Journal of Membrane Science 367 (2011) 325–331

Contents lists available at ScienceDirect

Journal of Membrane Science

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

Design of an effective methanol-blocking membrane with purple membrane for

direct methanol fuel cells
Yan Xiang ∗ , Jin Zhang, Yang Liu, Zhibin Guo, Shanfu Lu ∗
School of Chemistry & Environment, Beihang University, 37# Xueyuan Road, Haidian District, Beijing 100191, PR China

a r t i c l e i n f o a b s t r a c t

Article history: In this study, purple membrane (PM) was applied as a methanol-blocking agent on Nafion® mem-
Received 5 September 2010 branes. A series of well-organized poly(diallyldimethylammonium chloride)/PM multilayer films were
Received in revised form 5 November 2010 obtained on a Nafion® 212 surface (PDDA/PM/Nafion® ) to form composite membranes by the elec-
Accepted 7 November 2010
trostatic layer-by-layer (LbL) self-assembly method. The effect of the PDDA/PM/Nafion® heterogenic
Available online 13 November 2010
interface on proton conductivity and methanol permeability was studied by alternating the deposited
surface. With five PDDA/PM bilayers, double-sided modification (PDDA/PM)DF-5 and single-side modifica-
Keywords:
tion (PDDA/PM)SF-5 resulted in excellent methanol blocking with a 73.4% and 64.7% reduction in methanol
Layer-by-layer
Purple membrane
permeability in comparison with unmodified Nafion® 212, respectively. Indeed, the involvement of PM
Nafion modification suppressed the methanol crossover. With regard to the selectivity factor (PDDA/PM)DF-1 composite mem-
Methanol crossover branes showed approximately 2-fold improvement as compared with unmodified Nafion® membranes.
The cell performance of (PDDA/PM)SF-1 composite membranes achieved a power density of 27.0 mW cm−2
which is 48.4% increase compared to cells using unmodified Nafion® 212 membranes. Moreover, with
(PDDA/PM)DF-1 , we achieved a power density of 34.4 mW cm−2 . This study highlights the potential appli-
cation of PM as multifunctional protein membrane in direct methanol fuel cells.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction
Direct methanol fuel cells (DMFCs) have received much atten-
tion as power source for vehicles and portable devices due to their
highly efficient energy conversion, low environmental emissions,
simple structural design, and convenient fuel storage [1]. Perfluo-
rosulfonic polymers such as Nafion® are commonly used in DMFCs
as electrolytes because of their high proton conductivity, thermal
and chemical durability, and oxidative stability [2]. However, the
problem of methanol crossover limits the commercialization of
DMFCs. Methanol migrates from the anode through the Nafion®
membrane to the cathode where it not only reduces the open circuit
potential (OCP) but also poisons the electrocatalyst in the cathode
[3]. An active approach to suppress methanol crossover involves
exploring new sulfonated aromatic polymers including sulfonated
polyimide (SPI) [4], sulfonated poly(ether ether ketone) (SPEEK) [5],
and sulfonated polysulfone (SPSf) [6]. Unfortunately, membranes
that require a high level of sulfonation to achieve sufficient pro-
ton conductivity result in undesirable high swelling and excessive
methanol crossover. An alternative approach for blocking methanol
permeation is the modification of a Nafion® membrane by blend-
ing inorganic components, such as silica [7], zirconium phosphate
∗ Corresponding authors. Tel.: +86 10 82339539; fax: +86 10 82339539.
E-mail addresses: xiangy@buaa.edu.cn (Y. Xiang), lusf@buaa.edu.cn (S. Lu).
[8], zeolite [9], and montmorillonite [10], and various polymers,
such as chitosan [11], poly(vinylidene fluoride) [12], polybenzim-
idazole [13], and poly(vinyl alcohol) [14], to fabricate a composite
membrane.
The layer-by-layer (LbL) self-assembly method is a powerful and
simple method for depositing a thin multilayer film on a substrate
using sequential electrostatic adsorption between negatively and
positively charged polyelectrolytes [15]. Recently, the LbL tech-
nique has successfully been used in DMFCs to suppress methanol
crossover [16–18].
Purple membrane (PM) is localized in the cell membrane of
Halobacterium salinarium and consists solely of bacteriorhodopsin
(BR). PM patches are of irregular shape with a lateral dimensional
of up to 5 ␮m and a constant thickness of 5 nm [19]. PM occurs as
a two-dimensional, hexagonal crystalline array of uniformly ori-
ented bacteriorhodopsin trimers. It is a very promising biomaterial
that has attracted considerable interest in the fields of biooptics
[20] and bioelectrics [21]. This is because BR exhibits long-term
stability against thermal, chemical, and photochemical degrada-
tion, and it maintains its biological activity even when immobilized
on a solid substrate. The self-assembly of PM on polyelectrolyte
surfaces has been studied in recent years [22]; our laboratory simul-
taneously investigated the properties of PM [23] and used the LbL
method to prepare the biomaterial/polymer composite membrane
for DMFCs. PM was found to show promise for facilitating proton
transfer in Nafion® 117 membranes by preventing proton linkage

