic acid, and

one carboxylic acid of lactic acid.

 21

However, the degraded PAM was not fully mineralized to CO2 or other inorganic final products

(e.g. ammonia, nitrate/nitrite and carbonate/hydrocarbonate). TOC studies revealed that only 24.51

± 1.32% of total organic carbon was converted into inorganic products throughout the 24-h reaction

period (Fig. S2). This may be attributed to the fact that there were some other active species (e.g.

detectable H2O2) rather than ● OH which could be involved in the stepwise decomposition of PAM

to smaller intermediates. ●OH is the strongest oxidant following fluorine, which would be acting as

the dominant force in complete PAM mineralization. The less powerful species of H2O2 and HO2●,

which are much weaker than ●OH, could degrade PAM via different reaction pathways to some

organic metabolites (e.g. benzoquinone and hydroquinone) but imposed weak mineralization of the

generated metabolites [33]. The maximum concentration of ammonia (product of amide group) in

catholyte mixtures was 0.21 mg/L, accounting for only 18% of the initial nitrogen concentration in

PAM. Another possible reason for the partial conversion of organic substrates to inorganic products

was the relatively moderate bio-electrochemical process on the anode of MFCs. In this study,

low-strength wastewater was fed to the electricigens for releasing bio-electrons and enhancing

Fenton reactions. The average biodegradable organic concentration was 220 ± 34 mg L-1 (as BOD),

which was equivalent to 28 mM bio-electrons at the maximum power output capacity. In a number

of previous contributions (Table S2), nearly complete decolourization and mineralization of organic

dyes could be obtained within several hours. But those highly efficient MFC-Fenton processes were

driven by high BOD loading or pure chemical fuels such as 20 mM lactate (i.e. 360 mM

bio-electrons) [16], in which the avaible bio-electrons were dozen of times as high as in our

domestic wastewater. Subsequent studies reported a decrease in pollutant removal efficiency when
22

using real wastes such as swine wastewater as bio-fuels. The TOC removal efficiency was as low as

5% in up to 2 days when employing MFC-Fenton to treat coking wastewater [34].

Fig. 6. Proposed pathway of cathodic PAM degradation (possible intermediate in dashed box was

derived from the other identified compounds).

From an environmental point of view, the domestic wastewater-powered MFC-Fenton system for

PAM degradation has several advantages over conventional technologies. When comparing

cathodic degradation with anodic oxidation of PAM in MFCs, Fenton-MFC could prevent potential

toxicity of this pharmaceutical and its metabolites to electricigens on anode [35]. When comparing

traditional Fenton process with MFC-Fenton process, our results showed that bio-electron fluxes

extracted from the organic pollutants in domestic wastewater facilitated the regeneration of Fe2+,
 23

thus no continuous addition of iron source was required. The cathodic degradation of PAM

degradation has been proved to be enhanced by the bio-electrons, and the voltage output of MFC

can be improved by providing a stable cathode potential [ECathode(O2/H2O2) = 0.28 V] [36] with

constant generation of hydrogen peroxide in reverse. The average voltage output from Fenton-MFC

was 20-30% higher than that without addition of Fenton reagents in this study.

3.3.2 Kinetic analyses

The PAM existed in two forms: immobile species absorbed onto the solid graphite plate and free

species in the bulk catholyte. Since the bio-electrons was directly transferred from anode to cathode

via the conductive circuit, accelerated reaction rate in the transformation between Fe3+ and Fe2+ as

well as the generation of ● OH could be obtained in the electron-abundant zone near the cathode

surface. As a result of the reduced mass transfer resistance, the absorbed PAM onto or near the

surface sites of cathode was attacked and converted. Undergoing a stirring force exerted by the air

bubbles, the PAM concentration dropped off quickly in the whole cathode chamber during the first

few hours. Plotting the experimental data on natural logarithmic-linear scale, the regression

coefficients (R2) are 0.98 or greater (Table 1) in this stage, suggesting the fast PAM decomposition

followed the first order reaction. The apparent rate constants of fast PAM degradation come to 0.13

h-1 and 0.14 h-1 for the two sets of systems running in parallel. But the apparent rate constants drop

rapidly when the PAM concentration decreased to a lower level after 9 h. This result suggests that

the migration velocity of PAM in the diffusion layer near the cathode is not the limiting factor for

the fast degradation process when PAM concentration was higher than 3 mg L-1, whereas the mass
24

transfer of PAM from bulk solution to the cathode became the rate-limiting factor for slow stage at

lower concentration.

