i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5

Available online at www.sciencedirect.com

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

Performance of a microfluidic microbial fuel cell

based on graphite electrodes

Dingding Ye a,b, Yang Yang b, Jun Li a,b, Xun Zhu a,b,*, Qiang Liao a,b,
Bowen Deng b, Rong Chen a,b
a
Key Laboratory of Low-grade Energy Utilization Technologies and Systems, Chongqing University,
Ministry of Education, Chongqing 400030, China
b
Institute of Engineering Thermophysics, Chongqing University, Chongqing 400030, China

article info abstract

Article history: A microfluidic microbial fuel cell utilizing the laminar flow to separate the anolyte and
Received 16 January 2013 catholyte streams based on graphite electrode is proposed. The co-laminar flow of the two
Received in revised form streams inside the microchannel is visualized under different flow rates. The effects of the
2 May 2013 acetate concentration and flow rate on the cell performance are investigated. The results
Accepted 4 May 2013 show that the cell performance initially increases and then decreases with increasing
Available online 14 June 2013 influent COD concentration and the anolyte flow rate. The microfluidic microbial fuel cell
produces a peak power density of 618
4 mW/m2 under the conditions of 1500 mg/L
Keywords: influent COD and an anolyte flow rate of 10 mL/h. The low internal resistance of fuel cell
Microfluidic microbial fuel cell results from elimination of the proton exchange membrane and high surface-to-volume
Laminar flow ratio of the microfluidic structure. Moreover, the thickness of biofilm decreases gradually
Membraneless along the flow direction of the microchannel due to the diffusive mixing of the catholyte.
Internal resistance Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.

1. Introduction Previous studies have demonstrated that the poor perfor-

mance of the micro-sized MFC results from the high internal
Microbial fuel cells (MFCs) are considered as a promising resistance and non-anode electron sinks [2e4]. It is found that a
strategy for electricity generation from wastes and renewable high specific surface area of anode is an effective method to
biomass. Due to its potential in wastewater treatment and reduce the internal resistance [5]. The performance of the
reducing environmental pollution problems, macro-scale MFCs micro-sized MFC can be improved by employing an embedded
have been widely studied [1]. Recently, micro-sized MFCs have micropillar electrode [6], a graphite felt electrode [2] or opti-
emerged as an attractive power source for portable electronic mizing the electrode shape [7]. Adding L-cysteine into the anode
devices and sensors and received increasing attention because chamber can scavenge the dissolved oxygen which is mostly
of its advantages such as mild operation conditions, ability of considered as the terminal electron acceptor at anode [5].
producing power from indigenous resources without noble To fabricate the micro-sized microbial fuel cell, photoli-
metal catalysts, etc [2e13]. However, the practical application thography [7e9] and soft-lithography [6,10e12] are widely used
of micro-sized MFCs is limited by the low performance and due to its low cost, short manufacturing time and high repro-
expensive fabrication process. ducibility. However, in most of these studies, gold was used as

* Corresponding author. Key Laboratory of Low-grade Energy Utilization Technologies and Systems, Chongqing University, Ministry of
Education, Chongqing 400030, China. Tel./fax: þ86 023 65102474.
E-mail address: [email protected] (X. Zhu).
0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijhydene.2013.05.034
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5 15711

