ORIGINAL PAPER

Strong correlation between membrane effective fixed charge

and proton conductivity in the sulfonated polysulfone
cation-exchange membranes
Bambang Piluharto & Veinardi Suendo & Tati Ciptati &
Cynthia L. Radiman
Received: 10 December 2010 / Revised: 29 January 2011 / Accepted: 16 February 2011 / Published online: 1 April 2011
# Springer-Verlag 2011
Abstract The effectiveness of fixed charge in sulfonated
polysulfone membranes and its correlation with proton
conductivity and physicochemical properties have been
investigated in this work. The membranes were prepared
with various concentrations of sulfonating agent (6% to
10% v/v) and followed by the characterizations that include
membrane potential measurements, proton conductivity,
and physicochemical properties (contact angle, water
uptake, and ion-exchange capacity). Here, the effective
fixed-charge concentrations of the membranes were
obtained based on the data of membrane potential measure-
ments using the TeorellMeyerSiever equation. The analy-
sis results exhibit that a strong correlation between effective
charge concentration and proton conductivity, which is
expressed by the linear increase of proton conductivity with
QX. This correlation is also supported by the membranes
physicochemical data, such as water uptake, ionic exchange
capacity, surface contact angle against water and functional
analysis using FTIR. Finally, it was also developed an ionic
conductivity equation that describes the correlation between
proton conductivity and QX values.
Keywords Effective charge
.
Sulfonated polysulfone
.
Membrane potential
.
Proton conductivity
.
TMS equation
Introduction
Transport properties of a charged membrane deal with its
ability to transfer both ionic and non-ionic matter from
one phase to another. The transport processes are
affected by two effects, namely, the sieving and Donnan
effects. For non-ionic species, the sieving effect is more
dominant, while the Donnan effect is for ionic ones [1].
In its applications, ionic transport becomes an interesting
study that covers many industrial processes such as
electrodialysis, nanofiltration, metal ion recovery, and fuel
cells [16].
Sulfonated polysulfone (SPSF) is a charged membrane
that is widely used as a proton exchange membrane (PEM)
in fuel cell systems .This membrane plays the selective role
in the transport of protons from the anode to the cathode
[46]. The main factor that affects the ionic transport in a
charged membrane is the properties of fixed-charge groups.
The effectiveness of charged groups depends on its
physicochemical properties such as water content, ionic
valency, ion-exchange capacity (IEC), and counter-ion radii
[3, 7, 8]. To understand the mechanism of ionic transport,
many studies about PEM concerning the relationship
between physicochemical and ionic transport properties
have been carried out. Some factors such as the nature of
polymer matrices, solvent polarity of polymer, and degree
of sulfonation were considered [310], however studies
about the effective charge of SPSF and its correlation with
B. Piluharto
:
V. Suendo
:
C. L. Radiman (*)
Inorganic and Physical Chemistry Research Division,
Institut Teknologi Bandung (ITB),
Bandung, Indonesia
e-mail: cynthia@chem.itb.ac.id
B. Piluharto
e-mail: bampito@students.itb.ac.id
V. Suendo
e-mail: vsuendo@chem.itb.ac.id
B. Piluharto
Department of Chemistry, Universitas Jember,
Jember, Indonesia
T. Ciptati
Organic Chemistry Research Division,
Institut Teknologi Bandung (ITB),
Bandung, Indonesia
Ionics (2011) 17:229238
DOI 10.1007/s11581-011-0537-3
proton conductivity and physicochemical properties are still
limited [8].
One of the methods to characterize ionic transport
phenomena in a charged membrane is through membrane
potential measurements [1113]. Theoretically, to obtain
ionic transport properties, the membrane potentials are
analyzed using the TeorellMeyerSiever (TMS) equation
(Eq. 5). In this theory, membrane potential is formulated as
the contribution of the Donnan and diffusion potentials.
Through TMS analysis, we can confirm the effective charge
density in the charged membrane. Therefore, this analysis
can give a fundamental approach to understanding the
relationship between effective charge and physicochemical
and ionic transport properties in SPSF membranes.
In this study, we prepared SPSF-charged membranes
with various concentrations of sulfonating agent (6% to
10% v/v). The study emphasized on the analysis of
membrane effective charge using TMS equation and its
correlation with proton conductivity. Moreover, this corre-
lation will be developed into an ionic conductivity equation
and discussed from both theoretical and empirical point of
view.
Theoretical background
The ionic transport system of a membrane in this study is
presented as schematically shown in Fig. 1.
Electrolyte solutions with different concentrations sepa-
rated by charged membrane lead to a difference in
membrane potential. According to the TeorellMeyer
Siever equation (Eq. 5), the membrane potential is the
sum of Donnan potentials between the membrane surface
and the bulk solution, and the diffusion potential inside the
membrane. To understand in detail, the membrane potential
will be described as the contribution of Donnan potentials
due to Donnan equilibrium theory and the diffusion
potential based on the Nernst-Planck equation [3, 14, 15].
According to Donnan equilibrium, both the ionic concen-
tration and electrical potential are assumed to be discontinu-
ous at the membrane-solution interfaces. In the case of
negatively charged membrane that faces 1:1 electrolyte
systems, the electric potential at the left (f
Don(0m)
) and
right (f
Don(dm)
) parts of the membrane are
$f
Don 0!m