Table 1 Kinetic equations with regression coefficients (R2) and apparent rate constants (kapp) for

PAM degradation.

Reactor Kinetic equation R2 kapp (h

-1
) Half-life (h) Lifetime (h)

MFC-1 -ln(C/C0) = 0.13t + 0.07 (0 ≤ t ≤ 9) 0.98 0.13 4.87 58.79

-ln(C/C0) = 0.07t + 0.49 (12 ≤ t ≤ 24) 0.99 0.07

MFC-2 -ln(C/C0) = 0.14t + 0.05 (0 ≤ t ≤ 9) 0.99 0.14 4.59 58.07

-ln(C/C0) = 0.07t + 0.54 (12 ≤ t ≤ 24) 0.99 0.07

Note: lifetime was determined as the amount of time required for PAM concentration to fall to 1%

(99% removal) of its initial value.

Half-life, i.e. time duration needed for PAM to decrease to 50% of its initial concentration, was

calculated to be 4.87 h and 4.59 h for MFC-1 and MFC-2, respectively. A much longer lifetime, i.e.

around 58 h for a degradation efficiency of higher than 99%, would be taken for the complete

decomposition of PAM from catholyte. Herein, a combination of bio-electrochemical and

photo-electrochemical reactions, i.e. coupling photo-degradation of PAM to MFC-Fenton process,

is proposed for ultra-pure effluent. By doping Fe3+ (Fenton reagent) into immobilized TiO2 particles,

the responsive wavelength range of irradiation source for photocatalysis could be expanded from

UV to that of visible light which is abundant in sunlight and less harmful to the electricigens in
 25

MFCs [37]. Self-loading Fe3+ could also enhance the photocatalytic performance of TiO2 as an

electrons accepter to mitigate the recombination of holes and electrons [38, 39].

4. Conclusion

This study provided novel results on mainly three aspects: investigation into the feasibility of

PAM degradation by electro-Fenton reactions driven by domestic wastewater as fuels, optimization

of operating parameters for the MFC-Fenton system, and elucidation of the possible decomposition

pathways and kinetic processes of PAM in MFC-Fenton reactions. In the cathode chamber of an

MFC, PAM was mainly mineralized by hydroxyl radicals and decomposed synergistically by the

weaker species generated in the presence of iron and oxygen. The optimal operating conditions

were determined as: total iron dosage of 5 mg L-1, initial pH of 2.0 and external resistance of 20 Ω

for the degradation of 10 mg L-1 PAM in aqueous solution. In-situ generation of free hydroxyl

radicals was enhanced by bio-electrons released from pollutants oxidation in the anode chamber. By

increasing the input flux of electrons, bio-electrochemically degradation of PAM was promoted. At

optimal working conditions, the efficiency of PAM degradation could be enhanced by up to 70%

within 9 h. PAM was proposed to be firstly reduced to p-aminophenol, and then quickly converted

to PNP which could be further degraded to smaller dicarboxylic/carboxylic acids.

Acknowledgments
26

The authors gratefully acknowledge the financial support from the National University of

Singapore, National Research Foundation and Economic Development Board (SPORE,

COY-15-EWI-RCFSA/N197-1), and Ministry of Education (R-143-000-582-112).

References

[1] P.E. Stackelberg, E.T. Furlong, M.T. Meyer, S.D. Zaugg, A.K. Henderson, D.B. Reissman,

Persistence of pharmaceutical compounds and other organic wastewater contaminants in a

conventional drinking-watertreatment plant, Sci. Total. Environ. 329 (2004) 99-113.

[2] D.W. Kolpin, E.T. Furlong, M.T. Meyer, E.M. Thurman, S.D. Zaugg, L.B. Barber, H.T. Buxton,

Pharmaceuticals, hormones, and other organic wastewater contaminants in US streams, 1999-2000:

A national reconnaissance, Environ. Sci. Technol. 36 (2002) 1202-1211.

[3] C.G. Daughton, Environmental stewardship and drugs as pollutants, Lancet 360 (2002)

1035-1036.

[4] A. Joss, E. Keller, A.C. Alder, A. Gobel, C.S. McArdell, T. Ternes, H. Siegrist, Removal of

pharmaceuticals and fragrances in biological wastewater treatment, Water Res. 39 (2005) 3139-3152.

[5] A. Joss, S. Zabczynski, A. Gobel, B. Hoffmann, D. Loffler, C.S. McArdell, T.A. Ternes, A.

