DOI: 10.1002/cssc.
201100678
Scaling-Up Microbial Fuel Cells: Configuration and
Potential Drop Phenomenon at Series Connection of Unit
Cells in Shared Anolyte
Daehee Kim,[a] Junyeong An,[a] Bongkyu Kim,[a] Jae Kyung Jang,[b] Byung Hong Kim,[c] and
In Seop Chang*[a]
To scale-up microbial fuel cells (MFCs), installing multiple unit
cells in a common reactor has been proposed; however, there
has been a serious potential drop when connecting unit cells
in series. To determine the source of the loss, a basic stack-
MFC (BS-MFC) has been devised, and the results show that the
phenomenon is due to ions on the anode electrode traveling
through the electrolyte to be reduced at the cathode connect-
ed in series. As calculated by means of the percentage poten-
tial drop, the degree of potential drop decreased with an in-
crease in the unit-cell distance. When the distance was in-
creased from 1 to 8 cm, the percentage potential drop in BS-
MFC1 decreased from 46.76  0.90 to 45.08  0.70 % and in BS-
Introduction
Microbial fuel cells (MFCs) are bioelectrochemical systems that
generate electricity via the activity of microorganisms on the
anode electrode.[1] Thanks to its versatility, MFC technology has
been found in a wide range of applications, including biosen-
sors,[2, 3] small power generation systems, such as sediment bat-
teries,[4, 5] and wastewater treatment systems.[6, 7] However, most
have only been implemented on lab-scale.
The scalability of MFCs is crucial for real applications, not
only in terms of increased capacity of fuel utilization, but also
to increase the power production. Therefore, this concept has
recently become a critical issue pertaining to their feasibility;
a number of researchers have focused on various MFC configu-
rations, electrode/separator materials, catalysis, etc.[8]
To fulfill these needs, MFC stacks have been investigated,
and two approaches have been attempted so far: 1) Serial
stacking individual MFCs,[9–14] and 2) installing multiple anodes
in a single compartment.[15–19] In terms of ease-of-construction,
the second approach ensures simpler fabrication than the first,
which requires a delicate design for the feed and discharge,
that is, the second approach has advantages in terms of con-
struction and maintenance (viability requirements). In addition,
the installation of multiple anodes in a reactor can increase the
ratio of electrode area per volume of anode compartment, and
it was previously verified that a high ratio leads to a high
power density.[16]
However, it has also been reported that a potential drop
occurs when anolyte-sharing unit cells were connected in
series,[19] which was expected to be due to ionic cross conduc-
tion. Ionic cross conduction was also reported in hydrogen fuel
1086  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
MFC2 from 46.41  0.95 to 43.82  2.23 %. As the p-value of
the t-test was lower than 0.05, the difference was considered
significant; however, if the unit cells are installed far enough
from each other to avoid the potential drop phenomenon, the
system will be less dense, consequently reducing the ratio of
electrode area per volume of anode compartment and de-
creasing the power density of the system. Finally, this study
suggests design criteria for scaling-up MFC systems: Multiple-
electrode-installed MFCs are modularized, and the unit cells
are connected in series across the module (connecting each
unit cell does not share the anolyte).
cell stacks (planar banded configuration) to be due to the par-
asitic current generated between serially connected anodes
and cathodes when two unit cells shared an electrolyte mem-
brane.[20]
Herein, the issue of scaling-up MFCs will be addressed, espe-
cially with regard to their potential configuration. In addition,
the mechanism of the potential drop phenomenon is identi-
fied and alleviation methods suggested. Lastly, a design of
MFC plants is drawn according to suggestions made in the
study.