0376-7388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.memsci.2010.11.013
326 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331
Fig. 1. Scheme of multilayer formation via electrostatic layer-by-layer self-assembly of PDDA and PM on a Nafion® membrane.
by approximately 8% [24]. These studies encouraged us to explore
the application of PM in DMFCs.
In the present study, we report a novel approach for prepar-
ing an effective multilayer methanol blocking film on a Nafion®
membrane by using the electrostatic LbL self assembly method of
polycation PDDA and negatively charged PM. The methanol perme-
ability and selectivity factor of the composite membranes has been
studied in detail for potential application of PM in DMFCs.
2. Experimental
2.1. Chemicals
Purple membrane (PM) was isolated from the R1 M1 strain
of Halobacterium salinarium using standard procedures [25]
and was suspended in a tris-hydroxymethyl aminomethane
(TRIS) buffer solution at 4 ◦ C. In electrostatic experiments, PM
suspensions were used at a concentration of 19 ␮mol L−1 at
various pH values adjusted by aqueous HCl and NaOH solu-
tions. Poly(diallyldimethylammonium chloride) (PDDA) (Mw:
10,000–20,000) was obtained from Sigma–Aldrich. PDDA was
dissolved in 20 mmol L−1 Milli-Q water (pH 6.8) for multilayer
adsorption in all experiments. Nafion® 212 membranes were pur-
chased from Dupont, and TRIS from Sigma–Aldrich. All reagents
were of analytical grade and did not require further purification.
Milli-Q ultrapure water was used in this study (Millipore; 18.2 M
at 25 ◦ C).
2.2. Preparation of self-assembled Nafion® membranes
There are 2 different self-assembled multilayer films: double-
surface modification and single-surface modification. The prepa-
ration of self-assembled double-surface multilayer composite
membranes is shown in Fig. 1. Before self-assembly, Nafion® 212
membranes were treated according to the standard procedure [26].
The treated Nafion® membrane was dipped into a PDDA solution for
20 min. Subsequently, the support was washed and immersed into
a PM solution for another 20 min. Multiple repetition of the adsorp-
tion steps lead to the formation of double-surface self-assembled
polyelectrolyte multilayer films (PDDA/PM)DF on the supports.
The single-surface self-assembled multilayer composite mem-
branes were prepared using a homemade apparatus which
contained 2 compartments designated A and B. Initially, com-
partment B was filled with pure water only and compartment
A was filled with PDDA solution. The Nafion® membrane was
tightly sandwiched between the compartments for 20 min. Then,
the membrane was washed with pure water and tightly sand-
wiched again for 20 min between the compartments; however, the
PDDA solution in compartment A was changed to PM solution,
while the compartment B still contained pure water. Multiple repe-
tition of the adsorption steps lead to the formation of single-surface
self-assembled polyelectrolyte multilayer films (PDDA/PM)SF on
the supports.
2.3. Characterization
The zeta-potential () of the PM solution at different pH values
was determined using a Malvern Zetasizer 2000 series instrument
to optimize the deposition condition of pH. The formation of LbL
self-assembled multilayers on the membrane effectively prepared
from polyelectrolytes was monitored by a GBC Cintra 10e series
UV–vis spectroscopy. In addition, the cross-sectional structure of
the multilayer membranes was sputtered with gold and measured
using a JEOL JSM-5800 scanning electronic microscope (SEM) at an
acceleration voltage of 5 KV. The samples were prepared by freezing
the membranes in liquid nitrogen followed by breaking the frozen
membrane with forceps.
Proton conductivity () was measured using a two-point probe
electrochemical impedance spectroscopy technique over a fre-
quency range of 1 Hz to 100 kHz using a Princeton Applied Research
PARSTAT 2273 potentiostat/galvanostat. The amplitude of the AC
voltage was 10 mV. A sheet of self-assembled multilayer films was
sandwiched between 2 carbon papers (Toray) and was set in a cell
model. The cells were maintained at ambient pressure and humid-
ity.
The methanol permeability of the membranes was determined
using a diaphragm diffusion cell as stated above. The membranes
were equilibrated in Milli-Q pure water overnight with stirring. The
initial concentration of methanol in one side of the cell (compart-
ment A) was 2 M, while the other side of the cell (compartment B)
contained pure water. The increase in methanol concentration over
time was determined by a CP-3800 gas chromatograph.
To evaluate the methanol crossover and the performance
of the DMFCs, a membrane electrode assembly (MEA) was
constructed by sandwiching a LbL self-assembled composite
polyelectrolyte–Nafion® membrane between a Pt/Ru anode and
a Pt/C cathode. Pt/Ru black and 50% Pt/C (both from E-TEK)
were used for the anode and cathode respectively. The loading
of Pt in the catalyst layers was about 2.0 and 1.0 mg cm−2 in the
anode and cathode layers respectively. Carbon paper treated with
 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331 327