RT
F
ln
g
k C
0

X
2

2

C
0
Q

2

X
2

1
$f
Don d!m

RT
F
ln
g
k C
d

X
2

2

C
d
Q

2

X
2

2
where R is the ideal gas constant, T is the absolute
temperature, g

is the activity coefficient of cations in the

membrane, F is the Faraday constant, Q
2

, where
k
i
is the standard equilibrium constant between the bulk
solution and inside the membrane for the ith ion.
In the case of proton exchange membrane, we have the
negatively fixed-charge group, thus C

, where
X

is the concentration of fixed-charge groups inside the

membrane for negatively charged membranes, and C
o
and
C
d
are the concentrations of the solution on the left and
right sides, respectively.
The diffusion potential can be obtained from the Nernst
Planck equation as follows [14]
$f
diff

RT
F
W ln

1
2C
d
QX

1
2C
0
QX

W
3
Left solution Right solution
Membrane
0

diff
C
o
C
d

Don(d-m)
d

Don(0-m)
Fig. 1 Scheme of ionic
transport in a negatively
charged membrane
230 Ionics (2011) 17:229238
where W is defined as (
+

)/(
+
+

),
+
and

are the
absolute ionic mobility of positive and negative ions,
respectively, and QX

is the effective charge density. The

membrane potential is the sum of Donnan potentials and
diffusion potential that are given by Eqs. 1, 2, and 3. [1315]
$f $f
Don 0!m
$f
Don m!d
$f
diff
4
$f
RT
F
ln
C
d
C
0