[a] Dr. D. Kim, J. An, B. Kim, Prof. I. S. Chang
School of Environmental Science and Engineering
Gwangju Institute of Science and Technology (GIST)
261 Cheomdan-gwagiro (Oryong-dong), Buk-gu
Gwangju 500-712 (Republic of Korea)
Fax: (+ 82) 62-715-2434
E-mail: ischang@gist.ac.kr
[b] Dr. J. K. Jang
Energy and Environmental Engineering Division
National Institute of Agricultural Science
Rural Development Administration
150, Seodun-dong, Gwonseon-gu, Suwon 441-707 (Republic of Korea)
[c] Dr. B. H. Kim
Water Environment Research Centre
Korea Institute of Science and Technology
39-1 Hawolgok-dong, Sungpook-gu, Seoul 136-791 (Republic of Korea)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201100678.
ChemSusChem 2012, 5, 1086 – 1091
Scaling-Up Microbial Fuel Cells

Results and Discussion unit cells was devised. The unit cells (UC) were enriched, and
the maximum OCVs from 1-UC-A, 1-UC-B, 2-UC-A, and 2-UC-B
were 0.75, 0.73, 0.70, and 0.78 V, respectively. The OCV values
Multiple unit cell installed MFC (prism-shaped)
from each unit cell were similar to each other, showing that
A prism-shaped MFC (Pri-MFC) with three membrane electrode similar amounts of microbial cells were attached to each
assemblies (MEAs) was designed; advantages of a Pri-MFC anode electrode. Next, the unit cells were operated under
arise from the cell configuration and usage of MEAs. If six closed circuit mode, with an external resistance of 1000 W. The
prism-shaped cartridges are used for a cylindrical reactor, the respective maximum closed circuit voltages (CCVs) from each
ratio of the electrode area per volume of anode compartment unit cell were 0.56  0.01, 0.54  0.00, 0.54  0.01, and 0.56 
can be maximized (see the Supporting Information, Table S1), 0.01 V.
one of the key aspects for scaling-up an MFC. The MEA makes
the cathode surface area comparable to the anode surface
Series connection of unit cells in BS-MFC (un-shared vs.
area, an important design parameter because cathodes are
shared anolyte)
often regarded as a major bottleneck for MFC operations.[21]
Here, an air-breathing cathode is used to reduce the opera- The duplicated systems were operated in unshared mode by
tional costs compared to a cathode system with an aerated placing an acrylic plastic between the anode electrodes of UC-
buffer. A and UC-B. The re-developed OCVs from each unit cell were
Three MEAs in the Pri-MFC were enriched and displayed as noted in Table 2. After connection in series, no voltage
stable open circuit voltages (OCVs), as noted in Table 1. When
the three MEAs were connected in parallel, each showed the
same OCV level (0.55 V) as in a connected circuit. Various Table 2. OCVs of each unit cell and in series-connected stack with un-
orders of series connections were then tested (a–b–c, a–c–b, shared anolyte.
b–a–c, b–c–a, c–a–b, and c–b–a), which each had different Cell Mode OCV [V]
OCV values. These experiments were duplicated and the re- UC-A UC-B in series
sults were similar. BS-MFC1 un-connected 0.73 0.73 –
in series 0.73 0.73 1.46
BS-MFC-2 un-connected 0.74 0.72 –
Table 1. OCVs of three Pri-MFC MEAs and their parallel/series connection. in series 0.74 0.72 1.47

Connection mode OCV [V]

MEA-a MEA-b MEA-c stack
changes were observed in each unit cell, and the in series-con-
unconnected 0.55 0.55 0.56 –
parallel 0.55 0.55 0.55 0.55
nection voltages were 1.46 and 1.47 V for BS-MFC1 and BS-
series a–b–c 0.27 0.09 0.32 0.50 MFC2, respectively. In principle, when two cells are connected
a–c–b 0.30 0.41 0.21 0.51 in series (with voltages VCell A and VCell B), the total cell voltage
b–a–c 0.19 0.15 0.16 0.49 (Vtotal) can be calculated by using the following equation:
b–c–a 0.45 0.21 0.12 0.54
c–a–b 0.16 0.26 0.06 0.49
VTotal = VCell A + VCell B. Based on this principle, the total voltages
c–b–a 0.36 0.07 0.09 0.52 of BS-MFC1 and BS-MFC2 (in series connection) were estimated
to be 1.459 V (0.728 + 0.731 V) for BS-MFC1 and 1.466 V
(0.744 + 0.724 V) for BS-MFC2. In agreement with the estima-
It is well-known that a parallel connection results in an in- tion, the actual values were 1.46 and 1.47 V for BS-MFC1 and
crease of current while maintaining the voltage, whereas BS-MFC2, respectively (rounded to two decimal places). There-
a series connection results in an increase of the voltage while fore, it has been confirmed that, if the anodes of each unit cell
maintaining the current.[18] However, this theory was not exact- are not shared in a reactor, the voltage of in series connected
ly applicable to the three MEAs in the Pri-MFC that shared a re- cells follows this principle with no performance degradation.