the electrode material, which is not only an expensive material and two polymethylmethacrylate (PMMA) plates. Graphite
but also cause a high internal resistance due to the poor con- plates (1 mm thick, EDM Supplies Inc., poco grade EDM-3),
tact between exoelectrogenic bacteria and the gold anode which acted as electrodes, were placed side by side to form
[5,13]. It is interesting to note that, unlike most micro-sized a 2 mm wide, 40 mm long microchannel in which the anolyte
microbial fuel cells with a proton exchange membrane, a (culture medium) and catholyte (potassium ferricyanide so-
membraneless microbial fuel cell applying microfluidic flow lution in phosphate buffer) flowed in parallel without
control was reported recently [14]. The biological fuel cell with convective mixing. A wedge made of PMMA (1 mm thick) was
a total volume of 0.3 mL produced a maximum current density used at one end of the microchannel to separate the inlet
of 25.42 mA m2 by Shewanella oneidensis. Li et al. [15] reported a channels of the two streams, and a PMMA flake (1 mm thick)
laminar flow-based microfluidic microbial electrolysis cell was placed at the other end to seal the microchannel. Two
(MEC) to analyze the microbial electrochemical activity. In the inlets and one outlet were drilled out of the top and bottom
two microfluidic bioelectrochemical systems (BESs) PMMA plate (10 mm thick), respectively. The multilayer as-
mentioned above, the anolyte and the catholyte are separated sembly was held together by six M4 screw joints.
by the laminar flow without a physical barrier (such as a proton
exchange membrane). The two streams are mixed by trans- 2.2. Inoculation and operation
verse diffusion alone. The microfluidic structure can not only
avoid the membrane-related problems in MFCs and block of The microfluidic MFC was inoculated from the effluent of an
the proton transport by other cations, but also realize easily a acetate-fed MFC. The culture medium, containing 1.13 g/L
high surface-to-volume ratio, which benefits to reduce the CH3COONa$ 3H2O (COD of 500 mg/L), 15.3 g/L Na2HPO4$12H2O,
internal resistance [16e18]. However, gold electrodes were also 3 g/L KH2PO4, 0.5 g/L NaCl, 0.1 g/L NH4Cl, 0.1 g/L MgSO4$7H2O,
used in [14,15] leading to the high cost and internal resistance 11 mg/L CaCl2 and 1.0 mL/L of a trace elements solution, was
problems as discussed before. fed continuously. Prior to all experiments, pure nitrogen or
In the present study, a microfluidic MFC based on graphite argon was purged into the culture medium for 15 min to
electrodes is proposed and fabricated. The laminar flow in the remove the dissolved oxygen. The anolyte and catholyte were
microchannel is visualized at varied flow rates. Then the supplied by two syringe pumps, respectively. The potassium
distribution of the biofilm along the flow direction of the ferricyanide concentration is fixed to 20 mM. An Ag/AgCl
microchannel is observed. Moreover, effects of the fuel con- reference electrode was employed at the outlet of the micro-
centration and flow rate on the cell performance are dis- channel to obtain the individual electrode potentials of the
cussed. Finally, Electrochemical Impedance Spectroscopy anode and cathode. The distribution of the biofilm during the
(EIS) tests are conducted to determine the internal resistance steady-state operation was visualized by an inverted fluores-
of the microfluidic MFC. cence microscope (IX81, Olimpus).

2.3. Data acquisition and analysis

2. Materials and methods

Cell voltages (U), anode and cathode potentials of the fuel
cell were recorded every 15 s by an Agilent 34970 data
2.1. Construction of the microfluidic MFC
acquisition unit. The external resistance was varied in a
range of (0.2e10)  105 U to obtain the polarization curves.
As shown in Fig. 1, a microfluidic microbial fuel cell consisting
After each change in the electrical resistance, the MFCs were
of a Y-shaped channel was constructed by two graphite plates
held until the cell voltage reached steady-state, which
meant the fluctuation of the cell voltage was lower than
2 mV in 5 min and the steady time maintained at least
20 min. Current (I) can be calculated via Ohm’s law and
output power density by the equation: P ¼ U  I/A, where A
was the surface area of the electrode. Current and power
densities were normalized to the anode area (0.4 cm2). All
the potentials were quoted versus Ag/AgCl reference elec-
trode. All tests were conducted in a temperature-controlled
chamber at 25  C and repeated three times to check repro-
ducibility of the data.

2.4. EIS measurement

Electrochemical Impedance Spectroscopy (EIS) was per-

formed on the whole cell using a Zennium electrochemical
workstation (Zahner, Germany). Two electrode system was
adopted on the whole cell with the anode as the working
electrode and the cathode as the counter electrode and
Fig. 1 e Schematic of the microfluidic microbial fuel cell the reference electrode. All the EIS tests were carried out
based on graphite plate electrodes. under the closed circuit condition connecting an external
15712 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5

3. Results and discussion

3.1. Visualization of laminar flow in the microchannel

In this study, the anolyte and the catholyte are maintained at

the same flow rate. The laminar flow in the microchannel at
different flow rates (5, 10, 20, 40, 60 mL/h) are visualized as
shown in Fig. 2 (red ink was added into the anolyte to show it
distinctly). It can be seen that for all flow rates, two streams
containing anolyte and catholyte flow side by side in the
microchannel. The laminar flow can be characterized by low
Reynolds number, Re ¼ ruDh/m, where r is the fluid density, u
the average velocity, Dh the hydraulic diameter, and m the
dynamic viscosity. It is reported that the laminar flow gener-
ally occurs in a microchannel at the Reynolds number less
than 200 [19]. In this paper, Reynolds numbers are in the range
of 1.16e13.85 corresponding to flow rates from 5 to 60 mL/h.
Therefore, the results indicate that the anolyte and catholyte
flow laminarly in parallel without convective mixing in the
microchannel. Moreover, the diffusive mixing of the two
streams can be observed at the end of the microchannel.