Fig. 2. (A) Zeta-potential of PM solution as a function of pH and (B) UV adsorption spectra of (PDDA/PM)DF-4 with various pH values of PM solution.

poly(tetrafluoroethylene) was used as the gas diffusion layer. The
MEA was hot-pressed at 135 ◦ C and 4 MPa for 90 s. DMFC perfor-
mance tests were performed with single-cell hardware (active area:
4 cm2 ) by feeding methanol solution and dry oxygen at flow rates
of 2.5 and 150 mL min−1 (without backpressure) respectively. The
steady-state fuel cell polarization data were collected after oper-
ating the fuel cell for 2 consecutive days including the shut-down
period overnight. Methanol crossover was measured with a voltam-
metric method the steady-state limiting current density arising
from the oxidation of the permeated methanol at the cathode side.
During the measurement, methanol solution was fed at a flow rate
of 2.5 mL min−1 into the anode side of the MEA, while the cath-
ode side was kept in an inert humidified N2 atmosphere at a flow
rate 100 mL min−1 . The steady-state limiting current density result-
ing from complete electro-oxidation at the membrane/Pt catalyst
interface at the cathode side was measured by applying a positive
potential at the cathode side.
3. Results and discussion
Fig. 1 shows the schematic of the ideal configuration of the poly-
electrolyte and the PM fragment on a Nafion® membrane in order
to inhibit methanol permeation. It is known that Nafion® has a dual
structure with a hydrophobic region interspersed with ion-rich
hydrophilic domains; methanol primarily diffuses through these
domains, which are primarily associated with the sulfonic acid
group on the Nafion® membrane surface [26]. As shown, a mono-
layer of the polycation, PDDA, was first electrostatically adsorbed
328 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331

onto the pretreated surface of a negatively charged Nafion® mem-

brane covered by sulfonic acid group. Thus, the charge inversion
took place on the Nafion® surface. Next, a monolayer of PM (a
polyanion) was electrostatically adsorbed onto the PDDA layer.
Multiple repetition of the adsorption steps lead to the formation of
Nafion® -based composite membranes. In this study, it is expected
that the self-assembled multilayer films on the Nafion® membrane
would minimize methanol permeation without a significant detri-
mental effect on proton conductivity. PDDA/PM multilayers were
formed in the aqueous PM solution with pH between 6 and 10;
the pH of the PDDA solution was held constant at neutral in all
experiments.

3.1. Effect of pH

It is known that the rate of coagulation mainly depends on the
stability of colloid particles and that pH is an important param-
eter that affects colloid stability [27]. The stability of colloid is
often investigated in terms of its zeta-potential (), which reflects
their surface chemical properties and the way in which colloids
interact with each other [28]. The -potential of PM solutions of
varying pH values is shown in Fig. 2A; notably, the isoelectric point
(IEP) falls at approximately pH 6 where the -potential is 0. The
-potential data are as expected: positive at relatively low pH val-
ues due to excess protons. As the pH value increases above the IEP,
the -potential gradually decreases and becomes negative after the
solution becomes alkaline; this occurs until a plateau of −66.7 to
−70.8 mV is reached at approximately pH 9. This result suggests
that PM patches are more readily adsorbed onto the polycation
layer as a negatively charged clustered multi-counterion under
more alkaline conditions. Thus, pH 9 is sufficient for the PM frag-
ments to avoid agglomeration, resulting in their smooth and tight
assembly on the PDDA layer [29]. Moreover, the absorbance of
(PDDA/PM)DF-4 at 568 nm increases sharply with increases in the
pH of the PM solution until it reaches a maximum at pH 9.4 (Fig. 2B).
A possible explanation for this behavior is that the charge density
of PM in acidic and neutral media is insufficient to generate enough
Coulombic force to sufficiently interact with the cationic charges in
the PDDA layer. However, if the negative charge density of the PM
is great enough at high pH, the electrostatic interaction becomes
the dominant driving force [30]. Thus, we chose a pH of 9.4 for the
aqueous PM solution for all of the following experiments, whereas
the pH of the PDDA solution was kept constant at neutral in all
experiments.
Fig. 3. UV–vis adsorption spectra of (PDDA/PM)DF depending on the number of
deposited bilayers. The curves (from bottom to top) correspond to the adsorption
of 0, 1, 3, 5, 10, 15, 20, and 25 alternative PDDA/PM bilayers deposited in dou-
ble Nafion® membrane surfaces. The inset graph shows the linear increase of the
absorbance at 568 nm (BR characteristic absorbance) with the number of bilayers;
the linear relative factor = 0.9973.
3.3. Proton conductivity () and methanol permeability (PM ) of
multilayer composite membranes
The proton conductivity of the Nafion® -based composite mem-
branes was measured at room temperature using a two-probe
technique (Tables 1 and 2). The proton conductivity of the unmod-
ified Nafion® 212 was tested in our system and was found to
be 19.2 mS cm−1 . The conductivity of the (PDDA/PM)DF mem-
branes decreased slightly with an increase in the number of the
self-assembled PDDA/PM bilayers. The proton conductivity of the
self-assembled (PDDA/PM)SF multilayer films exhibited a similar
tendency (Table 2). The reduction in proton conductivity of the self-
assembled composite membranes is attributed to the formation of
very thick individual polyelectrolyte bilayers deposited on Nafion®
membrane surfaces, which hinders the mobility of protons in the
self-assembled composite membrane [31].
Methanol permeability is a major drawback that leads to a
marked reduction in cellular efficiency; it is a critical problem that
needs to be solved for DMFC applications. Methanol primarily dif-
fuses through the hydrophilic water channels of Nafion® , which are