1
2Cd
QX

2

1
2Cd
QX

2

RT
F
W ln

1
2Cd
QX

2

1
2Cd
QX

2

W
5
The last equation is more well known as TeorellMeyer
Siever (TMS) equation.
Experimental
Materials
Udel PSF (polysulfone) (Mw=35,000 Da) was obtained
from Union Carbide. Chloroform, chlorosulfonic acid,
hydrochloric acid, sulfuric acid, sodium hydroxide, and N,
N-dimetylacetamide (DMAc; analytical grade reagent) were
purchased from Merck. All chemicals were used without
any further purification.
Preparation of the charged membranes
The PSF was sulfonated according to the procedure previously
described by Nagarale, et.al [16]. Ten grams of PSF were
dissolved into 100 mL chloroform. Each solution was kept in
the ice bath at 0 C. A portion of 50 mL chlorosulfonic acid
in chloroform with different concentrations (6%, 8%, 9%,
10% v/v) was added dropwise to the PSF solution at constant
temperature of 0 C. The resulting mixture was then stirred
for 30 min. After the complete addition of the solution, the
polymer was precipitated in 500 mL of methanol, followed
by washing with methanol until it was free of acid and then
dried in an oven at 60 C for 4 h.
The original and sulfonated PSF (SPSF) solutions were
made in DMAc so that the concentration of the polymer
was 20% w/w. The membranes were formed by solvent
evaporation and then dried in an oven at 60 C to
completely remove the residual solvent. The resulting PSF
and sulfonated PSF membranes with various chlorosulfonic
acid concentrations were designated as PSF, SPSF 6, SPSF
8, SPSF 9, and SPSF 10, representing the concentration of
sulfonating agent, namely, 0%, 6%, 8%, 9%, and 10% v/v,
respectively.
Characterization
Physicochemical properties
Functional group analysis by Fourier transform infrared The
Fourier transform infrared (FTIR) spectra of the membranes
were obtained using a Shimadzu Prestige 21 at room
temperature using a deuterated triglysine sulfate detector.
Before measurement, all samples were dried at 60 C for 24 h
to remove all water from the membrane surface. Film samples
around 0.010.03 mmthick were placed directly in the sample
holder, and then measured in transmission mode in the range
of 4,500400 cm
1
.
Contact angle () measurement The contact angle of the
membrane surface against the water was measured using a
contact angle meter, the Cam-plus Micro (Tantec Inc.).
Here, deionized water was dropped onto the surface of the
membrane and then the contact angle was measured.
Water uptake Membranes were dried in an oven at 60 C
for at least 18 h before weighing to obtain the dry sample
weight [17]. They were then soaked in deionized water for
at least 18 h and the excess water on the surface of the
wetted membrane was removed using tissue paper before
weighing again. The water uptake can be calculated by:
WU
W
w
W
d
W
d
100% 6
where WU, W
w
, and W
d
represent water uptake and the
masses of wet and dried sample in grams, respectively.
Ion-exchange capacity The ion-exchange capacity (IEC)
was determined through titration previously described by
Smitha, et.al [18]. The following equation was used:
IEC
B P 0:01 5
W
d
7
where IEC is the ion-exchange capacity (milliequivalents
per gram), B is the required volume of sulfuric acid to
neutralize the blind sample soaked in NaOH (milliliters), P is
the required volume of sulfuric acid to neutralize the
sulfonated membrane soaked in NaOH (milliliters), 0.01 is
the normality of the sulfuric acid solution, 5 is the factor that
corresponds to the ratio between the amount of NaOH taken
to dissolve the polymer and the amount that is used for
titration and W
d
is the mass of the sample in grams.
Ionics (2011) 17:229238 231
Membrane potential measurement
Prior to any measurement, the membrane was immersed in
a KCl 1-M solution for overnight to ensure that the counter
ions were exchanged by K
+
, and then the membrane was
rinsed several times by distilled water to ensure there is no
excess ions left inside the membrane. The membrane with a
surface area of 3.14 cm
2
was placed between two compart-
ments in a measurement setup as described previously [13].
In the left compartment, a KCl solution was kept constant at
0.001 M, while in the right compartment, the concentration of
the KCl solution was varied from 0.001 to 2 M. The
measurement setup was prepared to have the same volume
of the KCl solution in both compartments to avoid mass or
volume transfer due to the hydrostatic pressure. Two Ag/AgCl
reference glass electrodes were placed in two separated
chambers filled with a saturated KCl solution which were
connected to the two compartments using KCl salt bridges.
The salt bridges were inserted into the compartments in a very
short time during the measurement to avoid the concentration
perturbation of the system. The membrane potential measure-
ment was carried out by connecting the two Ag/AgCl glass
electrodes to a digital multimeter (Sanwa CD800a). In order to
avoid the boundary layer effects, continuous stirring was
employed to the solutions during each measurement. In every
measurement, the temperature in both compartments was kept
constant at 25 C. The membrane potential is then plotted to
the concentration of the KCl solution in the right compart-
ment. The effective charge (QX) and mobility ratio between
cation and anion (
+
/