3.2. Distribution of the biofilm

Fig. 2 e Visualization of the laminar flow in the The distribution of the biofilm during the steady-state opera-
microchannel at different flow rates: (a) 5 mL/h, (b) tion is recorded as shown in Fig. 3. It is clear that the thickness
10 mL/h, (c) 20 mL/h, (d) 40 mL/h, (e) 60 mL/h. of the biofilm decreases along the flow direction of the
microchannel. This result is mainly attributed to the inhibi-
tion of the biofilm formation on the anode resulting from the
diffusive catholyte. It is found that mixing between the two
resistance of 40,000 U between the anode and cathode. The co-laminar flow streams in the center of the microchannel
AC amplitude was 10 mV and frequency was varied from depends only on transverse diffusion. The mixing region
100 kHz to 10 mHz. The EIS data was analyzed using Zview2 shows an hourglass shape with maximum width (dx) near the
software. top and bottom walls scaling as the following equation [20]:

Fig. 3 e Distribution of the biofilm during the steady-state operation: (a) at the beginning of the microchannel, (b) at the
middle of the microchannel, (c) at the end of the microchannel.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5 15713

 1=3 lead to a decrease in the cathode potential (cf. Fig. 5c) and
DHz
dx f (1) lower the cell performance. In addition, it can be clearly seen
u
that the maximum current density increases with increasing
where D denotes the diffusion coefficient, H the channel
the anolyte flow rate from 5 mL/h to 60 mL/h. This result can
height and z the distance the fluid flows downstream.
be also attributed to the mass transport enhancement at the
Therefore, it can be easily understood that the diffusive mix-
anode.
ing region broaden gradually along the flow direction of the
microchannel according to Eq. (1) leading to an enhanced
catholyte diffusion toward the anode and further a serious
inhibition of the biofilm formation.

3.3. Effect of influent COD concentration on the cell

performance

The influent fuel concentration has been recognized as an

important factor affecting the performance of MFCs [21].
Therefore, the performance of the microfluidic MFC is
investigated with different influent COD concentrations as
shown in Fig. 4. It is clear that the poorest performance is
observed at an influent COD concentration of 100 mg L1 due
to the seriously insufficient fuel supply. As the influent COD
concentration increases from 500 mg L1e2000 mg L1, the
cell performance remains almost the same at low current
densities (<1200 mA/m2), while it first increases and then
decreases at high current densities. The maximum power
density of 618
4 mW/m2 is obtained at the influent COD
concentration of 1500 mg L1. The increased cell performance
is expected because a high COD concentration will enhance
the mass transfer of substrate from the bulk fluid to the
anode biofilm. It can be also seen from Fig. 4c that the mass
transfer of the acetate at high current densities limits the
performance of the microfluidic MFC. However, the exces-
sively high COD concentration will lead to a substrate inhi-
bition phenomenon [22,23], and hence a drop in the cell
performance.

3.4. Effect of anolyte flow rate on the cell performance

The anolyte flow rate also plays a significant role in elec-

tricity generation of the microfluidic MFC. To investigate the
effect of the anolyte flow rate on the performance of the
microfluidic MFC, the influent COD concentration is kept at
500 mg/L, and the catholyte flow rate is maintained at 10 mL/
h. The results are presented in Fig. 5. When the anolyte flow
rate is increased from 5 mL/h to 40 mL/h, the maximum
power density increases from 129
1 mW/m2 to 510
2 mW/
m2. Such a result can be ascribed to the enhanced mass
transport of the acetate resulting from the decreased mass
transport resistance between the bulk solution and the bio-
film at a high anolyte flow rate, which can be confirmed by
the variation of anode potentials from Fig. 5c. Meanwhile, a
high acetate flow rate facilitates the product expelling from
the microchannel resulting in a high reaction rate at the
anode. Therefore, the abundant acetate supply and the rapid
product removal at high anolyte flow rates lead to the Fig. 4 e Effect of the influent COD concentration on the
improved cell performance. However, when the anolyte flow cell performance: (a) power density curves, (b)
rate is further increased to 60 mL/h, the maximum power polarization curves, (c) anode and cathode potentials.
density decreases to 484
4 mW/m2. This result can be Both the anolyte and catholyte flow rates are 10 mL/h,
explained by the fact that a too high anolyte flow rate may and the potassium ferricyanide concentration is fixed to
result in hydrodynamics instability [24], which will further 20 mM.
15714 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5

