Bioresource Technology xxx (2011) xxx–xxx
Contents lists available at ScienceDirect
Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech
Short Communication
A comparison of membranes and enrichment strategies for microbial fuel cells
Olivier Lefebvre a, Yujia Shen a, Zi Tan a, Arnaud Uzabiaga a, In Seop Chang b, How Y. Ng a,⇑
a
b
Department of Civil and Environmental Engineering, National University of Singapore, 1 Engineering Drive 2, Singapore 117576, Singapore
Department of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), 261 Cheomdan Gwagi-ro, Buk-gu, Gwangju 500-712, Republic of Korea
a r t i c l e
i n f o
Article history:
Received 18 November 2010
Received in revised form 31 January 2011
Accepted 1 February 2011
Available online xxxx
Keywords:
Biomax
Isopore
Microbial fuel cell
Selemion
Oxygen diffusion
a b s t r a c t
The external resistance is perhaps the easiest way to influence the operation of a microbial fuel cell
(MFC). In this paper, three enrichment strategies, whereby the external resistance was fixed at: (1) a high
value in order to maximize the cell voltage (U strategy); (2) a low value in order to maximize the current
(I strategy); and (3) a value equal to the internal resistance of the MFC to maximize the power output (P
strategy), were investigated. The I strategy resulted in increased maximum power generation and the
likely reason is that electron transfer was facilitated under low external resistance, which in turn, favored
the development of an electrochemically active biofilm. This experiment was conducted in a singlechamber MFC system equipped with a membrane electrode assembly, and a comparison of the performance achieved by five different membranes is also provided. Selemion was found to be a suitable
alternative to Nafion.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
In a microbial fuel cell (MFC), the maximum power is generated
when the external resistance applied to the circuit equals the internal resistance of the cell (Logan et al., 2006). Consequently, in order
to maximize electricity generation, it would make sense to operate
an MFC under such value of external resistance (P strategy). However, when operated under lower external resistance, the resistance to electron transfer is lowered which might favor the
development of electrochemically active microbes on an MFC anode. Under these conditions, the current is maximized (I strategy).
On the contrary, under high external resistance, the cell voltage is
higher (U strategy). To date, only Regan et al. (2009) addressed that
topic and showed that a low external resistance (i.e., 10, 50 and
265 X) resulted in a higher cell density over the anode, while
filamentous bacteria grew preferentially under high external resistance (i.e., 1000 and 5000 X). However, the difference in morphology was not accompanied by a difference in terms of performance,
and the maximum power density remained at the same level. As a
consequence, the ideal enrichment strategy for an MFC with
regards to the resistance applied to the external circuit remains
to be determined.
Membrane electrode assemblies (MEAs) have shown potential
for MFCs, maintaining the electrodes close to one another while
Abbreviations: DO, dissolved oxygen; MEA, membrane electrode assembly; MFC,
microbial fuel cell; PTFE, polytetrafluoroethylene.
⇑ Corresponding author. Tel.: +65 6516 4777; fax: +65 6774 4202.
E-mail address: esenghy@nus.edu.sg (H.Y. Ng).
preventing ambient air to come in contact with the anode, which
results in increased Coulombic efficiency and sometimes improved
power generation (Kim et al., 2009a; Pham et al., 2005). MEAs
proved particularly efficient to enhance the sensitivity of BOD sensors (Kim et al., 2009b). However, the membrane can contribute
largely to the electrolyte resistance and the search for an ideal
membrane is still required (Kim et al., 2007; Rozendal et al., 2008).
As a consequence, there were two objectives in this study. First,
a variety of membranes were tested for MEA–MFC application and
the optimal one was selected for enrichment experiments. In the
second phase, the efficacy of I, U and P strategies described above
for MEA–MFC enrichment were assessed.
2. Methods
2.1. Membrane testing
The anode consisted of non-wet-proofed plain carbon cloth
(Designation B, E-Tek, USA) and the cathode was made of nonwet-proofed carbon cloth coated with Pt at a standard load of
0.5 mg cmÀ2 (E-Tek, A6 ELAT V2.1). Five different membranes were
used for comparison: Nafion 117 (DuPont Co., USA), Selemion HSF
(Asahi Glass Co., Japan), polytetrafluoroethylene (PTFE) membrane
(Sartorius Stedim, Germany), Isopore membrane filter (Millipore,
USA) and Biomax ultrafiltration disc (Millipore, USA). These will
be referred as Nafion, Selemion, PTFE, Isopore and Biomax, respectively, hereafter. In all cases, the membrane was pressed against
one cathode and one anode to form an MEA and a layer of reverse
0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2011.02.003
Please cite this article in press as: Lefebvre, O., et al. A comparison of membranes and enrichment strategies for microbial fuel cells. Bioresour. Technol.