Thomsen, H. Siegrist, Biological degradation of pharmaceuticals in municipal wastewater treatment:

Proposing a classification scheme, Water Res. 40 (2006) 1686-1696.

[6] L.M. Yang, L.E. Yu, M.B. Ray, Photocatalytic Oxidation of Paracetamol: Dominant Reactants,

Intermediates, and Reaction Mechanisms, Environ. Sci. Technol. 43 (2009) 460-465.

 27

[7] P.K. Dutta, S.O. Pehkonen, V.K. Sharma, A.K. Ray, Photocatalytic oxidation of arsenic(III):

Evidence of hydroxyl radicals, Environ. Sci. Technol. 39 (2005) 1827-1834.

[8] M. Jagannathan, F. Grieser, M. Ashokkumar, Sonophotocatalytic degradation of paracetamol

using TiO2 and Fe3+, Sep. Purif. Technol. 103 (2013) 114-118.

[9] B. Manu, S. Mahamood, Enhanced degradation of paracetamol by UV-C supported photo-Fenton

process over Fenton oxidation, Water Sci. Technol. 64 (2011) 2433-2438.

[10] H.C. Tao, X.Y. Wei, L.J. Zhang, T. Lei, N. Xu, Degradation of p-nitrophenol in a BES-Fenton

system based on limonite, J. Hazard. Mater. 254 (2013) 236-241.

[11] C.E. Reimers, L.M. Tender, S. Fertig, W. Wang, Harvesting energy from the marine

sediment-water interface, Environ. Sci. Technol. 35 (2001) 192-195.

[12] F. Rezaei, T.L. Richard, R.A. Brennan, B.E. Logan, Substrate-enhanced microbial fuel cells for

improved remote power generation from sediment-based systems, Environ. Sci. Technol. 41 (2007)

4053-4058.

[13] X.P. Zhu, J.R. Ni, Simultaneous processes of electricity generation and p-nitrophenol

degradation in a microbial fuel cell, Electrochem. Commun. 11 (2009) 274-277.

[14] Y. Li, A. Lu, H. Ding, X. Wang, C. Wang, C. Zeng, Y. Yan, Microbial fuel cells using natural

pyrrhotite as the cathodic heterogeneous Fenton catalyst towards the degradation of biorefractory

organics in landfill leachate, Electrochem. Commun. 12 (2010) 944-947.

[15] C. Feng, F. Li, H. Liu, X. Lang, S. Fan, A dual-chamber microbial fuel cell with conductive

film-modified anode and cathode and its application for the neutral electro-Fenton process,

Electrochim. Acta 55 (2010) 2048-2054.

28

[16] C.-H. Feng, F.-B. Li, H.-J. Mai, X.-Z. Li, Bio-Electro-Fenton Process Driven by Microbial Fuel

Cell for Wastewater Treatment, Environ. Sci. Technol. 44 (2010) 1875-1880.

[17] L. Fu, S.J. You, G.Q. Zhang, F.L. Yang, X.H. Fang, Degradation of azo dyes using in-situ

Fenton reaction incorporated into H2O2-producing microbial fuel cell, Chem. Eng. J. 160 (2010)

164-169.

[18] D.Y. Huang, S.G. Zhou, Q. Chen, B. Zhao, Y. Yuan, L. Zhuang, Enhanced anaerobic

degradation of organic pollutants in a soil microbial fuel cell, Chem. Eng. J. 172 (2011) 647-653.

[19] X.Q. Wang, C.P. Liu, Y. Yuan, F.B. Li, Arsenite oxidation and removal driven by a

bio-electro-Fenton process under neutral pH conditions, J. Hazard. Mater. 275 (2014) 200-209.

[20] M.A.F. de Dios, O. Iglesias, E. Bocos, M. Pazos, M.A. Sanroman, Application of benthonic

microbial fuel cells and electro-Fenton process to dye decolourisation, J. Ind. Eng. Chem. 20 (2014)

3754-3760.

[21] X.P. Zhu, B.E. Logan, Using single-chamber microbial fuel cells as renewable power sources of

electro-Fenton reactors for organic pollutant treatment, J. Hazard. Mater. 252 (2013) 198-203.

[22] B.E. Logan, Exoelectrogenic bacteria that power microbial fuel cells, Nat. Rev. Microbiol. 7

(2009) 375-381.