Fig. 6 e Nyquist plot for the whole cell. The equivalent

circuit model is shown in the inset. The Warburg
impedance describing the diffusion resistance is
considered. CPE is the constant phase element related to
the double layer capacitance.

Table 1 e Comparison with previous studies.

Total Maximum power Internal Reference
volume (mL) density (mW/m2) resistance
(U)

8 6.25a 16 K [12]
9 47 10 K [5]
25b 95 0.55 K [26]
30 4.012 0.5 K [6]
100c 330 32 K [13]
80 618 2.35 K Present work

a Calculated based on the reported data.

b Anode volume.
c Per cell.

where RU, RCT, and RD represent the ohmic, charge transfer, and
diffusion resistance of the MFC, respectively. The results of EIS
measurements and the equivalent circuit model are shown in
Fig. 6. The EIS fitting data are listed as follows: RU ¼ 45.5 U,
Fig. 5 e Effect of the anolyte flow rate on the cell
RCT1 ¼ 493 U, RCT2 ¼ 1807U and RD ¼ 0.00033 U. Therefore, the
performance: (a) power density curves, (b) polarization
internal resistance RINT ¼ RU þ RCT þ RD z 2350 U, which is far
curves, (c) anode and cathode potentials. The influent COD
lower than the reported values at the same order of magnitude
concentration is 500 mg/L, the catholyte flow rates are
of maximum power density (Table 1), resulting in a higher cell
10 mL/h, and the potassium ferricyanide concentration is
performance in the present study. The low internal resistance
fixed to 20 mM.
can be mainly attributed to elimination of the proton exchange
membrane and high surface-to-volume ratio of the microfluidic
structure. Moreover, it is clear that the charge transfer resis-
3.5. Internal resistance tance dominates the internal resistance of the microfluidic MFC.

For microscale MFCs, high internal resistance is another crit-

ical factor deteriorating the cell performance [25]. The internal 4. Conclusions
resistance of a micro MFC can be defined as:
In this work, a membraneless microbial fuel cell via micro-
RINT ¼ RU þ RCT þ RD (2) fluidic flow control is proposed based on graphite electrodes
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 1 5 7 1 0 e1 5 7 1 5 15715