(2011), doi:10.1016/j.biortech.2011.02.003
2
O. Lefebvre et al. / Bioresource Technology xxx (2011) xxx–xxx
osmosis spacer was added on each side of the MEA to confer rigidity
to the assembly. The MEA was further assembled into a cylindrical
single-chambered MFC (28 mL) with the cathode-side of the MEA
facing the air. The MFCs were operated at an external resistance of
400 X and in continuous flow mode of 2 mL minÀ1 – providing a
hydraulic retention time of 14 min – using domestic wastewater
as the inoculum and as the substrate. All membrane tests were
performed in duplicate.
Oxygen mass transfer coefficients for each membrane were further determined by following the protocol of Kim et al. (2007) that
was adapted to a single-chamber MFC system. Briefly, a dissolved
oxygen (DO) probe was inserted in an un-inoculated MFC fitted
with the desired membrane and filled with distilled water previously sparged with nitrogen gas to remove DO. The mass transfer
coefficient of oxygen in the membrane, KO (cm sÀ1) was determined by monitoring the DO concentration over time and using
Eq. (1)
KO ¼ À
V
C 1;0 À C 2
ln
At
C 1;0
ð1Þ
where V is the working volume of the MFC, A is the membrane
cross-sectional area, C1,0 is the saturated oxygen concentration in
water and C2 is the measured DO in the MFC at time t. The oxygen
diffusion coefficient (DO cm2 sÀ1) for each membrane was further
calculated as DO = KO Lt, where Lt is the membrane thickness as reported by the manufacturer.
2.2. Enrichment strategies
After selection of the adequate membrane – i.e., the membrane
that achieved the highest power generation over a period of 100 d
– three new sets of experiments were conducted in duplicate using
the same MFC design under the same conditions except for the
external resistance applied. The three strategies applied aimed at
optimizing either voltage (U), current (I) or power (P) production.
Hence, the U strategy was conducted at a high external resistance
(5000 X), the I strategy at a low resistance (5 X) and the P strategy
at a resistance that matched the internal resistance of the MFC – as
determined by polarization curves – because the maximum power
is produced when the internal resistance is equal to the external
resistance (Logan et al., 2006).
2.3. Analytical methods and calculations
The cell voltage was measured with a multimeter connected to a
computer by a data acquisition system (M3500A, Array Electronic,
Taiwan). Polarization curves were obtained by varying the applied
external resistance at a time interval of 30 s and recording the pseudo steady-state voltage, after the MFC was allowed to reach its open
circuit voltage (after about 1 h). The current was then determined
using the Ohm’s law. The cell internal resistance was determined
using a linear regression (least squares method) on the linear part
of the polarization curve that corresponds to the Ohmic zone. The
electromotive force was estimated as the intercept of the regression
with the Y-axis whereas the internal resistance was the opposite of
its slope. Besides polarization curves, power curves were also drawn
in order to determine the maximum power supplied by the MFC.
3. Results and discussion
3.1. Membrane testing
Nafion, Selemion, PTFE, Isopore and Biomax membranes were
operated in an MEA–MFC configuration over a period of 100 d. In
average, Selemion, Nafion and Isopore membranes produced a
maximum power of 0.12 ± 0.02, 0.09 ± 0.02 and 0.02 ± 0.01 mW,
respectively. The difference observed in power was not due to a
difference of electromotive forces, which were quite similar with
values of 0.69 ± 0.03, 0.65 ± 0.04 and 0.62 ± 0.06 V for Selemion,
Nafion and Isopore membranes, respectively. However, the internal resistance averaged 1082 ± 193 X with a Selemion membrane,
but was 33% higher with a Nafion 117 membrane (1437 ± 193 X)
and 383% higher with an Isopore membrane (5228 ± 224 X). PTFE
and Biomax membranes failed to generate significant power. The
amount of power generated with a Selemion membrane
(4.3 W mÀ3) is in accordance with the other few studies using
domestic wastewater as a substrate. Specifically, Liu et al. (2004)
generated up to 1.5 W mÀ3 in a 388 mL MFC and, more recently,
Ahn and Logan (2010) produced 7.9 W mÀ3 with an MFC of a size
comparable to those used in this study (28 mL). In the latter publication, the internal resistance of their MFC was not provided
but can be calculated to be 408 X according to the polarization
curves published by the authors. This is about half the internal
resistance of the MFC used is this study; however, the electromotive force was a bit lower (0.6 V, still according to their polarization
curve). The reduced internal resistance in their device might be explained by the total absence of a membrane in their system.