actor. Even though the parallel mode showed the same OCV as The acrylic plates were then removed to allow UC-A and UC-
each of the three MEAs—following the theory—in the in series B to share the anolyte. The OCVs were again developed; the
mode, various connection orders displayed a similar or even values were 0.74, 0.61, 0.73, and 0.59 V for 1-UC-A, 1-UC-B, 2-
lower OCV level compared to parallel mode. It was also found UC-A, and 2-UC-B, respectively (Figure 1 a and b). The unit cells
that the OCVs of each MEA dropped to a certain extent. This were connected in series, and according to the series-connec-
phenomenon was also observed in other MFC stacks with mul- tion principle, the total cell voltages of in series-connected BS-
tiple electrodes in a shared reactor.[18, 19] MFC1 and BS-MFC2 were estimated to be 1.35 and 1.32 V, re-
spectively; however, their measured total voltages in the in
series mode were 0.86 and 0.75 V, which is considerably lower
Basic-stack MFCs to investigate the potential drop phenom-
than the estimation. This discrepancy arises from the potential
enon
drop of each unit cell. As depicted in Figure 1 a and b, when
To verify the potential drop phenomenon in the in series con- the circuit was connected in series, the voltage of each unit
nection mode, a basic-stack MFC system (BS-MFC) using two cell decreased to a certain degree. The voltages of 1-UC-A and

ChemSusChem 2012, 5, 1086 – 1091  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemsuschem.org 1087

Figure 1. OCV levels of each unit cell and in series-connected stack with
shared anolyte of a) BS-MFC1 and b) BS-MFC2. c) Electrode potentials (vs.
Ag/AgCl) of 1-UC-A and 1-UC-B connected in series with shared anolyte (ver-
tical dashed line indicates the time of series connection).
1-UC-B decreased from 0.74 to 0.71 V, and from 0.61 to 0.15 V,
respectively; therefore, the voltage of in series-connected BS-
MFC1 was 0.86 V (0.71 + 0.15 V). In the same manner, the vol-
tages of 2-UC-A and 2-UC-B decreased from 0.73 to 0.72 V and
from 0.59 to 0.03 V, respectively, thereby yielding an in-series
voltage of 0.75 V (0.72 + 0.03 V).
Performance degradation was observed not only for the
open circuit mode, but also for the closed circuit mode. The
maximum power density (Pmax) of BS-MFC1 decreased from
1184.80  10.31 to 282.63  3.29 mW m2 after the anolyte was
shared. Similarly, Pmax of BS-MFC2 also decreased from
1084.12  15.19 to 206.78  8.51 mW m2.
1088 www.chemsuschem.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
I. S. Chang et al.
Mechanism of the potential drop phenomenon
The behavior of each electrode potential (vs. Ag/AgCl) was
measured before and after connection in series. As described
in Figure 1 c, the anode potential of UC-A immediately in-
creased before returning close to its initial value (0.07 V
higher), while the cathode potential of UC-A increased slightly
(from 0.23 to 0.25 V). Therefore, it is proposed here that the
anode potential contributed mainly to the UC-A voltage
change. On the other hand, the cathode potential of UC-B dra-
matically decreased to a negative potential (from 0.12 to
0.34 V), whereas almost no change was observed in the
anode potential of UC-B (0.49 to 0.51); this shows that the
cathode potential was the main reason for the voltage change
of UC-B. Interestingly, the potentials of only two electrodes
changed (the cathode of UC-B and anode of UC-A)—the ones
directly connected in series mode. It can be anticipated that
some phenomenon occurred between these electrodes.