3.2. UV measurements

The stepwise assembly of the bilayer films was characterized

by UV–vis absorption spectroscopy (Fig. 3). Since PDDA is almost
transparent in the UV–vis spectrum, the increased absorbance was
mainly attributed to the adsorption of BR (568 nm) in PM on the
previously deposited PDDA layer. The absorbance increased lin-
early with an increasing number of PDDA/PM bilayers (see inset
graph of Fig. 3), which indicates that similar amounts of PDDA
and PM were deposited in every dipping cycle and that well-
organized PDDA/PM multilayers were obtained. In other words,
these results show that PM pitches were well fixed onto the Nafion®
surface.
To confirm the formation of multilayer films on Nafion® mem-
branes, cross-sectional SEM images of (PDDA/PM)DF-30 composite
membranes were investigated (Fig. 4). The image clearly shows that
dense, homogenous multilayers formed on both sides of the mem-
brane support after alternating deposits of polycation (PDDA) and Fig. 4. SEM cross-sectional image of (PDDA/PM)DF multilayer film with 30 bilayers
polyanion (PM). deposited on Nafion® membrane surface.
 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331 329

Table 1
Comparison of the proton conductivity (), methanol permeability (PM ), and mem-
brane selectivity (ϕ) values of (PDDA/PM)DF depending on the number of deposited
bilayers.

Multilayered Number of  (mS PM ϕ (×104

composite deposited bi-layers cm−1 )a (×10−7 S s cm−3 )b
membrane in double surfaces cm2 s−1 )

Nafion® 212 19.2 11.9 1.61

1 10.2 4.19 2.43
2 3.73 4.29 0.87
(PDDA/PM)DF-n 3 3.91 3.62 1.08
4 3.12 3.03 1.03
5 3.29 3.16 1.04
a
Measured at room temperature (23 ◦ C).
b
Membrane selectivity (/PM ).

also used to transport protons. Therefore, the methanol-blocking Fig. 5. The methanol permeability of (PDDA/PM)SF multilayers deposited on Nafion®
effect of PDDA/PM bilayers deposited on Nafion® membranes membranes with modified () and unmodified () surfaces facing the methanol
needs to be evaluated. Table 1 shows the variations in the solution.