) can be determined by fitting the

plotted curve with TeorellMeyerSievers (TMS) equation
presented in Eq. 5.
Proton conductivity measurement
Prior to any measurement, membrane was immersed in
distilled water overnight to get the membrane in a swollen
state. Thereafter, water on the membrane surface was
removed gently with tissue paper and then the membrane
was immediately sandwiched between two chromium-wire
electrodes in a two-probe configuration. The length of each
chromium-wire electrodes is 0.4 cm and both of them are
separated by the distance of 1.5 cm. The size of the membrane
used in this configuration is approximately 12 cm.
The proton conductivity was measured by AC imped-
ance technique using a frequency response analyzer with
working frequencies of 20 kHz2 MHz and a 100 mV
amplitude applied signal. All measurements were carried
out at room temperature (25 C). The proton conductivity
of membrane was calculated by the equation, =L/RS,
where L, R, and S are the distance between two chromium-
wire electrode, the measured resistance, and the surface
area of the electrode, respectively. The surface area of the
electrode is obtained by multiplying the length of the
electrode with the membrane thickness.
Results and discussion
All membranes were successfully prepared with varied
concentrations of chlorosulfonic acid except for 10% v/v
(SPSF 10) that resulted in fragile membranes due to the
significant increase of sulfonic groups in the polymer
backbone. This led to the lowering of the mechanical
stability of the cast films due to the change of the balance
between the hydrophilicity and hydrophobicity. Moreover,
increasing of sulfonic groups leads to the change of
molecule conformations through the repulsions among the
sulfonic groups that hindered the interaction among
polymer chains. Nevertheless, in order to confirm this
hypothesis, the functional group analysis were carried out
using FTIR as described below.
The FTIR spectra of unmodified PSF and sulfonated
PSF were compared to provide the information of func-
tional groups in polymer before and after sulfonation. As
shown in Fig. 2, after sulfonation, there are two character-
istic peaks in around 1,025 and 3,500 cm
1
that do not
appear in PSF. This peaks correspond to symmetric O=S=
O stretching of the sulfonate group and OH stretching of
the sulfonic group, respectively [17, 19]. This proved that
the SO
3
H group was introduced successfully into the
polymer chain. The increase of sulfonating agent concen-
tration tends to broaden this band indicating an increase of
membrane polarity. Broadening of the absorption bands
around 3,3503,500 cm
1
is due to increasing intermolec-
ular interactions via hydrogen bridging in each sulfonic
group leading to larger variation of vibration modes
(Fig. 2ad).
The sulfonation of PSF usually occurs at the ortho position
in the phenyl ring relative to the ether bridge of the bisphenol
due to an electrophilic substitution (Fig. 3). This could be
explained by the high electron density at this part of the
molecule (resonance effect, +M from ether bridge and
induction effect, +I from isopropylidene group), in contrast
to the diarylsulfone part of the repeating unit, which has
lower electron density due to the presence of electron-
withdrawing group, SO
2
(I and M effects) [16, 20].
In the case of ionic behavior, membrane potential is used
as a tool to analyze the ionic transport properties of the
charged membrane. The data analysis was carried out using
the TeorellMeyerSievers (TMS) equation. This equation
uses a fixed-charge theory to describe the ionic transport in
ion-exchange membranes [1113]. This theory combines
the Donnan equilibrium theory and the NernstPlanck ionic
flux equation, where the fixed-charge groups are assumed
to be distributed homogeneously in membrane phase and
232 Ionics (2011) 17:229238
the effect of the mean activity coefficient of electrolytes in
the external solution is negligible. Based on this theory, the
membrane potential of an ion-exchange membrane can be
described as the sum of two Donnan potentials at the
solution-membrane interfaces and the diffusion potential
between two solution phases.
Generally, the membrane potential increases with larger
differences of KCl concentration in the range of 0.01 and
0.1 M (Fig. 4ad). This phenomenon is due to the Donnan
exclusion effect. In this phenomenon, when two electrolytes
with different concentrations are separated by an ion-
exchange membrane, they will have a tendency to carry
out electrolyte diffusion. When the bulk electrolyte con-
centration is higher than the concentration of the fixed
charge inside the membrane, the ions from the bulk
electrolyte start to pass through the membrane, so the
membrane potential will decrease.
Based on the curves shown in Fig. 4, the membrane of
SPSF 8 has the highest membrane potential. It indicates that
the potential barrier for ionic diffusion (membrane potential)
is the highest. The magnitude of membrane potential
corresponds to the number of fixed-charge groups. Never-
theless, it does not mean that they also have high activity.
The activity of fixed charge group is determined by the
effective charge (QX) obtained from fitting the experimental
data with TMS equation. According to the data in Table 1
and 2, there is a strong correlation between QX and IEC,
where SPSF 8 also has the highest value. This correlation
indicates that the introduction of more sulfonic groups into
the PSF chain results in a stronger ionic polymer leading to a
more effective fixed-charge distribution as shown by a steep
increase of QX as a function of IEC (Fig. 5).
Besides acting as fixed-charge groups, the sulfonic
groups in SPSF membranes also affect the ion transport.
The data
+
/
-
in Table 2 shows the ratio of mobility
between cation and anion inside the membrane. If the ratio
is less than one, it means that the anions are more mobile
than the cations. On the contrary, if the ratio is more than
one, it means that the cations are more mobile than the
anions. In our experiment, all
+
/
-
values are less than
one. It indicates that the anions are more mobile than the
cations, hence, proving that the sulfonic group in SPSF
C
CH
3
CH
3
O S
O
O
O
S O O
O
-
H
+
n
Fig. 3 Chemical structure of
sulfonated PSF polymer
Fig. 2 The FTIR absorption
spectra of a PSF, b SPSF 6,
c SPSF 8, d SPSF 10
membranes
Ionics (2011) 17:229238 233
membranes act as a negative fixed-charge group or in other
words, our SPSF membrane is a cation-exchange mem-
brane. The negative values of Walso indicate that the fixed-
charge group is negatively charged. Furthermore, the
increase of sulfonating agent concentration up to 8% v/v
(SPSF 8) tends to decrease the
+
/
-
value, indicating that
the membrane becomes more negatively charged.
The data of f
max
and C
o(max)
in Table 2 shows the
maximum values of membrane potential and bulk solution
concentration, respectively. The maximum membrane
potential is defined as the potential when the membrane
acts as a complete barrier to electrolyte diffusion, while
the maximum concentration is the maximum concentra-
tion of the bulk solution in which the bulk electrolyte
cannot pass through the membrane due to the Donnan
exclusion effect. According to both data in Table 2,
increasing of sulfonating agent concentration leads to
similar tendencies on both parameters, f
max
and C
o(max)
.
This result also supports previous data such as the
relationship between QX and IEC shown in Fig. 5.
The overall data of transport parameters reveals that the
SPSF 8 membrane has the highest value of both fixed-
charge groups and effective charges. However, QX values
are much lower than IEC, which means that not all of the
fixed-charge groups play an active role in the ion transport.
The ineffective fixed charges lead to lower proton conduc-
tivity. The presence of ineffective fixed charges may be due
to branching or crosslinking of sulfonic units which further
hinder the proton mobility as previously described [16].
Figure 6 shows the relationship between sulfonating agent
concentration and both proton conductivity and QX. Here,
the similar trends are due to the presence of partial
ineffective fixed charges. The presence of ineffective fixed
charges can be explained by the formation of ion pairs.
Here, the fixed-charge group is said to be completely active
when all charge groups play a role in ion transport. Hence,
when there are some fixed-charge groups that are bounded
ionically to counter ions, they will form ion pairs, which do
not contribute to the transport of ions [3, 21]. Figure 6 also
shows a strong correlation between proton conductivity and
QX, regardless of different measurement techniques or ionic
species involved in each type of measurement. In proton
conductivity measurement, protons were used as the
1E-3 0.01 0.1 1
0
10
20
30
40
50
60
70
80
90
(d)
(c)
(b)
(a)
PSF
SPSF 6
SPSF 8
SPSF 9