The proposed potential drop mechanism is as follows.
1) When the cell is not connected, electrons are charged at the
anodes of UC-A and UC-B via microbial activity. 2) Once the
UC-A anode and UC-B cathode are connected, the electron
charge at the UC-A anode is transferred to the UC-B cathode.
Concurrently, the protons at the electrical double layer on the
UC-A anode surface also travel through the UC-B anolyte and
cation exchange membrane to the UC-B cathode for electro-
neutrality. 3) As the kinetics of the bacterial metabolism on the
anode are faster than the oxidation–reduction reaction at the
cathode,[21] the potential of the UC-A anode decreases, al-
though the potential of UC-B cathode does not recover; the
main reason is the conduction of ions through the electrolyte,
referred to as ionic cross conduction.[20] Parasitic currents form
between the UC-A anode and the UC-B cathode when con-
nected in series, resulting in a type of short-circuited cell.
Possible methods to alleviate the potential drop phenomen-
on
To overcome this performance degradation phenomenon, sev-
eral approaches were suggested by researchers in the conven-
tional fuel cell field: Thickening the membrane to increase the
resistance, decreasing the potential difference between unit
cells, etc.[20] However, many of them cannot be applied to MFC
systems due to its special characteristics. An MFC system is op-
erated with bulkier solutions than conventional fuel cells, and
produced sludge can be accumulated in the system; thus,
these aspects cause the discrepancy of substrate concentration
and microbial (biocatalysts) enrichment on each anodic sur-
face, which leads to the difficulty of maintaining similar poten-
tials between unit cells.
As the low resistance of the bulky anolyte between unit
cells caused ionic cross conduction, which led to a drop in po-
tential, increasing the solution resistance can be one way to
overcome it. Decreasing the ionic strength of the anolyte
would subsequently increase the ionic resistance of the liquid.
However, it was reported that decreasing the anolyte ionic
strength also decreased the MFC power output.[22] Hence, this
ChemSusChem 2012, 5, 1086 – 1091
Scaling-Up Microbial Fuel Cells
is not a practical method because a small loss of power direct-
ly affects the operational costs of the system.
Another approach could be to increase the distance be-
tween facing anodes, that is, as the distance of ion conduction
increases, the resistance is consequently increased. Our results
show that increasing the distance between UC-A and UC-B
(from 1 to 8 cm) decreased the potential drop. At a distance of
1 cm, the percentage potential drops of BS-MFC1 and BS-
MFC2 were 46.76  0.90 and 46.41  0.95 %, respectively (see
the Supporting Information, Table S2). However, this decreased
when the distance was increased to 8 cm to 45.08  0.70 and
43.82  2.23 % (see the Supporting Information, Table S3). The
difference was significant, as the p-values of the t-test were
0.048 and 0.038 for BS-MFC1 and BS-MFC2, respectively. Even
for distances tested in this study (1 to 8 cm), the degree of po-
tential drop decreased with increased distance. However, if the
unit cells were installed far enough from each other to avoid
the potential drop phenomenon, the system will be less
dense; consequently, reducing the ratio of electrode area per
volume of the anode compartment leads to a decrease in the
systems power density. As previously verified, the potential
drop almost disappeared when thin rubber tubing (20 cm long
and 0.3 cm inner diameter) was installed between each unit
cell.[19] However, when considering MFC scalability, this design
is not a practical option.
Design recommendations for scaling-up MFC systems
Note that there are other possible approaches for overcoming
the potential drop phenomenon at the scaling-up stage. First,
separating the anolyte of each unit cell can be an alternative.
As verified in the previous section, no potential drop occurred
if the unit cells did not share a common anode reactor. There-
fore, by simply using blocks to separate the unit cells into indi-
vidual sections, the potential drop can be avoided while in-
Figure 2. Proposed methods to overcome the potential drop phenomenon (for simplification, each unit cell in-
cluding the anode and the cathode is described as a rectangle and only two unit cells are depicted).