The details of mass transfer coefficients (KO) and diffusivities of
oxygen (DO) for the various membranes used in this study are provided in Table 1. KO was the highest for Isopore (3 Â 10À4 cm sÀ1);
however, this was compensated by the very thin membrane and DO
was the highest for Nafion (0.9 Â 10À6 cm2 sÀ1). Kim et al. (2007)
found a slightly higher value of DO for Nafion (2.3 Â 10À6 cm2 sÀ1)
probably due to the fact that they monitored oxygen diffusion in a
dual-chamber MFC where the cathode chamber was actively
sparged with air, whereas the oxygen diffusion was monitored in
a single-chamber device with passive aeration in this study. The
highest DO value observed with Nafion as compared to the other
four membranes suggests that this membrane allowed the most
O2 to be diffused into the MFC set-up. On the contrary, DO was
an order of magnitude lower for Selemion (0.08 cm2 sÀ1) which allowed the maximum power generation in the MFC set-up used in
this study. It is a well-known fact that electrochemically active
bacteria are facultative anaerobes that will switch from anodic to
aerobic respiration in the presence of O2 (Logan and Regan,
2006). Lower oxygen diffusion can consequently explain the superiority of Selemion membrane over Nafion membrane. Another
advantage of Selemion is its competitive price – about $400 mÀ2
as compared to $1500 mÀ2 for Nafion. Isopore had a lower DO
(0.45 Â 10À6 cm2 sÀ1) than Nafion; however, it was only ranked
as the third best membrane in terms of power generation. This
lower performance can be explained by the structure of the membrane being conceived as a screen filter whereas Nafion and Selemion are actually proton exchange membranes – resulting in
considerably enhanced proton transfer from the anode to the cathode and consequently, in increased power generation – even
though it has been demonstrated that the selectivity of proton
exchange membranes is not perfect and that other cations are
allowed to cross-over in an MFC system (Rozendal et al., 2006).
The PTFE membrane is also conceived as a screen filter and its
DO (0.24 Â 10À6 cm2 sÀ1) was even lower than that of Isopore;
however, it failed at generating significant power and this is
probably a result of the hydrophobic nature of PTFE that prevented
proton transfer. Finally, Biomax also failed at generating significant
power and a possible cause could be the relatively large pores of
this ultra-filtration membrane. With a nominal molecular weight
limit (NMWL) of 50 KDa, this is significantly higher than that of
the ultra-filtration membranes – also made of polyethersulfone
as for Biomax – tested by Kim et al. (2007). The NMWL of the
ultra-filtration membranes tested in their study was in the range
of 0.5–3 KDa and even though the DO observed in their case
Please cite this article in press as: Lefebvre, O., et al. A comparison of membranes and enrichment strategies for microbial fuel cells. Bioresour. Technol.
(2011), doi:10.1016/j.biortech.2011.02.003
3
O. Lefebvre et al. / Bioresource Technology xxx (2011) xxx–xxx
Table 1
Mass transfer coefficients and diffusivities of oxygen for various membranes tested in single-chamber MFC set-ups.
Selemion
Isopore
PTFE
Biomax
Nafion
180
0.5
0.90
Hydrocarbon
150
0.05
0.08
Polycarbonate
15
3
0.45
PTFE
80
0.3
0.24
Polyethersulfone
120
0.4
0.48
3.2. Enrichment strategies
Three other sets of MEA–MFCs using Selemion membranes
were operated in duplicate, differing only by the external resistance applied to the system, in order to maximize either the power,
current or voltage output (P, I and U strategies, respectively). The
enrichment of the different MFCs under the various strategies is
displayed in Fig. 1 over the course of 50 d. All MFCs started generating electricity immediately and the voltage went up very fast for
the MFCs connected to a high external resistance in the first few
days, reaching as high as 500 mV with 5000 X over 3 d (Fig. 1a).