(
m
V
)
C
0
(mol L
-1
)
Fig. 4 The effect of concentration of sulfonating agent (0% to 9%) on
the shape of membrane potential curves: The data points show the
membrane potential that is observed experimentally, while the solid
lines are obtained theoretically by fitting the experimental data using
TMS equation as expressed by Eq. 5
Table 1 Physicochemical properties and proton conductivity of
membranes
Membrane WU (%) IEC (mEq/g) (10
4
) (S/cm)
PSF 75 41 0.53
SPSF 6 50 70 1.52 1.60
SPSF 8 30 157 2.11 7.69
SPSF 9 21 194 1.94 3.43
Nafion 117
a
36 0.97 90
contact angle (degree); WU water uptake; IEC ion-exchange
capacity; proton conductivity
a
Data is obtained from literature [22]
Table 2 Parameters of transport properties of the charged membranes
Membrane QX
(10
2
mol/L)
W
+
/
-
f
max
(mV)
C
0 (max)
(10
2
mol/L)
PSF 0.69 0.20 0.67 35.95 1.72
SPSF 6 2.29 0.46 0.37 50.29 1.91
SPSF 8 7.73 0.52 0.32 82.68 5.99
SPSF 9 3.54 0.29 0.54 63.77 3.51
QX is the effective charge; W is (
+
-
-
)/(
+
+
-
);
+
/
-
is the ratio of
cation and anion mobilities; f
max
is the maximum membrane
potential; C
0(max)
is the concentration at maximum membrane potential
0.0 0.5 1.0 1.5 2.0 2.5
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Q
X

(
m
o
l
/
L
)
IEC (meq/g)
Fig. 5 The membrane effective charge (QX) as a function of ion-
exchange capacity (IEC)
234 Ionics (2011) 17:229238
probing charges, while a KCl solution was used as the
electrolyte in membrane potential measurement. The sim-
ilarity between these two techniques is that they only take
into account the effective fixed charges that are involved
directly in ionic transport through the membrane, either for
protons or K
+
. Thus, the results in Fig. 6 show strong
evidence that the results of membrane potential measure-
ment can be used as a tool to reveal the ionic transport
properties of charged membrane, including the tendency of
proton conductivity.
The correlation between QX and contact angle, and
between QX and water uptake were shown in Figs. 7 and 8.
There is a good agreement between WU and contact angle
in which increasing concentration of sulfonating agent
increases water uptake, and decreases contact angle. This
is explained by the fact that introducing more sulfonic
groups into the membrane matrix leads to the increase of
the membrane polarity as well as hydrophilicity of the
sulfonated membrane [17]. Beside sulfonic groups, the
other polar groups, such as sulfone, SO
2
, and ether, O,
also contribute to higher interaction between water and
polymer. Nevertheless, correlation between QX and WU,
and between QX and contact angle was not in good
agreement in SPSF 9. This phenomenon is due to the
probability to undergo branching or crosslinking reactions
that convert the intermediate sulfonic groups into partially
branched or crosslinked sulfone units [16]. This reaction is
assumed to occur in SPSF 9 membranes. In this case, all
branching or crosslinking between ionic groups always
result in polar functional groups, but they are not effective
charges anymore. Here, we note that the formation of
partially branched or crosslinked sulfone units might induce
the decrease of QX obviously, but not on their hydro-
philicity or even their interaction with solvent such as the
solubility due to the lack of interconnectivity among
polymer chains as a whole solids. In other words, this
situation can be described by a system of polyelectrolyte
with high molecular weight, high concentration of polar
groups, but low concentration of ionic groups. Moreover,
we observed that the SPSF 9 is still soluble in several
solvent, i.e., dimethylacetamide and chloroform, which
indicates their lack of interconnectivity due to the partial
branching. This is also supported by the work of Siu et al. if
we correlated this phenomenon with proton conductivity.
Siu et al. described about the presence of two types of water
state in the membrane, namely free water and fixed water,
which are defined as the water molecules that move freely
overall the membrane and that are bound to the sulfonic
groups, respectively, where the latter contributes more
dominantly in proton transport [22]. Thus, the formation
of partially branched or crosslinked sulfone units directly
induces the free water to become more dominant than the
fixed water that leads to decrease of both QX and proton
conductivity (Fig. 6).
Figure 9 shows the empirical and simulation relationship
between the membrane effective charge (QX) and proton
conductivity (). Linear curves a and c in Fig. 9 represent
these relations for Nafion and SPSF membranes, respectively,
while linear curves b and d are the simulation results that
differ in slope value. Here, we emphasized on the physical
meaning of slope value, A, especially between Nafion and
SPSF membranes. Theoretically, the value of proton conduc-
tivity will be proportional to the number of effective charge
density that can be expressed in the following formula,
s
H
A QX 8
where A is the empirical proportional factor in the protonated
case. In order to explain the physical meaning of the slope
exhibited in Fig. 9, we refer to the equation proposed by Choi
et.al [23]. They proposed a model of proton diffusion in a
Nafion membrane with incorporated factors that affect proton
0 2 4 6 8 10
0
1
2
3
4
5
6
7
8
Q
X