ChemSusChem 2012, 5, 1086 – 1091  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
creasing the voltage via a series connection (Figure 2 a). How-
ever, this approach might also increase the fabrication com-
plexity because each unit cell requires subsidiary systems, such
as influent/effluent channels and pumps, thereby incurring
enormous costs on a large scale.
To cut such extra costs, an MFC system using a common
anolyte in several unit cells would have a big advantage.
Therefore, we should determine possible methods for collect-
ing the current from each unit cell and develop the voltage via
a series connection. As one solution, the charging systems
(e.g., batteries) can be directly connected to each unit cell (Fig-
ure 2 b). However, this method also requires extra costs for the
battery (installation and maintenance costs).
Employing several MFC stacks is another prospective
method for overcoming the inherent drawbacks of MFCs (Fig-
ure 2 c). By connecting unit cells from different MFC stacks in
series, the potential drop phenomenon can be avoided as they
do not share an anolyte. In addition, employing plural MFCs
via a hydraulic connection has been shown to increase sub-
strate utilization and, consequently, the overall conversion effi-
ciency.[11]
A conceptual MFC module system was subsequently pro-
posed from the concept shown in Figure 2, using a multiple-
electrode MFC (ME-MFC) as the basic platform (Figure 3). Multi-
ple ME-MFCs are installed in each module, and the liquid flows
from the top to bottom by gravitational force. As ME-MFCs in
the same module share the anolyte (liquid), a series connection
is conducted between the cells in different modules. It is ex-
pected that a system with a vertical liquid flow and horizontal
series connection would utilize the whole substrate feed and
at the same time increase the voltage for the operation of cer-
tain electric devices. Moreover, this system can be operated in
both continuous and sequential batch modes. To realize this
type of system, the number of ME-MFCs in a module and
number of modules needs to be calculated.
Conclusions
In this study, installing multiple
unit cells in a common reactor is
suggested as being an advanta-
geous method for scaling-up mi-
crobial fuel cell (MFC) systems.
One of the biggest drawbacks of
the suggested system, however,
is the potential drop phenomen-
on that occurs in the series con-
nection of each unit cell. As rec-
ommended by a previous study,
increasing the resistance (espe-
cially distance) between the unit
cells decreases the potential
drop; however, this cannot be
applied to a large-scale MFC.
Therefore, a design criterion—
modularization—was proposed
to increase the capacity of fuel
www.chemsuschem.org 1089
 I. S. Chang et al.

Figure 4. Pri-MFC configuration (dimensions are not to scale).

into the Pri-MFC to develop an OCV in batch mode. After 10 days,

Figure 3. Conceptual diagram of the modularized MFC system for scale-up.
fresh medium containing 5 mm acetate was continuously fed to
the Pri-MFC at a dilution rate of 0.05 h1. After that, external resis-
tances of 10 W were connected to a circuit for closed circuit mode.
Note that Pri-MFCs were operated at a constant room temperature
(30 8C).

utilization, concomitant with avoiding the potential drop phe- Basic stack-MFCs: The BS-MFC stack used was prepared from poly-
nomenon. acrylic plastic (Figure 5). Graphite felt (5 cm  5 cm  0.4 cm), cation
exchange membranes (Nafion N424), and platinum wires were
used as the electrodes, separators, and electron collectors, respec-
Experimental Section tively. Two unit cells (UC-A and UC-B) had separated cathode com-
partments (125 mL; 5 cm  5 cm  5 cm) and a common anode
Construction, start-up, and operation of MFCs compartment (a set of two BS-MFCs was prepared; unit cells were
Prism-shaped-MFCs: The Pri-MFC was prepared by using poly- designated as 1-UC-A, 1-UC-B, 2-UC-A, and 2-UC-B). The horizontal
acrylic plastic, and had a cylindrical shape with three MEAs (desig- lengths of the anode compartments were 1 and 8 cm, yielding
nated as MEA-a, MEA-b, MEA-c). As shown in Figure 4, a prism-like void volumes of 25 (5 cm  5 cm  1 cm) and 200 mL (5 cm  5 cm 
cartridge was inserted into a reactor containing the anode com- 8 cm), respectively. An acrylic plate could be installed in the middle
partment (light gray region around the triangle in Figure 4 a). MEAs of the anode compartment to separate the anolyte from the unit
were mounted on each side of the cartridge (three MEAs per car- cells. An anaerobic digester fluid from a brewery wastewater treat-
tridge; Figure 4 b). Note that the electrodes facing the outside of ment plant (Gwangju, Korea) was used as the inoculum source.