methanol diffusion coefficients in the self-assembled PDDA/PM

bilayers in (PDDA/PM)DF multilayer films. As seen in one of the
self-assembled PDDA/PM bilayers, the methanol permeability of
the composite membrane was 4.19 × 10−7 cm2 s−1 – a decrease
of 64.7% in comparison to 1.19 × 10−6 cm2 s−1 measured in an
unmodified Nafion® 212 membrane. This indicates that the for-
mation of PDDA/PM bilayers on Nafion® membranes has positive
effect on blocking the diffusion of methanol. Moreover, when
the number of self-assembled PDDA/PM bilayers increased from
2 to 5, the methanol permeability of the composite membranes
decreased from 4.29 × 10−7 to 3.16 × 10−7 cm2 s−1 respectively. It
is known that Nafion® membranes have both ion-rich hydrophilic
and hydrophobic domains, and that methanol primarily diffuses
through the hydrophilic water-rich domains. When the surface of
a Nafion® membrane was deposited by the PDDA/PM multilayers,
the electrostatic interactions between the positively charged poly-
cation (PDDA) and the sulfonic acid function sites (–SO3 − ) at the
Nafion® membrane surface block the water-rich domains. Once
the methanol diffusion channel is sealed by the polyelectrolyte,
the cross-over of methanol through the Nafion® membrane is dras-
tically reduced [32]. Furthermore, PM is a biomacromolecule that
forms a huge stereospecific blockade covering the Nafion® surfaces
and is able to impede the permeation of methanol.
While considering the heterogenic interface effect between LbL
multilayers and Nafion® membranes on the proton conductivity
of the Nafion® -based composite membranes, we presume that
2 sides of the PDDA/PM multilayers on the substrate have dif-
ferent methanol permeabilities. For confirmation, the methanol
permeability of (PDDA/PM)SF multilayer films with modified and
unmodified surfaces facing the methanol solution were measured
(Fig. 5). It is apparent that the modified surface has a lower
methanol permeability than the unmodified one. The methanol
Table 2
Comparison of the proton conductivity (), methanol permeability (PM ), and mem-
brane selectivity (ϕ) values of (PDDA/PM)SF depending on the number of deposited
bilayers.
Multilayered Number of  (mS PM
composite deposited bi-layers cm−1 )a (×10−7
membrane in single surface cm2 s−1 )
Nafion® 212 19.2 11.9
1 12.2 5.35
2 8.93 4.95
(PDDA/PM)SF-n 3 5.46 4.69
4 4.63 4.53
5 3.91 4.20
a
Measured at room temperature (23 ◦ C).
b
Membrane selectivity (/PM ).
permeabilities of 1 and 9 self-assembled PDDA/PM bilayers with
modified surfaces facing the methanol solution are 5.35 × 10−7
and 2.97 × 10−7 cm2 s−1 respectively – a decrease of 32.1% and
21.4% compared with the unmodified one, respectively. In other
words, when PDDA/PM multilayers are deposited on the modi-
fied surface of Nafion® membranes, methanol must pass through
the PDDA/PM multilayers first, significantly reducing its perme-
ability. On the contrary, when methanol first passes through the
Nafion® membrane followed by the multilayer films, the effect of
the methanol-blocking will diminish.
Regarding the different modified surfaces, the methanol
permeability of (PDDA/PM)SF composite membranes with a mod-
ified surface facing the methanol solution is comparable to
that of (PDDA/PM)DF composite membranes. When comparing
methanol coefficient data between Tables 1 and 2, a signifi-
cant improvement in methanol permeability was observed for
all (PDDA/PM)SF composite membranes relative to (PDDA/PM)DF
composite membranes with the same number of bilayers; a sim-
ilar trend was observed with respect to proton conductivity. For
example, with 2 self-assembled bilayers, the methanol perme-
abilities of (PDDA/PM)SF-2 and (PDDA/PM)DF-2 are 4.95 × 10−7 and
4.29 × 10−7 cm2 s−1 respectively. The different methanol-blocking
properties between (PDDA/PM)SF and (PDDA/PM)DF suggest that
the 2 sides of PDDA/PM multilayers on the substrate have dif-
ferent effects on methanol blocking, which is consistent with the
presumption stated above.
3.4. Limiting methanol permeation current density
Fig. 6 shows the limit current density of methanol oxi-
dation for self-assembled (PDDA/PM)DF multilayers. With one
self-assembled PDDA/PM bilayer, the limiting methanol crossover
current density was 58.3 mA cm−2 – a 35.2% decrease compared
to 90.0 mA cm−2 measured for a pristine Nafion® 212 membrane.
The drastic reduction in the methanol crossover current demon-
ϕ (×104 strates the effective blockage of the sites for methanol crossover
S s cm−3 )b
through the self-assembly of PDDA/PM multilayers at the Nafion®
membrane surface. When the number of PDDA/PM bilayers was
1.61 increased to 4, the limit methanol current density decreased fur-
2.28
ther to 54.3 mA cm−2 . This indicates that the LbL self-assembled
1.80
1.16 (PDDA/PM)DF multilayers reduce the methanol crossover cur-
1.02 rent, which in turn decreases with an increase in the number of
0.93 deposited PDDA/PM bilayers. This result is consistent with the
methanol diffusion coefficients measured by gas chromatography
as reported above.
330 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331

Fig. 6. Methanol crossover current density of the composite membranes as a func-

tion of the deposited bilayers.