(
m
o
l
/
L
)

+

(
1
0

4
S
/
c
m
)
Concentration of sulfonating agent (%)
0.00
0.02
0.04
0.06
0.08
QX
Fig. 6 The membrane effective charge (QX) and proton conductivity
(s
H
) as functions of concentration of sulfonating agent
-1 0 1 2 3 4 5 6 7 8 9 10
20
40
60
80
100
120
140
160
180
200
Q
X

(
m
o
l
/
L
)
WU
W
U

(
%
)
Concentration of sulfonating agent (% v/v)
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
QX
Fig. 7 The membrane effective charge (QX) and water uptake (WU)
as functions of concentration of sulfonating agent
Ionics (2011) 17:229238 235
transport such as porosity, turtuosity, and diffusion coefficient
of proton. The proposed equation is expressed by
s
H

"
i
t
F
2
RT
D

H
C

H
D
G
H
C
H

D
W
H

1 d
c
C
H

9
where in "
i
l
i
= l
i
r , l
i
is the number of moles of water
sorbed per acid site, r is the ratio of the partial molar volumes
of the membrane to that of the water, and t is the turtuosity
factor. C

H
and C
H
are proton concentrations in the surface
and bulk regions inside the membrane pore, respectively.
D

H
,D
G
H
, and D
W
H
are the surface diffusion coefficient,
Grotthuss diffusion coefficient, and StefanMaxwell diffusion
coefficient of hydronium ion inside the membrane pore,
respectively. d
c
is the diffusion coefficient ratio,
d
c

2
p
l
i
r
2=3
, R is the ideal gas constant, T is the absolute
temperature, and F is the Faraday constant.
There are two contributors to proton conductivity, namely,
proton conductivity at the membrane surface and in the bulk
membrane pore. If we assume that the protons on the pore
surface obey Langmuir isotherm where all protons adsorbed
on the surface as a uniform monolayer. In this assumption, the
total concentration of protons, C
H

total
in the membrane is the
sum of proton concentrations in the pore surface, C

H
and in
the bulk phase of membrane pore, C

C
H

total
C

H
C
H
10
The correlation between proton concentration on the
surface, C

H
, and total proton concentration, C
H

total
, is
expressed as C

H
q
H
C
H

total
, thus we can rewrite Eq. 9
into
s
H

"
i
t
F
2
RT
D

H
q
H
C
H

total
D
G
H
C
H

D
W
H

1 d
c
C
H

11
where q
H

k
a
k
a
1
, q
H
; and k
a
are the Langmuir isotherm
factor of proton concentration on the pore surface and the
proton dissociation constant, respectively. We assumed that
C

H
(C
H

total
then we have C
H

total
C
H
, thus Eq. 11
can be rewritten as
s
H

"
i
t
F
2
RT
D

k
a
k
a
1

C
H
D
G
H
C
H

D
W
H

1 d
c
C
H

12
where k
a
is obtained from the correlation between concen-
tration of effective charge, X, and degree of ionic
dissociation, a formulated as k
a