the cartridge were the anodes; there were void volumes inside the Note that the modified artificial wastewater containing 2 mm phos-
prisms, and the electrodes facing the inside were operated as air- phate buffer (pH 7.0) was used as the medium, and 10 mm acetate
breathing cathodes (area inside the triangle in Figure 4 a). The lid was added as the substrate. The catholyte solution consisted of
and base of the reactor each had one hole and faced each other 50 mm potassium ferricyanide(III) in a 25 mm phosphate buffer
through the inside of the cartridg- (pH 7.0). The MFC stack was operated under batch mode during
es to expose the cathode electro-
des to air. The actual volume of
the anode compartment equipped
with six cartridges was 150 mL.
Each MEA was prepared as previ-
ously described,[23] with Nafion 114
(2.1 cm  9.2 cm) used as the
cation exchange membrane and
two carbon cloths (1.5 cm  8.5 cm;
non-wet-proofed for anode and
10 % wet-proofed for cathode)
used as the electrodes. Platinum
was used as the cathode catalyst
(0.1 mg cm2). An anaerobic digest-
er fluid was diluted two times with
an anaerobically prepared medium
(artificial wastewater) containing
acetate (5 mm as final concentra-
tion after dilution) as the sub-
strate.[24] The mixture was then fed Figure 5. Schematic diagram of BS-MFC.

1090 www.chemsuschem.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemSusChem 2012, 5, 1086 – 1091
Scaling-Up Microbial Fuel Cells
the enrichment step and was then switched to continuous mode
(dilution rate = 0.15 h1 for the main tests. All experiments were
carried out at a constant room temperature (30 8C). For the series
connection of two unit cells, the UC-B cathode was connected to
the UC-A anode by using an external metal wire (UC-B [Anode j j
Cathode]–[Anode j j Cathode] UC-A). An acrylic plate was placed
between the unit-cell anodes (UC-A and UC-B) during the enrich-
ment step to separate the anode compartments (unshared ano-
lyte). The unit cells were then inoculated and operated under OCV
mode for approximately 9 days. Note that the system was operat-
ed in fed-batch mode to replenish both the medium and substrate.
Fresh medium containing 10 mm acetate was fed three times
throughout the enrichment period.
Analytical methods and calculations
The pH value, acetate concentration, and cell voltage were routine-
ly monitored throughout the study. The acetate concentration and
voltage were measured as described previously.[25] The electrode
potentials were measured by installing Ag/AgCl reference electro-
des close to the cathode of Cell B and the anode of Cell A. The po-
tentials of counter electrodes were calculated by using the follow-
ing equation: VCell = ECatEAn, where EAn and ECat are the anode and
cathode potential, respectively, and VCell is the cell voltage.
Acknowledgements
This work was supported by grants from Doyak (formerly, the Na-
tional Research Laboratory Program; R0A-2008-000-20088-0); the
Pioneer Research Center for Nano-morphic Biological Energy Con-
version and Storage (2009-0082812) funded by the National Re-
search Foundation (NRF) of the Ministry of Education, Science
and Technology; and the Research Program for Agricultural Sci-
ence & Technology Development (Project No. PJ008517032012),
Republic of Korea.