[23] H. Liu, S.A. Cheng, B.E. Logan, Power generation in fed-batch microbial fuel cells as a function

of ionic strength, temperature, and reactor configuration, Environ. Sci. Technol. 39 (2005)

5488-5493.

[24] G.C. Gil, I.S. Chang, B.H. Kim, M. Kim, J.K. Jang, H.S. Park, H.J. Kim, Operational parameters

affecting the performance of a mediator-less microbial fuel cell, Biosens. Bioelectron. 18 (2003)

327-334.
 29

[25] S. Cheng, H. Liu, B.E. Logan, Power densities using different cathode catalysts (Pt and

CoTMPP) and polymer binders (Nafion and PTFE) in single chamber microbial fuel cells, Environ.

Sci. Technol. 40 (2006) 364-369.

[26] M. Ghasemi, W.R.W. Daud, S.H.A. Hassan, S.E. Oh, M. Ismail, M. Rahimnejad, J.M. Jahim,

Nano-structured carbon as electrode material in microbial fuel cells: A comprehensive review, J.

Alloy. Compd. 580 (2013) 245-255.

[27] B.E. Logan, Microbial Fuel Cells, John Wiley & Sons, Inc., Hoboken, New Jersey, 2007.

[28] C. Comninellis, Electrocatalysis in the electrochemical conversion/combustion of organic

pollutants for waste water treatment, Electrochim. Acta 39 (1994) 1857-1862.

[29] E. Moctezuma, E. Leyva, C.A. Aguilar, R.A. Luna, C. Montalvo, Photocatalytic degradation of

paracetamol: Intermediates and total reaction mechanism, J. Hazard. Mater. 243 (2012) 130-138.

[30] H. Zhang, C. Fei, D. Zhang, F. Tang, Degradation of 4-nitrophenol in aqueous medium by

electro-Fenton method, J. Hazard. Mater. 145 (2007) 227-232.

[31] W.P. Kwan, B.M. Voelker, Rates of hydroxyl radical generation and organic compound

oxidation in mineral-catalyzed Fenton-like systems, Environ. Sci. Technol. 37 (2003) 1150-1158.

[32] APHA, Standard methods for the examination of water and wastewater, 18th ed., American

Public Health Association, Washington, DC., 1992.

[33] H.C.A. Valdez, G.G. Jimenez, S.G. Granados, C.P. de Leon, Degradation of paracetamol by

advance oxidation processes using modified reticulated vitreous carbon electrodes with TiO2 and

CuO/TiO2/Al2O3, Chemosphere 89 (2012) 1195-1201.

30

[34] B.G. Wu, C.H. Feng, L.Q. Huang, Z.S. Lv, D.H. Xie, C.H. Wei, Anode-biofilm electron transfer

behavior and wastewater treatment under different operational modes of bioelectrochemical system,

Bioresour. Technol. 157 (2014) 305-309.

[35] J. Chen, S.K. Yang, W.K. Wang, B. Li, L.C. Gao, Y.Y. Liu, Effect of Graphene Modified

Cathode on Paracetamol Removal in Microbial Fuel Cell, Asian. J. Chem. 25 (2013) 8022-8026.

[36] R.A. Rozendal, E. Leone, J. Keller, K. Rabaey, Efficient hydrogen peroxide generation from

organic matter in a bioelectrochemical system, Electrochem. Commun. 11 (2009) 1752-1755.

[37] W. Tang, X. Chen, J. Xia, J. Gong, X. Zeng, Preparation of an Fe-doped visible-light-response

TiO2 film electrode and its photoelectrocatalytic activity, Mater. Sci. Eng., B 187 (2014) 39-45.

[38] W. Baran, A. Makowski, W. Wardas, The influence of FeCl3 on the photocatalytic degradation

of dissolved azo dyes in aqueous TiO2 suspensions, Chemosphere 53 (2003) 87-95.

[39] J.W. Zhang, D.F. Fu, H.Y. Gao, L. Deng, Mechanism of enhanced photocatalysis of TiO2 by

Fe3+ in suspensions, Appl. Surf. Sci. 258 (2011) 1294-1299.

31
32

Highlights

• A microbial fuel cell (MFC)-Fenton system without power supply was developed.

• Paracetamol (PAM) degradation was enhanced by bio-electrons from wastewater.

• PAM mineralization could be achieved by in-situ generation of free radicals.

• Possible pathways and kinetics of PAM degradation by MFC-Fenton were proposed.

• MFC-Fenton was proved to be both energy-saving and performance-efficient.