Xa
1a
. Therefore Eq. 12
can be rearranged as follows
s
H

"
i
t
F
2
RT
D

Xa
aX 1 1
C
H
D
G
H
C
H

D
W
H

1 d
c
C
H

13
To obtain a relationship between the effective charge
density, QX and the components of proton concentration in
the Eq. 13, the proportional factor is required. Since the
effectiveness of a charge group depends on the magnitude
of the ionic dissociation constant and the type of counter
ion in the membrane. In the case of proton, we can use the
degree of ionic dissociation, , as the proportional factor,
then the concentration of proton can be expressed as
C
H
a QX . Thus, we can rewrite Eq. 13 to become
s
H

"
i
t
F
2
RT
D

Xa
aX 1 1
aQX D
G
H
aQX
D
W
H

1 d
c
aQX

14
Finally, in order to show the relationship between s
H

and QXin a better perspective, Eq. 14 can be simplified into

s
H
k
0
a QX 15
0.0 0.1 0.2 0.3 0.4 0.5 0.6
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
(c)
(d)
(b)
(a)
A = 0.005
A = 0.007
A = 0.09
A = 0.1678

+
(
S
/
c
m
)
QX(mol/L)
0.0 0.1 0.2
0.000
0.005
0.010
A = 0.005
A = 0.007
A = 0.09
A = 0.1678
Fig. 9 Proton conductivity (s
H
) as function of membrane effective
charge (QX)
-1 0 1 2 3 4 5 6 7 8 9 10
20
30
40
50
60
70
80
Q
X

(
m
o
l
/
L
)

(
d
e
g
r
e
e
)
Concentration of sulfonating agent (% v/v)
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
QX
Fig. 8 The membrane effective charge (QX) and surface contact angle
against water () as functions of concentration of sulfonating agent
236 Ionics (2011) 17:229238
where, k
0

"
i
t
F
2
RT
D

Xa
aX11
D
G
H

D
W
H

1d
c

. By
substituting k
0
a A, we would then return to Eq. 8.
Implicitly, the factors that are covered in k have been
reported elsewhere [6, 24, 25]. Notably, Kreuer reported the
differences of proton transport in Nafion and SPEEK
(sulfonated poly ether-ether ketone) from the microstructural
point of view [24]. Compared to SPEEK, Nafion has wider
hydrophilic channels, smaller separation between SO
3
/SO
3

