Академический Документы
Профессиональный Документы
Культура Документы
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
, 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
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
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 (
+
/
(
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
+
(
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