(PDDA/PM)SF multilayer films show the same tendency

towards methanol permeation current density as (PDDA/PM)DF . As
expected, an increase in the number of PDDA/PM bilayers resulted
in regular decrease of the limit methanol current density (Fig. 6).
For example, one self-assembled PDDA/PM bilayer, had a limit- Fig. 7. Membrane selectivity of the layer-by-layer-assembled composite mem-
ing methanol crossover current density of 65.0 mA cm−2 , which is branes as a function of deposited bilayers. (A) (PDDA/PM)DF and (B) (PDDA/PM)SF .
slightly greater than (PDDA/PM)DF composite membranes. When
the number of PDDA/PM bilayers was increased to 5, the limit
methanol current density decreased to 56.0 mA cm−2 . All of the
results demonstrate that the PDDA/PM multilayers have a strong
effect on methanol blocking after assembly on the surface of
Nafion® membranes.

3.5. Membrane selectivity of the LbL-assembled composite

membranes

The selectivity factor (ϕ) is defined as the ratio of proton

conductivity to the methanol permeability and is used as an
indicator of the suitability of a given membrane for DMFC appli-
cations. A high selectivity factor indicates that the membrane has
high proton conductivity and low methanol permeability; thus,
it would be expected that a membrane with a higher ϕ-value
would perform better in a DMFC [33]. Fig. 7 shows the ϕ-values
of the self-assembled Nafion® -based composite membranes with
single and double modified surfaces as a function of deposited
PDDA/PM bilayers. The ϕ-value of unmodified Nafion® 212 is
1.61 × 104 S s cm−3 . With one PDDA/PM bilayer, the selectivity fac-
tor of (PDDA/PM)DF composite membranes reaches a maximum
value of 2.43 × 104 S s cm−3 – an increase of 50.9% compared with
that of an unmodified Nafion® 212 membrane. With an increase in
the number of PDDA/PM bilayers, the ϕ-value gradually decreases
(Fig. 7A). This can be attributed to the decrease in proton con-
ductivity and the roughly constant methanol permeability. On the
other hand, the selectivity factor of (PDDA/PM)SF systems exhibits
a similar tendency to that of (PDDA/PM)DF , reaching a maximum
value of 2.28 × 104 S s cm−3 with 1 self-assembled PDDA/PM bilayer
(Fig. 7B) – an increase of 42.2% compared with that of a pristine
Nafion® 212 membrane. In conclusion, Nafion® -based compos-
ite membranes exhibit higher ϕ-values with one self-assembled
PDDA/PM bilayer than unmodified Nafion® , suggesting that greater
performance should be expected.
Fig. 8. Polarization curves and power output of a DMFC using an unmodified Nafion®
212 membrane, Nafion® -based (PDDA/PM)SF-1 and (PDDA/PM)DF-1 composite mem-
branes measured at room temperature. Operating conditions: anode, 1.5 M CH3 OH,
2 mL min−1 ; cathode, oxygen, 150 mL min−1 .
(PDDA/PM)SF-1 or (PDDA/PM)DF-1 at room temperature is shown
in Fig. 8. The cell performance of (PDDA/PM)SF-1 composite mem-
branes achieved a power density of 27.0 mW cm−2 – a 48.4%
increase compared to cells using unmodified Nafion® 212 mem-
branes (18.2 mW cm−2 ). Moreover (PDDA/PM)DF-1 exhibits higher
performance than (PDDA/PM)SF-1 , achieving a power density of
34.4 mW cm−2 . Such a clear increase in cell output results from
decreased methanol crossover, because the PDDA/PM bilayers
clearly decrease the methanol crossover as stated above. The
high OCV and performance of cells with (PDDA/PM)DF-1 composite
membranes can be attributed to the fact that the modified mem-
brane has a lower methanol permeability and a higher conductivity
than (PDDA/PM)SF-1 composite membranes; this is consistent with
the selectivity of the PDDA/PM multilayer films.