As expected the voltage increased with the external resistance applied but variations were observed in the course of time. Since replicability was very good between duplicates and all MFCs followed
the same trend, the variations observed were most probably
caused by the variability of domestic wastewater itself. For instance, on day 22 the voltage in all MFCs dropped due to significantly lower COD content in the domestic wastewater. After the
COD went up again, the voltage rapidly recovered. This emphasizes
the sensitivity of MEA–MFCs as BOD sensors, as evidenced by others (Kim et al., 2009b). As a consequence, the voltage ranged from
130 to 590 mV with an external resistance of 5000 X and from 2 to
10 mV at 5 X. With a variable resistance, and following the strategy described in the Section 2, the resistance varied between
5000 and 400 X and the voltage output varied between 50 and
360 mV. Here again, the impact of the composition of domestic
wastewater was stronger than that of the external resistance applied. In terms of current production, the external resistance of
5 X (I strategy) allowed generation of between 0.4 and 2 mA,
5000 X (U strategy) resulted in between 0.03 and 0.1 mA and varying the resistance (P strategy) led to intermediate current production between 0.1 and 0.8 mA (Fig. 1b). Finally, in terms of power
production the P strategy obviously allowed maximizing the power
produced between 0.01 and 0.3 mW, whereas the power was in the
range of 0.001–0.02 mW at 5 X (I strategy) and 0.004–0.06 at
5000 X (U strategy) (Fig. 1c). The combination of Figs. 1a, b and c
shows the efficiency of the three strategies to reach their goal.
The optimum strategy was assessed by measuring the maximum power output as well as the electromotive force and internal
resistance of the various set-ups used in this study by drawing
Cell voltage (mV)
a
600
400
200
0
0
10
20
30
40
50
30
40
50
Time (d)
b
2
1.5
Current (mA)
(0.51–1.1 Â 10À6 cm2 sÀ1) was in the same range as Biomax
(0.48 Â 10À6 cm2 sÀ1), they showed that (i) acetate diffusion
through the membrane increased considerably with the pore size
in their dual-chamber MFC system and (ii) acetate diffusion was
significantly larger than that observed with other cation and anion
exchange membranes due to the absence of selectivity (apart from
the size) of ultra-filtration membranes. In the single-chamber system used in this study, it was not possible to monitor substrate
cross-over but it can be easily hypothesized that it was considerably higher for Biomax than for any other type of membrane,
resulting in decreased cathode potential as observed by Harnisch
et al. (2009).
Overall, the performance of the system used in this study validated its proper functioning on domestic wastewater and Selemion
was selected for enrichment experiments because it provided the
optimum power generation.
1
0.5
0
0
10
20
Time (d)
c
0.4
P strategy
0.3
Power output (mW)
Material
Thickness (lm)
KO (Â 10À4 cm sÀ1)
DO (Â 10À6 cm2 sÀ1)
Nafion
I strategy
U strategy
0.2
0.1
0
0
10
20
30
40
50
Time (d)
Fig. 1. Evolution of the (a) voltage (b) current and (c) power obtained while
acclimating MFC systems to domestic wastewater under different strategies.
polarization curves on a twice weekly basis. The electromotive
force was comparable in all cases, averaging 0.72 ± 0.04,
0.70 ± 0.05 and 0.72 ± 0.08 V for the P, I and U strategy,
respectively. However, the internal resistance was found to be
Please cite this article in press as: Lefebvre, O., et al. A comparison of membranes and enrichment strategies for microbial fuel cells. Bioresour. Technol.
(2011), doi:10.1016/j.biortech.2011.02.003
4
O. Lefebvre et al. / Bioresource Technology xxx (2011) xxx–xxx
significantly lower when the MFCs were running with a lower
external resistance (962 ± 261 X at 5 X) as compared with
2316 ± 239 X at 5000 X and 1296 ± 315 X at a varying resistance.
As a consequence, the maximum power generated averaged
0.14 ± 0.02 mW at 5 X, 40% higher than what was obtained with
a varying resistance (0.10 ± 0.02 mW) and 133% higher than that
achieved at 5000 X (0.06 ± 0.02 mW). This shows the superiority
of I strategy as compared to the P and mostly to the U strategy
in terms of enrichment. Regan et al. (2009) had already shown that
the live cell density was inversely proportional to the external
resistance; however, the present study further shows that the
internal resistance of an MFC is directly correlated to the external
resistance applied to the electrical circuit.
