Supplementary Material
Enhanced Nitrogen Removal in Single-Chamber
Microbial Fuel Cells with Increased Gas Diffusion Areas
Hengjing Yan and John M. Regan
Ammonium Conc. (mg N L -1)
150
C
CD
CC
120
90
60
30
0
0
100
200
300
400
500
Time (hr)
Figure S1. Ammonia concentrations during the enrichment stage.
-1-
Ammonia removal eff. (%)
100
80
60
40
20
C
CD
CC
0
0
2
4
6
8
10
12
14
16
Batch
Figure S2. Ammonia removal efficiencies in C, CD, and CC reactors during fed-batch
mode.
1.0
C
Mean Voltage (V)
0.8
CD
CC
0.6
0.4
0.2
0.0
1
3
5
7
9
11
13
15
COD: N Ratio
Figure S3. Cell voltages (mean values at each COD/N ratio condition) of C, CD, and
CC MFCs.
-2-
Estimation of ammonia assimilation rates using total redox reaction with cell
synthesis for the combination of both non-exoelectrogens and exoelectrogens
We estimated the ammonia assimilation rate for the COD/N ratio of 13,
assuming that autotrophic assimilation was minor in comparison with heterotrophic
assimilation and that methanogenesis was negligible. The stoichiometric relation
between ammonia assimilation and COD removal rates is shown in (eq. S-1), where
and
respectively,
are the molecular weights of nitrogen and sodium acetate,
and
are the stoichiometric coefficients for ammonium
and sodium acetate in the total redox reaction with cell synthesis for the combination
of both non-exoelectrogens growing on the air cathode and exoelectrogens, and 0.78
is the mass conversion factor from COD to sodium acetate.
[eq.
The
S-1].
values for C, CD, and CC reactors are 0.216, 0.221, and 0.217,
respectively. The detailed calculations are shown below.
Electron balance: in our mixed-culture single-chamber MFC system, we assume that
fraction a from the total acetate-derived electrons was used by exoelectrogens, with fs1
of this fraction used for cell synthesis and fe1 transferred to the anode for energy
generation (Fig. S4). Therefore,
is simply the Coulombic efficiency. Bacteria
growing nonexoelectrogenically on the air cathode use the balance (i.e., 1-a) of the
total acetate-derived electrons, with fs2 of this portion invested in cell synthesis and fe2
-3-
transferred to oxygen. fs2 and fe2 were adopted from empirical values for aerobic
heterotrophs (fs2 = 0.6 and fe2 = 0.4) (Rittmann and McCarty 2001).
Figure S4. Chart of electron flow from the substrate acetate to different pathways in
MFC systems.
An energy balance between catabolism and anabolism was then coupled to the
electron balance to calculate fs1, fe1, and a (Rittmann and McCarty 2001) .
Energy balance:
G p is the energy required to convert the sodium acetate to pyruvate, which is
G p  35.09  27.40  7.69 kJ / e eq [eq. S-2];
G pc is the energy required to convert pyruvate to cellular carbon, estimated to be
G pc  18.8 kJ / e  eq [eq. S-3];
The term  was used to account for energy–transfer efficiency. Therefore, the total
energy requirement for cell synthesis ( Gs ) is
Gs 
G p  G pc

[eq. S-4].
-4-
Gr is the free energy released per each equivalent of sodium acetate oxidized
for energy generation. To supply the energy required for cell synthesis ( Gs ), A
equivalents of sodium acetate must be oxidized, which yields AGr . Assuming the
same energy transfer efficiency as cell synthesis (  ), the energy balance is
AGr  Gs  0 [eq. S-5],
A 
Gs

Gr
G p  G pc

Gr

G p  G pc
 2 Gr

73.58 kJ / e  eq
[eq. S-6].
Gr
At the anode, sodium acetate was oxidized to carbon dioxide and the electrons
went from sodium acetate to the anode electrode. The free energy released from this
process ( Gr ) could be calculated from the difference of the reduction potentials
between the CO2/acetate redox couple ( ECO2 / acetate   0.28 V ) and the anode
( Eanode ) (Madigan 2006) as follows:
Gr  Gr  RT ln
0'
([CO2 ])1/ 8 ([ HCO3 ])1/ 8 [ H  ]
([ Ac  ])1/ 8
([CO2 ])1/ 8 ([ HCO3 ])1/ 8 [ H  ]
 nFE0  RT ln
([ Ac  ])1/ 8
'
([CO2 ])1/ 8 ([ HCO3 ])1/ 8 [ H  ]
 nF ( Eanode  E
)  RT ln
([ Ac  ])1/ 8
 1  (96485 C/mol )  ( Eanode  (0.28 V))
'
0 CO2 / acetate
 (8.314 J/(mol  K))  (303.15K ) ln
([CO2 ])1/ 8 ([ HCO3 ])1/ 8 [ H  ]
.
([ Ac  ])1/ 8
[eq. S-7]
The values of Eanode , sodium acetate effluent concentration [ Ac  ] , CO2
produced from the oxidation of the sodium acetate, [CO2 ] , [ HCO3 ] , and pH for C,
CD, and CC reactors at a COD/N ratio of 13 are shown below:
-5-
C
CD
CC
Eanode (V)
-0.254
-0.244
-0.250
[ Ac  ] (mol/L)
0.146
0.064
0.072
* [CO2 ] (mol/L)
0.0040
0.0053
0.0053
[ HCO3 ] (mol/L)
0.00048
0.00056
0.00054
pH
7.23
7.28
7.30
*The calculation of
[CO2 ] and [ HCO3 ] : CO2 is produced from sodium acetate. We assume a
portion X of the sodium acetate removed was converted to CO2 and HCO3-. Therefore,
X  a  f e1  (1  a)  f e2  CE  f e2  a  f e2  CE  0.4  0.4a [eq. S-8].
Since there is no headspace, we further assume that all CO2 was converted to H2CO3. Hence, we have
at 30 °C:
pKa 
[ HCO3 ][ H  ]
[ H 2CO3 ]

3
 5.01187  10 7 (Snoeyink and Jenkins, 1980)
[eq. S-9, 10]

[ H 2CO3 ]  [ HCO ]  2 X (Ac )
We used the trial and error method to solve this problem for a and the
Using the relationships f e1 
[CO2 ] and [ HCO3 ] values.
A
1
, f s1 
, and a  CE / f e1 , the Gr , A ,
1 A
1 A
f e1 , f s1 , and a values of C, CD, and CC reactors were calculated as the following:
C
CD
CC
Gr (kJ/e- eq)
-46.60
-47.72
-47.20
A
1.579
1.542
1.556
fe1
0.612
0.607
0.609
f s1
0.388
0.393
0.391
a
0.282
0.235
0.278
1
' 
The vNH
in the following cell synthesis reaction is
:
4
20
-6-
1
1
1
1
9
CO2 
HCO3- 
NH4  H   e   C5 H 7 O 2 N 
H O , and the
5
20
20
20
20 2
1
'
 in the following reduction reaction is
vCH
:
COO
3
8
1
1
1
3
CO2  HCO3-  H   e  CH3COO-  H 2O .
8
8
8
8
Therefore, the stoichiometric relationship between NH4 and CH3 COO- in the
system is
vNH 
4
vCH COO 
3
 a
' 
f s1  vNH
4
'
vCH
COO 
3
 (1  a) 
' 
f s 2  vNH
4
'
vCH
COO 
3
, and the values were calculated as
below:
vNH  vCH COO 
4
3
C
CD
CC
0.216
0.221
0.217
-7-