3.6. Single fuel cell performance 4. Conclusions

The cell performance of DMFCs prepared with an unmodified PDDA/PM multilayer films deposited on Nafion® 212 mem-
Nafion® 212 membrane and Nafion® 212 membranes modified by branes were successfully fabricated using the electrostatic LbL
 Y. Xiang et al. / Journal of Membrane Science 367 (2011) 325–331
self-assembly technique for both (PDDA/PM)DF and (PDDA/PM)SF
composite membranes. Although PDDA/PM multilayer films have
some negative effects on the proton conductivity of compos-
ite membranes, the methanol crossover is significantly reduced
with these membranes as compared to unmodified Nafion® mem-
branes. With 5 PDDA/PM bilayers, the methanol permeability of the
(PDDA/PM)DF and (PDDA/PM)SF composite membranes decreases
by 73.4% and 64.7% to 3.16 and 4.19 × 10−7 cm2 s−1 , respectively.
The main reason underlying this result is that PDDA/PM multilay-
ers with PDDA/PM on their Nafion® membranes associate with
SO3 − clusters, thus blocking water channels and plugging the
hydrophilic domain; this in turn suppresses methanol crossover.
Moreover, both (PDDA/PM)DF and (PDDA/PM)SF multilayer films
with a single PDDA/PM bilayer exhibited the greatest membrane
selectivity. (PDDA/PM)SF-1 multilayer film achieved a power den-
sity of 27.0 mW cm−2 , which is 48.4% higher than that with an
unmodified Nafion® 212 membrane, and (PDDA/PM)DF-1 exhibits
higher performance than (PDDA/PM)SF-1 , achieving a power den-
sity of 34.4 mW cm−2 . Consequently, Nafion® -based PDDA/PM
multilayer films have potential application in DMFCs with low
methanol permeation and high fuel cell performances.
Acknowledgements
This project was supported by grants from the National Natural
Science Foundation of China (20773008, 21003007, and 21073010),
the National High Technology Research and Development Pro-
gram (863 program, 2077AA05Z146), Beijing Novel Program (No.
2008B12) and the Fundamental Research Funds for the Central Uni-
versities (YWF-10-02-004).
References
[1] W. Li, A. Manthiram, Sulfonated poly(arylene ether sulfone) as a methanol-
barrier layer in multilayer membranes for direct methanol fuel cells, J. Power
Sources 195 (2010) 962–968.
[2] K.A. Mauritz, R.B. Moore, State of understanding of Nafion, Chem. Rev. 104
(2004) 4535–4585.
[3] A. Bettelheim, L. Soifer, E. Korin, Electropolymerized porphyrin films as
methanol barriers in direct methanol fuel cells, J. Electroanal. Chem. 571 (2004)
265–272.
[4] Y. Woo, S.Y. Oh, Y.S. Kang, B. Jung, Synthesis and characterization of sulfonated
polyimide membranes for direct methanol fuel cell, J. Membr. Sci. 220 (2003)
31–45.
[5] J.C. Tsai, H.P. Cheng, J.F. Kuo, Y.H. Huang, C.Y. Chen, Blended Nafion® /SPEEK
direct methanol fuel cell membranes for reduced methanol permeability, J.
Power Sources 189 (2009) 958–965.
[6] F. Lufrano, V. Baglio, P. Staiti, A.S. Arico, V. Antonucci, Polymer electrolytes based
on sulfonated polysulfone for direct methanol fuel cells, J. Power Sources 179
(2008) 34–41.
[7] S.Z. Ren, G.Q. Sun, C.N. Li, Z.X. Liang, Z.M. Wu, W. Jin, X. Qin, X.F. Yang, Organic
silica/Nafion® composite membrane for direct methanol fuel cells, J. Membr.
Sci. 282 (2006) 450–455.
[8] M.P. Rodgers, Z. Shi, S. Holdcroft, Ex situ characterisation of composite Nafion
membranes containing zirconium hydrogen phosphate, Fuel Cells 9 (2009)
534–546.
[9] X. Li, E.P.L. Roberts, S.M. Holmes, V. Zholobenko, Functionalized zeolite A-
Nafion composite membranes for direct methanol fuel cells, Solid State Ionics
178 (2007) 1248–1255.
[10] R.F. Silva, S. Passerini, A. Pozio, Solution-cast Nafion® /montmorillonite com-
331
posite membrane with low methanol permeability, Electrochim. Acta 50 (2005)
2639–2645.
[11] Y. Xiang, M. Yang, Z.B. Guo, Z. Cui, Alternatively chitosan sulfate blending
membrane as methanol-blocking polymer electrolyte membrane for direct
methanol fuel cell, J. Membr. Sci. 337 (2009) 318–323.
[12] S.W. Choi, Y.Z. Fu, Y.R. Ahn, S.M. Jo, A. Manthiram, Nafion-impregnated electro-
spun polyvinylidene fluoride composite membranes for direct methanol fuel