4. Conclusions
In this study, a variety of membranes in MEA–MFC configuration was compared and Selemion was found to be a suitable alternative to Nafion. In the second phase, the efficacy of three different
strategies for enrichment of electrochemically active bacteria was
assessed and it was found that a lower resistance resulted in increased maximum power generation. The likely reason is that electron transfer is facilitated under low external resistance, favoring
the development of an electrochemically active biofilm over the
anode. The results of this study suggest that MFCs should be
started under maximized current conditions even though the
power output is lower during the enrichment period.
Acknowledgements
References
Ahn, Y., Logan, B.E., 2010. Effectiveness of domestic wastewater treatment using
microbial fuel cells at ambient and mesophilic temperatures. Bioresource
Technol. 101, 469–475.
Harnisch, F., Wirth, S., Schroder, U., 2009. Effects of substrate and metabolite
crossover on the cathodic oxygen reduction reaction in microbial fuel cells:
platinum vs iron(II) phthalocyanine based electrodes. Electrochem. Commun.
11, 2253–2256.
Kim, J.R., Cheng, S., Oh, S.E., Logan, B.E., 2007. Power generation using different
cation, anion, and ultrafiltration membranes in microbial fuel cells. Environ. Sci.
Technol. 41 (3), 1004–1009.
Kim, J.R., Premier, G.C., Hawkes, F.R., Dinsdale, R.M., Guwy, A.J., 2009a. Development
of a tubular microbial fuel cell (MFC) employing a membrane electrode
assembly cathode. J. Power Sources 187, 393–399.
Kim, M., Hyun, M.S., Gadd, G.M., Kim, G.T., Lee, S.J., Kim, H.J., 2009b. Membraneelectrode assembly enhances performance of a microbial fuel cell type
biological oxygen demand sensor. Environ. Technol. 30, 329–336.
Liu, H., Ramnarayanan, R., Logan, B.E., 2004. Production of electricity during
wastewater treatment using a single chamber microbial fuel cell. Environ. Sci.
Technol. 38, 2281–2285.
Logan, B.E., Regan, J.M., 2006. Electricity-producing bacterial communities in
microbial fuel cells. Trends Microbiol. 14, 512–518.
Logan, B.E., Hamelers, B., Rozendal, R., Schrorder, U., Keller, J., Freguia, S., Aelterman,
P., Verstraete, W., Rabaey, K., 2006. Microbial fuel cells: methodology and
technology. Environ. Sci. Technol. 40, 5181–5192.
Pham, T.H., Jang, J.K., Moon, H.S., Chang, I.S., Kim, B.H., 2005. Improved performance
of microbial fuel cell using membrane-electrode assembly. J. Microbiol.
Biotechnol. 15, 438–441.
Regan, J.M., Ren, Z., Carpenter, W., Ramasamy, R.P., Mench, M.M., 2009. External
resistance effects on anode biofilm architecture and performance. in: The 2nd
Microbial Fuel Cell Conference. Gwangju.
Rozendal, R.A., Hamelers, H.V.M., Buisman, C.J.N., 2006. Effects of membrane cation
transport on pH and microbial fuel cell performance. Environ. Sci. Technol. 40,
5206–5211.
Rozendal, R.A., Sleutels, T., Hamelers, H.V.M., Buisman, C.J.N., 2008. Effect of
the type of ion exchange membrane on performance, ion transport, and
pH in biocatalyzed electrolysis of wastewater. Water Sci. Technol. 57,
1757–1762.
This work was supported by a Grant from the Environment &
Water and Industry Development Council, National Research Foundation, Singapore (MEWR 651/06/159).
Please cite this article in press as: Lefebvre, O., et al. A comparison of membranes and enrichment strategies for microbial fuel cells. Bioresour. Technol.
(2011), doi:10.1016/j.biortech.2011.02.003
... comparison of membranes and enrichment strategies for microbial fuel cells Bioresour Technol (2 011 ), doi :10 .10 16/j.biortech.2 011 .02.003 O Lefebvre et al / Bioresource Technology xxx (2 011 ) xxx–xxx... comparison of membranes and enrichment strategies for microbial fuel cells Bioresour Technol (2 011 ), doi :10 .10 16/j.biortech.2 011 .02.003 O Lefebvre et al / Bioresource Technology xxx (2 011 ) xxx–xxx... (MEWR 6 51/ 06 /15 9) Please cite this article in press as: Lefebvre, O., et al A comparison of membranes and enrichment strategies for microbial fuel cells Bioresour Technol (2 011 ), doi :10 .10 16/j.biortech.2 011 .02.003