clusters, and more separated hydrophilic and hydrophobic

units. Since basically the structures of SPEEK and SPSF are
similar, we assumed that the structure of SPEEKcan represent
the SPSF structure. These factors cause Nafion to have a
higher proton conductivity than SPSF. By correlating the
microstructure model with the extended proton conductivity
equation (Eq. 14), we can show that the parameters in
Kreuer's model such as the structure of formed channels, the
distance between -SO
3
-
clusters and the location of hydrophilic
and hydrophobic units, can be represented by the porosity
factor, "
i
, turtuosity factor, t, and, the proton diffusion
coefficients, D
H
, respectively, as described in Eq. 14.
Therefore, we can explain why Nafion has higher proton
conductivity than SPSF using this equation.
In the case of SPSF membranes, increasing the sulfo-
nating agent increases the QX values as Fig. 9c confirms
their linear relation. This indicates that the increase in QX
only affects the concentration of mobile charge carriers in
the membrane, but not the membrane structural factors
covered in k (Eq. 15). Here, every structural feature that
provides delocalization in the charge distribution in the
membrane will facilitate the conduction of proton, i.e., the
presence of electron-withdrawing groups along the polymer
chains, the motion of fixed-charge groups, and the motion
of side-chain [25, 26]. Among these three examples, only
one feature is provided by the intrinsic structure of SPSF,
namely the presence of electron-withdrawing groups that
delocalize the sulfonic groups, i.e., phenyl sulfone and
phenyl ether. The delocalization of fixed-charge groups by
electron-withdrawing groups is much less important than
the motions of fixed-charge groups and side-chains.
Moreover, the introduction of sulfonic groups directly onto
the polymer main chain through sulfonation reaction tend to
localize the fixed charges, hinder their motions and provide a
small spatial separation between hydrophilic and hydrophobic
units. Consequently, the transport of protons through SPSF
membranes becomes more difficult due to high energy barrier
for a proton to hop from one hydrophilic cluster to another
because of their structural features [6, 25, 26].
Conclusion
The results of TMS analysis show that the effectiveness of
membrane fixed charge has a good correlation with the
physicochemical measurements. The effectiveness of fixed
charges in the membrane is represented by QX quantita-
tively. The value of QX shows a good agreement with both
IEC and proton conductivity. However, the water uptake
(WU) gives less agreement with QX, which can be
explained due to the formation of crosslinks between
sulfonic groups. Moreover, the developed ionic conductiv-
ity equation provides a good correlation between QX and
proton conductivity with a proportional factor that depends
on the chemical structure of polymers, membrane porosity,
membrane turtuosity, and their proton diffusion coeffi-
cients. Analyzing the experimental data using this equation
provides a new perspective about the contribution of fixed
charge on the proton conductivity. It can be concluded that
the increase of the sulfonating agent in the case of SPSF
membranes only enhances the QX value, but not the intrinsic
structural factors of membrane covered in k, and consequent-
ly, results in the low proton conductivity. In order to
significantly enhance the proton conductivity, further study on
the modification of SPSF membranes has to be carried out to
provide the structural features that facilitate the fixed charge
delocalization, i.e., the introduction of ionic pendant groups, the
formation of acidbase complex structure, and the formation of
more percolated hydrophilic clusters in the hydrophobic
vicinities of the membrane.
Acknowledgments The authors would like to thank the Ministry of
Research and Technology of Republic of Indonesia for the Incentive
Program under Contract No. 050/RT/D.PSIPTN/Insentif/PPK/I/2009
and ITB for the Institution Research Program under Contract No.
1437a/K01.7/PL/2010, B. Piluharto also would like to thank the
Directorate of Higher Education, Department of National Education,
Republic of Indonesia for the Doctor Programs Grant and to the
Department of Chemistry, Universitas Jember, for partial funding of this
research. V. Suendo acknowledges partial funding from Riset KK-ITB
2010 under contract No. 244/K01.7/PL/2010.
References
1. Matsumoto H, Chen YC, Yamamoto R, Konosu Y, Minagawa M,
Tanioka A (2005) J Mol Struct 739:99104
2. Nagarale RK, Gohil GS, Shahi VK (2006) Adv Colloid Interface
Sci 119:97130
3. Yamamoto R, Matsumoto H, Tanioka A (2006) J Phys Chem B
107:1061510622
4. Fu YZ, Manthiram A (2006) J Power Sources 157:222225
5. Kerres JA (2001) J Membr Sci 185:327
6. Karlsson LE, Jannashch P (2004) J Membr Sci 230:6170
7. Chou TJ, Tanioka A (1998) J Membr Sci 144:275284
8. Tanioka A, Matsumoto H, Yamamoto R (2004) Sci Technol Adv
Mat 5:461468
9. Blanco JF, Nguyen QT, Schaetzel (2002) J Appl Polym Sci
84:24612473
10. Benavente J, Zhang X, Garcia-Valls R (2005) J Colloid Interface
Sci 285:273280
11. Meyer KH, Sievers JF (1936) Helv Chim Acta 19:649664
Ionics (2011) 17:229238 237
12. Lakshminarayanaiah N (1962) Transport phenomena in membranes.
Academic, New York
13. Suendo V, Minagawa M, Tanioka A (2002) J Electroanal Chem
520:2939
14. Hanai, T (1992) Maku to Ion: Busshitsu Idou no Riron to
Keisan, Kagaku Doujin, Tokyo 230232
15. Kawaguchi M, Murata T, Tanioka A (1997) J Chem Soc Faraday
Trans 93:13511356
16. Nagarale RK, Gohil GS, Shahi VK, Rangarajan R (2005) J Appl
Polym Sci 96:23442351
17. Piluharto, B, Jayusman, Y, Radiman, CL, Achmad, S, Suendo, V
(2008) In: Proc. International Conference on advanced and sustainable
polymer, Bandung, August 45
18. Smitha B, Sridhar S, Khan AA (2003) J Membr Sci 225:6376
19. Naim, R, Ismail, AF, Saidi, H, Saion, E. (2004) In: Proc. Regional
Symposium on Membrane Science and Technology
20. Johnson BC, Yilgor I, Iqbal M, Wightman JP, LLyod DR (1986) J
Polym Sci 22:721737
21. Mafe S, Ramrez P, Tanioka A, Pellicer (1997) J Phys Chem B
101:18511856
22. Siu A, Schmeisser J, Holdcroft S (2006) J Phys Chem B 110:6072
6080
23. Choi P, Jalani NH, Datta R(2005) J ElectrochemSoc 152:E123E130
24. Kreuer KD (2001) J Membr Sci 185:2939
25. Spohr E, Commer P, Kornyshev AA (2002) J Phys Chem B
106:1056010569
26. Jang SS, Molinero V, ain T, Goddard WA III (2004) J Phys
Chem B 108:31493157
238 Ionics (2011) 17:229238