cells, J. Power Sources 180 (2008) 167–171.
[13] R. Wycisk, J. Chisholm, J. Lee, J. Lin, P.N. Pintauro, Direct methanol fuel cell mem-
branes from Nafion-polybenzimidazole blends, J. Power Sources 163 (2006)
9–17.
[14] N.W. DeLuca, Y.A. Elabd, Nafion® /poly(vinyl alcohol) blends: effect of compo-
sition and annealing temperature on transport properties, J. Membr. Sci. 282
(2006) 217–224.
[15] G. Decher, Fuzzy nanoassemblies: toward layered polymeric multicomposites,
Science 277 (1997) 1232–1237.
[16] M. Yang, S.F. Lu, J.L. Lu, S.P. Jiang, Y. Xiang, Layer-by-layer self-assembly of
PDDA/PWA-Nafion composite membranes for direct methanol fuel cells, Chem.
Commun. 46 (2010) 1434–1436.
[17] S.C. Mu, H.L. Tang, Z.H. Wan, M. Pan, R.Z. Yuan, Au nanoparticles self-assembled
onto Nafion membranes for use as methanol-blocking barriers, Electrochem.
Commun. 7 (2005) 1143–1147.
[18] C.J. Zhao, H.D. Lin, Z.M. Cui, X.F. Li, H. Na, W. Xing, Highly conductive, methanol
resistant fuel cell membranes fabricated by layer-by-layer self-assembly of
inorganic heteropolyacid, J Power Sources 194 (2009) 168–174.
[19] U. Alexiev, T. Marti, M.P. Heyn, H.G. Khorana, P. Scherrer, Surface charge of
bacteriorhodopsin detected with covalently bound pH indicators at selected
extracellular and cytoplasmic sites, Biochemistry 33 (1994) 298–306.
[20] D. Ho, B. Chu, H. Lee, E.K. Brooks, K. Kuo, C.D. Montemagno, Fabrication of
biomolecule-copolymer hybrid nanovesicles as energy conversion systems,
Nanotechnology 16 (2005) 3120–3132.
[21] Y.D. Jin, N. Friedman, M. Sheves, D. Cahen, Bacteriorhodopsin-monolayer-based
planar metal–insulator–metal junctions via biomimetic vesicle fusion: prepa-
ration, characterization, and bio-optoelectronic characteristics, Adv. Funct.
Mater. 17 (2007) 1417–1428.
[22] M.-b. Saab, E. Estephan, T. Cloitre, R. Legros, F.J.G. Cuisinier, L. Zimaı̌nyi, C.
Gergely, Assembly of purple membranes on polyelectrolyte films, Langmuir
25 (2009) 5159–5167.
[23] Y. Xiang, M. Yang, T. Su, Y.Y. Chen, L.J. Bi, K.S. Hu, Glycolipid biotinylation on
purple membrane with maintained bioactivity, J. Phys. Chem. B 113 (2009)
7762–7766.
[24] D. Ho, B. Chu, H. Lee, C.D. Montemagno, Protein-driven energy transduction
across polymeric biomembranes, Nanotechnology 15 (2004) 1084–1094.
[25] D. Oesterhelt, W. Stoeckenius, Isolation of the cell membrane of Halobacterium
halobium and its fractionation into red and purple membrane, Methods Enzy-
mol. 31 (1974) 667–678.
[26] S.P. Jiang, Z.C. Liu, Z.Q. Tian, Layer-by-layer self-assembly of composite
polyelectrolyte-nafion membranes for direct methanol fuel cells, Adv. Mater.
18 (2006) 1068–1072.
[27] P. Narong, A.E. James, Effect of pH on the ␨-potential and turbidity of yeast
suspensions, Colloids Surf. A: Physicochem. Eng. Aspects 274 (2006) 130–137.
[28] J.A. Brant, A.E. Childress, Assessing short-range membrane-colloid interactions
using surface energetics, J. Membr. Sci. 203 (2002) 257–273.
[29] T. Jussila, M.L. Li, N.V. Tkachenko, S. Parkkinen, B.F. Li, L. Jiang, H. Lem-
metyinen, Transient absorption and photovoltage study of’ self-assembled
bacteriorhodopsin/polycation multilayer films, Biosens. Bioelectron. 17 (2002)
509–515.
[30] J.A. He, L. Samuelson, L. Li, J. Kumar, S.K. Tripathy, Oriented bacteri-
orhodopsin/polycation multilayers by electrostatic layer-by-layer assembly,
Langmuir 14 (1998) 1674–1679.
[31] H. Deligoz, S. Yilmazturk, T. Karaca, H. Ozdemir, S.N. Koc, F. Oksuzomer, A.
Durmus, M.A. Gurkaynak, Self-assembled polyelectrolyte multilayered films
on Nafion with lowered methanol cross-over for DMFC applications, J. Membr.
Sci. 326 (2009) 643–649.
[32] H.L. Tang, M. Pan, S.P. Jiang, Z.H. Wan, R.Z. Yuan, Self-assembling multi-layer Pd
nanoparticles onto NafionTM membrane to reduce methanol crossover, Colloids
Surf. A: Physicochem. Eng. Aspects 262 (2005) 65–70.
[33] H. Ahmad, S.K. Kamarudin, U.A. Hasran, W.R.W. Daud, Overview of hybrid mem-
branes for direct-methanol fuel-cell applications, Int. J. Hydrogen Energy 35
(2010) 2160–2175.