Keywords: electrochemistry · energy conversion · fuel cells ·
microbes · scale-up
ChemSusChem 2012, 5, 1086 – 1091  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[1] B. H. Kim, I. S. Chang, G. M. Gadd, Appl. Microbiol. Biotechnol. 2007, 76,
485 – 494.
[2] H. J. Kim, M. S. Hyun, I. S. Chang, B. H. Kim, J. Microbiol. Biotechnol.
1999, 9, 365 – 367.
[3] I. S. Chang, J. K. Jang, G. C. Gil, M. Kim, H. J. Kim, B. W. Cho, B. H. Kim,
Biosens. Bioelectron. 2004, 19, 607 – 613.
[4] J. An, H. Moon, I. S. Chang, Environ. Sci. Technol. 2010, 44, 7145 – 7150.
[5] D. R. Bond, D. E. Holmes, L. M. Tenders, D. R. Lovley, Science 2002, 295,
483 – 485.
[6] H. Liu, R. Ramnarayanan, B. E. Logan, Environ. Sci. Technol. 2004, 38,
2281 – 2285.
[7] R. A. Rozendal, H. V. M. Hamelers, K. Rabaey, J. Keller, C. J. N. Buisman,
Trends Biotechnol. 2008, 26, 450 – 459.
[8] B. E. Logan, Appl. Microbiol. Biotechnol. 2010, 85, 1665 – 1671.
[9] P. Aelterman, K. Rabaey, T. H. Pham, N. Boon, W. Verstraete, Environ. Sci.
Technol. 2006, 40, 3388 – 3394.
[10] A. Dekker, A. T. Heijne, M. Saakes, H. V. M. Hamelers, C. J. N. Buisman, En-
viron. Sci. Technol. 2009, 43, 9038 – 9042.
[11] A. Glvez, J. Greenman, I. Ieropoulos, Bioresour. Technol. 2009, 100,
5085 – 5091.
[12] I. Ieropoulos, J. Greenman, C. Melhuish, Int. J. Energy Res. 2008, 32,
1228 – 1240.
[13] I. Ieropoulos, J. Greenman, C. Melhuish, Bioelectrochemistry 2010, 78,
44 – 50.
[14] S. E. Oh, B. E. Logan, J. Power Sources 2007, 167, 11 – 17.
[15] D. Jiang, X. Li, D. Raymond, J. Mooradain, B. Li, Int. J. Hydrogen Energy
2010, 35, 8683 – 8689.
[16] Z. Li, L. Yao, L. Kong, H. Liu, Bioresour. Technol. 2008, 99, 1650 – 1655.
[17] T. Shimoyama, S. Komukai, A. Yamazawa, Y. Ueno, B. E. Logan, Appl. Mi-
crobiol. Biotechnol. 2008, 80, 325 – 330.
[18] B. Wang, J. I. Han, Biotechnol. Lett. 2009, 31, 387 – 393.
[19] L. Zhuang, S. Zhou, Electrochem. Commun. 2009, 11, 937 – 940.
[20] R. O’Hayre, T. Fabian, S. J. Lee, F. B. Prinz, J. Electrochem. Soc. 2003, 150,
A430 – A438.
[21] H. Rismani-Yazdi, S. M. Carver, A. D. Christy, O. H. Tuovinen, J. Power
Sources 2008, 180, 683 – 694.
[22] H. Liu, S. Cheng, B. E. Logan, Environ. Sci. Technol. 2005, 39, 5488 – 5493.
[23] T. H. Pham, J. K. Jang, H. S. Moon, I. S. Chang, B. H. Kim, J. Microbiol. Bio-
technol. 2005, 15, 438 – 441.
[24] J. K. Jang, T. H. Pham, I. S. Chang, K. H. Kang, H. Moon, K. S. Cho, B. H.
Kim, Process Biochem. 2004, 39, 1007 – 1012.
[25] D. Kim, I. S. Chang, Bioresour. Technol. 2009, 100, 4527 – 4530.
Received: October 29, 2011
Revised: February 14, 2012
Published online on May 8, 2012
www.chemsuschem.org 1091