and investigated under different operation conditions. The co- [9] Kamitani A, Morishita S, Kotaki H, Arscott S. Microfabricated
laminar flow in the microchannel and the biofilm along the microfluidic fuel cells. Sens Actuators B 2011;154:174e80.
flow direction are visualized. It is found that the performance [10] Hou H, Li L, Cho Y, Figueiredo P, Han A. Microfabricated
microbial fuel cell arrays reveal electrochemically active
of MFC first boosts but then decreases with improving the
microbes. PLoS One 2009;4(8):e6570.
influent COD concentration. The similar result is observed [11] Qian F, Baum M, Gu Q, Morse DE. A 1.5 mL microbial fuel cell
under various flow rates. The microfluidic MFC produces a for on-chip bioelectricity generation. Lab Chip
peak power density of 618
4 mW/m2, which is higher than 2009;9:3076e81.
other related studies due to a lower internal resistance. In [12] Qian F, He Z, Thelen MP, Li Y. A microfluidic microbial fuel
addition, the thickness of the biofilm during the steady-state cell fabricated by soft lithography. Bioresourc Technol
operation decreases along the flow direction of the micro- 2011;102:5836e40.
[13] Choi S, Chae J. An array of microliter-sized microbial
channel because of inhibition of biofilm formation caused by
fuelcells generating 100 mW of power. Sens Actuators A
the enhanced diffusive mixing of the two streams. 2012;177:10e5.
[14] Li Z, Zhang Y, LeDuc PR, Gregory KB. Microbial electricity
generation via microfluidic flow control. Biotechnol Bioeng
2011;108(9):2061e9.
Acknowledgments [15] Li Z, Venkataraman A, Rosenbaum MA, Angenent LT. A
laminar-flow microfluidic device for quantitative analysis of
This work is supported by the National Natural and Science microbial electrochemical activity. ChemSusChem
Foundation of China (Nos. 51006130, 51176212), National 2012;5:1119e23.
Natural Science Funds for Outstanding Young Scholar (No. [16] Qian F, Morse DE. Miniaturizing microbial fuel cells. Trends
51222603), the Natural Science Foundation of Chongqing Biotechnol 2011;29(2):62e9.
[17] Wang HY, Bernarda A, Huang CY, Lee DJ, Chang JS. Micro-
(Grant No. cstcjjA90009) and the Program for New Century
sized microbial fuel cell: a mini-review. Bioresourc Technol
Excellent Talents in University (NCET-11-0551). 2011;102:235e43.
[18] Yang J, Ghobadian S, Goodrich PJ, Montazami R, Hashemi N.
Miniaturized biological and electrochemical fuel cells:
references challenges and applications. Phys Chem Chem Phys 2013.
http://dx.doi.org/10.1039/C3CP50804H.
[19] Chen F, Chang MH, Hsu CW. Analysis of membraneless
[1] Logan BE, Hamelers B, Rozendal R, Schröder U, Keller J, microfuel cell using decomposition of hydrogen peroxide
Freguia S, et al. Microbial fuel cells: methodology and in a Y-shaped microchannel. Electrochimi Acta
technology. Environ Sci Technol 2006;40(17):5181e92. 2007;52:7270e7.
[2] Ringeisen BR, Henderson E, Wu PK, Pietron J, Ray R, Little B, [20] Kjeang E, Proctor BT, Brolo AG, Harrington DA, Djilali N,
et al. High power density from a miniature microbial fuel cell Sinton D. High-performance microfluidic vanadium redox
using Shewanella oneidensis DSP10. Environ Sci Technol fuel cell. Electrochimi Acta 2007;52:4942e6.
2006;40(8):2629e34. [21] Du Z, Li H, Gu T. A state of the art review on microbial fuel
[3] Fan Y, Hu H, Liu H. Enhanced coulombic efficiency and cells: a promising technology for wastewater treatment and
power density of air-cathode microbial fuel cells with an bioenergy. Biotechnol Adv 2007;25:464e82.
improved cell configuration. J Power Sources [22] Kim JR, Jung SH, Regan JM, Logan BE. Electricity
2007;171(2):348e54. generation and microbial community analysis of alcohol
[4] Richter H, McCarthy K, Nevin KP, Johnson JP, Rotello VM, powered microbial fuel cells. Bioresourc Technol
Lovley DR. Electricity generation by geobacter sulfurreducens 2007;98:2568e77.
attached to gold electrodes. Langmuir 2008;24(8):4376e9. [23] Sharma Y, Li B. The variation of power generation with
[5] Choi S, Lee HS, Yang Y, Parameswaran P, Torres CI, organic substrates in single-chamber microbial fuel cells
Rittmann BE, et al. A mL-scale micromachined microbial fuel (SCMFCs). Bioresourc Technol 2010;101:1844e50.
cell having high power density. Lab Chip 2011;11(6):1110e7. [24] Choban ER, Markoski LJ, Wieckowski A, Kenis PJA.
[6] Siu CPB, Chiao M. A microfabricated PDMS microbial fuel Microfluidic fuel cell based on laminar flow. J Power Sources
cell. J Microelectromech Syst 2008;17(6):1329e41. 2004;128:54e60.
[7] Chen YP, Zhao Y, Qiu KQ, Chu J, Sun M, Liu XW, et al. An [25] Ren H, Lee HS, Chae J. Miniaturizing microbial fuel cells for
innovative miniature microbial fuel cell fabricated using potential portable power sources: promises and challenges.
photolithography. Biosens Bioelectron 2011;26(6):2841e6. Microfluid Nanofluid 2012;13(3):353e81.
[8] Chiao M, Lam KB, Lin L. Micromachined microbial and [26] Biffinger JC, Ray R, Little B, Ringeisen BR. Diversifying
photosynthetic fuel cells. J Micromech Microeng biological fuel cell designs by use of nanoporous filters.
2006;16(12):2547e53. Environ Sci Technol 2007;41:1444e9.